International Congress on Renewable Energy (ICORE–2013) Renewables for Development of Rural Areas
www.groupexcelindia.com
Inte ernatiional Congre C ess on Ren newab ble Energy (ICOR ( RE–201 13) Reenewables for Developm ment of Rural Areas
Editors
Dr. M. M Kumarravel D S.M. Ali Dr. A Dr. S..K. Samd darshi Jha Dr. Ranjana R Mr. Jagat S. Jawa
Jointtly Publish hed by
EXCEL INDIA N PUBLISSHERS NEW DELHI
Solar S Energgy Society of India (SE ESI) New Delhi
First Impression: 2013 © Solar Energy Society of India (SESI) International Congress on Renewable Energy (ICORE–2013) Renewables for Development of Rural Areas ISBN: 978-93-82880-80-6
No part of this publication may be reproduced or transmitted in any form by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the copyright owners. DISCLAIMER The authors are solely responsible for the contents of the papers compiled in this volume. The publishers or editors do not take any responsibility for the same in any manner. Errors, if any, are purely unintentional and readers are requested to communicate such errors to the editors or publishers to avoid discrepancies in future. Published by EXCEL INDIA PUBLISHERS 91 A, Ground Floor Pratik Market, Munirka, New Delhi 110067 Tel: +91-11-2671 1755/ 2755/ 3755/ 5755 ● Fax: +91-11-2671 6755 Email: [email protected] Web: www.groupexcelindia.com & Solar Energy Society of India 2nd Floor, Central Board of Irrigation and Power Building Malcha Marg, Chanakyapuri New Delhi–110 021 Tel: 011-65649864 ● Fax: 011-40537412 E-mail: [email protected]/ [email protected] Website: www.sesi.in/ www.icoreindia.org Typeset by Excel Publishing Services, New Delhi–110067 E-mail: [email protected] Printed by Excel Printing Universe, New Delhi–110067 E-mail: [email protected]
Sri William Bilung, IAS Chief Executive OREDA
Message
Brief Bio-Data a of Dr. Ach hyuta Sama anta Educattion • •
:
M.Sc. (Chemistry) from Utkal University, Va ani Vihar, Bhhubaneswar. Ph.D. (Social Science))
Esstablished Kalinga Institute of o Industrial Teechnology (KIITT)-A multi-disciplinary promissing University with global vissibility having 21,000 studennts’ intake from m across the glo obe for pursuinng education in 25 professionnal courses. It is spread over 25 2 sq. km. of area in 20 sprawling beautifuul campuses witth 90 lakh sq. fft. built-up area. Esstablished Kalinnga Institute of Social Sciencces (KISS)-A ho ome for 20,24 42 poorest of tthe poor tribal(indigenous) chhildren from kinndergarten to post-graduatio p on, supposed to o be the largesst fully free ressidential tribal institute
Presently Associated A w with • • • • • • • • • • • • • • •
Coommission Mem mber, Universitty Grants Com mmission (UGC), Ministry of HRD, Governm ment of India (2008-11 ( & 20 011-14 (2nd Coonsecutive term m). M Member, Nationnal Council for Teacher Educa ation (NCTE), Ministry M of HRD, Govt. of India a. M Member, Coir Booard, Governm ment of India, Kochi. K M Member, Nationnal Executive Council, C Indian Society S for Tecchnical Educatioon. M Member, Executive Committee, Indian Sciencee Congress Asssociation (ISCA A). National Secreta ary, Gandhi Global G Family (G GGF). Foormer Member, Executive Committee, C All India Counciil for Technica al Education, (AICTE), Ministtry of HRD, G Government of India. I M Member, Acadeemic Council of Assam Centrall University, Silchar (President of India Nom minee). M Member, Acadeemic Council, Ceentral Universitty, Odisha (Nominee of the MHRD, M Govt. off India). M Member, Ministry of Human Resource and d Developmennt, Govt. of India I (Committtee on ‘Round d Table’ on Diisadvantaged Section, Womeen, and SC/ST for expansionn of educationa al opportunitiess). Foormer Memberr, High-level coommittee on Higher Educatio on, constituted by Planning C Commission, Go overnment of India. Foormer Memberr, Ministry of Environment & Forest and Tribal Affairss, Govt. of Ind dia (Committe ee on Forest Right Act). Foormer Memberr, General Bod dy, Centre forr Advancement of People’s Action and Ruural Technology (CAPART), G Govt. of India. Foormer Memberr, Executive Committee, Indian Red Cross So ociety, Odisha State Branch (Nominee of thhe Governor off Odisha). Foormer Vice-Preesident, Bharat Scouts & Guid des, Odisha Sta ate.
Member in n Internationa al Bodies • • • • • • •
Intternational Asssociation of University Presideents (IAUP), US SA. Intternational Insttitute of Educattion (IIE) New York. Y Asssociation of University of Asiia Pacific (AUA AP). Unniversity Mobility in Asia & thhe Pacific (UMA AP), Bangkok (Thailand). Assia-Pacific Aca ademic Consortium for Public Health (APACP PH). Unnited Nations Academic A Impa act. Assia Economic Foorum (AEF).
Award an nd Recognition Honorary Causa C from Six S Universitiees • • • • • •
D..Litt. Degree (H Honoris Causa) from Rashtriya a Sanskrit Vidy yapeetha (Central University)), Tirupati (A.P..)-2011. Deegree of Doctoorate (Honoris Causa) from Hanseo H Universiity of South Koorea-2010. Deegree of Doctoorate (Honoris Causa) from National N Univerrsity of Cambodia-2009. Hoonorary Ph.D. (2002) ( and D.SSc. (2005) from m OIU, Colomb bo. Deegree of Doctoorate (Honoris Causa) from National N Formossa University, Taiwan-2012. T Hoonorary Highest Degree of thhe University ‘G Goddess of Ha arvest OPS’ frrom the Czech University of Life L Sciences, Prrague-2013.
Have visited v several countries, excluusively for official work.
Brief Bio-Data of V..S. Verma Member Cenntral Electriccity Regulatory Commisssion V.S. Verma is a known sp pecialist in the thermal power and in the fie eld of planning g for generatioon capacity in the country. Mr. Verma graduated in Mechanical Engineering from m IIT, Roorkee (erstwhile ( University of Roorkkee) in the yea ar 1971 and completed his Masters Deegree in Appliied Thermoscieence in Mechanical Engineeriing from Roorkkee in the yea ar 1975. He also holds B.Sc B Degree froom Agra Univeersity and is ann FIE. He is Member, Central Electricity Regulatory Commiission (CERC) since Februa ary, 2009. Prioor to taking ovver as Memberr, CERC, Mr. Ve erma held the position of Meember (Planning g) in Central Electricity Authority A and Ex-Officio E Add ditional Secretary to Govt. of o India. Mr. Verma V has alsoo held charge of Member (Hydro) in CEA C for a brieef period. He has h been Direcctor General of o Bureau of Energy E Efficienccy (BEE) for three years in recent past.. Mr. Veerma belongs too Central Poweer Engineering Services of 19 971 batch. In his h long standinng career of ovver 36 years in the poweer sector in varrious formationns of CEA, Mr. Verma acquire ed wide and valuable v experience in plannning, thermal power plannt engineering, power project monitoring, project p construcction, supervisioon, operation m monitoring, hum man resource developmennt, grid operattion, renovationn and modernization of powe er plants and other o policy asspects. Planning g for power, load foreca asting, conserva ation and efficiency, national electricity plan, CDM, baseline data, etc. w were some of his h important responsibilitties as Member (Planning), CEA. Mr. Verma a also looked after a the fuel management, m R& &D and IT in power p sector. Mr. Verma took important initiatives to promote ennergy conserva ation, standard ds & labeling and energy efficiency in various secttors in the country. Mr. Veerma headed various v committtees set up by the Governme ent including woorking group oon National Action Plan for Climate Cha ange under thee National Misssion of Enhanceed Energy Efficciency. ‘Task Foorce of Formula ation of the action plan for developmennt of energy seector in the North Eastern Reg gion, Expert Co ommittee appoointed by MNRRE to study the geo-thermal based pow wer generating g potential in the Puga geoo-thermal field ds of Ladakh, J&K, Working g Group of re esearch and developmennt of energy sector s for 11th plan, 17th pow wer survey com mmittee and others, Memberr-Secretary of the working group on power for 11th pan set up by Planning Commission, played d a lead role in i 50,000 MW W of hydro pow wer initiative announced by Hon’ble Priime Minister, Publication of C02 C Baseline data in the India an Power sectoor and mapping of thermal power stations in the country for optimizzing the efficienncy of operatio on were spearhheaded by him m. Mr. Veerma has beeen a Member of the Standiing Committee e on research and developm ment in the Po ower Sector constituted by the Planninng Commission and the comprehensive R&D D perspective plan p was prepared under hiss leadership. Mr. Verma visited UK, USSA, USSR, Viettnam, Kenya, Guyana, Nige eria, Poland, Brussels and Geermany on various official assignmentss. More than 50 technical papers in the fielld of power se ector have beeen published annd presented by b him in the various natiional and interrnational seminnars and worksshops. Mr. Verrma has been responsible for power system m monitoring and grid operation o in thhe Eastern Reg gional Electricitty Board dealing with optim mization of geeneration and transmission capacities, inter-state and d inter-regiona al exchange off energy, gene eration scheduling and accouunting, etc. Mr. Verma has handled human resource management development d a system ma and anagement at Power System m Trading Instittute and Hot Line Training Center at Ba angalore for two years. Mr. Verma has be een conferred life-time achieevement award d by Central Board of Irrrigation & Pow wer and also byy the Bhopal Technological University. U Mr. Verma has alsso been on Governing G Coouncil/Board of o Directors of o various insstitutions like CPRI, NPTI, CWet, DVC C, etc.
Brief Biio-Data of Mr. M Rabindrra Kumar Satpathy S He holds a Bachelor degrree in Electrical Engineering. He has a professional experrience of more than 28 yearss in the area of renewab ble energy, i.ee. in the field d of solar phootovoltaic (PV) power system ms, wind poweer generation and others. Currently, he h is playing a key role in thee developmentt of the Solar PV P market in Inndia, South Asia a and central Asia A through his work at Trina Solar, Singapore. At present, he is the General Manager-Solar Energy Business Commercial Development off Trina Solar at a Singapore Office. Trinna Solar is the 4thlargest Solar PV Company in the world and is a Neew York listed company. He spearheads Trina Solar’’s initiatives and efforts in thee solar PV markket in India, So outh Asia and central c Asia. Prior too this, he was PresidentP RIL Solar S Group foor last +5.5 ye ears (2007-201 13). He was insstrumental to le ead a strong team, whichh has designed d, installed and commissioneed 2 prestigiouus solar projeccts in India–onee at Thyagara aja Stadium, New Delhi with a capacitty of 1MWp power p generattion and anothher at Khimshar, Rajasthan, Inndia, with a ca apacity of 5 MWp solarr power genera ation, having at a the highest CUF C amongst alll operating solar pv power p plants in India. He hass been instrumeental in building Solar PV Bussiness from inittial steps to larrge Solar PV ccompanies, suchh as Tata BP Solar India Limited (Now w Tata Power Systems Limiteed) between 1990 to 2002..Tata BP Solarr, a JV compa any between Tata’s and BP Solar, beccame the leading manufacturing and solarr system solutioon provider annd have been a dominant solar pv com mpany in India a and the SAARRC region. He hass extensive sola ar pv business experience ass GM of Solar World Asia /Shell Solar at Singapore offfice between 2002 to 20 007 and have been instrumenntal in Solar PV V business devvelopment in Soouth Korea, Chhina, Hong Konng, Australia, Thailand, Malaysia, M Vietnam and other Asia-Pac A Countries. He is currently c the Prresident of Sollar Energy Socciety of India (SESI) ( as well as Board Mem mber of Interna ational Solar Energy Sociiety (ISES). He is also membeer of Renewable Energy Committee of Confederation of Inndian Industry(CII).
B Brief Bio-da ata of Prof. P.P. Mathu ur Prof. Premeendu P. Mathur, Ph.D., Vice-C Chancellor, KIIIT University, Bhubaneswar, B India was Dea an, School of Life L Sciences and Professsor of Biocheemistry & Mollecular Biologyy and Head, Center of Exxcellence in BBioinformatics, Pondicherry University, Pondicherry. P H received M.Sc.and Ph.D. degrees He d from Banaras B Hindu University, Va aranasi. He hass contributed significantlyy in the field of o male reprod duction, reproductive toxicolo ogy and bioinfoormatics and hhas more than 33 years of teaching annd research exxperience. He has published d around 150 papers/ revieews in variouss high impact journals j and books and participated inn more than 10 00 national annd international scientific confferences. He is on the Editoria al Boards of Asian Journnal of Androloogy (Nature Puublishing Grouup), Spermatog genesis (Landees Biosciences), Open Androllogy Journal (Bentham), Advances A in Andrology A (Hind dawi), Asia Pa acific Journal of o Reproductionn, Everyman’s Science (ISCA)) and Peer J (USA). He is Reviewer for more than seeventy five nattional and inte ernational journnals. He is reccipient of many y prestigious awards suchh as Young Sciientists’ Award (ISCA), Rockefeller Foundatiion Special Posstdoctoral Felloowship Award, Rockefeller Foundation Biotechnologyy Career Awarrd, INSA-Germ man Academy (DFG) Exchange Programmee, Dr. P.N. Sha ah Memorial US Vitamin (India) Oratioon Award, ICM MR Internationa al Fellowship Award A for Sennior Indian Bioomedical Scienttist, Subhash Mukherjee Memorial M Infar India Orationn Award and Dr. D K.K. Iya Me emorial Oratioon (NDRI). He hhas travelled widely w within India and abroad. He is i on the Com mmittee on Reeproduction annd the Environnment of the Society for thhe Study of Reproductioon, U.S.A. He iss President of the Section of Animal, Veterrinary and Fishhery Sciences oof Indian Sciennce Congress 2014. Prof Mathur M has beeen Visiting Prrofessor/ Scientist at The Population P Couuncil, Rockefeller University, New York, Cleveland Clinic, Ohio, USA U and W.W W. Universitat, Muenster, Ge ermany. He has been Chairman/ Membe er for many national annd Internationa al Committees. He has beeen member/ Coordinator C off NAAC peer teams to mo ore than 50 institutions and a has beenn Chairman/ Member M of ma any academic/ / scientific com mmittees. He w was member of o Executive Committee of Indian Sciennce Congress, National N Task Force F on Bioinfformatics and Infrastructure FFacilities of the Department of Biotechnoology (DBT) annd Departmentt of Informationn Technology, Government of o India. He is oon the Nationa al Task Force on Fertility Regulation & Expanding Coontraceptive Chhoices of Indian Council of Medical M Researrch and Memb ber, Scientific Advisory Group (SAG) for Division of Reeproductive annd Child Healthh of Indian Couuncil of Medica al Research.
*************** *****
Brief Bio-Data of Prof. (Dr.) Sasmita Rani Samanta Prof. (Dr.) Sasmita S Rani Samanta, is a young, y dynamicc, dedicated, hard h working and a astute adm ministrator havving one and half decadees of rich and diverse experrience in Adminnistration of accademic institution in various capacity. Herr outstanding managemennt ability, perssonal and interpersonal effecctiveness, enterrprising qualityy in policy implementation, leadership etc have been appreciated inn different foruum. She held leeadership roless in KIIT Univerrsity in various fields starting from Human Resource Management, M T Talent Acquisitiion, General Administration, A Institute Goveernance and C Competency Management. M She has takken several innoovative steps for f engaging and a retaining people, p bringinng transparencyy in system lea ading to high motivation, ownership annd commitment among the employees, e thuus increasing their t productivvity for accomplishment of organizatioonal goals. She accted as the Loccal Secretary of o 99th Indian Science Congrress, the mega scientific evennt of the world d held in KIIT University Campus C in 20 012, which wa as attended by b large numb ber of delega ates including Nobel Laurea ates, Eminent Scientists, Policy Makers, Educationists, E e from across the Globe. etc Becausse of her sinceerity, dedicatioon, loyalty and d hard work, she was award ded Best Staff Award–2007 7 by the KIIT University. She received ‘Rajeev Smrutii Award 2010 0’ for Administration. She is also received Swami Viveka anand Youth Award 2013 for administration. Presently working as the Registrar and Director (A Admissions) of KIIT University, Bhubaneswarr. She is the life e member of ISTE, ISC, SEESI etc.
Brief Bio o-Data of Prrof. (Dr.) Deeba Kumar Tripathy Prof. (Dr.) Deba D Kumar Tripathy T has jooined as the Pro-Vice P Chanccellor of KIIT University. U He hhad a illustriouus career of more than 40 4 years with several s accomp plishments in Teeaching, Resea arch in the area a of Productionn Engineering and a Polymer Engineering g, heading of large Academic Institutions with tremendouss success markeed by various National and International Award, fellowship, patentt and publicatioons.
Previous Assignment A • • • • • •
Assst. Prof. in REC C Rourkela (19 971–1984). Assst. Prof./ Prof. IIT, Kharagpuur (1994–2011 1). Viisiting Scientist in North Londoon Polymer, UK K (1986). Deean (Student Affairs), A IIT, Kha aragpur (2007 7–2009) Diirector, Nationa al Institute of Technology T (NITT), Worangal (2003–2005). ( Viice Chancellor of VSSUT, Burla (2009–11).
Fellowship p • • •
Feellow of Instituttion of Engineers (India). Feellow of Indian Society of Theeoretical and Applied A Mecha anics (ISTAM). Feellow of Indian Rubber Institutte.
Award • • • •
Beest Technical Pa aper Award frrom Institute of Engineers (India). Diidactical Enviroonmental Educa ation Doctrine (DEED) ( Award from Coonfederation of o Indian Universities (CIU) annd United Natio ons. Liffetime Achieveement Award frrom World Insttitute Building Programme. P Diistinguished Aluumnus Award-2 2012, NIT, Rouurkela.
Details of Research Pub blications National & International Journals J 0, In Process–10 Published/ Accepted–160 National/ Innternational Coonference Published–1 120
Patent Aw warded A method foor Manufacturee of an Elastom meric Pad India an Patent, Pate ent Certificate 211264.
Administra ative Experien nce a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q. r.
Prro-Vice Chanceellor, KIIT Univeersity, Bhubaneeswar. Viice Chancellor, Veer Surendra a Sai Universitty of Technolog gy, Burla, Odisha. Diirector, Nationa al Institute of Technology T Wa arangal. Syyndicate Memb ber, North Odisha University, Baripada. Seenate Member, National Institute of Technollogy Rourkela. Seenate Member, National Institute of Technollogy Jamshedp pur. Heead, Rubber Teechnology Cenntre, IIT Kharag gpur. Chhairman, Instituution of Engineeers (India), Kha aragpur Centre e. Prresident, Indiann Society of Theeoretical and Applied A Mecha anics. M Member, UPSC Selection S Board. Chhairman, Comm mercial, Establishment and Liccensing Committtee, IIT Kharag gpur. Chhairman, Assesssment Committtee, AICTE, New w Delhi. M Member, Policy Planning Bodyy, Govt. of Odisha. Chhairman, Apexx Committee, Jooint Engineering g Entrance, Od disha. Chhairman, Task Force on Adva ance Technology, Departmentt of Science and d Technology, Govt. of Orissa. In-charge-Metroology Lab., Dynnamic Mechaniccal Analysis La ab, REC Rourkeela. Prrofessor-incharge Training annd Placement, Examination E att IIT Kharagpurr. M Member of accreditation Comm mittee, NIT Rouurkela, Orissa.
Brieef Bio-Data of Rajinder Kumar Ka aura CMD Berg gen Associattes Pvt. Ltd. An Exemplary academiccian, Shri Rajinder Kumar Kaura K comes from an imprressive backgrround in Electtrical Power Generation, Evacuation, Transmission T annd Distribution. An active purssuit in knowledge has alwayss guided Shri Rajinder R Ji to aim for the highest standa ards in Learninng, Teaching and Guiding hiss fellow peers and employeees, as he head ds one of the major Electrronics & Renew wable Energy firms f in India, Bergen B Associates Pvt. Ltd. His 30 0 plus years off experience in i the work fieeld comes from m the govt. ass well as the pvt. Sector, where w he has successfully brought world d class technoloogy to Indian grounds and enabled e growtth in the Indiann Automobile, Electronics E & Renewable Energy Field. Time and a again, his work w has prop pelled him to distinguished po osts such as Preesident SMTA ((Surface mountt Technology Association)); former presiident of IPCA (Indian Printed d Circuit Associiation); Foundeer Member of EDCI (Electroniic Designer’s Council of Inndia) and now the V.P. SESI. An outright Scholar since s early dayys, his NORAD Scholarship le ed him to Norw way, where he was influenced immensely by the pow wer of the sunn and its impa act in the lives of people evverywhere. Hee returned bacck with the aim m of 'doing something substantial s for his h country’’; a dream which today t is bringing light into the lives of counttless Indians. Having g successfully carried out turnnkey projects for f major indusstry giants like Videocon, Bha arti, Moser Bae er, Indosolar and activelyy engaged in upgrading thee education standards in IIT's, ITI's and moree, he has always believed in the blessing or Lord Sarraswati to enab ble people's am mbitions and sttrengthen their resolve in sustained develop pment for all mankind. A humb ble person by nature, his actiive pursuit lies in empowering g youth and rural India, whichh is a passion most m dear to his heart. Having H successffully executed nearly 40 rural electrificatio on programmees, Mr. Rajindeer Kumar Kaura lies at the pinnacle of what can be described d as ouur personal triuumph of light over o darkness.
*************** *****
Brief Bio-Data of Ballour Singh duation in Civil Engineering in the Year 19 982, working in i the field of Renewable Ennergy; Earlier with Punjab After Grad Agro Industtries Corporatiion and then transferred t to Punjab Energy Developmennt Agency duriing September 1991. The major Projeects undertakenn during this peeriod are Mini Hydel Projectts on Canals, Cattle C Dung based Biomethannation Power Generation Project, Bioma ass- Cogenerattion, Solar Pow wer Projects and d Energy Efficiiency Projects. Attended UNDP supp ported training g programme on o renewable energy with specific focus oon Bio-energy in i US during 1996 and two t week trainning course on energy e efficienncy in Japan during 2008.
Brief Bio-Data of Ajay Prakkash Shriva astava Maharishi Solar Techno ology Private Limited Mr. Ajay Prakash Shrivasstava is the Promoter and President P of Maharishi M Solarr Technology PPrivate Limited d (MST). The Company has set up a Factory in Andhra a Pradesh for manufacturing:: Multicrrystalline Wafeers, Mono/ Multicrystalline Soolar Cells, SPV V Modules, Sola ar Lighting Systtem, Solar Wa ater Pumping System, Solar Power Plantts. Maharrishi Solar is thee first Company in India to esstablish Multicrystalline Solar PV Wafers manufacturing fa acility. Maharrishi Solar has also establisheed a factory inn NOIDA, near Delhi for ma anufacturing Coollectors for So olar Thermal applicationss and Solar Steeam Generatinng systems.
Solar Energ gy Society of India (SESI) Mr. Shrivasstava is the im mmediate pastt President of Solar Energy y Society of India (SESI). Thhe Society is engaged in promoting the t awareness and knowledg ge of Solar Eneergy in India.. He is also a on the Boa ard of many ed ducational and charitable Insttitutions on natiional and international level.
*************** *****
Brrief Bio-Datta of Venka ateswara Ra ao C Chief General Manage er (TS)-IREDA A Mr. B. Venkkateswara Raoo is currently working as Chhief General Manager M with Indian Renewable Energy Development D Agency Ltd d. (IREDA) and was handling g different porrtfolios for the e last 19 yearrs at IREDA. He was involve ed in various activities ra anging from Co-ordination C with the Interrnational Agenncies for raising funds, Nettworking with National & Internationa al Organizationns, Consultancyy Development etc. Currently associated in financing off Bio Sector, Buusiness Develop pment, Recoverry and overall In-charge of IR REDA Branch and Camp Offices. Earlieer he was worrking with M/s. Tata Electric Company, Muumbai and sub bsequently worrked in M/s. Best & Crom mpton Engineeering Pvt. Ltd., Chennai and A.P. A Industrial and Technical Consultancy O Organization (APITCO) for nearly 13 years. y He is Engineering Graduate G from m Andhra University Engine eering Collegee, Vishakhapattnam and obttained Postgraduation in Mechanicall Engineering from f IIT, Mumb bai and MBA from f Institute of o Public Enterrprise, Osmania a University, Hyderabad d. He speent nearly threee decades in Renewable Ennergy field covvering various sectors includiing Manufactuuring, Project consultancy, Project Financcing, Policy Pla anning at differrent capacitiess.
Brief Bio-D Data of Rav vi Khanna Chieff Executive Officer-Sola O ar Power Buusiness Aditya Birlla Group. Mumbai India Ravi Khanna serves as Chhief Executive Officer of the Solar Power Business B of the Aditya Birla G Group, providing strategic leadership and direction to t the Solar Poower business. O and Presidennt of Scatec Soolar AS, Oslo, Norway N and Prior too joining the Aditya Birla Grooup, Ravi workked as the CEO served as Director D Scatecc AS. Ravi hass also served as a Chief Execuutive Officer of o Moser Baer Photovoltaic Ltd., L and PV Technologiees India Limited d, where he creeated one of thhe fastest manuufacturing capa acities in Asia. Beforee entering the solar s sector, Ravi worked wiith Delphi Corp poration US in various management positio ons globally, including Coountry Manager ASEAN and d MD for India a and Thailand d. Under his leeadership Delp phi Steering India won the coveted Toyyota best supp plier award forr manufacturing g excellence. Ravi ha as served as a member of thhe Indian PM’s solar technolog gy mission and d also engaged d very closely with w Ministry of New annd Renewable Energy (MNRRE), IREDA (Ind dian Renewab ble Developmeent Agency) and NORAD (D Development agency of Norway) N on deevelopment of the future reneewable and susstainable energ gy platforms inn India. He hass also served on the board off American firm ms HelioVolt Co orporation, Sol Focus, Solaria a and Stion Corrporation.
Qualificatiion Details •
M Master in Business Administration from Symbiiosis Inst. of Mg gmt, Pune in 19 982
•
B.E. (Mechanical) from Punjab Engg. College.., Chandigarh in i 1980
Additionall Professionall Qualification ns Trained in top-of-the-linne Leadership Programs administered by b the University of Michig gan Business School and Thunderbird d Business Lead dership Program. Certifieed by Lean Coollege at Ann Arbor, A Michiga an (2002) and Manufacturing g Academy (20 003) Delphi Headquarters Troy, Michig gan. Certifieed Innovation and a Continuouss Improvement Methodology (I & CIM) Six Sigma S Green BBelt.
*************** *****
Brief Bio-Da ata of Dr. Ranjana R Jha a Dr. Ranjana a Jha is presenntly working as an Associate Professor at Netaji N Subhas Innstitute of Techhnology at Delhhi University. Her area off research interest is Solar Ennergy Utilizatioon. Her researcch work mainlyy been on Solar Air, Water Heater, H Solar Passive Architecture and in Green Technnologies. She ha as more than 40 4 publications in internationa al journals and conferences in these fields. She has unndertaken seveeral research projects p from AICTE, A DST and d PCRA. She iss the life memb ber of Solar Energy Sociiety of India and Indian Chapter of Interna ational Centre of o Theoretical Physics, Italy. C Currently four students are registered under u her for their t Ph.D. work. She has atteended and pre esented her papers in variouss national and international conferencess in various couuntries like U.SS., U.K., Australia, Switzerland and Nepal. She has been reviewer of many m papers for many coonferences in U.S. U and Germa any.
Brief Bio--data Prof. Swati S Ray Dr. Swati Ray R was a sennior Professor of Indian Assoociation for the e Cultivation of o Science, Kollkata. She hass contributed significantlyy towards devvelopment of Amorphous A and d Nanocrystalline silicon bassed Thin Film ssolar Cells and d technology transfer to industry. She has h 5 patents, more m than 150 0 papers in inte ernational journals, 60 papers in Int. Conf. proceedings in her credit. She received d Material Research Society of India medal award in199 99 and Photovooltaic R&D awa ard of Solar Energy Socciety of India in i 2006. She is i a member of o international Advisory com mmittee of Int. PVSEC and was w Program chair of 18th Internationall PVSEC. She was w vice-chairp person of SESI for many yea ars and fellow of West Bengal Academy of science. She worked inn Tokyo Institutte of Tech., Ja apan; Ecole Po olytechnique, Frrance; Institutee of PV, Juelichh, Germany; National Reenewable Enerrgy Laboratoryy, Colorado, USA; Heriot Wa att Univ, Edinboourgh. She hass handled seve eral research projects havving international collaboratiion, attended many m international conferencces and visited d several resea arch institutes throughout the t world.
*************** *****
Brrief Bio-Data a of Shri Sa angeet Shukkla M.SSc. (Physics & Electronics), ICWA (Innter) CISA and d a Researchh Scholar in Physics at IIT, Kanpur A career banker, b joined State Bank off India as a Probationary P Officer O in 197 74 and retired d as its Deputy y Managing Director in 2011. Held important i and high profile assignments inn SBI. Deputy General Manngers of Industtrial Finance Branch, Indore and Overrseas Branch, Kolkata. Zona al Manager off Kolkata and d Mumbai Zonnes of the Bannk. Head of Syndicationns & Investmentts at Bank’s Lonndon Office foor more than 4 years. Genera al Manager in--charge of Bannk’s Network in the Statees of Maharashhtra and Goa as controller of o more 650+ branches. Chief General Ma anager (Personal Banking) at Corporate Office in chharge of the Retail R portfolio of Bank with portfolio exceeeding Rs. 240 0,000crores. De esigned and introduced many innovativve products annd internet secuurity trading plate form. Chieef General Ma anager in-charge of Bank’s Internal Aud dit functions at Bank’s Corporrate Office at Hyderabad. Ex xecutive Vice Chairman C and CEO of SBI Gllobal factors Ltd., an induustry leader in factoring. Second ded to Govt. of o Madhya Pradesh as Genneral Managerr (Project Finannce) in MP Sta ate Industries Development D Corporationn (1986-87) and a again as General G Manager (Finance and a Project) att MP State Eleectronic Corporration(198892). Membeer of the Projeect Managemeent Team of coountry’s first Fib bre Optic cablle and Opto-eelectronic proje ect for Optel Telecommunnication Ltd. Membeer SBI Mckinseey In-house Credit C Process Team. Worke ed on refinem ment of credit assessment, sanction s and monitoring processes p of both MSME and d Large Corporrates. Now Senior Advisor of o Risk Manageement and Corrporate Bankinng in Indian Bannks’ Associationn. Membeer of British Meensa, Avid read der and Bridgee player.
Brief Bio-Datta of Dr. S.K K. Samdarsshi Dr. S.K. Sam mdarshi, Professor, Central University U of Jhharkhand (CUJ)), Ranchi is an alumnus of Ind dian Institute off Technology Delhi, India a. He has been actively assoociated with the Solar Energ gy research inn the country and has made e substantial contributionn in design, sim mulation, materrials, applications and system ms’ research and a developmeent in the dive erse area of solar and reenewable energy technologies. These resullted in a number of importannt research pub blications (45 International, 22 Nationa al), one book and a one nationnal patent. He has guided six x Ph.Ds and foour more are w working under him. He has executed 07 0 research and a 04 developmental projects in differrent areas of solar energyy. Some of hiis significant contributionns in the area of visible photocatalysis include developm ment of highly active homojuunctions of titannia and zinc oxide nanoosystems and their consolida ation under doped, sensitized and comp posited conditioons. He has also a done a collaborativve work in the development of o bio-mimicked hierarchical morphology off the photocata alyst systems at a nanoscale. He has a nuumber of natioonal and internnational collaboorations. His cuurrent researchh interest is solar photocataly ytic and new generation photovoltaic materials, m solarr thermal systeems performannce parameterrs, and renewa able energy ed ducation. He has contribuuted substantia ally at nationa al and international level as member of different bodiess and also as Head of the Deptt of Energy, Tezpur Central C Universsity, Tezpur Asssam, India. A former f Governning Council meember of the Solar S Energy Society of India(SESI), Proof Samdarshi was w convener off ICORE 2011((International Congress C on Reenewable Energ gy 2011).
*************** *****
Brief Bio--Data of Drr. S.M. Ali Designationn
:
P Professor, Electtrical Engineeriing & Assoc. Deean (T&P), KIITT University, Bhubaneswar.
Qualificatioon
:
M. Tech from Calcutta Unive M ersity Ph.D in Electrical Enginneering from International U University, Caliifornia, USA in 2006. D in Electrica DSc al Engineering, International University, U Califfornia, USA in 2008.
Research & Development
:
His area of ressearch in the field of Renewa H able Energy b both Solar & Wind W Energy. H had also guided He g five no os. of Ph.D stud dents in his reesearch area. He has also p presented moree than 60 pap pers in differennt national & innternational co onferences in t field of Renewable Energ the gy apart from around 35 noos of paper alsso published in National and d International journals.
W Seminar & Workshop
:
He has conducted several nos. of Seminnar, Workshop H p and short te erm training p program for thhe Faculty mem mbers Engineerring College, PPolytechnic in collaboration c w AICTE, ISTE, MHRD DST, & Ministry of Industries, Govtt. of with
Professional Society
:
PPresently Vice--President (EAS ST) Solar Enerrgy Society off India National Executive C Council membeer of ISTE, New w-Delhi Secretary Institution of Engineers India, Orissa S State Center ExE Vice-Chairm man, Indian Insstitution of Industrial Enginee ering, Orissa C Chapter Ex-Chairman, ISTE, Orissa O Section Ex-Sectional C Committee mem mbers of the Indian Science Congress Association, Kolkata.
Awards Recceived
:
FFor outstanding g contribution inn the field of science and tecchnology including research a maintaining quality contrrol of technical education, prrofessor Ali wa and as felicitated m more than fiftteen National & Internationnal Awards like Sadananda Memorial A Awards, Madhhusudan Memo orial Awards, ISTE Calcutta a Conventationnal Awards, U UWA Life Time Achievementt Award and Leading Educators of the worlds w 2009 f from many preestigious organiizations of Indiia and abroad.
India Assocciation with th he
L Member of CSI, IETE, IIIE, AIIMA, ISCA, ISTE, SESI, & ISSLE Life
Briief Bio-data a Dr. Vinod Krishna Seethi Rector, RGP PV and Directo or, UIT-RGPV, Head Energy & Environmen nt 1. 2. 3.
4. 5.
BEE (Mechanical) from Roorkee University (IIT// R) 1971 batcch. Seerved Govt. off India, Ministryy of Power throough selection in IES–UPSC 1974 batch. Also have done a PG course from United King gdom and Ph.D D. from I.I.T., Deelhi in Mechaniical Engineering g (Energy / Thhermal Engineering). Exx-Director Ministry of Power, Government of o India, and have worked foor over 24 yea ars with the Ministry, out of which 14 yearss as Dy. Direcctor Consultanccy (Site) for Uttar U Pradesh State Electriccity Board, Annpara Super Thhermal Power Station (Singra aulli region). Was W responsible for overall execution e of 3 3x210 MW + 2x500 MW Annpara Thermall Power Plant In the field of ‘Academics’ ha ave worked att National Pow wer Training Innstitute (NPTI), Government of o India, for ab bout 9 years and a 13 years at Professor leevel in R.G.P.V V. Bhopal & Ossmania Universsity, Hyderaba ad. Presently Diirector UIT, RG GPV, Bhopal, M.P. M since Marchh 2004. Ass researcher inn the field of Energy & Envvironment, have worked oveer 4 years in tthe Departmennt of Atomic Ennergy and during last 40 yea ars carrier havee contributed about a 100 pap pers.
Membersh hips • • •
Member of ASM M ME, USA. Feellow of Institutte of Engineers (I), Chartered Engineers. Liffe member Gloobal Institute of Flexible Systems Managem ment.
Awards off Outstanding g Work in the Field of Energ gy Power • • •
Jeewel of India award. a Int. instittute of Education & Managem ment, New Delhi. Liffe Time achievement gold Meedal award… Int. Business Co ouncil, New Deelhi. M Mother Theresa Award (2006)). International Business Counccil, New Delhi.
Experiencee • • • •
Tootal: About 41 Years. Teeaching at PG level: 22 Yearrs. Industrial Consultancy (Power): 11 Years. ars. Reesearch: 8 Yea
*************** *****
Brieef Bio-Data of Mr. Gau utam Mohanka Mr. Gautam m Mohanka, a solar entrepreeneur for more than 11 yearss, has led Gauttam Polymers tto be a formid dable player in solar offf grid industry. A graduate from SRCC annd MDI, he ha as several prod duct patents a and innovationss in business models to his h credit. He passionately p chhampions the cause c of India to be a glob bal leader in solar off grid and a strongly believes tha at if the challenge is taken up earnestly byy the Industry & government, then India cann be a technolo ogy supplier and a net exporter e in the off-grid Solarr space.
Brief Bio-D Data of Dr. P.C. Pande Dr. Piyush Chandra C Pande, a well know wn scientist for his creativity, has been carrrying out resea arches in the field of solar energy utiliization for the last 37 years after his selecction in the firsst batch of Ag gricultural Reseearch Services of ICAR. Dr. Pande, a meritorious m schoolar, did his schhooling at Birla a Vidya Mandir, Nainital, gra aduation from DSB College Nainital, N Post graduation in Physics froom University of Roorkee (nnow IIT Roorke ee) and Ph.D on CdS solarr cells from University of Durham, United Kingdom m. He did adva ance course onn Photovoltaic systems s at SIESS, Urbino, Italy in 1983 and participated in Follow-up p programme on Internationa al Cooperationn on Renewable Energies at Finmeccanica, Rome in 1986 6. Dr. Pande was awarded Commonw wealth Fellowsship in 1981 foor his higher sttudies and Marie Curie Fello owship by Euro opean Union during 1995 to carry out advance reseearch work on the developme ent of thin film solar cells and integrated PV P system at University of o Northumbria, Newcastle upon Tyne, UK. U He was bestowed with the prestigiouus Hari Om Ashram A S.S. Bhatnagar Award in 200 01 for his meriitorious researcches leading to o the developm ment of novel solar devices for domestic agricultural and industrial applications. In 2006 he wa as honoured witth Aryabhatta Samman at N New Delhi. Dr. Pannde provided consultancy too army for usinng solar energy y in high altituude areas like Leh. He has presented p TV program on solar energy appliances for the nationnal network programme p of UGC. Dr. Pa ande is avid sportsperson, s represented d university in Cricket, C Hockey and Table Tennis. He was awarded coloour in Hockey b by University of Roorkee in 1974, won Hartfield Bridge Trophy in Durham D in 198 83, winner of Table T Tennis Chhampionship off ICAR in 1986 6 and 1990. He coordina ated the organnization of ICA AR Zonal Sportts Meet, Golde en Jubilee Ora ation Series annd Golden Jubilee Function at CAZRI. Jodhpur. J He has h been contrributing as meember of the main committeee of BIS for Non Conventional Energy Sources, Governing G Counncil of SESI (2009-10), Chhairman, Consuultancy Processsing Cell of CAZRI (2003-2012) and Chairman, Central C Works Committee. He continues to be a member of editorial Booard of Annalss of Arid Zone since 2001. Dr. Pande has h visited different laborattories in Francee, Germany, Belgium, B Switzeerland, Italy annd UK. He hass contributed some one hundred h and fifty f research papers. Dr. Pa ande chaired the Division off Energy Mana agement, Engineering and Product Proocessing during g 1990-93 and d at present is Principal Scienntist and Head d, Division of A Agricultural Eng gineering for Arid Producction Systems at a CAZRI. Jodhp pur.
*************** *****
Brief Bio-Data B of Prof. Indrajjit Mukhopa adhyay Prof. Mukhopadhyay is heading the School S of Solar Energy, Pa andit Deendayyal Petroleum University, Gandhinagar, G Gujarat. Beside, establisshing world class c laboratoory to carryouut both funda amental as w well as applie ed research, Dr. Mukhop padhyay is acttively involved in proliferatioon of solar ene ergy technolog gies among thee mass throughh the M.Tech and Ph.D. program p in the University. Earlier Dr. Mukhopa adhyay had worked w for CSSMCRI, Bhavna agar, one of the CSIR labooratories of Inndia for the developmennt of economiically viable Industrial I Proccesses. His exccellent contribution has beeen recognized by Vikram Sarabhai Award A in 2006 and CSIR Rura al Technology Award A in 2010 0, as team mem mber. Dr. Muukhopadhyay received r his Pd d.D. from the Chemistry C Dept. of IIT, Mumba ai. He had visitted many famo ous research Institutes likke Weizmann Institute of Scieence, Israel, Na agoya Institute e of Technologyy, Japan, Karlsruhe Universitty, Germany and Shinshuu University, Ja apan as visiting g Scientist and Post doc. Fello ow. He has five national and d international patents and more than fifty f research papers p in the Joournals of Interrnational Repute. His current research interests are: 1. 2. 3.
aterials for Efficient Energy C Conversion and d Storage. Syynthesis and chharacterization of novel Nanoo-structured Ma Noon-Vacuum based techniquess for fabricatioon of Thin film Solar S Cell and Sttudies on novel inorganic matterials for Li ionn secondary ba attery.
Contents Messages Brief Profile of Speakers 1. Transforming India into a Green Nation: Bottom up Approach Through Solar Energy Education, Training and Innovations J.P Kesari, Naveen Kumar, P.B. Sharma and Lavnish Goyal
v xiii
1
2. Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System T. Ramana, V. Ganesh, S. Sivanagaraju and K. Vasu
9
3. Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator Gezahegn Habtamu, S.C. Mullick and Sanjeev Jain
15
4. Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector Piyush Choudhary
28
5. A Prudent Study on Solar Water Heating System using RETScreen in India Sambeet Mishra, Sayed Majid Ali, D.P. Kothari and Pratyasha Tripathy
34
6. Solar Power: A New Paradigm to Combat Energy Poverty P. Natarajan and G.S. Nalini
41
7. Thermal Performance Evaluation Methodology of a High Rated Solar Distilled Water System Navneet Deval and Abhishek Saxena
8. Design Development and Performance of Solar Desalination Device for Rural Arid Areas N.M. Nahar, A.K. Singh, P. Sharma and G.R. Choudhary
9. Solar and Wind Power: Potential and Opportunities in Western Rajasthan K.R. Genwa, Sarita Boss and Shradhha
10. Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects Sameer Pikle
45
50 53
60
11. Performance Optimization of Solar PV Assisted Systems to Meet the Standards of Zero Energy Buildings G.R.K.D. Satya Prasad, K. Vijay Kumar Reddy and Ch. Saibabu
72
12. Experimental Studies on Biomass Pyrolysis using Microwave Radiation Lohit Kamble, Madichetty Sreedhar and A. Dasgupta
13. Role of Software Tools for Support of Renewable Energy Systems Deployment Srikant Mishra, G.R.K.D. Satya Prasad, Sujit Kumar Patro and Rati Ranjan Sabat
77 84
14. Cost Representation and Propsed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer Sthita Prajna Mishra, S.M. Ali and Prajnasmita Mohapatra
90
15. Empowering the Renewable Energy Sector Through Skilling Y.P. Chawla and R.S.P. Singh
97
16. Meeting India’s Energy Demand Through Concentrated Solar Power Technologies J.P. Kesari, Nadim Shams, P.B. Sharma and Rajinder Kumar Kaura
17. Development of a Photo Sensor, PIC Micro Controller based Solar Tracker A. Gokul Raj, T.V. Chavda and V. Siva Reddy
104 110
18. Analysis of the Effect of Nanostructured Paste Components on the Efficiency of Dye Sensitized Solar Cells: A Taguchi Approach Sarita Baghel and Ranjana Jha
117
19. Recent Developments in Standalone PV System: For Rural Electrification S.M. Ali, Punyashree Pattanayak, Sushree Subhadarsini and Prasun Sanki
121
ICORE 2013
20. Biogas—An Ideal Source of Energy S.S. Sooch and Jasdeep Singh Saini
21. Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still J.S. Gawande and L.B. Bhuyar
129
132
22. Nano Materials Coated Photovoltaic Device for Enhanced Solar Energy Conversion P.H.V. Sesha Talpa Sai, J.V. Ramana Rao, Devarayapalli K.C. and K.V. Sharma
141
23. Microalgae: Bio Energy Technologies for Sustainable Future Piyush Choudhary and Varteka Tripathi
24. Development and Performance Studies of a Community-Size Building-Material Solar Cooker R.C. Punia, V.K. Marwal, S. Mahavar, P. Rajawat and P. Dashora
146
155
25. Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households in Neh Region of India Mahendra S. Seveda
160
26. Designing and Testing of the Battery Performance of a PIC Microcontroller based Solar Wind Hybrid System During the Load and no Load Conditions B.K. Sahu and P.K. Choudhury
171
27. Performance Testing of Non-Edible Oil on Four-stroke Diesel Engine by using Lost Cost Additive A.K. Rout, M.K. Parida, S.K. Nayak and S. Das
28. Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell M.C. Adhikary and R.M. Pujahari
177
181
29. A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method K. Vimala Kumar, V. Ganesh and J. Suresh
187
30. Weibull Distribution Analysis of Sub-Hourly Surface Wind Data of Western Rajasthan P. Santra, P.C. Pande, Preeti Varghese, N.D. Yadav, P. Raja and N.K. Sinha
194
31. Design and Development of Low Cost Solar Cooker for Rural People Digvijay Singh, A.K. Singh and S.P. Singh
201
32. Short-Term Load Forecasting of an Integrated Renewable Energy using a Smart Grid Pradeep Kumar Mohanty, Neha Pasayat and Rakesh Mohanty
204
33. Green Roofing–Addressing Climate Change & Reducing Energy USE Sipra Das Mohapatra, Sushree Shataroopa Mohapatra and Swagatika Satpathy
34. Wind Energy Application in Water Pumping System by Hybrid System
S.M Ali, Ajanta Priyadarshinee, Deepti Priyanka Behera and Surya Prabha Biswal
213 220
35. Ni Doping Induced Structural and Optical Modifications in TiO2 Nano-Structure Photocatalyst for Efficient Use of Solar Energy Biju Mani Rajbongshi, Anjalu Ramchiary, Sunny Bharali and S.K. Samdarshi
224
36. MATLAB/ SIMULINK based Modeling of Solar Photovoltaic Module Babita Panda, Sulagna Das and Bhagabat Panda
228
37. Gender based Energy in the Development Strategy of Indian Scenario Bibhu Santosh Behera, Anama Charan Behera and Pradeep Mohapatra
235
38. Studies on Solar Drying of Kinnow (Citrus) Peel Tarsem Chand, Sukhmeet Singh, Sajeev Rattan Sharma and Suchi Gupta
246
xxvi ♦ Contents
ICORE 2013
39. Thermodynamic Analysis of Evacuated Tube (ETC) based Hybrid Ammonia-Water Refrigeration System A. Gupta, Anshu Sawhney, S. Anand and S.K. Tyagi
251
40. Application of a Brønsted Acidic Ionic Liquid as Catalyst in Esterification of Oleic Acid for Biodiesel Production Subrata Das, Ashim Jyoti Thakur and Dhanapati Deka
261
41. Enhanced Visible Light Active Silver Sensitized Nitrogen-Doped Titania for Photocatalytic Applications Anjalu Ramchiary, Bijumani Rajbongshi and S.K. Samdarshi
265
42. Off-grid Application of Solar PV S.M. Ali, Shubhra, Priyankari Pattanaik and Supriya Pattanayak
271
43. A Theoretical and Experimental Study of Glaze and Insulation Materials for Efficient Solar Thermal Appliances Mahavar S., Rajawat P., Punia R.C., Marwal V.K. and Dashora P.
280
44. Studies on PV Clad Structure for Controlled Environment P.C. Pande, A.K. Singh, P. Santra, S.K. Vyas, M.M. Purohit and B.K. Dave
294
45. Making Hydrogen Economy a Reality Smrithi Ajit
298
46. Advanced Technology in MPPT by using Minority Charge Inspired Algorithm (MCI) Applied to PV System Soumyadeep Ray, Sreedhar Madichetty and Abhijit Dasgupta
304
47. Heat Transfer in Heat Pipes for Solar Water Heating Vishal Bhasin, S.S. Sikarwar, J.K. Verma and Abhishek Saxena
313
48. An Experimental Study of an Alternative Energy Source for Greenhouse by a Hybrid Solar Photovoltaic Thermal System and Earth Air Heat Exchanger M.K. Ghosal, Sujata Nayak and N. Sahoo
325
49. India One Solar Thermal Power Plant: A Case Study Kumar Abhishek, Kandarp Mehta, Prashant Mishra, Abhijit Ray and Nanji J. Hadia
335
50. Fabrication and Experimental Study of a Solar Cooker with Electrical Back-up Rajawat P., Mahavar S., Verma M., Sengar N. and Dashora P.
347
51. Solar Thermal Technology in India: Issues and Opportunities Piyush Trivedi, V.K. Sethi and Mukesh Pandey
354
AUTHOR INDEX
362
Contents ♦ xxvii
Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations J.P Kesari1, Naveen Kumar2, P.B. Sharma3 and Lavnish Goyal4 1,2,3,4Delhi
Technological University, Bawana Road, Delhi–110042, India
INTRODUCTION India has tremendous energy needs and is finding it difficult to meet those needs through traditional means of power generation. Electricity consumption in India has been increasing at one of the fastest rates in the world due to population growth and economic development. India’s economic growth has been some what stalled because the amount of energy generated is inapt to keep the growth motor of India running at the desired speed and there are energy shortages almost everywhere in the country. Now, The question arises, What can India do to meet the future energy demands and help eliminate wide–ranging power outages in the future? To this the answer is, “India is the Saudi Arabia of renewable energy sources and if properly utilized, India can realize it’s place in the world as a great power which can not only take care of it’s own energy needs but also of other nations. Renewable energy also has the advantage of allowing decentralized distribution of energy– particularly for meeting rural energy needs, and thereby empowering people at the grass roots level. Solar electricity could also shift about 90 percent of daily trip mileage from petroleum to electricity by encouraging increased use of plug-in hybrid cars. For drivers in India this means that the cost per km could be reduced by a quarter in today's prices. RURAL SCENARIO IN INDIA Majority of India’s population lives in villages. The population in these communities has grown substantially in past fifty years. This growth has put tremendous pressure on resources needed to support this sector. In addition, it has resulted in enormous pollution of air, water and land. There is also substantial growth of population in India’s cities resulting in similar problems that are encountered by village and rural folks. Clearly, sustainable developments innovations and clean energy technologies are needed to accommodate the growing population in India for healthier environment. The presence of solar energy at any location in a village makes its uses attractive for such an environment. Most prominent of these are the drying of foods, vegetables and fruits for preservation and assured availability during off season periods of the year. The examples of technologies that are in use in rural India include mobile phones, home appliances, farm implements, vehicles for human, animal, and crop transport. The burning of wood and other materials for multiple uses continues. The cost and reliability for the supply of these energy sources has been a source of concern throughout India. The environmental impact from their use has also surfaced as a major concern. For this reason, solar technologies for rural applications have to be rugged, reliable and easy to use and repair. The ease of maintenance is essential for successful technology intervention. Furthermore, costs associated with repair and maintenance should be minimal. SOLAR ENERGY POTENTIAL AND TECHNOLOGIES Solar is the prime free source of inexhaustible energy available to all. And, India is one of the sun's most favoured nations, blessed with about 5,000 TWh of solar insulation every year. Even if a tenth of this potential is utilized, it could mark the end of India's power problems–by using the country's deserts and farm land to construct solar plants. Renewable energy has the potential to re-energize India's economy by creating millions of new jobs, allowing the country to achieve energy independence, reduce its trade deficits and propel it forward as a “Green Nation”. In short, renewable energy offers too many benefits for India to ignore, or delay its development. India should take full advantage of this golden opportunity because renewable energy has particular relevance in remote and rural areas, where there are around 289 million people who don't have access to reliable sources of energy. Solar energy is the most cost-effective option for India to reduce energy poverty without having to extend national grid services to provide power for individual homes and buildings.
ICORE 2013
India's present generation capacity is about 226,000 MW. The country could potentially increase grid-connected solar power generation capacity to over 226,000 MW and wind energy to over 100,000 MW by 2030 if the right resources (and more importantly, energy policies) were developed. India can develop massive commercial wind farms to harness the strong onshore costal area and offshore wind to boost the country's supply of clean renewable energy. But, to tap this vast resource, India must develop and implement smart business models and favourable policies as quickly as possible. Solar Powered Battery Well The concept of Solar Powered Battery Well is similar to the village Community water well. In this case, the solar energy is harvested and stored and inhabitants can charge devices or energy storage systems from energy stored in batteries for a fee. The installation and maintenance of this system can be done through a village cooperative arrangement so that it can be made accessible to inhabitants at a lower cost per watt. Solar Powered Food Preservation Facility Farm products degrade in normal climatic conditions. Low temperatures and humidity allows these products to extend their usable life. Storages with compartmental design can be built in rural areas and their environment can be regulated with solar energy technology. These storages can be built through village cooperatives and then rented to farmers for their use. Food preservation will greatly help to satisfy India’s growing demands. Solar Powered Water Wells In a typical village in rural India, the villagers have community water wells drawing water from underground. They also have water wells for irrigating fields. Most of these are operated manually and those that operate on electric power are limited by the availability of such power. The power availability in rural India for use by farmers is extremely limited. Solar powered water wells will undoubtedly increase crop yield and better the quality of life of the inhabitants. Solar lighted Streets/Corners/Farm ways Solar lights are very commonly used in many cities to provide safety and needed light to enable an individual to accomplish jobs. Now they need to be used in rural areas. Solar Powered Rural Homes/Huts The technologies to be used in home include lights, phones, refrigerators, food cookers, heaters, television and other electronic devices, air/water coolers, water well and fans. Solar Powered (Mobile) Tools/Implements Tools are essential for farmers. From digging to harvesting the crops, several implements/tools are needed. Many of these are manually operated presently to save energy use. Solar powered farming will realize great leap forward in productivity and quality of life of farmers. Solar Powered Vehicles For fast transportation of materials, crops, animals, and humans, rural areas now depend on vehicles that run on gas/patrol. These vehicles could be designed to run on solar power. Solar Powered Waste Management Farms and farm homes produce waste which largely pollutes land, water, and air thereby becoming a health hazard, in addition to being an eye sore. Solar powered compressors or waste conversion to useful products would benefit immensely the rural communities.
2 ♦ Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations
ICORE 2013
RENEWABLE ENERGY POLICIES IN INDIA India is one of the world leaders for installed renewable generation with a total capacity of almost 26 GW as of June 2013. The key drivers for renewable energy in India include: • • • •
Good levels of resource availability. The forecast growth in energy demand. Energy security concerns in light of increasing imports of fossil energy. Economic and quality of life costs associated with the environmental impacts of fossil fuel combustion. The Government of India has long recognised the potential business opportunities offered by growing demand for renewable energy in local and overseas markets. It established the world’s first ministry focusing on renewable energy, the Ministry for New and Renewable Energy, (MNRE). The Indian Renewable Energy Development Agency (IREDA) is administered by MNRE and was established in 1987 to operate a revolving fund for development and deployment of new and renewable sources of energy. MNRE also administers national institutions such as the Solar Energy Centre (SEC). India’s existing capability and potential for innovation is supported by a well educated, professional and skilled workforce. There is a thriving renewable energy sector with strong growth in biomass generation, wind turbine and photovoltaic manufacturing that offers opportunities for synergy with the CSP sector. MAJOR STEPS TO HARNESS RENEWABLE ENERGY Aggressively expand large-scale deployment of both centralized and distributed renewable energy including solar, wind, hydro, biomass, and geothermal to ease the strain on the present transmission and distribution system–and allow more off-grid populations to be reached. Facilitate growth in large scale deployment by installing 100 million solar roofs and large utility-scale solar generation, through both centralized and distributed energy within the next 20 years. Enact a National Renewable Energy Standard/Policy of 20 percent by 2020–to create demand, new industries and innovation, and a new wave of green jobs. Develop favourable Government policies to ease the project permitting process, and to provide startup capital to promote the exponential growth of renewable energy. Create and fund a national smart infrastructure bank for renewable energy. Accelerate local demand for renewable energy by providing preferential Feed-in-Tariffs (FIT) and other incentives such as accelerated depreciation; tax holidays; renewable energy funds; initiatives for international partnerships/collaboration incentives for new technologies; human resources development; zero import duty on capital equipment and raw materials; excise duty exemption; and low interest rate loans. Establish R&D facilities within academia, research institutions, industry, Government and civil society to guide technology development. Accelerate the development and implementation of Solar and Wind farms; utility-scale solar and wind generation nationwide. Initiate a move to electrify automotive transportation or develop electric vehicles and/or plug-in hybrids–such as the Nissan Leaf or Chevy Volt, etc. Develop and implement time-of-day pricing to encourage charging of electric vehicles at night. Adopt nationwide charging of electric cars from solar panels on roofs, and solar-powered electric vehicle charging stations around the country. Thousands of these solar-powered recharging stations could spread across India just like the present public call office (PCO), giving birth to the “Green Revolution.” These recharging connections could be deployed at highlyconcentrated areas including shopping malls, motels, restaurants, and public places where cars are typically parked for long periods. Aggressively invest in a smart, two-way grid (and micro-grid). Invest in smart meters, as well as reliable networks that can accommodate the two-way flow of electrons. Such networks need to be resilient enough to avoid blackouts and accommodate the advanced power generation technologies of the future. Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations ♦ 3
ICORE 2013
Develop large scale solar manufacturing in India (transforming India into a global solar manufacturing hub). Work towards a Hydrogen Economy development plan. Hydrogen can be fed into fuel cells for generating heat and electricity–as well as for powering fuel cell vehicles. Produce hydrogen using renewable energy with solar and wind power. If done successfully, hydrogen and electricity will eventually become society's primary energy carriers of the twenty-first century. SOLAR ENERGY RESOURCES FOR DESIGNING CSP IN INDIA CSP technologies rely on Direct Beam radiation for operation. That is radiation direct from the sun that has not been diffused or deflected by clouds or other atmospheric factors and so can be focussed by the mirrors. Ideally the data needed to assess potential sites is short interval, Direct Normal Irradiation (DNI) measurements collected over several years. Issues include: • • •
In designing a CSP plant, knowledge of the seasonal variation in DNI resource is needed to make an optimal economic assessment of the degree to which the solar field is oversized relative to the power block in the high season and undersized in the low season. Power blocks have an efficiency that decreases with part load (due to reduced solar input) and a threshold load under which they cannot operate. Thus to predict the power block’s total daily output, knowledge of the time dependence of input heat transfer fluid energy flow is required. The thermal receivers in collector fields take many minutes to reach operating temperature from cold, hence to accurately predict their output, data at time resolutions of one minute or less is required to predict performance in situations of intermittent cloud.
EDUCATION AND TRAINING ON RENEWABLE ENERGY: DELHI TECHNOLOGICAL UNIVERSITY INITIATIVES Objectives of the Centre of Excellence Each technical university must establish a centre of excellence on renewable energy technology with following objectives: • • • • •
To provide education on Renewable Energy Technology; To engage in Research & Development in Solar Photovoltaic and Concentrated Thermal Energy Technologies and their various applications for power generation, air conditioning, process heat industries. To promote Renewable Energy Innovation Park for general public, RWA, Women, children, NGOs etc. To carry out training programme(s) in the field of Renewable Energy Technology for students of other Engineering/Polytechnic/ ITIs/Arts, Science and Commerce colleges. To Collaborate with Industries and national and international agencies related to promotion of education and research.
Major Activities of the Proposed Centre Higher Education: • • • •
B. Tech. (Energy Technology) M. Tech. (Renewable Energy Technology PhD Programme with specialization in Solar thermal and photovoltaic energy technology. PGDM in relevant area of Renewable Energy
Research & Development in the Proposed Centre Solar Photovoltaic • • •
Use of solar photovoltaic in the development of solar rickshaw, passenger car Fundamental research in PV-related materials and cost reduction; Development of PV cells from several material systems;
4 ♦ Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations
ICORE 2013
• • • •
Characterization of PV cells, modules, and systems to improve performance and reliability; Development of standardized tests and performance models for PV devices; Developing advanced concentrating photovoltaic technologies Helping the PV industry to accelerate manufacturing capacity and commercialization of various PV technologies.
Solar Flat/ Concentrated Thermal Energy System • • • • • • • •
Cross Linear Concentrated Solar Power technology for solar process heat, refrigeration and electrical power generation. Solar process heat and space cooling. Concentrating solar systems for steam generation for cooking. Development of new designs and manufacturing processes for solar components and systems with an emphasis on improved performance, reliability and service life. Prototype development of systems through modeling and optimization. Characterization of the system’s performance, and accelerated materials durability testing. Cost and performance of solar air heating thermal systems. Research on lowering the cost of solar water heating systems.
Education and Training Programmes • • • •
The Centre will conduct specialized training programmesfor: Participants from Industries, RET users, Students and faculties from Polytechnics/ Engineering colleges/ Universities NGO/RWA/central/state govt. employees etc.
Electric Vehicle Modeling, Wireless Charging Systems • • • • • • •
Use of solar photovoltaic in the development of electric rickshaw, passenger car Fundamental research in PV-related materials wireless battery charging and cost reduction; Development of PV cells from several material systems; Characterization of PV cells, modules, and systems to improve performance and reliability; Development of standardized tests and performance models for PV devices; Developing advanced concentrating photovoltaic technologies Helping the PV industry to accelerate manufacturing electric vehicles.
Bio Energy Technologies • • • • • • •
Carbon capture and sequestration. Bio Diesel production using MERS Technology Biomass densified briquettes and pallets Gasifier and related Technologies CH4 Production from Sanitary Land fill Biogas Generation from Kitchen, vegetables and Animal waste All kind of Vermi Composting productions
Renewable Energy Innovation Park • • • • •
This will be the front face of the proposed Centre and showcase the latest trends & technologies in Renewable energy area of relevance to the world. Solar Water Heating Applications Solar Air Heating Systems Applications. Large Scale grain Dryers. Residential and Office Space Heating. Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations ♦ 5
ICORE 2013
• • • • • • •
Fruit and Vegetable Drying for off seasons Applications. Solar Steam Cooking in DTU’s Boys and Girls Hostels. Solar Cooling and Refrigeration for Library Buildings 1 Megawatt rooftop solar power generation in DTU’s campus in 4 phases. LED street lighting systems. Information and Communications Technologies powered by SPV plant. Solar operated cars and vehicles and solar energy charged battery Operated vehicles including solar rickshaw having charging stations at DTU.
Knowledge Partners of the proposed Centre of Excellence The Centre shall work in close cooperation with MNRE, & Solar Energy Center (SEC) and shall collaborate with world class Universities and Institutions in India & abroad. Special features of the collaborative program shall be focus on development of new and efficient Solar technologies & collaborative research in new horizons of Renewable Energy. We should collaborate among major IIT’s, DTU, NIT’s and other Institutes of higher learnings and generate funds from AICTE, and other bodies such as Deptt of Environment & Forest of GNCT Delhi, Ministry of Environment & Forest (MoEF), GOI & Ministry of New and Renewable Energy (MNRE), GOI, UGC,DST, and International organizations such as UNESCO, UNDP, IDRC, SDA, DANIDA, GTZ, IDRC, SIDBI, etc Indian Knowledge Partner • • • • • • • •
DTU and other Universities & R&D Institutions Solar energy center, MNRE Bergen Group Sunborne Systems Thermax Rajiv Gandhi Technological University, Bhopal National Thermal Power Corporation Maharishi Solar Technologies Pvt Ltd
International Knowledge Partners • • • • •
Tokyo Institute Of Technology, Japan Asia Pacific Sunbelt Development Association, Japan Toyo Engineering, Tokyo, Japan, Richo, Japan University of South Florida
CSP PROGRESS IN INDIA Details of the major Solar Thermal Projects recently undertaken at Solar Energy Centre (SEC), MNRE: National Solar Thermal Power Testing, Research and Simulation Facility: The facility envisages a grid connected Solar Thermal Power Plant of 1 MW capacity. This will also include a test set up that enables companies and research institutions to test the performance of different solar concentrators, coatings and materials, components and system for a Solar Thermal Power Plant. The project is being implemented by IIT Mumbai and a consortium partners consisting of Tata Power, Tata Consulting Engineers, Larsen & Toubro, Clique, KIE Solatherm. Concentrated Solar Thermal Energy Technology based on Parabolic Dish collectors The project is a cooperative effort between Megawatt Solutions Pvt. Ltd. and SEC under MNRE’s initiative to promote research, development and demonstration of indigenously developed renewable energy systems and technologies under a cost-sharing basis. It involves demonstration and evaluation of 4 interconnected dish concentrators each of 90m2 aperture area providing heated thermic fluids upto 4000C. 6 ♦ Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations
ICORE 2013
Solar Thermal Stirling Engine The project has been taken up in collaboration with ONGC Energy Centre (OEC). Three units of engines of 3 kW capacities each have been installed and commissioned in the campus. The objective of the project is to carryout long term performance evaluation under Indian conditions. The engines have been connected to the local grid and the electricity produced during the sunshine hours is being utilized in the Technical Block. The rated output of the facility is 9 kW (peak power) at solar insulation of 850 W/m2 at 200C ambient temperature. Development of a Modular Central Receiver Concentrated Solar Power Plant for Decentralized Power Generation Sun Borne Energy with support from MNRE will be setting up a 1 MWTh CSP Central Tower pilot facility in partnership with SEC. The main objective is to develop the optimized designs of the heliostat field, volumetric air receiver and thermal storage, the three major components of a Concentrated Solar Power (CSP) Central Receiver plant and also to develop the local sources for all the key components of the plant with a focus on lowering costs. High Efficiency Solar Thermal Air-conditioning Systems–a collaborative project of Thermax Limited and Solar Energy Centre The project (100 KW cooling capacity) is being implemented by M/s Thermax Limited at Solar Energy Centre with an objective to integrate solar collectors, vapour absorption machine (VAM) and appropriate thermal storage system to achieve consistence performance of the system with coefficient of performance (COP) (1: 1.7) Cold Storage with Solar–Biomass Hybrid System It is an APP project in partnership with TERI, Thermax Limited, SEC and CSIRO Australia with an objective to develop cold storage particularly in rural areas utilizing exhaust heat of biomass gasifier engine/solar scheffler dish. MITIGATION OF CLIMATE CHANGE AND LOW CARBON INITIATIVES Impacts of Climate Change are already being faced by each one of us. However, every individual needs to take measures in cutting carbon emissions through awareness campaigns that help in bringing out a change in attitude and consumption patterns. Sustainable use of natural resources is demonstrated in the traditional practices. It is a very important to preserve the traditional knowledge and practices for providing effective sustainable life-solutions. Some of the initiatives at Institutional and Individual level are as follows: use of solar energy for cooking, lighting, air conditioning, power generation etc, replacing home appliances with the energy efficient one, switching lighting systems to CFLs/LEDs, switch off the lights and other electrical gadgets when not in use, temperature of AC’s to be maintained above 24 degree C, use warm clothes instead of room heaters in the winter Low Carbon Technologies and Promotion of Renewable Energy Sources are two important thrust areas identified by Govt. of India as part of energy security missions. We are living in a world of rapidly depleting fossil fuel resources with access to conventional energy resources becoming constrained. Jawaharlal Nehru National solar Mission is one of the eight national missions which comprise India’s National Action Plan on Climate Change. We have to build India’s core competence in Renewable Energy Sector and we must get ready to meet the challenges in 21st century. Our dream is that India should become a country which can generate its own power through green sources, where everyone has access to clean and green energy and India should become a torch bearer of eco-friendly habitats, which aim at complete carbon neutrality.
Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations ♦ 7
ICORE 2013
CONCLUDING REMARKS India has emerged as a global power in the past decade. With an animal GDP growth rate of 8.6%, it is the world’s fourth largest country in terms of purchasing power, the second most populous country in the world and a leading player in IT, telecom and business outsourcing. The government of India recognizes that in order to sustain this growth rate and strengthen its national competitiveness, solar energy innovation has a critical role to play in India. The Government of India’s Solar Mission is a visionary and inspiring policy measure that has the potential to be a leading example for the world. Harnessing the potential of solar energy for the benefit of rural areas of India will enlighten the hearts and minds of the rural population. Research on all aspects of solar energy including science, technology, engineering, economics, and management will engage diverse body of faculty and students in a team environment. This can energize the entire University for a worthy cause. India’s population lives in rural areas and the quality of life will positively be affected by innovative use of solar energy. REFERENCES [1] KESARI J.P. (2010) Combating Climate Change Through Renewable Energy Technologies: Solar Drying Experiences, International Conference on global warming, climate change, sustainable development, secular spirituality, Santhigiri Ashram, Thiruvananthapuram 09-11 September 2010. [2] Kesari J.P. (2011) Solar Thermal Power Generation in India: An overview of recent policies under National Solar Mission, Proceedings of the National Conference on Management of Technologies for Advancing Rural India, feb. 5-6, 2011. at SDCOE at Muzaffarnagar. [3] Kesari J.P., P.B. Sharma, JASPREET SINGH, SUSHANT KUMAR SINGH (2011) Development of Passenger Solar Electric Vehicle, paper Published in International Solar Energy Society Congress proceedings 2011 in Kassel, Germany. [4] Kesari J.P. (2011) Empowering India with Green Energy Technology, paper presented in 53rd National Convention of Indian Institution of Industrial Engineering and National conference on India as a Technology Hub, Dec. 16-17, 2011. [5] Kesari J.P., Jaspreet Singh, Sushant Kr. Singh, Rohit Gupta, P B Sharma (2011) Development of Solar Electric Hybrid Vehicle (Ref. 2012-28-0025), Paper published under SAE International and presented at SAE India International Mobility Conference 2012. [6] Kesari J.P., P.B Sharma (2012) Concentrated Solar Power: Technology Roadmap at International Symposium on Concentrated Solar Power: Opportunities for India 23th-27th Jan, 2012 at Delhi Technological University. [7] Kesari J.P., P.B. Sharma(2012) Empowering India with Solar Energy Technology: A Roadmap at International Conference on Climate Change and Sustainable management of natural Resources Feb 5th to 7th, 2012 at ITM University Gwalior. [8] Kesari J.P., Kumar Krishen and P B SHARMA (2012) Solar Power is a Panacea for Changing Rural India; The role of CSP under Indo-Japan Collaboration, Proceedings of 2nd International Symposium–collaborative research project of CSP for India and Japan, 25 April 12, Tokyo Institute of Technology, Tokyo, Japan. [9] KESARI J.P., (2012) Eco-spirituality for sustainable living–Role of Solar Energy Technology‘, SSTech2012 National Conference at Mumbai University Oct 5th and 6th 2012. [10] Kesari J.P., P.B. Sharma (2012) Developing Solar Energy Projects: A Doorstep Opportunies for India. Paper presented at Clean Energy and energy conservation 2012, organised by The Institution of Engineers(INDIA) 13-14 Oct,2012 Nagpur. [11] Kesari J.P. and P.B. Sharma (2013) Changing the face of India with Solar Power Technologies: Generation of Energy, Employment and Entrepreneurship. Paper presented at National Seminar on IT Applications in Energy Management, Rajiv Gandhi Institute of Information Technology, Amethi, A campus of IIIT Allahabad, April 15, 2013. [12] Kesari J.P. and P.B. Sharma (2013) Changing the Face of India with Solar Power Technologies: Generation of Energy, Employment and Entrepreneurship, Paper published in GRIDTECH 2013, April 3-5, Pragati Maidan, New Delhi. [13] Kesari J.P. and P.B. Sharma (2013) Changing the face of India with solar power Technology, keynote address in International Conference on Role of technology in Nation Building, 27 April 2013, Subharti university, Meerut, UP.
8 ♦ Transforming India into a Green Nation: Bottom up Approach through Solar Energy Education, Training and Innovations
Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System T. Ramana1, V. Ganesh2, S. Sivanagaraju3 and K. Vasu4 1Research Scholar, JNT University, Anantapuramu, A.P Professor & HOD, Dept. of EEE, JNTUA College of Engineering, Pulivendula, Kadapa (DT), A.P 3Professor & HOD, Dept. of EEE, University College of Engineering, Kakinada, JNTUK, E.G (DT), A.P 4Assistant Professor, Dept. of EEE, MITS, Madanapalle, Chittoor(DT), A.P
2Associate
Abstract—In this present modern world, one cannot assume life without electric power, as everyone is habituated to use electrical power for all the applications. As the population is increasing day by day, the electrical power demand is also increasing. The cost of renewable energy technologies is on a falling trend as demand and production increases. There are many renewable energy sources such as solar, biomass, wind and tidal power etc. In this work an efficient Integrated Wind/ Solar Hybrid Electric Power System is developed. It provides theoretical studies of Photo Voltaic (PV) and its modeling and also investigates in detail the Maximum Power Point Tracker (MPPT), Power electronic devices that significantly increase the system efficiency. In this work it is focusing on modeling and simulation of Photo Voltaic (PV)/ Wind Turbine Generator (WTG) Hybrid Electric Power System (HEPS) interconnected to the electrical utility through grid taking into account of all irradiation data, temperature, wind speeds and variation of the load demand during the day. A computer simulation program is developed to simulate all quantities of HEPS such as phase voltage of the inverter circuit and current for PV and WTG. The circuit also going to simulate AC output current of the inverter that injected to the load/grid, load current, grid current, power output from PV and WTG, power delivered to or from grid. This proposed computer simulation interconnected system is tested with laboratory model of interconnected circuit and compared with existing methods. Keywords: Wind Turbine Generator, Back to Back Converter, LC Filter, Electrical Utility
NOMENCLATURE WTG: Wind turbine generator. WES: Wind Energy System. EU: Electric Utility. IRPT: Instantaneous Reactive Power Theory. PWES (t): The output power of WTG's. Pg(t): The power of EU. PL(t): The hourly load demand. t: The hourly time over one year. INTRODUCTION The rapid depletion of fossil fuel resources on a world-wide basis has necessitated an urgent search for alternative energy sources to meet to the present day demands. Alternative energy resources, such as solar and wind energies, are clean, inexhaustible and environment friendly potential resources of renewable energy options. It is prudent that neither a standalone solar nor a wind energy system can provide a continuous supply of energy due to seasonal and periodical variations [1-3]. Some projections indicate that the global energy demand will almost triple by 2050. Renewable energy sources currently supply somewhere between 15% and 20% of total world energy demand. PV and Wind Energy System are the most promising as a future energy technology. A 30% contribution to world energy supply from renewable energy sources by year 2020. Pedro Rosas [4] presented the basic influences of wind power on the power system stability and power quality. It has introduced also an aggregate wind farm model that support power quality and stability analysis from large wind farm. Koch F, Erlich I. and Shewarega F. [5] presented simulation results using a representative network containing wind power generations of up to 30%. Furthermore, modeling and simulation of different types of wind generators integrated into a multimachine power system discussed. Koch F., et. al. [6] described the effect of large wind parks on the frequency of the interconnected system on which they are operating. Additionally, the effect of the landscape and atmospheric condition at the location of the wind unit on the output power incorporated into the simulation. With increased penetration of WES various researches for modeling of WTG connected to the Electric Utility(EU) proposed as in [7]-[11].
ICORE 2013
THE PROPOSED SYSTEM MODEL The system model shown in Fig. 1 represents Wind Energy conversion system connected to a 50 Hz, 22 kV EU.WTG connected to EU through back to back converter, LC filter and step-up transformer [13]. The load connected to 22kV Bus through a step-down transformer. The variables which will
FIG. 1: POWER AND CONTROL CIRCUIT OF WECS INTERCONNECTED WITH EU TO FEED THE LOAD
be sensed for the controller of WTG are DC link voltage, Vdcw, inverter filter output currents Iinvaw, Iinvbw, Iinvcw, load currents ILa, ILb, ILcand load phase voltages Va, Vb, Vc. To provide the active filtering function, the filter output currents are controlled to ensure that the utility line currents and load current are sinusoidal and in phase with the phase voltage. The filter output currents are also controlled to pass power from the WES to the load and/or EU. The DC link voltage, Vdcpv and Vdcpw must be controlled to be higher than the peak line voltage of the EU. The proposed system control scheme for the system under study usually uses the Instantaneous Reactive Power Theory, IRPT. The load currents and load voltages are sampled and transformed into the two-axis -coordinate system and then into the rotating dq-coordinate system. IRPT uses the park transformation, as in (1) to generate two orthogonal rotating vectors α and β from the three-phase vectors a, b and c. This transformation is applied to the voltages and currents and so the symbol x is used to represent volt or current. IRPT assumes balanced three-phase loads and does not use the x0 term [13], [14], [15].The instantaneous active and reactive powers p and q are calculated from the transformed voltage and current. Then the reference compensating currents have been determined as in(2).In a balanced three-phase system with linear loads, the instantaneous real power p and imaginary power q are constant and equal to the three-phase conventional active power P3 and reactive power Q3Φ respectively. So, the inverse park transformation is applied to iα* and iβ* and this gives the output currents in standard three-phase form, as in (3) [13], [14], [15]. ⎡ x0 ⎤ ⎢ ⎥ ⎢ xα ⎥ = ⎢ xβ ⎥ ⎣ ⎦
⎡ ⎢ ⎢ 2⎢ ⎢ 3⎢ ⎢ ⎢ ⎣⎢
1
1
1
2
2 1
2 1
1 0
⎡i* ⎤ 1 ⎢α ⎥= * ⎢iβ ⎥ v 2 + v2 ⎣ ⎦ α β
−
2 3 2
−
2 − 3 2
⎤ ⎥ ⎥ ⎡x ⎤ ⎥ ⎢ a⎥ ⎥ ⎢ xb ⎥ ⎥ ⎢x ⎥ ⎥⎣ c⎦ ⎥ ⎦⎥
⎡vα − vβ ⎤ ⎡ pw ⎤ ⎥ ⎢ v vα ⎥ ⎢⎣ qw ⎥⎦ ⎦ ⎣⎢ β
10 ♦ Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System
(1)
(2)
ICORE 2013
⎡ ⎢ 1 ⎡ x0 ⎤ ⎢ 2 ⎢ ⎥ ⎢ 1 ⎢ xα ⎥ = 3 ⎢− 2 ⎢ xβ ⎥ ⎢ ⎣ ⎦ ⎢− 1 ⎣⎢ 2
0 3 2 3 2
⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦⎥
⎡i* ⎤ ⎢α⎥ ⎢i*β ⎥ ⎣ ⎦
(3)
There are two modes of operation: •
Mode 1: When the generated power from WEPS is lower than the load demand then the deficit power will be supplied from the EU. Presumably, the power factor will be within the allowed limits. • Mode 2: When the generated power from Wind greater than the load demand then the surplus power will be transmitted to the EU. To solve this problem, the coupling converter should also supply or absorb active power and reactive power simultaneously. Then a variable reactive power reference q*should be included in the inverter control. Thus, the AC source can operate at the allowed or unity power factor. That is, if q* is made equal to qL, the source power factor can be kept equal to unity under different load conditions. Then the proposed control strategy is supposed to be capable of generating any output imaginary power, that is, the source power factor may be set at any desired value. In case of choosing a particular value for the source power factor, the imaginary power reference q* should no longer be set to q but to the following value [15]:
(
q* = qL− PL− P*
) tan (
)*
(4)
Where; * is the source desired reference displacement angle,cos ( )* is the reference power factor. WTG INTERCONNECTED WITH EU Figure 1 shows an overview of the power and control circuit of the proposed WTG interconnected with EU. The power circuit of Fig. 1 has been simulated using Matlab/Simulink as shown in Fig. 2.
FIG. 2: OVERVIEW OF THE POWER AND CONTROL CIRCUIT OF THE PROPOSED WTG INTERCONNECTED WITH EU
Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System ♦ 11
ICORE 2013 Wind Inverter Voltage 4000 3000 2000
Voltage(v)
1000 0 -1000 -2000 -3000 -4000
0
0.05
0.1
0.15
0.2
0.25 Time(s)
0.3
0.35
0.4
0.45
0.5
FIG. 3: SIMULATED OF THE WTG VOLTAGE Wind Inverter Current 30
20
Current(A)
10
0
-10
-20
-30
0
0.05
0.1
0.15
0.2
0.25 Time(s)
0.3
0.35
0.4
0.45
0.5
0.4
0.45
0.5
FIG. 4: SIMULATED OF THE WTG CURRENT GRIDE VOLTAGE 4000 3000
VOLTAGE(V)
2000 1000 0 -1000 -2000 -3000 -4000
0
0.05
0.1
0.15
0.2
0.25 TIME(S)
0.3
0.35
FIG. 5: SIMULATED OF THE GRID VOLTAGE
12 ♦ Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System
ICORE 2013 Load Current Vs Time 300 200
Load Current(A)
100 0 -100 -200 -300 -400
0
0.05
0.1
0.15
0.2
0.25 Time (sec)
0.3
0.35
0.4
0.45
0.5
0.4
0.45
0.5
0.4
0.45
0.5
FIG. 6: SIMULATED OF THE LOAD CURRENT LOAD VOLTAGES 4000 3000
Load Voltage(v)
2000 1000 0 -1000 -2000 -3000 -4000
0
0.05
0.1
0.15
0.2
0.25 Time(s)
0.3
0.35
FIG. 7: SIMULATED OF THE LOAD VOLTAGE GRID CURRENT 300 200
CURRENT(A)
100 0 -100 -200 -300 -400
0
0.05
0.1
0.15
0.2
0.25 TIME(S)
0.3
0.35
FIG. 8: SIMULATED OF THE GRID CURRENT
SIMULATION RESULTS The variation of the generated power from hybrid PV/WTG according to radiation and wind speed variation is observed in this work. The proposed model has a purely sinusoidal controlled ideal voltage Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System ♦ 13
ICORE 2013
source at the inverter terminals. Due to the small width of the hysteresis band the voltage generated by the proposed model is nearly sinusoidal when seen at this bus. Figure 4 shows the current injected by the WTG with total harmonic distortion 1.4 %. The load line current of the load demand is shown in Fig. 6. Fig. 8 shows the simulated of grid line current with total harmonic distortion of 2.33% that injected to or drawn from grid and Fig. 4 displays the simulated voltage of WTG. Also, from these Figures 5 and 7 it can be seen that the simulated grid voltage and load voltage of the system. CONCLUSION From the results obtained above, it can be observed that: • • • •
A novel of WES interface with EU for solving modeling and simulation problems by using Matlab/Simulink environment have been proposed. A new computer program has been proposed for modeling and simulation of any WTG interconnected to EU. The total harmonic distortion at the local bus is within acceptable limits and reached to 1.4 % for the inverter current from WTG, and 2.33% for the grid current. Perform the necessary preliminary studies before investing and connecting WEPS to the grid where purchased and sold power from EU have been calculated.
REFERENCES [1] B. S. Borowy and Z. M. Salameh, “Optimum Photo-voltaic Array Size for a Hybrid Wind/PV System,” IEEE Transactions on Energy Conversion, Vol. 9, No. 3, 1994, pp. 482-488. [2] B. S. Borowy and Z. M. Salameh, “Methodology for Optimally Sizing the Combination of a Battery Bank and PV Array in a Wind/PV Hybrid System,” IEEE Transac-tions on Energy Conversion, Vol. 11, No. 2, 1996, pp. 367-374. [3] A. N. Celik, “Optimisation and Techno-Economic Analy-sis of Autonomous Photovoltaic-Wind hybrid Energy Systems in Comparison to Single Photovoltaic and Wind Systems,” Energy Conversion and Management, Vol. 43, No. 18, 2002, pp. 2453-2468. [4] Pedro Rosas, “Dynamic Influences of Wind Power on the Power System”, Ph.D. Thesis, Section of Electrical power Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark, 2003. [5] Koch F., Erlich I. and Shewarega F., 'Dynamic Simulation of Large Wind Farms Integrated in a Multi Machine Network', IEEE PES General Meeting, Toronto, Ontario, Canada, July 13-17, 2003. [6] Koch F., et. al.,'Simulation of the Dynamic Interaction of Large Offshore Wind Farms With the Electric Power System', Owemes, Naples, Italy, April 10-12, 2003. [7] A. S. Neris, N. A. Vovos and G. B. Giannakopoulos, 'A variable Speed Wind Energy Conversion Scheme for Connection to Weak AC Systems', IEEE Transactions on Energy Conversion, Vol. 14, No. 1, March 1999, pp. 122-127. [8] KimJohnsen, and Bo Eliasson, 'Simulink Implementation of Wind Farm Model for use in Power System Studies', Nordic Wind Power Conference, 1-2 March, 2004, Chalmers University of Technology. [9] Mariusz Malinowski and Steffen Bernet,' Simple Control Scheme of PWM Converter Connecting Wind Turbine with GridSimulation Study', Nordic Wind Power Conference, Chalmers University of Technology, March 2004, pp. 115-121. [10] JanPierik, Johan Morren and Sjoerd de Haan, 'Dynamic Models of Wind Farms for Grid-integration Studies Case study Results', Nordic Wind Power Conference, 1-2 March 2004, Chalmers University of Technology. [11] FlorinIov et al.,'Advanced Tools for Modeling, Design and Optimization of Wind Turbine Systems ', Nordic Wind Power Conference, 1-2 March, 2004, Chalmers University of Technology. [12] Debra J. Lew et. al., 'Hybrid Wind/Photovoltaic Systems for Households in Inner Mongolia', International Conference on Village Electrification through Renewable Energy, 3-5 march New Delhi. [13] H. H. El-Tamaly and Adel A. El-baset, 'Computer Modeling and Simulation of Wind Energy System Connected to Utility Grid' Proceedings on 2004 International Conference on Electrical, Electronic and Computer Engineering ICEEC’04, 5-7 September 2004, Cairo, Egypt. [14] H. H. El-Tamaly and Adel A. El-baset Mohammed, 'Computer Simulation of Photovoltaic Power System Interconnected With Utility Grid', Al-azhar Engineering Eighth International Conference, Vol. 8, No. 7, December 24-27, 2004, pp. 57-64 [15] P. G Barbosa. et al., “Novel Control Strategy for Grid-Connected DC/AC Converters with Load Power Factor and MPPT Control”, Congresso Brasileiro de Eletrônica de Potência, COBEP '95-III, Paulo 1995.
14 ♦ Modeling and Simulation of Grid Interconnected Electric Utility with Wind/ Photovoltaic Hybrid Electric Power System
Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator Gezahegn Habtamu1, S.C. Mullick2 and Sanjeev Jain3 1,3Department of Mechanical Engineering, Indian Institute of Technology, Hauz Khas, New Delhi–110016, India 2Centre for Energy Studies
Abstract—This paper investigates the optical performance of a single glazed, a double glazed and an unglazed liquid desiccant solar regenerator and also compares the thermal performance of glazed and unglazed solar regenerators. The optical performance is evaluated in terms of glass transmittance, solution transmittance, solution absorptance and optical efficiency. The mean solar transmittance of single and double glass were found to be 0.812 and 0.659, respectively. The average solar absorptance of 3 mm and 8 mm calcium chloride solution of concentration 0.4 were found to be 0.11 and 0.16, respectively. The optical efficiency of unglazed, single glazed and double glazed solar collector regenerators were found to be 94.9%, 77% and 62.5%, respectively. Regeneration experiments were conducted on 2.2 m x 1.84 m glazed and unglazed solar collector cum regenerators using calcium chloride solution to evaluate their thermal performance and compare with each other. Experiments carried out on glazed and unglazed solar collector cum regenerators are presented in terms of evaporated mass of water, energy of evaporation of water, increase in concentration of desiccant, solar collector/regenerator efficiency, figure of merit and concentration effectiveness. The mean irradiance, over the regeneration period, received by glazed and unglazed solar regenerators were 0.81 and 0.69 kW/m2. A total of 17 kg and 15.2 kg of water were evaporated from the calcium chloride solution which resulted increase in concentration from 0.31-0.47 and 0.3-0.45 for glazed and unglazed regenerators, respectively. The peak solar collector cum regenerator efficiency of glazed and unglazed regenerators were 51 and 53%, respectively. The experimental results showed that the thermal performance of glazed regenerate is slightly better than the unglazed one. Keywords: Solar Regenerator, Optical Efficiency, Liquid Desiccant, Thermal Performance
INTRODUCTION Direct solar regeneration of liquid desiccants using a collector cum regenerator may be a good alternative over indirect regeneration using hot water. The solar collector cum regenerator must convert the solar radiation into heat to regenerate the liquid desiccant flowing over the absorber surface since regeneration of a liquid desiccant is predominantly a thermal process. During the conversion process, the solar energy is attenuated by optical resistance of the constituent materials of the collector like glass cover, dust on the glass & absorber surface, shading due to glass supporting frame and the liquid desiccant flowing over the absorber. In addition to these, some fraction of the solar irradiation is reflected from the solution-absorber surface before it is absorbed by the black painted absorber. The net thermal energy absorbed is then used for evaporating water out of the liquid desiccant solution that flows over the absorber while encountering surface heat losses and sensibly heating the ventilation air & the solution. Experimental and theoretical research work on solar regeneration of liquid desiccants using solar collector cum regenerator were reported by several researchers (Mullick & Gupta [1]; Peng & Howell [2]; Johannsen & Grossman [3]; Gandhidasan [4, 5]; Kaushik et.al. [6, 7]; Kaudinya and Kaushik [8]; Pradip & Devotta [9]; Hawlader et al. [10]; Yang & Wang [11, 12, 13]; Fagbenle & Karayiannis [14]; Alizadeh & Saman [15, 16]; Kabeel [17]; Elsarrag [18]; and Yutong & Yang [19]). Most of these work focused on estimation of the thermal performance in terms of rate of water evaporation from liquid desiccants and regeneration efficiency. Only few researchers used the increase in concentration of desiccant in the liquid desiccant as performance indicating parameter such as Kaushik et.al. [6] and Kabeel [17]. Thermal performance of solar collector cum regenerator was also evaluated theoretically when it is integrated with open cycle absorption cooling systems by Collier, [20]; Bolzan & Lazzarin [21]; Mattarolo [22]; Wang & Yang [23]; Hawlader [10] and Haim et al. [24]. In addition, Gandhidasan [25] theoretically evaluated the cooling performance of a liquid desiccant system integrated with a solar collector cum regenerator of size 1 m x 1 m, using calcium chloride as desiccant. Katejanekarn & Kumar [26] and Katejanekarn et al. [27] evaluated the cooling performance of a lithium chloride cooling system theoretically and experimentally using four solar collector cum regenerators of size 1.72 m x 0.85 m. There are various designs of solar collector/regenerator used by researchers to convert solar energy into heat and control heat loss from absorber surface. For instance, some researches lined the absorber surface with black cloth to ensure complete wetting (Yang & Wang [11]; Kabeel [17]; and Yutong & Wang [19]). Some used unglazed solar regenerator (Collier [20]; Kabeel [17]), some single glazed
ICORE 2013
regenerator (Mullick & Gupta [1]; Yang & Wang [11, 12]) and others double glazed regenerator (Yang & Wang, [13, 23]) to produce concentrated desiccant using the solar collector/regenerator. Yang & Wang [13, 23] reported that a double gazed forced convection C/R provided better performance than single glazed solar regenerator. However, double glazing the solar regenerator increases the cost of the solar regenerator, makes it heavy and creates difficulty in cleaning the absorber surface regularly in addition to reducing the optical efficiency or the net solar energy that could be used for regeneration. But, glazing of the solar regenerator has several advantages such as preventing contamination of the liquid desiccant by dust, leaves and insects and also helps in reducing the top heat loss from the absorber surface by radiation and convection. All these conditions of operating the solar regenerator affects its optical as well as thermal performance. Thus, the advantage of glazing (either single or double) over the optical energy losses should be weighed and compared with photo-thermo performance of unglazed solar regenerator. In addition to these, the liquid desiccant too affects the thermal as well as the optical performance because the liquid desiccant also absorbs solar energy directly and reflect some fraction. In previous research work, the optical efficiency of a solar collector/regenerator was usually assumed by several researchers (Gandhidasan [5]; Kaushik et.al. [6]; Fagbenle & Karayiannis [14], and Alizadeh & Saman [16]) and is give negligible attention. Only Hawlader et al. [10] measured the solar absorptance of an asphalt shingle absorber over which a lithium chloride solution is flowing by measuring the reflected light from the absorber-solution surface using Gier-Dunkle spectrophotometer. The above are some of the foremost issues connected to the collection and conversion of solar energy into heat for regeneration of a liquid desiccant using a solar collector cum regenerator as observed from reviewed literature. These entail further experimental evaluation of the actual solar energy conversion capability of the solar regenerator and its thermal performance to develop the technology further and improve the design. In this work, the optical efficiency and thermal performance of a glazed and an unglazed solar collector/regenerator is experimentally evaluated taking into consideration a calcium chloride solution flowing over the absorber surface. EXPERIMENTATION Solar Collector/ Regenerator Setup A solar collector/regenerator was fabricated from black painted, corrugated GI sheets with 12 mm thick bottom insulated by thermocol material; a wooden frame (with support for glass) having ventilation air openings at its front & rear sides, and 15 cm diameter PVC discharge pipe at IIT Delhi. The solar absorber area of the collector/regenerator is 2.2 m x 1.84 m and its aperture area is 2.6 m x 1.84 m. The glazing is detachable and the same solar regenerator is used for studying unglazed regeneration performance of a calcium chloride solution. The assembly of the solar collector/regenerator was mounted on a steel structure inclined at 14o from the horizontal. The liquid desiccant regeneration system consists of a solution tank, a solution pump, a concentric pipe solution distributor and a glazed or unglazed solar collector/regenerator (refer Fig.1). The experimental procedure adopted was that certain quantity of calcium chloride solution was kept in the solution tank and its initial temperature, volume and density are recorded to calculate the initial mass of the desiccant and its concentration in the solution. The solution was dispensed over the absorber surface of the solar collector/regenerator through the distributor with the help of a centrifugal pump and the liquid desiccant coming out of the solar regenerator was allowed to mix with the solution in the tank until the temperature of the solution in the tank and the solar absorber were same. The mixture was then continuously circulated over the absorber surface from about 9:00 am up to 4:00 pm with manual measurement of solution density and temperature at every half an hour interval. The instruments used in the experimentation were pyranometer, hand held densitometer, thermocouple wires fixed on the absorber surface, and RTDs. Data needed to evaluate the regeneration performance of the solar collector/regenerator were collected every 10 seconds (30min for the solution density) with the help of a data logger over the period of experimentation.
16 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
FIG. 1: PHOTOGRAPH OF THE SOLAR COLLECTOR CUM REGENERATOR SETUP
Optical Setups Solar absorptance of absorber and solution, and solar transmittance of glass and solution were estimated using separate experimental arrangements as indicated in Fig.2a and Fig.2b. The solar absorptance of the black painted plate was measured by keeping one pyranometer upside down to detect the reflected light from the absorber surface and the other one exposed to the sun to measure the incident solar flux for dry as well as wet absorber surface as shown in Fig.2a. The position of the pyranometer was fixed at a particular point on the absorber surface to detect the general reflected light from the surface which changes when the position of the sun relative to the absorber is changing. The optical transmittance of glass and absorptance of solution were estimated based on measurements taken using three pyranometers; the first one exposed to the sun, second one under the glass covered with 3/8 mm thick solution, and the third one under a single glass.
FIG. 2: OPTICAL PERFORMANCE STUDY SETUPS: A) SOLAR ABSORPTANCE MEASUREMENT B) SOLAR TRANSMITTANCE & ABSORPTANCE
DATA REDUCTION Optical Performance The optical resistance of the glazed-solar collector/regenerator includes light transmission resistance of the glass and solution, reflection of light from the solution and absorber surface, shading of the absorber surface by glass supporting frames and absorption of light energy by solution and absorber plate. For Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 17
ICORE 2013
unglazed collector the optical resistance is due to absorption and reflection of the solar energy only from the solution-absorber surface. The optical efficiency of a glazed and unglazed solar collector/regenerator is given by Eq. 1 and Eq. 2, respectively. (1) (2) Two glass supporting frames were shading the absorber area: vertical support 5 cm wide and 2.2 m long and horizontal support 5 cm wide and 1.84 m long. Taking the ratio of the areas of these frames and the area of the absorber, the shading coefficient was found as 0.049 approximately for noon time. Actually the shading varies as the position of the sun changes relative to the absorber. The solution also collects solar energy when it is in the gutter which is 15 cm in diameter and 1.87 m long. The projected area is equivalent to 0.069 times the absorber area which offsets the shading effect by the glass frames. For clean glass surface, the dust coefficient is taken as zero; for unglazed solar regenerator, the dust coefficient is taken as approximately zero, which is also difficult to measure. The transmittance of glass and combined transmittance of the glass & solution can be obtained by measuring the incident light energy using a pyranometer exposed to the sun and the transmitted radiation using two pyranometers, one under a glass and the other under a glass covered with 3/8 mm thick solution as (refer Fig. 2b) (3)
(4) Neglecting the reflected light from the solution surface being small, the solar absorptance of solution can be found as (5) Thus, the transmittance of light through the solution can be found from the solar absorptance of solution (neglecting reflectance) as (6) The combined absorption of light by the solution and black painted absorber surface can be obtained using the principle of conservation of energy for unglazed solar absorber over which solution is flowing. For any opaque surface, energy from all directions are either absorbed or reflected. The input energy is the incident light on the absorber surface which is reflected from the absorber-solution surface and absorbed by the absorber-solution. In terms of the intensity of solar flux, the energy conservation is expressed as (7) Dividing Eq. (7) by the incident solar flux and rearranging gives an expression for the combined solar energy absorptance coefficient of the solution and the absorber plate as (8) The reflected and absorbed light coefficient are defined, respectively as (9)
18 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
(10) Thermal Performance The net heat absorbed by the collector plate is the product of the optical efficiency of the solar regenerator, ηopt, mean solar flux (over time Δt), incident solar radiation, I, solar absorber plate area, Aab, and the length of solar energy collection time, ∆t. (11) The total heat loss from the solution-absorber surface is a fraction of the absorbed solar energy and it comprises of sensible heating of the ventilation air & solution; surface heat loss by radiation and convection in top direction, and heat loss by conduction in the bottom direction. It is given by (12) The mass of water evaporated, increase in concentration of the desiccant solution, and energy of evaporation of water are evaluated using the procedure discussed by Gezahegn et al. [28] with introduction of slight modification on estimating the energy of evaporation of water which is the sum of the energy of evaporation of pure water and the differential enthalpy of dilution (Conde [29]).
(13)
(14) For aqueous solution of calcium chloride the energy of evaporation is given by (15) The thermal performance of the solar collector cum regenerator can also be assessed using figure of merit and concentration effectiveness (Fagbenle & Karayiannis [14]). These are, respectively given by (16)
(17) The average concentration is defined by Fagbenle & Karayiannis [14] as
(18) The solar collector cum regenerator efficiency is a critical parameter that decides the actual energy conversion capability of the solar collector, given by (19)
Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 19
ICORE 2013
RESULTS AND DISCUSSION Optical Performance Three pyranometers (A, B and C) with sensitivity of 13.35, 13.65 and 12.58/ were used to study the absorptance and transmittance of the solar collector, glass and solution. The relative error in detecting irradiance using pyranometer B compared to A is 0.04-5% for irradiance level greater than 0.60 kW/m2, 5-10% for irradiance level between 0.2 and 0.6 kW/m2. Taking pyranometer A as a reference (used for thermal performance evaluation), the errors increase with decreasing irradiance intensity and become nearly zero at maximum irradiance intensity. Pyranometer C is also calibrated against pyranometer A. The transient trend of the solar flux detected by these pyranometers over a period of approximately7 hours are depicted in Fig. 3.
FIG. 3: VARIATION OF SOLAR FLUX AS DETECTED BY 3 PYRANOMETERS WITH TIME
The solar transmittance of glass covered by 3/8 mm thick solution and solar absorptance of the solution are depicted in Fig. 4 and Fig. 5. The solar transmittance is more or less constant as shown in these figures. The average combined transmittance of 3 mm thick solution and glass at desiccant concentration of 0.4 and 0.3 were 0.75 and 0.76, respectively. It is 0.66 for 8 mm solution thickness. These values indicate that the optical transmission resistance of the solution is mainly a function of thickness of the solution rather than the desiccant concentration as shown in Fig.6. The absorber plate of the solar regenerator used is a corrugated one and half of its area approximately holds 7-8 mm solution in the trough and about 3 mm of solution is flowing over the crest. Hence the optical transmission of the solution and glass can be taken as the average of these two which is 0.71.
Fig. 4: Variation of Iinc, Itra, and τg_sol, with Time at ξ=0.4 20 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
Fig. 5: Variation of Iinc, Itra, τg_sol, τg and αsol with time at ξ=0.4
FIG. 6: VARIATION OF 3/8 MM THICK SOLUTION & GLASS TRANSMITTANCE WITH TIME
FIG. 7: VARIATION OF 3/8 MM THICK SOLUTION ABSORPTANCE AT ξ 0.4 WITH TIME
Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 21
ICORE 2013
The transient variation of solar absorptance of calcium chloride solution at desiccant concentration of 0.4 with solution thickness of 3 mm and 8 mm is shown in Fig.7. As seen, the solar absorptance varies from 0.15 to 0.2 and 0.05 to 0.11, with average value of 0.165 and 0.11 for 8 mm and 3 mm thick calcium chloride solution, respectively. Thus, the light transmittance of the solution alone is 0.835 and 0.89 for solution thickness of 8 mm and 3 mm, respectively and the average value is 0.862. The transient values of solar transmittance of single and double glass varies from 0.84 to 0.79 and from 0.7 to 0.6, respectively as shown in Fig.8. The glasses are 3.2 mm thick and the average solar transmittance of the single and the double glass are found as 0.812 and 0.659, respectively. This implies that the use of double glass reduces the optical performance of the solar regenerator by 18.8% as compared to single glass.
FIG. 8: VARIATION OF DOUBLE AND SINGLE GLASS SOLAR FLUX TRANSMITTANCE WITH TIME
FIG. 9: VARIATION OF SOLAR ABSORPTANCE OF ABSORBER WITH TIME
The value of absorptance of the plate ranges from 0.919 to 0.966 over a period of one hour as shown in Fig. 9. It is closer to the measured value of solar absorptance of a black painted collector by Hawlader et al. [10] which was 0.95. The overall optical efficiency of the solar collector/regenerator when it is unglazed, single glazed, or double glazed with corrugated absorber which can hold 3/8 mm thick solution over its surface are calculated using Eq.2 and Eq. 1, and the results are 94.9, 77.0 and 62.5 %. The optical efficiency of unglazed solar regenerator is greater than the optical efficiency of a single and double glazed C/R by 17.8 % and 32.4%, respectively. But, the advantage of better optical efficiency of unglazed regenerator is countered by accumulation of dust on the absorber surface and contamination of the solution by dust, 22 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
leaves and other foreign bodies which clog the liquid desiccant distributor holes. These cause reduction of coverage of the absorber surface with liquid desiccant, reducing the mass transfer of vapor in addition to increasing surface heat losses. THERMAL PERFORMANCE The thermal performances of glazed and unglazed solar C/R were evaluated in terms of net heat absorbed, heat loss from the collector, mass of water evaporated, increase in concentration, figure of merit, concentration effectiveness and overall solar collection and regeneration efficiency. The experimental thermal performance of glazed and unglazed regenerators on solar regeneration of calcium chloride solution carried out on two different days are elucidated below. The initial mass of solution in the tank for glazed and unglazed experiments were 49.54 and 48.18 kg, respectively.
Fig. 10: Variation of Iinc and mv,∆t for a Glazed Solar C/R with Time
Fig. 11: Variation of Iinc and mv,∆t for an Unglazed Solar C/R with Time
Fig. 10 and Fig.11 show the variation of irradiance and evaporated mass of water for glazed and unglazed solar regenerator, respectively. The average solar flux received by the glazed and unglazed solar regenerator over the period of regeneration were 0.81 and 0.692 kW/m2, respectively. The total mass of water evaporated by the glazed regenerator was 17 kg and for unglazed one it was 15.2 kg.
Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 23
ICORE 2013
FIG. 12: VARIATION OF INCREASE IN CONCENTRATION OF DESICCANT AND CONCENTRATION EFFECTIVENESS OF GLAZED AND UNGLAZED SOLAR REGENERATORS WITH TIME
As indicated in Fig.12 the increase in concentration of the desiccant for the glazed solar regenerator was almost the same as the increase in concentration of the unglazed regenerator up to 12 pm, but increases slightly later in the afternoon. The concentration of calcium chloride increased from 0.31 to 0.47 for glazed regenerator, while it increased from 0.30 to 45 for unglazed regenerator. The percent increase in concentration of the glazed and unglazed solar regenerator were 52.2 and 46.4%, respectively and the concentration effectiveness of the solar regenerators are more or less the same. It is varying between 0.5 and 0.51 as shown in Fig.12 which is similar to the results of Fagbenle & Karayiannis [14]; they reported that the maximum and minimum values of concentration effective at noon and at 4 p.m. as 51.9% and 49.9%, respectively. Fig.13 shows the variation of incident solar energy intercepted by the glazed and unglazed solar regenerators together with the energy of evaporation of water. The incident solar energy on the glazed collector/regenerator is greater than the unglazed one by 18.1%, due to experimentation of different days. The energy of evaporation of water from the glazed solar regenerator is greater than the unglazed one due to more evaporation of water as a result of higher irradiance.
FIG. 13: VARIATION OF INCIDENT SOLAR ENERGY AND ENERGY OF EVAPORATION OF WATER WITH TIME
In Fig. 14, the variation of net thermal energy absorbed and heat loss from the absorber surface are shown for glazed and unglazed solar regenerators. The net thermal energy absorber by the glazed and unglazed regenerators are almost same at the beginning and end of experimentation and it is slightly higher for the unglazed regenerator around noon. But the heat loss from the surface of the absorber is greater for the unglazed solar regenerator than the glazed one. The total heat loss, over the period of regeneration, from the unglazed solar regenerator is 26327.8 kJ and from glazed one is 18333.8 kJ, 24 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
which represent 37.3% and 22% heat losses compared to the total incident solar energy received by the regenerators (70551.4 kJ and 83307.9 kJ for unglazed and glazed regenerators, respectively). These results show that total heat loss from the absorber surface of unglazed regenerator by radiation and convection was more than the total heat loss from glazed one by 7994 kJ. The total loss of light energy from glazed regenerator (due to single glazing transmission resistance) was more than the total optical energy loss of unglazed one by 11781.4 kJ which is greater than energy loss by heat from unglazed regenerator. Since heat energy is mainly needed for regeneration of liquid desiccants, glazing the solar regenerator is advantageous in addition to obtaining clean & strong liquid desiccant.
FIG. 14: VARIATION OF NET THERMAL ENERGY ABSORBED AND HEAT LOSS FROM ABSORBER WITH TIME
Fig. 15 illustrates the variation of glazed and unglazed solar collector cum regenerators’ figure of merit and overall solar collector/regenerator efficiency. The overall solar collector cum regenerator efficiency of unglazed and glazed solar regenerators are in the range of 40 to 53% and 36 to 51.4%, respectively between 10 am and 2:30 pm. Hawlader et al. [10] reported regeneration efficiency of 11 m x 11 m unglazed solar regenerator as varying between 37 to 67% from 10 am to 2:30 pm. The regeneration efficiency of unglazed regenerator is almost equal to the glazed one. The solar collector/regenerator efficiency is a critical parameter that decides the cooling performance of a liquid desiccant air conditioning system with a solar collector cum regenerator as liquid desiccant regenerator. The average figure of merit of the glazed and unglazed regenerators are 70.4% and 59.8%, respectively. According to Fagbenle & Karayiannis [14] the figure of merit of unglazed regenerator are reported to be up to 0.4, but from the results obtained in this study it can go up 74 % for unglazed and 83.8% for glazed regenerator as indicated in Fig. 15.
FIG. 15: VARIATION OF FIGURE OF MERIT AND SOLAR COLLECTOR/ REGENERATOR EFFICIENCY WITH TIME Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 25
ICORE 2013
CONCLUSION In this study, light transmission resistance of a 3.2 mm thickness glass, solar absorptance of 3/8 mm thick calcium chloride solution of concentration 0.3 and 0.4, and thermal performance of glazed and unglazed solar regenerators were experimentally evaluated. The experimental results show that average transmittance of light through a double glass is 18.8 % less than a single glass. The optical efficiency of unglazed regenerator is found to be higher than single and double glazed solar collector/ regenerator. However, the solar absorber with single glazing helps to maintain clean liquid desiccant compared to unglazed one. The solar absorptance of a calcium chloride solution increases with thickness and solution concentration has negligible effect. The thermal performance of glazed and unglazed solar collector cum regenerators were compared based on experimental data collected on different days from 9 a.m. up to 4 p.m.. The initial masses of calcium chloride solution used for the experiment were 49.5 and 48.1 kg for glazed and unglazed solar regenerators, respectively. The total mass of water evaporated from the solutions were 17 kg and 15.2 kg which resulted 52.2 % and 46.45% increase in concentration for glazed and unglazed regenerators, respectively. The collector cum regenerator efficiency was ranging from 36-51% for glazed regenerator and 40-53% for the unglazed regenerator. ACKNOWLEDGEMENT The authors would like to acknowledge the financial support received from Ministry of new and renewable energy, Government of India. The contributions of all the past students and staff who were involved in this work are also acknowledged. NOMENCLATURE
26 ♦ Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator
ICORE 2013
REFERENCES [1] Mullick, S.C. and Gupta, M.C., 1974, Solar desorption of absorbent solution, Solar Energy; 6:19-24. [2] Peng C.P. and Howell, J.R. 1982, Analysis of open inclined surface solar regenerators for absorption cooling applications comparison between numerical and analytical models, Solar energy; 28:265-268. [3] Johansen, A. and Grossman, G., 1993, Performance simulation of regenerating type solar collectors, solar energy, Vol. 30, No.2:87-92. [4] Gandhidasan P., 1983, Theoretical study of tilted solar still as a regenerator for liquid desiccants, Energy convers. Mgmt.Vol 23 No. 2, (1983):97-101. [5] Gandhidasan P., 1983, A simple analysis of an open regeneration system. Solar energy Vol.3, No.31:343-345. [6] Kaushik S.C., Kaudinya, J.V., and Chandra, S. 1985, Diurnal response of an open roof solar regenerator absorption system. Energy Convers. Mgmt.; 25: 21-27. [7] Kaushik S.C., and Kaudinya, J.V., 1989, Open cycle absorption cooling-a review. Energy Convers. Mgmt.; 29: 89-109. [8] Kaushik S.C., Kaudinya, J.V., and Yadav, Y.K., 1992, Studies on some solar collector/regenerator systems for open cycle absorption air conditioning/liquid desiccant cooling systems, Heat recovery system and CHP Vol.12, No.4:357-363. [9] Pradeep K., and Devotta, S., 1989, Modelling of the thermal behavior of solar regenerator for open-cycle cooling systems. Applied energy; 33:.287-295 [10] Hawlader M.N.A., Novak K.S. and Wood B.D.,1993,Unglazed C/R performance for a solar assisted open cycle absorption cooling system, Solar energy; 50:59-73. [11] Yang R., and Wang P.L., 1994, Experimental study of a forced convection solar collector/regenerator for open-cycle absorption cooling, Journal of solar energy engineering; ASME 194/ vol.116, November 1994. [12] Yang R., and Wang P.L., 1995, The effect of heat recovery on experimental study of a forced convection solar collector/regenerator for open cycle absorption cooling, Journal of solar energy engineering; 54: 19-24. [13] Yang R., and Wang P.L., 1998, Experimental study for a double glazed forced convection solar collector/regenerator for open cycle absorption cooling system, Journal of Solar energy engineering, Vol. 120/257 November 1998 [14] Fagbenle R.L., and Karayiannis T.G., 1998, A thermodynamic analysis of a simple open flow solar regenerator, Applied thermal engineering; 18(1998): 1359-1374. [15] Alizadeh S., and Saman W.Y., 2002, An experimental study of forced flow solar collector/regenerator using liquid desiccant, Solar Energy; 73345-362. [16] Alizadeh S., and Saman W.Y., 2002, Modeling and performance of a forced flow solar C/R using liquid desiccant, Solar Energy; 72(2002): 143-154. [17] Kabeel A.E., 2005, Augmentation of the performance of solar regenerator of open absorption cooling system, Renewable energy; 30: 327-338. [18] Elsarrag E., 2008, Evaporation rate of a novel tilted solar liquid desiccant regeneration system. Solar energy; 82:663-668 [19] Yutong L., and Yang H., 2010, Experimental study of an open-cycle solar C/R using liquid desiccant for air conditioning, International Journal of green energy; 7:273-288. [20] Collier R.K., 1979, The analysis and simulation of an open cycle absorption refrigeration system, Solar Energy; 23:357-66 [21] Bolzan M., and Lazzarin R., 1979 Comparison between two absorption cooling systems of open type under different climatic conditions, International journal of refrigeration, Volume 2, number 3 May 1979 [22] Mattarolo L. 1982, Solar powered air conditioning systems: a general survey. Volume 5, numero 6, November, butterworth & Co (publisher) Ltd and IIR [23] Yang R., and Wang P.L., 2001, A simulation study of performance evaluation of single glazed and double glazed solar collectors/regenerators for open cycle absorption solar cooling system, Solar energy; 71:263-268. [24] Haim I., Grossman, G., and Shavit A., 1992, Simulation and analysis of open cycle absorption systems for solar cooling, Solar Energy, 49(1992):515-534. [25] Gandhidasan P., 1994 Performance analysis of an open-cycle liquid desiccant cooling system using solar energy for regeneration, International Journal of refrigeration, 1994 volume 17 number 7 [26] Katejanekarn T., and Kumar S., 2008, Performance of a solar-regenerated liquid desiccant ventilation pre-conditioning system, Energy and buildings; 40:1252-1267 [27] Katejanekarn T., Chirarattananon S., and Kumar S., 2009, An experimental study of a solar regenerated ventilation preconditioning system, Solar energy; 83:920933 [28] Gezahegn H., Rishi R., Mullick S.C., and Sanjeev J., 2011, Performance analysis of a liquid desiccant solar regenerator, ICORE2011 proceedings, Tezpur University, Assam, India, 2-4 Nov. 2011. [29] Conde M.R., 2004, Properties of aqueous solutions of lithium and calcium chlorides: formulations for use in air conditioning equipment design, International journal of thermal science; 43:367-382.
Optical and Thermal Performance of a Liquid Desiccant Solar Regenerator ♦ 27
Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector Piyush Choudhary Deputy Superintending Engineer (Electrical), Oil & Natural Gas Corporation (ONGC), New Delhi, India Department of Electrical Engineering, Indian Institute of Technology, Banaras Hindu University, Varanasi, India E-mail: [email protected]/ [email protected] Abstract—“Building Solar-India” is the vision of India supported by dedicated Jawaharlal Nehru National Solar Mission (JNNSM) [1] of Government of India. Based on fossil fuel reserve and abundance Solar Isolation, an innovative R&D project is conceptualized. The concept is oriented primarily on the use of Solar Energy for the crude oil heating purpose besides other many prospective industrial applications to reduce cost, carbon foot print and moving towards a sustainable and ecologically friendly Oil & Gas Industry. A Parabolic Dish [2] Concentrated Solar Power prototype system would be developed to substitute the presently used natural gas for heating purpose by solar thermal energy. The use of solar energy in the oil industry gives the opportunity for fossil fuels to be partly economized, and to improve the conditions of work, safety measures and ecology. Reduction on Natural Gas is predicted by 40% with this hybrid system. A positive implication for ecology, working conditions and safety measures is additive advantage. There could also decreasing air venting of CO2, CH4 and N2O by an average of 30-35%. Keywords: Solar Energy, CSP, Heater Treater, Oil & Gas, Sustainable Development, Parabolic Dish
INTRODUCTION India's vision is for economic development [3] with energy-efficiency. With fast growing economy India must pioneer a graduated shift from economic activity based on fossil fuels to one based on non-fossil fuels and from reliance on no-renewable and depleting sources of energy to renewable source of energy [4]. In this strategy, the sun occupies centre-stage, as it should, being literally the original source of all energy [5]. This mission shall need to pool scientific, technical and managerial talents, with sufficient financial resources, to develop solar energy as a source of abundant energy to power our economy and to transform the lives of our people. Our success in this endeavour will change the face of India. It would also enable India to help change the destinies of people around the world. With the push towards sustainable power production and the increasing realisation for the need to reduce CO2 emissions [6], renewable sources of energy are becoming an increasingly important element in the world energy balance. Concentrated Solar Power systems [7] have the potential to replace conventional fossil fuels. They will also help mitigate the possible effects of climate change. CONCENTRATING SOLAR POWER CONCEPT Concentrating Solar Power (CSP) [6.] is a renewable power generation technology that uses mirrors or lenses to concentrate the sun‟s rays to heat a fluid, e.g., water, which produces steam to drive turbines. CSP differs from solar photovoltaic (PV) technology, which directly converts the sun‟s ultraviolet radiation to electricity using semiconductors. In CSP systems the Sun‟s rays are focused through optical devices. These focused rays generate heat which can be used either to generate steam and electricity or to trigger chemical reactions. Dishes have been used for pilot scale natural gas reforming, steam generation, ammonia dissociation, Brayton Cycles [8.] and Stirling engines, as well as for concentrating photovoltaics. Because no input fuel is required, CSP plants release little or no carbon dioxide equivalent (CO2e) emissions [32.]Though CSP line-focusing parabolic trough technology is a proven technology with more than 350 megawatts (MW)[23.] of installed capacity operating commercially in the world, several emerging technologies that promise higher conversion efficiencies and cost-competitive generation have been demonstrated on a smaller scale are the areas of research. These technologies, such as pointfocusing parabolic dish and line-focusing Fresnel reflectors, may extend the ability of CSP to provide major industrial application on solar heat energy [27.] CONVENTIONAL APPLICATION IN OIL AND GAS SCENARIO While crude oil is being extracted from a well, accompanying gases in oil contain mechanical mixtures such as sand, clay, salt crystals and water. The crude oil`s high viscosity is area of concern for its flow ability [14.]. Separation of this mix is first need on processing of crude oil. Oil treatment [15] and transportation through pipelines becomes difficult if there is water and mechanical mixture in its contents. Oil separating from accompanying gas after being collected into barrels, water, salt and mechanical mixtures need to be refined. That‟s why oil moisture primary treatment is advantageous in oil fields. At present, according to globally trend, oil‟s physical-chemical properties, the oil must be heated till 80ºC95 ºC for initial treatment [17]. A 'Heater Treater' is used in the Oil and Gas production process and is
ICORE 2013
used to remove water and gas from the produced oil and to improve its quality for sale into a crude oil pipeline or for other transport [17]. A heater-treater tropical combines the following components inside the heater treater: a heater, free-water knockout, and oil and gas separator [17]. The heating section is provided to raise the temperature of the incoming crude stream to assist the coalescing of the oil/water emulsion [14,16,17].
FIG. 1: TYPICAL CRUDE OIL TREATMENT SYSTEM (SEPARATION, HEATING, DEHYDRATION, ETC.)
The process is being done in Heater treater assembly where first phase of crude oil heating is carried out by heating through burning of natural gas, followed by degasser section which separates gas emulsion. Balance first hand treatment is based on electrical polarization [17]. In heater-treater, a tank is divided into separate heating and treating sections and the heating section is in turn divided into a receiving zone and an enclosed heating chamber [17,19]. A pair of down comers conveys emulsion from the top of the receiving zone to the bottom of the heating chamber, and a portion of the emulsion flowing from the down comers is direction across a discharge passageway to flush sediment and free water from the chamber. The emulsion flows diagonally upwardly in the heating chamber past a pair of fire tunes and is conveyed from the top of the chamber to the bottom of the treating section by a conduit. The operation of the fire tubes is controlled by thermostat located in the upper portion of the heating chamber adjacent the inlet of the conduit. Within „energysaving technology and energy effectiveness‟ taking into consideration the energy situation, fossil fuel reducing (in the next 30-50 years), to apply our minds expediently for economical renewable fuels and environment safety due to the high-grade potential solar energy application, these expenses can be partially covered. Based on the above mentioned for realizing this process, need of creation and developing high grade temperature solar system with a modular parabolic dish concentrator corresponding to the technological process of primary crude oil heating. A schematic diagram of the experimental parabolic dish based crude oil heating system can be seen in Fig. 2.
Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector ♦ 29
ICORE 2013
FIG. 2: GENERAL SCHEMATIC VIEW OF THE PROTOTYPE DISH CSP CRUDE OIL HEATING SYSTEM
PILOT SCALE PROTOTYPE & SOLUTIONS The routine temperature required to heat the crude is 80ºC [17,19] (considering initial temperature is 30ºC) through three phase heater treater system. The natural gas is fuel for the Heater treater and helpful to achieve the required temperature. With help of a by–pass valve a system is proposed to heat the crude oil by solar energy through heat exchanger. Water is heated to 85–90°C by solar rays being reflected from the concentrator‟s surfaces [17]. Heated water circles in the solar thermal receiver-heat exchanger expansion tank system by going out from the solar receiver. A pump power shall be utilized for heat transfer to move. During a definite time (30–60minutes) after having got stationary regime oil coming from the oil tank untreated, it comes into the coil pipe in the heat exchanger. In this way, demulsifier is added to it. Here, the heat exchanging process is happening between oil and water oil, and is heated to demanded temperature (80°C) then leaving, and heat exchanger oil begins to go to the sedimentation tank. The oil mixture with demulsifier flows into the sedimentation tank. After having kept it approximately 24 hours, the separated water goes out the tank, and treated oil is sent to flow line. TABLE 1: DESIGN PARAMETERS OF PROTOTYPE SYSTEM AS PER INDIAN OIL AND GAS SECTOR Total flow of crude Pressure of Crude Uses of Natural Gas for Crude Heating Specific Heat of Crude Specific Gravity of Crude at 15°C Descific temperature of Crude Natural gas density into calculation 1kg of natural gas will generate
10m3/hr 4kg/sqcm 70m3/hr 0.5 0.85 80°C 0.8kg/m3 13.5KWh
Steam cycles work by expanding the high pressure steam through a turbine which converts the energy in the steam to mechanical work which drives a generator. The efficiency of these turbines is influenced by the pressure drop across the unit, which is a function of the ‘cold sink’ 30 ♦ Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector
ICORE 2013
[28] temperature i.e., the temperature at which thermal energy is rejected from the system through cooling. Wet (evaporative) cooling provides the lowest temperature for the cold sink but results in a significant usage of water. Dry cooling [27] is less efficient (by about 10%) but largely eliminates the consumption of water in the power cycle. However, this also results in increased capital cost for a given capacity and a higher electricity cost. Storage [29] is another important consideration for a CSP plant. The current trend is for plants with upwards of 6 hours of storage (i.e., the amount of thermal energy required to operate the power block at full capacity for 6 hours). This allows plants to be less impacted by variations in solar radiation through the day, and to continue operating after the sun has set. Though there are increased capital costs associated with this ability due to the need for extra solar collectors, a larger receiver system, and of course the storage system and medium itself, these costs are offset by the extra operating hours, particularly at periods of higher tariff [27,29]. Current practice is to use “solar salt”[30], a mixture of sodium and potassium nitrate which melts at around 220°C and is stable to about 590°C, although there is considerable research into new materials to extend the upper temperature limit.
FIG. 3: ENERGY FLOW DIAGRAM FOR A CSP POWER PLANT
SCOPE OF RESEARCH & SYSTEM DESIGN The prototype will be hybrid with existing natural gas fired crude oil heating facility. The performance of the prototype could be extensively evaluated and the results will provide the benchmark data on building solar thermal crude heater. Based on current gas values it could be estimated the total power (Electricity & Heat) in solar heating system that could result in annum saving in fuel cost and CO2 equivalent emissions reduction. The payback period of the proposed system could be estimated subsequently. During the prototype, the parameters which would be further be considered which includes: •
Direct solar radiation and wind speed.
•
Air temperature and temperature of skin of dish.
•
Temperature of heat transfer and oil and Flow rate of heat transfer and crude oil.
•
Thermal storage. Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector ♦ 31
ICORE 2013
The thermal energy generated by a CSP solar field does not need to be immediately used for power generation but can be stored for later use. Thermal energy can be stored much more efficiently than electrical energy, typically in the form of molten salt held in highly insulated storage tanks. Other alternative storage media, including concrete, water, synthetic oils, and phase-change materials, are being considered. Storage gives CSP technology several considerable advantages: •
Reliable operations during cloudy or night time conditions:
•
Near instantaneous dispatchable power to meet expected and unexpected peak demand.
•
The ability to shift electrical production from the natural peak of insolation to higher-priced peak demand, thereby increasing profitability and investment Returns.
•
The ability of the solar field to be oversized relative to turbine capacity, thereby decreasing turbine costs, increasing the capacity factor, and reducing the payback Period.
FIG. 4: FLOW DIAGRAM FOR DEVELOPING THE HYBRID PROTOTYPE SYSTEM
CONCLUSION A low Carbon energy initiative, reducing Carbon foot and enhancing energy security would be main objective in coming days and CSP based system would definitely a call for future in industrial aspects. This would be in view of Carbon Neutrality Mission [37.] Strive to reduce CO2 emissions for some industrial sectors. Economies are using some 40% of fossil fuels, with implications for ecology, working conditions and safety measures. There is also decreasing air venting of CO2, CH4 and N2O by an average of 30–35%. The computations in this experiment show that by the utilization of the proposed solar reactor based on a parabolic trough double modular concentrator; it is possible to achieve the temperature of water that reaches the vaporization temperature, which enables the given installation needed for various technical purposes. The benefits and these solutions could be game changes due to following add-ons in following areas: •
Social and environmental importance for reduced pollution & dependency on fossil fuels.
•
Importance of the project from climate change mitigation aspect [32].
•
Vision statement emphasizes for growth through sustainable development [2].
32 ♦ Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector
ICORE 2013
•
Unique and first of a kind-prototype can be patented.
•
Can accrue the Clean Development Mechanism (CDM) benefits [32].
REFERENCES [1] http://mnre.gov.in/file-manager/UserFiles/draft-jnnsmpd-2.pdf [2] http://www.mnre.gov.in/solar-mission/jnnsm/introduction-2/ [3] Stoddard, L.; Abiecunas, J.; O‟Connell, R. (2006). Economic, Energy, and Environmental Benefits of Concentrating Solar ower in California. NREL/SR-550-39291. Golden, CO: National Renewable Energy Laboratory. [4] http://planningcommission.nic.in/reports/genrep/pl_vsn2020.pdf [5] Gilman, P.; Blair, N.; Mehos, M.; Christensen, C.; Janzou, S.; Cameron, C. (2008). Solar Advisor Model User Guide for Version 2.0. NREL/TP-670-43704. Golden, CO: National Renewable Energy Laboratory. [6] Renewable Energy Technologies: Cost Analysis Series, The International Renewable Energy Agency (IRENA) 2012 [7] Anders, S. et al. (2005), Potential for Renewable Energy in San Diego Region, San Diego Regional Renewable Energy Group, San Diego, CA. [8] Strategic Research Priorities for Solar Thermal Technology European Technology Platform on Renewable Heating and Cooling [9] A.T. Kearney and ESTELA (2010), Solar Thermal Electricity 2025, ESTELA, Brussels. http://www. estelasolar.eu/index.php?id=22 [10] Cohen, G.E., D.W. Kearney, G.J. Kolb. (1999), Final Report on the Operation and Maintenance Improvement Program for Concentrating Solar Power Plants, SAND99-1290, Sandia National Laboratories, Albuquerque, NM. [11] CSP Today (2008), An Overview of CSP in Europe, North Africa and the Middle East, CSP Today, London. [12] Emerging Energy Research (2010), Global Concentrated Solar Power Markets and Strategies: 2010–2025, IHS, Cambridge, MA. [13] Ernst & Young and Fraunhofer Institute for Solar Energy Systems (Fraunhofer) (2011), MENA [14] http://pascagoula.chevron.com/home/abouttherefinery/whatwedo/processingandrefining.aspx [15] “Surface Equipment” Series Instructional Training Binder, New Mexico Junior College [16] http://www.oilfield.com/pumpers10e.html [17] http://ocw.utm.my/file.php/12/Chapter_7-OCW.pdf [18] Assessment of the Local Manufacturing Potential for Concentrated Solar Power (CSP) Projects, The World Bank, Final Report, Washington, D.C. [19] http://www.valerus.com/products-services/production-equipment/crude-oil-condensate-treating-stabilization/ [20] Fichtner (2010), Technology Assessment of CSP Technologies for a Site Specific Project in South [21] Africa Final Report, the World Bank and ESMAP, Washington D.C. [22] Hinkley, J. et al. (2011), Concentrating Solar Power-Drivers and Opportunities for Cost-competitive Electricity, Victoria. [23] International Energy Agency (IEA) (2010), Technology Roadmap: Concentrating Solar Power, IEA, Paris. http://www.iea.org/papers/2010/csp_roadmap.pdf [24] Kutscher, et al. (2010), Line-Focus Solar Power Plant Cost Reduction Plan, NREL, Boulder, CO. Muller-Steinhagen, H. (2011), Solar Thermal Power Plants-On the way to commercial market [25] ESTTP (European Solar Thermal Technology Platform): Solar Thermal Strategic Research Agenda 'Solar Heating and Cooling for a Sustainable Energy Future in Europe', 2008 www.rhc-platform.org/structure/solar-thermal-technologypanel/downloads [26] UNIDO (2011), Renewable Energy in Industrial Applications. An assessment of the 2050 potential, UNIDO, Geneva. [27] Technology Roadmap Solar Heating and Cooling. International Energy Agency IEA, 2012 [28] http://www.nrel.gov/csp/troughnet/pdfs/27925.pdf [29] http://www1.eere.energy.gov/solar/sunshot/csp_storage.html [30] http://www.solarnovus.com/potassium-calcium-nitrate-promising-for-csp-heat-storage_N6044.html [31] Macroeconomic impact of the Solar Thermal Electricity Industry in Spain, Study elaborated by Delloitte Pza. Pablo Ruiz Picasso, 1, Torre Picasso, 28020 Madrid, España October 2011 [32] Arvizu, D., et al., “Direct Solar Energy”, in: IPCC Special Report on Renewable Energy Sources and Climate Change Mitigation [O. Edenhofer, O. et al. (eds)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. [33] Dupeyrat, P., S. Fortuin, G. Stryi-Hipp (2011). Photovoltaic/Solar Thermal hybrid collectors: overview and perspective, ESTEC 2011. [34] European Solar Thermal Technology Platform (ESTTP) (2007) Solar Heating and Cooling for a Sustainable Energy Future in Europe, ESTTP Brussels. [35] Hoogwijk, M., W. Graus (2008), Global potential of renewable energy sources: a literature assessment. Background report prepared by order of REN21. Ecofys, Utrecht. [36] http://www.idfc.com/pdf/report/Chapter-24.pdf [37] http://www.tuv-nord.com/in/carbon-services/carbon-foot-print-neutrality-1117.htm
Substitution of Natural Gas by Solar Energy using Concentrated Solar Power Technology in Oil & Gas Sector ♦ 33
A Prudent Study on Solar Water Heating System using RETScreen in India Sambeet Mishra1, Sayed Majid Ali2, D.P. Kothari3 and Pratyasha Tripathy4 1.4School
of Electrical Engineering, KIIT University, Bhubaneswar, India School of Electrical Engineering, KIIT University, Bhubaneswar, India 3Director General, JB Group of Educational Institutions, Hyderabad, India E-mail: [email protected], [email protected], [email protected], [email protected] 2Professor,
Abstract—Having been able to obtain the solar energy at a high solar insolation for an average of 300 sunny days per year, the major problems of the country, India, can be easily solved using the innovative technologies of solar energy, which includes the solar water heating system. But the lack of awareness, concerns about solar energy’s efficiency, high cost, and queries on its reliability create the main hindrances in the customers’ point of view. This issue can be resolved when the efficiency of this system in a particular area can be found out before the actual installation of the solar water heaters; and even if the systems are installed then, such a technique is needed which may guide the customers with the correct tips for better efficiency by efficient calculations. This is done by the RETScreen software, which empowers better decision, facilitates better communication by being available in 40 different languages and reduces analysis costs dramatically among many other benefits. The prime objective of this research work is to implement the RETScreen software in the solar water heating system prior to its actual installation, so as to get the picturesque of whether or not to install a solar water heating system in the considered locality. And if yes, then what are the after-effects of the installation and what is the estimated payback period. A stepwise clearly guided picture of this software has not been provided in the earlier works; hence this paper gives the required guidance for the implementation of RETScreen with the data of India. Keywords: RETScreen Software, Solar Water Heating System, Batch Collector, Environmental Variables
INTRODUCTION Necessity drives the technology. Great deals of research works have been experimented on many conventional alternatives of solar energy to heating purposes [1], which can prove to be a better, environment friendly and cost-effective option. Solar thermal technologies hold quite significant promise for Bhubaneswar, India which has a high solar insolation of 4.5–5 kWh/meter2/day as per the data collected from Ministry of New and Renewable Energy, Government of India, for about 300 sunny days per year in India, as cited in the Wikipedia. Solar water heating system is a commercially viable and technologically mature product which has existed in the country for many years. But still, the market share of solar energy fails to fulfill the expectations of the people [1]. Hence, keeping in mind the way a customer thinks while buying a product, it can be said that if the cost of the solar products were a little less expensive along with the uninterrupted maintenance services offered by the companies, the willingness of people to opt for the solar products would have been more [1]. However, since this is not the case in the present, hence in this work, the implementation of RETScreen software has been shown. This software provides [2] the user with a broad range of options for assessing the technical, financial and environmental suitability for an investment in a 'clean energy’ project. It integrates a number of databases to assist the site assessor, including a global database of climatic conditions obtained from 4,700 groundbased stations and NASA's satellite data [2]. This software is available in 40 different languages, facilitating good communication with the users. This paper shall show the stepwise implementation of RETScreen software taking the conditions of India in view, with the help of flowcharts and screenshots. RETSCREEN RETScreen software was invented in 1998. According to the NASA Researcher News, RETScreen (Renewable Energy Technology Screen) is a tool developed by Natural Resources Canada (NRCan) CANMET Energy Technology Centre for renewable energy technologies analysis. The advantages of this unique software are summed up in the following points [3]: • Helps empower cleaner energy decisions. • Highlights results and impacts & outlines future directions for RETScreen. • Builds capacity. • Reduces analysis costs dramatically. • Empowers better decisions. A very interesting feature of the RETScreen software is the colour codes for different cells. There are four coloured cells namely blue, grey, yellow and white. The former 3 colours represent the input cells whereas white cells show the output directly.
ICORE 2013
OPERATION Among RETScreen4 and RETScreen plus, the former software helps in making the calculations regarding the efficiency of proposed case solar water heating system. RETScreen4 is based on Microsoft Excel spreadsheet. The following are the key worksheets to work with, in the RETScreen4 and each of them are described in the following sub-sections. Start Page The user has to input the simple data such as Project name, Project location, Prepared for and Prepared by. Now, the other data were given the inputs as ‘Heating’ as Project type, ‘Solar water heating’ as Technology, ‘English’ as Language and ‘Rs’ as the currency. The site location was given as Bhubaneswar and then the whole climate data appears on the screen. Figure 2 shows this page. Energy Model Worksheet This worksheet calculates the annual energy production for a solar water heating system based on local site conditions and system characteristics. Some data are entered by the user while others are calculated by the model. The model calculates the solar radiation, fuel consumption and fuel cost for ‘Electricity’ which would have been the source of heating water, had not the solar water heating system been present. NOTE: If the solar water heater is turned off for some days or months, then the energy demand for the corresponding period is not taken into consideration. Cost Analysis Worksheet It helps the user to take into account the initial/ investment cost standpoint and the annual/recurring cost standpoint associated with solar water heating project [4]. The user has to input the following: a. Initial costs: This includes the site investigation, resource assessment, environmental assessment, preliminary design, detailed cost estimate, GHG baseline study and monitoring plan, report preparation, project management, travel and accommodation. All these costs are calculated and provided by the user as input under the broad heading of ‘Feasibility study’ under Initial cost. Same is the case for Development, Engineering, Heating system, Balance of system and miscellaneous. The model then calculates the total initial costs. a. Annual costs: The user inputs the O&M costs which includes the parts and labour cost; while the model calculates the fuel cost–proposed case. b. Annual savings: This is calculated by the model, with the sub heading of fuel cost–base case, which is ‘electricity’ for our system. c. Periodic costs represent the recurrent costs that must be incurred at regular intervals to maintain the project in working condition. Emission Worksheet This is an optional worksheet for the user and helps the user estimate the greenhouse gas emission reduction (mitigation) potential of the proposed project. The user inputs the country in which the project is being implemented and the GHG emission factor. The rest all data are calculated by the model including fuel consumption, GHG emission in ton CO2. The annual GHG emission reduction is also calculated by the model. Financial Summary Worksheet Here the user provides the following financial parameters, for the considered location as input, such as fuel cost escalation rate, inflation rate, discount rate, project life, incentives and grant, debt ratio and some other simple data inputs. And the model then calculates the annual costs, annual savings and income, financial viability and the yearly cash flows. The model provides a graph which shows all these calculated data in graphical manner [5]. Hence, this section provides the financial indicator of the project analyzed, allows the user to visualize the stream of pre-tax, after-tax, and cumulative cash flows over the project life. The graph is shown in figure 9. A Prudent Study on Solar Water Heating System using RETScreen in India♦ 35
ICORE 2013
Sensitivity y and Risk Analysis A It helps the user to esttimate the sensitivity of im mportant fina ancial indicattors in relatioon to key tecchnical and financial parameters [4]. It includ des two main categoriess namely thee Sensitivity analysis and the Risk analysis. i screenshots of the whole RETScreen R sofftware imple ementation, Beforee we start going to the individual let us first discuss it through a flowcchart: Climate Loca ation Data
Startt Page
Input of Initial, annual and periodic costs
Energ gy Model Woorksheet
Cost Ana alysis Worksheeet
Finanncial Analysis Worksheet W
Annual saviings of system Other basic data
Cumulative cash flow diagrams d
Detail of fueel type in absennce of solar water heater FIG G. 1: FLOW CHART OF THE WORKSH HEETS OF RETSCR REEN
CASE STUD DY
FIG. 2: STARTT PAGE
FIG. 4: BASE CASE DATA OF ENERGY MODEL
FIG. 3: CLIMATE DATA
FIG. 5: SOLA AR RADIATION CA ALCULATED BY RET TSCREEN
36 ♦ A Prud dent Study on Solar Water Heating System using RETScreen in India
ICORE 20 013
FIG G. 6: PROPOSED CASE DATA OF ENERGY MODEL
FIG. 7: COST ANA ALYSIS DATA
FIG. 8:: FINANCIAL ANA ALYSIS
FIG. 9: CUMULATIIVE CASH FLOW DIIAGRAM
SOLAR WATER A HEATING G ENERGY MODEL FLOWCHA ART [5] Service hot water age (solved with stora using f-chart)
Calcullate Environm mental Variab bles
Calculate Sola ar Energy
Service hot h water without storage d using (solved utilisability y method)
Swimminng pool (calcula ate pool energy req quirements)
Other ca alculations
Calcuulate Renewablle Energy deliveered
1 SOLAR WATEER HEATING ENERGY MODEL FLOWCHART FIG. 10:
There are 3 modeels that cover the basic ap pplications co onsidered byy RETScreen [[5]: ater with storrage (calcula ated with the f-chart) 1. Seervice hot wa 2. Seervice hot wa ater without storage s (calcculated with the t utilisability method) A Prud dent Study on Solar S Water Hea ating System ussing RETScreen n in India♦ 37
ICORE 2013
3. Swimming pool (calculated by the ad-hoc method) 4. The description, calculation and implementation of Service hot water with storage are well focused in the further sections [7]. ENVIRONMENTAL VARIABLES The variables those are to be computed using the weather data are: •
Declination.
•
Solar Hour Angle.
•
Sunset Hour Angle.
•
Extra-terrestrial Radiation.
•
Clearness index.
•
Tilted irradiance.
•
Sky Temperature.
•
Cold water temperature.
•
Estimated load calculation.
F-CHART METHOD This method is used to calculate the service hot water with storage. It is the fraction of the hot water load that is provided by the solar heating system (solar fraction) [5]. f = 1.029Y–0.065X–0.245 + 0.0018 + 0.0215 Where the two dimensional groups X and Y are defined as [12]: X = [A F U T
T )/ L]
Y = [A F τα H N/ L] The terms used in the above equations can be described as [5, 12]: A is the collector area,F is the modified collector heat removal factor,U is the collector overall loss coefficient, T is the empirical reference temperature equal to 100°C, T is the monthly average ambient temperature, L is the monthly total heating load, τα is the collector’s monthly average transmittance-absorptance product, H is the monthly average daily radiation incident on the collector surface per unit area, N is the number of days in the month. PREVIOUS WORK DONE ON RETSCREEN IN INDIA The Hotel Golden Tower of Chennai, Tamil Nadu (India) has been taken into consideration for the implementation of RETScreen software, to decide whether opting for solar water heating. This hotel was completed in the year 2002 at the nearest weather location of Madras and at Latitude 13°N Longitude 80°E, consisting of a total of 70 guest rooms which include 60 rooms and 10 suites, with a demand of hot water at around 70°C for 24 hours a day for the whole year and an occupancy rate of about 90% and is estimated to consume around 3,000 liters of hot water per day. So, the ratio of storage capacity to collector area was set to 93L/m2. Had not been the use of solar water heaters possible, the hotel was planned to be installed with the conventional electric water heaters in the individual rooms. The manufacturer chosen for this task was the TATA BP SOLAR INDIA with model type TBPT 24M1. The following parameters were provided by the company: τα coefficient = 0.65, Fr UL coefficient = 4.20 (W/m2)/°C, Area of the collectors = 2.31 m2 each, Diameter of the pipe used = 38 mm. The solar water heating system project is capable of displacing the convention generation fuel mix of 50% coal and 50% large hydro, in the greenhouse gas analysis. The then cost of solar water heater was INR 5400.00. The inflation rate was projected to be 2.5% over the 20-year life of the project. The discount rate was taken as 12%. Concessional debt was available to cover up to 85% of the cost of the system at an interest rate 38 ♦ A Prudent Study on Solar Water Heating System using RETScreen in India
ICORE 2013
of 5% for a period of 5 years. The then electricity tariff was INR4.10/kWh and was expected to increase by 3% annually. The corporate tax rate payable by the hotel was 35%. After the implementation of RETScreen software to the site discussed above, it was discovered that the site was highly satisfactory on cost, environmental and technical aspects, with an average solar radiation of 5 KWh/m2 per day, when the collector surface is placed tilted at an angle of 25°C. This solar daily radiation can account for 1800 KWh/m2 when it is considered for a whole year. As per the system demand, a total of 14 number of glazed flat plate solar collectors were installed which covered an area of 32.3 m2.
FIG. 11: HOTEL, TAMIL NADU, INDIA
CONCLUSION It is quite well witnessed that renewable energy is gradually attaining higher levels of development with more people opting for renewable energy devices. A clear understanding and a detailed implementation has been provided in this paper with the use of RETScreen in solar water heating system in India-a country which is endowed with huge amount of solar radiation that can be effectively converted to energy in order to meet the daily needs of the people of the country. All steps concerned to each worksheet of the RETScreen software have been systematically shown in this paper, often rarely found in other papers in this manner. Ultimately, RETScreen in our project was found to be reducing the analysis costs dramatically and saving both time and money of the user by facilitating better decisions in the user. ACKNOWLEDGEMENT The authors would like to raise the vote of thanks to Dr. C. K. Panigrahi & KIIT University for providing a learning atmosphere to accomplish the needs of this paper & also Er. Sreedhar Madichetty for the constant support. REFERENCES [1] Panapakidis, D., Solar water heating systems study, Reliability, quantitative survey and life cycle cost method, Retrieved on October 18, 2012 from the website http://www.esru.strath.ac.uk/Documents/MSc_2001/dimitrios_panapakidis.pdf [2] Wikipedia contributors, RETScreen, Wikipedia, The Free Encyclopedia. Retrieved on February 2, 2013, from the website http://en.wikipedia.org/w/index.php?title=RETScreen&oldid=532261833 [3] Retrieved from the website http://sustainabledevelopment.un.org [4] RETScreen software online user manual, retrieved on September 26, 2012 from the internet, Retrieved from the website http://medilab.pme.duth.gr/C6/PV3.pdf [5] Jingcheng temperature control valve, Retrieved from the website http://www.jingcheng.co/content/?176.html [6] Retrieved from the website http://find-solar.org/pdfs/Textbook_SWH.pdf [7] Fantidis, J. G., et al, Financial Analysis of Solar Water Heating Systems during the Depression: Case Study of Greece, Engineering Economics, Kaunas University of Technology, Available from the website http://www.erem.ktu.lt/index.php/EE/article/view/1222
A Prudent Study on Solar Water Heating System using RETScreen in India♦ 39
ICORE 2013
[8] Lalwani, M., Kothari, D. P., Singh, M., 2012, Viability analysis by Techno-Economic aspects of Grid Interactive Solar Photovoltaic Project in India, IEEE-International Conference On Advances In Engineering, Science And Management (ICAESM-2012). [9] Singh, P., Kothari, D. P., Singh, M., Empowerment of the Consumers with Green. Energy Generation, National Seminar on “Solar Energy Empowering Rural India”,20.08. 11-21.08.11, Rajiv Gandhi Institute of Information Technology, Amethi. [10] D.P. Kothari, Rakesh Ranjan and K.C. Singhal, “Renewable Energy Sources and Emerging Technologies”, Prentice-Hall of India, 2007. Available from the website http://books.google.co.in/books?id=wIL12jArOoC&printsec=frontcover&source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false [11] D.P. Kothari and D.K. Sharma (Eds.), “Energy Engineering: Theory and Practice”, S. Chand and Co. Ltd., New Delhi, 2000. [12] Solar water heater, International Copper Promotion Council India, Retrieved on April 21, 2012 from the website http://www.copperindia.org/programs/sustainable-electrical-energy/solar-water-heater.
40 ♦ A Prudent Study on Solar Water Heating System using RETScreen in India
Sola ar Poweer: A Neew Para adigm to Comb bat Enerrgy Poverty P.Natarajjan1 and G..S. Nalini2 1
Professsor & Head, Deepartment of Coommerce, Pondiccherry Universityy, Puducherry–6 605014 2Ph.D Scholar, Depa artment of Commerce, Pondicheerry University, Puducherry–605 P 5014
Abstract—The Econom mic Poverty andd Energy Povertyy seem to be going hand in haand. It is difficuult to make a co onclusive determiination which onne drives the other. In the moderrn era, electricityy is perceived ass an indispensabble thing for the survival of human being and also a it ensures thhe basic Econom mic Developmennt. According to the World Bannk, in India 300 0 million people (24%) are nott connected to thhe national electtrical grid, and they are experiiencing frequentt disruptions. Sinnce solar p constaant and eco-frieendly source it acquires peoplee consideration over the years. Off grid solarr aids to being perennial, reach the grid unreachhed people so as a to ensure elecctricity for all. Subsidies S are wiidely acceptablee now than everr before. a the widespreead stimulus givven by the goveernment to boostt up Solar Poweer. Solar Capitall Subsidy and Innterest Subsidy are Subsidyy curtails goverrnment subsidy burden b on Keroosene, Diesel annd other fuels. Thus, T solar pow wer reduces govvernment subsidyy burden too. Thiis paper intends to address enerrgy poverty and d to swear energ gy for all by meaans of Solar Pow wer. Keeywords: Econoomic Poverty, Ennergy Poverty and Energy Accesss)
INTRODUCTTION According g to Internatiional Energyy Agency (IEA), Energy poverty p is a lack of acccess to mode ern energy services. These T services are defined as househoold access to o electricity and a clean coooking facilitie es. Modern energy seervices are crucial to huuman well-being and to o a country’ss economic d developmentt; and yet globally over o 1.3 billion people are a without access to elecctricity and 2.6 2 billion peeople are witthout clean cooking fa acilities. Morre than 95% % of these peeople are eitther in Sub-SSaharan Afriica or Develo oping Asia and 84% are in Rural Areas. Energy is a cross cutting c input that t facilitatees effective and a efficient delivery of most basic se ervices like safe drinkking water, public p lightinng, health ca are, educatio on, etc. and also enablinng better sta andards of household living whilst replacing inefficient and d polluting ke erosene lightiing. India is geographically a veryy large area and the villa ages are wid dely disperseed. A lot of population of the small villages may not be enough to economically y justify the extension off grid connection. Such areas aree perhaps beest served with w the distributed energ gy generated from locally available e resources like solar, biomass, sm mall hydro ettc. Despite people are co onnected with grid still energy poverrty remains by the wa ay of peak looad shedding g. Soolar being perennial, p connstant and ecco-friendly source it acquuires people consideratio on over the years. Offf grid solar aids to rea ach the grid unreached people so as a to ensure electricity fo or all. The Economic Poverty and Energy Povverty seem too be going hand h in hand. It is difficult to make a conclusive ation which one o drives thhe other. Thiss paper inte ends to addrress energy poverty and d to swear determina energy foor all by mea ans of Solar Power. P ENERGY ACCESS IN DEVELOPING COU UNTRIES In the mod dern era, eleectricity is peerceived as an a indispensa able thing foor the surviva al of human being and also it ensures the ba asic Economicc Developmeent. Lack of access to modern energ gy services iss a serious d . hindrance to economic and social development. 100% 50% 0% % Without Access A to Electriccity Traditional Use of Biomass
FIG. 1: STATUS OF ENERGY EXCLUSION IN N DEVELOPING COUNTRIES O
% of the peop ple do not ha ave access too electricity and also 66% of them Figuree 1 depicts inn India 25% are using traditional biomass b for cooking c purp poses. China is the only developing coountry which has 100% ple still depe end on tradittional biomass. Pakistan, Indonesia, electrification rate whhereas 29% of the peop Philippiness populace are a not havinng electricityy access of 33%, 3 27% annd 17% and d also using biomass of 64%, 55% % and 50% respectivelyy. Bangladessh has higher non electrification ratee of 54% annd 91% of people reely on traditioonal biomasss.
ICORE 2013
RENEWABLE ENERGY IN INDIA Renewable energy is imperative in our country because of lack of deposit of fossil fuels. Renewable energy accounts for 12% of the total installed power capacity in India. TABLE: 1 DEPLOYMENT OF RENEWABLE ENERGY IN INDIA Sources Wind Small Hydro Biogas Solar (PV) Biomass Waste to Power Total Source: MNRE
Cumulative Achievement up to 31.08.2013 (in MW) 19,779.15 3,711.75 2,337.43 1,968.84 1,264.80 99.08 29,161.05
Percentage 67.83% 12.73% 8.01% 6.75% 4.34% 0.34% 100%
Table 1 portrays the deployment of grid interactive renewable energy in India until 31st Aug 2013. Wind energy is the main renewable energy in India as it contributes 67.83% of the installed capacity. Following wind, small hydro has the contribution of 12.73% of the total installed capacity. Biogas, Solar (PV) and Biomass contribute to 8.01%, 6.75% and 4.34% respectively. Waste to power has minimal contribution of 0.34%. It is clear that wind energy hits the target so that preference moves towards other renewable energy. RELATIONSHIP BETWEEN ENERGY AND ECONOMIC INDICATOR Electricity consumption has direct link with economic indicator. India’s per capita consumption of electricity is one-fourth of the world’s average. TABLE: 2 IMPACT OF ELECTRICITY CONSUMPTION ON ECONOMIC INDICATOR Years Per Capita Consumption of Electricity (kWh) 2005-06 631.4 2006-07 671.9 2007-08 717.1 2008-09 733.5 2009-10 778.6 2010-11 818.8 2011-12 879.22 Source: CEA, World Bank and UNDP
GDP Per Capita (in $) 2,074.47 2,233.86 2,406.34 2,606.16 2,671.68 2,860.55 3,121.62
HDI 0.507 0.515 0.525 0.533 0.540 0.547 0.551
Table 2 describes the impact of electricity consumption on economic indicator. It is observed from the table that per capita consumption of Electricity, GDP Per Capita, and Human Development Index (HDI) move towards upward trend. Increase in per capita consumption of electricity result in high GDP Per Capita and HDI and also vice versa. TABLE 3: RELATIONSHIP BETWEEN ELECTRICITY CONSUMPTION AND ECONOMIC INDICATORS
Per Capita Consumption of Electricity
Pearson Correlation Sig. (2-tailed) N GDP Per capita (in $) Pearson Correlation Sig. (2-tailed) N HDI Pearson Correlation Sig. (2-tailed) N ** Correlation is significant at the 0.01 level (2-tailed).
Per Capita Consumption of Electricity 1 7
GDP Per Capita (in $) .993** .000 7 1 7
HDI .978** .000 7 .982** .000 7 1 7
Table 3 shows the relationship between Electricity Consumption and Economic Indicator. Correlation co-efficient between per capita consumption of electricity and GDP Per capita is.993, which indicates 99.3% positive relationship between the same. The correlation co-efficient between per capita 42 ♦ Solar Power: A New Paradigm to Combat Energy Poverty
ICORE 2013
consumption of electricity and Human Development Index (HDI) is.978, which indicates 97.8% positive relationship between the same. It is obvious that there is a significant relationship between electricity consumption and economic indicators. SOLAR POWER IN INDIAN CONTEXT India is blessed with an abundance of sunlight; and receives 4-7 KWh of solar radiation per sqm per day. It is uniquely placed to tap sunlight to meet up the rising energy requirements. It lends itself to a range of solutions, from a few watts as in a solar lantern or solar home lighting system to a few KW as in a rooftop solar power system to MW scale grid connected solar power plants. The fact is that more than 1,00,000 out of a total of 6,00,000 villages in India are still not electrified officially. According to the World Bank, in India 300 million people (24%) are not connected to the national electrical grid, and they are experiencing frequent disruptions. The opportunity cost of lack of electricity is high. In the southern part of India, many industries are struggling for the survival owing to frequent power cut. Thus, solar draws the attention of many industries to bring uninterrupted power supply. Germany, which has half the sunshine what India has, emerged as the biggest solar power country in the world with more than 30 per cent market share while India’s share is less than 1 per cent. This is because of the foresighted policy regime in Germany whereby solar power is encouraged on individual rooftops and can be sold to the local utility at pre-determined rates. The Jawaharlal Nehru National Solar Mission was launched on the 11th January, 2010 by the Prime Minister. The Mission has set the ambitious target of deploying 20,000 MW of grid connected solar power by 2022. Although Central and State governments have their own solar policy, the success lies on proper implementation only. In order to achieve the target of JNNSM, the Ministry of New and Renewable Energy (MNRE) has given the below incentives: •
MNRE would provide financial support up to 30% subsidy or Rs. 90/Wp with battery storage and Rs. 70/Wp without battery storage.
•
Interest bearing loans at 5%.
• Benchmark price for photovoltaic systems with battery back-up considered as Rs. 300. Subsidies are widely acceptable now than ever before. Capital Subsidy and Interest Subsidy are the widespread stimulus given by the government to boost up Solar Power. Solar Subsidy curtails the government subsidy burden on Kerosene, Diesel and other fuels. Thus, solar power reduces the government subsidy burden too. To facilitate MW projects new transmission lines should be placed as well as existing lines should be strengthened. The main limitation associated with solar power is heavy startup finance. To bring down the cost of solar panel, government should encourage the domestic panel manufacturers. Developers and Integrator should be paid in time to ensure the viability of solar power. BETTER PAY-OFF OF SOLAR POWER PROJECT Since solar requires high startup capital, people are reluctant to use domestic solar system. But the solar payback period indicates a positive sign and its pay-off is good. The central government provides the subsidy of 30% of the total project cost. Thus, the cost per KW solar power after deducting subsidy is about Rs.1.40 lakhs and it produces 5-6 units per day and the investment can be recovered in 12 years. In case of Industries, solar payback period is very less due to high electricity tariff. If the user has any excess production, he can sell it to the State Electricity Board at the pre determined tariff and also they are eligible to get Generation Based Incentive (GBI) from the government. Consequently, the payback period will further trim down. SUGGESTIONS The following are the suggestions recommended from this study to various stakeholders: •
To bring down the cost of solar power, the government should carryout various researches concerning solar power and also encourages domestic panel manufacturers with attractive packages. Solar Power: A New Paradigm to Combat Energy Poverty♦ 43
ICORE 2013
•
Voluntary organizations/ NGOs must come forward to induce awareness among household and Industries about solar power. Proper awareness facilitates the people to erect panel thereby drop in panel cost.
•
Solar power Policy paralysis should be addressed to ensure the viability of solar power.
•
The government should promote community based solar projects among the poor people.
•
Banks should come forward to offer collateral free credit to accelerate the growth of solar power.
CONCLUSION India is the fifth largest renewable energy rich country in the world. People who have grid connection do not have access to electricity for a long time of the day because of non-availability of power supply. To make the country self sufficient in energy sector, various alternative energy should be promoted. The available renewable energies are either seasonal or place bound. Among the non-conventional energy, solar is the perpetual source which ensures sustainable energy stock for country like India. By realizing the benefit of solar which could combat energy poverty, both central as well as state government framed their own solar power policies. The government should scrutinize the implementation process to enhance the viability of solar power. And also, the government has to propose a subsidy and tariff model exclusively for vulnerable people living in poverty. REFERENCES [1] Bharat Vasandani, Blessymol Thomas, 2012. Ministry’s Strategic Plan for Growth of Renewable Energy in India. Renewable Energy, Energetica India, May 2012. [2] Damian Miller, Chris Hope, 2000. Learning to Lend for Off-grid Solar Power: Policy Lessons from World Bank Loans to India, Indonesia, and Sri Lanka. Energy Policy, Vol. 28, Issue 2, Feb 2000, pp. 87-105. [3] Fridaus Muhammad Sukki et al, 2011. An Evaluation of the Installation of Solar Photovoltaic in Residential Houses in Malaysia: Past, Present and future. Energy Policy, Vol. 39, Issue 12, Dec 2011, pp. 7975-7987. [4] Mohd shahidan Shaari et al, 2013. Relationship between Energy Consumption and Economic Growth: Empirical Evidence for Malaysia. Business System Review, Vol. 2, Issue 1, pp. 17-28. [5] Slavica Robic et al, 2010. Understanding Energy Poverty–Case Study: Tajikistan. World Energy Council, Congress Paper. [6] Yannick Glemarec, 2012. Financing Off-grid Sustainable Energy Access for the Poor. Energy Policy, Vol. 47, June 2012, pp. 87-93.
44 ♦ Solar Power: A New Paradigm to Combat Energy Poverty
Thermal Performance Evaluation Methodology of a High Rated Solar Distilled Water System Navneet Deval1 and Abhishek Saxena2 1M.Tech 2Sr.
Scholar, Department of Mechanical Engineering, M.I.T., Moradabad–244001, India A.P., Department of Mechanical Engineering, M.I.T., Moradabad–244001, India E-mail: [email protected]
Abstract—In earlier days, a solar water heater was used for distilled the water through the process of pasteurization. But it has been found to be a slow process because of we could get only the temperature from 50°C-60°C of the water. Therefore, more time is required to get water hot (a good degree of hotness). Further studies revealed that if we use evacuated type solar water heater, then the temperature increases to 75°C-80°C. In this article a simple procedure has been shown to evaluate the performance analysis of a novel solar water distilling system integrated with low cost parabolic dish type solar cooker and a solar water heater. Keywords: Solar Energy, Solar Water Heater, Parabolic Cooker, Performance
NOMENCLATURE ηt =
Thermal efficiency
ηo = Optical efficiency
Ib = Mean insolation Ar = Area of receiver Aa = Area of aperture UL = Overall heat loss W = Electrical power input, W Tw = Cooker water temperature, oC Ta = Ambient air temperature, oC A = Surface area of cooking pot (cooker), m2 Qu = Useful energy gain, J G = Range of irradiation level a = Correlation coefficients c = Specific heat of water, J/kgK m = Mass of water, kg Twi = Absolute temperature of water, oC Twf = Absolute temperature (initial), oC I = Instantaneous solar energy, J η = Instantaneous energy efficiency of solar cooker
ψ = Instantaneous exergy efficiency of solar cooker As = Surface area of paraboloid aperture, m2 f = Focal length paraboloid concentrator, m d = Aperture diameter, m INTRODUCTION Around 1987, a new solar cooker was analyzed that allowed to be heated up from both (bottom and side walls) the sides. Different medium like oil and water was heated up. Food was also cooked up to show the promise of such type of cooker. In this cooker, cooking time was comparably short due to the use of concentrator. It was insulated and wind shielded so the thermal performance of such a cooker was also unaffected [1]. In 1991, parabolic solar cooker was particularly examined. It was tested in different form like when the pot (cooker) is without load i.e. empty and when the pot is full with water. At last test with full load was found to be quite satisfactorily [2]. In 2005, Jose carried out the testing of portable solar kitchen which reached an average power output of 175 W with an energy efficiency of 26.6% which
ICORE 2013
gives an ample a energ gy to cook a simple meal for 2 peop ple in averag ge 2 hours [3 3]. Another testing t was carried ouut of paraboolic solar cookker in 2003. The experim mental time period p was frrom 10:00 hrrs to 14:00 hrs of sola ar noon. Duriing this time period, it was w found outt that the da aily averagee temperature e of water was 333 K and the ennergy and exxergy efficieency of the so olar parabolic cooker weere in the rannge of 2.815.7% annd 0.4-1.25% % respectiveely [4]. Periood of Sun Shine is 2624 4 h/year witth a maximuum of 365 hrs/month in July and a minimum of o 103 h/moonth in December. Main solar radiatiion intensity was about 3.67 kWhh/m2 day which was sufficcient to provvide ample ennergy for sollar applicatioons. In 200 08, Abdllah, has been foound out thatt all types of solar cookeers to be ab ble to provide e sufficient temperatuure needed for f cooking. But the mostt common diffficulty is to have h an adeequate amouunt of solar energy throughout the whole proceess. For this, frequent f traccking of the sun s was requuired which ca an provide c with two axes suun tracking system was eexamined whhich was a ample sollar energy [5]. A solar cooker mechanica al system consisting two parts, one for f altitude tracking and d one for rootating arounnd vertical axes. A programming method of control with open loop wa as employed.. Continuous ttesting during g different f 8:30 hrrs to 16:30 hrs h was perfo ormed. All the experimenntal testing shhowed that time perioods in 2008 from the using of o cylindricall solar cookinng system witth two axes tracking t can give water uup to 90°C. EXPERIMEN NTAL SET-UP In order too carry out the t thermal performance p evaluation of o a high ratted solar water distilling system the investigatiion an out-dooor set-up was w installed which consistts with a sola ar water hea ater (fig 1) and a a solar parabolic cooker integ grated with copper c pipess for water circulation. The parabolic reflector with a normal apacity of around 5 literrs) was mounnted on its fo ocus and integrated with copper pipe es from the cooker (ca pre hot water w sourcee i.e. an eva acuated sola ar water hea ater. A plasstic bottle off 2 liters wa as directly connected d to cooker’s whistling cha annel through a copper pipe of 1cm m. This experimental workk has been carried ouut at M.I.T, Moradabad M 8o58/North and Longituude-78o47/East) Uttar Pradesh. To , (latitude-28 evaluate the thermal performancce evaluatioon of a high rated of distilled wa ater system (fig 2), a d in this article by using g energy ba alance equattion by whicch one can methodoloogy has beeen presented estimate the system’s performance. p .
FIG. 1: DIFFERENTT PARTS OF A SOLLAR WATER HEATEER
A coooker was moounted on the t focus of parabolic concentratorr, where thee maximum amount of radiation is projected because of the paraboolic shape of the collectoor. The key p phenomenon to use this parabolic dish is that the t incoming rays which are a parallel to t the axis of o the dish will be reflecte ed towards the focus, no matter where w on the dish, which arrives. On the other hand, evacuateed solar watter heaters are prefeerred over thhe flat plate heater beca ause evacuatted tubes aree made from m low emissivvity boronsilicate gla ass (glass witth very low iron content that t has supe erior durabiliity and heat resistance). ItI produces greater thhermal efficieency in brighht sunshine buut also produuces high effiiciency in oveercast conditiion. Water 46 ♦ Thermal Performance Evaluation Metthodology of a High H Rated Sola ar Distilled Wate er System
ICORE 2013
can be easily heated up to 60°C-70°C. This already hot water can be supplied to the solar dish. We can also use this system to make the water distill. The steam generated can be condensed and collected in a separate water bottle. So in this way the system can be utilized to do multiple jobs at a time. Evacuated solar water heater is also preferred over flat plate solar water heater to give hot water of temperature around 70°C-80°C which is not possible in the case of flat plate solar water heater because of heat loss [6]. That heat loss is minimized in evacuated solar water heater because of evacuated tubes. Concentrating type solar cooker as shown in figure can be made up of several elements like a disc which is in the shape of parabola [7]. The shape of the disc is parabolic so that the sun rays coming would be reflected towards the focus no matter where on the dish they arrive.
FIG. 2: EXPERIMENTAL SET-UP OF HIGH RATED SOLAR WATER DISTILLING UNIT
Besides this for measurement, a thermocouple meter was used for measuring the ambient and at different points in the system and water temperature, and a Solarimeter was used for measuring the solar radiation while an anemometer was used for measuring the velocity of air. THERMAL PERFORMANCE EVALUATION For thermal performance evaluation of a high rated of distilled water system, it is essential to estimate the thermal behavior of each involved systems viz; solar water heater, and parabolic solar cooker [8]. Thermal Performance of Solar Water Heater The average flow rate through the each tube can be calculated by;
m=
Qu C p (To − Ti ).ΔT 16
The Useful energy gain can be calculated by;
Qu = a1 + a2G + a3 [Twi − Ta ] The following rating factor can be used to describe the monthly performance of the system
f =
useful load − auxiliary load useful load
It defines the proportion of the useful load (energy delivered to outlet input) supplied by the solar input. Thermal Performance Evaluation Methodology of a High Rated Solar Distilled Water System ♦ 47
ICORE 2013
Thermal efficiency
useful load − auxiliary input
ηs =
solar radiation incident on collectors It specifies the efficiency of the system in terms of its ability to absorb and store the incident solar radiation and to deliver it at the required time. Thermal Performance of Solar Parabolic Cooker The overall heat loss factor can be determined by;
F 'U L =
W (Tw − Ta ). A
The thermal efficiency of the paraboloidal solar cooker can be determined by;
⎡
ηt = η o − ⎢U L ⎣
Ar Tw − Ta ⎤ ⎥ Aa Ib ⎦
The surface area can be calculated by; 3 ⎧ ⎫ 2 2 ⎛ ⎞ 8π f ⎪ ⎛ d ⎞ ⎪ 2 As = ⎨⎜ ⎜ ⎟ + 1⎟⎟ − 1⎬ m 3 ⎪⎝⎜ ⎝ 4 f ⎠ ⎪ ⎠ ⎩ ⎭ 2
The energy efficiency of solar cooker can be described by;
η=
mC (Tw f − Twi ) Δt / IA
The energy efficiency can be calculated by;
Tw ⎤ ⎡ m C ⎢ (Tw f − Twi ) − Ta ln f ⎥ / Δt Twi ⎦⎥ ⎣⎢ ψ= 4 ⎡ 1 ⎛ T ⎞ 4 ⎛ T ⎞⎤ I ⎢1 + ⎜ a ⎟ − ⎜ a ⎟ ⎥ A ⎢⎣ 3 ⎝ Ts ⎠ 3 ⎝ Ts ⎠ ⎥⎦ Thermal Performance of Solar High Rated Water Distilled Unit The thermal instantaneous efficiency of the system can be obtained by;
ηi =
q0005ew hew (Tw − Tg ) = I (t ) I (t )
The performance of a high rated solar distilled water system can be defined as the ratio of desired output to the required input. Here the desired output is the amount of distilled water, and the required input is of course the solar energy collected. Now by applying the concept as the production rate performance (PRP) of system, as; PRP =
T otal distilled water within interval tim e = T otal solar energy absorbed within interval time
∑ m0005 .Δ t ∑ I .Δ t i
48 ♦ Thermal Performance Evaluation Methodology of a High Rated Solar Distilled Water System
ICORE 2013
CONCLUSION As we all know that drinking water is one of the main necessities of every living being besides the food and shelter. Also there is an urgent requirement to use the alternate energy sources like solar energy in an efficient way so as to save traditional fuel source. So the use of solar energy on a large scale is required which is a clean source of energy, which also provides a pollution free environment. This solar unit is the simplest device to get potable/fresh distilled water from impure water using solar energy as fuel. There are many different designs of solar still system in the open literature. Researchers have modified the conventional solar still system to get a better performance, such as multi-basin, multi-slop solar still systems, and coupled with solar collector to increase the water temperature. But here we have used a solar water heater with a solar dish cooker which has become a unique combination to produce distilled water with such a high rate. This also looks like the best choice to obtain fresh drinkable water in remote area usage. In the locations where there is plenty of solar energy and where sources of brackish water are available, supplies of small amounts of fresh water can be produced at reasonable cost by such type of multipurpose arrangements (by which the food can be cooked and water can get be hot) can also be used for a large amount of distilled water, which are relatively inexpensive to build and easy to maintain. Also solar distillation is used on a small commercial scale to supply small communities in isolated areas. There is a wide scaled adoption of distilled water in hospital/dispensaries, D. C. batteries, factories, institutions, hostels and other field various energy demand authorities like NEDA, GEDA, HEREDA, and IREDA, can be installed these applications at a good subsidy and may be free of cost in villages where this setup can be made easily and tones of fuel and electricity can be saved. REFERENCES [1] A. M. A. Khalifa, M. M. A. Taha and M. Akyurt, Design, Simulation and testing of a new concentrating type solar cooker, Solar Energy, 1987, 38, 79-88. [2] S. Kumar, T.C. Kandpal, and S.C. Mullick, Experimental test procedure for determination of the optical efficiency factor of a parabolloid concentrator solar cooker, 1995, 145-151. [3] J. M. Arenas, Design development and testing of a portable parabolic solar kitchen, Renewable energy, 2007, 32, 257-266. [4] H. H. Ozturk, Experimental determination of energy and exergy efficiency of the solar parabolic-cooker, Solar Energy, 2004, 77, 67-71. [5] E. Abdallah, M. A. Soud, A. Akayleh, and S. Abdallah, Cylindrical Solar Cooker with Automatic Two Axes Sun Tracking System, Jordan Journal of Mechanical and Industrial Engineering, 2010, 4, 477-486. [6] G.L. Morrison, I. Budihardjo, and M. Behnia, Water-in-glass evacuated tube solar water heaters: Morrison, Solar Energy, 2004, 76, 135-140. [7] G. L. Morrison, and N.H. Tran, Long term performance of evacuated tubular solar water heaters in Sydney, Australia, Solar Energy, 1984, 32, 785-791. [8] J.A. Duffie and W.A. Beckman, Solar engineering of thermal processes, 4th ed. John Wiley & Sons, New York, 2012.
Thermal Performance Evaluation Methodology of a High Rated Solar Distilled Water System ♦ 49
Design Development and Performance of Solar Desalination Device for Rural Arid Areas N.M. Nahar1, A.K. Singh2, P. Sharma3 and G.R. Choudhary4 1,2,3,4Central
Arid Zone Research Institute, Jodhpur–342003 E-mail: [email protected]
Abstract—Solar desalination devices made of cement concrete hollow block, cement concrete, vermiculite cement, brick and stone masonry and plastered with cement have been designed, developed and constructed. These are basin type solar stills with absorber area 4.2 m2of each device and the bottom is painted with epoxy paint. The longer dimension of the device is in the east west direction so that it collects more solar radiation. One 3.5 mm thick clear window glass is provided over it having 20o tilt from the horizontal and two distillate channels are fixed for collection of distilled water. The performance evaluation of the devices made of cement concrete hollow block, vermiculite-cement, cement-concrete, brick and stone masonry were carried out by measuring distilled water obtained per day and average output was 1.671, 1.979, 1.90, 2.063 and 1.871 LM-2day-1 and efficiency was 24.61%, 29.54%, 28.21%, 30.25% and 28.55% respectively. The distillate output of solar desalination device is to be mixed with the available saline water in appropriate proportion to make it drinkable. In fact as much as 20 litres/day of potable water (1500 ppm TDS) can be made available in a day from raw water containing 3000 ppm TDS by a solar desalination device. Therefore these solar desalination devices can successfully used for desalination of saline water in rural arid areas for meeting requirement of potable water.
INTRODUCTION Water is a basic necessity of man along with food and air; the importance of supplying hygienic potable/fresh water can hardly be over stressed. The man has been dependent on rivers, lakes and underground water reservoirs for fresh water water requirements in domestic life, agriculture and industry. However, use of water from such sources is not always possible or desirable on account of the presence of large amount of salts and harmful organisms. The impact of many diseases afflicting mankind can be drastically reduced if fresh hygienic water is provided for drinking. There is acute shortage of drinkable water in the arid regions of India. Generally in summer season, villagers travel many miles in search of fresh water. It is observed that at least one or two family members are always busy in bringing fresh water from distant sources. The worst conditions are generated if the resources of water are not available and villagers are forced to take highly saline underground water containing nitrate and fluorides. This normally leads to cause the physical disorder of various kinds. Fortunately, India is blessed with abundant solar radiation. The arid parts of India receive maximum radiation i. e. 7600–8000 MJm-2 per annum, followed by semi arid parts, 7200–7600 MJm-2, per annum and least on hilly areas where solar radiation is still appreciable i.e. 6000 MJm-2 per annum [1]. Therefore, solar distillation seems to be a good substitute for conventional methods. The distillate output of solar still is to be mixed with the available saline water in appropriate proportion to make it drinkable. In fact as much as 20 litres/day of potable water (1500 ppm TDS) can be made available in a day from raw water containing 3000 ppm TDS by installing a solar still of capacity 10 litres/day. Solar distillation has been in practice for a long time. Solar distillation is carried out in solar still. Historical review of desalination of water was reported by Nebbia and Menozzi [2]. The basin-type solar still is in the most advanced stage of development. Several workers have investigated the effect of climatic, operational and design parameters on the performance of such a still [3]. Cooper [4,5] has proposed a computer simulation for analysing the performance of such a still. Frick [6] has also proposed a mathematical model for the still based on thermic circuits and the Sankey diagrams. Hirschmann and Roefler [7] have considered periodic insolation in estimating the effect of heat capacity on the performance of the still. The periodic transient analysis have also been presented by Baum et al. [8], Nayak et al. [9] and Sodha et al. [10]. Analysis of heat and mass transfer has been done by Tsilingiris [11]. Apart from the common basin-type solar still, several other type solar stills have also been propsed and studied viz. multiple effect stills [12,13], inclined-stepped solar stills [14], tilted, wick type and multiple wick type solar stills [15-17], solar film covered still and wiping spherical stills [18–20], solar still green house combination [21,22], indirectly heated solar stills [23,24], spray type solar still [25], air blown solar still [26], multi basin tilted type solar still [27,28]. Pasteurization of water using solar energy has been done [29–32]. DESIGN Solar desalination devices made of cement-concrete, cement hollow block, vermiculite-cement, brick and stone masonry and plastered with cement have been designed, developed and constructed (Fig. 1). These are basin type solar stills. The absorber area of each device is 4.2m2. The bottom is painted with epoxy
ICORE 2013
paint. The longer dimension of the device is in the east west direction so that it collects more solar radiation. One 3.5mm thick clear window glass is provided over it having 20o tilt from the horizontal. Two distillate channels have provided for collection of distilled water.
FIG. 1: SOLAR DESALINATION DEVICES MADE OF DIFFERENT CONSTRUCTION MATERIALS
PERFORMANCE The performance evaluation of the devices made of hollow block, vermiculite-cement, cement-concrete, brick and stone masonry have been carried out by measuring distilled water obtained per day and average output was 1.671, 1.979, 1.90, 2.063 and 1.871 LM-2day-1 and efficiency was 24.61%, 29.54%, 28.21%, 30.25% and 28.55% respectively. CONCLUSION Solar desalination device is very much useful in rural arid areas where potable water is not available but saline water is available. The device can provide 8 to 10litres of distilled water per day on clear sunny days. The solar still can successfully used for desalination of saline water in rural areas for meeting requirement of potable water. The distillate output of solar still is to be mixed with the available saline water in appropriate proportion to make it drinkable. In fact as much as 20litres/day of potable water (1500 ppm TDS) can be made available in a day from raw water containing 3000 ppm TDS by improved solar still. The use of solar desalination device would help in conservation of conventional fuels, such as firewood, cow dung cake and agricultural waste in rural areas of India. Conservation of firewood help in preserving the ecosystems and cow dung cake could be used as fertiliser, which could aid in the increase of production of agricultural products. Moreover, the use of this device would result in the reduction of the release of CO2 to the environment and getting CER under CDM mechanism of UNFCCC. ACKNOWLEDGEMENT The authors are grateful to the Director, Central Arid Zone Research Institute, Jodhpur and the Head, Division of Agricultural Engineering for Arid Production System for providing necessary facilities and constant encouragement for the present study. REFERENCES [1] IMD (1985). Solar Radiation Atlas of India. India Meteorological Department, New Delhi, India.1985. [2] Nebbia, G. and Menozzi, G. (1966). A short history of water desalination. Proc. International Symposium, Milano, April 1966, pp. 10–12. [3] Garg, H.P. and Mann H.S. 91976). Effect of climatic, operational and design parameters on the year round performance on single sloped and double sloped solar still under Indian and arid zone conditions. Solar Energy 18, 159–163. [4] Cooper, P.I. (1969). The absorption of radiation in solar stills. Solar Energy 12, 313–331. [5] Cooper, P.I. (1973). Digital simulation of experimental solar still data. Solar Energy 14, 451–457. [6] Frick, G.P. (1970). Some new considerations about solar stills. ISES Conference, Melbourne, paper 5/78. Design Development and Performance of Solar Desalination Device for Rural Arid Areas ♦ 51
ICORE 2013
[7] Hirschmann, J.R. and Roefler, S.K. (1970) Thermal inertia of solar stills and its influence on performance. Proc. Proc ISES Melbourne, p. 402. [8] Baum, V.A., Bayaramov, R.B. Annaev, M. and Rybakova, L.E. (1970) Thermal inertia of solar stills and its influence on performance. Proc. Proc ISES Melbourne, p. 420. [9] Nayak, J.K., Tiwari, G.N., and Sodha, M.S. (1980). Periodic theory of solar still Int. J. of Energy Research. 4, 41. [10] Sodha, M.S., Kumar, A. Singh, U. and Tiwari, G.N. (1980). Transient analysis of solar still. Energy Conversion 20, 91. [11] Tsilingiris, P.T. (2009).. Analysis of the heat and mass transfer processes in solar stills–The validation of a model. Solar Energy 83, 420–431. [12] Ottara, F. (1972). Saline water conversion and its stage of development in Spain, Publication of J.E. N. Madrid. [13] Bartali, E. A. (1976). Chimney and heated head solar still, Heliotechnique and Development, 11, 431. [14] Akhtamov, R.A., Achilov, B.M., Kamilov, O.S. and Kakharov, S. (1978). Study of regenerative inclined stepped solar still. Applied Solar Energy (Geliotekhnika). 14, 51. [15] Frick, G. and Sommerfield, J.V. (1973). Solar stills of inclined evaporating cloth. Solar Energy 14, 427–431. [16] Moustafa, S. M. A. and Brusewitz, G.H. (1979). Direct use of solar energy for water desalination. Solar Energy 22, 141. [17] Sodha, M.S., Kumar, A., Tiwari, G.N., and Tyagi, R.C. (1981). Simple multiple–wick solar still: analysis and performance. Solar Energy 26, 127. [18] Umarov, G.Y., Achilov, B.M., Djurayev, T.D. and Akhtamov, R. (1972). Water distillation by solar energy and winter cold. Applied Solar Energy (Geliotekhnika) 8, 62. [19] Menguy, G., Chassagne, G., Sfeir, A. and Saab, J. (1976). Experimental study and optimization of a solar still. Revue International d’ Heliotechnique p. 46. [20] Norov, E.Z. (1977). Expermental investigations of solar stills with various film surfaces. Applied Solar Energy (Geliotekhnika) 13, 61. [21] Selcuk, M.K. (1971). Green house and solar stills combines. COMPLEs Conference, Athens, Bull. No. 22. [22] Sodha, M.S., Kumar, A., Srivastava, A. and Tiwari, G.N. (1980). Thermal performance of still-on-roof system. Energy Conversion 20, 181. [23] Malik, M. A. S. and Tran, V.V. (1973).A simplified mathematical model for predicting the nocturnal output of a solar still. Solar Energy 14, 371. [24] Sodha, M.S., Singh, U. Kumar, A. and Tiwari, G.N. (1980). Enhancement of output in a double basin solar still. Proc. NSEC, Annamalai, TN, India. [25] Kaufmann, W. (1998). Device for desalination of sea water. Solar Energy 62, 10. [26] Mink, G, Mohamed, M.A. and Karmazsin (1998). Air-blown solar still with heat recycling. Solar Energy 62, 309–317. [27] Thanvi, K.P. (1996). Development of Solar Still for production of drinking water in arid region. National Seminar on New Strategies of Water Resources Management for 21st century, Department of Civil Engineering, JNV University, Jodhpur. pp. 233–241. [28] Thanvi K.P. (1997). Comparison of solar stills in different modes. Proc. Nat. Solar energy convention-97. Anna Univ. Chennai, pp. 316–321. [29] Y. A K. K. (2006). A solar kettle thermos flask: a cost effective sustainable & renewable water pasteurization system for the developing world. Proceedings of Solar cooking and Food Preservation International Conference, Organised by Solar Cooker International and Terra Foundation at Granada, Spain during July 12–16, 2006. [30] Robert, M. (2006). The microbiology of solar pasteurization, with applications in East Africa. Proceedings of Solar cooking and Food Preservation International Conference, Organised by Solar Cooker International and Terra Foundation at Granada, Spain during July 12–16, 2006. [31] Ligtenberg, A. (2006). Solutions in Nepal, solar cookers, water pasteurizers, sustainable devices & integrated approaches with efficient fuel wood stoves. Proceedings of Solar cooking and Food Preservation International Conference, Organised by Solar Cooker International and Terra Foundation at Granada, Spain during July 12–16, 2006. [32] Tyroller, M. (2006). Solar steam sterlizer for rural hospital. Proceedings of Solar cooking and Food Preservation International Conference, Organised by Solar Cooker International and Terra Foundation at Granada, Spain during July 12–16, 2006.
52 ♦ Design Development and Performance of Solar Desalination Device for Rural Arid Areas
Solar and Wind Power: Potential and Opportunities in Western Rajasthan K.R. Genwa1, Sarita Boss2 and Shradhha3 1,2,3Department
of Chemistry, Jai Narain Vyas University, Jodhpur–342005, Rajasthan E-mail: [email protected]
Abstract—The Indian hot arid zone, occupying an area of 31.71 Mha, spreads over Western Rajasthan, North Gujarat, South West Haryana and Punjab, some parts of Andhra Pradesh and Karnataka State, but the major part of it (61.8%) lies in the western part of Rajasthan, covering 12 districts commonly known as “Thar Desert”. The Thar desert population is mostly rural the western Rajasthan population density is 84 person/km2 and accounts for 38% of the total states population, although the population density is Low compare to other part of the country. It is repetitively high by arid zone standers. The livestock is around 25 million; these figures make this arid zone one of the densely populated desert of the world. Which implies considerable pressure on the regions natural resources, thus management of the natural resources assume great importance. According to the report maintained by the India Meteorological Department, the highest annual global radiation is received in Rajasthan (Solar insolation ranging between 6-6.4 Kwh/m2/day in about half of Rajasthan). In Rajasthan, large areas of land are barren and sparsely populated, making these areas suitable as locations for large central power stations based on solar energy. In Rajasthan, the State Government plans to establish a Solar Energy Enterprises Zone (SEEZ) in the districts of Barmer, Jaisalmer and Jodhpur by offering a package of incentives to private investors willing to develop various solar power technologies including solar thermal, solar photovoltaics (SPV), solar chimney etc. In Rajasthan, the Indian Government is about to complete a huge new power station using hybrid systems. Rajasthan is the second largest state in country covering about 10.4% land of area. The Rajasthan has limited traditional source of energy. It has only two perennial rivers, the Chambal and Mahi, whose hydroelectric potential has fully exploited. There are not much coil, oil and mines fields. However some gas and lignite reserve. The potential for using solar energy in western Rajasthan experiences a hot and dry with a serve summer, relatively clear skies, a short monsoon period and cold winter. Wind Power in western Rajasthan (Jaisalmer, Barmer) was a good roundup, which not only outlined the wind power potential in the area but also touched upon many facets that require consideration before opting for green power. With a view of diversification activities and to protect environment degradation, Rajasthan State Mines and Minerals Limited has entered into wind power generation business in 2001. Rajasthan State Mines and Minerals Limited (RSMML) commissioned 14 x 350 KW wind energy turbine on 70mtr. high lattice towers at Badabagh area of Jaisalmer. Keywords: Wind Energy, Solar Power, Energy Scenario, Renewable Resources
INTRODUCTION The consumption of non renewable sources of energy had caused more environmental damage than any other human activity. Electricity generated from fossil fuels such as coal, crude oil has led to high concentration of harmful gases in atmosphere. This has in turn led to problems such as ozone depletion and global warming. Due to the problems associated with the use of fossil fuels, alternative source of energy have become important and relevant in today’s world. These sources such as sun and wind can never be exhausted and therefore called renewable or non conventional source of energy, they cause less emission are available locally. Solar energy is currently thought to cost about twice as much as traditional sources (coal, oil etc.) obviously, as fossil fuel reserve become depleted, their cost with rise until a point is reached where solar cells become economically viable source of energy (1). Solar energy is the largest exploitable renewable resource as more energy from sunlight strikes earth in one hour than all the energy consumed by human in entire year. Solar energy intensity varies geographically in India. Rajasthan is the largest state of India which constitutes about 10.4% geographical area. Its location and favorable geographical conditions makes it best destination for solar energy development. Rajasthan contributes 7% of total state wise estimated potential for renewable power generation in India. If the recently discovered large hydrocarbon reserves of more than 3.6 x 109 barrel oil and oil equivalent in Barmer basin not considered, there limited available traditional sources of energy such as coal. In view of above, Rajasthan faces two unique challenges in power generation from conventional sources: one is only few hydropower projects due to non availability of large rivers and secondly, transportation of coal itself contributes 50% cost of energy production (2). The energy requirement of the state can be better achieved by exploitation of renewable energy sources. The state is blessed with abundant renewable energy sources like solar energy, wind, biomass, etc. Western Rajasthan has been emerged as a potential energy zone in India including solar and wind energy.
ICORE 2013
SOLAR ENEERGY: SCENARRIO AND POTEENTIALS Western part p of Rajassthan gets hig ghest solar ra adiation in thhe country. It receives solar radiation of 6.0-7.0 kWh/m2 has h more thhan 325 sunnny days evvery year with w rainy days. Exploitation of solar energy generation will meet growing eneergy needs of o the state resulting to sustainable developmentt in power sector. Out of o a total 1100MW new project alloocations, Raja asthan receivve the maxim mum share off 873 MW (i.e., 79.36% of all india allocatioons) through competitive bidding in thhe first phasee of Jawaha arlal Nehru National Solar Missioon (JNNSM). Furthermoree, 722 reputed companies have alrready registtered their interest foor setting up p of solar poower plants in Rajasthann. Welspun Energy E Ltd launches 55 MW solar plants in Rajasthan. R It targets t to invvests 15,000 0 crore to settup 1,750 MW W of renewa able energy projects in the next thhree years. At A present, it has an operrative capaciity of around d 150MW. Thhis would go up to 400 MW by thhe end of thee current fisca al (3). The neew solar prooject, located d near Phalod di in Jodhpurr district, wass developed in three pha ases of 15, 15 and 20 0 MW. The project p was completed c in 5 months. thhe project is one o of the hig ghest plant load factor generating g plants of thhe country. Itt touched 26 % DC plant load factor, a figure muuch higher tha an those of its neighb boring plantss. The solar farm will generate g 90 million unitss of electriciity annually.. With the commissionning of the first CSP plant allotted under the JNNSM J phase first, the total energy installed capacity in i Rajasthan has reached 608.5 MW. During the period April 2013–June 2013,a tota al of 55.75 MW was commissioned in the state (4). The latest status as per Rajasthhan renewab ble energy corporation c ltd (RRECLL) is given in Fig. 1
FIG. 1: INSTALLED N CAPACITY OF SOLAR POWER O IN RAJASTHA AN (MW)
Table 1 showing the t district wise solar proojects Jodhpuur contains too be the top districts in te erms of the 7) and total installed i capacity (334M MW). number off projects (47
54 ♦ Solar and a Wind Power: Potential and Opportunities in Western Raja asthan
ICORE 2013
TABLE 1: DISTRICT WISE SOLAR PROJECTS S. No. 1 2 3 4 5 6 7 8 9 10
District
Number of Projects 47 16 20 4 2 1 1 1 1 1 94
Jodhpur Jaisalmer Bikaner Nagaur Barmer Bhilwara Churu Jaipur Jhunjhunu Sirohi Total
MW 334 144 70.7 45.0 6.0 5.0 1.0 1.0 1.0 1.0 608.7
Source: RRECL
Energy production systems such as wind farms and solar systems are mostly located in desert districts such as Jaisalmer, Jodhpur, Barmer, but the load centres are situated away from these districts. A 400 kV network with associated 2220 and 132 kV strong transmission network in Barmer, Jaisalmer, Jodhpur, Bikaner area was created. RVPN is further strengthening the infrastructure with an investment of INR 2900 crore (29000 million) for 400 kV GSS at Jodhpur and Jaisalmer solar parks for transmission lines associated with solar lines. Grid connected solar power projects in Rajasthan in Financial year 2012-13 are shown in Table 2. TABLE 2: GRID CONNECTED SOLAR POWER PROJECTS IN RAJASTHAN (FY 2012-13) S. No.
Name of Solar Power Producer
Scheme
Capacity Technology (MW)
Location
Date of Commissioning
1.
Kanoria Chemicals and Industries Ltd
REC
2.5
PV
Baowri Barasingha
Jodhpur
22.06.2012
2.
R.H. Prasad & Company PVT. LTD.
REC
0.25
PV
Sarah Kishnyat
Bikaner
18.10.2012
3.
Fonroche Raajhans Energy NSM Ph-I PVT. LTD (BATCH-II)
5
PV
Gajner
Bikaner
23.12.2012
4.
Green Infra Solar Farms LTD.
NSM Ph-I
10
PV
BAP
JODHPUR
24.12.2012
Green Infra solar Farms LTD.
NSM Ph-I
5
PV
BAP
Jodhpur
24.12.2012
5.
(BATCH-II) (BATCH-II)
6.
Hasya Enterprises (P) LTD. REC
0.1
PV
Sarah
Bikaner
29.12.2012
7
Fonroche Raajhans Energy NSM Ph-I PVT. LTD (BATCH-II)
15
PV
Gajner
Bikaner
21.1.2013
8
Welspun Solar AP Private NSM Ph-I LTD. (BATCH-II)
15
PV
SRI Mandrup Nagar Jodhpur and Rawra
22.1.2013
9
Green Infra Solar FARMS NSM Ph-I LTD (BATCH-II)
10
PV
BAP
Jodhpur
30.1.2013
10
Welspun Solar AP Private NSM Ph-I LTD. (BATCH-II)
15
PV
SRI Mandrup Nagar Jodhpur and Rawra
31.01.2013
11
Impact Solar Power (P) LTD
REC
1.5
PV
Sarah Kishnyat
Bikaner
01.02.2013
12
Kanoria Chemicals AND Industries
REC
2.5
PV
Baowri Barasingha
Jodhpur
08.02.2013 Table 2(Contd.)…
Solar and Wind Power: Potential and Opportunities in Western Rajasthan♦ 55
ICORE 2013 …Table 2(Contd.)
13
SEI Solar Power LTD.
NSM Ph-I
20
PV
Ugaras
Jodhpur
11.02.2013
15
PV
Kathauti, &
Nagaur
12.02.2013
Nagaur
13.02.2013
Jaisalmer
18.02.2013
(BATCH-II) 14
Azure Solar PVT. LTD.
NSM Ph-I (BATCH-II)
15
Azure Solar PVT. LTD.
NSM Ph-I
Barnail 20
PV
(BATCH-II) 16
Gail (India)
NSM Ph-I
LTD
(BATCH-II)
Kathauti, & Barnail
5
PV
Raghawa
17
Welspun Solar AP Private NSM Ph-I LTD (BATCH-II)
20
PV
SRI Mandrup Nagar Jodhpur and Rawra
19.02.2013
18
Mahindra Suryaprakash PVT LTD.
NSM Ph-I
20
PV
Rawra
Jodhpur
20.02.2013
Mahindra Suryaprakash PVT LTD.
NSM Ph-I
10
PV
Rawra
Jodhpur
20.02.2013
Solarfield Energy TWO PVT LTD.
NSM Ph-I
20
PV
Rawra
Jodhpur
20.02.2013
Pokaran Solaire Energy PVT LTD
NSM Ph-I
5
PV
Bawdi Barsingha
Jodhpur
24.02.2013
NVR Infrastructures AND Services PVT LTD.
NSM Ph-I
10
PV
Kishanyat
Bikaner
25.02.2013
SAI Maithili
NSM Ph-I
Power Company PVT LTD
(BATCH-II)
LEPL Projects LTD.
NSM Ph-I
19 20 21 22 23 24
(BATCH-II) (BATCH-II) (BATCH-II) (BATCH-II) (BATCH-II)
Bhudhan 10
PV
Gurha
Bikaner
26.02.2013
5
PV
Bawdi
Jodhpur
26.02.2013
(BATCH-II) 25
Jakson Power PVT. LTD.
NSM Ph-I
Barsingha 10
PV
Manchitia
jodhpur
26.02.2013
10
PV
Manchitia
jodhpur
26.02.2013
10
PV
Manchitia
Jodhpur
5 MW ON 26.02.2013
(BATCH-II) 26
Jakson Power PVT. LTD.
NSM Ph-I (BATCH-II)
27
28
Symhony Vyapaar PVT LTD
Lexicon Vanijya PVT LTD
NSM Ph-I (BATCH-II)
NSM Ph-I
&BAL. 5MW 0N 27.02.2013 10
PV
Manchitia
Jodhpur
(BATCH-II)
6 MW ON 26.02.2013 &BAL. 5 MW 0N 01.03.2013
29
Giriraj Enterprises
REC
11
PV
Bavadi
Jodhpur
21.03.2013
30
Giriraj Enterprises
REC
19
PV
Bavadi
Jodhpur
21.03.2013
31
Giriraj Enterprises
REC
3
PV
Barsingha
Jodhpur
21.03.2013
Bikaner
22.03.2013
Bavadi 32
BMD Private LTD
REC
2.5
PV
Gajner
Table 2(Contd.)…
56 ♦ Solar and Wind Power: Potential and Opportunities in Western Rajasthan
ICORE 2013 …Table 2(Contd.)
33
Lahoti Overseas Limited
REC
2
PV
Barsingha
Jodhpur
26.03.2013
Bavadi 34
D J MALPANI
REC
13
PV
Lumbania
Jodhpur
26.03.2013
35
Chartered Gold Financial Services PVT LTD
REC
0.9
PV
Barsingha
Jodhpur
26.03.2013
36
Dindayal Commodities PVT LTD
REC
1
PV
Lumbania
Jodhpur
26.03.2013
37
LEPL Projects LTD
NSM Ph-I
5
PV
Bawdi
Jodhpur
26.03.2013
Bavadi
(BATCH-II) 38
Sunborne Energy Rajasthan Solar Private LTD
NSM Ph-I
39
Sanjeev Prakashan
40
Barasingha 5
PV
Gadhna
Jodhpur
26.03.2013
REC
2
PV
DEH
Bikaner
31.03.2013
Rajasthan Patrika
REC
2
PV
Sarah Kishanyat
Bikaner
31.03.2013
41
K C (India) LTD
REC
3
PV
Sarah Kishanyat
Bikaner
31.03.2013
42
RAJ Overseas
REC
1
PV
Sarah Kishanyat
Bikaner
31.03.2013
43
Vinay Corporation PVT LTD
REC
1
PV
Lumbania
Jodhpur
31.03.2013
44
Aman House Appliance PVT LTD
REC
1
PV
Lumbania
Jodhpur
31.03.2013
Total
(BATCH-II)
354.25
Source: RRECL
Rajasthan state is in the advanced stage of preparedness for installation of grid Interactive solar power plants (5). Encouraged by new initiatives such as single window clearance, solar power producers have registered with Rajasthan Renewable Energy Corporation under renewable energy policy 2004 and now Solar Energy Policy 2011. Electrification by minihybrid PV solar and wind enrgy system for rural, remote and hilly areas in Rajasthan has been demonstrated to be feasible and could prove to be boon for poor households. In light of these attractive features and the proactive initiatives the state received the first installment of large share of 583 MW, including three projects of 100 MW each and two projects of 50 MW based on solar thermal technologies. The total allocation as on Dec 2011 amounts to 873 MW (out of 1100MW in india). In Rajasthan, 41 MW solar photovoltaics power plants and 2.5 MW solar thermal power plants are already operational (6). WIND ENERGY POTENTIALS Wind is commercially and operationally the most viable renewable energy resource and accordingly emerging as one of the largest source in renewable energy sector. Wind energy is a clean energy source that holds out the promise of meeting energy demand in the direct grid connected mode as well as stand alone and remote applications(7). As a result of scientific assessment of wind resources (8-11) throughout the country, wind power has emerzed as a viable and cost effective option for power generation. In India, power generation from wind has emerzed as one of the most successful programme in the renewable energy sector and is making meaningful contribution s to over all power requirements in some of the states. India’s long coast line and strong winds in certain parts of the country creates an opportunity to develop 48,500 MW of energy through wind turbines. Due to the presence of world class private players in wind turbine manufacturing and investments wind power generation accounts for 75 per cent of the Solar and Wind Power: Potential and Opportunities in Western Rajasthan♦ 57
ICORE 2013
total installed renewable energy capacity. This is the fastest growing sector in renewable energy development as well as offers biggest potential among all other renewable segments. The Centre for Wind Energy Technology (C-WET) has set up 1244 monitoring stations across 31 states and Union Territories and has indentified 233 potential sites that can be developed to harness wind power. Wind energy has met and often exceeded the targets set for it under both the 10th Plan (20022007) and 11th Plan (2007-2012) periods. During the 10th Plan period the target set was of 1,500 M W whereas the actual installations were 5,427 MW. Similarly during the 11th Plan period the revised target was for 9,000 MW and the actual installations were much higher at 10,260 MW. The report of the subgroup for wind power development appointed by the Ministry of New and Renewable Energy to develop the approach paper for the 12th Plan period (April 2012 to March 2017) fixed a reference target of 15,000 M W in new capacity additions, and an aspirational target of 25,000 MW. Importantly the report recommends the continuation of the Generation Based Incentive scheme during the 12th Plan period. The report also prioritized the issue of transmission, which was a weak link in the value chain until now. A joint working group of the MNRE, the Ministry of Power, the Central Electricity Authority and the Power Grid Corporation of India is looking at this issue. However, for India to reach its potential and to boost the necessary investment in renewable energy it will be essential to introduce comprehensive, stable and long-term support policies, carefully designed to ensure that they operate in harmony with existing state level mechanisms so as to avoid reducing their effectiveness (12) The Wind Energy Development Programme in Rajasthan The programme began in 1999 when the then Ministry of Non-conventional Energy Sources now known as MNRE launched a scheme of demonstration projects of 2 MW capacity wind farms. A total of three demonstration projects were sanctioned by the MNRE for Rajasthan. With the successful establishment of the first 2 MW hub height were installed (each of 250 kW capacity), the dream of generating electricity through wind power came true (13). The second demonstration project with a total capacity of 2.25 MW was established in the year 2000 at Devgarh district in Chittorgarh (now in district Pratapgarh) with 3 NEG Micon wind machines of each of 750 kW. The hub height at this wind farm was 55 metres. The third demo project was set up at Phalodi district in Jodhpur with a total capacity 2.10 MW with 6 Suzlon machines each of 350 kW capacity. Subsequent to the overwhelming success of the three demonstration wind power projects, the first commercial wind farm was established by the RRECL at Jaisalmer in 2004 with a total capacity of 25 MW. This wind farm consists of 20 machines each of 1.25 MW capacity. The second commercial wind farm with a capacity 10.2 MW, was set up at Akal of Jaisalmer district in 2006. A third commercial wind farm with a total capacity of 10.2 MW was also set up at Jaisalmer. The total, including wind farms and demonstration projects, owned by RRECL has a capacity of 51.75 MW. Electricity generated by them is being supplied demonstration project at Amarsagar, Jaisalmer district where 8 wind machines of Nordex (BHEL) at a 40 metre (14). Total wind power project commission in the period of 1998 to 2013 are summarized in Table 3. It is seen that Maxmum wind power projects commissioned in period of December 2008 to March 2013 which is indication of good wind energy potentials in Rajasthan. TABLE 3: TOTAL WIND POWER PROJECT COMMISSIONED S. No. 1. 2. 3. 4. 5. 6. 7. 8. 9.
Name of Producer
Rrec Rsmml Suzlon Enercon Rrb Nepc Veer Regen Inox Total Source: RRECL
Dec. 1998 to Nov. 2003 (in MW) 6.35 9.80 54.05 40.67 0.00 0.00 0.00 0.00 0.00 110.870
Capacity Commissioned Dec. 2003 to Nov. 2008 (in MW) 35.20 10.00 252.05 229.70 16.20 0.675 0.00 0.00 0.00 543.825
58 ♦ Solar and Wind Power: Potential and Opportunities in Western Rajasthan
Dec. 2008 to March 2013 (in MW) 0.00 0.00 1076.25 524.00 13.20 0.00 28.90 121.50 264.00 2027.850
Total (in MW) 41.550 19.800 1382.350 794.370 29.400 0.675 28.900 121.50 264.00 2682.545
ICORE 2013
CONCLUSION With the acute need for electrification and higher energy production in the country, solar and wind energy is going to provide an increasingly significant share of the renewable based capacity in coming years. Solar and wind energy technologies are currently making a significant contribution to the electrical power generation system in Rajasthan. Now Rajasthan is one of the leading state in India for the development and utilization of solar and wind energy. Based on the growth trends, the future of solar and wind energy in Rajasthan is bright. The Government policies also providing help to the investors and producers. For India’s development to be truly inclusive, each and every citizen should be guaranteed a minimum standard of access to the electricity grid. Wind energy is a key solution offering India’s citizens access to clean, affordable and indigenous energy now. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]
Brown C. E., World Energy Sources, Eds. Springer, 2002. Pandey S., A Government Report on Success in scaling up solar energy in Rajasthan. Determinants of Success for promoting Solar Energy in Rajasthan, India. RSPCB Occasional paper No. 5/ 2012. Rajasthan Solar: status Report commissioned as of 30 June 2013. Ummadisingu A. and Soni M. S., Renewable and Sustainable Energy Rev., 15, 5169, 2011. Anjana P. and Tiwari H. P., In: 4th International Conference on Electric utility deregulation and reconstructing and power Technologies, Weihai, Shandong, 2011. Gupta D. and Gambhir J., Wind Energy: A Growing Potential of India In: Proceedings of the International Congress on Renewable Energy, Chandigarh, India, December, 2010. Ramchandra T.V., Shruthi B.V., Energy Convers. Manage, 46, 1561, 2005. Fernandez E., Carolin M. M., In: Proceedings of the International Congress on Renewable Energy, Pune, India, January 2005. Ramchandra T. V., Subramanian D. K. and Joshi N. V., Renew. Energy, 10, 585, 1997. Bakshi R. IEEE Power Eng. Rev., September, 2002. India Wind Energy Outlook, November, 2012. S. Wind Energy Potentials in Rajasthan, The Indian geographical Journal, 83, 131, 2008. Mathur S., Wind Energy Development in Rajasthan, Akshay urja, 6(2), 2012.
Solar and Wind Power: Potential and Opportunities in Western Rajasthan♦ 59
Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects Sameer Pikle Area Sales Manager, Software, GE Intelligent Platforms Pvt. Ltd. Delphi, Powai, Mumbai–400076 Abstract—This paper describes the basic technology of First Principal Modeling and Empirical Modeling. Power. How Solar Thermal Power works, the types of Solar thermal collectors and the process of Concentrated Solar Thermal Power (CSP). This paper also gives details of how cost of Operational and Maintenance of Concentrated Solar Power Projects is reduced by Predictive Diagnostics Systems. Keyword: Solar Thermal Power, Concentrated Solar Power, First Principle Modeling, Empirical Modeling, Predictive Diagnostics
INTRODUCTION Solar thermal power plants generate electricity indirectly. Heat from the sun's rays is collected and used to heat a fluid. The steam produced from the heated fluid powers a generator that produces electricity. It's similar to the way fossil fuel-burning power plants work except the steam is produced by the collected heat rather than from the combustion of fossil fuels. SOLAR THERMAL SYSTEMS There are two types of solar thermal systems: passive and active. A passive system requires no equipment, like when heat builds up inside your car when it's left parked in the sun. An active system requires some way to absorb and collect solar radiation and then store it. Solar thermal power plants are active systems, and while there are a few types, there are a few basic similarities: Mirrors reflect and concentrate sunlight, and receivers collect that solar energy and convert it into heat energy. A generator can then be used to produce electricity from this heat energy. The most common type of solar thermal power plants, including those plants in California's Mojave Desert, use a parabolic trough design to collect the sun's radiation. These collectors are known as linear concentrator systems, and the largest are able to generate 80 megawatts of electricity [source: U.S. Department of Energy. They are shaped like a half-pipe you'd see used for snowboarding or skateboarding, and have linear, parabolic-shaped reflectors covered with more than 900,000 mirrors that are north-south aligned and able to pivot to follow the sun as it moves east to west during the day. Because of its shape, this type of plant can reach operating temperatures of about 750 degrees F (400 degrees C), concentrating the sun's rays at 30 to 100 times their normal intensity onto heat-transfer-fluid or water/steam filled pipes [source: Energy Information Administration. The hot fluid is used to produce steam, and the steam then spins a turbine that powers a generator to make electricity. While parabolic trough designs can run at full power as solar energy plants, they're more often used as a solar and fossil fuel hybrid, adding fossil fuel capability as backup. CONCENTRATING SOLAR POWER TECHNOLOGIES Concentrating solar power (CSP) technologies, sometimes referred to as solar thermal electric technologies, have been developed for power generation applications. Historically, the focus has been on the development of cost-effective solar technologies for large (100 MWe or greater) central power plant applications. The U.S. Department of Energy’s (DOE) Solar R&D program focuses on the development of technologies suitable for meeting the power requirements of utilities in the southwestern United States. Numerous solar technologies and variations have been proposed over the last 30 years by industry and researchers in the United States and abroad. The leading CSP candidate technologies for utility-scale applications are parabolic troughs, molten-salt power towers, parabolic dishes with Stirling engines, and concentrating photo voltaics. PARABOLIC TROUGHS Nine independent power producer (IPP) parabolic trough plants were built during the California renewable energy boom of the late 1980s, and they sell power to SCE. These plants have established an excellent operating track record for this technology. They have delivered power reliably to SCE during the summer on-peak time-of-use period. A number of technology advances have been made in recent
ICORE 2013
years that are expected to make this technology more economically competitive in future projects. Key among these advances is the development of thermal energy storage. A number of new parabolic trough projects are currently in varying stages of project development around the world, some of these will include thermal energy storage. CONCENTRATING PHOTOVOLTAICS Several vendors are currently developing concentrating photovoltaic (CPV) systems. Similar to dish/Stirling systems these systems are considered attractive because of their modular nature (25 to 50kWe units) and their potential for high solar-to-electric efficiency (>30%). These systems also do not require water for cooling. Manufactures are currently providing CPV systems, but only at a few MWe per year and they are still have limited operational experience. Costs are currently somewhere between parabolic trough and flat plate PV. It is our judgment that CPV systems could be attractive for small distributed systems (25kWe and above). It is not clear at what size the economics of a small trough plant becomes the preferred option. NREL’S RECOMMENDATION FOR CSP Based on the assessment of CSP technologies above, parabolic trough technology is considered the only large-scale (greater than 50 MWe) CSP technology that is available for application in a commerciallyfinanced power project now and in the near future (5 years). The remainder of this report thus focuses on parabolic trough technology. PARABOLIC TROUGH POWER PLANT TECHNOLOGY The current state-of-the-art in parabolic trough plant design is an outgrowth of the Luz SEGS power-plant technology. Parabolic trough power plants consist of large fields of parabolic trough collectors, a heattransfer fluid/steam generation system, a Rankine steam turbine/generator cycle, and thermal storage or fossil-fired backup systems (or both). These systems are illustrated schematically in Figure 1.
FIG. 1: SCHEMATIC FLOW DIAGRAM OF PARABOLIC TROUGH PLANT
The technology can be described as follows. The solar field is modular in nature, and it comprises many parallel rows of solar collectors aligned on a north-south horizontal axis. The linear parabolicshaped reflector in each solar collector focuses the sun’s direct beam radiation on the linear receiver at the focus of the parabola as seen in Figure 2. The collectors track the sun from east to west during the day to ensure that the sun is continuously focused on the linear receiver. A heat transfer fluid is heated to 391ºC as it circulates through the receiver and returns to a series of heat exchangers in the power block, where the fluid is used to generate highpressure superheated steam (100 bar, 371ºC). The superheated steam is then fed to a conventional Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 61
ICORE 2013
reheat steam turbine/generator to produce electricity. The spent steam from the turbine is condensed in a standard condenser and returned to the heat exchangers via condensate and feedwater pumps to be transformed back into steam. Condenser cooling is provided by mechanical draft wet cooling towers. After passing through the HTF side of the solar heat exchangers, the cooled HTF is recirculated through the solar field.
FIG. 2: VIEW OF THE 30-MWE SEGS III SOLAR FIELD OF PARABOLIC TROUGH SOLAR COLLECTORS AT KRAMER JUNCTION, CALIFORNIA. THE FIGURE SHOWS THE LARGE FIELD WITH ROWS OF PARABOLIC TROUGH COLLECTORS
HEAT COLLECTION AND EXCHANGE More energy is contained in higher frequency light based upon the formula of, where h is the Planck constant and is frequency. Metal collectors down convert higher frequency light by producing a series of Compton shifts into an abundance of lower frequency light. Glass or ceramic coatings with high transmission in the visible and UV and effective absorption in the IR (heat blocking) trap metal absorbed low frequency light from radiation loss. Convection insulation prevents mechanical losses transferred through gas. Once collected as heat, thermos containment efficiency improves significantly with increased size. Unlike Photovoltaic technologies that often degrade under concentrated light, Solar Thermal depends upon light concentration that requires a clear sky to reach suitable temperatures. Heat in a solar thermal system is guided by five basic principles: heat gain; heat transfer; heat storage; heat transport; and heat insulation.[72] Here, heat is the measure of the amount of thermal energy an object contains and is determined by the temperature, mass and specific heat of the object. Solar thermal power plants use heat exchangers that are designed for constant working conditions, to provide heat exchange. Copper heat exchangers are important in solar thermal heating and cooling systems because of copper’s high thermal conductivity, resistance to atmospheric and water corrosion, sealing and joining by soldering, and mechanical strength. Copper is used both in receivers and in primary circuits (pipes and heat exchangers for water tanks) of solar thermal water systems. Heat gain is the heat accumulated from the sun in the system. Solar thermal heat is trapped using the greenhouse effect; the greenhouse effect in this case is the ability of a reflective surface to transmit short wave radiation and reflect long wave radiation. Heat and infrared radiation (IR) are produced when short wave radiation light hits the absorber plate, which is then trapped inside the collector. Fluid, usually water, in the absorber tubes collect the trapped heat and transfer it to a heat storage vault. 62 ♦ Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects
ICORE 2013
Heat is transferred either by conduction or convection. When water is heated, kinetic energy is transferred by conduction to water molecules throughout the medium. These molecules spread their thermal energy by conduction and occupy more space than the cold slow moving molecules above them. The distribution of energy from the rising hot water to the sinking cold water contributes to the convection process. Heat is transferred from the absorber plates of the collector in the fluid by conduction. The collector fluid is circulated through the carrier pipes to the heat transfer vault. Inside the vault, heat is transferred throughout the medium through convection. Heat storage enables solar thermal plants to produce electricity during hours without sunlight. Heat is transferred to a thermal storage medium in an insulated reservoir during hours with sunlight, and is withdrawn for power generation during hours lacking sunlight. Thermal storage mediums will be discussed in a heat storage section. Rate of heat transfer is related to the conductive and convection medium as well as the temperature differences. Bodies with large temperature differences transfer heat faster than bodies with lower temperature differences. Heat transport refers to the activity in which heat from a solar collector is transported to the heat storage vault. Heat insulation is vital in both heat transport tubing as well as the storage vault. It prevents heat loss, which in turn relates to energy loss, or decrease in the efficiency of the system. Heat Storage to Stabilize Solar-electric Power Generation Heat storage allows a solar thermal plant to produce electricity at night and on overcast days. This allows the use of solar power for base load generation as well as peak power generation, with the potential of displacing both coal-and natural gas-fired power plants. Additionally, the utilization of the generator is higher which reduces cost. Heat is transferred to a thermal storage medium in an insulated reservoir during the day, and withdrawn for power generation at night. Thermal storage media include pressurized steam, concrete, a variety of phase change materials, and molten salts such as calcium, sodium and potassium nitrate. OPERATIONS AND MAINTENANCE (O&M) OF SOLAR POWER PLANTS Parabolic trough solar power plants operate similar to other large Rankine steam power plants except that they harvest their thermal energy from a large array of solar collectors. The existing plants operate when the sun shines and shut down or run on fossil backup when the sun is not available. As a result the plants start-up and shutdown on a daily or even more frequent basis. Compared to a base load plant, this introduces additional difficult service requirements for both equipment and O&M crews. The solar field is operated whenever sufficient direct normal solar radiation is available to collect net positive power. This varies due to weather, time of day, and seasonal effects due to the cosine angle effect on solar collector performance; generally, the lower limit for direct normal radiation in the plane of the collector is about 300 W/m2. Since none of the plants currently have thermal storage7, the power plant must be available and ready to operate when sufficient solar radiation exists. The operators have become very adept at keeping the plant on-line at minimum load through cloud transients to minimize turbine starts, and at starting up the power plant efficiently from cold, warm or hot turbine status. The O&M of a solar power plant is very similar to other steam power plants that cycle on a daily basis. The plants are staffed with operators 24 hours per day, using a minimal crew at night; and require typical staffing to maintain the power plant and the solar field. Although solar field maintenance requirements are unique in some respects, they utilize many of the same labor crafts as are typically present in conventional steam power plants (e.g., electricians, mechanics, welders). In addition, because the plants are off-line for a portion of each day, operations personnel can help support scheduled and preventive maintenance activities. A unique but straightforward aspect of maintaining solar power plants is the need for periodic cleaning of the solar field mirrors, at a frequency dictated by a tradeoff between performance gain and maintenance cost. Early SEGS plants suffered from a large number of solar field component failures, power plant equipment not optimized for daily cyclic operation, and operation and maintenance crews inadequately trained for the unique O&M requirements of large solar power plants. Although the later plants and operating experience has resolved many of these issues, the O&M costs at the SEGS plants have been generally higher than Luz expectations. At the Kramer Junction site, the KJC Operating Company’s O&M cost reduction study addressed many of the problems that were causing high O&M costs. Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 63
ICORE 2013
Key Asset Failures Included •
HTF pump seal failures resulting from daily thermal and operational cycling of the HTF pumps.
•
HCE failures due to bad operational practices and installation procedures.
•
Lower Mirror reflectivity due to lack of reflectivity monitoring and improper wash methods followed by mirror wash crew.
•
High replacement costs. o Lower reliability and consistency in the power generation. o Lower Pumping Parasitics. Another significant focus of the study was the development of improved O&M practices and information systems for better optimization of O&M crews. In this area, important steps were: •
An update of the solar field supervisory control computer located in the control room that controls the collectors in the solar field to improve the functionality of the system for use by operations and maintenance crews.
•
The implementation of off-the-shelf power plant computerized maintenance management software to track corrective, preventive, and predictive maintenance for the conventional power plant systems.
•
The development of special solar field maintenance management software to handle the unique corrective, preventive, and predictive maintenance requirements of large fields of solar collectors.
•
The development of special custom operator reporting software to allow improved tracking and reporting of plant operations and help optimize daily solar and fossil operation of the plants, and
•
The development of detailed O&M procedures and training programs for unique solar field equipment and solar operations. As a result of the Operating Company O&M cost reduction study and other progress made at the SEGS plants, solar plant O&M practices have evolved steadily over the last decade. Cost effectiveness has been improved through better maintenance procedures and approaches, and costs have been reduced at the same time that performance has improved. O&M costs at the SEGS III-VII plants have reduced to about 25 USD/MWh. With larger plants and utilizing many of the lessons learned at the existing plants, expectations are that O&M costs can be reduced to below 10 USD/MWh at future plants. Maintaining critical equipment to ensure high levels of reliability, availability and performance is a primary focus of Power process engineers in every plant today. Without their attention, equipment failure would be rampant. Attention requires frequent, accurate assessment of equipment operating conditions to judge whether equipment meets current production demands and minimizes operational risks of unacceptable schedule interruptions or maintenance costs. In a typical plant, making this assessment involves the collection and effective analysis of reams of data about the health of complex production system elements, like compressors, turbines, pumps and fans. The amount of data every engineer needs to analyze effectively is growing steadily. Challenging plant economics frequently dictate relentlessly increasing operational demands on ever-aging critical equipment. At the same time, staff budgets are shrinking, experienced people are retiring and new staff need experience to maintain asset health efficiently. Equipment health and performance data come from periodic and real time systems. Periodic methods for making measurements and analyzing particular equipment elements include handheld vibration spectra analysis, oil analysis and thermography and boroscope inspection. Such solutions provide a great deal of information about important equipment elements prone to functional failure, but they are timeconsuming and intermittent by nature. Reducing O& M Costs by Better Monitoring & Diagnostics The service architecture of such an O&M operator is summarized in Figure 2.
64 ♦ Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects
ICORE 2013
FIG. 3: END TO END SERVICE ARCHITECTURE FOR SOLA ASSET OPERATION
Reporting Periodic reports provide owners/lenders with the necessary visibility in their assets. A detailed and accurate report allows them to account for every unit of energy that the asset is capable of producing. This includes actual energy produced and energy lost because of each failure event along with detailed root cause analysis of the failures. One of the biggest challenges in KPI (PR, CUF etc) reporting for solar assets is proper handling of data gaps. Data gaps could occur on the account of communication failures in the data acquisition system or, sometimes, because of equipment failures (like meter and weather station). There are various ways to handle the data gaps and appropriate analysis is required to either backfill the gaps (using advanced modeling techniques) or ignore the gaps depending upon the extent of the problem. Besides the data gap problem, other uncontrolled parameters (like temperature gradients within the plant, systematic and random errors in measuring instruments, random shading of modules by stray causes) introduce uncertainties in plant performance reporting. A perfectly operating solar asset can have 3-5% uncertainty in single day performance ratio. A longer time duration (a quarter or a year) performance ratio reduces this uncertainty. Hence, this uncertainty should be kept in mind while doing performance assessment within the same plant between two time periods or between two different plants. An accurate report should reflect this error for the benefit of the owners. Maintenance based on Real Time Monitoring of Site Conditions (Also known as condition based maintenance): This is a relatively newer concept in solar CSP assets and allows better efficiencies both in terms of cost reduction and improved energy production. Some of the examples of this approach are modifying the cleaning schedule based on real time soiling losses, scheduling visual checks based on inverter heat sink temperature etc. Such an approach is possible only if the asset is equipped with a robust monitoring and data acquisition system supported by a real time smart alerting engine. For larger sized assets, this is done at the plant location as well as from a remote monitoring center whereas for smaller size assets (like rooftop installations) remote monitoring center is almost always the only way to achieve this. Forecasting Solar assets require both short term and long term forecasting. SLDCs require next day (short term) forecasting at 15 minute intervals from solar assets. As the solar industry scales up to constitute a substantial fraction of the total energy supply, the forecasting will help SLDCs match solar production with Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 65
ICORE 2013
other power plants for efficient load dispatching. With the help of an independent weather forecasting service providers an O&M operator can predict energy production based on past behavior of the solar asset with reasonable accuracy. As performance modeling methods improve short term forecasting accuracy is bound to get better. Long term forecasting is needed by the lenders/owners to correct their financial models for revaluation of the solar asset. A full service operator must be equipped with statistical modeling tools to incorporate system reliability (derived from field failures) into probabilistic production models for accurate long term forecasting. IT Infrastructure and Software Platform as the Backbone of Solar Assets Management An end-to-end solar asset management requires best in class data management architecture that is smart and scalable. A large scale solar asset (say a 25MW plant with string monitoring capabilities) generates approximately 16GB of time-series data per year. A robust software application should be capable of handling such large amount of data and allow real time troubleshooting and exception handling to detect failures quickly. An O&M operator attends to a failure incident by asking what failed, when it failed, why it failed (root-cause), what is the correction schedule, and what is the level of impact (energy loss and number of equipment failed). Recording all of the above critical information for easy traceability and developing useful business insights through analytics also requires highly customized software tools. In addition, a unified spares management, service ticket management, warranty management applications brings in a complete tool set necessary for an operator to deliver quality service at an optimized cost. To address the problems that were causing high O & M costs, we need to understand the root cause of asset failures, predict very early such failures and mitigate the issues related to these failures. Technology exists to facilitate prediction of when assets will fail, allowing engineers to target maintenance costs more effectively. Real time systems focused on this area of equipment health monitoring are frequently referred to as equipment condition monitoring (ECM) or predictive asset management (PAM) systems. This can be done by way of First principle modeling which is a statistical modeling technique to predict failure by way of alarms. The other method and a better way is by empirical modeling, intelligent alarming, predictive diagnostics and alarm prioritization. First Principle Modelling This refers to modeling based on First Law of Thermodynamics. First law of thermodynamics: Heat and work are forms of energy transfer. Energy is invariably conserved, however the internal energy of a closed system may change as heat is transferred into or out of the system or work is done on or by the system. It is a convention to say that the work that is done by the system has a positive sign and connotes a transfer of energy from the system to its surroundings, while work done on the system has a negative sign. For example, changes in molecular energy (potential energy), are generally considered to remain within the system. Similarly, the rotational and vibrational energies of polyatomic molecules remain within the system. From the above, all the energy associated with a system must be accounted for as heat, work, chemical energy etc., thus perpetual motion machines of the first kind, which would do work without using the energy resources of a system, are impossible. Empirical Modeling (EM), is a novel approach to computer-based modeling that developed from research initiated in the early 1980s by Meurig Beynon of the Department of Computer Science at the University of Warwick, England. It has many critics who think of it as a broken type of Functional Programming. Early research within the group led to the development of a new language called Eden-an Evaluator for Definitive Notations. The first implementation of Eden was by Edward Yung in 1987 and a number of contributors have been leading the development of this tool ever since. The approach of modeling offered by Empirical Modeling (or EM as it is often known) centers on the concepts of Observation, Dependency and Agency. The importance of dependency has been particularly well researched with a number of software tools being developed that exploit dependency maintenance as a native concept. 66 ♦ Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects
ICORE 2013
Most useful among modern condition monitoring methods are empirical regression methods. The empirical methods use historical measurement data for a set of variables known to represent the “inputs” and “outputs” of a system. Parametric methods make some assumptions about the probability distribution from which a data sample is taken–in contrast to nonparametric methods, which do not. The nonparametric model structure is not specified by a functional form, but is determined only by empirical data. Nonparametric methods make few assumptions about the application in question, so they are very versatile. Empirical methods on which commercial condition-monitoring software applications have been developed include parametric methods such as Neural Networks and nonparametric methods such as Principle Components Analysis and Similarity-Based Modeling (SBM). Key Technology Elements of a Good Modeling Solution It was noted above that the choices made in selection of new tools for improving plant performance can be very important to future success of the plant operations. These choices require solid understanding of the problems to be solved and the advantages and trade-offs of potential solutions. A number of elements to be considered in making selection choices were identified and separated into “technology” elements–judged for suitability to the technical challenge of the application–and “engineering” elements that determine whether a potential solution is a good fit for a particular organization. Here is a brief review of the technology challenges: Core algorithm accuracy and robustness is a fundamental requirement for analyzing complex plant systems. The goal is to provide highly sensitive ability to detect impending problems at their earliest manifestations in the data. Complex plant equipment, like compressors, turbines, pumps and fans, presents wide variation in operating conditions and in the amount of real time instrumentation to indicate operating state, performance and health. The operations environment may strain the instruments’ ability to provide high quality data. Any modeling algorithm must, therefore, provide sufficient detection to give early warning of equipment problems despite operational variation, even if some fraction (as much as 25%) of the instruments drift or fail. Core algorithm execution speed is required for analyzing complex systems at high sample rates or when a large number of assets are monitored, such as in monitoring & diagnostics (M&D) centers or decision support centers (DSC). In these environments, instrument counts can be in the tens of thousands, with sample rates as fast as five minutes. Fast execution speeds are characteristic of all of the empirical methods at some scale. But the chosen method must be fast enough for the current monitoring scale requirements, as well as those anticipated in the future. Slower methods are useful only for post-mortem analysis. Simplicity of model design is a critical requirement for engineers integrating real time monitoring and analytics into their work processes. This requirement ensures that interpretation of analytical results does not require arcane knowledge that is limited to special enclaves in an organization. Models can be designed around ideas about how particular failure modes can be detected and diagnosed. Simplicity/speed of model training comes into importance during initial model building/implementation, during model retraining following major maintenance and during model rebuilding after equipment overhauls. Identifying the data for an empirical model that best represents full operating range variation must be an intuitive, streamlined process if the methodology is going to fit into an already complex, busy process of operating a plant or DSC. The training process can likely be automated to a large extent; the resultant quality should be easily validated. Simplicity of model results means that a plant engineer can use the model results directly to reduce the complexity of lots of data for his plant systems and quickly diagnose and prioritize problems-without having to consult experts in statistical or other methods to gain the proper level of understanding. The solution requires the ability not only to detect problems early, but to provide a platform for rapid, certain convergence to the right diagnosis and prognosis. Visualization/communication of model results is another critical element of the successful methodology applied to complex plant systems. A visual representation can carry a high information density that is easily understood and compactly communicated electronically across an organization. A clear representation of all the data being modeled means that analysis and diagnostics can be accomplished more easily than if the representation is buried in analytical details that have to be massaged out of statistics and then represented in an unfamiliar format. Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 67
ICORE 2013
FIG. 4: PROBLEM DETECTION TO PROBLEM RESOLUTION
Example of an incident life cycle, showing questions the engineering team asks in the path from problem detection to problem resolution. The Best Modeling Solution SBM for Predictive Analytics GE’s Proficy Smart Signal carefully considered the foregoing requirements for a good modeling solution and chose Similarity-Based Modeling (SBM) as the technology foundation for development of a Predictive Analytics solution to a broad spectrum of real time modeling needs. Other analytical methods failed to satisfy the key technology and engineering requirements outlined above–therefore, they were rejected. Validation of this conclusion may be inherent in the fact that no other analytic method has been applied effectively to ECM on such a broad scale as SBM today. Product development has focused on providing and improving the solution requirements for application in the Power industries. Application in these industries requires demonstration of cost-effective value from the single asset level to a level inclusive of fleets of complex assets across several divisions of global companies. In these industries, benefits to users include improved understanding of asset readiness, improved maintenance costs and improved resource utilization–all leading to improved business performance in highly competitive industries. Similarity-Based Modeling (SBM) is a particular form of nonparametric regression. SBM was built as a predictive modeling solution to the need for actionable intelligence from large amounts and diverse sources of current data on equipment like compressors, turbines, pumps and fans brought together in today’s complex production systems. SBM models are quickly built from an asset’s historical data, with a structure that mimics a natural engineering design. This produces a result that is quickly and efficiently implemented, from a single asset to the largest corporate scope. Using a sample of the data collected from a complex plant system, such as a compressor, a set of “normal” operating conditions can be defined that can be used to reconstruct normal operational behavior in real time and exclude or flag abnormal behavior. The SBM model provides the essential fidelity of the natural system, and it has the advantage of (A) utilizing simple selection guidelines for reference data and (B) requiring no model parameterization. Analytic computation can be done very rapidly, so Predictive Analytics can be applied as effectively to an entire enterprise as to a single asset in a plant. Model design can follow familiar engineering principles, which facilitates interpretation of results and post-processing operations, like application of diagnostic logic. 68 ♦ Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects
ICORE 2013
Some of the practical technology benefits of SBM that are not likely to be found in other methods used for equipment monitoring include: Fast and easy set-up-Fast and easy execution • Few model design decisions required. • Models based on engineering logic, not arcane statistical concepts. • Simple guidelines for reference data selection. • Computationally expensive modeling processes done off-line and stored. • A clear estimate of normal behavior. • Works for all equipment, all operating modes. • Easy to interpret results. • Supports automated diagnostics. • Very robust to typical data problems. • Bad data does not disrupt model. • Very tolerant of multiple sensor losses. Brief Review of SBM for Predictive Analytics Similarity-Based Modeling (SBM) is a kernel-based, pattern-reconstruction technique using multidimensional interpolation that is designed to exactly fit training data. SBM produces very stable estimates by using a non-linear and nonparametric kernel (similarity operator) to compare new measurements to a set of reference states (state matrix D)–without making stringent requirements on the smoothness and statistical distribution of the data. The SBM approach measures the input vector’s closeness (similarity) to the observation vectors (states) in the D matrix to generate the estimate for that input vector. This has the effect of deriving a current value estimate from the contents of the training space, normalized to the conditions of the current observation. Weight coefficients are computed by solving a system of equations formed using selected reference data points. Several studies have shown that SBM technology outperforms other candidate technologies in detecting faults (1–6). While other nonparametric techniques can produce estimates with similar accuracy metric (measure of how closely the estimates follow the actual), SBM outperforms them in robustness (likelihood that the estimates will over-fit a fault) and “spillover” (the influence a fault in one variable has in the estimates for the other variables) SBM was designed specifically for the problems of data analysis and diagnostics encountered in reallife equipment. It has been proven through modeling of tens of thousands of assets in power generation, oil & gas, aviation, and transportation applications. It can provide accurate estimates for any number of sensors, of any type, over any load range. Even if a quarter of these sensors fails in the course of operations, SBM still can generate accurate estimates for the remaining sensors without following the faulted signals, making it a very robust methodology. Analytical methods such as PCA and Clustering do not have the same level of accuracy and robustness as modeled signals fail. Model training is an area in which SBM is strong. Training data can be assembled based on subjectmatter expertise from empirical data collected in the plant historian. Because the training data can easily be collected over the range of any independent variables of the system, the effect of these variables can easily be normalized. Automated algorithms can be applied for quickly selecting a set of model conditions that represents the full operating range of the data very well. SBM training is a non-iterative, single-pass operation that involves a single matrixmultiplication and inversion [1]. A model matrix, D, represents the entire dynamic range of the reference behavior–selected from historical data, personalized to every piece of equipment. An automated Smart Signal-proprietary vector selection method is used to build D. The selection algorithm is able to rapidly sort through tens of thousands of observation vectors to construct it. SBM, with the Smart Signal training vector selection algorithm, exhibits a very consistent modeling behavior. As shown in Figure 4, there is a clear relationship between the model training data and the model (D) structure. Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 69
ICORE 2013
As with the initial model-training process, model retraining is a quick and simple process, taking several minutes to complete. The retraining for a new operating condition involves a simple inclusion of new training vectors into the D matrix from the new operating range via the same vector selection algorithm used to create the initial model. A distinct advantage of SBM is that model designs normally reflect the most sensible structural elements of the asset being modeled. A design for an SBM model is normally as straightforward as designing a physical model for First Principles methods. The available instruments for an asset can be partitioned into sub-systems that have physical meaning to the engineer. In the example of a steam turbine, the available sensors are grouped such that oil temperatures (OT), metal temperatures (MT) and vibrations (V) are collected to form a mechanical model. Temperatures (T), pressures (P), flows (F) and valve positions (VP) are collected to form individual models of the high-pressure turbine, intermediate pressure turbine and low-pressure turbine. Because SBM was built from the ground up with the detection and analysis of complex plant systems–like compressors, turbines, pumps, fans, and heat exchangers–in mind, it produces results that can be easily interpreted using the subject-matter expertise of the equipment expert, rather than requiring the subject expertise of a statistician or vibration expert. The simplicity of the SBM model results derives from the simple model structure–this simple structure is not characteristic of statistical methods that have been adapted to real-world modeling problems. Statistical methods such as Clustering-based, Neural-Network-based or Principle Components Analysisbased model systems all can suffer from the problem of design complexity unless this is well handled by the application. Model design simplicity, the accurate and robust production of estimated normal conditions for each modeled signal and the way this leads to straightforward interpretation of model results, creates another advantage for SBM. The product of the modeling is a set of estimates that mimics the actual data under normal conditions, and it easily shows the trend and magnitude of any differences from normal conditions using a chart format familiar to any plant engineer. This leads to a visualization that complements normal engineering structure, based on familiar failure mode and failure analysis representation. In the example of a tube leak in a heat exchanger, Figure 6, the pattern of the failure easily stands out from the normal condition of the equipment. The overall simplicity of SBM models facilitates development and usage of automated expert rules logic to distinguish between normal operating condition and a faulted condition. It can be used to identify a new operating condition, facilitating automated adaptation of models. Expert rules logic also can be extended to provide fault diagnostics in important cases, like the complex plant systems listed above, where “fingerprints” of failure modes are known from development of subject-matter expertise and knowledge capture. CONCLUSION Engineers are faced with the demanding responsibility of maintaining critical equipment to have high levels of reliability, availability and performance under tight budget constraints. To avoid operating surprises, accurate assessment of equipment operating conditions is needed to judge whether production demands can be satisfied while maintenance costs are controlled. Large volumes of data about the health of complex production system elements are generally available, and the amount of data is growing steadily. Pulling together large amounts of current data from diverse sources across a plant or an enterprise to create actionable intelligence is a challenge to the organization, from the plant engineer to the CIO. This article has been about the use of digital systems for ECM to create a Predictive Analytic solution. There are both technology challenges and engineering challenges to a successful outcome. The technology challenges involve selecting a solution that has the accuracy and robustness to provide early warning of failure under all operational conditions. It must have the speed for real time application across hundreds or thousands of assets across an entire enterprise. But, technology challenges include more than success at automated detection. The technology must also facilitate diagnostics and prognostics of problems by nature of providing simple connection of equipment design, model design and interpretation of results. One additional requirement is for a visual representation of results that fosters communication and understanding. 70 ♦ Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects
ICORE 2013
In addition to these technology elements, application choice must carefully consider engineering challenges that determine whether a solution will fit normal business practices and thus provide practical results for the business. To satisfy engineering requirements, the solution must be easy to use and have reasonable training demands. This lowers the bar to acceptance when introducing something new to an organization that already is busy. Because there is a wide variety of equipment types and ages in any plant in an enterprise, the solution must be capable of managing this variation, as well as the variation of operating conditions. A single solution that is flexible in its equipment scope and adaptable to the cultures across an organization improves its ROI. It reduces the cost of implementation, maintenance and updating application capabilities. It is critical that any solution chosen must possess the ability to grow with organizational vision–even to help lead it in the case of advanced technology solutions. This requires attention to not only the technology, but also the people and processes supporting its implementation and integration into your organization. GE’s Proficy Smart Signal has carefully considered the foregoing requirements for a good modeling solution and has chosen Similarity-Based Modeling (SBM) as the technology foundation for development of a Predictive Analytic solution to a broad spectrum of real time modeling needs. Product development has focused on providing the solution requirements for application in the Power industries. Successful application in these industries requires demonstration of cost-effective value from the single asset level to a level inclusive of several divisions of global companies. In these industries, benefits to users have included improved understanding of asset readiness, improved maintenance costs and improved resource utilization. The focus leading to the selection of SBM as a technology platform is recognition of the need to improve understanding of immediate and future health of assets under complex sets of operational and organizational demands. The choice of a technology designed from the ground up to provide Predictive Analytics provides the basis for continuing innovation of software that makes your bigger vision possible. REFERENCES [1] Hines, J., Seibert, R.: Technical Review of On-Line Monitoring Techniques for Performance Assessment. 2006, 1. [2] Wegerich, S., Smart Signal Corp.: Condition Based Monitoring using Nonparametric Similarity Based Modeling. Japanese Maintenology Society Meeting 2006. [3] Hines, J., Wrest, D.: Signal Validation Using an Adaptive Neural Fuzzy Inference System. Nuclear Technology 1997,: 181–193. [4] Autoassociative Model Input Variable Selection For Process Modeling: April 26–30 2004a; Virginia Beach, VA.; 2004a. [5] On-Line Monitoring Robustness Measures and Comparison: September 27–29 2004b; Halden, Norway.; 2004b. [6] Hines, J., Usynin, A.: MSET Performance Optimization through Regularization. Nuclear Engineering and Technology 2005, 38(2):177–184. [7] Wegerich, S., Singer, R., Herzog, J., Wilks, A.: Challenges Facing Equipment Condition Monitoring Systems. Maintenance and Reliability Conference Proceedings, May 8–10, 2001.
Empirical Modeling and Predictive Diagnostics to Reduce Operational & Maintenance Cost of Solar Thermal Power Projects ♦ 71
Performance Optimization of Solar PV Assisted Systems to Meet the Standards of Zero Energy Buildings G.R.K.D. Satya Prasad1, K. Vijay Kumar Reddy2 and Ch. Saibabu3 1Associate
Professor, Dept. of EE Gandhi Institute of Engineering and Technology, Gunupur Dept. of EEE and Director of Admissions in JNTUCEK, Kakinada
3Professor,
Abstract—Zero energy buildings are becoming popular in building design. A zero energy building (ZEB) combines stateof-the-art, energy-efficient construction and appliances with commercially available renewable energy systems to benefit its owner with annual net-zero energy consumption. With its reduced energy needs and renewable energy systems, a ZEB can return as much energy as it takes from the utility on an annual basis. The key objective of this paper is to analyze how solar PV generation can be integrated in to ZEB’s. The present paper will be carried out by analyzing buildings energy requirement and how a building can become self sustained and zero neutral with the integration of solar PV systems. For this an existing building will be simulated by considering the above said parameters and they will be compared with the modified building and to show that such a system feeding a load is carried out with the application of Homer software. Based on simulation results, it has been found that these renewable energy sources would be a feasible solution for zero energy buildings. A case study has presented to show how the renewable energy systems are supported to ZEB’s with economical and environmental features. Keywords: Zero Energy Building, Homer Software, HVAC Systems, Carbon Emissions
INTRODUCTION For the past half century, new building construction has greatly increased. Building energy consumption also has grown significantly as building energy loads became larger due to the introduction of new systems of heating, air-conditioning, ventilation, and artificial lighting. Currently, energy use by buildings accounts for 40% of total primary energy produced while energy consumption by commercial buildings is 18.4% of total primary energy [1]. In commercial buildings, energy is mainly used for heating (26.1%), lighting (14.7%), and cooling (9.3%) [1]. Building heating and cooling needs are affected by internal sources (electrical lighting, building equipment, and people) and external sources (solar radiation, air temperature, and wind) of heat gain and loss through facades. Transparent parts of building envelopes, or fenestration, are especially susceptible to large heat gain and loss because they are made from highly conductive materials and exposed to the direct heat gain from solar radiation. India is gifted with Renewable Energy (RE) potential like wind, hydro, solar, tidal, geothermal energy resources etc. There is increasing social acceptance of various solar gadgets with a potential of substantial conventional power savings. Innovations in Solar photovoltaic and thermal technology have made it feasible to harness grid and off grid solar power generation projects a reality. Scientific processing & treatment of municipal waste entails power generation besides environmental benefits. The Biodegradable Agro residue and waste (Biomass) offer de-centralized power generation potential coupled with opportunity of realizing organic fertilizer. Co-generation technology by the sugar, paper, fertilizer, chemical, textile, steel, cement industry etc, offers scope for power generation [2]. The utilization of RE sources for the generation of energy results in 'zero-carbon emissions', RE projects have a tremendous potential of generating carbon credits. By properly integrating Renewable energy sources with buildings the buildings can become nearly zero energy buildings. In this paper an analysis has been given on how a building can become a Zero energy building. AN APPROACH TO ZERO ENERGY BUILDING Several design failures have been repeated in new building designs which become costly to rectify once constructed. These misconceptions boil down to many aspects like not understanding the behavior of wind, major facades facing the east-west orientation and also not knowing the physics of building materials. There is then the need to educate the building professionals and the public at large to demand such renewed knowledge to be incorporated into the designs of their house or buildings. Therefore one of the sectors that contribute to the uneconomical consumption of energy is in the building industry. The sub-sectors that feed into the construction of a building depended highly on the design of buildings and this then refers mostly to the building designers, viz., the planners, the architects and engineers.
ICORE 20 013
FIG. 1: AN APPROACHH TO ZERO ENERG GY BUILDING
FIG. 2: ROAD MAP TO O ZEB
By loooking in to all a aspects diiscussed so far f leads to a very comp plex problem m to build ze ero energy building. Fig. F 1 will gives g a betteer idea abouut how to co ome for a coonclusion to ccome for a final result before to design & sim mulate “Zeroo Energy Building” (ZEB). But, a system matic approa ach shown inn fig. 2 will a ZEBs. A complete planned p mod del will not only give the solution for ZEB it also solve the complexity about enables too solve the problem of ca arbon emissioons. Buildinng energy performance p simulation is still rarely y used in building b desig gn, commissiioning and operations. The proceess is too cosstly and too labor intensive, and it ta akes too lonng to deliver results. Its ducible due to arbitrary y decisions and assumptioons made in simulation quantitativve results arre not reprod model deffinition, and can c be trusteed only under special circcumstances. This paper p descriibes the new w methodoloogy and its relationship p to the softtware compatibility. It identifies the key stepss that are criitical to its im mplementation, and showss what part oof the method dology can be applied today. k objective of the reesearch is to t produce new, innova ative, clean and efficie ent energy The key technologiies and systeems and to give g a better solution to build b “Zero Energy E Buildings”. The pre esent work will be ca arried out byy analyzing numerous pa arameters in the building gs energy peerformance in i terms of Electrical and Thermal context. Foor this an existing buildinng will be simulated by considering the above said parameters whichh are shown in the figure and they wiill be compared with the modified buuilding and to show thhat how the specific energy consump ption can deccrease with the t adoptionn of renewab ble energy sources. This T work ca an be carrieed by doing g the simulations in Hom mer softwaree and comp pared with measured values, and to arrive at certain geneeral design ruules for such systems. Visuallize a building that is noot only enerrgy efficient, but also prroduces its oown power. Just like a typical hoome, a Zeroo Energy building (ZEB) is connected d to and ta akes energy from the lo ocal utility. However, at times, thhe ZEB makees enough power p to send some ba ack to the uutility. Annua ally, a ZEB g in a net-ze ero annual produces enough enerrgy to offset the amountt purchased from the utiility; resulting energy biill. A ZEB com mbines state--of-the-art, energy e efficiient constructtion techniques and equip pment with renewablee energy sysstems to returrn as much ennergy as it ta akes on an annual a basis. With the eve er growing concern about global warming, envvironmental pollution and d the rising coost of fossil ffuels, there iss a greater interest in using renewable energy sources to meeting m grow wing energy demand d [5]. Integratin ng with PV power p The beneffits offered by PV poweer for the ecconomical annd sustainable deploymeent of renew wable solar energy also include: easy e Access to sunlight. Because the e roof surfacce is genera ally located above the “shade linne” for trees and adjacennt structures, this surface offers o virtually unobstructted access to o available Perforrmance Optimization of Solar PV P Assisted Sys stems to Meet th he Standards off Zero Energy Buildings ♦ 73
ICORE 2013
solar energy. It is placed secure location. In general, the users of the energy generated by rooftop solar are located directly beneath the rooftop, reducing transmission and operating costs. In addition, because rooftop solar is generally located directly within the current developed electric grid, no new transmission lines or controls are necessary.The roof top arrangement is the economics of rooftop solar are becoming increasingly attractive for both commercial and residential customers. For the homeowner, the cost of rooftop solar after available federal and local incentives may generate up to a 15% return on investment [6]. Currently the rooftop solar can be effectively combined into Building Integrated Photovoltaic (BIPV) systems, offering significant material, installation, and maintenance cost savings. About HOMER Software Homer is an abbreviation of “Hybrid Optimization Model for Electric Renewables.” It is a micro power optimization model developed and regularly improved by the American National Renewable Energy Laboratory. This software helps to find the best electricity generation system configuration that is to say the appropriated technologies, the size and number of each component, also comparing costs and environmental impacts. It models both conventional and renewable energy technologies in particular solar photovoltaic and wind turbines which are the options envisaged for energy efficient technologies. Homer is able to evaluate economics and technical feasibility of the system. First, Homer simulates the working power system by calculating the hourly energy balance for a year. Hour by hour, Homer determines the electric demand of the site and the local electricity supplied by the system. Comparing these energy flows, Homer is able to estimate if the configuration is feasible that is to say if the system can satisfy the electricity requirements. Then, Homer optimizes the results. Among the possible configurations defined by the simulation, Homer retains the most cost-effective in a table ranked by Net Present Costs (NPC). Homer can realize a sensitivity analysis by modifying some inputs in a range defined by the user in order to compare different possible scenarios. DESIGN OF ZERO ENERGY BUILDING The proposed ZEB system is designed with locally generated electricity for its energy uses. The ZEB has to generate as much energy as their yearly load and they are designed to be highly energy efficient and to utilize renewable energy resources. Building Simulation Tools The development of Zero energy building integrating with renewable energy sources simulation tools for building designers and energy analysis is a complex and expensive attempt. Modern simulation programs such as TRNSYS, Energy plus, ESP-r, RETSCREEN and Homer are having their own limitations and Homer is the best suitable option for renewable energy resources integrated with building energy analysis. A case study has been analyzed in GUNUPUR, Orissa and the options for providing power to an offgrid house which can be made as zero energy building. By considering power sources such as photovoltaic’s, and connected to grid, as well as battery storage. The optimal system type graph shows that HOMER recommends the generator and the batteries throughout the sensitivity space, and some renewable power for almost all the sensitivity space. Only for very cheap diesel and very light winds does HOMER recommend the generator/battery system without renewable power. That is because the generator power is quite expensive due to short lifetime and fairly low efficiency. To know the peak load and base load of the building, a preliminary audit of EEE block of GIET, GUNUPUR has been conducted with a systematic approach. After collecting the data from various sources the calculations and analysis of load pattern and load curves of various rooms has been checked with HOMER software for accuracy. Energy Auditing and Load Survey Process For the installation of solar panel; the energy audit in the EE&EEE department in order to calculate the number of loads and its power consumption. Here we considered all types of loads such as lights (20W/40W), AC (1700/2000W), CFL (15W), fans, systems (CRTs/TFT’s) and refrigerators. T No. 1 present a chart of different loads connected to each and every room.
74 ♦ Performance Optimization of Solar PV Assisted Systems to Meet the Standards of Zero Energy Buildings
ICORE 2013
TABLE1: CONNECTED LOAD IN A BLOCK
TABLE 2: ENERGY CONSUMPTION PER HOUR IN A DAY
Monthly consumption of load for 24hr in a weekday is represented in this T No: 2. This table is prepared by taking analysis of working hour of college which is given below. EEE department working hour starts at 7am and ends at 1pm.So maximum power consumption is taking place in between this time. From 1pm to 3pm lunch break are taken. So there is nil power consumption at that time. Again the department re-opens at 3pm and continues up to 6pm. Electrical Power Production and Consumption TABLE 3: YEARLY LOAD PROFILE
TABLE 4: SOLAR AND GRID INTEGRATION RESULTS
The maximum electrical power from PV array and purchased moderate power from grid to fulfill the load profile has shown in table 3 & 4. Table 4 shows monthly average electric power production in a year. Here PV is represented in yellow color and grid is represented in gray color. Here colors represent PV and grid power consumption in kW every month. In the month of January, May, June, July and December PV is alone enough to produce power and remaining month depends upon both PV and grid. TABLE 5: PERCENTAGE OF PV & GRID Component PV Array Grid Purchases Total
Production (kWh/yr) 46,443 13,169 59,612
Fraction 78% 22% 100%
Performance Optimization of Solar PV Assisted Systems to Meet the Standards of Zero Energy Buildings ♦ 75
ICORE 2013
TABLE 6: GRID SALES Load AC Primary Load Grid Sales Total
Consumption (kWh/yr) 27,083 27,836 59,612
Fraction 49% 51% 100%
T No: 5 & 6 shows about detail the power consumption in a year. 49 percent of power is consumed by ac primary load and rest of the power is sold to grid which is produced by the PV cell. Hence a base building can be evaluated with the energy performance index (EPI). If the building can be supplied with PV power the requirement of PV is large and consequently the cost of installation is also very high. Therefore a building EPI can be decreased by optimization processes shown in fig.2. After reducing the EPI to a lower level then PV power has been integrated then the cost of the installation is also less and properly integrated with grid then there is a chance to import the power to the grid when the load is low. A building can become a zero energy building if the power exported and imported are equal. Hence calculating and optimizing the EPI plays an important role while designing the Zero energy buildings. CONCLUSION The results obtained by using HOMER software can be very realistic and gives very promising results for zero energy buildings. The environmental friendly nature of the hybrid system with the integration of buildings will lead to zero energy buildings. Communities with integrated energy systems that depend heavily on renewable energy sources are coming. The success depends on a sincere and programmatic understanding of not only integrated energy system benefits, but also, the long term planning towards zero energy buildings. The two renewable energy sources, comprising photovoltaic system, considered for ZEB with grid connected option. The simulation results show potential results for favoring Zero energy buildings. The building simulation results, found that these renewable energy sources would be a feasible solution for zero energy buildings. However, as the consideration of equipments was done optimistically for the desired building load and net profits will be higher than the results in simulation by considering carbon credits and subsidies from governments. REFERENCES [1] Climate Change Impacts on Residential and Commercial Loads in the Western U.S. Grid, Ning Lu, Senior Member, IEEE, Todd Taylor, Wei Jiang, IEEE Transactions on Power Systems, VOL. 25, NO. 1, february 2010 [2] Harnessing High-Altitude Solar Power, Guglielmo S. Aglietti, IEEE, IEEE Transactions On Energy Conversion, VOL. 24, NO. 2, JUNE 2009 [3] Techno-Economic Optimum Sizing of a Stand-Alone Solar Photovoltaic System, Mohan Kolhe, IEEE Transactions on Energy Conversion, VOL. 24, NO. 2, JUNE 2009 [4] Electricity, Resources, and Building Systems Integration at the National Renewable Energy Laboratory, David Mooney National Renewable Energy Laboratory, Golden, Colorado 978-1-4244-4241-6/09/ 2009 IEEE [5] Creating Low-Cost Energy-Management Systems for Homes Using Non-Intrusive Energy Monitoring Devices, Rebecca L. Sawyer, Jason M. Anderson, Edward Lane Foulks, John O. Troxler, and Robert W. Cox 978-1-4244-2893-9/09/ ©2009 IEEE [6] New Technologies for Rural Lighting in Developing Countries: White LEDs Adoniya Ben Sebitosi, Member, IEEE, and Pragasen Pillay, Fellow, IEEE Transactions on Energy Conversion, VOL. 23, NO. 1, MARCH 2008 [7] Control Algorithm of Fuel Cell and Batteries for Distributed Generation System, Phatiphat Thounthong, St´ephane Ra¨el, and Bernard Davat, Member, IEEE Transactions on Energy Conversion, VOL. 23, NO. 2, JUNE 2008 [8] Analysis of the Cost per Kilowatt Hour to Store Electricity, Piyasak Poonpun, Student Member, IEEE, and Ward T. Jewell, Fellow, IEEE Transactions on Energy Conversion, VOL. 23, NO. 2, JUNE 2008 [9] Parallel Operation of Battery Power Modules, Chin-Sien Moo, Member, IEEE, Kong Soon Ng, Student Member, IEEE, and Yao-Ching Hsieh, Member, IEEE Transactions on Energy Conversion, VOL. 23, NO. 2, JUNE 2008 [10] Power Management of a Stand-Alone Wind/Photovoltaic/Fuel Cell Energy System Caisheng Wang, Senior Member, IEEE, Transactions on Energy Conversion, VOL. 23, NO. 3, September 2008 [11] Simulation Model for Discharging a Lead-Acid Battery Energy Storage System for Load Leveling Igor Papiˇc, Member, IEEE Transactions on Energy Conversion, VOL. 22, NO. 1, March 2007 [12] Compressed Air Energy Storage in an Electricity System With Significant Wind Power Generation Derk J. Swider, Member, IEEE TRANSACTIONS ON ENERGY CONVERSION, VOL. 22, NO. 2, JUNE 2007
76 ♦ Performance Optimization of Solar PV Assisted Systems to Meet the Standards of Zero Energy Buildings
Experimental Studies on Biomass Pyrolysis using Microwave Radiation Lohit Kamble1, Madichetty Sreedhar2 and A. Dasgupta3 2,3KIIT
1NIT, Tirichi University, Bhubaneswar
Abstract—The concerns for energy supply and the pollution problem caused by burning fossil fuels become more pronounce, more attention has been paid to using renewable and clean fuels at present, including the use of biomass. Among alternatives of using biomass as an energy source, the thermochemical conversion appears to be a promising route, which includes pyrolysis, combustion, and gasification. Pyrolysis is not only the initial process of gasification and combustion, in which biomass goes through physical and chemical changes to produce gas, char and bio-oil when it is heated in an inert atmosphere. In this we are using microwave radiations for the pyrolysis rather than conventional heating system and also study effect of microwave power, microwave irradiation time and effect of microwave absorber on production of bio oil from bagasse. The advantage of this technique includes substantial reduction in consumption of energy, time and cost in order to produce bio-oil from biomass materials. Large biomass particle size can be used directly in microwave heating, thus saving grinding as well as moisture removal cost. The pyrolysis of bagasse (sugarcane waste) under microwave irradiation was performed with microwave absorbers glycerol and char. With biomass to microwave absorber (glycerol) ratio 1:0.75, microwave heating for 30 min, the bio-oil yield from bagasse reached maximum 19 % at power 800W. By using char as microwave absorber, microwave heating for 20 min, the bio-oil yield from bagasse reached maximum 23 % at power 640W.with. With biomass to microwave absorber (char) ratio 1:0.5. Keywords: Microwave Pyrolysis, Biomass, Microwave Absorber
INTRODUCTION The term ‘biomass’ refers to wood, woody crops, agricultural wastes, herbaceous species, wood wastes, bagasse, industrial residues, waste paper, municipal solid waste, sawdust, bio-solids, grass, waste from food processing, aquatic plants and algae animal wastes. Biomass is the name given to all of the earth’s living matter. Biomass, as the solar energy stored in chemical form in plant and animal materials, is among the most precious and versatile resources on earth. It is a rather simple term for all organic materials that originates from plants, trees, crops, and algae. The components of biomass include cellulose, hemicelluloses, lignin, extractives, lipids, proteins, simple sugars, starches, water, hydrocarbons, ash, and other compounds. The “pyrolysis” is defined as a thermal degradation in the absence of oxygen, which Converts a raw biomass into solid (char), liquid (Heavy molecular weight compounds) and gaseous products (light molecular weight gases). There are mainly two stages of pyrolysis, The first stage mainly involves dehydration, dehydrogenation, decarboxylation reactions. The second comprises of processes such as cracking (thermal or catalytic), where heavy compounds further break into gases, or char is also converted into gases such as CO, CO2, CH4 and H2. Pyrolysis process has the ability to provide three end products: gas, oil and char, which all have the potential to be refined further if required. When microwave heating is applied to a pyrolysis process different chemical profiles of the volatiles in both heating systems are obtained allowing for modification of final pyrolysis products. Conventional heating means that heat is transferred from the surface towards the center of the material by convection, conduction and radiation. microwave heating represents the transfer of electromagnetic energy to thermal energy. Because microwaves can penetrate materials and deposit energy, heat can be generated throughout the volume of the material, rather than from an external source (volumetric heating). Therefore, microwave heating is energy conversion rather than heat transfer. In microwave heating the material is at higher temperature than the surrounding area, unlike conventional heating where it is necessary that the conventional furnace cavity reach the operating temperature, to begin heating the material. The differing performance between conventional and microwave heating is also translated into differential heating rates of the material. Microwave heating reveals higher heating rates due to the fact that microwave energy is delivered directly into the material through molecular. EXPERIMENTAL Biomass and Catalyst The representative bagasse collected from Trichy Distilleries, Trichy, Tamil Nadu, India. The proximate analysis of the biomasses was performed according to the ASTM standards. The analysis results are listed in Table 1. TABLE 1: PROXIMATE ANALYSIS OF BAGASSE Moisture Content % 6.66
Ash Content % 13.33
Volatile Matter % 60.01
Fixed Carbon% 20.01
ICORE 2013
Two microwave absorbers glycerol and char is used in these experiments. Microwave absorbers are used in this experiment because it increases the absorption of microwaves resulting in high temperature distribution in biomass, thereby yielding higher percentage of bio oil and gas. Microwave Pyrolysis Process The pyrolysis of biomass was conducted in a domestic microwave oven (LG MS 2029 UW India) with the maximum output power of 800W and frequency of 2.45 GHz, with hole of size (30mm) for gas inlet and gas outlet were made on the top board of the oven. 5gm of bagasse was weighed and bagasse was dried in hot air oven for 5min to remove the moisture from bagasse, the bagasse was mixed with microwave absorber (glycerol) in different proportions. For each experiment, the ratio of biomass to microwave absorber was varied (1:0.25, 1:0.5, 1:0.75) charged into the microwave oven. Nitrogen gas (99.96% pure) at about 20ml per minute was supplied before the initiation of the experiments to ensure an inert environment. During the experimental run, nitrogen gas was continuously supplied at flow rate of about 10ml per minute to maintain the inert environment as well as to sweep the vapor out of the reactor. The nitrogen gas plays an important role in avoiding any possible explosion or hazard, thus taking care of safety issue during the experiments. Microwave power was kept constant for each batch. Then production of bio oil for time period of 5min, 10 min, 20min, 30min, and 60min was observed. The vapor generated out of the reactor was condensed into bio-oil (liquid) using water cooled condenser at temperature of about 5–8oC. Bio-oil remained in the equipment was also determined via weight difference of the equipment before and after the experiments. The bio-oil was determined from reactor chamber, connecting tube and condensing unit. The experimental set up was made up of glass materials and can be easily dismantled in order to weight them individually and reconnect back for further experimental runs. Thus total bio-oil, and solid char residue was weighted at the end of the experiments to investigate the yield. The yield of the gas was measured by water displacement technique. The experiments were repeated for various power ratings such as 480W, 640W and 800W and for each experiment, the ratio of biomass to microwave absorber (glycerol) was varied as 1:0.25, 1:0.5, and 1:0.75. The amount of bio oil produced was observed for different time periods such as 5min, 10min, 20 min, 30min and 60min. EXPERIMENTS USING CHAR AS MICROWAVE ABSORBER The microwave absorber used in this set of experiments is char. The bagasse was mixed with microwave absorber (char) in different proportions. For each experiment, the ratio of biomass to microwave absorber was varied (1:0.25, 1:0.5, 1:0.75) and charged into the microwave oven. Microwave power of was kept constant for each batch. Then production of bio oil for time periods of 5min, 10min, 20min, 30min and 60 min was observed. The vapor generated out of the reactor was condensed into bio-oil (liquid) using water cooled condenser at temperature of about 5–8 oC. Bio-oil remained in the equipment was also determined via weight difference of the equipment before and after the experiments. The bio-oil was collected from reactor chamber, connecting tube and condensing unit. The experiments were repeated for various power ratings such as 480W, 640W and 800W and for each experiment, the ratio of biomass to microwave absorber (char) was varied as 1:0.25, 1:0.5, 1:0.75 and amount of bio oil produced was observed for different time periods such as 5min, 10min, 20min, 30 min and 60min. The amount of liquid bio-oil produced (ml, g) was evaluated by the sum of the weight increased in the cooling bath and the weight of the residual liquid in the flask. The amount of gaseous products (mg, g) was determined by subtracting the weight of nitrogen from the total weight of the gas collected in the bag. If the initial amount of biomass was m0 (g), the yields of bio-oil and gas were calculated as follows: 1
100
ml,-the amount of liquid bio-oil produced (g) m0-initial amount of biomass (g)
78 ♦ Experimental Studies on Biomass Pyrolysis using Microwave Radiation
(1)
ICORE 2013
100
(2)
mg, amount of gaseous products produced (g) m0-initial amount of biomass (g) RESULTS AND DISCUSSION Experiments were carried out using glycerol as microwave absorber with biomass to microwave absorber the Production of bio oil at 5min, 10min, 20min, 30min and 60min was analyzed. By using glycerol as microwave absorber, with a biomass to absorber ratio of 1:0.25,1:0.05,1:075 and microwave power of 480W, the following results were obtained for different time periods. The entire graphs shows that at microwave power 480W the maximum yield of bio oil obtained at biomass to microwave absorber ratio of 1:0.75 and irradiation time of 30min and the maximum yield of bio oil obtained is 12%. Maximum percentage yield is observed at 1:0.75 than at 1:0.5 and 1:0.25 ratios, because as amount of absorber increases heat dissipated in biomass increases and hence production of bio oil increases. The added microwave receptor is able to absorb microwaves and produce heat, which allows the nearby particles of the materials to be heated by conventional methods (convection, conduction and radiation). Higher amount of absorber produces more heat. Therefore at higher ratio of biomass to microwave absorber, yield of bio oil is more. A similar observation was obtained by Xiaoyaet.al. (2010). By using char as microwave absorber with a biomass to absorber ratio of 1:0.25, 1:.0.5, 1:0.75 and a microwave power of 480W, the following results were obtained for different time periods. The Fig. 4 shows that at microwave power 480W, the maximum yield of bio oil is obtained at biomass to microwave absorber ratio 1:0.5 and irradiation time 20min, the maximum yield of bio oil obtained being 20%.The production of bio oil increased at a ratio of 1:0.5. Production of bio oil is highest (20%) at ratio of 1:0.5 at a power 480W. Microwave receptor is able to absorb microwaves and produce heat, which allows the nearby particles of the materials to be heated by conventional methods. Then, the removal of volatiles produces char, which will act as a microwave absorber, so that the pyrolysis process can then be sustained. This sustains higher temperature uniformity and faster heating rates, improving the pyrolysis rate of the material. At ratio of 1:0.5, the pyrolysis is sustained for long time which will help to produce more amount of bio oil. (Fig. 1)
FIG. 1: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF BIOMASS AT POWER 480W
Fig. 2 shows the plot of percentage of bio oil and gas produced at 640W and for a biomass to microwave absorber ratio of 1:0.25 1:0.5, 1:075 The graph shows that at microwave power of 640W, the maximum yield of bio oil is obtained at biomass to microwave absorber ratio of 1:0.75 and irradiation time of 30min and the maximum yield of bio oil obtained is 15%. It was found that the catalysts, apart from working as microwave absorbents to speed up heating, participate in the so-called in situupgrading of pyrolytic vapors. Therefore the amount of production of bio oil depend on the dispersal of the initial microwave receptor, first, second, third, etc. generations of microwave receptors may be found. Higher amount of initial use of microwave absorber Experimental Studies on Biomass Pyrolysis using Microwave Radiation ♦ 79
ICORE 2013
will produce more generations of receptors which help to increase production of bio oil at higher biomass to microwave pyrolysis ratio. For the microwave power of 640W the % production of bio oil increases compared to 480W at same biomass to microwave ratio, at higher power, the production of bio oil increases. At 480W of microwave power temperature of microwave oven is less, it is 200oC and as microwave power is increased to 640W, the microwave oven temperature increases up to 300oC, therefor the production of bio oil is increased at 640W even if biomass to microwave absorber ratio is same because more heat will degrade biomass more rapidly resulting in more yield. Experiments were carried out by using char as microwave absorber with biomass to microwave absorber ratio (1:0.25, 1:0.5, 1:0.75) and at microwave power of 640W. The production of bio oil at 5min, 10min, 20min, 30min and 60min was analysed. It can be seen that with char as absorber most of the bio oil formation takes place rapidly so most of volatile matter will be vaporized in 20min. The production of bio oil increased at 640W because at this power the temperature is more compared to 480W because microwave energy is concentrated in the receptor, whose rate of absorption increases with temperature, leading to an exponential increase in heating rate. At 640W the microwave radiations are absorbed more, higher temperature at 640W leads to an increase in heating rate and increase in production of bio oil. Maximum yield of bio oil obtained at 640 W power and biomass to microwave absorber ratio 1:0.25 is 20 % at irradiation time of 20min.
FIG. 2: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF BIO OIL AT POWER 6400W
From all the experiments performed using char as microwave absorber with power 640W and different biomass to microwave absorber ratio, it shows that the maximum production of bio oil was obtained at power of 640W and microwave irradiation time of 20min. The study of dielectric parameters, which influence the absorption through heat dissipation, the char have high electric loss coefficient the maximum absorption of microwaves in oven takes place is at 640W for bagasse, which leads to higher temperature thus the maximum production occurs at 640W. Fig. 3 shows the plot of percentage of bio oil and gas produced at 800W and for a biomass to microwave absorber ratio of 1:0.25, 10.5,1.75 The graph shows that at biomass to microwave absorber ratio of 1:0.25, the maximum % of bio oil produced is about 15% at an irradiation time of 30min. There is no further increase in production of bio oil after 30min because most of cellulose and hemicellulose material will vaporize and for further degradation of lignin material very high temperature is required. The production of bio oil increased by 5% atthe ratio of 1:0.25 with power 800W than at power 480W and 640W. The production of bio oil at power 800W is more because the temperature of microwave oven at this power is 450oCwhich is much higher compared to temperature at 480W and 640W. At higher temperature the degradation of biomass is more and thus production of bio oil is more, which attributed to the higher percentage of bio oil and gas from the biomass. It was assumed that as the amount of microwave absorber increases, the production of bio oil also increases at same microwave power. Therefore the amount of bio oil depends 80 ♦ Experimental Studies on Biomass Pyrolysis using Microwave Radiation
ICORE 2013
on the dispersal of the initial microwave receptor. Higher amount of initial use of microwave absorber conductivity and permittivity of sample increases and hence the strength of electric field and power dissipated in it which increase the temperature of microwave oven and because of it, the production of bio oil increases.
FIG. 3: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF BIO OIL AT POWER800W
From all the experiments performed using glycerol as microwave absorber with power ratings are 480W, 640W, 800W and different biomass to microwave absorber ratio it shows that the maximum production of bio oil was obtained at 800W and microwave irradiation time of 30min (19%). So the optimum ratio of biomass to microwave absorber is 1:0.75 and irradiation time is 30min. By using char as microwave absorber (biomass to absorber ratio) 1:0.25 1.0.5, 1.075 and a microwave power of 800W the following results were obtained for different time periods. It can be seen that the production of bio oil at power 800W and at biomass to microwave absorber ratio 1:0.25 is same as the production of bio oil at 640W.This is because most of the volatiles vaporizes at temperature which is attained at 640W, further increase in power rating will not show increase in the production of bio oil. Gas Production from Microwave Pyrolysis Production of gas from pyrolysis process was analyzed by conventional water displacement technique. Fig 4, 5 and 6 shows % gas production at different power ratings and using glycerol and char as microwave absorber. Maximum yield of gas at 480 W power using glycerol as microwave absorber and biomass to microwave absorber ratio 1:0.25 is 12 % at 30 min of irradiation time. It can be seen that for the initial period, the gas production is low and as time increases, the production of bio oil also increases.
FIG. 4: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF PYROLYSIS GAS AT POWER 480W Experimental Studies on Biomass Pyrolysis using Microwave Radiation ♦ 81
ICORE 2013
Maximum yield of bio oil obtained at 480W power using char as microwave absorber and ratio 1:0.25 is 18 % at irradiation time of 20min and maximum gas obtained is 23% at 20min irradiation time, after which there is only slight increase with time Fig. 4 For the microwave power of 640W and using glycerol as microwave absorber the % production of bio oil increased by 2% compared to 480 W at same biomass to microwave ratio, at higher power, the production of bio oil increases and gas production increases by 3%.
FIG. 5: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF PYROLYSIS GAS AT POWER 640W
Maximum yield of bio oil obtained at 640W power and and using glycerol as microwave absorber, biomass to microwave absorber ratio of 1:0.5 is 24 % at irradiation time of 20min and maximum gas obtained is 30% at 20min irradiation time. Fig 5. The entire Fig. 6 shows that at microwave power of 800W, the maximum yield of bio oil is obtained at biomass to microwave absorber ratio of 1:0.75 and irradiation time of 30min and it is 25%.
FIG. 6: EFFECT MICROWAVE ABSORBER ON PRODUCTION OF PYROLYSIS GAS AT POWER 800W 82 ♦ Experimental Studies on Biomass Pyrolysis using Microwave Radiation
ICORE 2013
The Fig. 6 shows that at microwave power 800W, the maximum yield of gas obtained at biomass to microwave absorber ratio 1:0.5 and irradiation time of 20min the maximum yield of gas obtained is 31%. CONCLUSION Experiments were carried out using glycerol and char as microwave absorber with biomass to microwave absorber ratio 1:0.25, 1:0.5, 1:0.75 and at microwave power 480W, 640W and 800W. The production of bio oil at 5min, 10min, 20min, 30min and 60min was evaluated and with this it is concluded that Maximum bio oil is 19% using glycerol as microwave absorber, and it reaches to maximum value of 23% when char is used as microwave absorber thus char proves to be a better microwave absorber due to high loss coefficient of char, which means that it dissipate more power into the biomass material and hence more production of bio oil. As microwave power increases the production of bio oil increases, the optimum production of bio oil is 19% at a power of 800W when glycerol is used as microwave absorber as compared to other two power ratings. The optimum production of bio oil is 23% at microwave power 640W. Thus more energy is consumed when glycerol is used as microwave absorber. Maximum production of bio oil using glycerol as absorber is obtained at biomass to microwave absorber ratio 1:0.75 (19%) and oil production is maximum (23%) at biomass to absorber ratio 1:0.5 using char as microwave absorber. Thus less quantity of absorber is required when char is used. Optimum production of bio oil is achieved in 30min using glycerol as absorber and production is maximum at 20min using char as absorber material thus char reduces the microwave heating time. The changes in microwave receptors can alter conductivity and permittivity of sample and hence the strength of electric field and power dissipated in it. REFERENCES [1] Arshad A.S., Ani, (2010) A.F Microwave induced pyrolysis of oil palm biomass Bio resource Technology (2010) 3388–3395. [2] Arshad A.S, Ani N.F. Heating characteristics o/biomass and carbonaceous materials under microwave radiation 2011 IEEE First Conference on Clean Energy and Technology CET. [3] Asadullah M., Rahman, M. Ali, Rahman M.S, Motin, M.A (2007) Production of bio-oil from fixed bed pyrolysis of bagasse Fuel 86 (2007) 2514–2520. [4] Basu, P. Biomass gasification and pyrolysis: practical design and theory/ ISBN 978-0-12-374988-8. [5] Duponta C, Commandre J.M, Gauthier P, Guillaume, Boissonnet A, Salvador B, Schweich. D. (2007) Biomass pyrolysis experiments in an analytical entrained flow reactor between 1073 K and 1273 K Fuel 87 (2008) 1155–1164. [6] Demirbs A.(2009) Pyrolysis Mechanisms of Biomass Materials Energy Sources, Part A, 31:1186–1193, 2009. [7] Fernandez, Yolanda, Ana Arenillas, and J. Angel. 'Microwave Heating Applied to Pyrolysis', Advances in Induction and Microwave Heating of Mineral and Organic Materials, 2011. [8] Salema, A.A. 'Microwave induced pyrolysis of oil palm biomass', Bioresource Technology, 201102. [9] DU, J. 'Fast pyrolysis of biomass for bio-oil with ionic liquid and microwave irradiation', Journal of Fuel Chemistry and Technology, 201010. [10] Salema, A.A. 'Microwave-assisted pyrolysis of oil palm shell biomass using an overhead stirrer', Journal of Analytical and Applied Pyrolysis, 201207.
Experimental Studies on Biomass Pyrolysis using Microwave Radiation ♦ 83
Role of Software Tools for Support of Renewable Energy Systems Deployment Srikant Mishra1, G.R.K.D. Satya Prasad2, Sujit Kumar Patro3 and Rati Ranjan Sabat4 1Asst.
Professor, EEE Department at Gandhi institute of engineering & Technology Professor, Dept. of EE in Gandhi Institute of Engineering and Technology, Gunupur 3Asst. Professor, Department at Gandhi institute of engineering & Technology 4Associate Professor, Dept. of EEE & HOD in Gandhi Institute of Engineering and Technology, Gunupur 2Associate
Abstract—The global requirement for sustainable energy provision will become increasingly important over the next fifty years as the environmental effects of fossil fuel use become apparent. Therefore, the issues surrounding integration of energy supplies need to be considered carefully. The focus of this paper was the development of a decision support framework that will aid the design of sustainable energy systems through simulation software. Hence there is a need of real time software which integrates with real data and different local climatic conditions. Hence the simulation results will gives the promising results and economically they have the flexibility to plan any energy system to meet the conditions of different climatic conditions with mixed energy opportunities. This paper highlights the importance of such software’s work has succeeded in defining and evaluating software tools for analyzing the feasibility of integrated energy systems and also the universities curriculum to be updated by introducing such software at UG level. This will allow the future technocrats informed decisions to be made about the technical feasibility of supply mix and control strategies, to plan any energy system with full pledged analysis.
INTRODUCTION The use of energy efficient and renewable energy technologies (RETs) has increased greatly over the past several decades. The applicability of clean energy technologies has been identified almost in every sector at small, medium and large scales. Additionally, many renewable energy technologies are also suited to rural and remote areas, where energy is often crucial in human development. It was seen that for many renewable energy technologies, worldwide capacity grew at rates of 10–60% annually. All forms of energy are expensive, but as time progresses, renewable energy generally gets cheaper and on the other hand fossil fuels generally get more expensive. Setting up a project based on renewable energy or incorporation energy conservation measures at existing site certainly depends on the viability of the project. In order to benefit from these technologies, potential users, decision and policy makers, planners, project financiers, and equipment vendors must be able to quickly and easily assess whether a proposed clean energy technology project makes sense[1]. Some time it takes lengthier time span for manual checking the viability of the project and also it is not a simple process. Therefore, a number of software tools have been developed and tested for the above purpose. In recent years, renewable energy resources have become significant contributors to energy usage among both developed and developing countries. New textbooks dealing with alternative and renewable energy resources have been published recently. Many universities have also started offering courses on renewable and alternative energy sources at both undergraduate and graduate level. Simulation software’s and analysis tools are very useful to analyze the practical usage of renewable energy sources. Some of the software’s using in college level is not met with the requirements and compatibility of renewable energy sources. Renewable energy based software’s requires innovative clean energy project analysis, modeling with dynamic weather conditions and geographical conditions. These tools are key enablers to assess various types of renewable energy and energy efficient technologies to meet clean air and greenhouse gas emissions objectives [2]. With the help of those software tools we can increase the ability to optimize integrated energy efficient design in domestic and international markets, reduce operating costs, comply with code requirements, and qualify for funding and incentive programs. Presently plenty of software’s are available to analyze economical feasibility of renewable energy sources. Once economical feasibility of the renewable energy sources known, then the implementation of these sources can be implemented effectively. In this paper the complete information on these software’s availability and their importance will be analyzed. At the end one of the software called RETSCREEN will be analyzed along with all its simulation features. AVAILABLE RENEWABLE ENERGY SIMULATION SOFTWARE’S Modern energy systems are characterised by increasingly complex interactions between energy supply, distributors, and consumer demand. High quality analytical data and tools1 are vital for predicting and understanding these interactions in order to enable informed energy system design, implementation and operation decisions. Effective energy analytics to inform public policy requires that three key conditions be
ICORE 2013
met: (1) validated models must be available and appropriate for the target environment; (2) suitable data must be available for input into the model and for verifying model-based results; and (3) models must be operated by people trained in the use of the tools and in interpreting the outcomes for local conditions. Developing these resources can be expensive and time consuming. Even when data and tools are intended for public re-use they often come with technical, legal, economic and social barriers that make them difficult to adopt, adapt and combine for use in new contexts [3]. In this paper, the main focus is on how publically accessible software and data (i.e., open energy data, open source energy tools and open educational resources) can help the design engineers meet the above conditions. Energy is central to improved social and economic well-being, and is indispensable to most industrial and commercial wealth generation [3]. Adequate national capacity to track progress towards universal modern energy access represents a crucial element of energy poverty alleviation and sustainable development strategies. For developing countries-which frequently lack established infrastructure, data, and software tools-there are significant potential benefits from rigorous analyses enabled by open source tools and data. A reliable and comprehensive energy information base is required to set targets and monitor outcomes, to design strategies and policies, to make evidence-based decisions as well as to enable citizens to make informed choices. In addition, poor quality statistics limit multi-country analysis and undermine efforts to implement global or regional programmes. Ultimately, these open energy resources, combined with open innovation processes, can be harnessed to better inform energy decision-making and rapidly develop low-cost, high-quality and localized energy resources. Open Data Open data refers to that subset of data that is freely available to everyone to use without restrictions. Complex energy decisions often draw upon data from a wide range of economic, technical, policy and social sources. Portions of this data may be particular to a local context while other data sources may be less context-sensitive. Some of this data is available through open government initiatives, but increasingly it is coming from other institutions including non-profits, industry and research institutions [3]. The ability to easily and effectively reuse a data set can eliminate considerable redundant work often associated with assembling datasets [4]. Governmental acceptance and adoption of open data has been growing rapidly with examples ranging from all weather conditions with geographical information. As another example, the World Bank’s Open Data Initiative includes multiple platforms through which one can access and process data, including mobile ‘apps’, Application Programming Interfaces (APIs), catalogue listings of resources, data visualization tools, a knowledge repository, and development of metadata standards. The impetus for these open data government initiatives is transparency, accountability and the belief that opening this data to the public will lower the barriers to innovations that will benefit society [5]. A large amount of data on US national energy consumption has been made available online. This information is presented in a variety of forms, often requiring pre-processing before use. More structured information can be made available through web services that provide machine-accessible mechanisms for retrieving data. The concept of linked open data takes this further and aims to provide information sources that can be easily combined together using standard tools. Machine-accessible access to linked open energy data has the potential to greatly enhance the productivity of design engineers. Crowd Sourcing Crowd sourcing has proven to be an effective and efficient way to generate and maintain valued datasets, tools and educational resources. It was first introduced as a term in a Wired Magazine article (Howe, 2006) and has since been adopted by a huge range of projects, which maintain common principles. Popular examples include Wikipedia, which provides an online encyclopedia built from community volunteers, and Amazon’s Mechanical Turk, which provides outsourcing and a payment mechanism for small tasks [6]. Since quality energy data is often not available for particular local needs, Crowd sourcing can be an effective method for distributing the task among the broader community. The Role of Software Tools for Support of Renewable Energy Systems Deployment ♦ 85
ICORE 2013
application of Crowd sourcing to data collection has several benefits including the potential for reduced cost, reduced time and higher quality given that contributors are often reviewers and users of the data. Acquiring empirical observations or measurements is costly because it requires the presence of observers at the physical (or temporal) location of interest. The intention with Crowd sourcing is to minimize cost by making use of observations from community members who have access to model-relevant data that they are already collecting. Researchers have shown that Crowd sourcing can produce results comparable to those produced by experts, but that in some cases this relies on correctly formulating and structuring the requested task. Crowd sourced input might also play a role in the larger view of policy-making or determining research directions. There are a number of tools for Crowd sourcing collective decision making processes, which have been shown to outperform expert panels, particularly with regard to forecasting. EDUCATION AND CAPACITY BUILDING The complexity of energy generation and distribution systems coupled with technological advances means that energy analysis tools must be continuously adapted and extended to new contexts. These uses apply to countries at all levels of socio-economic development. In developing countries, however, access to energy services remains a critical issue for large segments of the population. This requires the significant enhancement of, and adaptation to traditional energy planning tools. The selection and sourcing of input data are key considerations for the modeling process. However, the collection and maintenance of these data sets is often a costly and time-consuming process. Data quality requirements for energy modeling can be measured along two dimensions. First, temporal fidelity is required. When contemplating electricity consumption for example, it is important to ask if the model requires measurements at high frequency resolution or whether long-term average values are acceptable. Second, one must identify the measurement accuracy required. Data with a high level of accuracy is unnecessary if the model itself represents a weak approximation. Increasing the data resolution will normally increase the cost of data collection. As such, researchers should aim for the minimum acceptable fidelity, which is especially important in developing economies where modeling resources are limited. Energy modeling software can be proprietary or Open Source. Open Source Software is now a common paradigm for software distribution and is rapidly being adopted in many sectors. The conventional form of OSS project has been developed by volunteers who contribute for reasons, which might include learning, career concerns or satisfying functional needs, covering a wide range of platforms and programming languages. TABLE 1: RENEWABLE ENERGY RELATED SOFTWARE’S: (FREE OF CHARGE) Software Name Manufacturer/ Developing Institution RETScreen Natural Resources Canada HOMER National Renewable Energy Laboratory, USA NREL Solar National Renewable Advisor Energy Laboratory, Model(SAM) Washington ESP-r 11.5 University of Strathclyde, Scotland
Main features of the Software
Cost/ License
Solar PV, Wind, Biomass, Fuel Cell Solar PV, Wind, Biomass, Fuel Cell
Free of Charge www.retscreen.net Free of Charge www.nrel.gov/homer
Solar PV and Energy analysis
Free of Charge https://www.nrel.gov/ analysis/sam/backgroun d.html Free of Charge http://www.esru.strath. ac.uk/Programs/ESPr. htm
Energy analysis
Website
The above software’s are available with free of charge and most of the software’s operations are liked with software provider’s database. Simulation modeling and analysis is becoming increasingly popular as a technique for improving or investigating process performance. Simulation has got a lot of applications in today’s world scenario i.e. in the area of health care, computer and communication system, manufacturing and material handling system, automobile industry, Logistics and Transportation system, Service system, Military and Scheduling. Since simulation is a main tool for modeling and analysis so technology advancement is occurring in this field. Recent advances in simulation methodologies, availability of software, and technical developments have made simulation one of the most widely used and accepted tools in system analysis and operation research. 86 ♦ Role of Software Tools for Support of Renewable Energy Systems Deployment
ICORE 2013
TABLE 2: RENEWABLE ENERGY RELATED SOFTWARE’S (PAY VERSIONS) Software Name
Manufacturer/ Main features of Cost/ License Developing the Software Institution TRNSYS University of Wisconsin, Solar and Energy $2100 for educational Madison, US Simulation use INSEL Insel Company, Germany Solar PV 1700 Euro for full version; PV F-Chart University of Wisconsin, Solar PV $400 for single user, Madison, US SolarDesignTool Verdiseno, Inc, Santa Solar PV expert version Cruz, USA with a monthly fee SolarPro Laplace System Co., Ltd, Solar PV $1,900 for educational Japan use
Website http://sel.me.wisc.edu/trnsys/ http://www.insel.eu http://www.fchart.com/ http://www.solardesigntool.com/ http://www.lapsys.co.jp/english/
The reach of simulation has become wide because the software’s are becoming cheap as well as containing easy to learn languages. That’s the reason why, market contains specific simulation software for a specific task. In the field of Solar PV system, if one is to design stand alone PV system, then there will be separate simulation software for corresponding task as par the requirement of the users. There are twelve major types of software’s for simulating Solar PV System these are RETScreen, PV F-Chart, SolarDesignTool, INSEL, TRNSYS, NREL Solar Advisor Model, ESP-r 11.5, PVSYST 4.33, SolarPro, PV DesignPro-G, PV*SOL Expert, HOMER and many others available are DDS-CAD PV, Polysun, APOS photovoltaic StatLab, PV Designer, SolarNexus, Valentin Software, PV Cost Simulation Tool, NREL's In my Backyard, SolmetricIPV, Solmetric Suneye, Blue Oak Energy and Solar Pro Magazine's Solar Select, Seneca Software & Solar, Inc., Sombrero, PV Potential Estimation Utility, Horizon, Panorama master, METEONORM, GOSOL, Shadows, Shadow Analyser, SPYCE, ECOTECT, Tetti FV, Kerychip, PV Professional, Pvcad, Meteocontrol etc. [Solar photovoltaic software, www.appropedia.org, [Modeling Software, www.energycommunity.org, November, 2010][PV resources, www.pvresources.com, December, 2010].The desirable features of PV Simulation software include consideration of Hardware and software, Simulation Capabilities, Modeling capabilities and Input Output issues. ANALYSIS OF RETSCREEN SOFTWARE OF A BUILDING ENERGY MODEL The energy performance of a building envelope is influenced by a number of factors. For example, these may include design elements such as the physical orientation of the building and the amount of sunlight that penetrates into the interior living or work spaces. Other factors may also include the heat transfer characteristics (both losses & gains) and the location of the building envelope components, including walls, windows, doors, floors and the roof. And the energy performance of a building may also be influenced by any natural air infiltration through the building envelope The RETScreen Software Building Envelope Model can be used worldwide to evaluate the energy use and savings, costs, emission reductions, financial viability and risk for building envelope energy efficiency measures. The software can model a wide variety of projects ranging from passive solar designs for houses, to energy efficient window use in commercial buildings, to complete energy efficient construction practices on the entire building envelope for large institutional and industrial buildings. The model also considers the impact of one or more operating schedules for the base case and the proposed case building, and it models the cross-effects of other energy efficiency measures employed in the facility (e.g. heating and cooling impact of installing efficient lighting systems). In addition, the effective thermal resistance (R-value) or thermal conductance (U-value) of a specific building envelope assembly can be calculated using an optional Building envelope properties tool and a detailed Window properties tool is provided to determine the effective window properties (i.e. area, U-value and solar heat gain coefficient) and individual window costs. The software also includes product, project and climate databases, and a detailed user manual. The RET Screen software has been validated in a number of ways. For example, RET Screen has been compared with the Renewable Energy Laboratory's HOMER simulation tool, which uses hourly solar Insolation data. On an annual basis, the two tools agreed to within a few percent for their predictions of PV array energy production and genset fuel consumption. For all months considered individually, the two tools agreed to within about 10%. Role of Software Tools for Support of Renewable Energy Systems Deployment ♦ 87
ICORE 2013
There are 163 Universities and colleges involved in development of this software. Some of the universities are Indian Institute of Technology–Delhi, Indian Institute of Technology–Roorkee, KJ Somaiya College of Engineering,Tamilnadu Agricultural University The Energy and Resources Institute and soon are participated with co ordination of NASA. CASE STUDY WITH SIMULATION RESULTS In this case the analysis is considered for one engineering college building. For this analysis the procedure is as follows: Step 1: Starting worksheet: The project type is of Energy efficient Measures, this is of institutional facility type and the analysis method is considered as method 1. The Heating value reference is considered as LHV. Higher heating value is typically used in Canada and USA, while lower heating value is used in the rest of the world. Hence LHV is selected. This is of International wide software hence we can get the currency what we required. The climatic data conditions are obtained for this location which is included in Site reference conditions.
FIG. 1
FIG. 2
For selecting climate location To access the RETScreen Climate Database click on the 'Select climate data location' hyperlink or use the RETScreen menu or toolbar shown in Fig. 2. This is the extension of first window which show the data of climatic conditions of the considered location. While the latitude and longitude values are entered, the values of Air temperature, Relative humidity, Daily radiation-horizontal, Atmospheric, wind speed, Earth temperature, Heating Degree days, Cooling degree days are obtained for monthly basis. The data obtained is only for reference purpose not to run the model Step 2: Energy Model: The fuel type is of Electricity at a unit rate of 4.0 Rs/Kwh. The scheduled working days are of 24 days on an average for a month. The base case and proposed cases have working hours of 8 per a day. Therefore, out of 100%, 29% occupancy is utilized. The climatic conditions are always hotter as described in the previous worksheet; hence the cooling temperatures are required for 365 days. Step 3: Facility characteristics: In this section, enter the information about the facility characteristics, for the base case and the proposed case facilities. The user clicks on the blue hyperlinks (e.g. Heating system, Cooling system, Building envelope, etc.) to access the data entry forms used to describe the facility. In addition, the key results of the model are displayed in this section (e.g. fuel saved, simple payback, etc.). In this analysis the electrical equipments like fans, tube lights and computer are taken. For these equipments the base case and proposed case are considered. By using efficient devices the power saving is of 36.2%. Although the initial cost may be high, at last there is a lot of saving. For lights the payback period is 3 years and for electrical equipments there is no payback period means when we invest the money immediately we can get the savings.
88 ♦ Role of Software Tools for Support of Renewable Energy Systems Deployment
ICORE 2013
Step 4: Summary: This section summarizes key information for the base case and proposed case facilities, including detailed information for each fuel type used, as well as fuel consumption and annual energy use information for heating, cooling and electricity. This section also provides a tool to allow the user to benchmark their project for various energy and reference units. In financial analysis, the fuel rate for base case with proposed cases are compared even though the initial investment is high the payback period is very less in some times it may be neglected also. Here the project life is considered for 20 years.
FIG. 3
FIG. 4
CONCLUSION Energy analytics is essential to informing the design, implementation and operation of energy systems. This is particularly true in countries, which are undergoing rapid transformations of their energy systems. Developing these resources can be expensive and time consuming. In this context, open energy data and energy applications may induce better-informed energy decisions. These tools, models and applications are often easily adapted to local needs and can be a catalyst for innovation. This paper highlighted the importance of such software’s work and how it can be succeeded in defining and evaluating software tools for analyzing the feasibility of integrated energy systems and also the universities curriculum to be updated by introducing such software at UG level. This will allow the future technocrats informed decisions to be made about the technical feasibility of supply mix and control strategies, to plan any energy system with full pledged analysis. REFERENCES [1] US Department of Energy. Annual Energy Review 2006 27 June 2007. Accessed 27 April 2008. [2] Torcellini et al. Zero Energy Buildings: A Critical Look at the Definition.National Energy Renewable Laboratory (NREL). June 2006. [3] Investigation of Solar Photovoltaic Simulation Softwares Mahendra Lalwani, D.P.K othari, Mool Singh [4] Natural Resources Canada, www.retscreen.net, accessed on 2nd October, 2010. [5] National Renewable Energy Laboratory, https://www.nrel.gov/analysis/sam [6] Details and four basic functions of Solar Pro, http://www.lapsys.co.jp, accessed on 29th October, 2010. [7] Screen Shot of PV-Design Pro-One of the solar energy programs on the Solar Design Studio”, http://www.mauisolarsoftware.com, accessed on 3rd November, 2010. [8] Details of PVSOL Expert”, http://www.valentin.de/en/products/photovoltaics/12/pvsolexpert, [9] Solar photovoltaic software”, http://www.appropedia.org, accessed on 14th November, 2010. [10] The software tools and databases that are useful for sustainable energy analysis,http://www.energycommunity.org, accessed on 27th November, 2010
Role of Software Tools for Support of Renewable Energy Systems Deployment ♦ 89
Cost Representation and Proposed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer Sthita Prajna Mishra1, S.M. Ali2 and Prajnasmita Mohapatra3 1Research
Scholar, SOA University, Bhubaneswar Professor, KIIT University, Bhubaneswar 3Research Scholar, KIIT University, Bhubaneswar
2Associate
Abstract—The renewable energy is the most efficient and challenging clean and green energy now a days. Only stand alone system have some draw backs where the grid connectivity system and hybrid system are more sustainable form of clean energy where we can get without any disturbance of power 24 hours. This Paper has a study and model design of Hybrid system comprising of solar and wind technology for a small locality rural area and proposed design for the stand alone hybrid system is designed with the help of HOMER software to calculate the exact load with the money and cost estimation by the project and This chapter comprises the simulation data of each component used in Project work as well as the cost effectiveness of project during working and during setup. This is known as Hybrid Optimization Model for Electric Renewable. In this both PV and Wind set has been taken care for the balanced load sharing and to feed the local home connectivity system. Keywords: PV, HOMER, TGB, Wind Turbine
INTRODUCTION In the present scenario the whole world is suffering with energy crisis where in the population growth in India this problem is very much. So for recent trends the conventional energy is going for shortage and the water resources are going for end but only there is one way to overcome this shortage circumstances with the help of renewable sources as solar energy, which is clean and green energy form. As per the geographical location of India and the proposed site in this project has a tremendous available of solar energy resources from nature throughout the year as well as wind energy due to the global wind direction in the upper half of equator and to the earth planetary moment. Hybrid Renewable Energy Research Methodology The proposed hybrid renewable is consisting of wind turbine and solar photovoltaic (PV) panels with battery, generator and inverter are address as part of back-up and storage system. The proposed system is shown in below Figure 1.
FIG. 1: HYBRID MODEL
Load Characteristics Data The load is assumed constant all year. The renewable energy supplied is based on hourly basis as the fluctuation of parameters involved in wind turbines and solar PV. Load characteristics 6 5 4 3 2 1 0
Series1
FIG. 2: GRAPH FOR LOAD OF PROPOSED SITE
ICORE 2013
Solar PV Panels As per the proposed work we are consider that there are ten PV panels with each has a capacity of 1KW.For the proposed site 4KW solar panel is estimated and the cost is 4363$ where replacement cost is 4151$ as per the HOMER software calculation. The lifetime of the panels will consider being 20 years. The monthly average daily solar radiation in Kantapalli is between 5.05Kwh/m2/day with the monthly average daily sunshine duration ranging from six to eight hours as shown in following Table 1. These values are important for sizing of solar energy system. TABLE 1: ANNUALLY SOLAR RADIATION DATA FOR THE SITE Solar Resource Input (Jharsuguda) Clearness Index 0.619 0.63 0.603 0.635 0.588 0.461 0.345 0.392 0.463 0.606 0.616 0.631
Month Jan Feb Mar April May June July Aug Sep Oct Nov Dec
Daily Radiation (Kwh/m²/day) 4.539 5.26 5.766 6.671 6.43 5.089 3.778 4.164 4.563 5.263 4.65 4.414
8 7 6 5 4
Clearness Index
3
Solar radiation
2 1 0 Jan
Feb
March
April
May
June
July
August
Sep
Oct
Nov
Dec
FIG. 3: PLOTTED GRAPH
Wind Turbine The wind turbine has a capacity of 3KW, its initial cost is $1641 per KW and its replacement at $1559. Annual operation and maintenance cost is $82 per year. Its hub and anemometer is proposed to locate at 35 meter height. Lifetime is assumed for 15 years. The average wind speed for this location shown in Table 2. TABLE 2: WIND DATA FOR PROPOSED SITE Wind Speed Data (Kantapalli) Month Wind Speed m/Sec Jan 3.12 Feb 3.27 Mar 3.9 April 3.85 May 3.59 June 3.54 July 3.63 Aug 3.49 Sep 2.86 Oct 2.76 Nov 3.06 Dec 3 Cost Representation and Proposed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer ♦ 91
ICORE 2013
Wind speed m/secc 4.5 4 3.5 3 2.5 2
Winnd speed
1.5 1 0.5 0 Jan
Feb
March
April
May
Junne
July
A Aug
Sep
Oct
Nov
Dec
FIG. 4: GRAPH FO OR WIND ESTIMATEED DATA FOR YEA AR
SYSTEM ARCHITECTURE R DESIGN AND SIMULATION USING U “HOMER” (HYBRID OPTIMIZATION N MODEL FOR ELECTRIC RENEWABLE) The System m Architecturre Comprisess of: •
PV V Array: 4 kW W DC.
•
W turbine1 Wind 1 Generic: 3kkW AC.
•
Ba attery: 24 Tuubular Gel battery.
•
Innverter.
• Reectifier. The proposed p da ata comprisees of 3KW wind w turbine e who is givving AC supp ply through Alternator connected d to wind mill. Converter is connected d which has the property of changing g AC to DC as a Rectifier and DC too AC as Inveerter. PV celll is connected d as of 4KW W load in forrm of DC whhich is converrted to AC before feeding to the load becausse the load iss AC load. Thhe batteries are connecteed in strings (24 ( nos) of d 1250Amp Hr. H The wholee system architecture is designed d for the off-grid rural electriffication. As 2Volt and per the syystem the ba atteries is firsst charged and a store the electric eneergy then thee supply is given to the load. In his h case Tubular Gel ba atteries are used for the e high efficiiency and a an operational time of 10 years.
FIG. 5: 5 SHOWS THE PROPOSED MODEL FOR F THE SITE (HO OMER)
Net Cash Flow Chart for Total Sysstem Configuration For the ca ash flow studyy in the projeect system thee total config guration is ma ade up using g HOMER sofftware and average cost c in each unit is as givven below inn graph. The maximum exxpenditure oof this projecct is for PV panel, duee to its cost iss very high. The T 2nd higheest cost is in battery b due to t the storag ge of electricity and for the longevvity period the tubular gel battery ha as taken the most expenssive cost. Theen the 3rd hig ghest is the windmill, only o the purrchase and establishment e t cost is therre in this secction. The 4thh lowest costt is for the 92 ♦ Cost Representation R a Proposed Model and M of PV & Wind W Hybrid Sys stem for Off-grid d Rural Electrificcation using Hom mer
ICORE 20 013
converter which conveerts the enerrgy form. Thee lowest cosst of the tota al system is DG-Set. It is one time me is low of DG set because the usee of this is in only at the scarcity or investmentt cost and thhe running tim failure of hybrid systeem time. Acccording to prroject the be elow is givenn graphical rrepresentatio on data of cash flow study for prooject.
FIG. 6: NET CASH FLOW CHART FOR TOTAL SYSTEM CONFIG GURATION
Estimated d Cost for thee Project The total project cost is estimated in $ becausse the HOMEER software version is onnly able to calculate c in “$” till now n Indian rupees convversion has not come in software.. So accord ding to marrket value (1$=Rs.68 8.80) as on 28.08.13 2 TABLEE 3: NET PRESENT COSTS Comp ponent PV Generic 3kkW Generator 1 Tubular Geel battery Converter System
Capital ($) 17,452 1,641 780 6,432 1,000 27,305
Replacemen nt ($) 5,170 651 0 5,597 417 11,835
O&M ($) 0 1,048 0 3,988 1,278 6,315
Fueel ($)) 0 0 0 0 0 0
Salvage ($) -2 2,897 -121 -159 -749 -78 -4 4,005
Total ($) 19,724 3,219 621 15,268 2,618 41,450
TABLE 4: POWER O PRODUCTION N ESTIMATION PER E YEAR TO MEET THE DEMAND Component PV array Wind turbine Generator 1 Total
Productio on (kWh/yr) 6,751 1,362 0 8,113
Fraction 83% 17% 0% 100%
The above table is i shown as the t total estiimated load is 8113KW Wh/yr. This esstimated load d is totally A shown the 83% load means m 6751K KWh/Yr. The e rest 17% given by the generation from PV and wind. As load is 13 362Kwh/Yr is supplied by b wind turb bine. So ove erall the entire load is suupplied by the t Hybrid Renewable energy sysstem.
FIG. 7 Co ost Representattion and Proposed Model of PV V & Wind Hybrid d System for Offf-grid Rural Elecctrification using g Homer ♦ 93
ICORE 2013
The above figure shown as the load profile of the selected data. The DMAP is shown as daily mapping of load survey throughout the year. The total average load survey is 3.6KWh/day. The above simulated data is taken from the HOMER. TABLE 5: TOTAL LOAD AND LOAD CONSUMPTION PATTERN Load
Consumption Fraction (kWh/yr) AC primary load 1,314 100% Total 1,314 100%
The total AC primary load is estimated as the 1314Kwh/yr. and the total load is provided by the generation from wind and solar hybrid system. TABLE 6: ELECTRICITY CALCULATION Quantity Value Units Excess electricity 6,575 kWh/yr Unmet load -0.0000000671 kWh/yr Capacity shortage 0.00 kWh/yr Renewable fraction 1.000 TABLE 7: SOLAR POWER GENERATION DATA (ESTIMATED) Quantity Rated capacity Mean output Mean output Capacity factor Total production Quantity Minimum output Maximum output PV penetration Hours of operation Levelized cost
Value Units 4.00 kW 0.771 kW 18.5 kWh/d 19.3 % 6,751 kWh/yr Value Units 0.00 kW 4.21 kW 514 % 4,366 hr/yr 0.229 $/kWh
Simulation Result of Solar Radiation Data for 1 Year
FIG. 8: SIMULATION RESULT OF SOLAR RADIATION DATA FOR 1 YEAR TABLE 8: WIND TURBINE OUTPUT Variable Total rated capacity Mean output Capacity factor Total production Variable Minimum output Maximum output Wind penetration Hours of operation Levelized cost
Value 3.00 0.155 5.18 1,362 Value 0.00 2.78 104 5,596 0.185
Units kW kW % kWh/yr Units kW kW % hr/yr $/kWh
94 ♦ Cost Representation and Proposed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer
ICORE 2013
Wind Turbine Simulation Data
FIG. 9: WIND TURBINE SIMULATION DATA
Battery Simulation Report The use of battery in 1st phase of work for energy storage is a vital requirement for night purpose when the solar is not generated the stored battery can supply the whole load including the wind turbine. The red colour indicates that the battery charge storage is 100% throughout the year and on the demand at night time it discharged. When energy is generated the former work is to give supply to battery for storage purpose then we can send the power for the utility.
FIG. 10
Converter Work Operation The below table mentioned the conversion in process in 1KW rage so as to design the converter for requirement purpose. TABLE 9 Quantity Capacity Mean output Minimum output Maximum output Capacity factor Quantity Hours of operation Energy in Energy out Losses
Inverter 1.00 0.09 0.00 0.37 9.1 Inverter 6,664 882 794 88
Rectifier 1.00 0.01 0.00 0.93 0.6 Rectifier 460 66 56 10
Units kW kW kW kW % Units hrs/yr kWh/yr kWh/yr kWh/yr
Rectifier Simulation Result
FIG. 11: RECTIFIER SIMULATION RESULT Cost Representation and Proposed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer ♦ 95
ICORE 2013
Inverter Simulation Result
FIG. 12: INVERTER SIMULATION RESULT
CONCLUSION It has been concluded that rural electrification presents different load patterns in relation to that of urban loads in terms of the daily variation expected and the yearly variation: rural patterns are smoother. Besides, in case of remote areas that present small incomes (common parameters for most rural areas), the extension of utility grids is not feasible and the total dependence on imported fossil fuels is economically unaffordable, fuel transport costs become prohibitive. REFERENCES [1] Gutierrez-Vera, J. “Use of renewable sources of energy in Mexico case: San Antonio Agua Bendita” Energy Conversion, IEEE Transactions on Volume: 9, Issue: 3 , Publication Year: 1994, Page(s): 442–450. [2] Vera, J.G. “Options for rural electrification in Mexico” Energy Conversion, IEEE Transactions on Volume: 7, Issue: 3 Publication Year: 1992, Page(s): 426–433 [3] Pradhan, N. ; Karki, N.R.,” Probabilistic reliability evaluation of off-grid small hybrid solar PV-ind power system for the rural electrification in Nepal” North American Power Symposium (NAPS), 2012 ,IEEE conference publication 2012, Page(s): 1–6. [4] Ilakkia, T. ; Vijayagowri, G. “Hybrid PV/wind system for reduction of harmonics using artificial intelligence technique” International Conference on Advances in Engineering, Science and Management (ICAESM), 2012 IEEE conference publication, Publication Year: 2012, Page(s): 303–308. [5] Pradhan, N. ; Karki, N.R. ; Pokhrel, B.R.,” Reliability evaluation of small standalone hybridsolar PV-wind power system” IEEE Third International Conference on Sustainable Energy Technologies (ICSET), 2012 Publication Year: 2012, Page(s): 259–264. [6] Anjana, P. ; Tiwari, H.P. “Electrification by mini hybrid PV-solar/wind energy system for rural, remote and Hilly/ trible areas in Rajasthan (India) ”4th International Conferenceon Electric Utility Deregulation and Restructuring and Power Technologies (DRPT),IEEE conference publication 2011, Publication Year: 2011, Page(s): 1470–1473.
96 ♦ Cost Representation and Proposed Model of PV & Wind Hybrid System for Off-grid Rural Electrification using Homer
Empowering the Renewable Energy Sector through Skilling Y.P. Chawla1 and R.S.P. Singh2 1Advr
Jt. Elect. Regulatory Commission & National Jt. Secy. IIPE 2Asso. Prof, School of Vocational Education & Training–IGNOU Abstract—The Renewable Energy (RE) has shown it’s prominence in recent years in India. Like any other sector RE has also shown a Skill Gaps due to the expansion in the sector. There is an urgent need to set up Skill Standards tied to quantitative measurements of qualitative Skill parameters of Human Capital Assets required for the sector needing Capacity Building. The Skill Standards act as bench mark and compare two independent individuals and is useful to establish a correlation, measuring the same, testing the significance of the variation establish cause and effect relationship for course correction as required. The Skill Standards also can establish correlation between Institutions providing Skilling inputs, Skilling facilitators and the individuals from diversified regions and countries as well as other stakeholders. Process of Setting Skill Standards or the concept of Skill councils is a recent phenomenon in India that got attention after the Skill gaps started getting reflected in the requirements of the Industry, due to expansion in various industries, demographic concerns of aging work force in various countries, skilling curriculum remaining unchanged and not catching up with the dynamic technology changes. This resulted in on the job learning or self-learning which at times an expansive proposition is and Return of Investment in Skilling was not considered. Skilled manpower is needed for execution of an ambitious RE program. The Skills needed in this sector are broadly covered in this paper. The skilling domain in this paper will also include Soft Skills within the ambit along with Technical Skills. Keywords: Skills, Renewable Energy, National Occupation Standards, Skill Standards, Skill Gaps, Skilling ROI, Technology needs.
RENEWABLE ENERGY-SKILL STANDARDS India a signatory to the Kyoto protocol for containing Global Warming, increasing Renewable Energy contribution for meeting Energy needs has made this sector vibrant. India’s Grid Connected New & Renewable Energy accounts for 28,905 MWs as accounted for till July 2013, The capacity includes 835 MWs installed during the Year 2013-14 against a total target of 4325 MW with 1839MWs targeting for SPV. This capacity addition includes Solar Photovoltaic (SPV), Small Hydro Power, Biomass Power, Bagasse Cogeneration, and Waste to Power. This however does not include Wind Power. This Paper is focused on Skill Standards on SPV with peripheral touching upon the other RE technologies. India has set up Sector Skill Councils for initiating work on Skill Standards which are called National Occupational Standards (NOS) in India, while a few other countries call it National Skill Standards. It is common to hear from employers reporting serious academic and job skill gaps among new hires and job applicants, especially among younger and less-experienced workers and job applicants. Basic math, writing and communication skills are often described as inadequate for entry-level employment. This is true for various nations including India. Challenge of Skill Gaps is getting compounded by changing demographic patterns world over. As such the Skill Standards are required for every job occupation in every Industry and every sector and that are applicable nationwide irrespective of regional geography. SKILL STANDARDS-WHAT ARE THESE Skill standards or Occupational Standards are set up towards Specification performance identifying knowledge, skills, job competencies and soft skill abilities of an individual needs to succeed in the workplace and make contribution to enhance Productivity of the Organization. These Skill Standards are critical for improving workforce skills, raising living standards of individuals, and improving the competitiveness in the economy. Skill Standards/ National Occupational Standards have to be developed in most industries (Including the RE Sector). Skill Standards in RE Sector • • •
Bring together the skills, knowledge and values necessary to do the work in RE as statements of competence Provide managers with a tool for a wide variety of workforce management in RE sector, quality control and Specification tasks and recruitment in RE Act as a basis of training and qualifications in RE.
ICORE 2013
Principally, Industry based Skill Standards in Any Sector Including RE Sector are Required to be • • • • • • •
Responsive to changing work organizations, technologies and market structure. Benchmarked to world-class levels of industry performance and free from gender, racial, or other forms of bias. Tied to measurable, competency-based outcomes that can be readily assessed. Useful for qualifying new hires and continuously upgrading employees’ skills. Applicable to a wide variety of education and training providers, based on National Vocational Education Qualification Framework (NVEQF). Developed in consultation with various stake holders, independent from prototype training/education provider. Provides opportunities to the employee for horizontal migration to other technologies like from SPV to Bio mass or Power Sector or any another sector altogether say Steel Sector by acquiring requisite modules of Skills as per the required standards in that sector. These Skills Standards also allow vertical mobility to the employee by acquiring higher competencies of laid down skill standards through learning in line with Life Long Learning (LLL) approach
WHY IS SKILL STANDARD IMPORTANT? In today’s work environment: • Jobs require high performance work processes & enhanced skills for cost Competiveness. • Skill standards reflect changing workplace realities. • Skill Standards-a tool for various stake holders. • Nationally recognized Skill standards in identified sectors form a common platform for certifying achievement benchmarked to these standards, allowing portability (even across Industry Sectors) of skills across geographical areas, companies and careers and also to abroad. • Skill and knowledge updating is a lifelong endeavor, forcing employers and employees to spend more effort, time/money on education and training. • Skill standards are benchmarks for deciding education and training, shaping curriculum, and directing funds toward highest value education and training investments. These Skill Standards are useful to the perspective employee to help him know what is expected from him. The Institutions of Training take a reference of these to align the Training/ Skilling Programs accordingly. The Employers take a cue to decide the training programs of their employees in line with dynamically changing technologies and hence the Skill Standards. Status of Skill Council As already indicated above, in India Skill Standards are yet to be set up. Let RE industry be proactive to initiate the process and offer its services to the concerned organizations like NSDC/ FICCI to launch the RE Sector Skill council. A reference was drawn from USA for Electrician for Plant and based on the parameters set by them a study was conducted in India to understand the Skill gaps that exist between USA & Indian Electrician for similar Project occupation. A Plant Electrician comes under hard to find category. And the technicians are required more in numbers.. Objectives of Skill Standards The Specific Objectives of setting Skill Standards for RE technologies are listed hereunder: • to identify the particular skills required to install and maintain renewable energy systems; • to review the present training practices in the RE sector and consider the availability of training facilities for education and training; • to design curriculum of training as per skill standards in RE sector
98 ♦ Empowering the Renewable Energy Sector through Skilling
ICORE 2013
•
to devise a process of engaging the training institutions, learning and Sector Skills Councils and other stake holders involved in the planning of RE sector training with a aim of raising awareness about skills required in the renewable energy sector in India keeping in view expansion program of RE sector.
Way Forward in Setting Skill Standards The RE sector may include the following to begin with: • Solar Photovoltaic (SPV). • Wind. • Biomass-fuelled Power Plants/ combined heat and power (CHP) and heat pumps. The Manpower requirements for RE Sector in India1 are as under TABLE 1: MANPOWER REQUIREMENTS FOR RE Growth Scenario Moderate High
Estimated Current Employment(Nos.)
Estimated Employment by 2015 (nos.) 589,000 699,000
350,000
Estimated Employment by 2020 (Nos.) 10,51,000 13,95,000
The employment potential in Green jobs in India is as under, with India's potential of grid-connected solar power generation capacity to over 200,000 MW and wind energy to over 100,000 MW by 2030 if the right resources (manpower and more importantly, energy policies) are implemented. India can develop massive commercial wind farms to harness the strong onshore costal area and offshore wind to boost the country's supply of clean renewable energy. But, to tap this vast resource, India must develop and undertake the steps for Skilling the manpower, along with smart business models and favorable policies as early as possible. TABLE 2: GREEN JOBS MANPOWER REQUIRED/ MW OF SOLAR PLANT Employment as per 2010 By the Year 2013 (Automation & Skilling) By the Year 2017 By the Year 2022 Off Grid Solar Thermal–Thumb Rule Study
Manufacturing 20 9
O&M Very low 4
2.5 4.5 30 3.86
1.4 5.9
A Solar Developer having executed Projects-Solar Thermal
Steam Cycle additional 7
Indirect
Ancillary
Total 20
60 23.18
8.1
35.16
5
TABLE 3 Skills
No. of Days Involved in One No. of Installations/ year Installation SPV-Domestic Installer/ Electrician 6 days for a typical roof top 27 SPV-Commercial For a Typical 1 MW Ground Mounted Engineer 20 2 Installer/ Electrician 10 11 Structural Technician 10 35 Large-Wind Administrative/ Design 40 2 Installer 15 11 Maintain Wind Engineer/Technician 2 35 Days available Installer 180 (Technicians with 70% of available working time) Engineer 70 (Engineer with 30% of available working time) The above data is sourced from London Renewables Studies with an objective to initiate the discussions in India.
Generally for a direct employment of 24,000, an indirect employment of 48,000 is generated by the Solar Industry for PV off grid. Direct Employment of Manufacturing takes 20%, Installation takes 10%, O&M 20%, Marketing 40% and others 10%. The data as available has been presented. Solar roof Top heaters being another sub sector has not been discussed here.
Empowering the Renewable Energy Sector through Skilling♦ 99
ICORE 2013
With Solar Power getting an infusion of USD 3.25 Mn., Cabinet giving a nod to 75o MWs of Solar Energy, Govt. announcing a few lacs Solar roof tops to be getting ready, Solar Power has a big future. TABLE 4 Broad Green Jobs Skills Requirements to Install an RE System Solar Photovoltaic Industry Electrical/Electronics, Roofing & scaffolding for Roof Top Installation Structural Work for Ground mounted installation Health & Safety (implications of converting electricity from DC to AC) Wind Industry Engineering/Design, Construction skills (for foundations) Crane driver (for larger turbines), Electrical/Electronics Biomass-fuelled CHP Electrical fitters, Mechanical fitters, Mechanical engineer
The above manpower numbers in table 2 & 3 for Green Jobs some skills, which are broadly listed hereunder, These Skill requirements are for all Job occupations related to site activities in the Industry that include Engineers to Technicians. The demand for skilled manpower in the sector is on the increase in Solar Sector with Govt. of India push for Solar Cities. Chandigarh has taken steps in setting up Grid Connected Roof tops Solar Plants in the city. Gujarat also has announced to set up 60 MW Solar Roof Tops projects that will be grid connected. RE BUSINESS AND TRAINING/ SKILLING Many of the businesses have been talked to; these were found to be headed by self-taught Engineers, who view training as the only way to both widen the renewable energy market and to employ staff with the necessary skills. The education and training in renewable energy is to be geared towards, qualified Electrician/ Structural technicians/ Welders/ plumbers wishing to move in for horizontal mobility into renewable energy installations. Based on the Skill Standards2 set by Texas Skill Standards Board in respect of Solar Projects is as under in table no.5: TABLE 5 Skills/ Function Area 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
a Listening Speaking Infn. & Commn. Usage Collect/Analyze Info Prod. Analyze/ Solving Judgment/Decision Organize & Plan Social Skills Usage Adaptability Team Work Leadership Consensus Building Enhancing Skill, Self Career Development Writing Reading Maths. Science
Site Assessment Work b 4 4 4 4 4 4 3 3 3 2 2 2 2
Design for Site c 3 3 3 4 4 4 3 2 3 3 2 3 3
Coordinating Resources d 3 3 3 3 3 3 3 3 3 3 3 3 2
System Installation e 3 3 3 3 4 4 4 3 3 4 3 3 3
System Maintenance f 3 3 3 4 4 3 3 2 3 2 2 2 2
Average
3 3 4 3
3 3 4 3
3 3 3 2
2 3 3 3
2 3 3 3
2.6 3 3.4 2.8
g 3.2 3.2 3.2 3.6 3.8 3.6 3.2 2.6 3 2.8 2.4 2.6 2.4
In various other Skill Standards only one Skill Rating on 1-5 Scale is indicated. But in this case Skill ratings have been varied as per the Job activity in the header row. Thus the table no.5 also covers the Skill levels of various activity roles.
100 ♦ Empowering the Renewable Energy Sector through Skilling
ICORE 2013
The numbers and the Skills referred above led to a study by the authors in respect of an Electrician for a Plant. The results of the study from a sample collected from various organizations employing Plant Electricians with 1-2 years’ experience in Power Sector initially and then expanded to the sectors where similar work profile exists like Refinery, Cement & Steel Sector etc. Wherein under this Para RE Business and Training/ Skilling has been talked about, it may be noted that an International program attended by RE professionals from 7 countries attended. The RE programs are in great demand and SESI can invest in training units to be run by retired professionals from the Industry to fill the Skill Gaps in this sector. The Experts may fend for themselves out of the programs. The investment of Skilling infrastructure can come from manufacturer-members of SESI. This will keep the pipeline of supply chain of the skilled manpower healthy. Methodology of the Study During an interaction with a focus group belonging to HR Managers of Power Sector utilities from various parts of India, a questionnaire was requested to be filled in to establish skill gaps of a Plant Electrician so as to benchmark this vis.-a-vis. Plant Electricians of USA for which results were available. 38 Skill Parameters in 7 broad groups (various Skills covering Basic, Thinking, Personal, Information usage, Team Participation, Plant system understanding, Use of Technology) were compared with USA average Score on 1-5 Scale, of 33 respondents and 26 respondents in case of an Indian sample of the population as under in Fig 1.
FIG. 1: SKILLS OF PLANT ELECTRICIANS IN USA-INDIA COMPARISON 38 SKILL PARAMETERS COMPARED
There are some surprises in the study, which is the advantage of these studies and fixing the Skill Standards, made it interesting to explore further and more requests were made to get the questionnaire filled in from Plant Managers. Initial respondents are from various State Utilities, and additional responses were received from organizations like NTPC, Power Grid, BHEL and one of the respondents is an Indian in Dubai. The responses were collected in person and through Email and the population sample of respondents was increased to 40 numbers. Selected few parameters of the study are presented hereunder in Fig 2. The Colored bands in Red & Green are USA study and the arrows are for Sample in India. The numbers at the right end of the table are USA Score/ Indian Sample Score.
FIG. 2: COMPARISON OF TECHNICIANS SKILLS AS PER THE RESPONDENTS IN INDIA VS. USA
Empowering the Renewable Energy Sector through Skilling♦ 101
ICORE 2013
FIG. 3
Correlation between various parameters of the Skill Gaps that affect other secondary factors and vice versa was also done by the authors in respect of a plant electrician. A few of these correlation factors that can help us deciding the intervention for improving the Skill Standards are hereunder in Table 6 FIG. 6: CORRELATED SKILL PARAMETERS OF PLANT ELECTRICIAN Primary (Inducing) Factor Interpreting/Communicating Information Effective Speaking Capability Acquisition/Usage of Information Time management Computer Usage-Information Processing Monitor & Controlling System Performance Idea Generation & Creative thinking Reasoning Skills Explains to Other members Selection Right Technology for Working Recognizing & Solving Problems Interpreting/Communicating Information Trouble Shooting Technology Issues Organising/maintaining/Retrieving Info.
Correlation Co-eff. $$ 0.83 0.82 0.82 0.82 0.77 0.76 0.75 0.75 0.74 0.74 0.74 0.73 0.72 0.72
Secondary (Indicator) Factor Organising/ Maintaining/ Retrieving Information Effective Listening Skills Reasoning Skills Integrity & Honesty Acquisition/ Usage of Information Explains to Other Members Effective Speaking Capability Recognising & Solving Problems Participates as a Part of Team Capability of Visualization Decision Making Skills (Upto the Mandate) Managing Human Resources Selection Right Technology for Working Handling Machine/Material Inventory
RESULTS Some of the results are contrary to normal perception, but these were based on the data collected. Reasons for its being contrary to the perception were found out and got interesting findings that Indian Electrician is better in Inventory Management as compared to an American Plant Electrician, as the economy there is a replacement economy, while in India the Spare Parts and other inventory is all to be kept ready by the technician as getting the inventory could be a problem. An American Electrician will go by rule book provided to him and would not go beyond that in experimenting anything, which in a way is good but Indians known for Jugad make the system work somehow. SUMMARY 1. The above study indicates that though there is a Skill Gap, but it is not do sharp. This can be corrected if the training is modified appropriately. The Soft Skills and High Technology Skills need attention at our end. 2. The Skill Standards can help the HR Professionals to develop suitable training packages and measure the Skills for improving the Plant productivity. 3. The Skill Standards can act as tools for rewarding the employees. 4. The Skill Standards, if set by SESI in collaboration with Employers, Technical Institutions will provide a guide to the perspective employees and can help promotions of the RE sector. 5. The Skill Standards drawn to fit in NVEQF program will motivate the employees.
102 ♦ Empowering the Renewable Energy Sector through Skilling
ICORE 2013
Recommendations to SESI for taking up proactively in setting up a Skill Standard Council for RE Sector helping in fixing the RE standards as under 1. Identifying the Occupation and Reviewing Resources. 2. Helping Setting up of the Industry Technical Advisory Committee (Employers/ Researchers/), Associating Colleges Community & Technical institutions to develop Skill Standards. 3. Setting up a Plan, Method of Analysis and Strategy for Validation of the Skill Standards. 4. Developing Process/ Work-Oriented Information. 5. Developing Worker-Oriented Information. 6. Analyzing, Synthesizing and Organizing Data. 7. Integration of skill standards into a curriculum of Community Colleges & Institutions. 8. Setting up a Skilling institution for Indian manpower as well as for providing Skills abroad which may lead to RE business expansion overseas. 9. Recognizing the Application Process and Apply for program recognition, renewing program recognition getting the Skills achieved accredited. 10. Dynamically up-dating of the Skill Standards in line with Technology advancement. 11. The Skilling may also include understanding critical scenarios of RE Power evacuation based on system demand, having or not having a backing of Conventional power etc., micro grids etc. CAVEAT DGE&T had tried to introduce a program at ITIs but there were no takers. Motivation and job prospects have to be propagated to the job aspirants by SESI appropriately.
Empowering the Renewable Energy Sector through Skilling♦ 103
Meeting India’s Energy Demand through Concentrated Solar Power Technologies J.P. Kesari1, Nadim Shams2, P.B. Sharma3 and Rajendra Kumar Kaura4 1,2,3,4Department
of Mechanical Engineering, Delhi Technological University, Bawana Road, Delhi–110042, India
Abstract–India’s Solar Mission is devoted to harness solar power in India in areas of high solar radiation. The mission aims at establishment of 22000 MW of solar power of which 20000 MW is for grid connected solar power and 2000 MW for stand alone solar power plants. The mission is currently focused on solar photovoltaic SPVs for solar power development. The concentrated solar power, CSP can provide low carbon renewable energy resources in countries or regions with strong direct normal irradiance. It is estimated that by 2050 CSP could provide 11.3% of global electricity with 9.6% from solar power and 1.7% from back up fuels such as fossil fuels or biomass as per technology road map for CSP by International Energy Agency, 2010. The biggest advantage of CSP is that it has integrated thermal storage so as to provide 24x7 availability of energy. CSP besides solar electricity can also provide high temperature heat for industrial processes and can also be utilized for solar refrigeration and solar air conditioning. Current technologies for CSP include CSP parabolic troughs, linear Fresnel reflectors, solar towers and parabolic dishes. With new and emerging technologies CSP can be used to generate high temperatures such as 600-700 degree centigrade. Further it is important to examine the importance of CSP in the context of solar thermal and solar refrigeration and air conditioning.
INTRODUCTION The basic concept of concentrating solar power is relatively simple: CSP devices concentrate energy from the sun’s rays to heat a receiver to high temperatures. This heat is transformed first into mechanical energy by turbines and then into electricity. As of early 2010, the global stock of CSP plants neared 1 GW capacity. Projects now in development or under construction in more than a dozen countries including China, India, Morocco, Spain and the United States are expected to total 15 GW. Parabolic troughs account for the largest share of the current CSP market, but competing technologies are emerging. Some plants now incorporate thermal storage. The solar energy that CSP plants use is measured as direct normal irradiance (DNI), which is the energy received on a surface tracked perpendicular to the sun's rays. It can be measured with a pyrheliometer.DNI measures provide only a first approximation of a CSP plant’s electrical output potential. At present, there are four main CSP technology families, which can be categorized by the way they focus the sun’s rays and the technology used to receive the sun’s energy. Parabolic trough systems consist of parallel rows of mirrors (reflectors) curved in one dimension to focus the sun’s rays. The mirror arrays can be more than 100 m long with the curved surface 5 m to 6 m across. Stainless steel pipes (absorber tubes) with a selective coating serve as the heat collectors. The coating is designed to allow pipes to absorb high levels of solar radiation while emitting very little infrared radiation. Linear Fresnel reflectors (LFRs) approximate the parabolic shape of trough systems but by using long rows of flat or slightly curved mirrors to reflect the sun’s rays onto a downward-facing linear, fixed receiver. Solar towers, also known as central receiver systems (CRS), use hundreds or thousands of small reflectors called heliostats to concentrate the sun’s rays on a central receiver placed atop a fixed tower. Some commercial tower plants now in operation use DSG in the receiver; others use molten salts as both the heat transfer fluid and storage medium. The concentrating power of the tower concept achieves very high temperatures, thereby increasing the efficiency at which heat is converted into electricity and reducing the cost of thermal storage. In addition, the concept is highly flexible; designers can choose from a wide variety of heliostats, receivers, transfer fluids and power blocks. Parabolic dishes concentrate the sun’s rays at a focal point propped above the centre of the dish. The entire apparatus tracks the sun, with the dish and receiver moving in tandem. Most dishes have an independent engine/generator at the focal point. This design eliminates the need for a heat transfer fluid and for cooling water. The storage and backup capabilities of CSP plants offer significant benefits for electricity grids. Losses in thermal storage cycles are much smaller than in other existing electricity storage technologies including pumped hydro and batteries, making the thermal storage available in CSP plants more effective and less costly.CSP plants can enhance the capacity of electricity grids to accommodate a larger share of variable energy sources, thereby increasing overall grid flexibility. As demonstrated in Spain, connecting CSP plants to some grid sub-stations facilitates a greater share of wind energy. CSP plant backup may
ICORE 2013
also eliminate the need to build fossil-fired “peaking” plants purely to meet the highest loads during a few hours of the day. The CSP roadmap is meant to be a process, one that evolves to take into account new technology developments, policies and international collaborative efforts. Ref: International Energy Agency (IEA) Technology roadmap (CSP) OCT-2010. TYPES OF CONCENTRATED SOLAR POWER A b s o rb e r Tube
C u rv e d m irro r
A b s o rb e r tu b e a n d r e c o n c e n t r a to r
C u r v e d m ir r o r P ip e w it h t h e r m a l f lu id
P a r a b o lic T r o u g h
L in e a r F re s n e l
R e c e iv e r / E n g in e
S o la r R e c e iv e r
R e f le c t o r
H e lio s t a t s
D is h / E n g in e
C e n tra l R e c e iv e r
FIG. 1
SOLAR ENERGY RESOURCES FOR DESIGNING CSP IN INDIA CSP technologies rely on Direct Beam radiation for operation. That is radiation direct from the sun that has not been diffused or deflected by clouds or other atmospheric factors and so can be focussed by the mirrors. Ideally the data needed to assess potential sites is short interval, Direct Normal Irradiation (DNI) measurements collected over several years. Issues include:  

In designing a CSP plant, knowledge of the seasonal variation in DNI resource is needed to make an optimal economic assessment of the degree to which the solar field is oversized relative to the power block in the high season and undersized in the low season. Power blocks have an efficiency that decreases with part load (due to reduced solar input) and a threshold load under which they cannot operate. Thus to predict the power block’s total daily output, knowledge of the time dependence of input heat transfer fluid energy flow is required. The thermal receivers in collector fields take many minutes to reach operating temperature from cold, hence to accurately predict their output, data at time resolutions of one minute or less is required to predict performance in situations of intermittent cloud.
FIG. 2 Meeting India’s Energy Demand through Concentrated Solar Power Technologies  105
ICORE 2013
FIG. 3
FIG. 4
INDIAN ENERGY SECTOR & CSP The shape of our future will be largely determined by how we generate and apply technological innovations to the energy sector and also to simultaneously enforce compliance of energy efficiency and environment management regulations. CL-CSP Technology of Concentrated Solar Power provides a promise to achieve high enough temperature of about 600 degree C good enough for Power Cycle Optimization of a Mega scale CSP Plant. The technology is however needs to be proved on a moderate size Pilot plant of 30 kW. The present DPR is aimed at Design and development of 30 kW CSP Pilot plant at RGPV campus for proving the CL-CSP technology, the most innovative Solar technology of today, for further up-gradation to Mega scale plants. Despite the fact that we in India have taken a giant leap forward in increasing the installed capacity from a mere 1713 MW in 1950 to 1, 83,670 MW as on date, the renewable energy sources however contribute merely 11% with total capacity of over 19,000 MW.But this contribution has a major social and economic impact on rural and remote area population. Solar Energy is the most promising Renewable Energy in India having high average solar intensity. This has prompted Indian RE sector for planning a target of 20,000 MW under Jawaharlal Nehru National Solar Mission by 2022. By virtually all accounts, renewable energy resources will be an increasingly important part of the power generation mix over the next several decades. Not only do these technologies help reduce global carbon emissions, but they also add some much-needed flexibility to the energy resource mix by decreasing our dependence on limited reserves and overseas sources of fossil fuels. Experts conclude that small hydropower and biomass will continue to dominate the renewable arena for some time. However, the rising stars of the renewable world-wind power and Solar are on track to become strong players in the renewable energy market of the next century. We shall see an emergence of the Concentrated Solar Power (CSP) as major power sources for decentralized as well as grid connected power generation. Coal substitution is another focus area for Conventional coal fired Power Plants, where rising cost of Imported coal and Coal allocation issues have set aside many power plants from taking off the ground The CL-CSP Technology once proven can also pave way for Power Cycle Optimization by interposing CSP in the feed water stream for Feed heater replacement.
106  Analysis of the Effect of Nanostructured Paste Components on the Efficiency of Dye Sensitized Solar Cells
ICORE 2013
CONCENTRATED SOLAR POWER–ROADMAP FOR INDIA Concentrated Solar Power (CSP) or solar thermal power plants is renewable energy's answer to bulk power supply. The CSP market has witnessed a global uproar in the past year not only in terms of installations but also in terms of developer interest. Currently, the world cumulative pipeline has reached up to 20 GW which accounts for the projects that are under development and also that are under construction. The International Energy Agency (IEA) projects that by 2050, CSP plants would supply up to 11.3% of global electricity. The Indian home ground is also not far behind with an ambitious target of 10 GW by 2022 from solar thermal power plants alone, with plans to install 550 MW of CSP by 2013. India had planned to set up a 35 MW plant based on CSP in Rajasthan way back in 1994, but the project was postponed indefinitely because no qualified contractors were able to submit a bid. Now, of course, the situation has changed. The Jawaharlal Nehru National Solar Mission recognizes CSP as a key source of renewable power. The Ministry of New and Renewable Energy (MNRE) is setting up a string of demonstration projects. A 10 MW CSP plant is already under construction near Bikaner by ACME, a private company. Ministry officials said CSP could be an attractive option for the country, given the fact that large areas of northwest India fall in the high radiation zone. The costs can be brought down if concentrating mirrors and receiving tubes are manufactured locally. The vast Rajasthan Desert is very similar to the Sahara desert in Africa, and has the potential to become the largest solar power plant in India. Due to high levels of available sunlight, CSP plants in Rajasthan could satisfy most of India’s energy needs in just a few years. India’s potential benefits from solar power are as numerous as the sands of Rajasthan desert, and include reduced dependence on fossil fuels and a cleaner environment. These benefits can be realized by installing renewable energy technologies, such as CSP, to protect the environment while diversifying energy resources and helping to lower prices. Solar power can also reduce strain on the electric grid on hot summer afternoons, when air conditioners are running, by generating electricity where it is used National Thermal Power Corporation has announced plans to install two 50MW solar thermal plants in Gujarat and could extend this to 300MW. It is also building 5MW and 1MW installations on the Andaman and Nicobar islands. Pune Gadhia Solar, an OEM, has announced plans for a 100 MW plant in Kutch, Gujarat and is in discussions with the Maharashtra state government to install 1.0–1.5 MW CSP systems, to extend power to rural areas. CSP–WORLDWIDE  
More than ten different technology combinations for CSP installations. More than 500 MW installed capacity using Parabolic Trough technology, 40 MW from Solar Tower, 5 MW from Fresnel technology and 0.5 MW from Dish technology working worldwide.  More than 800 MW capacities project has been announced worldwide under CSP technology.  CSP is beneficial in co-generation mode like Power & Steam and more so in Tri-generation modes like Power, De-salivation and Heat Generation.  CSP can supplement conventional power plant for feed water heating capacity. The important industrial segments where industrial process heat between 100-300 deg C is required are: Applications for CL-CSP System      
Pulp & Paper Textile Dairy Leather Food Processing Electroplating Meeting India’s Energy Demand through Concentrated Solar Power Technologies  107
ICORE 2013
 
Fertilizer Drug and Pharmaceutical
FIG. 5
The above diagram shows the application of high value process heat in industries. Cross Linear system has the capability to produce temperature in excess of 700 deg C. CONCENTRATED SOLAR POWER (CSP) TECHNOLOGY Solar photo voltaic (PV) technology has been used for long to generate power from the sun. But its use has remained limited because of its low efficiency and other constraints. Now, concentrating solar power (CSP) is emerging as a viable alternative. Unlike PV cells which are flat, CSP involves the use of parabolic mirrors in long troughs, which concentrate solar irradiation into a centrally placed special tube which absorbs radiation. The heat generated into the tube is then used to produce electricity. CSP plants generate electricity from sunlight by focusing solar energy, collected by an array(s) of suntracking mirrors called heliostats, onto a central receiver. There are different types of CSP technologies currently available. In one typical type (parabolic trough with liquid salt), liquid salt (a mixture of sodium nitrate and potassium nitrate) is circulated through tubes in the receiver, absorbing the heat energy gathered from the sun. The heated salt is then routed to an insulated tank where it can be stored with minimal energy losses. To generate electricity, the hot molten salt is routed through heat exchangers and a steam generation system. The steam is then used to produce electricity in a conventional steam turbine. After exiting the steam generation system, the now cool salt mixture is circulated back to the “cold” thermal storage tank, and the cycle is repeated. There are various other types heat transfer medium which are used One of the biggest benefits of CSP is its ability to store thermal heat which ensures continuous supply of power even when the sun is down and when it gets cloudy. The core of a CSP plant is the special tube placed at the focal point of parabolic mirrors. The mirrors have a motorized system that enables them to keep tracking the sun. Inside the glass tube is a steel tube with absorber separated by vacuum insulation. The ceramic metallic tube contains synthetic oil which gets heated up to 400 deg Celsius. The oil flows into a heat exchanger which generates steam, which in turn, is used to generate power using conventional steam turbines. The receivers have to achieve maximum solar absorption and at the same time minimal emission of heat. The surface of the glass tube remains cold, while the temperature in the steel tube could go up to 400 deg Celsius.
FIG. 6 108  Analysis of the Effect of Nanostructured Paste Components on the Efficiency of Dye Sensitized Solar Cells
ICORE 2013
Pre-Requisites for CSP CSP Technologies rely on Direct Beam radiation for operation, i.e. direct radiation from the Sun that has not been diffused or deflected by clouds or other atmospheric factors and so can be focused by the mirrors. Ideally the data needed to assess potential sites is of short interval duration and hence Direct Normal Irradiation (DNI) measurements are collected over several years. The considerate are: 
 
In designing a CSP plant, knowledge of the seasonal variation in DNI resource is needed to make an optimal economic assessment of the degree to which the solar field is oversized relative to the power block in the high season and undersized in the low season. Power blocks have an efficiency that decreases with part load (due to reduced solar input) and a threshold load under which they cannot operate. Thus to predict the power block’s total daily output, knowledge of the time dependence of input heat transfer fluid energy flow is required. The thermal receivers in collector fields take many minutes to reach operating temperature from cold, hence to accurately predict their output, data at time resolutions of one minute or less is required to predict performances in situations of intermittent cloud cover.
REFRENCES [1] 1Detailed Project Report, 30 kW-thermal Cross linear-CSP System test unit. A joint Project by RGPV-DTU-BERGEN, submitted to MNRE 2012-13. [2] 2. International Symposium on Concentrated Solar Power(CSP): Opportunities for India,23rd-27th January 2012.organised by DTU-RGPV_TOKYO TECH. [3] 3.2nd symposium on concentrated solar power Indo-Japan joint collaborative research project.18th-22nd, june2012. Organized by DTU-RGPV-BERGEN &TOKYO TECH. [4] 4. Kesari J P and P.B. Sharma (2013) Changing the face of India with Solar Power Technologies: Generation of Energy, Employment and Entrepreneurship. Paper presented at National Seminar on IT Applications in Energy Management, Rajiv Gandhi Institute of Information Technology, Amethi, A campus of IIIT Allahabad, April 15, 2013. [5] 5. KESARI J P and P B SHARMA(2013) Changing the Face of India with Solar Power Technologies: Generation of Energy, Employment and Entrepreneurship, Paper published in GRIDTECH 2013, April 3-5, Pragati Maidan, New Delhi. [6] 6.Kesari J P and P B Sharma (2013) Changing the face of India with solar power Technology, keynote address in International Conference on Role of technology in Nation Building, 27 April 2013, Subharti university, Meerut, UP.
Meeting India’s Energy Demand through Concentrated Solar Power Technologies  109
Development of a Photo Sensor, PIC Micro Controller based Solar Tracker A. Gokul Raj1, T.V. Chavda2 and V. Siva Reddy3 1,2,3Sardar
Patel Renewable Energy Research Institute, Vallabh Vidhyanagar, Gujarat–388120, India
Abstract—Sun tracking systems are, in general, timer controller based (15o per hour rotation) and require manual intervention for starting and stopping the system because solar time changes throughout the year. Available electro-optical automatic controlled sun trackers are expensive and local expertise is not available for their repairs and maintenance. Keeping this in view, a solar tracking system using PIC micro controller and commercially available low cost motor has been designed and developed at SPRERI. The solar tracker actuates the PV panel to receive maximum solar radiation. PV system output has been connected to a DC pumping system and efficiency of the tracking system was worked out. Total electrical energy outputs in the fixed and tracking modes were 3.12 and 3.64 kWh respectively, for solar radiation of 5.24 kWh/ m2/ day. Variations in the electrical energy output were far lower in the tracking mode than the fixed mode. The sun tracking system was found more effective during morning and late afternoon hours. Tracking system efficiency was found to be more than 15 % with a tracking error of ± 2 %.
INTRODUCTION
Installed capacity (MW)
India has 1718MW of solar PV installed capacity as on May 2013. The annual growth rates between 40 % and 90 % have been reported in PV panel production and 4023MW installed capacity for solar PV in India is expected at 2016 (Fig. 1) [1]. The solar tracking devices maximize the solar panel collected energy by placing the PV panel perpendicular to the sun’s direct radiation throughout the day. Compared to timer controller based trackers, the electro-optical trackers are accurate. In those active controllers light sensors are used to determine the sun position. 4500 4000 3500 3000 2500 2000 1500 1000 500 0
4023 3522 2834
1174
1252
2012
2013
2014 Year
2015
2016
FIG. 1: EXPECTED INSTALLED CAPACITY FOR SOLAR PV IN INDIA TILL 2016
Nader Baroum [1] developed a tracker using LDR light sensors, PIC16F84A microcontroller and 12V DC motor having a torque of about 18.3Nm to 23.73Nm. A chain drive with 4:1 gear reduction ratio was used between motor socket and frame socket [6]. Tanvir Arafat Khan et al., [2] has developed a tracking system using ATMEGA32 microcontroller. The total angle of rotation (1500) covered by the tracker from east to west was achieved by 40 rotations of the stepper motor at 3.750 rotations for every 15 minutes. Kais I. Abdul-lateef et al., [3] developed a single axis tracker with PIC16F84A flash microcontroller, permanent magnet geared DC motor, H-Bridge motor driver and PV cells (as optical sensor). C. Saravanan et al. [4] integrated a three phase 0.5 HP induction motor to their solar tracking system with gear reduction ratio of 25:1. VFD was used for controlling the speed of the induction motor by controlling the frequency of the electrical power supplied to the motor. Romy Kansal [5] developed a single axis sun tracking system using PIC microcontroller and stepper motors. 8 % tracking efficiency was achieved for the prototype tested at field. Development of tracking system using commercially available low cost motor with PIC microcontroller has not been reported so far. An attempt has been made at SPRERI to develop a high efficiency tracking system using commercially available sub-components, which will facilitate prompt repair and maintenance of the system by local technicians. Design, development and performance evaluation of the tracking system has been presented here. Experimental performance analysis of the system was carried out using 1kW solar PV system.
ICORE 2013
DESIGN AND DEVELOPMENT 16F877A microcontroller was selected as controller and program was written in PICBASIC language using MPLAB tool according to the program algorithm shown in the Fig 2 [10]. The microcontroller receives the signal from the photo sensors based on light intensity. The microcontroller compares the photo sensor signal in percentage of the first sensor with respect to the second sensor. It can be noted that when the difference in intensity of light is less than 98 % and greater than 60 % the microcontroller drives the motor in forward direction (towards west) and if the difference in intensity of light is more than102 % and less than 150 % it drives the motor in reverse direction (towards east). Fig. 3 shows the schematic diagram of the control circuit. The control circuit comprised of, step down transformer (230V AC to 12V AC supply), rectifier (12V AC to 12V DC), DC voltage regulator (for constant 5V DC), PIC 16F877A micro controller, LCD display, LED indicators, relays, control switches and photo sensors. An AC to DC adaptor of 12V, 5A was used to supply power to the low cost motor. The relays were used for changing the direction of rotation of the motor shaft based on the micro controller instruction. 1 kW capacity solar panel was mounted on the tracker. The system comprises of a solar PV panel mounting structure with tilting adjustment and a rigid pole at the centre. The photo sensor used is a photo diode. The semicircular structure used for the smooth and controlled movement of the tracking system. The chain slides over the semi-circular structure. A compound wheel reduction drive was connected between the DC motor and the chain.
Development of a Photo Sensor, PIC Micro Controller based Solar Tracker ♦ 111
ICORE 2013
FIG. 2: FLOWCHART ILLUSTRATING THE OPERATION OF THE MICROCONTROLLER
FIG. 3: SCHEMATIC DIAGRAM OF THE CONTROL CIRCUIT
The schematic view of the solar tracking system was shown in Fig. 4. The design features of the tracking system are as follows.
112 ♦ Development of a Photo Sensor, PIC Micro Controller based Solar Tracker
ICORE 2013
FIG. 4: SCHEMATIC LINE DIAGRAM OF THE SOLAR TRACKING SYSTEM
•
The tracking system makes use of a photo diode as the light sensor, which is locally available at a very reasonable cost and easy to replace.
•
The gear reduction of 30:1 was achieved using a compound gear reduction mechanism.
•
A PIC16F877A microcontroller is used in the control unit and is commercially available.
•
The program was written in PICBASIC (.bas). The program code is compiled (.asm) and assembled (.hex) using MPLAB IDE. A PIK KIT2 programmer was used for programming the micro controller.
•
An economical and commercially viable DC wiper motor was used. The rated capacity of the wiper motor is as follows. Rated voltage : 12V Rated current : 5A Shaft diameter : 8mm RPM : 60 Rated torque : 10 N-m Braking torque : 30 N-m •
The usage of battery for driving the DC wiper motor was eliminated by using an AC to DC adaptor (Rated capacity: 220 V AC to 12 V DC at 5 A).
•
The developed solar tracker can be upgraded to hold 5 kW PV panel installed capacity.
PERFORMANCE ANALYSIS The experiment to determine the performance of the solar tracker was performed. Various parameters of the tracker system were monitored and the relative data was recorded. Two pyranometers were selected and used to measure the solar radiation incident on the solar panel. One pyranometer was placed on the tracking panel to measure the tracking position solar radiation and another was placed at the centre on the beam to measure fixed position solar radiation simultaneously.
Development of a Photo Sensor, PIC Micro Controller based Solar Tracker ♦ 113
ICORE 2013
Output voltage (V)
30.00 25.00 20.00 26 Apr-2013 (TM)
15.00
29 Apr-2013 (FM) 10.00 5.00 0.00 10:30
11:30
12:30
13:30
14:30
15:30
16:30
Time (h)
FIG. 5: OUTPUT VOLTAGE VS. TIME IN TRACKING MODE AND FIXED MODE
Output current (A)
25.00
20.00
15.00
26 Apr-2013 (TM)
10.00
29 Apr-2013 (FM) 5.00
0.00 10:30
11:30
12:30
13:30
14:30
15:30
16:30
Time (h)
FIG. 6: OUTPUT CURRENT VS. TIME IN TRACKING MODE AND FIXED MODE
The variation of output voltage and output current with respect to time in tracking mode and fixed mode is shown in Fig. 5 and Fig. 6 respectively. The output voltage and current fluctuation were less in tracking mode compared to fixed mode. In tracking mode the PV panel surface was controlled by the tracker to face the high solar radiation region throughout the day thereby reducing the cosine losses at the PV panel surface. Fig. 7 shows the gain in input solar radiation in tracking mode compared to fixed mode. There was a considerable drop in input power before and after noon in fixed mode compared to that of the tracking mode. The maximum input power gain was during 1 PM to 5 PM.
114 ♦ Development of a Photo Sensor, PIC Micro Controller based Solar Tracker
ICORE 2013
Power Input (TM) - 26 Apr 2013 Power Input (FM) - 30 Apr 2013
8.00
Power input (kW)
7.00 6.00 5.00 4.00 3.00 TM - Tracking Mode FM - Fixed Mode
2.00 1.00 0.00 10:30
11:30
12:30 13:30 Time (h)
14:30
15:30
16:30
FIG. 7: GAIN IN INPUT POWER IN TRACKING MODE COMPARED TO FIXED MODE Power Output [TM] - 26 Apr 2013 0.60
Power Output [FM] - 29 Apr 2013
Power output (kW)
0.50 0.40 0.30 0.20 TM - Tracking Mode FM - Fixed Mode
0.10 0.00 10:30
11:30
12:30
13:30
14:30
15:30
16:30
Time (h)
FIG. 8: GAIN IN OUTPUT POWER IN TRACKING MODE COMPARED TO FIXED MODE
Fig. 8 shows the gain in output power in tracking mode compared to fixed mode. The deviation in output power was found to be less in tracking mode compared to fixed mode. The maximum drop in output power was during 1 PM to 5 PM. Throughout the day the deviation in input power and output power was very less in tracking mode compared to fixed mode. Table 1 shows the experimental results of the tracking system from 26-April to 2-May, 2013. In fixed Pyranometer shown approximately equal energy input for April 26 (TM) and May 1 (FM). Tracking efficiency of 16.67 % was achieved with tracking error of ± 2 %. TABLE 1: EXPERIMENTAL RESULTS OF THE SUN TRACKING SYSTEM Day
Energy Input [FM] kWh 26-Apr 40.98 29-Apr 41.23 30-Apr 39.76 1-May 40.78 2-May 40.44 * FM: Fixed mode; TM: Tracking mode
Energy Input [TM] kWh 47.00 ------46.12
Mode of Operation TM FM FM FM TM
Energy Output kWh 3.64 3.23 3.10 3.12 3.49
Development of a Photo Sensor, PIC Micro Controller based Solar Tracker ♦ 115
ICORE 2013
CONCLUSION A photo sensor PIC micro controller based tracking system was developed using commercially available components and was found working satisfactorily. Its performance was evaluated using 1 kWp solar PV panel. The fluctuation in PV panel voltage and current were found reduced considerably with the use of the tracking system. The tracking efficiency was found to be 16.67% for a tracking error of ±2 %. The tracking efficiency is likely to increase for lower level of tracking error. The system is easy to repair and maintain as the components used in the solar tracker are commercially available at a very low cost. REFERENCES [1] Nader Barsoum. Implementation of dual-axis solar tracking pilot project. Transaction on Energy, biotechnology, planning and Environment, 2011; ET-E33/GJTO: 2229–8711. [2] Md. Tanvir Arafat Khan, S.M. Shahrear Tanzil, Rifat Rahman, S M Shafiul Alam. Design and construction of an automatic solar tracking system. 6th International Conference on Electrical and Computer Engineering ICECE 2010. [3] Kais I. Abdul-lateef. A low cost single-axis sun tracker system using PIC microcontroller. Diyala Journal of Engineering Sciences, 2012; 5:65–78. [4] C. Saravanan, M.A. Panneerselvam, I.William Christopher. A Novel Low Cost Automatic Solar Tracking System. International Journal of Computer Applications, 2011; 31:0975–8887. [5] Romy Kansal. PIC Based Automatic Solar Radiation Tracker (M. Tech thesis). Thapar University, Patiala, Punjab; 2008. [6] P. Roth, A. Georgive, H. Boudinov. Cheap two axis sun following device. Energy Conservation and Management, 2005; 46:1179–1192. [7] Okan Bingol, Ahmet Altintas, Yusuf Oner. Micro controller based solar-tracking system and its implementation. Journal of Engineering Sciences, 2006; SAYFA: 243–248. [8] Arindam Bose, Sounak Sarkar, Sayan Das. Helianthus–A low cost high efficient solar tracking system using AVR microcontroller. International Journal of Scientific & Engineering Research, 2012; 3(10). [9] Hemant Kumar Nayak, Manoj Kumar, Nagendra Prasad, Rashmi Rekha Behera. Fabrication and experimental study on twoaxis solar tracking. International Journal of Applied Research in Mechanical Engineering, 2011; 1(1). [10] Avipsa Dey. Design, development and performance evaluation of a PIC micro controller based commercially viable solar tracker (M.E thesis). B.V.M Engineering College, Vallabh Vidyanagar, Anand; 2013.
116 ♦ Development of a Photo Sensor, PIC Micro Controller based Solar Tracker
Analysis of the Effect of Nanostructured Paste Components on the Efficiency of Dye Sensitized Solar Cells: A Taguchi Approach Sarita Baghel1 and Ranjana Jha2 1,2School
of Applied Sciences, Netaji Subhas Institute of Technology, University of Delhi, Delhi, India
Abstract—The properties nanostructured porous film of the photoelectrode of DSSC is highly influenced by the amount of solvent, dispersant and surfactant present in the paste. We investigate the effect of ethanol, acetylacetone, and Triton–X100 in TiO2 nanopowder on the photoelectrode of DSSCs using Taguchi method-based robust design technique. It was found that the solvent has highest effect on the efficiency whereas dispersant is the least affecting factor. Keywords: Nanostructured, Optimization, Surfactant
INTRODUCTION Dye sensitized solar cells (DSSCs) are characterised by relatively high efficiency and low cost of fabrication. Due to these properties they have attracted lot of attention [1-3]. Since the original model cell reported by O’Regan and Gratzel in 1991[4], the DSSCs have been studied to bring improvements in its performance as well as stability [5, 6]. Owing to such decades long effort, DSSCs convert more than 11% of incident solar light into electricity [7]. The DSSC consists of mainly four components namely nanoporous metal oxide film, sensitizing dye, electrolyte and counter electrode consisting of catalyst. The metal oxide film has a significant role in the enhancement of photoelectric conversion efficiency of DSSC, and many studies focus on the relation between film structure and power conversion efficiency of DSSC [8-10]. It is well known that photovoltaic performance of DSSCs is strongly affected by the quality of the nanostructured film. It is therefore crucial to prepare a high quality photoanode for highly efficient DSSCs. There are many methods through which nanocrystalline metal oxide film can be prepared, including screen–printing deposition [11, 12], spin coating [13], spray pyrolysis [14, 15] and chemical vapor deposition [16, 17]. The desirable qualities of the film are high surface area and high porosity. Large surface area leads to greater amount dye adsorption whereas high porosity results in better permeability of the electrolyte. Such film can be easily prepared by the “doctor blade” technique. In this method, a solvent, dispersant and surfactant are mixed with commercial nanopowder to form a paste which is then deposited on ITO coated glass using a glass rod and scotch tape. Adsorption of dye and permeability of the electrolyte, depend upon the structure of the photoelectrode, which changes with different proportion of the quantities of the paste components. In this paper, we investigate the effect of ethanol, acetylacetone, and Triton-X100 on the efficiency of DSSC using Taguchi approach. This optimization technique not only enables us to find the component which affects the most but also the percentage contribution each component holds on the efficiency of DSSC. TAGUCHI METHOD Taguchi method was introduced by Taguchi, a Japanese quality engineer [18]. Taguchi method attempts to optimize a process or product design. Despite having some limitations, it has been able to solve single response problems effectively. The following are the steps to be followed for parameter optimization Step 1 : Determination of the characteristic to be optimized. Step 2 : Identification of influencing factors and their levels. This refers to the selection of control factors and the determination of optimal levels for each of the factors. Control factors are those variables in a process that management can manipulate. Step 3 : Design of orthogonal array and conduct of experiments using the factors and factor levels. Step 4 : Analysis of the data and determination of optimum levels of factors. Orthogonal Array (OA) As three levels and three factors are taken into consideration, L9 OA is used in this analysis. Only the main factor effects are taken into consideration and not the interactions. The degrees of freedom (d.f) for each factor is 2 (numbers of levels-1 i.e 3-1=2) and therefore, the total d.f. will be 3×2=6. Usually, the d.f. of the O.A should be greater than the total d.f of the factors. As d.f. of L9 is 8 it is chosen for this study.
ICORE 2013
In table.1 [19], Ethanol, Acetylacetone, and Triton X-100 are three controllable factors added in 2.5g of TiO2 nanopowder, influencing the response (Efficiency). Each factor has been assigned three levels (Quantity).Efficiency is the characteristic that needs to be optimized The results tabulated in Table.1, were then analysed to determine the optimum level of factors as well as their relative ranking TABLE 1: L9 (33) ORTHOGONAL ARRAY DESIGN EXPERIMENTS [19] Trial Number 1 2 3 4 5 6 7 8 9
A 1 1 1 2 2 2 3 3 3
B 1 2 3 1 2 3 1 2 3
C 1 2 3 3 1 2 2 3 1
Response 4.8 5.0 6.1 4.9 5.4 5.0 4.9 4.3 3.9
Column 1, number of trials Column 2, ethanol(A); level 1, 20ml; level 2, 40ml; level 3, 60ml. Column 3, acetylacetone(B); level 1, 0.4ml; level 2, 0.6ml; level 3, 0.8ml. Column 4, Triton-X100(C); level 1, 0.2ml; level 2, 0.4ml; level 3, 0,6ml. Column 5, (Response), the photoelectric conversion efficiency of DSSCs,η (%) RESULTS AND DISCUSSION Signal to Noise Ratio The signal to noise S/N ratio is calculated based on the quality of characteristics intended. The objective function described in this investigation is the maximum of the efficiency, so the larger the best S/N ratio is calculated using the formula given below
1 n ⎛1 1⎞ Larger the Best: S / N = −10log 10 ∑ ⎜ × ⎟ n i =1 ⎝ yi yi ⎠
(1)
Where n= number of replications and yi=observed response value The efficiency value of DSSCs is analyzed to study the effect of components of the paste used to make nanostructured film. The experimental data are converted into S/N ratios which are tabulated in Table 2. S/N values of all levels are calculated and listed in Table 3. Analysis of Variance The purpose of analysis of variance (ANOVA) is to find the significant factor statistically. It gives a clear picture how far the process parameter (quantity of paste components) affects the response (efficiency) and the level of significance of the factor considered. TABLE 2: S/N VALUE n 1 2 3 4 5 6 7 8 9
A 20 20 20 40 40 40 60 60 60
B 0.4 0.6 0.8 0.4 0.6 0.8 0.4 0.6 0.8
C 0.2 0.4 0.6 0.6 0.2 0.4 0.4 0.6 0.2
Efficiency 4.8 5 6.1 4.9 5.4 5 4.9 4.3 3.9
118 ♦ Analysis of the Effect of Nano Structured Paste Components on the Efficiency of Dye Sensitized Solar Cells
S/N 13.62 13.97 15.7 13.8 14.64 13.97 13.8 12.66 11.82
ICORE 20 013
TABLE A 3: RESPONSEE TABLE FOR SIGNAL TO NOISE RATTIOS Level 1 2 3 Deltta Rannk
A 14..4369 14..1437 12..7648 1.6 6721 1
B 13.7442 13.7655 13.8357 0.0915 3
C 13.3646 6 13.9209 9 14.0599 9 0.139 2
TABLE 4: ANALYYSIS OF VARIANCEE FOR S/N RATIO O Sourrce Df SEQ Q SS ADJJ SS A 2 4.7 7832 4.78 832 B 2 0.0 0138 0.01 138 C 2 0.8 8123 0.81 123 Residual error 2 4.0 0304 4.03 304 total 8 9.6 6397 9.63 397 df, deg gree of freedom m; SEQ SS, sequuential sum of sq quares;
ADJ MS 2.39 916 0.00 069 0.40 061 2.0152
F 1.186 67 0.003 34 0.201 15
(%)Con ntribution 49 9.6198 0 0.1431 8 8.4266 41.8105 100 0
Rank 1 3 2
Adj SS, adjusted sum of o squares; ADJJ MS, adjusted mean m square annd F, Fisher ratio
The ANOVA A tablee for S/N ratios is calcula ated and listed in Table 4. The F-testt is carried out o to study the signifiicance of process param meter. The high F value indicates i tha at the factor is highly sig gnificant in affecting the t responsee. The main effects for S/N ratio are plotted p in Fig g. 1.
FIG. 1: MAIN EFFECTS PLOTS L AND RESIDUAL LOTS FOR S/N N RATIOS Analysis of o the Effect of Nano N Structured d Paste Compon nents on the Effficiency of Dye S Sensitized Solar Cells ♦ 119
ICORE 2013
Optimum combination of factor levels can be obtained from main effects plots in Fig. 1. The combination that yields largest value of S/N ratio is determined from these graphs to be as follows: Factor A B C Optimal level 1 3 3 Thus, A1, B3, C3, provide theoretically the best combination for the highest possible response value, which is in agreement with the experimental result. CONCLUSION During fabrication of dye-sensitized solar cells various controllable and uncontrollable factors arise which can affect its energy conversion efficiency. This research was an attempt to investigate the TiO2 layer fabrication through (statistical design of experiments) Taguchi method. The use of this method allowed analysis and ranking of the importance of the factors relative to one another. In this study, the role of ethanol as the key factor influencing the efficiency of DSSC is established. Ethanol contribution is highest at 49% followed by Triton X100 with over 8% contribution. Acetyl acetone is the least affecting component with 0.1% contribution and ranked lowest in the Anova table. Moreover, not only Efficiency but other characteristics like fill factor can also be taken into consideration. This method can also be used for optimizing TiO2 film with other fabrication technologies like screen printing, spin coating and spray pyrolysis. REFERENCES [1] T. S. Kang, A. P. Smith, B. E. Taylor and M. F. Durstock, Nano Lett. 9 (2009) 601-606. [2] Gratzel. 2001. “Photoelectrochemical cells-MIT.” Nature 414: 338. http://web.mit.edu/andrew3/Public/Papers/2001/Gratzel/2001_Nature_Photoelectrochemical%20cells_Gratzel-3.pdf [3] P.T. Chou, Q.F. Zhang, G.E. Fryxell, G.Z. Cao, Adv. Mater. 19 (2007) 2588. [4] O’Regan and Gra¨tzel. 1991.” A low-cost, high-efficiency solar cell based on dye sensitized colloidal TiO2 films.” Nature 353:737–40. doi:10.1038/353737a0 [5] Snaith and Mende.2007.” Advances in liquid electrolyte and solid state dye sensitized solar cells.” Advanced Materials 19:3187–3200. http://onlinelibrary.wiley.com/doi/10.1002/adma.200602903 [6] Hamann, Jensen, Martinson, Ryswyk and Hupp. 2008. “Advancing beyond current generation dye-sensitized solar cells.” Energy and Environmental Science 1: 66–78. http://pubs.rsc.org/en/Content/ArticleLanding/2008/EE/b809672d [7] Chiba, Islam, Watanabe, Komiya, Koide and Han. 2006.” Dye-sensitized solar cells with conversion efficiency of 11.1%.” Japanese. Journal of Applied Physics 45: L638–L640. http://jjap.jsap.jp/link?JJAP/45/L638 [8] Hara, Miyamoto and Abe.2005.”Electron Transport in Coumarin-Dye-Sensitized Nanocrystalline TiO2Electrodes” Journal of Physical Chemistry B 109: 23776-23778. http://pubs.acs.org/doi/abs/10.1021/jp055572q [9] Jorge, Claudio, Calixto, Pedro and Victor. 2007.” The influence of surfactants on the roughness of titania sol–gel films” Material Characterization 58: 233-242. http://www.sciencedirect.com/science/article/pii/S1044580306001227 [10] Yoo, Kim, Kim, Hahn, Lee and Cho. 2007. “Effects of annealing temperature and method on structural and optical properties of TiO2 films prepared by RF magnetron sputtering at room temperature” Applied Surface Science 253:3888. http://www.sciencedirect.com/science/article/pii/S0169433206011196 [11] Ito, Nazeeruddin, Liska, Comte, Charvet, Pechy, Jirousek, Kay, Zakeeruddin and Gratzel.2006.” Progress in Photovoltaics 14: 589. [12] Nazeeruddin, De Angelis, Fantacci, Selloni, Viscardi, Liska, Ito, Takeru and Gratzel. 2005.” M. G. Combined Experimental and DFT-TDDFT Computational Study of Photoelectrochemical Cell Ruthenium Sensitizers”, Journal of. Americal Chemical. Society 127, 16835-16847. http://pubs.acs.org/doi/abs/10.1021/ja052467l [13] Cai, Wang, Yuan and Duan. 2012.”Magnetic-field effect on dye-sensitized ZnO nanorods-based solar cells”, Journal of Power Sources 216: 269-272, http://www.sciencedirect.com/science/article/pii/S0378775312009019 [14] Okuya, Nakade and Kaneko. 2002.” Porous TiO2 thin films synthesized by a spray pyrolysis deposition (SPD) technique and their application to dye-sensitized solar cells.” Solar Energy Materials and Solar Cells 70: 425.http://www.sciencedirect.com/science/article/pii/S0927024801000332 [15] Okuya, Nakade, Osa, Nakano, Kumara and Kaneko. 2004. “Fabrication of dye-sensitized solar cells by spray pyrolysis deposition (SPD) technique”, J. Photochemistry and Photobiology 164:1-3 http://www.sciencedirect.com/science/article/pii/S1010603004001194 [16] Somani, Somani, Umeno and Sato. 2006.” A Concept and demonstration of all organic Gratzel solar cells (dye sensitized solar cell)”. Applied Physics Letters 89: No. 083501. http://dx.doi.org/10.1063/1.2337563 [17] Miyasaka and Kijitori. 2004.”Low-Temperature Fabrication of Dye-Sensitized Plastic Electrodes by Electrophoretic Preparation of Mesoporous TiO2 Layers”, Journal of Electrochemical Society 151: A1767. http://jes.ecsdl.org/content/151/11/A1767. [18] Taguchi. 1986. Introduction to quality engineering: Designing quality into products and processes, Asian productivity organisation: Tokyo. [19] Xu, Zhou, Yang, Hu, Sebo, Chen, Tai, and Zhao. 2011.” Effects of Ethanol on Optimizing Porous Films of Dye-Sensitized Solar Cells” Energy and Fuels 25: 1168–1172 http://pubs.acs.org/doi/abs/10.1021/ef101546a. 120 ♦ Analysis of the Effect of Nano Structured Paste Components on the Efficiency of Dye Sensitized Solar Cells
Recent Developments in Standalone PV System: For Rural Electrification S.M. Ali1, Punyashree Pattanayak2, Sushree Subhadarsini3 and Prasun Sanki4 1,2,3,4School of Electrical Engineering, KIIT University, Bhubaneswar E-mail: [email protected], [email protected], [email protected], [email protected]
Abstract—Solar electricity supply system has grown and now growing also at very rapid pace in India since last few years. A total of 1047.84 MW of grid connected photovoltaic projects and 160.8 MW of off-grid systems have been commissioned Between 2010 to 2012. The standalone system is much more used in rural areas and helpful in rural electrification. Because almost 40% of the country’s population, most of which lives in the villages, is still deprived of reliable supply of electricity. The conventional power system with large power plants connected to grids, failed to reach the last mile of the vast rural regions. Even where electricity has been officially achieved, the supply is erratic, inadequate and not upto the mark. Therefore standalone PV systems should provide a good quality electricity services to be considered as an alternative to conventional grid extension, for places with no access to electricity. Quality requirements must aim at daily energy supply to the users, but also to ensure long system lifetimes. This paper includes the components, technologies used and recent developments in standalone PV system. Keywords: Grid-Connected, Conventional Power System, Standalone PV System
INTRODUCTION Rural electrification is an integral component of poverty alleviation and rural growth of a nation. In India, electricity has not played effective role in the socio-economic growth of village. Government of India has ambitious target of providing electricity to all villages by 2008 and all rural households by 2012. Steps are initiated with Rural Electric Corporation, State Electricity Boards, Reforms in Power sector. Ever increasing demand of electrical energy is causing a large gap in generation and load demand. All the requirement of energy cannot be fully met with conventional grid supply so, an alternative energy source has to be found out for this purpose. India has one of the fastest growing economics in the world and ranked sixth place in the worldwide consumer of energy. Being the seventh largest country in the world, six thousand villages inhabit 72.2 percent of its human resource (census 2001). About, 40 percent of the total energy is in rural areas. Domestic sector constitutes major energy demand and its consumption accounts for 60 percent of energy used. The main energy sources are coal and oil, whilst hydro, wind, nuclear and biomass provide additional sources [1]. • Both the traditional energy and commercial energy are in short supply and the demand supply gap is in increase. • Pressure on traditional energy resources such as wood is continuously increasing due to growing population. • Heavy dependence on commercial fuels such as coal and oil as a short term measure for meeting increasing demand is alarming in view of depleting fossil fuels and pollution. • Energy supply to far–off rural areas is associated with high transportation and transmission losses of about 22.4%. Thus emphasis should be laid on the auditing of the energy in such a way that ensures affordable, ecofriendly and clean energy. There are two ways of supplying electrical energy to rural areas-one extending grid to near by areas and other distributed generation particularly in deep rural areas. Majority of rural electrification is carried out by the existing grid extension. The need of modifying the existing grid connected supply in rural India [2][3]. Importance of electricity as a crucial infrastructure input for economic development of the country has been well established. Recent studies of rural electrification indicate the following broad consensus concerning the impact of electrification in the rural areas. [4] FEATURES OF RURAL ELECTRIFICATION Rural electrification is an important component of Integrated Rural Development. In India, it has been given less importance with respect to urban, because of the following reasons. •
Villages are located from 3-80 km away from existing grid or even more.
•
They are located in difficult terrain areas like forests, hill areas and deserts.
•
The number of households may range between 2 to 200, with a majority of villages having a population below 500.
ICORE 2013
•
60% of the 5 67000 villages in India have a population of 500 and under and 8% of these are inaccessible. (Source: Census 2001).
•
Power demand in villages is quite low due to dispersed distribution of loads. Also rural domestic consumers are mainly peak time consumers, contributing for poor load factors of 0.2-0.3. The Electricity act 2003 (EA03) marked an increase in urgency attached to the problem at the national level, codifying the requirement to supply electricity to all villages. To implement the law, the government launched the Rajiv Gandhi Grameen Vidyutikaran Yojana initiative in 2005, with the aim of achieving universal electrification by 2012.In theory, electricity produced under the Rajiv Gandhi scheme is supposed to cover operational costs (except for households below the poverty line). The Rajiv Gandhi guidelines also emphasize distributed generation options in cases where grid extension is not feasible, with individual states required to submit proposals to the Ministry of Non–conventional Energy sources (MNES). MNES envisions a major role for renewable sources in meeting the 2012 electricity requirement (Banerjee 2006)[5]. CRITERIA FOR ELECTRIFICATION IN AN ISOLATED VILLAGE The criteria for village electrification are as: (a) identification of parameters such as: 1. Economic: cost of product, maintenance and operating cost, prevailing subsidy, tax benefits, benefits due to absence/ lesser amount of social/ scarcity/opportunity cost etc-all in annualized quantities. 2. Social: Energy habit of the customer, social custom, aesthetic value of the product customers goodwill for reasons such as lowering of pollution by use of these 'green system', political goodwill/ propaganda, population density & accessibility of the location, grid connectivity, etc. 3. Environmental: Availabili ty of solar radiation and other environmental conditions that would significantly affect the performance of the SPV system in consideration. 4. Supply of time. A. A: 24 hours supply. B. B: fixed time supply. C. C: Any time supply.
FIG. 1: STANDLONE PV IN VILLAGE HOMES AND SCHOOLS
NEED OF RURAL ELECTRIFICATION The aim is implement tele-education in rural schools, as well the adult night classes for a population that is essentially illiterate and would otherwise have little chance to learn to read. This program is especially 122 ♦ Recent Developments in Standalone PV System: For Rural Electrification
ICORE 2013
important since it will provide electricity to these schools for the first time. The academic environment will be enhanced with adequate lighting and with the ability to provide educational links through television and satellite communication. It is expected to raise the living standard in the school and a positive effect on health-related issues. Depending on the distance from the grid and the concentration of the communities, PV alternative can be cost-effective compared to grid connection. Then a defined PV rural-electrification market was identified based on the existence of a substantial dispersed, nongrid-connected community population and the potential to provide electricity with costs below comparable grid extension options[6]. The criteria for the schools selection include: the school should be at least 10 Km away from the power distribution lines; the municipality should be inside areas with the worst quality of life classification in the State; the municipality prefectures must accept to participate of the program. Rural electrification policy by Government of India(Source: Ministry of Power)has identified the importance of this. Electricity is an essential requirement for all facets of our life and it has been recognized as a basic human need. It is the key to accelerating economic growth, generation of employment, elimination of poverty and human development especially in rural areas. •
Previous definition of village was(Source: Ministry of Power)-A village will be deemed to be electrified if electricity is used in the inhabited locality, within the revenue boundary of the village, for any purpose whatsoever.
Modified definition of village from 2004-05 is,-A village would be declared as electrified ifa. Basic infrastructure such as Distribution Transformer and Distribution lines are provided in the inhabited locality as well as the Dalit Basti/ hamlet where it exists. (For electrification through Non Conventional Energy Sources a Distribution transformer may not be necessary). b. Electricity is provided to public places like Schools, Panchayat Office, Health Centers, Dispensaries, Community centers etc. and c. The number of households electrified should be at least 10% of the total number of households in the village. CONFIGURATION OF STAND-ALONE PV SYSTEM A stand-alone solar house system (SHS) proposed for adoption is an immediate interim measure toward satisfying the less than modest requirements of electricity of the inhabitants of remote rural communities. Fig. 2 shows the layout of a typical stand alone PV system for rural houses.
FIG 2: SCHEMATIC VIEW OF PV/ DIESEL HYBRID SYSTEM FOR RURAL ELECTRIFICATION Recent Developments in Standalone PV System: For Rural Electrification ♦ 123
ICORE 2013
Design Considerations In designing the system a number of factors need to be taken into account. These include technical specifications, sizing of individual system components, safety considerations as well as system economics. The following is a condensed excerpt from the design process. The proposed system comprises an inverter and batteries. Thus the efficiency of these system components, along with that of wiring, must be taken into account. The gross daily energy demand for a typical remote household can be determined from E
gd =
E nd η i ×η b ×η w
where Egd and End denote gross and net energy demand per day respectively. ηis efficiency with subscripts i, b and w referring to inverter, battery and wiring respectively. Typical efficiencies are 85% for the inverter, 85% for batteries and about 98% for the system wiring in a well esigned PV system [7].
FIG. 3: ELEMENTARY MODEL OF STANDALONE PV SYSTEM
System Voltage It has been suggested that if the daily energy demand is more than 1 kWh, a system voltage of 24 Vdc should be considered, whereas 12 V is deemed to be a better choice if the daily energy demand is less than 1 kWh [8. On that basis, a system voltage of 12 V is adopted for design. With a daily energy demand of some 800 Wh, this into having to deliver about 70 Ah of daily charge capability (DCC). PV Modules In choosing the PV modules for implementation, several factors need to be considered including the daily demand, solar insolation at the geographic location as well as the method of mounting the modules. It seems that the modest daily household energy demand of some 800 Wh can be easily met with the PSH of 5.5 of the Northeast. There are basically two options as far as mounting the PV cells is concerned: (a) fixed mounting, and (b) mounting on a tracker. The latter option may yield up to 20% more output but costs more and requires skilled maintenance. However, both in view of budgetary considerations and the villagers’ lack of familiarity with the technology involved it is deemed best to opt for fixed mounting. The number of modules is derived on the basis of daily energy demand and the commercial availability of the module specified[9]. Charge Controller, Inverter and Batteries Charge Controller A charge controller is essential to the effective and safe operation of a SHS. The controller must be sized to handle maximum input currents produced by the PV modules and maximum output currents delivered to battery and load. Currents in the three main current circuits dictate the sizing of the charge controller [7]. They are the i) array-to-controller current, ii) controller-to-load current, and iii) battery-to-controller current. The array-to-controller circuit should be able to handle at least 125% of total short circuit current 124 ♦ Recent Developments in Standalone PV System: For Rural Electrification
ICORE 2013
(Isc) [7]. This yields a maximum array-to-controller current of 17.3A. The maximum controller-to-load current is obtained as 15.3 A from the total system power of 184 W and the system voltage of 12 V. Of these two, whichever is the larger is to be chosen as the maximum battery-to-controller current. Allowance needs to be made to meet these current requirements. In the present case, a 20 A controller with a maximum 22 A input and output current will be chosen for the 12 V system. [10]. Inverter The size of the inverter depends on the total load demand on the system. The inverter capacity (IC) is obtained from: IC =
P SHS η i × cos ϕ × k loss
where PSHS is the total estimated energy demand as determined before (Table I), ηi is the inverter efficiency, cosϕ is the power factor and kloss accounts for reduction due to of other system losses. Batteries Battery specifications are to be derived from the daily energy demand, the number of days of storage required–the so called days of autonomy (DOA)–and the maximum depth of discharge (DOD). DOA depends on where the system is to be located. BSC =
DCC × DOA DOD
The maximum depth of discharge (DOD) refers to the lowest point to which the battery can be discharged before having to be recharged. For a deep discharge battery, DOD can be as low as 40% state of charge [11]. Wiring and Protection 1.
2.
Wiring: Wiring design aims at determining the best possible layout and wire gauges for circuits by considering the currents carried by each circuit and ensuring that voltage drops are not significant [7]. A voltage drop of ∆V= 2% is considered most suitable for the best energy transfer although in practice 5% is also considered quite acceptable [7]. However, occasionally voltage drops of up to 10% voltage drops are tolerated in small power SHS schemes [11]. Protection: Fuses and a circuit breaker are used to provide protection for the PV modules, charge controller, inverter, batteries and appliances against system faults Fuse sizes are usually determined on the basis of fault currents which exceed rated currents by 20–56%.[10][12].
ADVANTAGES OF PV SYSTEM PV technology is identified as most environment friendly technologies. It requires only sunlight and no other energy fuel. [13][14] Being modular in design, the capacity can be increased to meet additional demand. It is easy to dismantle and reconfigure these systems for other applications. PV Systems require little maintenance. These components can be manufactured and assembled locally. 1. Environment friendly as they do not emit gaseous and liquid pollutants. 2. Can be easily transported, assembled and installed in remote areas. 3. Produce DC electricity that can be stored in batteries. 4. Zero fuel usage and Noise free. 5. Robust, reliable, weather proof and having long life of 25 years with proper maintenance.
Recent Developments in Standalone PV System: For Rural Electrification ♦ 125
ICORE 2013
PV PANEL FOR A RURAL HOUSE 1. The average energy consumption of a household is influenced by many factors-Like construction and size of house, climate, season, size of house, and size of family (census 2001 average 5.3 persons per house in rural area with 3 rooms). 2. Arid regions in India receive plentiful solar radiation of 4-7kwh per square meter with the potential availability of 20 MW per square kilometer. Average solar insolation available is 5 kWh per square meter. (Source: Indian Renewable Energy Development Agency). IREDA is planning to electrify 18,000 villages by year 2012 mainly through solar PV systems. 3. Availability of solar insolation in Dharwad, (Karnataka state) rural indicates that there is ample scope for using PV inrural areas (fig. 4) and additional 2.71 kWh energy can be saved in tubular batteries with one panel of 35 W, i.e. the system is able to supply power to the load for 3 days even if the charging is drastically reduced due to rain or other natural calamities. Also the depth of discharge for the battery is taken to be 75%.[15]
FIG. 4: SOLAR RADIATION MAX [KWH/SQUARE METER] AND LOAD [KWH] VERSES TIME OF DAY IN INDIAN VILLAGE
4. Government of India provides central financial assistance for remote village electrification programs. 5. Every kWh of generated solar power prevents the release 0.7 kg of carbon dioxide. Crystalline silicon PV panels prevent 2,000 kg-3,000 kg of carbon dioxide emissions per square meter for a 25years life expectancy of the panel. Using PV in areas of high solar insolation could reduce carbon emissions by as much as 450 Million Metric tons during the next 25 years Every gigawatt of electricity generated by PV rather than coal was estimated to prevent up to 10 tons of sulfur dioxide, 4 tons of nitrogen oxide, 0.7 tons of particulate matter (including cadmium and arsenic), and up to 1,000 tons of carbon dioxide being emitted into the air. 6. Electric lighting (up to 200 times brighter than kerosene lamp directly improves the quality of life. It allows children to study in the evening and women to gain some precious time for them or to extend income generating work into the evening hours (Domdom et al2000). Scarce availability of kerosene and high cost of diesel, solar power is making inroads amongst villagers. 7. Mathematical modeling of 75 Wp array using MATLAB programming is shown in the fig. 5. 8. PV cost includes two components-PV system cost and Balance-of-system. System cost mainly includes module, delivering d c supply and Balance of system includes other ancillary equipments like power storage, power conditioning, mounting and site specific requirements. Decreasing trend of PVSHS cost by 20% and increasing trend of growth are resulting in encouraging results in India. 9. The main barriers to large penetration of solar house system (SHS) consists of (Purohit and Kandpal et al2005):
126 ♦ Recent Developments in Standalone PV System: For Rural Electrification
ICORE 2013
a. Nature of the solar energy resource high initial investment and still higher costs of these systems to the end users as comparedto the conventional. b. Lack of strong marketing network and excellent market-support infrastructure. c. The market for SPV home lighting systems is currently supply driven as user’s needs have not been full. d. Emphasis is more on technology development rather than product development to effectively meet the user’s need and. e. Awareness level of the benefits of SHS is not up to expected level.
FIG. 6: MATLAB MATHEMATICAL MODEL OF 75WP SOLAR ARRAY
CONCLUSION Rural electrification is a ‘selective catalyst’ to improve essential for many rural activities. It works best when it is complemented by social and economic infrastructure development. In India, 5 lakh SHSs have been deployed till December 2007, and Government of India started National Solar Mission 2010. This aims promote programmes for off grid applications, reaching 1000 MW by 2017 and 2000 MW by 2022, in three different phases. Recent programs are showing good results but more promising new approaches need to be tested to determine if they can address poverty, equity, environmental and public health concerns in the context of ongoing global restructuring of energy industries. There is a vast scope for utilising of solar photovoltaic energy in India. With continuing research and development and cost reduction it can become the most potential energy source. With a clear renewable energy policy in place, India is the forerunner in this sector. There is room for manufacturers, foreign investors, local financial and institutional agencies and others. Solar energy can be one of the thrust areas due to its accessibility through the country in sufficient quantity. For we owe it to ourselves and our children to provide for sustainable development with due regard to our ecology. REFERENCES [1] Kamalapur G D, Udaykumar R Y: ‘Electrification in rural areas of India and consideration Of SHS’,5th International Conference on Industrial and Information Systems, ICIIS 2010, Jul 29-Aug 01, 2010, India [2] D P Sen Gupta (1989) Rural Electrification in India: ‘The achievements and the shortcomings TENCON apos;89.Fourth IEEE Region 10 International Conference, 22-24 Nov 1989 pp 752-755 [3] Siyambalpitiya D J T etal, (1991) Evaluation of grid connected rural electrify-cation projects in developing countries, IEEE Transactions on Power Systems, Vol, 6, No.1 February pp332-338 [4] Mohan Munasinghe, Rural Electrification in the third world, July 1990, power engineering journal, pp-189-202. [5] Manas Mondal and Satyabrata Mandal: ‘Remote Village Electrification through Renewable Solar energy: a Case Study of Sagar Island, West Bengal, India’, The International Journal of Engineering And Science (IJES),Volume 2, Issue 01, Pages 201-205,2013. [6] A.S.A.C. Diniz, E. D. França, C. F. Câmara, P. M. R. Morais, L. Vilhena: ‘The important contribution of photovoltaics in a rural school electrification program’, IEEE,2006 Recent Developments in Standalone PV System: For Rural Electrification ♦ 127
ICORE 2013
[7] Messenger, R. A. and Ventre, J., Photovoltaic systems engineering, CRC Press LLC, USA, 2004. [8] A Zahedi, The engineering and economics of solar photovoltaic energy systems, The New World Publishing, Melbourne, Australia, 2004. [9] Solartron Co., Ltd., Solar modules, http://www.solartron.co.th (last viewed 20 January 2007). [10] Leo Electronics Co., Ltd., Solarcon SE-T series,http://www.leonics.co. th/support/ brochure/2_set_en.pdf (last viewed 20 January 2007). [11] Hankins, M., Solar electric systems for Africa: a guide for planning and installing solar electric lighting systems in rural Africa, Commonwealth Science Council & AGROTEC, United Kingdom & Zimbabwe, 1995. [12] Hiranvarodom, S., Hill, R. and O’Keefe, P., ‘A strategic model for PV dis-semination in Thailand’, Progress in Photovoltaics: Research and Applications, Vol. 7, No. 5, 1999, pp. 409-419. [13] Ijumba N M, Utilization of Renewable Sources in Deep Rural Areas, IEEE AFRICON 2004, vol 2 pp 745-748 [14] Macias E, Ponce A, Photovoltaic Solar Energy in Developing Countries, IEEE Conference, May 2006, Vol. 2 pp. 2323-2326 [15] Kamalapur G D, Udaykumar R Y: ‘Electrification in rural areas of India and consideration of SHS’, 5th International Conference on Industrial and Information Systems, ICIIS 2010, Jul 29-Aug 01, 2010.
128 ♦ Recent Developments in Standalone PV System: For Rural Electrification
Biogas—An Ideal Source of Energy S.S. Sooch1 and Jasdeep Singh Saini2 1Director, School of Energy Studies for Agriculture, Punjab Agricultural University, Ludhiana–141004, Punjab, India 2Extension Assistant, Department of Civil Engineering, Punjab Agricultural University, Ludhiana–141004, Punjab, India E-mail: [email protected], [email protected]
Abstract—The recent scarcity of traditional sources of energy has generated an intense interest locating alternate fuels. One such fuel which can be made available both in the urban as well as rural areas is methane gas produced by the anaerobic digestion of the organic wastes. Anerobic digestion not only provides valuable fuel and enhances the fertilizer value of the waste, but also provides a convenient, safe, aesthetic and economical waste disposal method. Keywords: Waste Disposal, Biogas, Development of Biogas Technology
INTRODUCTION The anaerobic sludge digestion process is well known and is often practised in municipal sewage treatment plants. Now it has been extensively applied to the treatment of livestock and crop wastes. On a moderate scale it has been tried in Europe and in Africa. The experimental production of biogas was first attempted at the Indian Agricultural Research Institute (IARI) in 1939 in India. The gas can be put to various uses. Schmidt has reported its use as a source of power for farm tractors. The gas was first compressed into large storage tanks at pressure upto 5000 lbs/in and then pumped into two fuel cylinders fitted on each tractor and pressurized to 2840 lbs/in. Other uses for the gas including cooking, heating of farm buildings, drying, refrigeration and production of electricity. In India, the use of biogas as a fuel for an I.C. Engine, cooking and for lighting purposes has been reported. MATERIALS AND METHODS The amount of gas production as well as the fertilizer value of the final product of digestion depend to a great extent on the physical, chemical and biological properties of animal wastes. These properties, in turn, depend upon the characteristics (size, sex and breed) of the animal, the nature of the feed ration (its digestibility, the content of protein, fibre and other elements) and the environment. Nitrogen contained in proteins vary in digestibility depending on the source of protein. The animal wastes will contain all the ingredients of the feed, some of them in their original and some in chemically simpler form. On the average about 75 % of the nitrogen, 80 % of the phosphorous (as P2S5), 85 % of the potassium (as K2O) and about 40 to 50 % of the organic matter of the feed can be recovered in the manure. Although the moisture content of the excreta is only 70 to 80 %, ruing constitute 40 to 70 % of the fertilizer value of the total excrement. THE DIGESTION PROCESS The anaerobic digestion is a complex biological process during which the organic matter is decomposed by anaerobic bacterial organisms. It is neither necessary to use a pure culture for inoculation nor to maintain such a culture for inoculation. The bacteria capable of decomposing organic substances and of producting methane gas are found universally and abundantly in nature, particularly in decaying matter, e.g. human and animal excreta, The organic substrate need not be pure-any kind of mixture of organic substance is decomposed if the process is allowed to continue long enough. The entire substrate– carbohydrates, fats and proteins–with the possible exception of a small amount of fibre is broken during the digestion process and yields methane and carbon dioxide. During the digestion of organic matter, two phases of decomposition occur. These arc the liquefaction stage and the gasification stage. The first stage is brought about by a highly mixed culture of bacteria (the acid fermenters), the majority of which are saprophytes. These bacteria are capable of rapid reproduction and are not as sensitive to changes in environment as the bacteria responsible for the gasification stage. Extra cellular enzymes excreted by a saprophytes bring about the liquefaction of the organic matter. Complex carbohydrates are converted to simple sugars and alcohols; and fatty acids and proteins to peptides and amino acids. The first culture of bacteria carries the decarboxylation even future by converting all the alcohols, fatty acids and amino acids to volatile acids and water. The gasifying organisms (the methane fermentrers) are also a mixed culture of bacteria. They are strict anaerobies, the majority of them are nonmotile have low rates of reproduction, are extremely
ICORE 2013
sensitive to pH and temperature changes and require carbon dioxide for the reduction of the volatile acids to methane. The conversion of acids to methane is accomplished with interacellular enzymes secreted by the methane forming bacteria. In the absence of methane bacteria, the digestion process succeeds only in liquefying the sludge, often rendering it more offensive than the raw sludge. On the other hand, if under certain conditions, liquefaction proceeds at a faster rate than gasification, the resultant accumulation of acids inhibits further the methane bacteria, and the digestion process malfunctions. Thus both type of bacteria must be in proper balance. However, the optimum conditions for the gasifying bacteria are also suitable for the liquefying bacteria. Anaerobic digestion, in contrast to most fermentations, is a continuous process. Once the process starts, it is possible to feed the raw waste at one point and withdraw sludge from another continuously, while methane and carbon dioxide are given off at a steady rate. OPTIMUM CONDITIONS FOR GAS PRODUCTION The quantity and the composition of gas produced is greatly affected by temperature, the loading rate, the consistency of the waste fed to the digester, the accessibility of the substrate, the detention period, and the concentration of volatile acids in the digester. The optimum temperature for gas production is about 95°F (35°C). Digestion is not affected materially by an increase in temperature, but a small drop in temperature may result in an excessive accumulation of volatile acids and in digester failure. Maintenance of constant temperature is essential. The loading rate of a digester is expressed as the weight in Kgs of volatile (organic) solids fed per day per m3 of digester capacity and usually varies from 25 Kg/day/m3 to 30 Kg/day/m3. Mixing of the digester contents aids the digestion process by establishing uniform distribution of bacteria and food and by breaking up the floating scum layer formed from the accumulation of fibrous material at the top of the digester. The objection to the scum layer is that it constitutes a zone of substrate in which a high concentration of volatile acids develop. Usually if the volatile acids concentration rises much above, 200 mg/litre as acetic acid methane gas formation drops, and digestion ceases within two to three days. The optimum detention period is the digestion time required to obtain maximum gas production and figure of 40 days is often used. The range of solid concentration for satisfactory digestion is 8 to 10 % and the optimum pH range is 7 to 9. Presence of toxic substances in excessive quantities is deterimental to digestion because the growth of bacteria is retarded. PROPERTIES OF BIOGAS Biogas is a mixture of methane, carbon dioxide, nitrogen and hydrogen sulphide gases as well as some water vapours. The actual ratio of these gases depend upon the organic material used in the digestion. On an average, gas contains 55-60 % methane, 30–35 % carbon dioxide and 5 % nitrogen and hydrogen sulphide. The calorific value of the biogas is about 4700 Kcal. One m3 of biogas is equivalent 3.50 Kg of wood, 12.30 Kg of dung cakes, 1.6 Kg of coal, 0.62 litres of kerosene oil, 0.43 Kg of LPG, 0.52 litre of diesel etc. PROPERTIES OF MANURE LEFT AFTER DIGESTION The manure obtained after the anaerobic digestion of organic wastes is superior to the farm yard manure in several ways in the preparation of farm yard manure by the usual methods, losses of organic matter of the order of 50 per cent are known to take place, and over of the original nitrogen is also lost to the temperature. By digestion in the gas plant about 25% of the dry matter of the dung is converted into cumbustible gas and there is no loss of nitrogen. The digested manure contains 1.5 to 1.8 per cent nitrogen as compared to 0.5 per cent in farm manure. The farm yard manure as obtained from compost pits is a source of weeds. All weeds are, however, destroyed during the digestion process in the gas plant. Also the digested manure is available in only 30– 40 days while it takes 4 to 6 months to compost it in the usual manner.
130 ♦ Biogas—An Ideal Source of Energy
ICORE 2013
DISCUSSION Although experiments on biogas production have been reported to be started at the IARI in India in 1939. According to the information available from Ministry of New and Renewable Energy (MNRE), New Delhi, at present number of family size biogas plants installed in India are 45.45 Lakh and that in Punjab are 1.25 Lakh. The capacity of these plants vary from 2 m3 to 6 m3 of biogas per day and almost all of them are fixed dome type biogas plants. Also larger plants have also been installed in India for power generation, bottling of biogas and thermal application of biogas for cooking/Industrial use. At present, in Punjab around 100 biogas plants of larger capacity of fixed dome type have been operating for decentralized power generation applications at the dairy farms. One bigger biogas plant of capacity 1MW grid power generation has been operating at Ludhiana since the year 2000 and one more bigger biogas plant of capacity 4MW capacity grid power generation is under construction in Hoshiarpur District. Similarly two bigger biogas plants (one of capacity 600 m3/day at Abohar and one of capacity 1000 m3/day at Mukatsar) have been operating for bottling of biogas for the last 2-3 years and one more biogas plant of capacity 5000 m3/day for bottling of biogas is under construction at Bathinda. The main problems which restricts the use of such a technology are as follows: 1. The cost of installing such a plant is high. The recent (March, 2013) experience at the Punjab Agricultural University (P.A.U) shows that it may cost upwards of Rs. 35,000/-for a 4 m3/day capacity plant. 2. Since the gas is supplied at atmospheric temperature and pressure, the gas supply system gets cumbersome and inconvenient very soon as the distance of the plant from the point of use. This may either not be available in the residential area or it may not be preferred by the inhabitants. The sources of wastes and water should also be close to the site of the gas plant. 3. The products of digestion contains water vapours. These condense into the gas delivery pipe and clog it. Care must, therefore, be taken that this water be removed in some suitable manner. 4. The use of biogas is limited to cooking purposes. The large percentage of CO2 present in the biogas must he removed if it is to be efficiency used for producing mechanical energy. 5. The disposal of slurry from the digestion pit is another problem. Either it must be dried which means more space for drying beds is needed or it must be transported in the fluid from to the field which is not easy except when the plant is located close to the field and at such elevation that the slurry can flow to the field by gravity. CONCLUSION The relevance of the biogas in the modem technological World stems from considerations of energy scarcity and environmental protection both of which have become important recently. The digestion process for producing biogas has the following main advantages: 1. Methane gas which has commercial value is produced during the process. 2. The organic content of the waste is reduced by 50%. 3. The digested waste is a thick free-flowing fluid with no offensive order. 4. The waste is well stabilized and thus, needs no further treatment before final disposal. 5. The fertilizing constitutions of the raw waste are conserved and the fertilizer value of the digested solids is higher than that of the raw waste. Weeds in the waste are completely destroyed. 6. Rodents and files are not attracted to the end product of digestion. REFERENCES [1] [2] [3] [4] [5] [6]
Mittal, K.M. 1996 Biogas Systems: Principles and Applications, New Age International (P) limited Publishers, New Delhi. Biogas–A Rural Energy Source. (1985), Ministry of Non-Conventional Energy Sources Publication, New Delhi. Grewal N.S., Sooch S.S., Ahluwalia S and Brar G.S. 2000. Hand Book Biogas Tech, PAU, Ludhiana. Sooch S.S. 2010. Biogas Plants for Rural Masses. School of Energy Studies for Agriculture, PAU, Ludhiana. Akshyay Urja 2012. Energing the re way, Ministry of Renewable Energy Government of India www.mnre.gov.in.
Biogas—An Ideal Source of Energy ♦ 131
Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still J.S. Gawande1 and L.B. Bhuyar2 1Professor, Department of Mechanical Engineering, Rajarshi Shahu College of Engineering, Buldana, SGB Amravati University, Amravati, Maharashtra, India 2Professor, Department of Mechanical Engineering, PRM Institute of Technology & Research, Badnera, SGB Amravati University, Amravati, Maharashtra, India E-mail: [email protected], [email protected]
Abstract—The research work carried out so far in the field of solar desalination is related to the single basin solar still only. The effect of changes in design, climatic and operational parameters on the distillate yield have been studied but limited to the single basin solar still. In this work, we have tried to improve the performance by modifying flat basin of a solar still into a stepped type structure. The stepped type solar still selected in this case is having 8 number of steps of size 620mm(L) x 100mm(W) and total absorber area equal to 0.5093m2. The characteristic feature of stepped type solar still is that it provides an additional 40% absorber area in the same volume as compared to the single basin solar still. Some important parameters affecting the performance of a solar still were studied. The effect of depth of water, glass cover thickness, shape of the absorber area and various enhancements provided on the distillate yield were evaluated. While testing a particular parameter at an instant, other parameters were kept constant. Different depths of saline water over the basin (5mm, 7.5mm and 10mm); different thicknesses of the glass cover (3.5mm and 4mm); different shapes of the absorber plate (flat, convex and concave); different enhancements provided over the basin (sponges and fins) were tested under the same climatic conditions. After conducting the various tests under the same weather conditions; it has been observed that the productivity of the solar still increases with decrease of depth of water and glass cover thickness. Also higher distillate yield is obtained when convex shape is provided to the absorber surface and fins are provided over absorber surface. The economic analysis reveals that payback period of the various configurations is reduced to less than two years due to the modifications suggested in the design of a single basin solar still. Keywords: Stepped Type Solar Still, Depth of Water, Productivity, Convex Surface, Concave Surface, Sponges and Fins, Condensing Glass Cover
INTRODUCTION Desalination of ground brackish water by solar powered systems is a practical and promising technology for producing potable water in the regions which suffers from water scarcity especially in arid areas. In remote and arid areas in India, the abundant solar radiation intensity along the year and the available brackish water resources are two favorable conditions for using the desalination solar technology to produce the fresh water, especially for domestic use. Based on these conditions, a small scale solar powered desalination system has been constructed and operated. A solar still is a device that produces clean, drinkable water obtained from brackish water using the energy from the sun. This inexpensive device can easily be built using local materials. The basin type solar stills are the most popular solar distillation systems from a technical standpoint. However, it is known that, for the traditional solar stills, there are three serious shortcomings; the latent heat of condensation is not reused in the condensation-evaporation process; small evaporation surface and limited condensation process. The basin type solar still is tested by many researchers concluding that the maximum productivity is about 2–3 liters/m2day [1]. While basin type solar stills may be used for meeting individual requirements, they are not suitable for large distillation systems as may be required for a small village or a small industrial unit. Against this backdrop, research and development in the field of solar stills can be directed to obtain higher distillate yield of the required quality with maximum still efficiency by using advanced stepped type solar still. This is precisely what the proposed work is aimed at with the scope limited to the development of an efficient stepped type solar still with maximum distillate yield. This study aims to introduce the applicability of a stepped type solar still for supplying the fresh water to a small household residing in remote settlements where salty water is the only type of source available and electricity is scarce. The performance characteristics of a basin plate with two different depths of trays have been investigated by V. Velmurugan, S. Senthil Kumaran, V. Niranjan Prabhu and K. Srithar [2]. Fin type, sponge type and combination of fin and sponge type stepped solar still are analyzed in terms of productivity. Production of water increases by 80% when fin and sponge type stepped solar still is used
ICORE 2013
than flat type stepped solar still. Result shows that integration of fins in the basin plate gives more evaporation rate than adding sponges. Maximum productivity occurs when both these effects are combined. Experimental results show that productivity increased by 76%, 60.3% and 96% when fins, sponges, and combination of both the fins and sponges are used respectively. The parametric study of an active and passive solar stills using computer based thermal models by G.N. Tiwari, Vimal Dimri and Arvind Chel [3] indicates that the daily yield decreases with the increase of water mass in case of passive solar still. Whereas in case of active solar still it is found that as the depth of water is increased, the daily yield increases and after reaching the peak value at 11 cm; the daily yield decreases. These models were developed based on mass and energy balance equations. The variation of the daily yield for different thicknesses of the glass cover in the active and passive modes indicates that the daily yield decreases with the increase of the glass cover thickness. The performance of a solar still with different size sponge cubes placed in the basin of the still has been verified by B. A. Hijleh and H. M. Rababa’h [4]. The increase in distillate yield of the still ranged from 18% to 27% compared to an identical solar still without sponge cubes under the same conditions. The effects of sponge cube size, per cent volume of sponge, water depth, water salinity and the use of black coal and black steel cubes were also investigated. The study showed that the daily production of such a still can be greatly enhanced using sponge cubes. EXPERIMENTAL SET UP The experimental setup consisting of one of the eight different stepped type solar stills mounted on an iron stand with a saline water storage tank is as shown in Fig. 1. The relationship between the size of a solar still and its capacity depends upon its design and efficiency. On cloudy or rainy days, production stops so it is necessary to build a solar device to anticipate this drawback. Because this still is quite small, it is designed so that water collected can be drained into bottles. A separate hole was drilled in the sidewall of the solar still to fix thermocouples to sense the temperatures of water in the basin, absorber plate temperature and inner glass cover temperature. The entire unit was placed on an angle iron stand inclined at an angle of 20.50equal to the latitude of Buldana to the horizontal. The still requires unobstructed sunshine from early morning to late in the evenings. The solar still was oriented due south as the location lies in the northern hemisphere to receive maximum solar radiation throughout the year. The selected site at Buldana which is in the state of Maharashtra in India has an average sunshine ranging from 4.88kWh/m2/day in January to 6.63kWh/m2/day in April. A multi channel digital temperature indicator was provided to measure these temperatures. The collecting vessel is used for measuring distillate yield and a vane type digital anemometer is used for measuring wind velocity. The experiments were performed during the months of January 2012 to April 2012 when the sky was clear i.e. on sunny days. All experiments were started at 9 am local time and lasted up to 5 pm amounting to eight hours duration. The solar still A was used as a reference still to gauge the changes in still performance due to variation in design, climatic and operational parameters. All results presented herein are discussed in terms of distillate yield of the individual solar still. This was very important, since the experiments are conducted outdoors, and thus, the solar and ambient conditions could not be controlled or repeated from one experiment to another. Thus, no solar isolation measurements were made nor included in this work. The use of a reference solar still negated the effect of changes in ambient conditions. This facilitated the comparison between the different solar still configurations tested even though the experiments were conducted at different days under varying environmental conditions. When studying the effect of a given parameter, the other parameters were fixed at a reference value.
Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still ♦ 133
ICORE 2013
FIG G. 1: SCHEMATIC DIAGRAM OF THE EXPERIMENTAL SETUP
The sp pecifications of eight diffferent config gurations of stepped s typee solar still a and design parameters p of stepped d structure are illustrated d in Table 1 and a Table 2 respectively. The reeference valuue for each parameter p is shown in bold text in thee table. TABLE 1: SPECIFIICATIONS OF SOLA AR STILLS A TO H Type of Solar Still A B C D E F G
Depth of Water in mm 5 7.5 10 5 5 5 5
Glass Cover Thickness in mm 4 4 4 3.5 4 4 4
H
5
4
Design Pa arameters SShape of Abso orber Type of Enhancement Surface Flatt Flatt Flatt Flatt Connvex Conncave Flatt with sponges 48 no. of sponge cub bes of 70mm X 70mm 7 X 10 mm (H (6 nos. n per step) H) size are provvided over the to op surface of the basin Flatt with circular fins 552 no. n of circular alluminium fins of 4 mm (23x3=69 nos. per step) diameeter and 12 mm m length are sold dered at the top surface of thee basin
134 ♦ Effectt of Design, Clim matic and Opera ational Parametters on the Perfo ormance of Stepped Type Sola ar Still
ICORE 2013
TABLE 2: DESIGN PARAMETERS OF STEPPED STRUCTURE S. No. 1 2 3 3 4 5 7
Item Location Latitude of the location Longitude of the location Steps provided to the basin Size of each step Area of free water surface, Aw Area of still basin, Ab
Description Buldana (Maharashtra) 20.50480 N 76.21550 E 8 Nos. 620 mm(L) x 100 mm(W) = 0.062 m2 620 mm(L) x 100 mm(W) x 8 = 0.4960 m2 620 mm(W) x 101 mm(L) x 8 = 0.5093 m2
The block diagram representing the complete set up consisting of eight different stepped type solar stills interconnected with each other as well as with a saline water storage tank is shown in Fig. 2.
SALINE WATER STORAGE TANK
TH
TH
SOLAR STILL B
SOLAR STILL A
BK
BK
SOLAR STILL D
SOLAR STILL C
BK
BK
TH
TH
SOLAR STILL H
SOLAR STILL G
SOLAR STILL E
SOLAR STILL F
BK BK
BK
BK
FIG. 2: BLOCK DIAGRAM OF THE EXPERIMENTAL SETUP
RESULTS AND DISCUSSION The comparison of distillate yield for the stepped type solar stills of different configurations have shown in the following figures. The distillate yield of solar still A with flat basin surface was observed to be 1060 ml including the distillate obtained during off sunshine hours i.e. from 5pm to 10am on a typical day of 5th March, 2012. Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still ♦ 135
ICORE 2013
VARIATION N OF DISTILLATE YIELD FOR SOLAR STILL A, A B AND C The hourlyy variation off distillate yiield for solarr stills A, B annd C with va arying depthss is shown in Fig. 3. The effect of depth d of watter [5] on thee solar stills indicates thatt the distillate yield decreases with thhe increase of depth of o water. This is due to thhe fact that at a lower wate er depths, the specific heeat capacity of water is less due too decreased water mass.. This results in i increase inn water temp perature caussing faster evvaporation of water in the basin. Hence, the distillate yield d increases at lower wateer depths.
FIG. 3: DISTILLLATE YIELD WITH TIME OF THE DAYY FOR SOLAR STILLS A, B AND C N OF DISTILLATE YIELD FOR SOLAR STILLSS A AND D VARIATION
The hourlyy variation of distillate yield [6] for solar s stills A and D with varying v glass covers is re epresented in Fig. 4. It indicatess that, 3.5 mm m glass coover thicknesss of solar still s D increa ases distillate e yield as compared d with 4 mm glass g cover thhickness of soolar still A. The heat dissipatioon from the glass cover to the atmo osphere is duue to natural convection as well as radiation. As thermal coonductivity off glass cover is i low, overalll heat transfeer coefficient is very less re educing the heat transsfer betweenn glass coveer and the environment. e Hence part of latent heat of cond densation is accumulateed in air vapoor mixture, this phenomena a in thermal science is calleed as thermall inertia.
FIG. 4: DISTTILLATE YIELD WITHH TIME OF THE DAY A FOR SOLAR STIILLS A AND D
136 ♦ Effectt of Design, Clim matic and Opera ational Parametters on the Perfo ormance of Stepped Type Sola ar Still
ICORE 20 013
VARIATION N OF DISTILLATE YIELD FOR SOLAR STILLSS A, E AND F The hourlyy variation of distillate yield for solar stills A, E and a F with different shapes of absorb ber area is indicated in Fig. 5. Thhe distillate yield y obtaineed in solar still s E is 56.6 60% higher tthan still A. Due to the convexity of the absoorber surfacee, the depth of water is the least at the middle oof the basin steps. The water dep pth at the outter edges of the basin is the most. The e water masss at the midd dle of basin steps s being thinner; itss specific heat capacity is less. This results in increase in the water temp perature causing faster evaporation of at the middle of the basin stepss. t the chang ge in shape of the bassin in the form of conca ave surface, the distillatte yield is Due to 29.24% higher h than the solar still A. Due to thhe concavity of the absorber surface, the depth of o water is the most at a the middle of the bassin steps. Thee water dep pth is the lea ast at the outter edges off the basin steps. Thee water masss at the outeer edges of the basin steps being thhinner; its sp pecific heat capacity c is less. This results in increase in the water temp perature cauusing faster evaporation e of water att the outer edges of the basin steeps.
FIG. 5: DISTILLLATE YIELD WITH TIME OF THE DAYY FOR SOLAR STILLLS A, E AND F
As thee mass of wa ater is lesser in case of coonvex basin steps as com mpared to cooncave basinn steps; the distillate yield y obtaineed in case off convex typ pe solar still is higher in case c of convvex type sola ar still than that of conncave type solar s still. VARIATION N OF DISTILLATE YIELD FOR SOLAR STILLSS A, G AND H The hourlyy variation of o distillate yield y for sola ar stills A, G and H provvided with various enhanncements is shown in Fig. F 6. Whenn sponges arre used, the water surfacce area increeases and diistillate yield d increases. For consta ant solar inteensity in the days d of experiments withh and withouut usage of ssponges in thhe stepped solar still, the distillatee yield obtainned is 25.47% % higher tha an the distillate yield of thhe solar still A. A pe absorber plate absorrbs more therrmal energy due to increa ase in exposure area and d increases Fin typ the sensib ble heat in sa aline water, which in turnn increases thhe distilled yield y which is 33% highe er than the distillate yield y of the solar s still A.
Effecct of Design, Cliimatic and Operrational Parame eters on the Perrformance of Ste epped Type Sollar Still ♦ 137
ICORE 2013
FIG. 6: DISTILLATE YIELD WITH TIME OF THE DAYY FOR SOLAR STILLLS A, G AND H
VARIATION N OF DISTILLATE YIELD WITH H WIND VELOC CITY FOR SOLA AR STILLS A, D, D E AND H The effectt of wind veloocity on the distillate d yielld of four of the eight moost efficient solar stills A, D, E and H during seccond week off March 2012 is shown inn Fig. 7. For comparison c p purpose, distiillate yield obtained o at 1:00 pm during d the weeek is selecteed. It is observed from the figures tha at increasing the wind velocity tends to increasee the distillatte yield of thhe solar stills. As thee wind velociity increases, the convectiive heat trannsfer coefficieent from the glass cover to ambient air increasses and simuultaneously thhe glass coveer temperatuure decreasess. This happeens due to thhe fact that the coefficcient of external heat excchange is deependent on the t wind veloocity according to the relation (1) 5.7 3.8 5 Due to this, the teemperature difference between b watter surface and a the glasss cover incrreases and ultimately the yield of the solar stilll increases as a compared to stagnant ambient air cconditions.
FIG. 7: 7 VARIATION OFF DISTILLATE YIELDD WITH WIND VELLOCITY FOR SOLARR STILLS A, D, E A AND H
COST AND ECONOMICS The produuction rate of o distilled water is propoortional to thhe area of the t solar still; which means that the cost per liiter of waterr produced iss nearly the same regard dless of the size s of the fa abricated stiill. The cost of distilled d water prooduced by thhe stepped type solar still s dependss upon threee main facto ors namely, capital cost; energy soource and op peration and maintenance e cost. 138 ♦ Effectt of Design, Clim matic and Opera ational Parametters on the Perfo ormance of Stepped Type Sola ar Still
ICORE 2013
The capital cost gets reduced if locally available materials are used. As the solar energy is available free of cost, it has no effect on the total cost of the solar still. The operation and maintenance cost in this case is almost negligible. The payback period of the experimental setup depends on overall cost of fabrication, maintenance cost, operating cost and cost of feed water. The operating and maintenance costs are negligible. The overall fabrication cost was ` 8723. It was not necessary to take into account the operation and maintenance cost and the cost of feed water, which was almost negligible. Compared to purchasing comparable quantities of bottled water, the average return on investment on different configurations of solar still is shown in Table 2. If the health costs of contaminated water are taken into consideration, the payback period can further be reduced to a greater extent. Solar distillation is the cheapest way to clean water for a household and is quite economical as compared to reverse osmosis and electric distillation. TABLE 2: PAYBACK PERIOD CALCULATIONS S. No. 1 2 3 4 5
Item Overall fabrication cost to be considered for 0.5 m2 absorber area (`) Cost per liter of distilled water (`) Productivity of the solar still Cost of distilled water produced per day (`) Payback period
A 8723
B 8723
C 8723
D 8723
E 8723
F 8723
G 9023
H 9423
10 10 10 10 10 10 10 10 1.06 0.91 0.82 1.39 1.66 1.37 1.33 1.41 Liters Liters Liters Liters Liters Liters Liters Liters Per Day Per Day Per Day Per Day Per Day Per Day Per Day Per Day 10.6 9.1 8.2 13.9 16.6 13.7 13.3 14.1 823 Days
958 Days
1064 Days
628 Days
525 Days
637 Days
678 Days
668 Days
Where,
SUMMARY AND CONCLUSION In this work, eight different configurations of stepped type solar stills with different modes of operations have been analysed. The various conclusions drawn from the tests carried out are as follows. It has been observed that the distillate yield of solar still A was greater than solar still B and C by 14.15% and 22.64% respectively. As depth of water goes on increasing, the distillate yield produced per unit area of absorber surface goes on decreasing. Also the distillate yield of solar still D with glass cover thickness 3.5mm was 31.13% higher than that of solar still A with 4mm glass cover thickness. It can be concluded that lesser the glass cover thickness, higher is the distillate yield. The distillate yield of convex type solar still E was 56.6% higher than that of A and distillate yield of concave type solar still F was 29.24% higher than that of A. The maximum yield was obtained when the convex type of absorber surface was provided. In addition, the distillate yield of concave type absorber surface was greater than that of flat type absorber surface. Hence, the convex type surface area gives the maximum distillate yield for the given design and climatic conditions of the stepped type solar still. It has been further observed that the distillate yield of sponge type solar still G was 25.47% higher than that of still A and distillate yield of fin type solar still H was 33% higher than that of still A. Result shows that integration of fins in the basin plate gives more evaporation rate than adding sponges. The wind velocity is related to climatic conditions of the stepped type solar still. The effect of wind velocity on the distillate yield of the four of the eight most efficient solar stills A, D, E and H indicates that wind blowing over the glass cover causes faster evaporation. As the wind velocity over the solar still increases, the distillate yield of the solar still increases continuously. Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still ♦ 139
ICORE 2013
The average return on investment on eight different configurations of solar stills shows that due to the modifications introduced in the design of the stills; the payback period has been reduced from 823 days for solar still A to 525 days for solar still E. The fresh water requirement is 3–5 liters per person per day. Considering a small family of four persons consisting of two adults and two kids, total fresh water required is 15 liters per day. The distillate yield obtained by the stepped type solar still provided with convex shape of the absorber area is the highest and it was 1660ml on a typical day of March 2012. The area of the still basin in this case was 0.5108m2. Hence if a stepped type solar still having convex surface of area 4.5m2 is provided; the fresh water requirements of a small family residing in remote settlements where salty water is the only type of source available and electricity is scarce can be met. Since other desalination plants are uneconomical for low capacity fresh water demand, under these situations, stepped type solar stills are viewed as a means to attain self-reliance and ensure regular supply of water. ABBREVIATIONS •
TH–Thermocouple
•
BK-Beaker
REFERENCES [1] H.E. Gad, S.M. El-Gayar and H.E. Gad, “Performance of a solar still with clothes moving wick”, 15th International Water Technology Conference, IWTC, 28–30 May 2011, Alexandria. [2] V. Velmurugan, S. Senthil Kumaran, V. Niranjan Prabhu and K. Srithar, “Productivity enhancement of stepped solar still–Performance analysis”, Thermal Science, Vol. 12, No. 3, 2008, pp. 153–163. [3] G.N. Tiwari, Vimal Dimri and Arvind Chel, “Parametric study of an active and passive solar distillation system: Energy and exergy analysis”, Desalination, 242, 2009, pp. 1–18. [4] Bassam A/K Abu-Hijleh and Hamzeh M. Rababa’h, “Experimental study of a solar still with sponge cubes in basin”, Energy Conversion and Management, 44, 2003, pp. 1411–1418. [5] Gawande J.S., Bhuyar L.B. and Deshmukh S.J., “Effect of depth of water on the performance of stepped type solar still”, International Journal of Energy Engineering, World Academic Publishing, Vol.3, Iss.4, 2013, pp.137–143. [6] Gawande J.S. and Bhuyar L.B., “Effect of glass cover thickness on the performance of stepped type solar still”, International Journal of Innovative Research in Technology and Science, Vol. 1, No.3, 2013, pp. 19–26.
140 ♦ Effect of Design, Climatic and Operational Parameters on the Performance of Stepped Type Solar Still
Nano Materials Coated Photovoltaic Device for Enhanced Solar Energy Conversion P.H.V. Sesha Talpa Sai1, J.V. Ramana Rao2, Devarayapalli K.C.3 and K.V. Sharma4 1Department
of Mechanical Engineering, Jawaharlal Nehru Technological University, Hyderabad–500075, India for Nano Science and Technology, Jawaharlal Nehru Technological University, Hyderabad–500075, India 3Bhaskar Engineering College, Moinabad, Hyderabad, India 4Department of Mechanical Engineering, JNTUH College of Engineering, Manthani, Karimnagar–505 212 E-mail: [email protected]
2Center
Abstract—The objective of the project is to develop a prototype nano material coated device aimed at improving the efficiency of the photovoltaic cell. So far the low conversion efficiency has been the cause for limited applications of this technology, even an increase in conversion efficiency by a few percent makes this technology a viable energy source. Experiments with sol-gel prepared TiO2 spray coatings on silica solar cell base are conducted to evaluate the performance of the solar cell. A comparatively low cost sol-gel preparation and coating techniques were experimented in this work by taking into consideration of nano material properties, processes and energy transfer acquired through modeling. The proposed nano material coated photovoltaic device can be easily and economically manufactured with better performance and enhanced conversion efficiency even under cloudy conditions. Surface morphology was studied using SEM and the electrical properties of nano coated photovoltaic cell are characterized using sun simulator under standard test conditions to validate the results of the experiments. Proposed device will be useful for fabrication of low cost rural solar installations with enhanced conversion efficiency. Keywords: Photovoltaic Cell, Titanium Dioxide, Sol-Gel, Nano Coatings
INTRODUCTION Among various renewable energy resources, solar energy is reliable, clean and environment friendly [1]. This promising and sustainable energy is abundant and can replace the fast degrading fossil fuels. Need for new and clean energy is further enhanced due to the critical situation of environmental pollution that is caused due to the conventional energy production [2]. Even now many rural areas are having shortage of electricity either in full or partial. Renewable energy devices like customized solar power systems, Solar panels, biomass energy systems, biogas plants can be installed in rural areas to provide clean and cheaper energy for meeting their lighting, cooking, agricultural and other power requirements [3]. Solar photo voltaic power generation is fast emerging as an alternative to solve these energy issues. A further growth in PV power generation can be expected with increased technologies and researches on improving the efficiency of the PV power devices. Sun produces 3.9 X 10 26 watts of energy per second and 1386 watts of that fall on a square meter of Earths atmosphere. Even less of that reaches the earth’s surface. This energy that fall on earth’s surface is many times more than the energy consumed by the world. This energy can be utilized to generate electricity without producing pollution and dangerous wastes. Photovoltaic (PV) cells can directly convert solar incident energy to electricity by absorbing incident photon energy from sun [4,5] as represented in Fig. 1.
Sun light
n-type junction p-type
Photons Electron flow - ++ Hole flow
FIG. 1: SCHEMATIC REPRESENTATION OF PHOTOVOLTAIC DEVICE FOR ENERGY CONVERSION
ICORE 2013
Multi crystalline cells domina ate the marrket and arre widely used u in PV devices manufactured elatively low w and they a are reliable with good worldwidee as the cost of producttion of mc-Sii cells are re conversionn efficiency. How ever thhe conversionn efficiency of o mc-Si silica a cells used iin PV device es is limited to 10% due to various factors inclluding optica al losses caussed by reflecction of solarr incident rad diations on ditions (shading effect byy buildings, trrees and clouuds) causes the cell surface. The efffect of non-uniform cond change in the current-vvoltage (I-V)) characteristtics of the sysstem, and coonsequently thhe device pe erformance is affected d [6,7]. Anti Reflection R (ARR) coatings on o the mc-Si cells is emerrging as bestt technique to reduce the e reflection losses, imp prove transm mittance and thereby increase the connversion efficciency of the cell. It is a mandatory m to coat the solar cell with w at least a single layyer anti refle ection (SLAR) coating at tthe front surfface of the duce the losss due to refleection as thee light generated current density out put of the cell c largely cell to red depends on the absorption and transmittance t e of the incid dent light. Using SLAR w with optimum refractive index and d thickness ca an reduce reeflectance loosses to minim mum at speccific wave length. The die electric AR coating fiilm shall alsso compatible with the cell surface e and provide good suurface passivvation [8]. Schematicc representation of the effect of AR cooating on the e cell surface is shown in FFig. 2.
FIG. 2:: GRAPHICAL. REPPRESENTATION OFF IRRADIANCE. AN ND REFLECTANCE ON PLAIN (A) AN ND TIO2
Titaniuum dioxide is having many adva antages like optimum range of reefractive ind dex values (1.9 to 2..3), high mecchanical and d chemical reesistance and d longer stability. More over it has over 90% transmittance over sun’s broad bannd spectral range. r It is no on toxic, non corrosive annd can be de eposited at low tempeeratures at atmospheric a pressure typ pical amorphous TiO2 film ms can be deeposited belo ow 350oC. o Meta stab ble crystallinne structure of o anatase can c be form med above 350 3 C and sstable crysta alline rutile phase cann be obtaineed at temperratures abovve 800oC. Thhe phase of the t material strongly inflluences the functional properties of o the TiO2 films. Anatasee TiO2 having g optimal refractive indeex as mentionned above u in solarr cell AR coatting applicattions. Stoichio ometric thin film f AR coatings by spray y paralysis is widely used exhibit exxcellent optical propertiess for a singlee layer anti reflection coa ating on multii crystalline silicon s solar cell. [9, 10 0] Sol-geel TiO2 AR coatings are significant and most widely used forr solar cell applications and a having potential to replace more expennsive conventtional vacuum processes. It is becauuse of the fact f that it s equipm ment with neeither vacuum m nor high te emperature processes. High homogenneous final requires simple films can be produced d and many kinds of coa atings can be e obtained by b varying thhe process parameters p [11,12]. MATERIALSS AND METHO ODS TiO2 filmss were prepa ared by sol--gel spray coating technnique. Titaniuum iso propooxide as a precursor p is purchased d from Sigma a Aldrich. Ab bsolute isoproopyl alcohol was used as a solvent annd hydrochlo oric acid as a catalystt. The mixturee was sonicated vigoroussly in a ultra sonicator for 1 hour. Thee solution obttained was sprayed on o the mc-Si silicon cell using u fine sp pray pilot guun. Before thhe spray thee cell was tre eated with diluted HFF solution to eliminate thee unwanted contaminatio c ons on the cell surface annd the cell was washed 142 ♦ Nano Materials Coatted Photovoltaicc Device for Enh hanced Solar En nergy Conversio on
ICORE 2013
with ethanol and dried for 10 min [13,14]. TiO2 was sprayed on the cell as explained above and the cell is subsequently annealed at 350oC to obtain homogeneous anatase structure. Different kinds of films were prepared by changing the parameters of the sol-gel synthesis process and the films were coated and heat treated and tested for optimizing the desired results. Sol-gel flow chart and reaction pathway for synthesis of TiO2 solution is schematically represented in Fig.3 and Fig. 4 respectively. Precursor
Alcohol
Acid + Water
Sonication
Sol
Gel FIG. 3: FLOW-CHART OF SYNTHESIS ANTI REFLECTION SOLUTIONS
FIG. 4: REACTION PATHWAY OF TITANIUM DIOXIDE ANTI REFLECTION SOLUTION
Two solutions of concentrated and diluted TiO2 are synthesized and coated on mc-Si solar cell of size 50 x 25 mm and the results are compared with plane and TiO2 coated cells. RESULTS AND DISCUSSION The physical appearance of the plain and TiO2 coated cells and the corresponding SEM iages are shown in Fig 5. SEM images are obtained to access the presence of the anatase TiO2. Fig. 5A shows the surface morphology of plain cell. Fig. 5 B and Fig. 5 C shows the images and surface morphology of the cells coated with concentrated and diluted TiO2 solution prepared by sol-gel acid catalyzed process. (A )
(B )
(C )
FIG. 5: EDS AND SEM MICROGRAPHS OF PLAIN MC-SI SOLAR CELL (A), TIO2 DILUTED AR SOLUTION COATED MC-SI SOLAR CELL (B), TIO2 CONCENTRATED AR SOLUTION COATED MC-SI SOLAR CELL (C)
Manufacturing of mc-Si solar cells involves the laser texturisation on the top surface of the silica wafer.During this treatment the top surface is subjected to some damages and few unwanted materials may be deposited on the cell surface which causes the increased recombination or junction shunting that Nano Materials Coated Photovoltaic Device for Enhanced Solar Energy Conversion ♦ 143
ICORE 2013
results in reduced cell performance. These defects can be neutralized by post laser etching process. Selection of proper etching agent with desired dilution ratio is essential in maintaining the texturized surface of the cell with improved optical and electrical properties. Two high texturization some times becomes a hurdle for homogeneous deposition of the materials on the surface. Titanium dioxide exhibits good adherence and compatibility to the textured surface of the mc-Si surface. Fig. 5 B shows the formation of anatase TiO2 on the cell coated using diluted solution prepared by sol-gel synthesis. The improved optical and electrical properties of this cell coated with diluted TiO2 are attributed to the increased surface area and the reduction in the reflectance of the incident light. The phenomenon of the destructive interference of light is the reason for the reduced reflection losses and the bluish physical appearance of the TiO2 coated cell. Intermediate cracks observed in between the particles are helpful in scattering of incident light with in the cell surface and in reorientation of the electron hole pairing. However the cell coated with concentrated TiO2 shows relatively higher efficiency than the plain cell and lower efficiency than the cell coated with diluted TiO2 solution. The additional layer thickness due to larger particle size of the TiO2 is the reason for lower efficiency of the cell than the cell coated with diluted TiO2. Electrical parameters like short circuit current (Isc), open circuit voltage (Voc), Maximum power point current (Impp), Maximum power point voltage (Vmpp), fill factor (FF) and conversion efficiency (η) are obtained for plain cell and cells coated with concentrated and diluted TiO2 are tabulated in the table 1. Cells coated with concentrated and diluted TiO2 shows conversion efficiencies of 11.34% and 13.05% respectively compared with 10% efficient plain cell. TABLE 1: 1-V CHARACTERISTICS OF PLAIN AND TIO2 COATED MC-SI SOLAR CEL Solar Cell Plain silecon cell TiO2(concentrated)-coated cell TiO2(diluted)-coated cell
Jsc (mA/cm2) 29 30 31
Voc (mV) 500 500 500
Impp 25 27 29
35
Fill factor (%) 68.9 75.6 82.5
Efficiency (%) 10 11.34 13.05
Plain mc-Si solar cell TiO2 conc. mc-Si solar cell
30 Current (mA)
Vmpp 400 420 450
TiO2 dilut. mc-Si solar cell
25 20 15 10 5 0
100
200
300
400
500
600
Voltage (mV)
FIG. 6: I-V DATA OF PLAIN MC-SI SOLAR CELL, TIO2 CONCENTRATED COATED MC-SI SOLAR CELL AND TIO2 DILUTED COATED MC-SI SOLAR CELL
Corresponding I-V curves are plotted in Fig. 6 which shows the conversion efficiencies of the plain cell and cells coated with TiO2 solutions. CONCLUSION This study presents the low cost sol-gel methodology as an alternative to expensive vacuum processes used for preparation of conventional AR coatings. Diluted TiO2 single layer AR coating with moderate concentration has shown 30.5 % improvement in conversion efficiency of a mc-si solar cell. In future, different types of other nano materials can be experimented for better enhancement in conversion efficiency. A prototype nano material coated device with enhanced conversion efficiency can be easily fabricated with the findings of this study which is useful in fabrication of low cost rural solar systems. 144 ♦ Nano Materials Coated Photovoltaic Device for Enhanced Solar Energy Conversion
ICORE 2013
ACKNOWLEDGEMENT The authors sincerely acknowledge the valuable support extended by Sri J Bhaskar Rao, Chairman and Sri J V Krishna Rao, Secretary, J.B. Educational Society Hyderabad, India. The authors thank Mr.Chandra Sekhar CEO, Asses solar Ltd-Hyderabad, India for their technical support. Support extended by Dr K. Venkateswar Rao, and Prof. Shilpa Chakra, Center for Nano Science and Technology, JNTU, Hyderabad and Dr Sarala Devi, IICT-Hyderabad, is greatly acknowledged. REFERENCES [1] D.Hocine, MS. Belkaid, M.Pasquinelli, L.Escoubas, J.J. Simon, G.Riviere, A.Moussi, international conference on Renewable Energies and Power Quality, Spain-2012. [2] Naresh kumar malik, Jasvir Singh, Vineet Singla, International journal of latest trends in engineering and technology, 3 (2013) 167. [3] www.sesi.in [4] Roger A. Messenger, Jerry Venture, Photovoltaic systems Engineering, CRC press, 2nd edition 2005. [5] Antonio Luque, Steven Hegedus, Handbook of Photovoltaic Science and Engineering, Vol-1, Wiley, 2003. [6] J. BANY and J. APPELBAUM, Solar Cells, 20 (1987) 201. [7] L.A.Dobrzanski,A.Drygala, Jounal of achievements in Materials and Manufacturing Engineering 31 (2008) 77. [8] Barbara Swatowska, Tomasz Stapinski, Kazimierz Drabczyk, Piotr Panek, Optica Applicata, Vol. XLI (2011) 487. [9] L.Andronic, S.Manolache, A.Duta, Journal of optoelectronics and advanced materials 9 (2007) 1403. [10] C. Martinet, V.Paillard, A.Gagnaire, J.Joseph, Jounal of non-crystalline Solids 216 (1997) 77. [11] G. San Vicente, A.Morales, M.T. Gutierrez, Thin Solid Films 391 (2001) 133. [12] G. San Vicente, A.Morales, M.T. Gutierrez, Thin Solid Films 403-404 (2002) 335. [13] L.A. Dobrzanski,A.Drygala, Jounal of achievements in Materials and Manufacturing Engineering 17 (2006) 321. [14] A. Ibrahim, A.A. El-Amin, International Journal of Renewable Energy Research 2 (2012) 356.
Nano Materials Coated Photovoltaic Device for Enhanced Solar Energy Conversion ♦ 145
Microalgae: Bio Energy Technologies for Sustainable Future Piyush Choudhary1 and Varteka Tripathi2 1Oil
and Natural Gas Corporation, New Delhi & Indian Institute of Technology, BHU Varanasi 2Center for Environmental Planning and Technology’ (CEPT), Ahmedabad, India E-mail: [email protected], [email protected]
Abstract—Fossil fuels represent around 80% of the world energy demand and are relatively cheap to produce. On the other hand they have caused some socio-political problems and added CO2 into the atmosphere, contributing to the climate change. Modern Bioenergy sources are viewed as components of a low-carbon, energy-secure future. By reducing dependence on imported fuel and providing new employment opportunities, bioenergy production has the potential to stimulate local economies in developing countries. In general, there is growing consensus that if significant emission reductions in the transport sector are to be achieved, biofuel technologies must become more efficient in terms of net lifecycle greenhouse gas (GHG) emission reductions while at the same time be socially and environmentally sustainable. This paper focuses on microalgae biomass based Bioenergy solutions which has high potential of making sustainable energy future for days to come. Microalgae-to energy is being studied in the last decades and some industrial plants are being built to start its commercialization. Improved sustainability assessment of biomass energy project types is guided based on priority-setting for sustainable future energy reform efforts. This paper intends to assess qualitatively the microalgae bioenergy production. Keywords: Bioenergy, Sustainability, Climate Change, Carbon Emissions, Sustainable Development, Microalgae etc
INTRODUCTION Based on the global classification on the basic needs of the society, sustainance has importance and relevance. For the developing countries relevance of sustainability has been prioritized in the sequence of social, economic and at last environmental aspects. There is a sustainability gap between the three pillars of sustainable parameters which affects humanity in the area of health, energy dependence and environment concern. To protect the world and reduce the gap there is an urgent need to focus on environmental sustainability i.e. globally the consideration of shift of focus from fossil fuels to various other options of renewable energy which could maintain the energy sustainability gradually without affecting the demand and growth. To protect the 2ºC rise in global temperature, there is urgent need to reduce the CO2 emissions to the level of 14 Giga ton by 2050, which could be achieved by increasing the share of global renewable energy and the rate of energy efficiency. In order to cut the energy-related CO2 emissions to half of the current levels by 2050, the International Energy Agency has suggested that bioenergy use should triple by 2050, to approximately 135 exa joule (EJ) per year (IEA 2010). A wide range of additional conversion technologies are under development, offering prospects of improved efficiencies, lower costs and improved environmental performance. However, expansion of bioenergy also poses some challenges. The potential competition for land and for raw material with other biomass uses must be carefully managed.
FIG. 1: 2°C TRAJECTORY
Bioenergy could sustainably contribute between a quarter and a third of global primary energy supply in 2050. It is the only renewable source that can replace fossil fuels in all energy markets–in the production of heat, electricity, and fuels for transport. Many bioenergy routes can be used to convert a range of raw biomass feedstocks into a final energy product. Technologies for producing heat and power from biomass are already well-developed and fully commercialised & are called as 1st generation routes to biofuels for transport. However the increasing criticism of the sustainability of many first-generation biofuels has raised attention to the potential of so-called second-generation biofuels. Depending on the feedstock choice and the cultivation technique, second-generation biofuel production has the potential to provide benefits such as consuming waste residues and making use of abandoned l and.
ICORE 2013
Among biodiesel, bio ethanol, biogas which compromise the various form of Bioenergy, the micro algae system emerges as the important area for exploration. The usage of micro algae for the various Bioenergy solutions is one of the sustainable approaches to lessen the dependence on the fossil fuels, thus indirectly leading to reduction in greenhouse gas emissions and increasing the environmental sustainability. CONCEPTS OF BIOENERGY AND ITS CLASSIFICATIONS Renewable energy sources will play an important role in the sustainable development in the future, protection of the environment and the security of energy supply being the main driving forces in the short term. During the various COP conclusions there were an agreement to “urgently and substantially increase the share of renewable energy sources” while the Kyoto protocol implies for a reduction of 8% of the greenhouse gas emissions (corresponding to around 600million tons of CO2-equivalent) between 2008 and 2012 (compared to 1990 level). To reach the objective of increasing the share of renewable energy sources to 12% in 2010, all the different technologies, including geothermal, solar and ocean, have to be supported. Further increasing the use of biomass will be necessary, and biomass is expected to cover as much as 8% of the energy supply in 2010. Biomass accounts for 98% of total renewable heat production. Furthermore, biomass is the only renewable energy source that can produce competitively priced liquid fuels for transport. Reduced need to import oil (67% of oil is for road transport purposes), increased security of supply, reduction of emissions, improved local environment and new jobs are the primary benefits.
FIG. 2: BIOMASS RESOURCES VS. BIO ENERGY UTILIZATION CHART
Bio-energy programme clearly spells on “Energy, environment and sustainable development” and “Cleaner energy systems and economic and efficient energy for a competitive era.” Priority would be to the proposals, which employ an innovative approach to the large-scale production and use of bioelectricity including CHP applications, and to innovative technologies that result in gains in conversion efficiency. Research and technological development are crucial for the development of bio-energy. Biomass based energy systems can be built on a wide variety of feedstocks and use many different conversion technologies to produce solid, liquid or gaseous fuels. It is possible to upgrade biomass to obtain fuels that are identical to or have properties close to those of fossil fuels. This minimises the need to adapt end-use technologies. The whole of Bioenergy growth is to maintain the sustainable development of the society as whole. TABLE 1: PROJECTS CLASSIFICATIONS OF BIOENERGY IMPLEMENTATION PROCESS Biomass Project Socioeconomic drivers Short rotation crops Sustainable forestry Combustion & co-firing Thermal Gasification of biomass Pyrolysis From municipal solid waste Biogas and landfill gas Bioenergy system analysis
Outline Socio-economic impacts and improving impact assessment process Assessing sustainable short rotation crops & market development Conventional forestry to stakeholders for sustainable production Biomass consumption for the use for CHP plants Biomass gasification & promoting cooperation Resolving technical issues & implementation of fast Pyrolysis Creating a network for information exchange & dissemination Network on anaerobic digestion to upgrade biomass as source Supply unbiased analyses & strategic decision policy issues Microalgae: Bio Energy Technologies for Sustainable Future ♦ 147
ICORE 2013
BIOENERGY FOR SUSTAINABLE DEVELOPMENT Sustainable development is oriented towards meeting the needs of the present without compromising the ability of future generations. The relevance of SD is based on two key concepts first concept of essential needs on priority and second the idea of state of technology limitations on the environment sustainability. The progress of any nation today is measured in terms of its efforts towards the achievement of the Millennium Development Goals (MDGs). Many factors converge in making bioenergy a key component and a viable opportunity in the great effort towards the achievement of the Millennium Development Goals (MDGs). Although the sustainable access to energy is not treated as a priority in itself in the MDGs, most of them have a direct energy implication, particularly Goal 1 (Eradicate extreme poverty and hunger) and Goal 7 (Ensure environmental sustainability) (FAO, 2005).According to the WSSD Johannesburg Declaration, energy must be considered a human need on a par with other basic human needs (clean water, sanitation, shelter, health care, food security and biodiversity) Energy Services, running water lighting cooking heating cooling.
FIG. 3: BIOENERGY—A SUSTAINABLE DEVELOPMENT PERSPECTIVE
Social, Economic and environment are the three important pillars of sustainable development. It is the need of hour to act in all these areas, finding long-term development solutions that combine economic growth with environmental protection and energy efficiency Millions of people worldwide live below the poverty line, who have no access to basic needs such as food, health care, and have inadequate shelter. With the rapid increase of the global population, the disparity between rich and poor is becoming greater. Secondly the large dependence on fossil fuels is contributing towards emissions which has already crossed the limit of 450ppm.Thus in order to overcome that disparity and reduce emissions and to maintain energy security, sustainable development is the only path to be followed. Energy dependence shows how much an economy relies on the imports to meet its domestic energy demand. Bioenergy is the most widely used renewable source of energy in the world, providing about 10% of the world primary energy supplies. Out of renewables, one of potential RE source is pressumed as conversion of Biomass to Energy, i.e. Bio Energy. While fossil fuels still dominate the global energy supply with a combined share of 81%, renewable energy sources have the potential of becoming the dominant sources of energy for coming generations. Bioenergy is already around twice as large as nuclear energy in the world. The Renewable energy corresponds to 13% of global energy supply and consists of 10% bioenergy, 2% hydropower and 1% wind, solar and geothermal energy. The Biomass is considered as greenhouse gas (GHG) neutral. The carbon dioxide (CO2) released from biomass during production of bioenergy is from carbon that circulates the atmosphere in a loop through the process of photosynthesis and decomposition. Therefore, production of bioenergy does not contribute extra CO2 to the atmosphere like fossil fuels. This leads to its benefits as contribution to global primary energy supply; reductions in greenhouse gas emissions, environmental benefits; improvements in energy security and trade balances, opportunities for economic and social development in rural communities; and scope for using wastes and residues, reducing waste disposal problems, and making better use of resources.
148 ♦ Microalgae: Bio Energy Technologies for Sustainable Future
ICORE 2013
BIOENERGY CONVERSION TECHNOLOGIES Bioenergy is the most widely used renewable source of energy in the world, providing about 10% of the world primary energy supplies. Biomass energy is derived from plant-based material whereby solar energy has been converted into organic matter. Sources include forestry and agricultural crops and residues; byproducts from food, feed, fiber, and materials processing plants; and post-consumer wastes such as municipal solid waste, wastewater, and landfill gas. There are many bioenergy routes which can be used to convert raw biomass feedstock into a final energy product. Biomass can be used in a variety of energy-conversion processes to yield power, heat, steam, and transportation fuels (Figure 1). Traditional biomass already provides the main source of energy for household heating and cooking in many developing nations. It is also used by food processing industries, the animal feed industry, and the wood products industry, which includes construction and fiber products (paper and derivatives), along with chemical products made by these industries that have diverse applications including detergents, fertilizers, and erosion control products. Several conversion technologies have been developed that are adapted to the different physical nature and chemical composition of the feedstock, and to the energy service required. Upgrading technologies for biomass feedstocks (e.g. pelletisation, torrefaction, and pyrolysis) are being developed to convert bulky raw biomass into denser and more practical energy carriers for more efficient transport, storage and convenient use in subsequent conversion processes. The production of heat by the direct combustion of biomass is the leading bioenergy application throughout the world, and is often costcompetitive with fossil fuel alternatives. The bioenergy conversion technologies are explore based on three important paramters viz, the available feedstock, technology for conversion and end product requirement. Fundamentallay there are four types of conversion technologies are currently available, each appropriate for specific biomass types and resulting in specific energy requirement: TABLE 2: DIFFERENT CONVERSION TECHNOLOGIES AND DESCRIPTION Conversion Technology Thermal Conversion Thermochemical conversion Biochemical conversion Chemcial conversion
Description Direct combustion, pyrolysis and torrefaction Heat and chemical processes as gasification (example as syngas). Break down biomass into liquid fuels, chemicals, heat, and electricity using enzymes, bacteria, and other microorganisms.eg anaerobic digestion and fermentation. Transform biomass into other forms by chemical agents. Eg. Transesterfication process
To make use of energy available in biomass, it is necessary to utilize technology to either release the energy directly or to convert it into other forms. Various technologies exist to convert biomass into power, heat, and fuels is explained as under:
FIG. 4: BIOENERGY CONVERSION TECHNOLOGIES AND ENERGY PRODUCTS Microalgae: Bio Energy Technologies for Sustainable Future ♦ 149
ICORE 2013
For a more energy efficient use of the biomass resource, modern, large-scale heat applications are often combined with electricity production in combined heat and power (CHP) systems. Different technologies exist or are being developed to produce electricity from biomass. Co-combustion (also called co-firing) in coal-based power plants is the most costeffective use of biomass for power generation. CRITERIA AND ASSESSMENT OF VARIOUS SUSTAINABLE DEVELOPMENT BIOENERGY METHODS Bioenergy is the most widely used renewable source of energy in the world, providing about 10% of the world primary energy supplies. Sources include forestry and agricultural crops and residues; byproducts from food, feed, fiber, and materials processing plants; and post-consumer wastes such as municipal solid waste, wastewater, and landfill gas. Biomass can be used in a variety of energy-conversion processes to yield power, heat, steam, and transportation fuels. Sustainability indicators were developed under the following three pillars, noting interlinkages between them. To take forward the sustainable methods for Energy solutions requires a detailed process by which bioenergy opportunities can be assessed along with a set of few of the requisite of planning. These planning must incorporate the various associated aspects of the planning of a project on Bioenergy. The triple bottom approach is core for the assessment of the bioenergy methods for overall sustainable development. The indicators were intentionally crafted to report on the environmental, social and economic aspects of sustainable development. The indicators are meant to guide analysis at the domestic level and to inform decision-making that encourages the sustainable production and use of bioenergy as a means towards meeting national goals of sustainable development. TABLE 3: TRIPLE BOTTOM APPROACH SUSTAINABILITY INDICATORS FOR BIOENERGY Environmental Social Economic Themes Indicators for Triple Bottom Approach: Based on Bio Energy • Greenhouse gas emissions, • Social Accountability • Resource availability • Productive capacity of the land • Stakeholder participation • Conversion technologies • Air quality & ecosystems, • Price and supply of a national food basket, • Financing • Water availability, • Labour conditions • Economic viability • Efficiency • Rural development • Technological capabilities, • Biological diversity, Land-use change, • Access to energy • Energy security • Human health and safety. • Diversification of Energy Sustainability Indicators Lifecycle GHG emissions Land allocation Productivity Fuel substitution Soil quality National food supply Improved energy balance Resource harvesting Income generation GDP added Emissions of non-GHG Job security Shift of fuel economy Water use and efficiency Time spent for collection of Biomass by people Training projection Water quality Bioenergy access Diverse energy portfolio Biological diversity Mortality reduction due to non-smoke fuel Distribution of Bioenergy Land use change Occupational health Flexibility of Energy
For way forward action in Bio energy a process by which bioenergy opportunities can be assessed along with a set of resources to assist for zeroing the project/ activitiy. The major perquisites for the assessment are mainly four steps briefed at Figure–XX. In order to propose long-term RD&D priorities, it is necessary to define a long-term vision of the energy supply. This vision currently focuses on a biofuel economy, recognizing that there are several pathways to biofuel production. However it is necessary to continue focusing on sustainable ways to cultivate large amounts of biomass worldwide without hampering food supply, the local ecology and biodiversity. At the same time, efforts should continue on the bio refinery approach, through which biomass for products, food and energy would become an integral part of the economy. Based on these aspects, one of potential area of bioenergy is under aggressive Research and Development through the effective and efficient exploitation of Microalgae as a source of Biofuel globally.
150 ♦ Microalgae: Bio Energy Technologies for Sustainable Future
ICORE 2013
FIG. 5: FOUR STEP ASSESSMENT TOOL FOR THE IDENTIFICATION OF BIOENERGY PROJECT
REVIEW OF THE POTENTIAL OF MICRO ALGAE AS A SOURCE OF BIOFUEL Transport fuels dervided from Biomass feedstock can enable reneable energy to be used in transport without the need to sustanaitially modify exising vehicles or the fuel distribution infratsturure. Assessment tool proposes that Conversion of Microalagae into Bioenergy is the high potential way forward to serve the dual purpose of cambating the climate change and maintaining the energy security issues globally. Algae to biofuel has been widely discussed among experts in the petroleum industry and conservationist who are looking for a more reliable and safer source of energy that is both renewable and easy to attain. One of the key reasons why algae are considered as feedstock for oil is their yields. It has reported that algae yield 30 times more energy per acre than land crops such as soybeans, and some estimate even higher yields up to 15000 gallons per acre. Aside from keeping the earth clean and free form pollution, these algal biofuels help to utilize a resource that is available in abundance just waiting to be harnessed and exploited.
FIG. 6: TECHNOLOGICAL PATH FOR PRODUCING BIOFUEL AND OTHER PRODUCTS Microalgae: Bio Energy Technologies for Sustainable Future ♦ 151
ICORE 2013
Once the algae are grown and harvested, there are a different ways of extracting the oil. Whichever method is used for extraction, the resulting product is a vegetable oil called 'green crude', similar to crude oil, which is further transformed into biodiesel fuel through a process called 'transesterification'.
FIG. 7: SCHEMATIC OF MICROALGAE BIOFIXATION OF CO2 TECHNOLOGY
FIG. 8: INTEGRATED LIPID PLATFORM FOR SECOND GENERATION BIOFUEL PRODUCTION
CO2 sequestration by cyanobacteria and green algae are receiving increased attention in alleviating the impact of increasing CO2 in the atmosphere. They, in addition to CO2 capture, can produce renewable energy carriers such as carbon free energy hydrogen, bioethanol, biodiesel and other valuable biomolecules. Biological fixation of CO2 are greatly affected by the characteristics of the microbial strains, 152 ♦ Microalgae: Bio Energy Technologies for Sustainable Future
ICORE 2013
their tolerance to temperature and the CO2 present in the flue gas including SOX, NOX. However, there are additional factors like the availability of light, pH, O2 removal, suitable design of the photobioreactors, culture density and the proper agitation of the reactor that will affect significantly the CO2 sequestration process. This review underlines the existing technical viability for the development of biofuels from microalgae as a renewable energy resource and for mitigation of GHG related impacts of petroleum derived fuels. The achievable high yields for both lipids and biomass, combined with some useful co-products if purposefully exploited, could enhance algae’s economic viability as a source for biofuels. Phototrophic production is the most effective in terms of net energy balance. Therefore, large-scale production of microalgae for biofuels will increase the availability of these products. Overall, with the current demand for renewable fuels, especially for use in the transportation sector, there is a need to develop a range of sustainable biofuels resources as the combined mix will be a significant step towards the replacement of fossil fuels. Continued development of technologies to optimise the microalgae production, oil extraction and biomass processing has the capacity to make significant contributions towards this goal. CONCLUSION Bioenergy projects will provide greater diversification and income opportunities for agriculture, agroindustries and forestry: they will increase the access of small rural industries to energy services; and will enhance the value of rural resources, encouraging private and public sector participation and investments. Modern bioenergy systems will increase both access to and reliability of energy services for households in rural areas, thus improving the quality of life. Algae can play an important role in the bio based economy. Algae are efficiently cultivated in places that are unsuitable for agriculture and where nature is not harmed. Sustainable production of biodiesel, but also many other products such as proteins, colorants and raw material for bio plastics is achievable. To achieve profitable cultivation of algae, the production efficiency must be increased three times and costs must be reduced ten times. In addition, besides oil for biofuel, other useful substances such as proteins must also be extracted from the algae. Optimization of algae cultivation could also be explored by research team compare them for effective and efficient results. Based on these results and data obtained from the laboratory, the team will develop a new reactor design for application on commercial scale. The use of renewable energy resources is an action for economic development, which will bring benefits in the coming decades. It is a consequence of striving towards sustainable economic development, stimulated by a growing concern about the impacts of global warming. Renewable energy will play a growing role in the world’s primary energy mix. It is estimated that, Bioenergy renewables, will grow faster than any other primary energy source, at an average rate of 7 % per year over the period to 2030. Last but not least Bioenergy sustainability needs to be considered in broader context of integrated land-use planning/natural resource management, new lessons can be learned, and old knowledge revived and shared. REFERENCES [1] Bioenergy Australia 2011 Conference IEA Bioenergy ExCo Workshop November 2011. [2] Bioenergy–A Sustainable And Reliable Energy Source: A review of status and prospects. [3] Advancing Bioenergy for Sustainable Development & Guideline for Policymakers and Investors, The International Bank for Reconstruction and Development/THE WORLD BANK, 2005. [4] The Global Bioenergy Partnership Sustainability Indicators for Bioenergy, First Edition, Dec 2011. [5] DOE, National Algal Biofuels Technology Roadmap, MAY 2010. [6] Bioenergy Assessment Toolkit, NREL/TP-6A20-56456, October 2012. [7] A Review of the Potential of Marine Algae, as a Source of Biofuel in Ireland, February 2009, Sustainable Energy Ireland [8] Environmental Engineering and Management Journal, http://omicron.ch.tuiasi.ro/EEMJ/ September/October 2008, Vol.7, No.5, 617–640. [9] ALGAE-BASED BIOFUELS: A Review of Challenges and Opportunities for Developing Countries, May 2009. [10] Eurec, Research Priorities, For Renewable Energy, Technology By 2020 And Beyond. [11] European Commission community research 1999-2002, European Bioenergy projects. [12] Daniel Fishman, Rajita Majumdar 2008, National Algal Biofuels Technology Roadmap. Microalgae: Bio Energy Technologies for Sustainable Future ♦ 153
ICORE 2013
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22]
Atmanand, Amit K. Gupta, and Rishabh Raman 2009, Energy and Sustainable Development-An Indian Perspective. Uwe R. Fritsche Oeko-Institute 2012, Sustainable Bioenergy Key Criteria & Indicators. U.S Department of Energy, Energy Efficiency & Renewable Energy 2011, Biomass program. Al Darzins, Philip Pienkos 2010, Current Status and Potential for Algal Biofuels Production. IUCN 2008, Implementing Sustainable Bioenergy Production a Compilation of Tools and Approaches. Baljeet Singh Saharan, Deepansh Sharma Microbial Resource Technology Laboratory, Department of Microbiology, Kurukshetra University, Kurukshetra 2013, towards algal biofuel production: a concept of green bio energy development Uwe R. Fritsche, Katja Hunecke WWF 2008, Sustainable standards of Bioenergy. Global Bioenergy Partnership (GBEP) 2011, The global bioenergy partnership sustainability indicators for bioenergy first edition. Bio-global/ con CISE net Aquatic Biomass: Sustainable Bio-energy from Algae?-Issue Paper–November 2009 Energy Production from Microalgae Biomass: The Carbon Footprint and Energy Balance, São Paulo–Brazil–May 22nd to 24th-2013
154 ♦ Microalgae: Bio Energy Technologies for Sustainable Future
Development and Performance Studies of a Community-Size Building-Material Solar Cooker R.C. Punia1, V.K. Marwal2, S. Mahavar3, P. Rajawat3 and P. Dashora3 1Govt.
Polytechnic College, Sikar, R.C. Khaitan Polytechnic College, Jaipur 3,4,5Department of Physics, University of Rajasthan, Jaipur E-mail: [email protected] 2Govt.
Abstract—This paper presents a study of community-size building-material solar cooker. This community solar cooker is fixed structure with housing made of local masonry bricks and cement plaster. This community solar cooker is capable of cooking for about 25 persons. Thermal performance studies of community solar cooker have been carried out at Jaipur (26.92°N, 75.87°E). The highest plate stagnation temperature, without reflector and no-load condition reached 142.1°C and for most of the time of test, the absorber plate temperature remains above 100°C. The obtained results has been used to calculate two figures of merit F1 and F2, as well as overall utilizable efficiency ηu, specific ts and characteristic tc boiling times. Two figures of merit F1 and F2 are found 0.117°C-m2/W and 0.416 respectively which are as per recommended values of BIS. The adjusted cooking power 209W and the loss coefficient 1.91W/°C, place this cooker in the region of large size and good insulation cooker as per International Standards. Keywords: Community Solar Cooker, Building-Material, Figures of Merit, Cooking Power
INTRODUCTION Looking in to the present energy scenario solar cookers seems to be a good alternative for cooking, as it saves the fuel which may be used for other purpose. Various types of solar cookers developed so for can be categories in mainly three categories viz. hot box type, heat transfer type and concentrating type. Hot box types of cookers are the most favored ones in the society due to their easy operation and reasonable cost structure. The solar cookers generally studied are suitable for meeting the food requirement of about 4-5 persons. Theses cookers are not suitable for community purposes e.g. hostels, hotels, temples, canteens, and restaurants etc. where various types of conventional fuels are used for cooking. Literature survey reveals that very few efforts have been made towards the development of nonconcentrating community size solar cookers, several attempt were made in early 1990’s [1, 2] after that no attention has been paid towards the development of this type of solar cookers. The casing of systems studied earlier has been made of mild steel. As the use of metal increases the cost of the system and production of these metals requires a large amount of fossil fuel which adds the burden on limited fossil fuel resources. With this aspect in mind authors have also developed and studied purpose specific cookers [3, 4]. Some of the researchers have developed; solar collector’s mainly employing building material [5, 6, 7, 8, 9]. These systems were made using different type of building materials and were wholly prepared through these building materials. Metal absorber tray or separate insulation were not used hence the performance of these systems was poor, though cost was also less. A building material animal feed cooker has been developed by Singh [10].The cooker was made of RCC and it had an aluminum tray as absorber surface. The cooker was without reflector and was capable of cooking 5 kg of animal feed per day. Further two models of the solar cooker for animal feed were developed by Nahar et al. [11] through clay, locally available material, exfoliated vermiculite and cement tiles. The cooker was capable of preparing 2 kg of animal feed per day. The cost of this animal feed cooker was quite low but these had low efficiencies, as the stagnation temperatures are around 80°C. Dasora et al. [12] developed a building material housing solar cooker in which the absorber tray was fabricated through used oilcans and glass wool as insulation. The maximum plate Stagnation temperature reached in this cooker without load was around 135°C which is fairly high. The thermal performance studies of designed solar cooker have been conducted as per the test procedure suggested by Mullick [13], Funk [14] and Khalifa [15]. DESIGN DETAILS OF COMMUNITY SIZE SOLAR COOKER This community size solar cooker is fixed structure so tracking is not possible. The cooker has been designed in such a way that length to width ratio for the reflector and glass window is about four, so maximum radiation reflected from the reflector falls on the glass window. This has helped in eliminating the tracking, which is required in squared apertures. The system has been installed at an open place at the rooftop of Department of Physics, UOR, and facing south. The housing of this cooker has been built with local masonry bricks and cement plaster. The inner dimensions of the casing are 205 cm × 67 cm × 16 cm, thickness of walls is 15cm. A trapezoidal tray of aluminum sheet of thickness 0.3 mm is used as absorber tray, which has aperture area 195 × 57 cm2 and base plate area 185 × 47 cm2. The slant height of
ICORE 2013
walls is 12 cm and walls are inclined at 26.92° with vertical. The upper surface of the absorber plate is painted with black board paint; therefore it is capable of absorbing most of the incident solar radiation. The top aperture of the solar cooker is covered with double glass covers separated by 13 mm. The inner toughened glass sheet is of 5 mm thickness and upper glass is a plane glass sheet of thickness 4mm. Since aperture of the system is very large therefore, for convenience of handling the glaze has been designed in two pieces, each of dimension 100 × 60 cm2. The double glass cover minimizes the rate of heat loss through the top of the cooker. Around 5 cm thickness glass wool insulation of thermal conductivity 0.034 W/m-K is filled at the bottom and sides of the cooker to reduce the heat losses through the bottom and sides of the cooker. Three plane mirrors each of dimension 70×74 cm2are fixed in the lid of the box. The lid is hinged at the top of cooker and used to reflect the solar radiation onto the cooker aperture. The mirror tilt angle can be adjusted through a mechanical arrangement to maximize the amount of solar radiation onto the cooker top. Cooking vessels are cylindrical in shape with diameter of 19 cm and depth 7 cm and these are made of stainless steel. Fourteen such vessels can be kept inside the cooker. Fig. 1 gives the sectional view of community size solar cooker. EXPERIMENTAL STUDY The on field studies of the community solar cooker were performed on the rooftop of Physics Department at University of Rajasthan, Jaipur (26.92°N, 75.87°E). Thermal performance test were undertaken under no-load conditions to determine F1 and sensible heat tests were also carried to determine F2. Sensible heat test were repeated for different loads to see the effect of Mf on F2. Experiments began at 10:00 a.m. and were continued up to 4:00 p.m. Indian Standard time (IST). Measurements were taken at intervals of 10 min for no-load and sensible heat tests. During all the experiments, the solar radiation intensity on a horizontal surface was measured using a pyranometer (Nation Instruments Ltd. Calculta, instrument no. 0068), CIE-305 thermometer with point contact thermo couples (Accuracy 0.01r C) was used to measure the temperatures of different parts of the cooker, absorber plate, cooking fluid, vessel, air enclosed, upper and lower glass. The ambient temperature was measured using mercury thermometer (Accurecy 0.1 C) placed in the ambient chamber and the wind speed was measured by an anemometer (Prova instruments inc. AVM-03).
FIG.1: SECTIONAL VIEW OF A BOX TYPE SOLAR COOKER
FIG. 2: PHOTOGRAPH ON-FIELD INSTALLATION OF DEVELOPED COMMUNITY SOLAR COOKER 156 ♦ Development and Performance Studies of a Community-Size Building-Material Solar Cooker
ICORE 2013
In order to calculate the cooking power and hence the adjusted cooking power, experiments were performed for community solar cooker. During the cooking power experiments fourteen containers were used and the temperature of water was measured in five pots placed at different positions in the system. The average temperature was then calculated and was used to calculate the cooking power and the adjusted cooking power using the method discussed earlier. One reflector on south age of the system was used whenever required as per test conditions. The system is off south oriented and no tracking is possible because it is fixed one. RESULTS AND DISCUSSION Fig. 2 represents the diurnal variation in the weather conditions, namely insolation and ambient temperature and the transient response of community solar cooker under no-load conditions. It is clear from the figure that transient response closely follows insolation patterns. The temperatures of cooker elements increase with time of day until they achieve their maximum values at 12.50 p.m. The maximum temperatures of the absorber plate Tp, lower glass Tgl and upper glass Tgu are 142.1 C, 93.5 C and 67.2 C, respectively. For most of the time of test, the absorber plate temperature remains above 100 C. At the time when absorber plate attains highest temperature the insolation and ambient temperature are 888 W/m2 and 37.9 C respectively. Using the above results first figure of merit has been calculated and is found to be 0.117 C-m2/W. This ensures the high optical efficiency and low heat loss factor for community solar cooker. The typical diurnal variation of temperatures of the cooker elements, insolation and ambient temperature for 8 kg water load are depicted in Fig. 3. The results of this figure are used to calculate the second figure of merit. The value of F2 using Tw1 = 63.8 C, Tw2 = 95.1 C, τ = 5400 sec., average values for the ambient temperature 36.4 C and insolation 806.6 W/m2is found to be 0.416 for the designed community solar cooker. The effect of load on F2has also been studied. Fig. 4 shows temperature variations of various cooker elements, with 10 kg of water load for this load the value of F2is 0.464, which is higher than previous value. The load dependence of various thermal performance parameters is evident in Table 1. Further the results of Fig. 3 and Fig. 4 are used to calculate the specific boiling time ts, characteristic boiling time tc and overall utilizable efficiency ηu. It is clear from Table 1 that ts and tc decrease with increasing load and ηu increases with increase in load. Values of ts and tc decrease from 25.13 and 21.72 (min-m2/kg) to 21.09 and 20.51 (min-m2/kg) when load increases from 8 kg to 10 kg. However ηu increases from 21.6% to 25.5% when load increases from 8 to 10 kg. In fact the system efficiency increase with increase in load but it should be noted that besides good efficiency, adequate rate of rise of temperature is also essential for good cooking performance. Therefore depending on the size of any cooker there would be an optimum load for which both rate of rise of temperature and efficiency have adequate values. Using Fig. 4 and relation obtained by Nahar [1] the efficiency were calculated and was found to be 27.6%, which is better than the only reported value 24.6% for non-concentrating community solar cooker. In the cooking experiments, the test standard requirements were applied. The system with intercept area of 1.5 m2 was loaded with 10.5 kg of water (7kg/m2). The adjusted cooking power and temperature difference between the cooking fluid and ambient air were calculated every 10 min. interval. Fig. 5 represents the adjusted cooking power as a function of temperature difference. The adjusted cooking power a function of temperature difference fits in the following relation 304.6 1.91 ∆ The value of regression coefficient (R2) of above equation is 0.907 which satisfies the test standard. The values for initial cooking power and adjusted cooking power and adjusted cooking power at a temperature different of 50 C are 304.6 and 209 Watts, respectively. The loss coefficient from the slope of the regression line is found to be 1.91 W/ C. The values of cooking power place the community solar cooker in the region of large size and good insulation cooker, as per international standard.
Development and Performance Studies of a Community-Size Building-Material Solar Cooker ♦ 157
ICORE 2013
TABLE 1: VARIOUS THERMAL PERFORMANCE PARAMETERS AND THEIR LOAD DEPENDENCE Parameter (i) F1 (ii) F2 (i) ts (ii) tc (iii) nu Adjusted cooking power (P) Efficiency (η)
3 4 Tp Tgu
Tamb Tgl I 07/04/12
150
Tamb 1000 900 800 700 600 500 400 300 200 100 0
130 110 90 70 50 30
150 Temperature (°C)
2
Radiation Intensity (W/m2 )
1
Temperature (°C)
Load 8 kg 0.117 C-m2/W 0.416 25.13 min-m2/kg 21.72 min-m2/kg 21.6% -
1000 900 800 700 600 500 400 300 200 100 0
110 90 70 50 30
1000
120
800 700
100
600 80
500 400
60
300 200
40
100
09/04/12
310
p
Linear (p)
290 270 250 230 210
y = -1.912x + 304.5 R² = 0.907
190 170
3:00
2:00
1:00
12:…
11:…
0 10:…
20
FIG. 3: MEASURED TEMPERATURE PROFILES AND RADIATION INTENSITY FOR THE VARIOUS COOKER ELEMENTS AND RADIATION INTENSITY DURING SENSIBLE HEATING TEST WITH BOILING 8 KG WATER LOAD.
Adjusted cooking power (W)
Tgl Tw
Ta
Radiation intensity (W/m2 )
02/05/12 Tgu
900
Temperature (°C)
Tw
Time (IST)
FIG. 2: MEASURED TEMPERATURE PROFILES AND RADIATION INTENSITY FOR COOKER ELEMENTS DURING STAGNATION TEST.
140
Tp 08/04/12
130
Time (IST)
Tamb Tp I
10 kg 0.117 C-m2/W 0.464 21.09 min-m2/kg 20.51 min-m2/kg 23.5% 209 27.6% Radiation Intensity (W/m2 )
S. No.
Time (IST)
FIG. 4: MEASURED TEMPERATURE PROFILES AND RADIATION INTENSITY FOR THE VARIOUS COOKER ELEMENTS AND RADIATION INTENSITY DURING SENSIBLE HEATING TEST WITH BOILING 10 KG WATER LOAD.
150 0
20
40
60
Temperature difference (°C)
80
FIG. 5: RELATIONSHIP BETWEEN ADJUSTED COOKING POWER AND TEMPERATURE DIFFERENCE FOR COMMUNITY SOLAR COOKER.
CONCLUSION Calculated values of F1 and F2 for the developed community solar cooker are found to satisfy BIS, which ensures good cooking performance. Efficiency increases with increase in load but it should be noted that besides good efficiency, adequate rate of rise of temperature is also essential for good cooking performance. The values of the initial cooking power, adjusted cooking power and heat loss coefficient at a temperature difference of 50 C are within the range of these parameters obtained by Funk for large size and good insulation cooker. The developed solar cooker would prove to be of great importance for community cooking at hotels, hostels and Mid-Day-Meal programme conducted in govt. schools. 158 ♦ Development and Performance Studies of a Community-Size Building-Material Solar Cooker
ICORE 2013
REFERENCES [1] Nahar N.M., 1991, Renewable Energy, 2, 421. [2] Nahar N.M., Gupta J.P. and Sharma P., 1994, Applied Energy, 48, 295. [3] Dashora P., Sengar N. and Gupta M, 2005b, In the Peer-Reviewed Proc. Of Solar World Congress organized by International Solar Energy Society (ISES), Orlando, Florida, USA Paper No. 1327. [4] Mahavar S., Rajawat P., Marwal V. K., Punia R. C. and Dashora P., 2012, Modelling and on-field Testing of a Solar Rice Cooker, Energy, 2012,404. [5] Hoffman M., Rodan K., Feldman M. and Saposnik D. S., 1983, Solar Energy, 30, 275. [6] Peck M.K. and Proctor D., 1983, Solar Energy, 31, 183. [7] Nayak J.K., Sukhatme S.P., Limaye R.G. and Bopshettty S.V., 1989, Solar Energy, 42, 45. [8] Chaurasia P. B. L., 2000, Energy, 25, 703. [9] Bilgen E. and Richard M.A., 2002, Solar Energy, 72, 405. [10] Singh R., 1992, in New Dimensions in Renewable Energy–Proc. of NSCE, Ed. N. K. Bansal, IIT, Delhi, 48. [11] Nahar N.M., Gupta J.P. Sharma P., 1996, Renewable Energy, 7 (1), 47. [12] Dashora P., Sengar N., Pareek S.K., 2005a, In the Peer-Reviewed Proc. of Solar World Congress organized by International Solar Energy society (ISES), Orlando, Florida, USA, Paper No. 1197. [13] Mullick S.C., Kandpal T.C. and Sexena A.K., 1987, Solar Energy, 39, 353. [14] Funk P.A., 2000, Solar Energy, 68, 1. [15] Khalifa A. M. A., Taha M. M. A., and Akyurt M., 1984, Solar and Wind Technology, 1, 81.
Development and Performance Studies of a Community-Size Building-Material Solar Cooker ♦ 159
Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households in Neh Region of India Mahendra S. Seveda College of Agricultural Engineering and Post Harvest Technology, Central Agricultural University, Ranipool, Gangtok-737135, Sikkim, India E-mail: [email protected] Abstract—A low cost photovoltaic powered forced convection solar dryer was developed and evaluated for drying of Chilli in the conditions of NEH region of India. The PV powered forced convection solar dryer consists of solar PV module of area (280×230) mm2 connected with drying chamber. The solar panel is tilted to an angle of 45 with respect to horizontal. It was connected to the exhaust fan provided at the opposite wall of the dryer with the help of electric wire. The drying chamber was made up of M.S. angle, G.I. sheet and glazing material with the frame size of (700×700) mm2, opposite wall size of (700×700) mm2 and front side size of (890×700) mm2 with the inclination of 45. The two drying trays were contained inside the drying chamber which is made up of aluminium angle, aluminium strip and steel wire mess. The lower and upper tray was fitted at the height of 150 mm and 350 mm from the base of the dryer. The size of lower and upper tray was (680×490) mm2 and (680×270) mm2. Air inlet is provided in the one fourth area of the base with the diameter of 600 mm. The drying chamber is insulated with thermocole of 10 mm thickness. At the one side of the dryer an insulated door is provided to facilitate the loading and unloading of the trays. The dryer was capable of holding about 6 kg of chillies per batch. Average air temperature attained in the solar dryer was about 40oC higher than the ambient temperature. Drying of chilli in a PV powered forced convection solar dryer reduces the moisture content from around 80.2% (wet basis) to the final moisture content about 10.00% in 32 h. Keywords: Solar Dryer, Forced Convection, Drying Rate, Chilli drying
INTRODUCTION Drying is an essential process used all over the world for the preservation of farm produce. It helps in reducing the water activity of the produce to a level below which deterioration does not occur for a definite duration. Sun drying is still widely used in many tropical and subtropical countries [1]. Sun drying is the cheapest method, but the quality of the dried products is far below the international standards. Improvement of product quality and reduction of losses can only be achieved by the introduction of suitable drying technologies [2]. However, this method of drying is extremely weather dependent and has the problems of contamination, infestation, microbial attacks, etc., thus affecting the product quality. The rate of drying depends on various parameters such as solar radiation, ambient temperature, wind velocity, relative humidity and initial moisture content, type of crops, crop absorptivity and mass of product per unit exposed area. This form of drying has many drawbacks such as degradation by wind, blown, debris, rain and insect infestation, human and animal interference that will result in contamination of the product [3]. Drying rate will reduce due to intermittent sunshine, interruption and wetting by rain. Solar drying is a well-known food preservation procedure used to reduce the moisture content of agricultural products, which reduces quality degradation over an extended storage period. Several Solar crop dryers are a viable alternate to open sun drying, and have several advantages. They offer better quality dried products as the products are protected; drying time is also significantly reduced. Several types of solar dryers have been developed and used to dry a variety of agricultural product. Solar dryers using natural convection or forced circulation have been investigated to overcome these problems [4]. For commercial applications, the ability of the dryer to process continuously throughout the day is very important to dry the products to its safe storage level and to maintain the quality [5]. Solar dryers may be classified according to the mode of air flow as natural convection and forced convection dryers. Natural convection dryers do not require a fan to blow the air through the dryer. Solar drying may also be classified into direct, indirect and mixed-modes. In direct solar dryers the air heater contains the materials and solar energy passes through a transparent cover and is absorbed by the materials. Essentially, the heat required for drying is provided by radiation to the upper layers and subsequent conduction into the material bed. In indirect dryers, solar energy is collected in a separate solar collector (air heater) and the heated air then passes through the material bed, while in the mixedmode type of dryer, the heated air from a separate solar collector is passed through a material bed and at the same time, the drying cabinet or chamber absorbs solar energy directly through the transparent walls or roof [6].
ICORE 2013
Active solar dryer are also called force convection solar dryer. It is generally agreed that forced convection distributed solar dryer are more effective and more controllable than the natural-circulation type. External force is provided to the solar dryer such as fans may be powered with utility electricity if available, or with a solar photovoltaic panel [7]. In a solar photovoltaic powered, fan is directly coupled to the solar module increasing the solar radiation increase the module’s output, thus speeding up the fan. Chilli is one of the most important commercial crops of India. It is grown almost throughout the country. There are more than 400 different varieties of chillies found all over the world. It is also called as hot pepper, cayenne pepper, sweet pepper, bell pepper, etc. It is botanical name is “Capsicum annuum”. The world’s hottest chilli “Naga Jolokia” is cultivated in hilly terrain of Assam in a small town Tezpur, India. India is the largest producer of chilli and annually production is around 1.1 million tones. India also has the maximum area dedicated to the production of this crop. Chilli is a universal spice of India [8]. It is cultivated in all states and union territories of the country. Chillies are cultivated seasonally but are consumed throughout the year. Chillies are grown on soils with light sands to well drained clay [9]. Silty and clay loam soils are better, while water-logged and alkali is not suitable. In Indian subcontinent, chillies are produced throughout the year [10]. Two crops are produced in kharif and rabi seasons in the country. Chilli grows best at 20–30 C. Growth and yields suffer when temperatures exceed 30 C or drops below 15 C for extended periods. Chilli is normally dried on open ground with no shelter provided. It is common to see rooftops covered with chilli during the summer months. This practice results in poor final quality associated with huge financial losses to the poor farmers. MATERIALS AND METHODS Working Principle of Photovoltaic Powered Forced Convection Solar Dryer This dryer was designed to make use of solar energy during day time. As the solar radiations falls on the glass surface some of incidence solar radiation is reflected back to atmosphere and remaining enter inside the dryer, the thermocole sheet inside the solar dryer was black painted to absorb the incident solar radiation. Solar radiation gets absorbed by the produce and inside drying chamber, resulting in an increase of dryer temperature [11]. This process produces temperature difference between the inside and outside cabinet air. Inside the chamber, heated air pick up moisture from the product as it passes through the trays and taken away by the air entering into the drying chamber from the inlet provided at the bottom of the dryer and comes out through the outlet provided on the opposite wall of the drying chamber with the aid of the supplied PV powered dc fan. Experimental Set Up The photovoltaic powered forced convection solar dryer consists of solar panel of area (280×230) mm2 connected with drying chamber. The solar panel is tilted to an angle of 45 with respect to horizontal. It was connected to the exhaust fan provided at the opposite wall of the dryer with the help of electric wire. The drying chamber was made up of M.S. angle, G.I. sheet and glazing material with the frame size of (700×700) mm2, opposite wall size of (700×700) mm2 and front side size of (890×700) mm2 with the inclination of 45. The two drying trays were contained inside the drying chamber which is made up of aluminium angle, aluminium strip and steel wire mess. The lower and upper tray was fitted at the height of 150 mm and 350 mm from the base of the dryer. The size of lower and upper tray was (680×490) mm2 and (680×270) mm2. Air inlet is provided in the one fourth area of the base with the diameter of 600 mm. The drying chamber is insulated with thermocole of 10 mm thickness. At the one side of the drier an insulated door is provided to facilitate the loading and unloading of the trays. The dryer was capable of holding about 6 kg of chillies per batch. The schematic diagram of PV powered forced convection solar dryer shown in Fig. 1.
Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households ♦ 161
ICORE 2013
Exhuast fan M.S. angle GI sheet Plain glass Solar pannel
Door
Electric wire
Tray Caster wheel Air inlet
FIG. 1: SCHEMATIC DIAGRAM OF PORTABLE PV POWERED FORCED CONVECTION SOLAR DRYER
Construction of the Photovoltaic Powered Forced Convection Solar Dryer The materials which have been used for the construction of photovoltaic powered forced convection solar dryer for chilli drying were easily obtainable in local market. Frame Structure The frame structure was fabricated with the help of cutting, grinding and arc welding. The structure consisted of legs, base frame and supporting frame. M.S angle of (20 × 20 × 3) mm was used to fabricate the structure. This frame structure was mounted on four legs of 800 mm height. Box The frame structure was covered with the G.I. sheet on the back, bottom and one side of the frame. Thermocole sheet were fitted on the same side with G.I. sheet for insulation. On the other side of the structure insulated door is fitted to facilitate loading and unloading of wire mesh trays containing the product to be dried. Insulation Thermocole sheet was fitted for the insulation on the back, bottom and sides of the dryer, which help in reduction of heat loss for the dryer. Glazing Glazed cover of plain glass is mounted on front of the frame for solar energy interception. 4mm thickness of plain glass was used. Drying Trays The solar dryer has two drying trays of wire mesh on base. The trays were constructed from wire mesh, aluminium angle of (25× 25 × 3) mm and aluminium strip of (20 × 3) mm. Exhaust Fan DC fan fitted on opposite wall of the drying chamber which help in faster removal of moisture from produce and it also help in circulation of hot air develop inside the drying chamber due to solar energy. 162 ♦ Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households
ICORE 2013
Solar Panel/ Module Manufacturer by Tata BP Solar India Limited in ISO 9001 and 14001. The module has a voltage of 8.5 V, current 0.6A mp and peak power of 5.0 W. Experimental Procedure Only good quality chillies were used in the experiment. 6 kg of fresh Chilli were dried until the required final moisture content was attained. The fresh chilli was located over the trays of drying chamber having about 90% perforation. The initial moisture content was calculated by taking three different samples. During the experiment moisture content was calculated at every interval of one hour during the drying hours. The exhaust fan was operated for reducing of relative humidity inside the solar dryer. The different parameters were observed during the experiment hours at every interval inside and outside the solar dryer. Open sun drying method was also conducted to know the performance of the dryer. Solar dryer was tested for its performance in drying of Chilli by conducting two sets of experiments. The tests were conducted from 08.00 to 16.00 hrs and the hourly data were recorded. The full load test was conducted for evaluate the performance of the photovoltaic powered forced convection solar dryer in actual loading condition. The dryer contained two trays upper and lower; 4 kg of fresh chilli was placed on the upper tray and remaining in the lower tray i.e. 2 kg of fresh chilli. Loading and unloading was done manually. One kg of fresh Chilli was placed in open air for drying for comparison purpose. The observations of the parameters were recorded at interval of one hours starting from 8 hrs to 16 hrs each day. The solar insolation was measured with lux meter, the inside and outside temperature was measured with the digital thermometer and the hygrometer was used to measure relative humidity. Keeping two thermometers and hygrometer inside and one outside the observation was recorded. By using oven drying method the initial moisture content of fresh chilli was determined. The drying was continued till the moisture content of the Chilli tends to a constant value. The performance of drying unit was evaluated in terms of moisture content variation, drying rate, etc. For this purpose, the hourly reductions in weight of representative sample were recorded. RESULTS AND DISCUSSION The photovoltaic powered force convection solar dryer was developed and fabricated for small farmers and households in NEH region of India in the workshop of the College of Agricultural Engineering and Post Harvest Technology, Central Agricultural University, Ranipool, Gangtok, Sikkim. The experimental data along with input environmental parameters such as solar insolation, ambient temperature, inside temperature, relative humidity were recorded. The recorded data were analyzed to evaluate the performances. Dimension of Solar Dryer The photovoltaic powered forced convection solar dryer was designed and developed in the NEH region of India. The dimensions are given below in Table 1. TABLE 1: DIMENSIONS OF PORTABLE PV POWERED FORCED CONVECTION SOLAR DRYER Components Aperture area Front base of solar dryer Width of solar dryer Nos. of trays Upper tray size Lower tray size Nos. of air inlet Diameter of air outlet Inclination of the dryer Loading per batch Drying time per batch
Specifications 0.595 m2 0.7 m 0.7 m Two (680 × 270) mm (680×490) mm 50 6 mm 45 6 kg 4 days
Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households ♦ 163
ICORE 2013
System Description of the Solar Dryer Figure 2: Shows a Pictorial View of Photovoltaic Powered Forced Convection Solar Dryer Designed and Developed for Drying of Chilli
FIG. 2: PICTORIAL VIEW OF PHOTOVOLTAIC POWERED FORCED CONVECTION SOLAR DRYER
Performance Evaluation of Solar Dryer The performance of solar dryer was evaluated during the month of December, 2011 for drying of chilli. The performance evaluation test includes measuring solar insolation, ambient temperature, ambient relative humidity, wind velocity, air flow rate inside the dryer, air temperature and relative humidity inside the solar tunnel dryer. During full load testing, four days are required for drying total amount of Chilli in the solar dryer from moisture content of 80.2 to 10 per cent (w.b.). The testing on full load was conducted for three consecutive days in the month of December, 2011. VARIATION OF TEMPERATURE AND SOLAR RADIATION First Day of Drying During the first day of drying in the month of December, 2011, it was observed that 65 C and 57 C was the maximum temperature inside the solar dryer in upper and lower tray at 14 hours respectively and the minimum temperature for the upper and lower tray was 28 C and 24 C at 8 hours respectively. The maximum ambient temperature was 28 C at 14 hours while minimum was 17 C at 8 hours respectively. It was observed that maximum solar radiation inside the dryer was 523 W/m2 at 13 hours and minimum was 19 W/m2 at 16 hours respectively. The maximum ambient solar radiation was 601W/m2 and minimum was 43 W/m2. The trend of results is plotted on graphs as shown in Fig. 3 and 4. Ambient
Solar radiation, W/m2
700
Inside the dryer
600 500 400 300 200 100 0 8
9
10
11
12
13
14
15
16
Time, h
FIG. 3: VARIATION OF SOLAR RADIATION UNDER FULL LOAD CONDITION IN THE FIRST DAY OF DRYING STAGE 164 ♦ Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households
ICORE 2013 Ambient
Upper tray
Lower tray
70 Temperature, °C
60 50 40 30 20 10 0 8
10
12 Time, h
14
16
FIG. 4: VARIATION OF AIR TEMPERATURE UNDER FULL LOAD CONDITION IN FIRST DAY OF DRYING STAGE
SECOND DAY OF DRYING During the second day of drying in the month of December, 2011, it was observed that 69 C and 57 C was the maximum temperature for the upper and lower tray inside the solar dryer at 14 hours respectively and the minimum temperature for the upper and lower tray was 30 C and 25 C at 8 hours respectively. The maximum ambient temperature was 28 C at 14 hours while minimum was 17 C at 8 hours respectively. It was observed that maximum solar radiation inside the dryer was 535 W/m2 at 13 hours and minimum was 19 W/m2 at 16 hours respectively. The maximum ambient solar radiation was 609W/m2 and minimum was 51 W/m2 respectively.The trends of results is plotted on graphs as shown in Fig. 5 and 6. 700
Outside
Inside
Solar radiation, W/m2
600 500 400 300 200 100 0 8
9
10
11
12
13
14
15
16
Time, h
FIG. 5: VARIATION OF SOLAR RADIATION UNDER FULL LOAD CONDITION IN THE SECOND DAY OF DRYING STAGE 80
Ambient
Upper tray
Lower tray
Temperature, °C
70 60 50 40 30 20 10 0 8
9
10
11
12
13
14
15
16
Time, h
FIG. 6: VARIATION OF AIR TEMPERATURE UNDER FULL LOAD CONDITION IN SECOND DAY OF DRYING STAGE Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households ♦ 165
ICORE 2013
THIRD DAY OF DRYING During the third day of drying in the month of December, 2011, it was observed that 71 C and 68 C was the maximum temperature for the upper and lower tray inside the solar dryer at 14 hours respectively and the minimum temperature for the upper and lower tray was 30 C and 24 C at 8 hours respectively. The maximum ambient temperature was 28 C at 14 hours while minimum was 16 C at 8 hours respectively. It was observed that maximum solar radiation inside the dryer was 586W/m2 at 13 hours and minimum was 30 W/m2 at 16 hours respectively. The maximum ambient solar radiation was 671W/m2 and minimum was 65W/m2 respectively. The trend of results is plotted on graph as shown in Fig. 7 and 8. 800
Ambient
Solar radiation, W/m2
700
Inside
600 500 400 300 200 100 0 8
9
10
11 12 Time, h
13
14
15
16
FIG. 7: VARIATION OF SOLAR RADIATION UNDER FULL LOAD CONDITION IN THE THIRD DAY OF DRYING STAGE Ambient
80
Upper tray
Lower tray
Temperature, °C
70 60 50 40 30 20 10 0 8
9
10
11
12 Time, h
13
14
15
16
FIG. 8: VARIATION OF AIR TEMPERATURE UNDER FULL LOAD CONDITION IN THIRD DAY OF DRYING STAGE
FOURTH DAY OF DRYING During the fourth day of drying, it was observed that 69 C and 58 C was the maximum temperature for the upper and lower tray inside the solar dryer at 14 hours respectively and the minimum temperature for the upper and lower tray was 30 C and 27 C at 8 hours respectively. The maximum ambient temperature was 29 C at 14 hours while minimum was 17 C at 8 hours respectively. It was observed that maximum solar radiation was 409 W/m2 at xx hours and minimum was 19 W/m2 at 16 hours respectively. The maximum ambient solar radiation was 671W/m2 and minimum was 65W/m2 respectively. The trend of results is plotted on graph as shown in Fig. 9 and 10.
166 ♦ Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households
ICORE 2013
Solar radiation, W/m2
700
Ambient
Inside
600 500 400 300 200 100 0 8
9
10
11 12 Time, h
13
14
15
16
FIG. 9: VARIATION OF SOLAR RADIATION UNDER FULL LOAD CONDITION IN THE FOURTH DAY OF DRYING STAGE
80
Ambient
Upper tray
Lower tray
70 Teperature, °C
60 50 40 30 20 10 0 8
9
10
11
12 Time, h
13
14
15
16
FIG. 10: VARIATION OF AIR TEMPERATURE UNDER FULL LOAD CONDITION IN FOURTH DAY OF DRYING STAGE
VARIATION OF MOISTURE CONTENT First Day of Drying
Moisture content,%
At the starting of first day the initial moisture content measured at upper and lower tray was 80.2 % and 80.2 % w.b. and at the end of the first day it was 65.1% and 67% w.b. respectively. The reduction in moisture content with time was shown in Fig. 11.
90 80 70 60 50 40 30 20 10 0
Upper tray
8
9
10
11
Lower tray
12
13
14
15
16
Time, h
FIG. 11: VARIATION OF MOISTURE CONTENT WITH TIME IN THE FIRST DAY OF DRYING STAGE Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households ♦ 167
ICORE 2013
Second Da ay of Drying g At the starting of Secoond day the initial moisture content me easured at upper u and lower tray wa as 64.65 % and 66.31% w.b. and d at the end d of the secoond day it was w 45.1% and a 47.63% w.b. respecctively. The reduction in moisture content with was w shown in Fig.12. 70 Moisture content,%
60 50 40 30 20
Uppeer tray
Lower tray
10 0 8
9
10
11
12 Time, h
1 13
14
15 5
16
FIG. 12: VARIATION OF O MOISTURE CONTENT O WITH TIME IN THE SECOND DAY OF DRYING STAGE
y of Drying Third Day
Moisture content,%
At the starting of third d day the initial moisturee content measured at up pper and low wer tray wass 44.63 % and 46.11 % w.b. and at the end of the thhird day it was w 24.1% and 26.9% w.b. respecctively. The reduction in moisture content with tiime was show wn in Fig. 13. 50 45 40 35 30 25 20 15 10 5 0
Upper tray t
8
9
10
11
Lower tray
12 Time, h
13 3
14
115
16
FIG. 13: VARRIATION OF MOISTTURE CONTENT WIITH TIME IN THE THIRD DAY OF DRYYING STAGE
Fourth Da ay of Drying At the sta arting of fourrth day the initial moisture content measured m at upper and llower tray was w 23.3% and 24.9% % w.b. and at a the end off the fourth day d it was 10 0% and 10% % w.b. respecctively. The reduction in moisture inn moisture coontent with was shown in Fig. F 14.
FIG. 14: VARIATION OF O MOISTURE CONTENT O WITH TIMEE IN THE FOURTH DAY OF DRYING STAGE 168 ♦ Development of a Ph hotovoltaic Powered Forced Co onvection Solar Dryer for Small Farmers & Hou useholds
ICORE 2013
VARIATION OF DRYING RATE Figure 15 shows the variation of drying rate against drying time. High drying rate (of about 0.5127 kg/hr) was observed during the initial stages of drying. This is because the higher drying rate of the chilli during initial stage of drying results in release of more moisture in the drying air. Drying rate gets decreased with increase in drying time. Drying occurs in the falling rate period with steep fall in moisture content in initial stages of drying and becomes very low in the later stages.
Drying rate, kg/hr
0.6 0.5 Upper Tray
0.4
Lower Tray
0.3 0.2 0.1 0 0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
Time, h
FIG. 15: VARIATION OF DRYING RATE WITH TIME
Open Sun Drying
Moisture content,% (w.b.)
It was observed that in open sun drying method the required final moisture content was attained in 56 hours of drying process. It was found that the final required moisture content was 11.9 at the end of the drying. Figure 4.19 showing the temporal variation of moisture content during the open sun drying process in the month of December, 2011. 90 80 70 60 50 40 30 20 10 0
First day Second day Third day Fourth day Fifth day Sixth day 8
9
10
11
12
13
14
15
16
seventh day
Time, h
FIG. 16: TEMPORAL VARIATION OF MOISTURE CONTENT DURING THE OPEN SUN DRYING PROCESS
CONCLUSION The photovoltaic powered force convection solar dryer was designed and developed at workshop and tested at solar yard of Collage of Agricultural Engineering and Post Harvest Technology, Central Agriculture University, Ranipool, Gangtok, Sikkim for small farmer and households in NEH region. The dryer had capacity to dry 6 kg fresh chilli in 32 hours. The tests were conducted from 0.800 to 16.00 hrs and the hourly recorded. The full load testing of dryer was conducted for evaluating the performance in actual loaded condition. In order to compare the efficiency of drying in photovoltaic powered force convection solar dryer, open sun drying was also conducted. The chilli was dried within 32 hrs from initial moisture content 80.2% to final moisture content about 10% w.b. In order to compare the efficiency of drying in solar dryer PV powered forced convection, open sun drying was also conducted and it was observed that within 56 hrs chilli was dried from moisture content 80.2% to 11.9% w.b. Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households ♦ 169
ICORE 2013
REFERENCES [1] Desai,S., Palled, V.K. and Anantachar, M. 2008. Performance evaluation of farm solar dryer for chilli drying. Journal of Agriculture Sciences 22(2): 382-384. [2] Fortson, F.K., Nazha, M. A. A., Akuffo, F.O. and Rajakaruna, H. 2006. Design of mixed mode natural convection solar crop dryer: Application of principles and rules of thumb. Journal of Renewable energy 32: 2306-2319 [3] Folarami, J. 2008. Design, constructional and testing of simple solar maize dryer. Leonardo Electronic Journal of Practices and Technology: 122-130. [4] Gatea, A. A. 2010. Design, construction and performance evaluation of solar maize dryer. Journal of Agriculture Biotechnology and Sustainable Development 2(3): 39-49. [5] Gatea, A. A. 2011. Performance evaluation of mixed-mode solar dryer evaporating moisture in beans. Journal of agriculture Biotechnology and Sustainable Development 3(4): 65-71. [6] Mumba, J., 1996. The development of a photovoltaic powered forced convection solar maize dryer for use in Sub-Saharan Africa. Journal of RERIC International Energy 18(1): 27-42. [7] Mohanraj, M. and Chandrasekar, p. 2009. Performance of a forced convection solar dryer integrated with Gravel as heat storage material for chilli drying. Journal of Engineering Science and Technology 4(3): 305-314. [8] Okeke. C. A. and Nwokoye A. O. C. 2008. A comparative study of cabinet solar dryer and open sun drying of agriculture produce: (A Case of Unripe Plantain, Breadfruit and Bitter Leaf). Journal of Natural and Applied Sciences 9 (1): 1-8. [9] Prasad, J., Vijay, V.K., Tiwari G.N. and sorayan V.P.S. 2004. Study on performance evaluation of hybrid drier for turmeric drying at village scale. Journal of Food Engineering 75(4): 496-502. [10] Sawhey, R. L., Pangavadia D.R. and Sarsavadia P. N. 2000. Design, development and performance testing of a new natural convection solar dryer. Energy 27: 579-590. [11] Yusuf, M. O. L. and Oghenervona, D. 2011. Design and fabrication of a direct natural convection solar dryer for Tapioca. Leonardo Electronic Journal of Practices and Technology: ISSN 1583-1078.
170 ♦ Development of a Photovoltaic Powered Forced Convection Solar Dryer for Small Farmers & Households
Designing and Testing of the Battery Performance of a PIC Microcontroller based Solar Wind Hybrid System during the Load and no Load Conditions B.K. Sahu1 and P.K. Choudhury2 1Gandhi
Institute for Education and Technology, Baniatangi, Khurda, Bhubaneswar, Odisha–756020, India 2Department of Energy, Tezpur University, Assam–784028, Napaam
Abstract—Increase in energy demand along with fossil fuel depletion has raised a critical issue worldwide on meeting the future demands. The exploitation of renewable energy is one of the alternative options in this context. Among all the renewable energy sources, both solar and wind energy is gaining much more importance across the globe. However, both are dependent on geographic locations with varying degree of unpredictability. In our present study, a PIC microcontroller based solar wind hybrid system has been designed for meeting the daily load requirement. Here, a boost type DC-DC converter from 11-12V I/P to 16V O/P is designed for efficient charging of the battery.
INTRODUCTION Access to affordable and reliable source of energy is a vital factor for overall developm-ent of a country. The supply of secure, clean and sustainable energy is the most important factor in the 21th century. It is a challenging to say that the energy security, national security, environmental security and economical security will meet together within 10-20 years [1]. With the rapid population growth, industrialization and urbanization of rural areas, particularly in the Asian and African countries, the demand of energy will be significantly increased in the coming years besides the constant rise in energy demand in developed countries. Till date, coal, liquid fuels and natural gas are the prime sources of energy and will continue to dominate the global energy market. Among these fossil sources, about 80% of world energy demand comes from fossil fuels [2]. Most of the coal is consumed in the power sector. Coal shares about 42% of global electricity generation. In 2010, the energy consumption in Japan grew up by 6.7%, in USA by 3.7% and more than 6% by both in India and China than 2009. In the European countries also, the demand has also increased by 4%. At present, China is world’s top energy consumer (11%) followed by the USA and India. Therefore, there is an urgent need for efficient exploitation of aalternative energy sources in order to meet the energy demands, at least partially if not cmpletely. The various alternative sources include solar, wind, biomass, geothermal and small hydropower. Out of these, both exploitation of solar and wind energy sources exhibits the fastest growth in the world, the rate of increase being 25-35% annually over the last decade [3]. However, geographic or seasonal dependence of these sources restricts them from meeting the load requirement continuously as independent sources. For example, during the period of available solar energy, wind energy may not be sufficient for supplying the required load and vice-versa. In such situations, an appropriately interconnected system of solar and wind energy sources can demonstrate an efficient load driving capability almost continuously by extracting energy from the available energy sources. The present study is focused on the characteristics of a battery under load and no load conditions which is fed from an integrated solar–wind system interconnected through a PIC microntroller based control system. SYSTEM CONFIGURATIONS OF SOLAR WIND HYBRID SYSTEM The configurations of stand alone solar wind are proposed that shown in fig.1. In these systems, one DCDC boost converter is designed with MC34063A IC, a lead acid sealed and maintenance free type of 12V,7 Ah, a solar PV module of 10Wp manufactured by Ritilka system, India and wind turbine manufactured by PAWAN URJA wind turbine charger,72W Manufactured by Jindesh International and one PIC microcontroller based charge controller. The designed boost converter is used as a voltage converter from 12V I/P to 16V O/P at one ampere of output current for charging of the battery. The boost converter was used used both the systems individually. The purpose for designing of the system is when solar energy is not able to supply the load at that time wind energy will supply, similarly when wind energy is not able to supply the load at that time solar energy connections, then it is able to give continuously 24hrs supply the load and the energy will store by the battery for satisfying the demands in the future. This hybrid system is most preferable in the remote areas where electrifications are not available. Of course, this complicated designing can be more complex as compared with single renewable energy sources. This complexity designing could be minimized through the optimizations of the whole systems. However, there are various issues about the effects of battery regarding with his life spans, size and costs.
ICORE 2013
DC-DC Boost Converter Solar Module 10 Wp
Wind Turbine 72 Watt
PIC 16F877A Microcontroller
Battery, 12V, 7Ah
DC Load FIG.1: A PIC MICROCONTROLLER BASED SOLAR WIND HYBRID SYSTEM
In the present study, we are designing a DC-DC boost converter and a PIC microcontroller based charge controller for a laboratory test of the solar wind hybrid system. Specially, we are given the focus on the characteristics of the battery at load and no load conditions. TABLE 1: OPERATING CONDITIONS OF SOLAR WIND HYBRID SYSTEMS Sl. No. 1 2
Solar On On
Wind On On
Load Conditions Generation > Load Generation < Load
Battery Effects Charges Discharge
3
On
Off
Generation> Load
Charge
4
On
Off
Generation Load
Charge
6
Off
On
Generation< Load
Discharge
Comments This is an ideal operating mode The generation of power is not sufficient to charge the battery The solar energy is available to supply sufficient power to the load The output power from solar energy is not sufficient to charge battery The wind energy is available and it sufficient to supply the load The wind energy is available but not sufficient to supply the load
BATTERY The battery is an electrochemical device which uses electrochemical reactions to store electricity in the form of chemical energy. The role of the battery is charging from the solar or wind systems and supply to the load or stored energy for the future works in the off grid systems. Further, there re number conditions for charging or discharging of the battery because both systems are unpredictable in nature which affects to the life span of the battery. So, for this system, one controller is to be required. Here, it is a microcontroller based charge controller which is fully programmable based. So, battery is a very important issue for the hybrid system. In this study, we are discussing different effects of battery such as charging current rate, the charging efficiency, the self charge rate as well as the battery capacity. Most of cases there are three important characteristics of battery such as the battery state of charge (SOC), the floating charge voltage and the battery life time. Battery State of Charge For the ideal condition, when solar or wind energy or both generations is more than the demand load then the battery is said to be charging. So, the battery capacity is changing with respect to the load and generations. The battery capacity is dependent the temperature. The battery capacity is changed by using the temperature coefficient . Cbat ´ = Cbat´´ 1 + δc(Tbat − 298.15)
172  Designing and Testing of the Battery Performance of a PIC Microcontroller Based Solar Wind Hybrid System
ICORE 2013
Where Cbat´ is the available or practical capacity of the battery when the battery temperature is Tbat, Cbat is the nominal or rated battery capacity which is given by the manufacturer as per the standard value that characterizes this battery, Ah; a temperature coefficient of δc (Berndet,1994). The state of charge simply calculates by σ∆t I ∆tη SOC(t + 1) = SOC(t) 1 − + 24 C Where is the self discharge rate of the battery which depends on the accumulated charge of the battery and a proposed value of 0.2% per day is recommended. The current rate of the battery at time t for the battery at time t for the solar wind hybrid system is ´´
P (t) + P (t) − I
(t) =
V
(
ɳ
)
−P
(t)
( )
Floating Charge of the Battery The floating charge of the battery responses under charging and discharging is modelled by the equation-fit method V ′ = A × (SOC) + B × (SOC) + C × SOC + D Where V is the battery floating charge voltages in order to take into account the effect on the battery voltage predictions. = V ′ + δ (T − 298.15) Where V is calibrated battery voltage after the effects of the temperatures. The temperature coefficient is constant of-4mV/0C per 2V cell for the considered battery temperature range and A, B, C and D are the constants depending upon the battery current values calculated by second degree polynomial equations. After estimation, if I>0, then the battery is charging and I15%) crystalline silicon solar cell. Stainless steel (SS316) screen printing is the most suitable process for its low cost, repeatibility, flexibility and high productivity. A comparative study of the application of normal two dimensional (2D) SS screen to three dimensional (3D) SS screen for front silver metallization of multicrystalline large area (125 mm × 125 mm) silicon solar cell is reported. Our study reveals the impact of the 3D screen printing to enhance the solar cell efficiency from 14.67% to 15.5% in the production plant. Keywords: Multicrystalline Silicon Solar Cell Texturization, Knitting of SS Wires in 2D & 3D Mesh, Roughness Analysis, Cell LIV and DIV Characteristics
INTRODUCTION In commercial large area crystalline silicon PV plants, different metallization schemes such as photolithography after vacuum evaporation, electroplating, and buried contact are used, they are more expensive as well as time consuming for large scale production process [1]. Screen printing technique provides a cost-effective alternative to those complicated schemes due to which it has already been a widespread method in solar cell industries [2]. For crystalline silicon, the well established metallization method is screen printing technology which results in the surface covered with suitable metal pastes. However, for metallization of the emitter surface, it is very much required to control the series resistance loss after cell fabrication. The main problem that is generally experienced in screen printing metallization is poor contact quality that results in poor fill factor of the fabricated cells. Metal contact with silicon surface depends on several parameters such as surface condition, emitter surface concentration and dopant impurity profile [3], anti-reflection coating on the emitter surface, screen parameters, metal pastes and above all, the firing profile. Optimization of front and back contacts for the solar cell is the ultimate task for process designers. People thus look for an approach which not only reduces the effects of series resistance by enhancing the cell fill factor (FF), but also yields a higher short circuit current (Isc) by the reduction of the shadowing loss without compromising with production output. In our present work, multicrystalline silicon solar cells are fabricated in a conventional industrial process. During cell front screen printing both the two dimensional (2D) and three dimensional (3D) stainless steel meshes are used. Our paper reports the impact of these screen printing processes on the performance of electrical parameters of the solar cell by studying illuminated voltage–current (LIV) and dark voltage–current (DIV) characteristics to ascertain the superiority of the 3D screen. Also knitting of the meshes and surface reflectivity of the cells are studied by the optical microscope and the spectrophotometer respectively. EXPERIMENTAL Cell Fabrication The starting material for the experiment is boron doped p-type mC-Si wafers of base resistivity 0.5 ~ 3.0 Ω-cm of brand SOLSIX from Deutsche Solar of size 125mm × 125mm square. and cells are fabricated using texturization with sodium hydroxide (NaOH)–sodium hypochlorite (NaOCl) based polishing solution [4]. The alkali based polishing solution has NaOH solution (20% by weight) and NaOCl solution in the ratio of 1:1 by volume [5]. The polishing bath (made with SS316 material) is filled with this solution and is heated by teflon heater (jacket type) from the bottom and a constant temperature of 80-82oC is maintained with the combination of thermocouple and PID controller. 40 wafers are loaded in a single teflon jig and the single polishing bath of NaOH-NaOCl solution can accommodate 6 jigs, i.e., 240 wafers at a time for 20 minutes. In the same solution of NaOH-NaOCl bath 10 batches, i.e., 2400 wafers are polished. Heating of the solution is required only for the first batches for bath temperature to reach up to 82oC, and later further heating for subsequent polishing batches are not required due to slow exothermic nature of silicon polishing reaction [6]. After texturing, the wafers are then doped with phosphorus (P) using optimized diffusion condition with phosphorus oxychloride as the source at 875ºC. After diffusion, the phosphosilicate glass (PSG) layer is removed in dilute hydrofluoric (HF) acid solution. The low temperature oxidation (LTO) is a pre requisite step to dissolve hard PSG which can not be removed by HF. Thermal SiO2 of thickness ~150-200 Å is grown on the sample surface at 7500C. After LTO, the grown oxide is
ICORE 2013
again removed so that it consumes some dead layer to enhance cell blue response [7]. The P-impurities diffused on the edges of the wafer need to be removed to prevent shunting in the wafer and to minimize the leakage current. In our cost effective process, the stack of wafers are edge-etched together using HFnitric (HNO3)-acetic (CH3COOH) solution in 3:5:3 ratio for 1 min at 7-8˚C bath temperature. In order to suppress reflections from a solar absorber surface an anti-reflection coating (ARC) layer is deposited on top of the solar wafer. For the AR coating to function well it should be made of a material having a lower refractive index than the underlying surface. ARC is particularly beneficial for multicrystalline material that cannot be easily textured. In our process ARC of 700Å of Si3N4 of refractive index 2.05 has been deposited at 450˚C by plasma enhanced chemical vapour deposition. The back and front sides are screen printed with silver-aluminium (DUPONT PV202), aluminium (DUPONT PV333) and silver pastes (DUPONT PV145) one by one, followed by baking of individual pastes of the printed wafers. For excellent screen printed metallization we require the setting of the optimum conditions for the paste, optimized screen mask specifications and optimized printing conditions. After depositing, the layer is sintered at temperatures of approximately >760oC in a conveyer belt furnace. The complete process flow chart is given in Fig.1. Metallization For the front emitter metallization, two batches of 40 wafers each, are taken after back aluminium and silver-aluminium printing. First 40 wafers are then screen printed on the front surfaces by 2D 325 and the other 40 wafers by 3D SS 325 screens. After printing and cofiring, one cell from each batch is taken as the representative for the 2D and 3D screen printing processes. NaOH-NaOCl Texturing Phosphorous Diffusion PSG removal Oxidation Edge Etching Silicon Nitride deposition Back Metal printing & Baking Front metal printing & Baking Co-firing FIG. 1: SOLAR CELL FABRICATION PROCESS
Characterization Small pieces of dimension 2 cm × 2 cm are cut from the representative cells of 2D and 3D screen printing categories by using Nd–YAG laser for the surface reflectance analysis by spectrophotometer. Before the surface analysis, wafer pieces (as samples) are cleaned ultrasonically in isopropyl alcohol followed by rinsing in DI water and drying. These samples containing only the grid lines are taken from the same areas of both the cells. The DIV and LIV characteristics of the cells are also measured. Their LIV characteristics are measured under 1 SUN intensity with AM1.5 Global spectrum.
182 ♦ Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell
ICORE 2013
RESULTS AND DISCUSSION The knitting of the SS wires in screens are observed by an optical microscope and the magnified pictures in the 2D-325 and 3D-325 screens are shown in Figs.2(a) and 2(b) respectively. For the 2D screen as in
Thin FIG. 2(A): THE KNITTING OF SS WIRES IN 2D MESH
Thick
FIG. 2(B): THE KNITTING OF SS WIRES IN 3D MESH
Fig. 2 (a), only two wires are used and maximum knitting height possible here is the total of diameters of the two wires, whereas, Fig. 2 (b) shows that the 3D mesh consists of three wires, thereby, the maximum knitting height possible is the strip, area of the metal strip and the radius of the Ag-grid fingers respectively. Although ‘R’, ‘ρ’ and ‘l’ are fixed for a total of diameters of the three wires. Table-I describes the comparison of different screen parameters for the 2D and TABLE 1: THE COMPARISON TABLE OF 3D AND 2D SS MESH PARAMETERS Mesh Properties
Mesh Type
Mesh count (nos./ inch) SS wire diameter ( μ m ) Maximum metal height possible ( μ m ) Open area (%) Maximum volume of the paste flow (cm3/m2)
325 28 45-56 41.2 23.07
2D 400 23 35-46 40.7 19.54
500 19 25-36 39.2 14.11
325 28 77 41.2 31.72
3D 400 23 66 40.7 26.86
500 19 53 39.2 20.78
Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell ♦ 183
ICORE 2013
and 3D screens for a mesh count of 325, 400 and 500 [8]. In all the three screens the diameters of the SS wires is same for both the 2D and 3D screens and for higher mesh counts, wires becomes thin. In the 3D mesh, the volume of the paste flow is much higher as compared to the 2D mesh even though the open areas are the same for the same mesh count. In our front Ag-printing, the 3D-325 mesh has been used. In this mesh, the maximum volume flow is nearly 30% higher as compared to the 2D-325 mesh. It results into the more height of the Ag-metal generated by 3D mesh than that by the normal 2D one. The front metal resistance is given by l l =ρ 2 A πr Here, R’, ‘l’, ‘A’ and ‘r’ are the resistance of the front Ag-metal, length of the metal particular type of Ag-paste and similar screen opening design, the value of ‘ r ’ becomes more for the 3D screen printed cell as compared to the 2D one. This only reduces the effective value of the ‘ R ’. The 3D-325 mesh is selected for our front printing and a silver grid finger height of ≈ 12–15 μm for Ag at front is achieved, while for the 2D-325 mesh, this height is only ≈ 8–10 μm. Hence this increase in the metal height has its impact on cell series resistance by the reduction of resistance of Ag-front metal. R=ρ
45
40
3D Screen printed (3DSP) cell 2D Screen printed (2DSP) cell
35
Reflectivity (%)
30
25
20
15
10
5 300
400
500
600
700
800
900
1000
1100
1200
Wavelength (nm)
FIG. 3: VARIATION OF SURFACE REFLECTANCE OF THE FRONT SURFACE OF THE 2DSP AND 3DSP CELL
The reflectivity comparison graph is shown in Fig. 3. There is a definite marginal decrease in average reflectivity from 14.60% to 14.47% in the wavelength range of 300 nm to 1200 nm which is contributed by the increase in finger height in the 3D printing. This small enhancement of the reflectivity is also caused by the marginal shadowing loss minimization in the 3D screen and this fact contributed to the increase of short circuit current of the cell as shown in the LIV characteristics in Fig.4. The uniformity of metal coverage in the 3DSP cell is ascertained by the comparison of the cell LIV parameters as shown in Table-II. A major impact of excellent emitter metallization is clearly visible in the significant improvement in FF of the 3DSP cell as compared to the 2DSP one. This upgradation of the Isc and FF results into an improvement of cell efficiency from 14.6% (2DSP) to 15.5% (3DSP) without any major changes in cell fabrication techniques. 184 ♦ Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell
ICORE 2013
TABLE 2: ELECTRICAL PARAMETERS OF SOLAR CELL FABRICATED USING 2D AND 3D SCREENS Screen used 2D 3D
Voc (V) 0.608 0.612
FF
Isc (A) 5.02 5.18
Rs (mΩ) 8.2 7.3
0.751 0.764
Rsh (Ω) 15 17
Efficiency (η), % 14.6 15.5
Also the DIV characteristics of the cells (shown in Fig.5) do not have much difference in nature and thus indicates no notable change in cell leakage current characteristics. 0.4
2D Screen printed (2DSP) cell 3D Screen printed (3DSP) cell 0.3
Current (A)
0.2
0.1
0 -1.2
-0.8
-0.4
0
0.4
0.8
Voltage (V) -0.1
FIG. 4: THE ILLUMINATED CURRENT–VOLTAGE (LIV) CHARACTERISTICS CURVES OF THE 2DSP AND 3DSP CELLS
Current (A)
4
2
2DSP Cell
3DSP Cell
Voc = 608 mV = 5.02 A Isc FF = 0.751 Effiiciency= 14.67% Rs = 8.2 m-ohm = 15 ohm Rsh
Voc = 612 mV = 5.18 A Isc FF = 0.764 Effiiciency= 15.5% Rs = 7.3 m-ohm = 17 ohm Rsh
2D screen printed (2DSP) cell 3D screen printed (3DSP) cell
0 0
200
Voltage (mV)
400
600
FIG. 5: THE DARK CURRENT–VOLTAGE (DIV) CHARACTERISTICS CURVES OF THE 2DSP AND 3DSP CELLS
CONCLUSION The increase in metal heights during the 3D screen printing of the emitter surface layer is quite large as compared to the 2DSP cell. Higher height of front silver decreases both the shadowing loss and metal resistance. It resulted into a marginal decrease of reflectivity by 0.13% in 300 nm to 1200 nm wavlength range. Low series resistance value (7.3 mΩ) of the 3DSP cell enhances cell FF upto 0.764. In a cumulative effect the cell efficiency enhances from 14.67% to 15.5% without any other change in the regular production line. Also higher height eases soldering with solder plated copper strip and creates strong bonding between them at tabbing step during module fabrication. This clearly reflects the supiriority of the 3D screen metallization processes. Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell ♦ 185
ICORE 2013
ACKNOWLEDGMENT The authors express their sincere thanks to P.G. Dept. of Applied Physics and Ballistics, FM University, Balasore India for continuous stimulation and support for the present research. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
A. Ebony, Y.H. Cho, M. Hilalii, A. Rohatgi and D. Ruby; Solar Energy Materials and Solar Cells, (2002), 74, p.51. M. Bohm, E. Urbanski, A.E. Delahoy and Z. Kiss; Solar Cells, (1987), 20, p.155. P.K. Basu, B.C.Chakravarty, S.N.Singh, P.Dutta and R.Kesavan, Solar Energy Materials and Solar Cells, (1996); 43, p.15. U. Gangopadhyay, S.K.Dhungel, P.K. Basu, S.K. Dutta, H.Saha and Junsin Yi, Solar Energy Materials and Solar Cells, 91, (2007), p.285–289. P.K. Basu, H. Dhasmana, D. Varandani, B.R. Mehta and D.K. Thakur, Solar Energy Materials and Solar Cells,Vol.93, Oct, 2009, P.1743–1748. P.K. Basu, H. Dhasmana, Udayakumar N. and D.K. Thakur, Renewable Energy, Vol. 34, Nov. 2009, p.2571–2576. J. Lindmayer and J.F. Allison, COSMAT Tech. Rev., vol.3 (1973), p.1. www.teiko-sino.com E. Vazsonyi, Z. Vertesy, A. Toth and J. Szlufcik, Journal of Micromechanics and Microengineering, (13), 2003, p.165–169. U. Gangopadhyay, S.K. Dhungel, P.K. Basu, S.K. Dutta, H. Saha and Junsin Yi, Solar Energy Materials and Solar Cells, 91, (2007), p.285–289.
186 ♦ Comparison of 2D and 3D Screen Printed Metallization on Multicrystalline Silicon Solar Cell
A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method K. Vimala Kumar1, V. Ganesh2 and J. Suresh3 1Assistant
2Associate
Professor, Dept. of EEE, JNTUA College of Engineering, Pulivendula, Kadapa (DT), A.P. Professor & HOD, Dept. of EEE, JNTUA College of Engineering, Pulivendula, Kadapa (DT), A.P. 3HOD, Dept. of EEE, Audisankara College of Engineering, Gudur, Nellore (DT), A.P E-mail: [email protected], 2email:[email protected]
Abstract—This paper presents simulation of ncremental conductance (IncCond) maximum power point tracking (MPPT) used in solar array power systems with direct control method. The main difference of the proposed system to existing MPPT systems includes elimination of the proportional integral control loop and investigation of the effect of simplifying the control circuit. Contributions are made in several aspects of the whole system, including converter design, system simulation, controller programming. The resultant system is capable of tracking maximum power point accurately and rapidly without steady state oscillation, and also, its dynamic performance is satisfactory. From the results acquired during the simulations, it was confirmed that, with a well-designed system including a proper converter and selecting an efficient and proven algorithm, the implementation of MPPT is simple and can be easily constructed to achieve an acceptable efficiency level of the PV modules. The results also indicate that the proposed control system is capable of tracking the PV array maximum power and thus improves the efficiency of the PV system and reduces low power loss and system cost. The results of the simulations are validated through latest version of MATLAB package. Keywords: Digital signal processor (DSP), incremental conductance, maximum power point tracking (MPPT), photovoltaic (PV) system.
INTRODUCTION Recently, energy generated from clean, efficient and environmentally-friendly sources has become one of the major challenges for engineers and scientists. Among them, the photovoltaic (PV) generation system has received great attention in research because it appears to be one of the possible solutions to the environmental problem [3]–[5]. The efficiency of solar cells depends on many factors such as temperature, insulation, spectral characteristics of sunlight, dirt, shadow, and so on. Changes in insulation on panels due to fast climatic changes such as cloudy weather and increase in ambient temperature can reduce the photovoltaic (PV) array output power. In other words, each PV cell produces energy pertaining to its operational and environmental conditions [6], [7]. In addressing the poor efficiency of PV systems, some methods are proposed, among which is a new concept called maximum power point tracking(MPPT). All MPPT methods follow the same goal which is maximizing the PV array output power by tracking the maximum power on every operating condition. Direct Control Method Conventional MPPT systems have two independent control loops to control the MPPT. The first control loop contains the MPPT algorithm, and the second one is usually a proportional (P) or P–integral (PI) controller. The IncCond method makes use of instantaneous and IncCond to generate an error signal, which is zero at the MPP; however, it is not zero at most of the operating points. The main purpose of the second control loop is to make the error from MPPs near to zero [8]. Simplicity of operation, ease of design, inexpensive maintenance, and low cost made PI controllers very popular in most linear systems. However, the MPPT system of standalone PV is a nonlinear control problem due to the nonlinearity nature of PV and unpredictable environmental conditions, and hence, PI controllers do not generally work well. In this paper, the IncCond method with direct control is selected. The PI control loop is eliminated, and the duty cycle is adjusted directly in the algorithm. The control loop is simplified, and the computational time for tuning controller gains is eliminated. To compensate the lack of PI controller in the proposed system, a small marginal error of 0.002 was allowed. The objective of this paper is to eliminate the second control loop and to show that sophisticated MPPT methods do not necessarily obtain the best results, but employing them in a simple manner for complicated electronic subjects is considered necessary. The feasibility of the proposed system is investigated with a dc–dc converter configured as the MPPT. PV MODULE AND MPPT The basic structural unit of a solar module is the PV cells. A solar cell converts energy in the photons of sunlight into electricity by means of the photoelectric phenomenon found in certain types of semiconductor materials such as silicon and selenium. A single solar cell can only produce a small amount of power. To
ICORE 2013
increase the output power of a system, solar cells are generally connected in series or parallel to form PV modules. PV module characteristics are comprehensively discussed in [3], [6],and [11], which indicate an exponential and nonlinear relation between the output current and voltage of a PV module. The main equation for the output current of a module is [6] I 0 = n p I ph − n p I rs [exp( K 0
v ) − 1] ns
Where Io is the PV array output current, V is the PV output voltage, Iph is the cell photocurrent that is proportional to solar irradiation, Irs is the cell reverse saturation current that mainly depends on temperature, Ko is a constant, ns represents the number of PV cells connected in series, and np represents the number of such strings connected in parallel. In (1), the cell photocurrent is calculated from
I ph = [ I scr + k i (T − Tr )]
s 100
Where, ISCR cell short-circuit current at reference temperature and radiation; Ki short-circuit current temperature Tr coefficient; cell reference temperature; S solar irradiation in milliwatts per square centimeter. Moreover, the cell reverse saturation current is computed from
I rs = I rr [
qE T 3 1 1 ] exp( G [ − ]) Tr KA T r T
Where, T r cell reference temperature; Irr reverse saturation at Tr; EG band-gap energy of the semiconductor used in the cell. For simulations and the experimental setup also, the KC85T module was chosen. The electrical parameters are tabulated TABLE I: ELECTRICAL PARAMETERS OF SOLAR MODULE Maximum Power (Pmax) Voltage at MPP (Vmpp) Current at MPP (Impp) Open Circuit Voltage (Voc) Short Circuit Current (Isc)
87W 17.4V 5.02A 21.7V 5.34A
In Table I, and the resultant curves are shown in Fig. 1.(a) and (b). It shows the effect of varying weather conditions n MPP location at I–V and P–V curves. Fig. 2 shows the current-versus-voltage curve of a PV module. It gives an idea about the significant points on each I–V curve: open-circuit voltage, shortcircuit current, and the operating point where the module performs the maximum power (MPP). This point is related to a voltage and a current that are Vmpp and Impp, respectively, and is highly dependent on solar irradiation and ambient temperature [7].
(a) (b) FIG. 1: MAXIMUM POWER WITH VARYING WEATHER CONDITIONS [−25 ◦C,−50 ◦C]. (A) I–V CURVES, (B) P–V CURVES 188 ♦ A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method
ICORE 2013
FIG. 2: CURRENT-VERSUS-VOLTAGE CURVE OF A PV MODULE
In Fig. 1, it is clear that the MPP is located at the knee of the I–V curve, where the resistance is equal to the negative of differential resistance.
V V =− I I
(3)
This is following the general rule used in the P&O method, in which the slope of the PV curve at the MPP is equal to zero
dP =0 dV
(4)
Equation (8) can be rewritten as follows:
dP dP dI = I. +V. dV dV dV dP dI = I +V. dV dV
(5) (6)
And hence
I. + V .
dI =0 dV
(7)
which is the basic idea of the IncCond algorithm. One noteworthy point to mention is that (3) or (4) rarely occurs in practical implementation, and a small error is usually permitted. The size of this permissible error (e) determines the sensitivity of the system. This error is selected with respect to the swap between steady-state oscillations and risk of fluctuating at a similar operating point.It is suggested to choose a small and positive digit. Thus,(6) can be rewritten as
I. + V.
dI =e dV
In this paper, the value of “e” was chosen as 0.002 on the basis of the trial-and-error procedure. The flowchart of the IncCond algorithm within the direct control method is shown in Fig. 3. According to the MPPT algorithm, the duty cycle (D) is calculated. This is the desired duty cycle that the PV module must operate on the next step. Setting a new duty cycle in the system is repeated according to the sampling time. A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method♦ 189
ICORE 2013
FIG. 3: FLOWCHART OF THE INCCOND METHOD WITH DIRECT CONTROL
SELECTING PROPER CONVERTER When proposing an MPP tracker, the major job is to choose and design a highly efficient converter, which is supposed to operate as the main part of the MPPT. The efficiency of switch-mode dc–dc converters is widely discussed in [1]. Most switching-mode power supplies are well designed to function with high efficiency. Among all the topologies available, both Cuk and buck–boost converters provide the opportunity to have either higher or lower output voltage compared with the input voltage. Although the buck–boost configuration is cheaper than the Cuk one, some disadvantages, such as discontinuous input current, high peak currents in power components, and poor transient response, make it less efficient. On the other hand, the Cuk converter has low switching losses and the highest efficiency among nonisolated dc–dc converters. It can also provide a better output-current characteristic due to the inductor on the output stage. Thus, the Cuk configuration is a proper converter to be employed in designing the MPPT. Figs. 4 and 5 show a Cuk converter and its operating modes, which is used as the power stage interface between the PV module and the load. The Cuk converter has two modes of operation. The first mode of operation is when the switch is closed (ON), and it is conducting as a short circuit. In this mode, the capacitor releases energy to the output. The equations for the switch conduction mode are as follows:
v L1 = v g ic1 = i2
v L 2 = −V1 − V2
ic 2 = i2 − V2 / R
FIG. 4: ELECTRICAL CIRCUIT OF THE CUK CONVERTER USED AS THE PV POWER-STAGE INTERFACE
FIG. 5: CUK CONVERTER WITH (A) SWITCH ON AND (B) SWITCH OFF 190 ♦ A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method
ICORE 2013
On the second operating mode when the switch is open (OFF), the diode is forward-biased and conducting energy to the output. Capacitor C1 is charging from the input. The equations for this mode of operation are as follows:
v L1 = V g − v1 vL2 = −V2
ic1 = i1 ic 2 = i2 − V 2 The principles of Cuk converter operating conditions state that the average values of the periodic inductor voltage and capacitor current waveforms are zero when the converter operates in steady state. The relations between output and input currents and voltages are given in the following: Some analyses of Cuk converter specifications are provided in, and a comparative study on different schemes of itching converters is presented in the literature. The cuk converter is developed with simulink and results are validated hrough simulations. The power circuit of the proposed system consists of a Cuk converter and a gate drive, and the control of the switching is done using the control circuit. The control tasks involve measuring the analog voltage and current of the PV module using current and voltage sensors, convert them to digital using an ADC, process the obtained information in a microcontroller, then them compare to the predefined values to determine the next step, revert the PWM to the gate drive, and hence control the switching of IGBTs. The control loop frequently happens with respect to the sampling time, and the main program continues to track the MPPs. SIMULATION RESULTS The diagram of the closed-loop system designed in MATLAB and Simulink, which includes the PV module electrical circuit, the Cuk converter, and the MPPT algorithm. The converter components are chosen according to the values presented in above. The PV module is modeled using electrical characteristics to provide the output current and voltage of the PV module. The provided current and voltage are fed to the converter and the controller simultaneously. The PI control loop is eliminated, and the duty cycle is adjusted directly in the algorithm. To compensate the lack of PI controller in the proposed system, a small marginal error of 0.002 is allowed. To test the system operation, the condition of changing irradiation was modeled. The first illumination level is 1000 W/m2; at t = 0.4 s, the illumination level suddenly changes to 400 W/m2 and then back to 1000 W/m2 at t = 0.8 s. An illustration of the relationship between the duty cycle and PV output power to demonstrate the effectiveness of the algorithm mentioned in the flowchart. the change in duty cycle adjusted by the MPPT to extract the maximum power from the module.
FIG. 6: GRID VOLTAGE AND GRID CURRENT A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method♦ 191
ICORE 2013
FIG. 7: VOLTAGE AND CURRENT AT THE LOAD SIDE
FIG. 8: PV DC VOLTAGE OF SOLAR MODULE
Fig. 6, 7and 8 shows the simulation resultsof converter voltage and converter current,load voltage and load current and solar module DC voltage results. From these results it is obsereved that the DC voltage of the panel is getting almost 500 volts. CONCLUSION In this paper a fixed step size incremental conductance MPPT with direct controlled method was employed and necessity of other control loop was eliminated and the proposed system was simulated, the functionality of the suggested concepts were studied. From these results it is obsereved that the DC voltage of the panel is getting almost 500 volts. REFERENCES [1] R.-J. Wai, W.-H. Wang, and C.-Y. Lin, ―High-performance stand-alone photovoltaic generation system, IEEE Trans. Ind. Electron., vol. 55, no. 1, pp. 240–250, Jan. 2008. [2] W. Xiao, W. G. Dunford, P. R. Palmer, and A. Capel,―Regulation of photovoltaic voltage, IEEE Trans. Ind. Electron., vol. 54, no. 3, pp. 1365–1374, Jun. 2007. [3] N. Mutoh and T. Inoue, ―A control method to charge series-connected ultra electric double-layer capacitors suitable for photovoltaic generation systems combining MPPT control method, IEEE Trans. Ind. Electron., vol. 54, no. 1, pp. 374–383, Feb. 2007. [4] R. Faranda, S. Leva, and V. Maugeri, MPPT Techniques for PV Systems: Energetic and Cost Comparison. Milano, Italy: Elect. Eng. Dept. Politecnico di Milano, 2008, pp. 1–6. 192 ♦ A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method
ICORE 2013
[5] Z. Yan, L. Fei, Y. Jinjun, and D. Shanxu, ―Study on realizing MPPT by improved incremental conductance method with variable step-size, in Proc. IEEE ICIEA, Jun. 2008, pp. 547–550. [6] F. Liu, S. Duan, F. Liu, B. Liu, and Y. Kang, ―A variable step size INC MPPT method for PV systems, IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2622–2628, Jul. 2008. [7] F. M. González-Longatt, ―Model of photovoltaic module in Matlab, in 2do congreso iberoamericano de estudiantes de ingenierıacute;a eléctrica, electrónica y computación, ii cibelec, 2005, pp. 1–5. [8] T. Esram and P. L. Chapman, ―Comparison of photovoltaic array maximum power point tracking techniques, IEEE Trans. Energy Convers., vol. 22, no. 2, pp. 439–449, Jun. 2007. [9] V. Salas, E. Olias, A. Barrado, and A. Lazaro, ―Review of the maximum power point tracking algorithms for stand-alone photovoltaic systems, Sol. Energy Mater. Sol. Cells, vol. 90, no. 11, pp. 1555–1578, Jul. 2006. [10] G. Petrone, G. Spagnuolo, R. Teodorescu, M. Veerachary, and M. Vitelli,―Reliability issues in photovoltaic power processing systems, IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2569–2580, Jul. 2008. [11] C. Hua, J. Lin, and C. Shen, ―Implementation of a DSP-controlled photovoltaic system with peak power tracking, IEEE Trans. Ind. Electron., vol. 45, no. 1, pp. 99–107, Feb. 1998. [12] T. Noguchi, S. Togashi, and R. Nakamoto, ―Short-current pulse-based maximum-power-point tracking method for multiple photovoltaic-andconverter module system, IEEE Trans. Ind. Electron., vol. 49, no. 1, pp. 217–223, Feb. 2002. [13] N. Mutoh, M. Ohno, and T. Inoue, ―A method for MPPT control while searching for parameters corresponding to weather conditions for PV generation systems, IEEE Trans. Ind. Electron., vol. 53, no. 4, pp. 1055–1065, Jun. 2006. [14] N. Femia, G. Petrone, G. Spagnuolo, and M. Vitelli,―Optimization of perturb and observe maximum power point tracking method, IEEE Trans. Power Electron., vol. 20, no. 4, pp. 963–973, Jul. 2005. [15] N. Femia, D. Granozio, G. Petrone, G. Spagnuolo, and M. Vitelli, ―Predictive & adaptive MPPT perturb and observe method, IEEE Trans. Aerosp. Electron. Syst., vol. 43, no. 3, pp. 934–950, Jul. 2007. [16] E. Koutroulis, K. Kalaitzakis, and N. C. Voulgaris, ―Development of a microcontroller-based,photovoltaic aximum power point tracking control system, IEEE Trans. Power Electron., vol. 16, no. 1, pp. 46–54, Jan. 2001. [17] S. Jain and V. Agarwal, ―A new algorithm for rapid tracking of approximate maximum power point in photovoltaic systems, IEEE Power Electron. Lett., vol. 2, no. 1, pp. 16–19, Mar. 2004. [18] A. Pandey, N. Dasgupta, and A. K. Mukerjee, ―Design issues in implementing MPPT for improved tracking and dynamic performance, in Proc. 32nd IECON, Nov. 2006, pp. 4387–4391. [19] K. H. Hussein, I. Muta, T. Hoshino, and M. Osakada―Maximum photovoltaic power tracking: An algorithm for rapidly changing atmospheric conditions, Proc. Inst. Elect. Eng.—Gener., Transmiss. Distrib., vol. 142, no. 1, pp. 59–64, Jan. 1995. [20] T.-F.Wu, C.-H. Chang, and Y.-H. Chen, ―A fuzzy-logic-controlled singlestage converter for PV-powered lighting system applications, IEEE Trans. Ind. Electron., vol. 47, no. 2, pp. 287–296, Apr. 2000. [21] M. Veerachary, T. Senjyu, and K. Uezato, ―Neural-network-based maximum-power-point tracking of coupled-inductor interleaved-boostconverter.
A New Approach for the Development of MPPT Algorithm with Short Circuit Method and Incremental Conductance Method♦ 193
Weibull Distribution Analysis of Sub-Hourly Surface Wind Data of Western Rajasthan P. Santra1, P.C. Pande2, Preeti Varghese3, N.D. Yadav4, P. Raja5 and N.K. Sinha6 1,2,3,5Division
of Agricultural Engineering for Arid Production Systems 4Regional Research Station, Bikaner 6Regional Research Station, Jaisalmer Central Arid Zone Research Institute, Jodhpur–342003 E-mail: [email protected]
Abstract—Renewable energy sources are considered as the key factor for future sustenance and development of human civilization because of two main reasons. Firstly, the conventional sources of energy are depleting at a fast rate so as to reach its finite limit in a quick time and secondly CO2 emission factor for using renewable energy sources is comparatively lesser than other energy sources, which is an important and positive favourable factor under present climate change situations. Considering the renewable energy as the future potential, surface wind energy resources in western Rajasthan has been assessed in this study for its future exploitation. Fine resolution surface wind speed data for the period 2010-2012 were analysed for assessing the wind resource potential. Wind speed distribution was fitted in Weibull probability distribution function and two Weibull parameters, k and c, were estimated on monthly basis for three locations, Jaisalmer, Chandan and Bikaner. The shape parameter, k, which indicates how peaked the distribution is, was observed to range from 1.01 to 3.01 among all three locations with maximum value during June. The scale parameter, c, which indicates how windy is the location, was observed 1.54-4.85, 1.01-5.55, and 1.16-2.24 m s-1 at Jaisalmer, Chandan and Bikaner, respectively. Furthermore, characteristics features of wind resource potential e.g. mean wind speed (vm), most probable wind speed (vmp), wind speed carrying maximum energy (vmax,E), wind power density (W m-2) and wind energy density (kWh m-2 month-1) were computed for four summer month (May-August) for each location. All these characteristics features revealed that surface wind energy potential is higher in Jaisalmer and Chandan than Bikaner and month-wise distribution showed June month is the most suitable. The results indicated that suitable vertical axis wind turbine (VAWT) may be designed and developed for harnessing available surface wind energy at Jaisalmer and Chandan region of western Rajasthan for off-grid power generation and thereafter utilization in agricultural farms and home appliances. Keywords: Wind Energy, Vertical Axis Wind Turbine (VAWT), Weibull Parameters, Automatic Weather Station (AWS)
INTRODUCTION Energy is a key ingredient for socioeconomic development and economic growth of a nation. The nonrenewable energy source like fossil fuels is depleting continuously and therefore attention is now focused on alternative source of energy. Renewable energy sources like wind energy is indigenous and can help in reducing the dependency on fossil fuels. The energy associated with wind is mainly due to its kinetic forces and increases with the cube of wind speed. The total annual kinetic energy of air movement in the earth atmosphere is estimated to be about 3 × 105 kWh or about 0.2% of the solar energy reaching the earth. The maximum technically usable potential is estimated to be theoretically 30 trillion kilo watt hours per year. Wind energy is expected to play an increasingly important role in the future national energy scene (Sen and Sahin, 1997; Sesto and Casale, 1998). Moreover, renewable energy resources such as wind, solar, biomass etc. are being highly demanded because these energy sources are environment friendly as it is not associated with much emission of CO2 in atmosphere as compared to others in the process of generating power. In western Rajasthan, most periods of a year has been found windy especially during summer season (~10-30 km hr-1) with an increasing gradient from east to west and thus have a great wind energy potential (Ramachandra and Shruthi, 2003). Such high wind speed causes erosion of top soil with frequent occurrence of dust storms during summer months. Utilization of such strong wind energy through converting it into electrical energy by installation of horizontal axis wind turbine (HAWT) at 70-80 m height has been progressed rapidly during last one decade with a total capacity of 2067 MW till 2012. Panda et al. (1990) made a stochastic analysis of the wind energy potential at seven representative weather stations in India and developed a probability model for the wind data and wind potential. It has been hypothesized that surface wind energy may also be extracted through designing suitable vertical axis wind turbine (VAWT) of Savonious or Darrieus design. Installation of such VAWT along mechanical barrier in field boundaries is expected to reduce the soil loss through wind erosion as well as to generate off-grid electric power. Considering this potential, surface wind energy resource [1.5-2 m above ground level (agl)] during summer months (MayAugust) at three selected locations in western Rajasthan has been assessed in this study. Weibull distribution parameters were calculated for these three locations and different wind resource parameters were computed for assessing the potential of surface wind energy.
ICORE 2013
MATERIALS AND METHODS Wind Speed Database Data on wind speed (m s-1) at hourly/sub-hourly interval for the period 2010-2012, recorded by cup anemometer sensors at three locations in western Rajasthan were collected and a database was prepared. At Jaisalmer and Bikaner station, wind speed was recorded at 1.5 m agl whereas at Chandan station, it was recorded at 2 m agl. The database on fine resolution wind speed data for western Rajasthan was further analysed in this study. The details of the database database are given in Table 1. Programming codes were developed in Matlab 7.1 for analysing the wind speed data. Diurnal variation as well monthly variation of wind speed for each location was computed. The sorted data was further analysed for estimating the Weibull parameters and computation of wind resource characteristics. TABLE 1: DETAILS ON WIND SPEED DATA FROM THREE LOCATIONS IN WESTERN RAJASTHAN, INDIA Station Name Geographical Wind speed Height of measurement location measurement system* Jaisalmer 1.5 m above ground 26°55' N, 70°57' Cup anemometer sensor of AWS surface E Chandan 2.0 m above ground 27°01' N, 71°01' Cup anemometer sensor of AMS surface E Bikaner 28º04' N, 74º03' Cup anemometer 1.5 m above ground E sensor of AWS surface *AWS = automatic weather station, AMS = Agro-meteorological station
Duration of Data
Data interval
April, 2010–December, 2012 December, 2011– December, 2012 April, 2010–December, 2013
5 minutes 30 minutes 1 hour
Weibull Parameter Estimation Weibull distribution was found as the best distribution to fit the measured wind speed data for different locations of the world (Ulgen and Hepbasli, 2002, Bansal et al., 2002; Keyhani et al., 2010). In this study, fine-resolution wind speed data was used to estimate Weibull parameters for each location. Weibull probability density function (pdf) of wind speed is written as follows:
⎛ k ⎞⎛ v ⎞ f (v) = ⎜ ⎟⎜ ⎟ ⎝ c ⎠⎝ c ⎠
k −1
⎡ ⎛ k ⎞k ⎤ exp ⎢ − ⎜ ⎟ ⎥ ⎣⎢ ⎝ c ⎠ ⎦⎥
(1) where f(v) is the frequency distribution of wind speed, v; k and c are the weibull shape parameter and weibull scale parameter, respectively. Both these parameters characterize the wind resource potential of any region. The shape parameter, k, indicates the peak height of the wind distribution. If the wind speeds tend to be very close to a certain value, the distribution will have a high k value and is very peaked. whereas the scale parameter, c, indicates how windy is the location and higher is the value more windy is the location or situation. Value of k has been reported to vary from 1.5 to 3.0 for most wind conditions. The Rayleigh distribution is a special case of the Weibull distribution in which the shape parameter, k is 2.0 (Spera, 1995; Persaud et al., 1999). Computation of Weibull Parameters From the pdf of weibull distribution (Eq 1), the cumulative distribution function (cdf) may be written as follows
⎡ ⎛ v ⎞k ⎤ F (v) = 1 − exp ⎢ − ⎜ ⎟ ⎥ ⎢⎣ ⎝ c ⎠ ⎥⎦
(2)
Solving the above weibull cdf (Eq 2), following form of equation may be obtained
ln {− ln [1 − F (v ) ]} = k ln(v ) − k ln c
(3) Linear plotting of ln(v) as abscissa variable vs ln{-ln[1-F(v)]} as ordinate variable will result into the slope of the linear trend as equal to the parameter k and intercept of the line equal to (-klnc). Thus, k and c parameter of the wind speed distribution may be estimated from measured frequency distribution of Weibull Distribution Analysis of Sub-Hourly Surface Wind Data of Western Rajasthan ♦ 195
ICORE 2013
wind speed. Estimation of these two parameters was further improved through optimization function of solver programme of excel spreadsheet by minimizing the fitting error. Wind Resource Potential Once Weibull parameters are estimated for different time periods of a location, several characteristic features of wind resources was calculated for real assessment of wind energy potential, which are as follows: Mean wind speed: It indicates average wind speed for a location for a specific time period and can be calculated from Weibull parameters as follows
⎛ 1⎞ vm = cΓ ⎜1 + ⎟ ⎝ k⎠
(4)
where, Γ is gamma function Most probable wind speed: The most probable wind speed denotes the most frequent wind speed for a given wind probability distribution and is expressed by 1
vmp
⎛ 1 ⎞k = c ⎜1 − ⎟ ⎝ k⎠
(5) Wind speed carrying maximum energy: Wind speed carrying maximum energy represents wind speed, which carries maximum wind energy and can be expressed as follows 1
vmax, E
⎛ 2 ⎞k = c ⎜1 + ⎟ ⎝ k⎠
(6)
vmax,E is also expressed as the optimum wind speed for a wind turbine, vop, which is the speed that produces the most energy. The wind turbine should be chosen with a rated wind speed that matches this maximum energy wind speed for maximizing energy output. Wind power density: The power of the wind that flows at speed v through a blade sweep area A may ଵ be calculated from kinetic energy of wind ( ߩ‫ ݒ‬ଷ ) and increases with the cube of the wind speed. ଶ Therefore, to assess wind resource potential of any area it is better to calculate wind power density rather than using wind speed as an indicator. Wind power density may be calculated from Weibull parameter as follows:
P 1 3 ⎛ k +3⎞ = ρc Γ ⎜ ⎟ A 2 ⎝ k ⎠
(7) The Betz limit states that a wind turbine would not extract more than 59.3% of the available wind power. Therefore, the maximum extractable power from the wind will be the product of the factor 0.593 and the calculated result from Eq. (7). Wind energy density: Once wind power density of a site is calculated, the wind energy density for a desired duration, T, can be calculated as
E 1 3 ⎛ k +3⎞ = ρc Γ ⎜ ⎟T A 2 ⎝ k ⎠
(8)
RESULTS AND DISCUSSION Diurnal Variation of Wind Speed Diurnal variation of wind speed for all three locations is plotted in Figure 1. In case of Jaisalmer, diurnal variation is plotted at 5 minute interval whereas for Chandan and Bikaner station it was plotted at 30 196 ♦ Weibull Distribution Analysis of Sub-Hourly Surface Wind Data of Western Rajasthan
ICORE 20 013
minutes annd 1 hr interval, which was actually the t measurem ment interval for each loccation. Each data point represent the averagee of (30×3) values correesponding to 30 days foor a month inn a year for a total of three years. The diurna al plot showss two peaks and a two trouughs for eachh location. Tw wo peaks are e generally observed during day time t roughly between 8:0 00 am and 4:00 4 pm and d night time rroughly betw ween 10:00 pm and 4:00 am. Cha aracteristic trooughs with loow wind spee ed are geneerally observeed during mo orning and evening tiime of a day. Among thhree locations, wind spee ed is higher in Chandan station than other two stations. Month-wise M distribution also showed thhat wind spe eed is higher in June montth than otherrs. It is also noted here that wind speed data of Chandann station was recorded at 2 m heightt from surfacce whereas e, it is not justifiable to ccompare thesse stations, for other two stations it was 1.5 m from surfacce. Therefore however, it can be sa aid that wind d resources in Bikaner station is comparatively leess than Jaissalmer and Chandan station. s
FIG. 1: DIUURNAL VARIATION N OF SURFACE WIND I SPEED (M S-1) AT THREE LOCATIONS IN WESTERN RAJASTHAN (EACH GRAPH REPRRESENTS THE AVEERAGE VALUE OF HOURLY AND SUB-HOURLY WIND SPEED DATA RECCORDED BY WIND SPEED SENSORS DURING 2010-20 012)
Monthly Distribution D of Wind Speeed Monthly average a wind d speed for three locatioons in westernn Rajasthan is given in Ta able 1. June month has been founnd more windy than restt months of a year for all a three loca ations. For Ja aisalmer and d Chandan station, avverage wind speed during June was >4 > m s-1 (~14.4 km hr-1), whereas forr Bikaner station, it was 1.92 m s-1. Winter moonths from October O to Ja anuary are calm c and thee wind speed d was 350 0.010 800 - 900 Air Stirling, Brayton Stirling engine High molten salt, ceramics, PCM Simple moderate Simple moderate moderate
TECHNOLOGY OF CONCENTRATED SOLAR POWER (CSP) Solar photo voltaic (PV) technology has been used for long to generate power from the sun. But its use has remained limited because of its low efficiency and other constraints. Now, concentrating solar power (CSP) is emerging as a viable alternative. Unlike PV cells which are flat, CSP involves the use of parabolic mirrors in long troughs, which concentrate solar irradiation into a centrally placed special tube which absorbs radiation. The heat generated into the tube is then used to produce electricity. CSP plants generate electricity from sunlight by focusing solar energy, collected by an array(s) of suntracking mirrors called heliostats, onto a central receiver. There are different types of CSP technologies currently available. In one typical type (parabolic trough with liquid salt), liquid salt (a mixture of sodium nitrate and potassium nitrate) is circulated through tubes in the receiver, absorbing the heat energy gathered from the sun. The heated salt is then routed to an insulated tank where it can be stored with minimal energy losses. To generate electricity, the hot molten salt is routed through heat exchangers and a steam generation system. The steam is then used to produce electricity in a conventional steam turbine. After exiting the steam generation system, the now cool salt mixture is circulated back to the “cold” thermal storage tank, and the cycle is repeated. There are various other types’ heat transfer media which are used. One of the biggest benefits of CSP is its ability to store thermal heat which ensures continuous supply of power even when the sun is down and when it gets cloudy. The core of a CSP plant is the special tube placed at the focal point of parabolic mirrors. The mirrors have a motorized system that enables them to 356  Solar Thermal Technology in India: Issues and Opportunities
ICORE 2013
keep tracking the sun. Inside the glass tube is a steel tube with absorber separated by vacuum insulation. The ceramic metallic tube contains synthetic oil which gets heated up to 400 deg Celsius. The oil flows into a heat exchanger which generates steam, which in turn, is used to generate power using conventional steam turbines. The receivers have to achieve maximum solar absorption and at the same time minimal emission of heat. The surface of the glass tube remains cold, while the temperature in the steel tube could go up to 400 deg Celsius. PRE-REQUISITES FOR CSP CSP Technologies rely on Direct Beam radiation for operation, i.e. direct radiation from the Sun that has not been diffused or deflected by clouds or other atmospheric factors and so can be focused by the mirrors. Ideally the data needed to assess potential sites is of short interval duration and hence Direct Normal Irradiation (DNI) measurements are collected over several years. The considerate are: In designing a CSP plant, knowledge of the seasonal variation in DNI resource is needed to make an optimal economic assessment of the degree to which the solar field is oversized relative to the power block in the high season and undersized in the low season. Power blocks have an efficiency that decreases with part load (due to reduced solar input) and a threshold load under which they cannot operate. Thus to predict the power block’s total daily output, knowledge of the time dependence of input heat transfer fluid energy flow is required. The thermal receivers in collector fields take many minutes to reach operating temperature from cold, hence to accurately predict their output, data at time resolutions of one minute or less is required to predict performances in situations of intermittent cloud cover. CONCENTRATED SOLAR POWER – ROADMAP FOR INDIA Concentrated Solar Power (CSP) or solar thermal power plants is renewable energy's answer to bulk power supply. The CSP market has witnessed a global uproar in the past year not only in terms of installations but also in terms of developer interest. Currently, the world cumulative pipeline has reached up to 20 GW which accounts for the projects that are under development and also that are under construction. The International Energy Agency (IEA) projects that by 2050, CSP plants would supply up to 11.3% of global electricity. The Indian home ground is also not far behind with an ambitious target of 10 GW by 2022 from solar thermal power plants alone, with plans to install 550 MW of CSP by next year. India had planned to set up a 35 MW plant based on CSP in Rajasthan way back in 1994, but the project was postponed indefinitely because no qualified contractors were able to submit a bid. Now, of course, the situation has changed. The Jawaharlal Nehru National Solar Mission recognizes CSP as a key source of renewable power. The Ministry of New and Renewable Energy (MNRE) is setting up a string of demonstration projects. A 10 MW CSP plant is already under construction near Bikaner by ACME, a private company. Ministry officials said CSP could be an attractive option for the country, given the fact that large areas of northwest India fall in the high radiation zone. The costs can be brought down if concentrating mirrors and receiving tubes are manufactured locally. The vast Rajasthan Desert is quite similar to the Sahara desert in Africa, and has the potential to become the largest solar power plant in India. Due to high levels of available sunlight, CSP plants in Rajasthan could satisfy most of India’s energy needs in just a few years. India’s potential benefits from solar power are as numerous as the sands of Rajasthan desert, and include reduced dependence on fossil fuels and a cleaner environment. These benefits can be realized by installing renewable energy technologies, such as CSP, to protect the environment while diversifying energy resources and helping to lower prices. Solar power can also reduce strain on the electric grid on hot summer afternoons, when air conditioners are running, by generating electricity where it is used National Thermal Power Corporation (NTPC) has announced plans to install two 50MW solar thermal plants in Gujarat and could extend this to 300MW. It is also building 5MW and 1MW installations on the Andaman and Nicobar islands. JAWAHARLAL NEHRU SOLAR MISSION (JNSM) The launch of the National Solar Mission has given a big impetus to solar energy in India. The highlights of the mission are given below: Solar Thermal Technology in India: Issues and Opportunities  357
ICORE 2013
 Installed capacity of 20000 MW by 2022  Establishment of a single window investor friendly mechanism  Envisages an investment of Rs 90,000 crores over the next 30 years  Initial investment of Rs 4,337 crores provided by the Government of India The first phase of the Jawaharlal Nehru National Solar Mission (JNNSM) is scheduled to be completed by 2014 only. However, as per progress available so far, all the application segments of the Mission are going on as scheduled and on target. The details of projects sanctioned under the Phase-I of the JNNSM are as under:  Projects under Migration Scheme (84 MW - Solar PV 54 MW & Solar Thermal 30 MW)  Projects under New Project Scheme (Batch-I) (Solar PV 150 MW & Solar Thermal 470 MW)  Projects under New Project Scheme (Batch-II) (Solar PV 350 MW) 
Projects under Roof Top PV and Small Solar Power Generation Programme (RPSSGP) – 98.05 MW.
CSP – WORLDWIDE More than ten different technology combinations for CSP installations are available world-wide. More than 500 MW installed capacity using Parabolic Trough technology, 40 MW from Solar Tower, 5 MW from Fresnel technology and 0.5 MW from Dish technology are already working. Over 800 MW capacities project has been announced worldwide under CSP technology.CSP is beneficial in cogeneration mode like Power & Steam and more so in Tri-generation modes like Power, De-salination and Heat Generation. CSP has supplement conventional power plant for feed water heating in several installations. CROSS LINEAR CONCENTRATED SOLAR POWER (CL-CSP) TECHNOLOGY This concentration system is situated between point and linear focusing concentration. Hence with the CL system we can get a higher temperature around 800 degree C by applying linear focussing method. A joint collaboration between Japanese and Indian industries, institutes and universities was started to develop Cross Linear technology in August 2011. The institutes and universities are Tokyo Institute of Technology, Delhi Technological University, and Rajiv Gandhi Technological University (RGPV), Toyo Corporation, Japan with support from Solar Energy Centre (MNRE). The levelized electricity production cost (LEC) is expected to be lower by 20-30% compared to the tower top system on account of Solar beam energy collected by CL is higher by nearly 20% as compared to Linear Fresnel. Thus, the CL system seems to be promising solar concentration system to use the solar heat in a wide temperature range of 300-800 degree C at the lowest production cost among the existing solar concentration systems, as evident from the following table. TABLE 3: CROSS LINEAR CONCENTRATED SOLAR POWER (CL-CSP) TECHNOLOGY Temperature
Cross Linear 300-800 deg C
Concentration
100-1000
Thermal Fluid
Liquid: Water, Oil Gas: Air, Steam, CO2
CL Heliostat
Axis: 1.01 Control precision: Moderate or Low Cavity, CPC, Pipes
CL Receiver
(Advantages of CL compared to Tower, Trough and Linear Fresnel)
358  Solar Thermal Technology in India: Issues and Opportunities
Tower, Trough, Linear Fresnel Tower: 600 deg C Trough: 400 deg C Linear Fresnel: 500 deg C Tower: 300-1000 Trough, Linear F

Thèse de Doctorat école doctorale sciences pour l’ingénieur et microtechniques
U N I V E R S I T É
D E
F R A N C H E - C O M T É
Robust control for manipulation inside a scanning electron microscope n Marcelo Gaudenzi de Faria
Thèse de Doctorat école doctorale sciences pour l’ingénieur et microtechniques
U N I V E R S I T É
D E
F R A N C H E - C O M T É
` THESE pr´esent´ee par
Marcelo
Gaudenzi de Faria pour obtenir le
Grade de Docteur de l’Universit´e de Franche-Comt´e Sp´ecialit´e : Automatique
Robust control for manipulation inside a scanning electron microscope Unit´e de Recherche : FEMTO-ST, UMR CNRS 6174
Soutenue publiquement le 17 F´evrier 2016 devant le Jury compos´e de : Florence
Pr´esident Rapporteur
Laroche Lutz Yann Le Gorrec
Rapporteur Directeur de th`ese Co-directeur de th`ese
Sergej
Fatikow Marchi
Edouard Philippe
Yassine
Haddab
Co-directeur de th`ese
Professeur a` l’Universit´e d’Oldenburg - DE Maitre de conf´erences-HDR a` l’Universit´e Grenoble Alpes - FR Professeur a` l’Universit´e de Strasbourg - FR Professeur a` l’Universit´e de Franche-Comt´e - FR ´ Professeur a` l’Ecole Nationale Sup´erieure de M´ecanique et des Microtechniques - FR Professeur a` l’Universit´e de Montpellier - FR
Acknowledgements
The work presented here were developed in the Automation and Micro-Mecratronic Systems (AS2M) department, at the FEMTO-ST Institute, which I would like to thank for receiving me during this period. I would like to thank Prof. Philippe Lutz, director of the graduate school and my advisor, for the constant support and motivation, allowing me to complete this thesis within the best possible conditions. My thanks also to my co-advisors: Prof. Yassine Haddab for the patience, friendship and the advices over several technical aspects for this work, which were essential, and Prof. Yann Le Gorrec for his constant support and guidance for my multiple questionings in the last years. I can not forget all the FEMTO-ST team, specially Patrick Rougeot, Jean-Yves Rauch and Olivier Lehmann, for all the aid offered with the experiments and the operation of the scanning electron microscope. My thanks to all my friends that, in the course of this thesis, contributed with extensive laughter and advice: Adrian, Ahmed, Alexandra, Am´elie, Andrey, Ar`ef, Babs, Bassem, Benoit, Bilal, Didace, Elodie, Francisco ’Paco’, H´ector, Haitem, Hussein, I˜nigo, Jo˜ao Stotz, Kamran, Louise, Marco, Margot, Marina ’Tava’, Marine, Martina, Max, Michelle, Mohamed, Mouloud, Nandish, Nathalie, Patricia, Racha, Sergio, Taha, Tristan, Valerian, Vincent, Viki, Vladmir, Wiem, Yannis, and many others. Finally, I would also like to thank my family that, despite the distance, supported me and helped to make this work possible.
To you all, my most sincere thanks.
v
Contents
General introduction
1
1
Challenges for robotics in nano-characterization and nano-manipulation
5
1.1
The need for micro- and nano-objects characterization . . . . . . . . . . . . . .
6
1.2
The NanoRobust project - a platform for nano-manipulation in SEMs . . . . . .
9
1.2.1
Common tasks performed in SEMs . . . . . . . . . . . . . . . . . . . .
9
1.2.2
Scope of the NanoRobust project . . . . . . . . . . . . . . . . . . . . .
11
Overview on nanorobotics system in SEM . . . . . . . . . . . . . . . . . . . . .
12
1.3.1
The scanning electron microscope . . . . . . . . . . . . . . . . . . . . .
12
1.3.1.1
Interaction between electron beam and matter . . . . . . . . .
13
1.3.1.2
SEM imaging characteristics . . . . . . . . . . . . . . . . . .
14
1.3.2
Nanomanipulation systems . . . . . . . . . . . . . . . . . . . . . . . . .
16
1.3.3
End-effectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
1.3.3.1
Needle tips . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
1.3.3.2
Micro-grippers . . . . . . . . . . . . . . . . . . . . . . . . . .
20
1.3.4
Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24
1.3.5
Examples of robotized tasks carried in SEMs . . . . . . . . . . . . . . .
26
1.3.5.1
Tele-operated manipulation and characterization in SEMs . . .
26
1.3.5.2
Development of autonomous tasks in SEM . . . . . . . . . . .
30
1.3.5.3
Improving performance and precision in SEM through the use of models . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
Thesis context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
33
1.4.1
Review of the issues presented . . . . . . . . . . . . . . . . . . . . . . .
33
1.4.2
Thesis objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35
1.3
1.4
2
Characterization and analysis of disturbance on SEM
37
2.1
Overview on sources of disturbances occurring in the SEM . . . . . . . . . . . .
37
2.1.1
Mechanical and acoustic disturbances . . . . . . . . . . . . . . . . . . .
38
2.1.2
Electromagnetic disturbances . . . . . . . . . . . . . . . . . . . . . . .
39
2.1.3
Thermal disturbances . . . . . . . . . . . . . . . . . . . . . . . . . . . .
41
vii
viii
CONTENTS
2.1.3.1
Thermal noise . . . . . . . . . . . . . . . . . . . . . . . . . .
41
2.1.3.2
Thermal effects on sensors and micro-structures . . . . . . . .
41
2.1.3.3
Heating due to electron beam . . . . . . . . . . . . . . . . . .
43
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
45
Vibration measurement inside a SEM . . . . . . . . . . . . . . . . . . . . . . .
46
2.2.1
Vibrometer characterization . . . . . . . . . . . . . . . . . . . . . . . .
47
2.2.2
SEM disturbances and noises identification . . . . . . . . . . . . . . . .
49
2.2.2.1
System under regular operation condition . . . . . . . . . . . .
50
2.2.2.2
System under external acoustic excitation . . . . . . . . . . . .
54
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
56
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
57
2.1.4 2.2
2.2.3 2.3 3
Micro-gripper dynamics identification in vacuum and at atmospheric pressure
59
3.1
The FT-G30 micro-gripper . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
59
3.2
Identification of actuated finger at atmospheric pressure . . . . . . . . . . . . . .
62
3.2.1
Experimental identification for the gripper in air . . . . . . . . . . . . .
63
3.2.2
Effects of electronic input disturbance over the electrostatic comb-drive in air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
65
Identification of actuated finger in vacuum . . . . . . . . . . . . . . . . . . . . .
66
3.3.1
Nonlinear stiffness effects . . . . . . . . . . . . . . . . . . . . . . . . .
66
3.3.2
Experimental identification for the gripper in vacuum . . . . . . . . . . .
68
3.3.3
Effects of electronic input disturbances over the electrostatic comb-drive in vacuum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
70
3.4
Resume of differences between air and vacuum . . . . . . . . . . . . . . . . . .
71
3.5
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
74
3.3
4
Position control in the micro-world
75
4.1
Control strategies for nano-positioning . . . . . . . . . . . . . . . . . . . . . . .
75
4.1.1
Examples of positioning tasks inside the SEM . . . . . . . . . . . . . . .
78
4.1.2
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
80
Robust position control strategies for a micro-gripper in vacuum . . . . . . . . .
80
4.2.1
Uncertain systems and the LFT representation . . . . . . . . . . . . . . .
82
4.2.2
µ analysis - a tool for robust stability and performance evaluation . . . .
83
4.2.3
H∞ controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
85
4.2.4
Extended state observer structure . . . . . . . . . . . . . . . . . . . . .
88
4.2.4.1
90
4.2
High-gain observer . . . . . . . . . . . . . . . . . . . . . . .
CONTENTS
4.3
4.4
4.5 5
ix
Controller design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
4.3.1
Controller synthesis using H∞ . . . . . . . . . . . . . . . . . . . . . . .
94
4.3.1.1
Selection of the weighting functions . . . . . . . . . . . . . .
94
4.3.2
Controller synthesis using ESO . . . . . . . . . . . . . . . . . . . . . .
97
4.3.3
Robustness analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . .
99
4.3.4
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
103
Experimental validation and results . . . . . . . . . . . . . . . . . . . . . . . . .
104
4.4.1
Reference tracking and steady state error under environmental disturbances 106
4.4.2
Disturbance rejection for external mechanical and acoustic perturbations
108
4.4.3
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
110
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
111
Conclusions and perspectives
113
5.1
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
113
5.2
Future perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
114
Personal publications
117
References
119
List of Figures
133
List of Tables
139
List of Definitions
141
I
143
Appendix
A Atohm SD28CR08F datasheet
145
B The Linear Fraction Transformation
147
C Noise impact on high-gain observer
149
List of Acronyms and abbreviations
AFM - Atomic force microscope CNT - Carbon nano-tube DO - Disturbance observer DoF - Degree of freedom EMI - Electromagnetic interference ESO - Extended state observer EHGO - Extended high-gain observer FIB - Focus ion beam FFT - Fast Fourier transform FT-G30 - FemtoTools Gripper 30, a model of micro-gripper G0, G1, G2 and G3 - Gains for the vibrometer’s analog output LFT - Linear fraction transformation LMI - Linear matrix inequality LQG - Linear quadratic gaussian MEMS - Micro electro-mechanical system MINIMAN - Miniaturized Manipulation, a family of robots MPC - Model predictive control Nano-FET - Nanowire field-effect transistor NCA - Nanoparticle chain aggregates OL - Open loop P/PI/PID - Proportional (P), integral (I) and derivative (D) controllers PMN-PT - Lead (Pb) magnesium (Mg) niobate (Nb) - lead titanate (Ti), a ceramic piezoelectric material PSD - Power spectral density RMS - Root mean square SEM - Scanning electron microscope SM - Sliding mode SNOM - Scanning near-field optical microscope SSV - Structured singular value STEM - Scanning transmission electron microscope xi
xii
STM - Scanning tunneling microscope TEM - Transmission electron microscopes UIO - Unknown input observer
LIST OF ACRONYMS AND ABBREVIATIONS
General introduction
Advances in nanotechnology allowed the development of new materials and the production of micro and nano-components with complex structures and geometries in large scales. Nanoobjects, structures with one or more dimensions under 100 nanometers, and nano-materials have gathered the attention of various research fields (i.e. biology and medicine, chemistry, physics, computer science, engineering, and others). Exploring the unique properties, effects, and phenomena only existing at these scales allowed to develop various devices, by improving existing ones or by exploring news designs. Despite the evident progress brought by the study and manipulation of matter at these minuscule scales on the past years, various obstacles hinder the advances in this field. The characterization of various nano-structures (single walled carbon nanotubes, graphene membranes, nano-particles, ...) is an essential step towards improving our knowledge over fabricated nano-structures. It is usually performed only on a reduced amount of samples, with the assumption of consistent properties over a larger number of individual. This method presents various limitations in face of current demands. Current fabrication methods can produce micro- and nano-objects in large quantities, and individual characterization (mechanical, chemical and electrical) becomes necessary. This more extensive process can deepen our knowledge over sample’s singular properties, improve fabrication methods and detect faults, and help in the selection of suitable individuals with uniform characteristics from a larger group. The minuscule operation scale required for manipulation and characterization of these nano-objects motivated the development of various strategies. When operating at micro and nanoscales, visual feedback is essential and commonly obtained with the aid of microscopes. Among them, the scanning electron microscope (SEM) is a popular tool. This device offers a large magnification range and can present relatively fast image acquisition speeds (up to tens of Hertz in some cases). Electron microscopes work by bombarding samples with a focused electron beam and are therefore not limited to the wavelength of visible light. As result, resolutions close to 1 nanometer can be achieved. In addition, the 3D manipulation of objects possible and often explored in this device. This microscope operates in high vacuum, necessary to reduce deviations of the focused electron beam during its path from source to sample, bringing some advantages and disadvantages. By performing various tasks in this environment, such as characterization and assembly of devices, the risk of contamination is reduced. At the same time, operating remotely with a restricted work volume brings up new challenges, linked to the limitations this imposes over the instrumentation, where compact yet accurate sensors are required and sensor placement can become an issue. Inside a SEM, all tasks are carried out with the aid of robotic manipulators and stages, using information from image, sensors and computational models. These tasks can be fully teleoperated or present various degrees of automation. The manipulators are usually equipped with micro-grippers or needle-like end-effectors, allowing to grasp, characterize and assemble various structures. Despite the success obtained so far in these tasks, problems do exist when performing them, specially when high precisions are needed. When component dimensions are reduced, they become more fragile and prone to the effects of disturbances and interaction forces. Sensors can 1
2
GENERAL INTRODUCTION
present unfavorable signal-noise ratios and their acquisition frequencies may not be adequate to capture the fast dynamic events in this scale. To reduce the chances of damaging both samples and end-effectors, movements should be carefully planned and slowly executed. Even when these strategies are employed, various researchers point out achieving high-precision positioning is not an easy feat. This can be partially attributed to the lack of reliable real-time information about the end-effectors position, specially over its dynamic behavior. The works described in this document were developed in the scope of the NanoRobust project. It aimed for the robust manipulation and characterization of micro/nano-components and structures inside a scanning electron microscope. The term “robust” here expresses the desire to reduce the occurrence of failures when performing these tasks. In the project framework, this thesis explored the dynamic characteristics of a micro-gripper, considering various factors that could contribute to the degradation of its precise positioning. Conjointly, an analysis of the pressure effects over the end-effector was performed, indicating a few particular consequences of operating in a vacuum environment. The thesis goal was to develop a robust control strategy for precise positioning of an endeffector inside the SEM. To measure the dynamic response of components in real-time, methods based on SEM image analysis are not well suited. For this reason, a dedicated setup was proposed, and included a high-resolution vibrometer to provide real-time position data used for measuring the end-effector and disturbance dynamics inside the vacuum chamber. The capacity to perform in situ identification of these elements is a singular characteristic of this work. This often unavailable and overlooked information was employed in the development of control laws for the end-effector position, aiming to achieve improved performance characteristics in terms of response speed, precision, accuracy and disturbance attenuation. This document is structured as follows: Chapter 1 introduces the importance of handling and characterizing micro and nanostructures, and how currently employed strategies experience difficulties towards achieving the required precisions and accuracies to manipulate nano-structures. The project NanoRobust aims to improve various aspects of the nano-manipulation, including the precise positioning of endeffectors and sensors. A short literature review is presented, covering various applications of manipulation in SEMs, in order to identify the causes of the reduced precision experienced and to delimit the objectives for this work. Chapter 2 describes the disturbance problem in a SEM through a survey on various perturbation sources affecting this system. Based on this review, an experimental setup was proposed to directly measure a structure displacement inside the vacuum chamber. The employed sensor, a laser vibrometer, was sufficiently fast to capture the most important dynamic effects in this environment. This setup studied the influence of external mechanical and acoustic disturbances inside the SEM. As consequence, a model for these disturbances considering its frequency spectrum and amplitudes was obtained. Chapter 3 presents the end-effector taken as case-of study for this work: the FT-G30 micro-gripper. This device was dynamically characterized in air and vacuum, and various differences in behavior related to pressure variation are commented. This highlights one of the common neglected aspects of SEM operation, as the modeling of devices in the vacuum can easily become a complex task. Chapter 4 covers the nano-positioning control problem, focusing on SEM applications. A short review over control strategies used for nano-positioning is presented. Based on this, and considering the desired performance requirements for this task, two control strategies were se-
3
lected: H∞ control and Extended State Observer based control. These controllers were developed based on the performance requirements, the identified models and the information about disturbances collected previously. Their robustness was tested through simulation and experimentally validated. Finally, Chapter 5 presents some final remarks about this work and summarizes its contributions. Also, a brief discussion over ideas and perspectives for future developments are proposed.
1 Challenges for robotics in nano-characterization and nano-manipulation
The importance of nano-technology, the capability of producing and handling micro and nano-sized objects and materials with novel unique properties, is undeniable. It has the potential to restructure current technologies applied for material science, manufacturing, computation, medicine and health, aeronautics, environment and energy, biotechnology, defense and security, as well as transportation and communication [Fukuda et al., 2013]. Miniaturization and nanotechnology can impact the production of virtually every man-made object. Figure 1.1 exemplifies some objects that can be found in the nature or manufactured, and where they location in the dimensional scale. water molecule
100pm
1nm
DNA helix
10nm
carbon nanotubes
virus
100nm
visible light spectrum
red blood cell
1um
10um
bacteria
MEMS
100um
1mm
human hair
tennis ball
10mm
100mm
grain of salt
Figure 1.1: Size comparison for various objects, in logarithmic scale. The International Organization for Standardizations (ISO) defines nano-objects as materials or structures which have at least one of its dimensions in the nano-scale, the range from approximately 1 nanometer to 100 nanometers. They can be classified as nano-plates, nano-fibers and nano-particles when one, two or its three dimensions are in this range, respectively. However, it is not rare to find in the literature the use of these terms for objects with dimensions up to a 5
6
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
few hundred nanometers. Similarly, the term micro-object includes structures with at least one dimension ranging from 0.1 to 100 µm. The properties of matter at nano-scales are not necessarily predictable compared to those observed at larger scales. The control over materials and the understanding of the phenomena taking place at these scales can lead to lighter, stronger materials and to the creation of devices exploring principles and architectures in ways not possible at macroscopic scales. Recent developments based on nano-technology include miniatures particle accelerators [Peralta et al., 2013], more efficient super-capacitors [Wu et al., 2013] and fuels cells, nanoparticles for recognition and marking of biological material [Otsuka et al., 2012], DNA injectors [Aten et al., 2014], quantum-dot solar cells[Semonin et al., 2011], smaller and more efficient transistors based on carbon nanotubes [Franklin et al., 2012], and various others. Through these technologies, miniaturized devices that integrate multiple functions for various application can be contemplated. Yet, to achieve this stage, further developments in the study, assembly and handling of micro- and, more importantly, nano-objects is necessary. This chapter expresses the importance of accurate handling and characterization for micro- and nano-objects, and reveal some of the conditions limiting the advances on this domain. These issues are the field of interest for the NanoRobust project, a group effort between various research centers in France. This project aims to tackle some of the existing challenges on nano-manipulation, including the challenge in obtaining the highaccuracies required when positioning structures (samples, end-effectors, sensors), which forms the background for this thesis. Several practical examples of micro- and nano-manipulation from the literature were reviewed, searching for indications that could help identify and evaluate the causes for the presented degradation in these tasks. This information guides and assist in outlining the main objectives for this work.
1.1/
The need for micro- and nano-objects characterization
Micro- and nano-structures and materials have been objects of study for many decades. The advances brought by these studies led to breakthrough in miniaturization, the discovery and development of new materials and a better understanding of the phenomena at these scales. Yet, our knowledge over the produced objects and our domain over fabrication methods is far from perfect. Multiple technologies currently employed in the fabrication of micro and nano-devices originated from existing ones developed for fabrications of macro-scale objects. Methods such as micro-machining and laser cutting were derived from their macroscopic counterparts, and can be traced back to the mid 1960’s through the early 1970’s. Fabrication methods were developed to follow the rapid growth of the micro electro-mechanical system (MEMS) industry. Today, several approaches can be used to produce a large variety of components and devices. These methods can be classified into two major groups: bottom-up and top-down. The bottom-up approach is based on using simple building blocks, sometimes at the molecular or atomic levels, to create more ordered, complex, 2D or 3D structures. Chemists and biologists have pioneered these methods, and fabrication strategies include chemical synthesis and self-assembly, such as the DNA origami [Zadegan et al., 2012]. One of the most popular examples of bottom-up approach is the creation of carbon nanotubes, nano-structures which are grown in a controlled way, over a substrate. Figure 1.2(a) shows an example of grown CNTs. The challenge in this method is to control the synthesis and assembly of molecules into the desired structures. Methods based on self-assembly are still being studied and are not yet fully understood
1.1. THE NEED FOR MICRO- AND NANO-OBJECTS CHARACTERIZATION
7
[Cui, 2009]. The top-down approach is that which produces a desired object from a starting, larger one. This approach includes various types of lithographic techniques and nano-imprinting methods to cut, etch and grind a surface into smaller structures with a desired shape, or by adding material to a negative mold. These methods are generally applied in the fabrication of structures ranging from a few hundreds of micrometers to hundreds of nanometers. Figure 1.2(b) shows an example of an electrostatic angular actuator fabricated using top-down strategies, with a honeycomb structure and maximum diameter of 2.6 mm. Recent advances in the fabrication process allowed for the reduction of the minimum feature size to 32 nanometers for beam lithography, and up to 12 nm through dip-pen nano-lithography [Ziaie et al., 2007].
(a) Carbon nanotube forest, grown over a substrate (b) Angular electrostatic micro-motor, fabricated [Gjerde et al., 2006]. through a top-down approach [Horsley et al., 1998].
Figure 1.2: Examples of micro- and nano-objects produces with different fabrication methods. Despite the current capacity to produce nano-structures in relative large quantities, in many cases their properties are not completely known. Furthermore, fabrication processes may produce unexpected results, a direct consequence of our limited knowledge and control over the various phenomena at this scale. The study and characterization of the produced objects allows to improve both aspects. The following motivation examples can be found in literature: • Carbon nanotubes (CNT) demonstrate interesting properties for micro and nano-system. They can present one (single-walled CNTs, with diameters between 0.4 and 3 nm) or multiple layers (multi-walled CNT, diameters between 1.4 and 100 nm). The measured Young modulus for those presents a large variation across the literature and is attributed to differences in the measurement strategies, fabrication methods and measurement errors [Fukuda et al., 2003]. Also, single-walled CNTs can be either metallic or semiconductor, each one with different applications, requiring its identification before using them. • A series of star-shaped structures produced by electron beam lithography, with sizes ranging from 1 to 4 µm and under 50 nm thickness, is shown in Figure 1.3. In the center of the stars, other suspended structures were added (circles and other stars, 500 nm width) [Zhu et al., 2015]. This study considered different fabrication parameters for each individual structure, to observe its influence on various scales. As a result, thousands of relatively complex three-dimensional structures were produced, each one slightly different from each other. Unfortunately, working with objects at these scales is not straightforward, as the difficulties in safely handling these structures increase with the reduction of their dimensions. Figure 1.4 demonstrates an example of how those samples can be easily damaged. In
8
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
Figure 1.3: Structures produced by electron beam lithography. The smaller ones were 500 nm wide and 50 nm tall [Zhu et al., 2015]. [Abrahamians et al., 2014], the characterization of thin membrane-like structures, close to 200 nm in thickness, is described with the aid of a contact probe. For a successful measurement, the same structure should be inspected multiple times at different points. Despite the high-quality experimental setup available and the attention paid during tele-operated manipulation, early results presented a low success rate as one single failure in a measurement set could permanently damage the sample, preventing the completion of the task. Possible causes for these results included the electrical charging of the sample and the tip as well as difficulties in precise tip positioning, resulting in collisions.
(a) InP membrane, 200 nanometers thick
(b) Damaged membrane and probe
Figure 1.4: Characterization of a nano-structure in the SEM. (a) Sample before starting the measurements. (b) Membrane damaged by the measuring probe [Abrahamians et al., 2014]. There is a demand for more precise, repeatable handling, processing and characterization of nano-objects. As seen in the examples above, the difficulties imposed by scale result in an increased failure rate and a high variability in the obtained measurements. [Sitti, 2001] cites, as a rule of thumb, that the accuracy of actuators and end-effectors should be at least 10 to 100 times better than the size of the manipulated objects, depending on the task. This means that, in some cases, accuracies of a few nanometers or less are necessary. The precise handling of objects at these scales is difficult and involves multiple, complex domains. This is something the NanoRobust project, introduced in the next section, sets to investigate. It aims to improve various aspects of the nano-manipulation, searching to develop a robotic platform with increased
1.2. THE NANOROBUST PROJECT - A PLATFORM FOR NANO-MANIPULATION IN SEMS9
performance, robustness and reliability.
1.2/
The NanoRobust project - a platform for nano-manipulation in SEMs
When working with nano-objects, the magnitudes of forces, displacements and other quantities often require the use of specialized robotic structures, actuators and sensors. Unfortunately, as seen earlier, some issues arise due to the scale, where various factors such as disturbances, interaction forces and other unknown effects can have major effects on the system, deteriorating its performance. These undesired, less known phenomena contribute to the challenges in studying and characterizing samples and processes at the micro- and nano-metric scales. The NanoRobust Project consists of a partnership between four research centers in France, to develop and study a robotic platform for nano-manipulation and multi-physical characterization in situ inside the scanning electron microscope. The centers involved in this project are: • Franche-Comt´e Electronique M´ecanique Thermique et Optique - Sciences et Technologies (FEMTO-ST) Institute, in Besanc¸on • Institut des Syst`emes Intelligents et de Robotique (ISIR), in Paris • Institut de Recherche en Informatique et Syst`emes Al´eatoires (IRISA), in Rennes • Laboratory for Photonics and Nanostructures (LPN), in Marcoussis In this section, the scope and objectives of the NanoRobust project are introduced, aiming to improve the quality of nano-manipulation tasks inside the SEM. Next, a brief introduction about the works commonly performed in SEMs is made, highlighting the importance of precise control over position during nano-manipulation tasks.
1.2.1/
Common tasks performed in SEMs
Robotic systems are employed to remotely work nano-structures inside a SEM. These systems should be compact, so fitting all necessary components in the limited space of the vacuum chamber is possible without limiting their mobility. More important, the robotic system task is to provide a link between vastly different scales, where quantities in the micrometric, nanometric or even picometric scales should be translated into values more suitable for human interaction, and vice versa. With the SEM, large image amplification is possible, making a bridge between subnanometric scales and human vision. However, producing and measuring forces, displacements, voltages, and other physical quantities, and linking them to the macro-world, are tasks attributed to nano-robotic systems and their components. In the micro and, especially, the nano-world, the mentioned tasks can become real challenges, as interaction forces, uncertainties and disturbances have an increased influence while precision and accuracy requirements grow to be more strict. When operating in a SEM closed environment, tasks need to be performed remotely. This usually involves the coordination of various elements (robots, platforms, end-effectors, ...) to achieve a complex goal. During manipulation and characterization inside SEMs, three main steps can be distinguished as illustrated in Figure 1.5:
10
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
• Preparation of nano-objects: This task includes identifying, isolating and holding the structure of interest. • Transport and positioning: Once the sample is held, it may be necessary to move it to a different location, where it can be processed and tested. The precise positioning, not only of samples but also of the tools, is essential to minimize faults in the subsequent steps. • Characterization/processing: Includes measurements of various properties of the sample, using a large variety of instruments, to obtain its mechanical, electrical and chemical properties. If the component is being processed, tasks such as adding or removing material, assembling components together, and others, can be performed.
Manipulators with various end-effectors
SEM electron beam
Test sample
Measuring / assembling stage
Sample magazine
Preparation
Transport
Characterization and processing
Figure 1.5: Representation of the characterization/assembly workflow of micro and nanostructures in SEM. The preparation process depends largely on visual surveying in order to identify the working environment, samples and tools. Once a sample is selected, the manipulator should approach with sufficiently accuracy to only capture the target, separating it from the substrate without damage. Various methods can be used to hold the sample, such as grasping (through a gripper) or gluing (i.e. with focus ion beam soldering) the structure on a probe. Once the sample is ready, the transport and positioning phase can start. The objective is to approach a target location without losing or damaging the sample. In this area, further tasks, such as inspection and assembly can be performed, also requiring precise positioning and orienting of the part. Strategies based on image processing and internal sensors are used to estimate the system position, orientation and forces during this task. Finally, to characterize the samples, various measurement methods can be performed through probes, embedded sensors and others devices. In some cases, such as during electrical characterization, the contact should be sustained between the sample and probe. This apparently simple task is not trivial at the nano-scale, where contacts may be difficult to detect, and excess forces may damage both the sample and the probe. If the part is being further processed, a good control over its position and the forces is important to ensure the good quality of the task, and to avoid damaging or losing components.
1.2. THE NANOROBUST PROJECT
1.2.2/
11
Scope of the NanoRobust project
The manipulation in situ seeks to realize a larger portion of the tasks in a same environment, in this case the SEM vacuum chamber. Operations such as sample preparation (selection of a single structure, detachment from the fabrication substrate, thinning of thick layers, ...), its transportation, positioning and finally processing and characterization would be ideally realized entirely inside the microscope with the aid of robotic platforms, manipulators and other computational tools. The in situ approach can reduce the risks of contaminating and damaging samples during the transport and preparation phases between different machines. Furthermore, the project aims to expand the information available to operators, improving throughput, robustness and repeatability of tasks requiring extremely high accuracy and precision, such as those often found when dealing with nano-objects. Commercially available platforms offer complete structures for manipulation and characterization, to be installed directly on SEMs. However, the success rate for these devices is often low, depending of the task complexity, and often present low flexibility, once tools for image, force or hybrid based control are not readily integrated. Robotic operation in SEMs, specially those dealing with samples with nanometric dimensions, often face the following obstacles: • The attainable resolution and repeatability of sensors and actuators may be not sufficient for the tasks. • Manipulation strategies containing integration of force/positioning control are rarely exploited. • The automation of tasks in SEM is still in its early state, with only a few works dealing with image based tracking under this kind of microscope. To achieve the project goals, the joint efforts of various expertise fields are required. Different strategies should be applied to manipulate, with sufficient dexterity, objects of different shapes and dimensions. Currently employed methods focus on studying particular cases. In addition, implemented systems present a reduced degree of automation and reliability, as most of the manipulation tasks inside a SEM are tele-operated. This project proposes the development of a more flexible approach through the fusion of different sensors to cover the multiple requirements imposed. The following four main branches of study are encompassed by this project, each dealing with a different aspect: • Development of strategies for micro- and nano-manipulation and characterization: Study of grasping strategies adapted to different objects in this scale and the interaction phenomena occurring (i.e. electrostatic, Van der Waals forces). Includes the development of collaborative manipulation with multiple end-effectors, and the use of different static and dynamic physical characterization techniques to study nano-objects. • Robust control for precise manipulation and characterization: Aims to study the working conditions found in SEMs (noise levels, disturbances, uncertainties) limiting the attainable performance of actuators and sensors. At the micrometric and nanometric scales, systems are sensible to environmental conditions (temperature, humidity, ...) and to disturbances, which imposes addition difficulties when trying to achieve nanometric precision and accuracy. Furthermore, the signal-to-noise ratio can be unfavorable. Typical actuators used
12
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
for operations in the micro-world (piezoelectric, electrostatic, ...) can present strong nonlinear effects to be considered when developing control laws for those systems. This branch also treats on the development of robust control laws for micro-robots and and-effectors inside the SEM environment, to improve their precision. • SEM imaging and vision: The image processing is essential for the manipulation inside SEMs. This branch deals with some aspects related to the image acquisition process on the electron microscope (i.e. drift, distortion, projection effects). Some imaging parameters may change as function of the amplification. These effects requires a multi-scale calibration and real-time image compensation (accordingly to the image acquisition frequency) to ensure the correctness of the information extracted from it. • Visual servoing on SEM: Intent to perform the manipulation of nano-objects, guiding the end-effectors through visual reference control while taking advantage of the tools developed in the other branches. It includes the selection of interest points to track in images. Objects of interest, such as samples and end-effectors, could be identified in the 3D environment based on a description of its geometry (CAD model of the components). Information from other sensors would help build a more precise representation of the events inside the vacuum chamber, improving the accuracy and reliability of such system. This project aims to have a direct impact on works dealing with nano-objects, as they could largely benefit from improved reliability and speed during manipulation and characterization. From an industrial point of view, the integration of developed tools to commercially available solutions is possible and desired, contributing to the automation of various tasks in this environment. The works developed in this thesis are part of the branch responsible for the robust control of manipulators and end-effectors in SEM, particularly their fine positioning to compensate for undesired behaviors. To better understand the uncertainties and disturbances in systems inside the SEM, it is important to know the components commonly used in robotic manipulations in this environment. Additionally, looking into research results in the domain help drawing a more accurate picture of the issues and how they deteriorate the quality of performed tasks. In the next section, the components of common nano-manipulation architectures are introduced, together with examples of their application found in the literature.
1.3/
Overview on nanorobotics system in SEM
When performing robotic tasks inside the SEM, various components are involved. To achieve the precisions aimed for in the NanoRobust project, all elements in the system possibly contributing to performance degradation should be analyzed. It is important to understand them and consider the advantages and limitations they bring into the system. In this section, the scanning electron microscope and the most commonly used components for manipulation and characterization of micro- and nano-structures are briefly summarized, together with some examples of its applications.
1.3.1/ The scanning electron microscope The electron microscopes are a popular tool among scientists due to its observation and characterization capabilities for samples ranging from nanometric to micrometric scales. The
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
13
family of electron microscopes include scanning, transmission and reflection electron microscope. SEMs are the most popular kind when manipulating micro and nano-structures. They can offer two-dimensional images with a large depth of field, over an amplification range between tens to hundred thousands times, and with relative short acquisition times. This section briefly introduces its working principle and main components, while a more detailed explanations can be found in [Newbury et al., 2003]. SEM column
Vacuum chamber
Sample exchange chamber
Power unity Workstation
Figure 1.6: Carl Zeiss Auriga 60 scanning electron microscope, used during this work, and its main components. The SEM is composed by a specimen (vacuum) chamber, a microscope column, a quick sample exchange chamber, a power unity, an operator workstation (shown in Figure 1.6) and vacuum pumps. The column contains a filament responsible for generating the electrons and a set of electromagnetic coils, responsible for accelerating and focusing them into a beam. The beam spot, usually less than 10 nanometers in diameter, is directly related to the image sharpness. The electron microscopes, differently from its optical equivalents, captures the image point-by-point, by dividing the sample intro micro-volumes and irradiating them sequentially with the focused electron beam. Electromagnetic scan coils are responsible for bending the beam, sweeping lines across the sample’s surface and defining the scan path. During this operation, the microscope column and the specimen chamber sustain a high-vacuum, normally in the range of 10−3 -10−4 Pa. This is necessary to minimize beam deflections due to collisions with gas molecules in the air. 1.3.1.1/ Interaction between electron beam and matter The interaction between electron beam and matter produces back-scattered (primary) electrons, secondary electrons, x-rays and other particles that are captured by sensors, supplying information about the sample, such as topography and composition. In an ideal scenario, a focused electron beam with infinitesimal spot area would interact with the first layer of atoms in the sample’s surface, resulting in the best possible spatial resolution. In reality, the beam spot has a diameter of a few nanometers and penetrates multiple atom layers
14
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
into the sample, resulting in a complex interaction volume. This volume, varying in depth from hundreds of nanometers to a few micrometers, depends on several factors (i.e. sample composition, incident electron beam energy, ...). Figure 1.7 shows, in a simplified diagram, two of the possible interactions between the incident electron beam and the sample: backscattering and the production of secondary electrons.
Backscattered electron Primary electron
-
Electron beam
Secondary electron -
-
-
-
+
+
-
+ +
-
-
+
+ +
-
+ +
+
+
+
-
-
-
-
-
Interaction volume
Figure 1.7: Interaction between the focused electron beam with the surface of a sample produces backscattered electrons and secondary electrons. The scan speed defines the irradiation time for each one of the micro-volumes. Increasing its speed results in faster image acquisition times, ranging from less than 0.1 seconds up to tens of seconds, while producing more noisy images. When entering the sample surface, the negatively charged electrons in the beam interact with protons and electrons in the atom. Backscattered electrons are generated by elastic scattering produced by inter and intra-atomic forces. This process have, in theory, no kinetic energy loss and can, after a series of interactions, alter the electron path sufficiently to cause its ejection from the sample. The probability of backscattering increases with the atomic number Z, as heavier atoms have stronger positive charges, and decreases as the incident electron beam energy augments. Simultaneously, inelastic collisions between the electron beam (primary electrons) and atoms occur in the volume, knocking electrons out of the atoms (secondary electrons), with loss of kinetic energy. These collisions propagate in the interaction volume, in a cascade effect, before losing all its energy. However, some electrons can be emitted from the surface before this happens. The sensors capture these ejected electrons and convert the recovered signal in a numeric value, proportional to the number of recover particles. This process is repeated sequentially for all the surface, and the recovered information is transformed into a gray-scale image.
1.3.1.2/ SEM imaging characteristics Through the operation station, it is possible to supervise and modify scanning parameters, i.e. beam energy and size, scan speed, sensors sensibility, and others. These parameters have a direct impact on the image characteristics (resolution, sharpness, contrast between features, ...)
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
15
and are selected accordingly to the sample composition and dimensions.
(a) Optical microscope
(b) SEM
Figure 1.8: Image of the mineral skeleton of a Radiolarian protozoa, acquired with optical and scanning electron microscopes [Newbury et al., 2003]. One of the most used SEM features is the acquisition of high-resolution image. Nowadays, SEMs with resolutions close to 1 nanometer are commercially available. Furthermore, its large depth of field helps to provide more information about the sample. This property is useful for micro-manipulation tasks. Figure 1.8 compares two images, obtained with optical and scanning electron microscopes. The SEM image has a higher resolution and a much larger depth of field, allowing to observe with improved sharpness a greater portion of the structure. This property, combined with the shadows and highlights produced by the electron beam/sample interaction, results in a notion of depth attractive for various applications. In addition, relative quick image acquisition can be obtained in SEMs, with sampling frequencies achieving tens of Hertz in modern devices. This allows operators to act on elements inside the vacuum chamber, (handling samples, assembling and characterizing components, ...) and obtain a visual feedback that is usually sufficiently quick for most of the desired tasks dealing with static and quasi-static phenomena. It is important to stress the compromise between speed and image quality. Figure 1.9 helps exemplifying this aspect, where the extremity of a micro-gripper was captured under different conditions. Images A and B contrast the influence of the image acquisition speed considering a static gripper. When a slow scan is used (A), a less noisy image is obtained. An approach for the SEM compromise between image quality and acquisition speed was proposed in [Sievers, 2011]. It explored the idea of increasing the scan speed by reducing the electron beam irradiation time at each point, resulting in a more noisy image. Using algorithms robust to noise and the dynamic selection of the scan area, automatic tracking of objects could be achieved. Still, this method of image processing was limited by the SEM image acquisition capabilities, in this case close to 13 Hz. Images C and D in Figure 1.9 compare, under fast acquisition speeds, the effects of movement. Image D indicates the fast dynamics of the studied object produces a blurred, incomplete photo of the gripper. The same effect can be observed at large magnifications, where small vibrations limit the accuracy on measurements performed directly by image analysis. The vibration of structures can be an issue when performing tasks requiring positioning precisions of a few nanometers. Methods based solely on image feedback encounter difficulties to achieve this level of accuracy, specially when fast dynamics are involved.
16
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
A)
B)
C)
D)
Figure 1.9: SEM images of a micro-gripper extremity.
1.3.2/
Nanomanipulation systems
At the micro- and nanometric scales, robots are used to produce the desired precise displacements. Micro-manipulators are systems that can generate movements with at least micrometric precision. To improve their flexibility, manipulation and testing capabilities, stations can be composed of one or more robots, each with multiple degrees of freedom (DoF). A large working range (up to a few millimeters) is also desired for transporting materials. Nowadays, various structures are commercially available, although customized solutions are not rare to find in laboratories. Figure 1.10 illustrate some examples of nanomanipulation systems. The Zyvex nProber system [zyv, ] is composed of 8 independent positioning arms and a positioning stage, each one capable of linear movements in the 3 directions. Each positioner is composed of a coarse and a fine positioning unit. By using separate actuating elements, this device can achieve a large fine motion range at the expense of a greater basis footprint [Zhang et al., 2013b]. In this device, the advertised movement resolutions for the positioning arms are under 1 µm in closed-loop, coarse mode and 2 nm when operating in open-loop, fine mode. To reduce the robot’s dimensions, it is common to only employ one piezomotor for both coarse and fine motions. The Kleindiek MM3A [kle, ] micro-manipulator is a 3 DoF arm (two rotations and one translation) actuated by piezomotors in open-loop configuration, with the nominal resolution close to 1 nanometer. To perform more complex tasks, multiple arms are often used. To overcome the lack of internal sensor in devices operating in open loop, methods such as modeling, characterization and visual servoing are often employed. This feedback suffers from shortcomings due to SEM imaging acquisition, discussed earlier, and the challenging of modeling and characterizing the device. The company SmarAct develops various different robotic structures, ranging from systems with 1 DoF up to complex 16 DoF workstations. As example, a 15 DoF was taken, composed of four 3D manipulators and a 3D stage. This structure, advertised as possessing 1 nm movement resolution when operating in closed loop with an integraded position sensor [sma, ]. The enterprise Physik Instrumente also offers various solutions for nano-positioning problems. One example is
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
(a) Zyvex nProber
(b) Kleindiek MM3A
(c) SmarAct 15D
(d) PI P-911KNMV
17
Figure 1.10: Examples of commercially available nano-manipulators.
the P-911KNMV [pi, ] station, a 6 DoF miniature parallel robot with large load capacity (1.5 kg) and a displacement resolution close to 100 nanometers. When internal sensors are available, other issues may arise. Difficulties in integrating sensing capabilities into small actuators may prohibitively increase its dimensions, the often reduced signal/noise ratios present, drift and thermal influence, are some of the issues limiting its use and performance. The calibration of such sensors in a SEM environment can be challenging, usually resorting to imaging-based methods. Additionally, some internal sensors may not be sufficiently fast. Even though these sensors may achieve higher sampling frequencies than those obtained by SEM imaging, the dynamics of components in these structures not rarely surpass its measurement limits. Figure 1.11 exemplifies the difficulty in obtaining precise positioning data, using a commercial micro-stage with three linear axis actuated by piezo-motors. The position data was acquired by the device’s internal sensor (optical encoder). The stage is controlled in closed loop by its own software. Via a computer interface, users can send reference positions to be tracked and recover its measurements, via specific commands supplied by the manufacturer. To verify the correctness of this results, an external sensor (laser vibrometer) was used, capable of faster acquisition (tens of kHz against hundreds of Hz in the internal sensor) with resolution close to 1 nanometer. The graph highlights that, even at a relative fast acquisition speed, this internal sensor does not fully capture the displacement dynamics. Various works also explore the design of custom made solutions [Yamamoto et al., 2003, Fukuda et al., 2003, Nakabayashi et al., 2007, Nakazato et al., 2009]. In addition, some researchers have moved towards the development of mobile manipulation systems
18
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
1.2 1
Internal sensor Laser vibrometer
Displacement (µm)
0.8 0.6 0.4 0.2 0 −0.2 −0.4 0
0.01
0.02
0.03
0.04
0.05
Time (s)
Figure 1.11: Comparison between position measurements given by the internal sensor and by an external laser vibrometer for a step in the desired reference position.
[Fatikow et al., 2000, Schmoeckel et al., 2001, Driesen et al., 2005]. These structures can present a large working range, interesting for real applications and an advantage over fixed designs. In addition, multiple robots can be used at the same time, each one with a specific task (manipulation, fixation, transport, ...). These systems could bring a large flexibility into nano-manipulation and fabrication. In [Fatikow et al., 2000], a flexible micro-assembly station is described to integrate different nano-components into hybrid micro-systems. The workstation was implemented using two robots from the MINIMAN (MINIaturized MANipulation robot) family (Figure 1.12). One of the employed robots (MINIMAN-II) was mobile, and could move freely over the platform by using the stick-slip principle to control its piezoelectric leg [Fahlbusch et al., 1999].
(a) MINIMAN-I
(b) MINIMAN-II
Figure 1.12: Piezoelectric nano-manipulators from the MINIMAN family. Multiple robots were used in [Fahlbusch et al., 1999] to increase the flexibility for this manipulation station. In addition, the MINIMAN-II robot is mobile and could move freely over the platform.
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
1.3.3/
19
End-effectors
Various different end-effectors can be coupled to the nano-robotic structures, accordingly to the sample characteristics and desired applications. When operating in SEM environment, additional restrictions are imposed, as these tools should be compatible with the requirements for operation in the vacuum. The most common end-effectors can be classified into two categories: grippers and needles. 1.3.3.1/ Needle tips Needle tips consist of a single, thin element with a sharp point, that aims to reduce its contact area with the sample. Some of the most known example of this group are the AFM tips, although other specialized structures are not uncommon. When using a single tip, 2D manipulation can be performed by pushing components across the surface. Various works report the use of a AFM tool to roll and slide CNT across surfaces. As results, translational displacement, bending, kinking and breaking were achieved [Hertel et al., 1998, Postma et al., 2000]. Figure 1.13 demonstrates a sequence of 2D manipulation tasks performed over a CNT in a surface using an AFM tip. The red marker indicates a reference point in the substrate. By deforming the tubes, it was also possible to investigate their behavior under large strain and bending [Wong et al., 1997, Falvo et al., 1997].
(a)
(b)
(c)
(d)
Figure 1.13: Manipulation of a CNT over a surface using a needle-type end-effector [Hertel et al., 1998]. These tools can be also used to cut and indent samples [Meyer et al., 2008]. Figure 1.14 demonstrates a cutting process over a thin (50 nm) polystyrene film using a thin needle probe inside a low-voltage SEM. The article remarks that, despite the brittle nature of this material, the sample does not crack or present further deformation due to the different behaviors of ultra thin films. When more complex trajectories or 3D manipulation are required, adhesion forces can be used. If the interaction forces produced by the needle tip are higher than the ones generated by the substrate, samples can be lifted. To deposit the structure to a surface, the inverse is true. This behavior brings more complexity to manipulation task with a needle. Some solutions to this problem include the use of additional tool tips [Sitti et al., 2001, Xie et al., 2009], and the control of the tip orientation, reducing the contact area during deposition tasks [Koyano et al., 1996]. In some cases, however, adhesion forces are not sufficient to ensure that the sample will be correctly held. To solve this issue, fixation strategies (through gluing process, electron beam induced deposition, ...) [Yu et al., 1999, Fukuda et al., 2002] can be used. In [Rong et al., 2006], two AFM tips were used to test the tensile strength and Young modulus of nanoparticle chain
20
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
Figure 1.14: Cutting of a ultra thin (50 nm) polystyrene film using a sharp needle. [Meyer et al., 2008] aggregates (NCA). These chains are formed of branches structures, composed of particles between 1 and 50 nm, and have applications in structure reinforcement, flexible electronics and gas sensors. To test these particles, a soft and a hard AFM tips were mounted on independent manipulators. The NCA were firmly attached to each tip through electron beam induced deposition, and the AFM tips were separated. Through inspection of the SEM images, it was possible to estimate the mechanical characteristics of the sample by measuring deflections in the soft tip. Figure 1.15 shows some steps of this process. In the left image, an individual NCA is captured by the soft cantilever. In the right image, the end-effectors are distanced, elongating the trapped structure.
(a)
(b)
Figure 1.15: Two AFM tips with different stiffness were used to hold and perform traction test on a nanoparticle chain aggregate [Rong et al., 2006].
1.3.3.2/ Micro-grippers Grippers are structures composed of two or more fingers, that can be closed around a sample to grasp it. This can be used to avoid the fixation and separation processes often employed when using needles for 3D manipulation. They are the most used tools for handling micro-objects [Cl´evy et al., 2006], and can be passive or active, with one or more degrees of freedom, and be equipped with different sensors. The use of passive, compliant grippers, was described in [Dechev et al., 2004, Tsai et al., 2005]. They were designed so that it can elastically deform when facing external forces.
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
(a) Passive gripper and component
21
(b) Components fixed in the base structure
Figure 1.16: A passive gripper used to assemble 3D structures inside the SEM [Dechev et al., 2004]. By adapting the structure of both gripper and sample, a passive grasping and releasing mechanism was produced. This kind of approach has a low degree of flexibility, as the end-effector design was strongly dependent of the manipulated object. In Figure 1.16, an example of this passive grasping mechanism is shown. To grab a part, the manipulator should approach the structure and exert enough force so that the mechanism firmly locks it. The sample could then be manipulated in 3D, and positioned over the substrate, where it was fixed with the aid of another especially conceived mechanism. The fixation force between samples and substrate should be large enough to overcome the gripper’s adhesion forces, so that the passive mechanism can open and releases the structure easily. Although possible to be used for manipulation and assembly tasks, these grippers may require large modifications in the end-effector, in the sample and in the deposit surface. Due to this added complexity, active grippers are more commonly used. In these structures, finger movements can be controlled. Their actuation system can be based on various methods, where the most commonly found are: • Electro-thermal: Takes advantage of the local thermal expansion due to Joule effects. They can be constructed by two principles: the difference in thermal properties between two materials (i.e. bimorph effect on beams) [Chu, 1994, Park et al., 2010] or the difference between dilatation on areas of the same material [Nordstr¨om Andersen et al., 2009, Sardan et al., 2008]. These methods allow producing large forces with relatively compact structures. Some designs are capable of active opening and closing movements, what can be convenient for real applications . They often require high temperatures (up to 1000 K), what can be dangerous to the gripper and to the manipulated sample. Also, when operating in different environments (air and vacuum), the actuator presents different behavior due to the thermal dissipation difference. In [Park et al., 2010], a displacement of 250 nm was produced for a given actuator under a constant current in air, while the same input induced a displacement of 900 nm in the vacuum. Figure 1.17 shows the end-effector developed in
22
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
[Nordstr¨om Andersen et al., 2009], where controlling the current flow in each of the three parallel connectors forming the fingers allows to open and to close the grasp.
Figure 1.17: Thermo-electric micro-gripper. By heating different parts of each finger, open and close movements could be produced [Nordstr¨om Andersen et al., 2009]. • Piezoelectric: these materials are capable of transforming mechanical energy into electrical energy. The reverse process is also possible, where an electric potential can generate deformation in certain directions. Piezoelectric grippers can achieve high resolution and fast response times. They are generally built by using piezoelectric actuation to deform a compliant structure [Salim et al., 1997, Goldfarb et al., 1999, Agnus et al., 2005], or by using a bimorph configuration [Haddab et al., 2000, P´erez et al., 2005, Tamadazte et al., 2012, Ivan et al., 2013]. When using piezoelectric actuators, the achievable stroke is usually reduced, requiring multiple piezoelectric elements to be stacked or an amplification system to be included, augmenting the size and complexity of the device. Furthermore, devices using this actuation method can also present hysteresis and creep effects. In [Ivan et al., 2013], a dual bi-morph gripper was developed using an alternative piezoelectric material (lead magnesium niobate - lead titanate, PMN-PT), allowing to achieve 6D of movement (Figure 1.18). This end-effector could produce larger displacements than equivalent constructions, with lower hysteresis effects thanks to the use of PMN-PT. This gripper was only tested in air, although its structure is also compatible for operation in high-vacuum environments. • Electrostatic: use the Coulomb attraction forces produced between differently charged plates, where one element is fixed and the other is free to move. By using this principle, different structures can be conceived. Planar parallel surfaces where used in [Chu et al., 1994]. Due to the nonlinear characteristic for this kind of actuators, multiple equilibrium points can appear, what is a challenge from the control point of view. After a certain displacement, the attraction force increases quickly, producing high accelerations and possible collisions between the two charged plates. This effect is called pull in instability. The use of interdigited comb-drives [Beyeler et al., 2007] can reduce this problem, by keeping the distance between plates constant. The direction of movement in this case is parallel to the plate surface. Figure 1.19 shows a monolithically fabricated micro-gripper, actuated by a comb-drive system. This device could achieve strokes up to 100 µm, and possessed an integrated force sensor, to manipulate and characterize delicate sample.
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
PMN-PT [011] beams
23
End-effector manipulation tips
30 µm
10 mm 1 mm SOI end-effectors
Figure 1.18: Piezoelectric actuated gripper. The arms are composed of dual bi-morph structures, giving the end effector 6 DoF [Ciubotariu, 2015].
Figure 1.19: Electrostatic comb-drive actuated gripper, located at the extremity of the circuit board (lower left corner of the image) [Beyeler et al., 2007].
Grippers have relatively large contact surfaces interacting with the samples. In addition, the capacity of exert prehensile forces on the object ensures they can be safely held. One of the challenges for this class of end-effector lies on the releasing phase. When the gripper opens, the adhesion forces between fingers and the manipulated object are generally large, resulting in failure to release the part. Various methods aim to facilitate this step, by reducing the grippers contact area or by introducing active releasing methods, such as the repulsive charges or mechanical aid [Arai et al., 1996, R´egnier, 2006, Gauthier et al., 2011, Chen et al., 2009]. It is important to consider that, when operating in a vacuum environment, some of the characteristics for those elements can be modified. It is clear, in the case of thermal based actuators, reduced heat dissipation to the environment should be examined to avoid damaging gripper and sample. However, the effects of pressure variation between air and vacuum may also affect these systems and its dynamics, similarly as occurs in cantilever structures
24
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
[Sandberg et al., 2005, Mertens et al., 2003], where the environment composition and pressure alter their vibration characteristics. These effects over system dynamics are often overlooked, and can contribute to their performance degradation.
1.3.4/
Sensors
The characterization of the different material and structure properties can be performed through various sensors inside the SEM. They can offer information over position and surface, forces, temperature, currents, ..., and may help during object handling (detecting contacts, deformations and other phenomena). The SEM itself is, primarily, a powerful sensing device, offering topographical and morphological information about the sample surface, detecting features up to a few nanometers. Through image, it is also possible to estimate the mechanical characteristics (stiffness, Young modulus, ...) of materials, by measuring its deformations under the effects of a known force. The image capabilities of the SEM are also widely explored to offer position information during robotic operation. As seen earlier in this chapter, the compromise between image acquisition speed and resolution is an important tradeoff to be considered. Despite this limitation, the SEM is an useful and flexible tool. It allows studying crystalline structures and its orientation on the surface of specimens (through the use of backscattered electron diffraction [Matteson et al., 2002]) and to acquire compositional information through the use of characteristic x-rays. Modern x-ray detectors are capable of detecting the x-ray emissions of all elements above atomic number 4 at typical electron beam currents used in SEMs [Newbury et al., 2003]. Force and torque sensors are also usual tools to be found during manipulation and characterization tasks. Some end-effectors possess integrated force sensing capabilities [Beyeler et al., 2007, Komati et al., 2014, Mokrane Boudaoud, 2014], (i.e. strain gauges, capacitive and piezo-resistive sensors, optical methods), but dedicated force sensors can be applied. In [Beyeler et al., 2009], a complete six-axis force-torque sensor was demonstrated (Figure 1.20). AFM tips are also com6 DoF force sensor monly used for force estimation, where the Figure 1.20: well known stiffness for this structure and the [Beyeler et al., 2009]. measurement of its deformation allows to estimate some sample’s characteristics. In some cases, the AFM can be used as both sensor and end-effector. The use of tuning forks has been reported in various SEM applications. When operating in vacuum, its quality factor increases drastically, resulting in more precise measurements. In [Friedt et al., 2007], a quartz tuning fork was used as a scanning probe, to obtain a 3D representation of a surface. As this device presents a stable resonance frequency, its variations can be precisely measured. The same principle was used to estimate forces and stiffness [Acosta et al., 2011, Abrahamians et al., 2014, Polyakov et al., 2014] of fragile and flexible materials. Figure 1.21 shows an example of a quartz tuning fork (left) and a possible application for stiffness mapping in a thin membrane surface (right), a task requiring high position accuracy over the 3 directions.
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
25
(a) Quartz tuning fork, uses for dynamic force (b) Stiffness chartography of a thin (200 nm) InP measurement and 3D topography mapping membrane [Abrahamians et al., 2014] [Friedt et al., 2007].
Figure 1.21: Tunning fork and application in characterization inside a SEM. The characterization of electric properties for nano-materials has been explored by various authors. This process can be performed with aid of testing probes [Lim et al., 2005, de Jonge et al., 2003, Hwang et al., 2009, Wei et al., 2010], or over specially constructed surfaces [Suga et al., 2009, Franklin et al., 2012]. The correct positioning of probes and samples, and the need to ensure sustained contact between, are essential. In the case of fragile samples, such as CNTs, the contact forces exert by the probe, pushing the sample against a surface, may cause deformations and alter the results [Yu et al., 1999]. Performing measurements in free space or in measurement stages can reduce this effect. In [Suga et al., 2009], nano-spheres with diameters between 50 and 200 nanometers were deposited over a specially designed junction between gold electrodes (Figure 1.22), allowing to measure the electric conductivity for multiple samples. However, these methods still rely on the careful manipulation of samples.
Figure 1.22: Electric characterization surface for nano-objects. It is composed of two gold covered pads, separated by 40 nm. Objects were carefully placed over the gap, and tested for its electric properties. [Suga et al., 2009]. The precise positioning of sensors is essential to ensure the quality of measurement, and
26
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
can be challenging when contact with fragile structures, such as nano-objects, is required. Controlling the position and forces exerted by these devices, statically and dynamically, is an important step towards improving the success rates.
1.3.5/
Examples of robotized tasks carried in SEMs
To successfully perform the different task required for a complete operation in SEM (preparation, transport and assembly/characterization), the elements previously described (manipulators, end-effectors, sensors, and SEM) are used and combined in different forms. The following examples of real applications help to illustrate the evolution of SEM manipulation and the current challenges faced in this field.
1.3.5.1/ Tele-operated manipulation and characterization in SEMs In the 1990’s, the first reported research for SEM-based nanomanipulation was presented in Japan [Hatamura et al., 1990]. The idea was to obtain a direct relation between manufacturing methods in the macro and micro-world, where objects with dimensions as small as 10 µm could be handled. It described the use of the stereo scanning electron microscope to provide a magnified 3-dimensional image of the working environment, while two nanorobotic systems, controlled by joysticks and emulating the opera- Figure 1.23: Proposed schematics for the tor’s left and right hand. An schematic of the first nano-manipulation station inside a SEM proposed experimental setup is presented in [Hatamura et al., 1990]. Figure 1.23. One hand was used to control the tools, while the other held the sample. A scaled feedback on the forces involved during the manipulation of micro-components was available. The article recognized that some nano-manipulation problems were too difficult to be tackled at the time. It highlighted issues related to the need of extended manipulation capabilities and additional challenges posed by operating in an isolated environment, such as tool changing. A different workstation was proposed in [Koyano et al., 1996], consisting of two manipulators (with 5 DoF each) and a workstation (3 DoF). The setup included an optical microscope, placed laterally to offer an improved vision over the interest zone. Improvements in the operation were achieved by using the additional angle of view, and the inclusion of rotational degrees of freedom in the manipulators and in the worktable. The article also proposed the use of different handling strategies to counter adhesive forces that become important at this scale. By cooperation of two tools with different tip diameters, this system could perform pick-and-place tasks of spheres between 20 and 30 µm. Figure 1.24 shows a schematic for the proposed manipulation system (left) and the actual implementation inside the SEM chamber (right), illustrating how most of the available space was cluttered by the workstation’s structure. In the following years, the ideas proposed by these pioneer articles were further developed. Quickly, the study of carbon nanotubes was extensively explored with the aid of SEMs and robotic interfaces. In [Yu et al., 1999], a dedicated platform was developed to investigate the
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
(a) Schematic of nanorobotic station
27
(b) Practical implementation
Figure 1.24: SEM nanorobotic station. (a) Scheme for the station, which included 1 moving stage, 2 manipulators and an optical microscope to give lateral vision over the system. (b) Actual implementation, where all the components should be arranged to fit the vacuum chamber tight space [Koyano et al., 1996]. properties of single CNT structures after liberating them from the substrate. It supposed the large differences in CNT characteristics found in literature at the time were due to the sample preparation process employed, that could deform the samples. It proposed to free the samples from its growth substrate and to characterize them in a free space, without the need of posing the CNTs against a rigid measurement surface. Figure 1.25 shows the developed robotic manipulator, with 3 linear and 1 rotational axis, and a SEM image of a carbon nanotube suspended between two AFM tips. This setup was used for simultaneous measurement of tensile strength and conductivity in the samples.
(a) Robotic manipulator
(b) CNT attached to two AFM tips.
Figure 1.25: Robotic manipulator developed in [Yu et al., 1999], to test CNTs in a free space, without the substrate interference. In [Akita et al., 2002], two carbon nanotubes were attached to an AFM tip and used as tweezers, to manipulate and characterize other nano-objects. By inducing a small DC voltage on this system, the separation between CNTs could be controlled. The tweezers assembly process was performed inside a SEM by a specially designed working platform, named nanofactory. It possessed tree independent stages, each one with four degrees of freedom. Figure 1.26 shows the fabricated tweezers, and how a small voltage induced the nanotubes approxima-
28
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
tion. During manipulation, the contact forces between the substrate and target strongly influenced its performance. Other articles dealing with the attachment of individual carbon nanotubes to probe tips indicate that this procedure, when teleoperated, can take up to 30 minutes [de Jonge et al., 2003, Fahlbusch et al., 2005].
(a) Open fingers
(b) Closed fingers
Figure 1.26: CNT nano-tweezer fabricated on an AFM tip. Fingers could be controlled by applying a small voltage [Akita et al., 2002].
Figure 1.27: SEM robotic station composed of 4 independent manipulators, for a total of 16 DoF. [Fukuda et al., 2003]. In [Fukuda et al., 2003], a robotic station composed of 4 independent manipulators totalizing 16 DoF was proposed. This robotic platform, shown in Figure 1.27, aimed to perform manipulation and in situ characterization of nano-structures inside the SEM. These nano-structures were intended to be used as the building blocks for the fabrication of more complex devices. Tests shown adequate manipulation on carbon nanotubes for tasks such as picking and connecting individual structures, as also destructive characterization tests. The different tasks necessary to execute one complete operation were carried out individually and manually. The article concluded that, although numerous operations could be performed by an operator in this nano-laboratory, devel-
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
29
opments in task automation and better knowledge over the nano-world phenomena are necessary to further develop the manipulation in SEM. CNT-based transistors (Figure 1.28) were produced in [Kim et al., 2003], using two commercial piezo-actuated micro-manipulators and carbon nanotubes based probes, by carefully placing the CNTs over pre-fabricated connections pads. The authors remarked that, even as the actuator used in the micro-manipulators possessed nominal resolutions close to 1 nanometer, a minimal attained displacement of 10 nanometers when moving objects was obtained during practical application inside the SEM. This degradation was attributed to the presence of surrounding vibrations. Although no further indication was offered in the article about the relation of external disturbance to the loss in performance, it serves as an important indication that these phenomena can have a significant impact over nano-manipulation systems in SEMs.
100 nm
(a) Robotic arm and positioning stage
(b) Assembled CNT transistor
Figure 1.28: Basic CNT transistor, build manually by placing the nanowire over pre-patterned electrodes [Kim et al., 2003]. Another example on the importance of precise assembling is found in [Aoki et al., 2008]. This paper dealt with the construction of 3D photonic crystals from 2D structures. Photonic crystals are composed of patterned structures capable of controlling and manipulating light flow, with large interest in fundamental and applied research. By changing the patterns on its otherwise periodic lattice, new functionalities can be added, allowing light to flow or to be blocked in particular paths, or even creating micro-resonators [Busch et al., 2004]. These 3D structures were built by stacking thin (12.5x12.5x0.2 µm) 2D Figure 1.29: 3D photonic crystal, built by stackcrystals with the aid of two triaxial positioning ing multiple planar structures. Scale bar of 5 µm stages installed inside a SEM. The stacks were [Aoki et al., 2008]. positioned with under 50 nm of precision, what was considered adequate for the task, yet far away from the sub-nanometric precision of the piezoelectric nanostages used for positioning. Figure 1.29 shows an example of one assembled 17 layers structure. The three square pillars are used to facilitate the alignment during stacking phase. The assembly strategy required careful tele-operation to capture, transport and depose correctly each
30
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
layer. These studies were based on tele-operated, static and quasi-static manipulation and characterization of micro- and nano-structures. Despite the various positive results obtained, researchers contemplated the automation of tasks to improve productivity during nanomanipulations, reducing times and the variability induced by human-operated methods. Furthermore, they highlighed the necessity of better understating the phenomena occurring inside the vacuum chamber, including the effects of disturbances, dynamics and interactions between components.
1.3.5.2/ Development of autonomous tasks in SEM Since the start of manipulation in SEM, researchers pointed the importance of autonomous tasks in this environment to achieve higher performances and speeds. A work dealing with image processing for automated precise positioning is presented in [Gong et al., 2013]. The authors were interested in obtaining the highest accuracy for automatic position of probes, intended to perform electric measurements on electronic devices. The probe tip radius were smaller than 100 nm, and the targets were concentric lines with 130 nm width separated by 125 nm spacings. It employed a high-accuracy robotic platform system with two independent probing systems, each with nominal fine positioning resolution close to 1 nm. After images were acquired and filtered to reduce noise, a visual tracking algorithm controlled the probes in the X-Y plane, to be positioned over selected points. Next, the Z axis was slowly moved, also automatically, until it touched the sample on desired spots. As no forces were directly measured, image processing was used to detect contact between probes and substrate automatically. Figure 1.30(a) shows the robotic station, with two independent 3 DoF Cartesian platforms used for the experiments inside a SEM. Figure 1.30(b) illustrates the automatic nanopositioning task, as seen by the SEM. It shows the test targets (I), followed by the selection of the probing points (II). Next, the system automatically approached the selected spots, in order (III and IV). The article indicated a 100% success rate at low-magnifications (1000x), and values between 80% for X-Y and 60% for Z positioning at large magnifications (80000x), where image drift and noise, shallower dept of field and external disturbances were more evident. Nevertheless, this setup was able to perform the task at least three times faster than a skilled operator. In [Zhang et al., 2012], a method for fabricating a nanowire field-effect transistors (NanoFET) inside SEMs was described. The Nano-FET is a highly sensitive device capable of molecular label-free sensing. The response sensitivity for this kind of sensor is affected by construction parameters, such as the number and diameter of employed nanowires bridging the electrodes. To improve the performance, multiple CNT were deposed over the electrodes, connecting them. By inspection, nanotubes with similar diameters bridging the electrodes were selected, and all the other connections were removed (Figure 1.31). With this method, the electrical connection properties across the device presented similar values, which increased performance. By introducing an automatic detection and selection of the appropriate wires, and later automatic removal of unwanted connections by a needle probe, success rates up to 95% were reported, a large gain when compared with a 40-60% rate during tele-operated manipulation. Furthermore, the production time was reduced, from 10 min/device for manual nanomanipulaton to 1 min/device by the proposed automated method. One of the critical parameters in the automation algorithm was the minimal distance between the target CNT (to be kept) and the other nanotubes. If this distance was too small, the chance of accidentally damaging the target increased, up to approximately 12% fail rate when a minimal distance of 500 nm was selected. By increasing this minimal distance, the change of failure was reduced. However, this imposed an additional constraint and reduced the number
1.3. OVERVIEW ON NANOROBOTICS SYSTEM IN SEM
(a) Nano-probing station
31
(b) Sequence of steps for sample probing. The positioning task was performed automatically.
Figure 1.30: Robotic probing station. Combining vision servoing, the testing probes are automatically positioned over the desired spots. Contact between probes and substrate were detected by image inspection [Gong et al., 2013].
of candidate nanowires meeting the requirements. As consequence, a balanced minimal distance (1 µm) value was experimentally found, to ensure the automated system would be able to find a successful combination of CNTs connecting the electrodes and the manipulation process had an acceptable risk of failure, estimated to be around 4% in this case.
(a) Before
(b) After
Figure 1.31: Automatic selection and isolation of single carbon nanotubes, signaled by the red arrow in the image. Needle tips were used to sewer undesired connections between two electrodes [Zhang et al., 2012]. These examples demonstrate how certain degrees of automation in positioning tasks can be achieved, but limitations on performance do exist when high-precision, fine motions are required. In the examples, positioning tasks find its limitations, indicated by the increase in failure rates, for values that are many times larger then the nominal actuator resolutions. Again, these issues might be associated with the difficulty in obtaining fast, reliable data over the end-effector position, where disturbances and dynamic effects are often neglected. Other approaches were developed, not only using images provided by the SEM but also integrating different sensory information and models, to obtain a more complete panorama over the system’s behavior.
32
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
1.3.5.3/ Improving performance and precision in SEM through the use of models The use of 3D representations and virtual reality tools were most commonly found in atomic force microscopy operations. This kind of microscope returns 3D information about the surface, but normally its acquisition times are slow (from a few seconds up to minutes). Using the provided information to build a virtual model allows to plan tasks more efficiently, an important feature in this kind of device [Sharma et al., 2005, Korayem et al., 2012]. However, in SEMs, these methods have found so far limited applications. In [Abrahamians et al., 2014], virtual environments were explored to create a replica Figure 1.32: Virtual manipulator model and inof the manipulation system (sample, position- teraction between virtual and real components. ing stage and micro-actuators) installed inside [Abrahamians et al., 2014]. the SEM. The virtual model used information from the manipulator internal sensors to estimate its configuration. This allowed to improve the planning in 3D characterization tasks, effectively increasing the throughput and reducing the occurrence of failures. Figure 1.32 shows the virtual robot an how the different components interacted during manipulation. To enhance the positioning accuracy on robotics devices, other alternatives to image position are available. The work of [Saeidpourazar et al., 2008] described a method to fuse image and force sensors. It was developed based on a micro-robot without any position sensor. In this kind of robot, feedback information should be entirely provided by images (operator or image processing algorithm), what limited its accuracy and speed, depending on image resolution and acquisition times. To ameliorate these characteristics independently from image-based methods, a force sensor was added to the end-effector. By measuring the interaction forces between effector and sample, distances separating both elements could be estimated based on an interaction force model. The proposed system presented, in simulations, faster responses (less than 1 second) with no oscillations, and allowed a certain degree of control over the end-effector dynamics. A custom micro-robotic structures was proposed in [Gorman et al., 2006, Gorman et al., 2007]. It consisted of a nanoassembly system to be contained in a chip, composed of four MEMS probe manipulators placed around a central port where samples were placed. The manipulators, with 3 linear axis and actuated by combination of thermal and electromagnetic elements, could be used individually or collaboratively to manipulate and characterize samples. Figure 1.33 shows the design of the structure and a SEM image for one of the manipulators mechanism. The positioning system was characterized, both dynamically and statically, with aid of external sensors in air only, and presented a nonlinear response and under-damped first mode characteristics. In open loop, the position resolution was found to be under 12 nm, with step repeatability around 7 nm for small displacements, with large degradations in those values when large steps were required, what limited its practical application. To improve the results, multiple control strategy were implemented, including the use of inverse models to compensate the non-linearity and a PI controller. The controllers were tested, in simulation only, considering uncertainties of up to 5% in the parametric estimation. Even though good tracking results could be achieved, it was clear that the performances depended on the correctness of the
1.4. THESIS CONTEXT
(a) CAD desing for the nanoassembly station
33
(b) SEM image of one of the 3 DoF nanopositioners.
Figure 1.33: Desing of a microscale nanoassembly system to be fully contained on a chip, for operation in small environments, such as the SEM [Gorman et al., 2007]. models. These examples bring up another issue: how to obtain adequate models considering environment effects over the system? In the considered examples, geometric parameters for the manipulators could be obtained with relative ease, yet the effects of low pressure operation in various components (manipulators, end-effectors, sensors and actuators) were often disregarded.
1.4/
Thesis context
In the past couple of decades, numerous examples of manipulation, characterization and assembly of micro- and nano-structures in SEMs were reported, covering a large number of fields: physics, chemistry, biology, material science, among others. The evolution of fabrication methods, allowing to produce more complex structures in larger quantities with relative easy, demands for higher throughput, precision and accuracies when handling and examining with these components. In this chapter, the main components employed in SEM manipulations were reviewed, together with real-applications issues and other limitations found in the literature. Based on this information, an approach for improving the position accuracy of end-effectors is delimited.
1.4.1/
Review of the issues presented
The manipulation and characterization of micro- and specially nano-metric structures inside the SEM still presents multiple challenges, as seen earlier. A brief summary of the issues found in the literature collaborating to reduce the attainable performance on nano-manipulation in SEMs is presented in Table 1.1. These issues effectively limit the processing speed of samples. SEM imaging, often takes as a ground-truth to detect displacements, struggle to capture dynamic events and require postprocessing strategies that may not be suitable for real-time applications. For this reason, the behavior of manipulators, end-effectors and sensors inside the vacuum chamber is not well known. As result of these uncertainties and reduced dynamic information, tasks inside the SEM are often limited in speed and in accuracy, in both tele-operated and autonomous/semi-autonomous
34
CHAPTER 1. CHALLENGES FOR NANO-ROBOTICS
Component
SEM
Issues Operation in the vacuum restricts samples/components. Limited volume available for instrumentation. Image acquisition rates of a few Hertz restricts its use to static or slow dynamic events. Localization in the 3D space can be difficult.
Nanomanipulators
Multiple/complex robots needed to improve the system’s dexterity. Difficult to fully characterize their behavior in vacuum. Nanometric positioning precision and accuracies depend on image feedback (slow) or internal sensors (limited by resolution and speed, drift effects).
End-effectors
Adhesion and other effects during manipulation. Subject to disturbances and uncertainties. Precise information over the positioning/orientation of this device is not often available.
Sensors
Signal-to-noise ratios can be unfavorable. Precise information over the positioning/orientation of this device is not often available.
Table 1.1: Summary of issues presented during nano-manipulation tasks inside the SEM.
task, with outcomes including damaged samples and components, an increase in operation time per sample and a higher variability in results. If more in depth data about the phenomena harming the performance of components in SEM is available, improved strategies could be implemented for nano-manipulation tasks, bringing a direct gain on the overall system performance. The project NanoRobust, of which this thesis works are part, aims to improve several different parts of the nano-manipulation task, including image processing, calibration and visual servoing, and also component manipulation and handing. The goal is to perform, with nanometric accuracy, the manipulation and characterization of fragile micro and nano-structures entirely in the SEM environment. By doing so, the risk of damaging and contaminating samples during transport between tools is reduced. Furthermore, performing locally multiple tasks helps to improve the processing speed of parts. In the scope of the NanoRobust project, various expertise fields were gathered together to enhance different aspects for the tasks, aiming to boost speed, accuracy, reliability and robustness of the complete nano-manipulation system. In this context, this thesis works on exploring directly the dynamic behavior of structures and the existence of disturbances in the vacuum chamber environment, to improve the quality of fine motions performed by end-effectors. By measuring displacements and vibrations inside the vacuum chamber, it is possible to compare results between both air and vacuum environments, helping to quantify its influence on the overall positioning performance. Disturbance effects from the environment were reported to have impact over the attainable performance, yet they have not been deeply studied and understood. If these information were available, precise positioning tasks inside SEMs could be improved in various stages: SEM room and nano-manipulation design and constructions, and during
1.4. THESIS CONTEXT
35
the development of dynamic control strategies for an end-effector for high accuracy tasks.
1.4.2/
Thesis objectives
This thesis aims to contribute to nano-manipulation by investigating the nano-positioning task inside the SEM. In this environment, various different phenomena occur and can have significant impact when dealing with nanometer precisions. The objectives for this thesis can be summarized into three main points: 1.
Characterization of the most important mechanical disturbances that affect components inside the scanning microscope: The investigation of these effects may allow to better tackle the issue and to attenuate it more effectively. By comprehending its frequency spectrum, systems can be developed with characteristics outside its range. Conversely, if the disturbance sources were known, it could be possible to actuate directly on them.
2.
Identification of the pressure influences over a micro-actuator dynamic response, and development of a representative model for its dynamics in vacuum: Structures operating in the vacuum of a SEM can behave differently from when operating in air. These uncertainties can have a large impact when developing controllers and estimation strategies based on models. By comparing a system response in both conditions, it would be possible to reduce those uncertainties in the future.
3.
Development of positioning control laws capable of improving the system performance in this environment: Through the better understanding of the mechanisms degrading the performance on high-precision positioning tasks, more suitable control laws can be obtained, integrating fast dynamic responses and real time disturbances attenuation strategies. These strategies can be then explored to achieve higher levels of repeatability and robustness for end-effectors during handling and characterization of nano-objects and fragile structures.
One of the challenges limiting the achievement of these goals is directly related to the difficulty in obtaining precise and, at the same time, sufficiently fast positioning data on components inside the SEM. As stated earlier in this chapter, strategies based on the electron beam bombardment and integrated sensors may be not adequate, due to its low resolution and acquisition speed, unfavorable signal/noise ratios and problems such as heating and measurement drift when operation in vacuum environment. A different strategy should be employed to obtain reliable real-time information about the displacement of the systems. For this purpose, a dedicated vibrometer was used, allowing to explore their dynamics directly inside the SEM chamber. In the next chapter, the proposed experimental setup is described, along with the study of mechanical disturbances, the first objective of this work towards improving nano-positioning tasks on SEM. Chapter 3 presents the end-effector that was studied during this work: the FT-G30 micro-gripper. This device was characterized in air and vacuum, to study the influence of environment pressure over its dynamics. Chapter 4 deals with the nano-positioning control problem of the studied gripper, focusing on SEM applications. Two strategies were considered during this work: H∞ and Extended State Observer controllers, developed based on requirements for SEM applications and the available information over disturbances and dynamic characteristics of the considered micro-gripper. The robustness of these controllers were tested through simulation and experimentally validated. Finally, Chapter 5 presents some final remarks about this work and its contributions, and a brief discussion about perspectives for future developments.
2 Characterization and analysis of disturbance on SEM
The SEM is, as any other system, prone to the effects of disturbances. Those disturbances can be of various origins and may have significant impact on the precision of tasks developed on this device. Large portion of the literature on this subject focuses on how those perturbations deteriorate the images, with less regard to their effects on the elements inside the SEM. This chapter sketches an overview on the different disturbance sources and their impact on components in the microscope. Later, an experimental investigation is performed, to quantify what are the acting disturbances and their contributions to the displacement of samples in the vacuum chamber. For this purpose, the use of a dedicated vibrometer was proposed.
2.1/
Overview on sources of disturbances occurring in the SEM
When acquiring images with a SEM or using it as a base for performing complex tasks such as manipulation and characterization of samples, the best possible performance is desired. However, it is not rare to obtain sub-optimal results for a given device. In imaging, this could mean that the resulting resolution is lower than a nominal advertised value or the failure to reproduce these precisions and accuracies when manipulating objects in the vacuum. This document will focus on the second issue and how those perturbations affect components inside the vacuum chamber (samples, sensors, actuators), without concerning about the problem of image acquisition. In practice, disturbances acting upon a system limit the achieved results, as environmental perturbation can not be eliminated, only attenuated. The following analysis aims to determine how sensitive the microscope and components placed inside it are to those perturbations and how it can affect its dynamic behavior. To organize the study, sources of perturbation are classified in three main groups: • Mechanical: this category includes ground vibrations, transmitted from the surroundings to the microscope’s base and acoustic disturbances. • Electromagnetic: varying magnetic fields can interfere with the precise electronic components that allow the microscope to obtain high quality images. • Thermal: fluctuations in the room temperature can have an impact on the microscope’s performance, i.e., producing drifts in images. Furthermore, due to poor heat transfer in the vacuum chamber, sample temperatures tend to increase. 37
38
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
The effects of vibration in SEM imaging are known and are somehow reflected by the manufacturer recommendations over the microscope room vibration levels, described in its manual. For the Auriga60 SEM, from Carl Zeiss manufacturer, ground vibration should be limited to under 0.04 mm/s2 for frequencies lower than 9.5 Hz, and under 50 dB for acoustic disturbances with frequencies under 200 Hz. Electromagnetic fields should be limited to no more than 2 mG peak-to-peak for frequencies under 10 Hz, and temperature variation to 0.5◦ C/hour [Aur, ]. The following subsections detail each one of the three disturbance kinds, exemplifying how they can alter a sample’s behaviors and corrupt measurements.
2.1.1/
Mechanical and acoustic disturbances
Mechanical disturbances can affect the microscope through floor (background vibration) and air (acoustic vibrations). The sensitivity characterization of a microscope to background and acoustic vibrations is a complex task, as it depends on the frequency and on each of the microscope’s resonances, varying from instrument to instrument. In general, vibrations with frequency above 100 Hz are attenuated, while those from 1 to 10 Hz are in the microscope resonant frequency range and can become a problem [Muller et al., 2006]. Many studies about vibration on SEMs are focused on their effects to the image produced by the SEM, where the microscope’s column is disturbed, resulting in low quality images. Figure Figure 2.1: Effect of a short mechanical impulse 2.1 shows the effect of mechanical disturbance to the SEM electron optical column (500000x applied to the microscope column during imamplification). From [Vladar, 2003]. age acquisitions. The work of [Vladar, 2003] summarized some practices and rules of thumb when preparing a room to receive a SEM. It recommended placing the microscope in an area with reduced structural vibration (ground floor or underground), and distant from air currents and water piping, avoiding movements that will degrade the images. However, no remark was made about the effect of vibrations on the elements in the vacuum chamber interior. Vibration measurement and its frequency analysis provides an efficient tool to detect, identify and isolate the noise (cooling fans, air conditioning, human speech, ...) and vibration sources (heavy machinery, vacuum pumps, human interference, ...). If those sources can be identified, it would be possible to eliminate them (turning off unnecessary equipment, rearranging instruments, ...) or to reduce their effects by adding acoustic shielding to attenuate a specific frequency band of interest. The later solution is generally more expensive and time consuming and implementing it can be challenging for low frequency acoustic noise. However, some disturbances are not constant, and do not work in a specific frequency range. This is the case of air currents, commonly caused in the laboratory environment by air conditioning system. Those distortions can be difficult to detect, as they are produced by turbulent air displacements and do not present themselves in a single band frequency of the spectrum. If not taken into consideration, those disturbances can have an important impact in the performance of manipulation tasks. A few articles found in the literature can help to describe the
2.1. OVERVIEW ON SOURCES OF DISTURBANCES OCCURRING IN THE SEM
39
importance of those phenomena in practical SEM applications. In [Kim et al., 2003], the electrical characterization of carbon nanotubes with diameters ranging from 30 to 50 nanometers was performed inside a Field Emission SEM. The tools were positioned with the aid of two commercial micro-manipulators, actuated by high-precision piezoelectric motors. While the nominal step size of the motors can be adjusted between 1 and 200 nm, the practical attained accuracy was close to 10 nanometers. The article concluded that external disturbances caused the response degradation, without further investigation on its possible sources. Vacuum pump vibrations on images were studied in [Jung et al., 2012]. It remarked that, for high magnifications, these disturbances were transferred to the chamber despite the use of vibration suppressors. Through an external accelerometer, vibrations in the range between 0.05 to 500 Hz were measured. It demonstrated that passive anti-vibration dampers, installed on the SEM table and the vacuum pump, were not enough to ensure high-performance imaging at large magnifications (40000x or more). To attenuate those effects, an image filtering method was proposed, based on the measured vibration frequencies. A different approach was proposed in [Iwaya et al., 2012] through the use of an active vibration damping system. The authors sought to reduce the influence of low-frequency vibrations transmitted from the ground to the vacuum chamber. Using accelerometers to estimate the disturbance frequencies acting in multiple directions, different foundation slabs were compared to infer and improve their vibration damping characteristics. Also, electronic components and human operator interference, that could produce acoustic noises, were re-positioned to minimize their effects on the microscope. Despite its effectiveness in reducing low-frequency disturbances, this method requires a large effort, including physical modifications to the SEM room, such as adequating its foundation bed and installing soundproof walls. More recently, the work of [Pluska et al., 2015] studied the effects of acoustic influence on the focus ion beam (FIB). Through the use of an external noise source, the identification of resonant frequencies for the microscope’s column, the FIB gun and the positioning table was achieved. It remarked that the SEM and its components can be a source of disturbance, and compared the noise levels before FIB installation (40 dB) and after (68 dB during its operation), corresponding to a 25 times increase. As conclusion, the study highlighted the microscope sensitivity to acoustic noises as it may contain significant harmonic components matching those of the mechanical elements of the instrument. In this case, oscillations of 90 nanometers were reported. In the reviewed literature, this was the first demonstration of external disturbances having effects on components inside the SEM, distinguished from the imaging system. The reviewed literature indicates that external mechanical and acoustic vibrations play an important role in applications that aim to achieve nanometric precision. Besides the environment natural disturbances, SEM components can also contribute significantly as perturbation sources. While vibrations transmitted by the ground are limited to low frequencies (up to a few hundred hertz), acoustic noises may have a much larger band (up to 20 kHz). When these disturbances match the resonant modes of mechanical components inside the microscope, resulting vibrations can become sufficiently large and thus surpass acceptable levels for the task.
2.1.2/
Electromagnetic disturbances
Environmental sources of alternated current can affect the SEMs imaging acquisition process, by distorting the electron beam path or the signals applied to the scanning coils caus-
40
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
ing periodic deformations in the registered image. In [Fatikow, 2007], the use of high-resolution SEM images as an alternative to on-board sensors for identification and manipulation of micro and nano-sized parts was reviewed. Taking images as ground truth when performing tasks in the microscope is a common practice. However, these strategies can be limited by noises, low acquisition frequency and the movement of objects in the scene (drift). In many cases, the images are used for measurements and even for calibration of sensors, what can lead to incorrect or imprecise results. In [Muller et al., 2006], alternated current electromagnetic interference (EMI) on the scan coils of a scanning transmission electron microscope (STEM) was studied. The microscope column should shield radio frequency noise from the electron beam, but lower frequencies (less than 3000 Hz) are less effectively screened. To quantify the EMI, a large coil was built and placed at approximately 1 meter from the microscope, and the generated field was measured close to the microscope column. Images were recorded in the presence of induced alternated fields (32 mG peak to peak) as well as in their absence (0.2 mG peak to peak). The scanned structure, five unit cells of S rT iO3 grown on silicon, was 1.96 nm thick. The results shown in the figure exemplify these distortions caused to the images by large magnetic fields. For this set-up, average deviations of 48 pm/mG (peak-to-peak) were measured. The effects of alternating magnetic fields were analyzed in [Pluska et al., 2009]. During this study, a separation between direct beam interference and coil signal distortion was made. It concluded that perturbations acting directly on beams were affected by the electron energy, where high energies are less susceptible to path deviations, while interference in the electromagnetic coil signals distorted the beam path, regardless its energy. An example of those effects are shown in Figure 2.2. A disturbance along the X scanning axis d x was induced to the system, producing periodic deformations close to 200 nm peak-to-peak in the image.
(a) Low magnification
(b) High magnification
Figure 2.2: Effect of induced EMI disturbances on images [Pluska et al., 2009]. Electromagnetic interference can be difficult to detect and to distinguish from other disturbance sources. In [Vladar, 2003, Pluska et al., 2009], some characteristics of an adequate SEM room are described. They include avoiding placing the microscope in a room susceptible to electromagnetic and radio-frequency interference. If this is not possible, it is recommended to use active and passive shielding techniques to reduce these effects. It also suggested removing metal objects from the room, such as chairs and tables, as moving metallic masses can change the EM field around the SEM. Electric cables should be shielded, and positioned far from the microscope’s column, as they can produce significant interference, depending on their currents.
2.1. OVERVIEW ON SOURCES OF DISTURBANCES OCCURRING IN THE SEM
41
Overall, the degradation over the SEM operation, due to EMI documented in the literature is related to the image acquisition process and is not a major concern on this thesis scope.
2.1.3/
Thermal disturbances
The effects of temperature on devices constitute a problem when considering objects of micro or nanometric dimensions, as its measurement becomes complex, and the precise data acquisition can be affected by the sensor. In addition, systems in vacuum have a reduced heat exchange and its accumulation can be an issue. A particularity of SEMs lies on electron beam effects during image acquisition. Those accelerated particles, constantly colliding with the sample, can also generate considerable heat. Those thermal-related phenomena are discussed in the next subsections. 2.1.3.1/ Thermal noise Johnson-Nyquist noise [Johnson, 1928], also known as thermal noise, is used to describe how temperature can induce random vibrations in a lattice structure. It is most commonly applied to electric circuits, describing the voltage fluctuations at the terminals of a conductor or semi-conductor material at equilibrium. As the noise is random in nature, it can affect the whole frequency spectrum [Stievater et al., 2002]. Its power spectral density Vt2 is given by Equation 2.1, and represents the expected voltage variance per Hertz of the bandwidth: Vt2 = 4kB T R
(2.1)
with kB the Boltzmann constant, T the absolute temperature in Kelvin, and R the resistance of the electronic circuit. The same principle can be applied to infer mechanical-thermal noises. In this case, the spectral density for the thermal driving force Ft2 is given by: Ft2 = 4kB T Rmech
(2.2)
but considering Rmech the mechanical resistance (damping). In [Babic et al., 2003], the Young modulus for CNTs are estimated based on the mechanical-thermal induced excitation, relating the sample length and vibration amplitudes. For a CNT of 6 µm length and 80 nm diameter at room temperature, displacements close to 2 nanometers were measured experimentally. 2.1.3.2/ Thermal effects on sensors and micro-structures The measurement of temperature variations inside the vacuum chamber is a complex challenge, even more difficult when dealing with microscopic samples, as their dimensions limit the direct non contact and contact method usage. Due to the reduced heat exchanges with the environment, it is important to consider the contribution of different elements to the heating problem (i.e. extended use of the electron beam, Joule effect on components, ...). An example of how heat can compromise components was shown in [Toda et al., 2010], where a comparison on the pressure effects for the heating dissipation of bi-material cantilevers was performed. In this study, pre-bent silicon nitride and gold cantilevers (with dimensions 400 µm x 20 − 70 µm x 0.4 µm) were used as high-precision temperature sensors. By heating the cantilever’s tip with a laser diode of varying power (from 0.5 to 13 mW) and measuring the resulting
42
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
deformation of the beam, temperature variations could be estimated. In experimental tests, an incident power of 4 mW resulted in a tip temperature increase from 19 to 143 ◦ C in air, while the temperature at the cantilever’s base only rose a few degrees. This produced an average radius variation of the pre-bent cantilever of a few µm. Increasing the incident beam power to its maximum produced an average radius variation of approximately 100 µm in air. Performing a similar study in the vacuum revealed that, for a beam incident power of only 2.5 mW, an average radius change of 400 µm is achieved, while a 5 mW incident beam causes permanent damage in the sample. This illustrates the importance of temperature exchanges with the air, and how accumulated heat can quickly become a problem when operating in vacuum environments. The effects of heat in an optical encoder during long operation inside the SEM were also noticed in [Zhang et al., 2013a], where the axis of a 3 degree of freedom Cartesian micro-robot was analyzed in vacuum through image processing. The sensor characterization performed indicates a minimal resolution of 2 nm, while the nominal value indicated by the manufacturer was 1 nm. Also, the authors remarked the existence of long term drift. Figure 2.3 demonstrates this effect after long usage, attributed to the thermal expansion of the optical scale. In the tests, position values were measured by the encoder at 100 Hz for 10 minutes, with no action from the actuators. In this setup, the measured drift was approximately 1.2 nm/s. The averaged encoder accuracy for the device was within 98% of the travel distance.
Figure 2.3: Drift quantification of optical encoders on long usage inside the vacuum chamber [Zhang et al., 2013a]. Some encoders allow a power save mode, where the sensor power supply is disabled most of the time, and activated in sparse time intervals or when a movement command is issued. This option is useful to reduce the generated heat [Sma, 2013] at cost of degrading the reaction time, as the sensor needs to be re-activated. A similar study was performed in [Jasper et al., 2010], describing the use of a linear nanopositioner from Physik Instrumente, based on stacked piezoactuators, to achieve nanometer resolution with a large working range. For closed loop positioning, a capacitive position sensor was used. The downsides of this measurement strategy include frequent recalibration for longterm drift compensation, due to thermal variations. The difference between the sensor measured positions and image analysis indicates a deviation of up to 15 nm in precision positioning tasks. The effects of temperature fluctuations on cantilevers in low-pressure environments are described in several articles. In [Mertens et al., 2003], the use of cantilevers as precise thermal
2.1. OVERVIEW ON SOURCES OF DISTURBANCES OCCURRING IN THE SEM
43
sensors was studied, describing how small temperature variations (∆T = 0...30 K) can affect the frequency response of bending beams (silicon nitrite cantilevers, dimensions 200 x 20 x 0.5 µm, with added gold layers of different thickness). In experimental tests, variations smaller than 0.5% in the first mode were found for the defined ∆T , and a relation between temperature and Young’s modulus E was established. The relation ∆E/∆T could be considered linear for mono-material beams, while on bi-material cantilevers a nonlinear behavior appeared. This can be related to the difference of thermal expansion coefficients between the two materials, resulting on changes on the bi-material cantilevers deflection and frequency. The results are summarized in Figure 2.4, where four cantilevers (A for a mono-material and B, C and D for bi-material, with increasing gold coating thickness, cantilevers) had theirs deflection and frequency measured under temperature variations.
(a)
(b)
Figure 2.4: Deflection (a) and first resonance frequency (b) as a function of temperature for mono (silicon nitrite, Cantilever A) and bi-material (silicon nitrite with goal coating of different thickness, Cantilevers B,C and D) samples. [Mertens et al., 2003]. In a related work, [Sandberg et al., 2005] analyzed the influence of temperature on the first five resonant modes of cantilevers. Two silicon dioxide cantilevers (dimensions 195µm x 10 µm x 850 nm) were tested. One of them was coated with a thin layer of gold (400 nm), and the differences in its behavior due to temperature changes (from 30o to 80o C) were observed. Figure 2.5 demonstrates how the resonant frequencies change as functions of temperature for (a) the uncoated and (b) coated cantilever. Material properties, such as the density ρ and the Young’s modulus of elasticity E, are linked to the temperature. For uncoated cantilevers, the linear dependence between frequency and temperature was clear, in accordance with previous results in the literature. It was possible to verify that the thermal expansion effects are small and most of the time negligible for these samples. The elasticity thermal dependence was presumed the dominating effect in this change. For coated cantilevers, the behavior of different modes were non-trivial and attributed to the nonhomogeneous coating structure (Figure 2.5(b)), that could influence internal frictions and affect the individual modes differently. 2.1.3.3/ Heating due to electron beam As explained in Section 1.3.1.1, the different image acquisition processes in a SEM and TEM are based on the interaction of high-energy electrons with matter. Those interactions, however, can have important effects over the studied sample. In [Egerton et al., 2004], the conse-
44
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
(a)
(b)
Figure 2.5: Normalized resonance frequencies of the first five flexural modes (M1–M5) of (a) the uncoated and (b) the gold-coated cantilever as functions of temperature [Sandberg et al., 2005].
quences of such interactions on TEMs and SEMs were summarized. These effects include heating, electrostatic charging, permanent sample damage, and others. As the common accelerating voltage for a SEM is much lower than the needed ones for TEM imagery (from 1 to 30 kV in the first against 100 to 300 kV in the later), most of the electron-irradiation problems have a reduced impact in scanning electron microscopes. However, heating and electrostatic charging can still have significant influence in prolonged exposure situations. When an electron beam is focused on a material, various elastic and inelastic interactions can occur between electrons and the material. During inelastic scattering, most of the energy created is dissipated as heat within the specimen, causing a local rise in temperature. Materials with low thermal conductivity, such as polymers, are more sensitive to heating and may experience thermal degradation. Another important aspect to be considered is the sample geometry, as bulk specimens benefit from a radial heat flow in three dimensions, something not true for thin structures. The works of [Polyakov et al., 2014] remarked that, during the manipulation of gold nano-particles (nominal radius between 100 and 280 nm), as the diameter of structures decreases, so does its melting temperature. In the example discussed in the article, partial sample deformation was reported during the image acquisition process inside the SEM, due to the electron beam radiation. For experimental measuring of the temperature variation due to electron beam, different methods can be applied. Thermocouples can be employed for direct measures, but with limited spatial resolution. Another method consists in using low fusion point samples bombarded with different beam powers. A third method was proposed in [Bouscaud et al., 2012], measuring the lattice parameters of materials under the electron beam using the Kossel micro-diffraction technique. Knowing the linear thermal coefficient of the material, temperature variations could be deduced from the lattice dilatation. This method is attractive, as modern techniques can determine lattice parameters with high precision. In this article, two materials (copper and germanium) were chosen, due to the differences in their thermal conductivity. The study consisted in changing the incident beam power P and measuring variations in the lattice structure. The local temperature rise ∆T induced by the electron beam on the bulk specimen can be calculated using Equation 2.3:
2.1. OVERVIEW ON SOURCES OF DISTURBANCES OCCURRING IN THE SEM
45
P −1 (q − Q−1 ) (2.3) 2πλ where λ is the thermal conductivity of the specimen, Q is the approximated sample dimension and q is the distance from the probed zone on the material corresponding to the x-ray source used for the micro-diffraction technique. The true lattice parameter without heating atrue and the measured lattice parameter with beam heating ameas are related by Equation 2.4, with αt the linear thermal expansion coefficient. ameas = atrue (1 + αt ∆T ) (2.4) ∆T =
Equations 2.3 and 2.4 established a linear relationship between lattice parameter and electron beam power. By measuring those quantities and its uncertainties, it was possible to determine the relation between local temperature variations and beam power. The experiments performed in [Bouscaud et al., 2012] led to the following relations between beam power and temperature variations for the two metals: ∆T (Cu) = 102(±22).P ∆T (Ge) = 606(±11).P As the normal operating power could vary from 0.003 to 0.03 Watts, the local temperature increase in bulk materials, due to the electron beam, could achieve values close to 3 degrees for copper and 27 for germanium. While these temperature variations were still small in this case, they could be greatly increased if the studied materials were less conductive and their geometry restricted the heat dissipation. It was reported in [Egerton et al., 2004], for the analysis of thin film of low conductivity samples, temperature increases of a few hundred degrees for a stationary scan, even considering lower energy electrons (from 0.5 to 2 keV). In this situation, the beam energy was deposited in the sample, where the heat flow is only two dimensional. Considering normal raster scan, the temperature rise could be reduced by a factor of ten.
2.1.4/
Conclusion
Despite the best efforts to attenuate their effects, external disturbances are still a limiting factor for high-performance imaging and manipulation in the SEM. Mechanical vibrations from the surroundings and the microscope itself can affect components placed inside, even when passive and active damping techniques are implemented. Acoustic disturbances can affect significantly systems in vacuum through mechanical coupling, and even relatively small noise levels may result in a large degradation of the positioning accuracy. A SEM is a complex system, and these perturbations can excite the vibration modes of various components (robotic stage, micro-manipulators, samples, ...). Electromagnetic interference can affect electronic devices in the SEM and the natural caution should be exerted. The major problem caused by this kind of perturbation is related to the image acquisition process. As the microscope imaging mechanism is based on large voltages and the use of magnetic fields, special attention should be given when preparing the SEM room, to avoid image deterioration. However, they represent artifacts in the image, and not real displacement of components. Thermal disturbances can become a serious issue in the vacuum, due to the low heat transfer characteristics and the reduced object dimensions. The effects of heat accumulation can degrade measurements and modify the characteristics of components, to the point of generating premature failure in the device. A special case should be made over the heating due to electron
46
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
beam incidence, a particularity in this family of microscopes. The incident beam can generate considerable amounts of heat during image scanning processes. If the beam bombardment is too frequent (i.e. when defining a small scanning area to improve the acquisition frame rate), and has enough energy, heat is added faster to the systems than they can dissipate. This effect is emphasized when considering thin or small samples and poorly conductive materials.
2.2/
Vibration measurement inside a SEM
From the previous section, there are clues indicating that disturbances on electron microscopes can be of various origins (mechanical, thermal and electromagnetic). A complete analysis of each one of those sources would be ideal for the development of precise and detailed nanomanipulation systems. In this work, only mechanical and acoustic excitations will be studied, as, based on the literature review, these are believed to be the major contributors for the positioning task degrading in the SEM environment. Disturbance sources, such as vacuum pump vibrations and acoustic noises, are present in most of SEMs operation environments. The influence of these disturbances poses a recurring challenge when performing precise micro- and nano-manipulation tasks, as the induced vibrations (from 0 to a few kilo-Hertz) can excite the vibration modes of components inside the microscope. These vibrations can affect samples, end-effectors and other components inside the vacuum chamber, which limit the achievable accuracy of a manipulation system. To characterize these disturbances, it would be interesting to measure how they can affect elements of the microscope. The two main vibration measurement methods found in the literature are based on: • Electron beam bombardment: Processes images (or parts of them) and estimate the vibration amplitudes and frequencies. However, it may be difficult to separate sample vibrations from electron gun column vibrations. This analysis may take a long time and requires multiple image acquisitions, which reduces usage in real-time applications. An example is the stroboscopic electron scanning microscopy [Fujioka et al., 1980, Ishikawa et al., 1993]. Reducing the scanned area (i.e. a single line, instead of the full image [Wong et al., 2007, Jasper, 2011]) improves the acquisition and processing speed, while losing the images capabilities of the SEM. • Accelerometers: The use of these instruments allows retrieving information in real time for multiple axes. So far this method is limited to measurement of disturbances of the vacuum chamber and other large components, as high-sensitivity sensors are bulky (up to hundreds of grams) and their measurement ranges may be limited. While accelerometers can offer an interesting approach to estimating vibrations effects over the vacuum chamber, this method does not capture the actual behavior of components inside the chamber. On the other hand, electron beam methods can measure information directly on components of interest inside the SEM. However, it may require extensive post-processing of a large amount of images to retrieve this information. Furthermore, these methods may not achieve enough spacial resolution to identify vibrations at the nanometric scale. To overcome these limitations, a different method is proposed based on the use of a dedicated laser vibrometer. This kind of sensor performs non-contact displacement measurements, with great accuracy over a large spectrum, without interfering with the image acquisition process.
2.2. VIBRATION MEASUREMENT INSIDE A SEM
47
Through this device, the displacement information of an object can be measured directly inside the vacuum chamber, and is available in real-time, with minimal need of post-processing. The setup consists of fixing a vibrometer (SP-120, vacuum compatible, from SIOS Meßtechnik GmbH) inside a SEM (Carl Zeiss SEM Auriga 60, shown in Figure 2.6). They are fixed at a relative 45◦ angle, so both the laser and the electron beam intersect. This allows acquiring images and perform displacement measurements simultaneously. For this purpose, a rigid metallic support was designed and fixed to the SEM door, independently from the microscope six degrees-of-freedom positioning stage (four translations and two rotations), allowing relative movement between Figure 2.6: Carl Zeiss SEM Auriga 60, from the both structures. EQUIPEX ROBOTEX project, where the experSamples were fixed through an approiments were performed. priate support to the stage, that positions it inside the vacuum chamber. The sample can be remotely positioned using a joystick or by a software interface, as it is necessary to adjust the laser beam’s incident angle and will be described later in this chapter. Finally, the sensor is connected to its external processing unit through the appropriate feed-through ports (optical fiber and electrical connection). This processing unit can then transmit the data to other devices, via digital or analog signals, in real time. Figure 2.7(a) illustrates the proposed setup, with the measured and control signals available to the user. Figure 2.7(b) shows its implementation inside the microscope, where the vibrometer head was attached to the SEM door through an aluminum support at a 45◦ angle. Specimen Chamber (vacuum) SEM Image Electron beam power
Vibrometer’s support
Vibrometer’s head
Vibrometer
Electron beam
Sample displacement
3D positioning table (6 axes) Actuators/sensors
Positioning Stage
(a)
SEM’s door
(b)
Figure 2.7: (a) Scheme of the experimental setup proposed and (b) implementation inside the SEM.
2.2.1/
Vibrometer characterization
To correctly estimate the disturbance levels in the system, it is necessary to first quantify the intrinsic noise levels existent in the measurement system and the practical attainable resolu-
48
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
tions achieved by the transducer. This can be estimated by measuring the vibration of a rigid metal block at different reflection levels. During the characterization process, the metal block and the vibrometer were fixed to a pneumatic anti-vibration table, to minimize ground vibration interferences. The measurements were acquired through the vibrometer analog output, and digitized by a 12 bits A/D converter (dSPACE DS1103 PPC Controller Board). When operating in analog mode, the vibrometer allows users to select an analog output gain. This gain regulates the measurement range (maximum vibration amplitude before saturating the output signal). These values, described in the manufacturer manual [SP1, 2012], are summarized in Table 2.1 for each output gain named as G0, G1, G2 and G3. Analog output gain G0 G1 G2 G3
Maximum oscillation amplitude (in µm)
Measurement gain K (in µm/V)
0.63 2.50 10.10 40.30
0.24 0.97 3.87 15.48
Table 2.1: The vibrometer offers several analog output gain options, with different operation ranges.
To estimate the noise levels for these gains, a series of measurements were performed for different reflection levels, representing a metric, indicating the laser beam percentage reflected back by the sample and captured by the sensor. It can be interpreted as the signal quality, where a higher value ensures a lower noise level. The measurements for the calibration tests were performed with an acquisition frequency of 25 kHz. For each one of the four analog output gains to be tested, and for each studied reflection level (30, 40, 50, 60 and 70%), ten measurements were performed. These twenty different scenarios intent to represent common values found in normal operation conditions. The manufacturer recommends reflection levels greater than 50% to ensure measurement precision. However, in real applications, these values depends on the studied surface and its roughness, were high reflection can be difficult to achieve. Each acquired measurement was processed using Welch’s power spectral density (PSD) method [Welch, 1967]. The ten resulting frequency spectra for each scenario were then averaged, to obtain an estimation of the sensor’s frequency response. Figure 2.8(a) shows, for the analog output gain G0, the reflection level influence over the measurements and how high levels can improve the signal quality, up to a limit. In this case, it is close to a 60% reflection level. The graph shows significant peaks around 700 and 4000 Hz. These frequencies were considered part of the sensor’s characteristic response (measurement noise). Figure 2.8(b) shows, for a close reflection level (50%), how the analog output gain influences the measurements. There is a loss in signal quality as the gain increases, reflecting the resolution reduction with measurement range increase. Furthermore, it is possible to notice for gains G2 and G3 peaks at 1600 and 3200 Hz, in addition to the ones mentioned before. Table 2.2 resumes the sensor’s analog output results, considering different reflection levels. The measures performed with gains G0 and G1 have shown similar noise levels, credited to
2.2. VIBRATION MEASUREMENT INSIDE A SEM
49
−18
10 30% 40% 50% 60% 70%
Welch’s PSD (m2/Hz)
10
−21
10
−19
10
Welch’s PSD (m2/Hz)
−20
−22
10
−23
10
G0 G1 G2 G3
−20
10
−21
10
−22
10
−23
10 −24
10
−24
0
500
1000
1500
2000
2500
3000
3500
4000
4500
10
0
500
Frequency (Hz)
1000
1500
2000
2500
3000
3500
4000
4500
Frequency (Hz)
(a) Averaged PSD for G0 input gain, considering different (b) Averaged PSD for a constant reflection level of 50%, reflection levels. considering different output gains.
Figure 2.8: Vibrometer PSD results, considering different reflection levels and analog output gains. Average RMS noise level and standard deviation (in nm) Gain
G0 G1 G2 G3
Resolution (nm)
Reflection level (%) 30
40
50
60
70
1.1 ± 0.07 0.9 ± 0.17 1.3 ± 0.02 4.6 ± 0.06
0.68 ± 0.12 0.56 ± 0.09 1.2 ± 0.02 4.4 ± 0.02
0.55 ± 0.16 0.57 ± 0.02 1.2 ± 0.01 4.4 ± 0.05
0.37 ± 0.06 0.48 ± 0.04 1.2 ± 0.04 4.4 ± 0.08
0.42 ± 0.14 0.48 ± 0.15 1.1 ± 0.03 4.4 ± 0.03
0.07 0.3 1.2 4.7
Table 2.2: Measurement noise RMS value and standard deviation for different output analog gains and reflection levels, and maximal measurement resolution for each gain.
residual disturbances not compensated by the anti-vibration table (i.e. acoustic perturbations form devices in the room, such as computers and cooling fans) and electronic noise. It is possible to see that, for the gains G2 and G3, resolution becomes a limiting factor. The effects of quantization are shown in Figure 2.9, where the loss of information becomes evident. Therefore, a trade-off should be achieved between desired resolution and measurement range. Hence, G1 will be used for all measurements, except when the desired displacement range exceeds the limit value of 2.5 µm. In that case, the gain value G3 is applied.
2.2.2/
SEM disturbances and noises identification
As stated in Section 2.1, components inside the SEM are subject to unknown disturbances that may produce undesired vibrations during micro- and nano-manipulation tasks. From the possible sources of disturbance, mechanical disturbances appear to be the major contributor. Its precise identification and quantification can offer valuable information to researchers aiming for extreme high accuracies for positioning tasks in SEM. To study the influence of external mechanical disturbances over components inside the SEM, two different tests were proposed. The first observed the perturbations constantly affecting
50
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
20 15
Displacement (nm)
10
G3 G2 G1 G0
5 0 −5 −10 −15 −20 0
0.2
0.4
0.6
0.8
1
Time (s)
Figure 2.9: Acquired signals for different output analog gains, at a constant reflection level of 50%. the microscope are under regular operation conditions. In the second test, a controlled external acoustic disturbance source was added in the environment to investigate how these excitations can be transmitted to structures inside the vacuum chamber.
2.2.2.1/ System under regular operation condition This step considered tests performed in conditions close to real practical tasks and aimed to identify the most important disturbances frequencies and to characterize its sources. It was performed in the SEM chamber at high-vacuum (10−4 Pa), with reduced human activity in the room. The sample, a set of silicon cantilevers of different lengths, was fixed in the sample holder and placed in the microscope positioning table (Figure 2.10).
Figure 2.10: Silicon cantilevers in the sample holder.
2.2. VIBRATION MEASUREMENT INSIDE A SEM
51
The cantilevers approximated dimensions are given in Table 2.3, together with the measured first resonant frequency in air, f1air . These samples are interesting during the identification process due to its reflective surface, allowing for better reflection levels and, consequently, improved displacement readings. In addition, the different cantilever lengths and first modes may emulate the behavior of various structures, as those are roughly similar to values found in microgrippers and other micro-positioning tools.
Name Cantilever 1 Cantilever 2 Cantilever 3 Cantilever 4
Dimensions (l x w x t, in µm)
f1air (Hz)
5000 x 400 x 20 4000 x 400 x 20 3000 x 400 x 20 2000 x 400 x 20
1162.0 2013.4 3649.1 8337.4
Table 2.3: Nominal dimensions (length l, width w and thickness t) and first resonance mode measured at atmospheric pressure for the different silicon cantilevers.
The silicon samples do not possess any kind of actuators or electrical connections, and therefore all the displacements measured were caused by mechanical vibrations (from inside and outside the SEM) transmitted to the cantilever, or sensor noise. The laser reflection calibration, as stated before, is essential to ensure good measurement quality. However, once that all the setup is sealed inside the vacuum chamber, the only visual hint to aid in calibration comes from an infrared camera, which can become tiresome when dealing with micro-elements. Therefore, an initial calibration process is performed manually while the SEM door is open, when rotation angles and positions are adjusted. Once the chamber is closed and sealed, only small adjustments can be performed. In this particular case, the high reflection for the silicon cantilevers simplifies the laser calibration process. In consequence, reflection values larger than 50% are usually encountered. For each silicon cantilever, twenty separated measurements were acquired and had Welch’s PSD method applied to each of them. To further reduce measurement noises and the effects of sporadic disturbances, the obtained PSD’s were averaged for each cantilever. Figure 2.11 shows the resulting curves for silicon cantilevers from 0 to 1 kHz. From the graph, six major peaks (56.1, 95.4, 140, 235.7, 688.55, and 955.2 Hz) in the spectrum frequency were identified, besides the cantilevers resonant modes (not shown in the figure) and other minor peaks with smaller contributions. The influence of sensor noise is present in the measurement through a high frequency peak close to 4010 Hz that is not shown in the graph. The root mean square (RMS) vibration level measured for the different cantilevers ranged from 1.07 to 2.55 nm, depending on the sample’s length. Performing similar measurements at the cantilevers’ base, to eliminate the first vibration mode influence, found RMS displacements values between 0.86 and 0.98 nm. As expected, both tip and base showed similar frequency spectra, with exception of the first cantilever mode, responsible for this difference in the computed RMS value. It is possible to infer that, for longer cantilevers, its first vibration mode presents a larger contribution to the total vibration. The following step in the disturbance analysis is to identify the origins for each one of the six main vibrations peaks. In this phase, new sets of measurements were acquired for the cantilevers, considering small variations in the environment. Comparing results obtained before and after these variations can help isolate each one of the disturbance’s origin.
52
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
−19
10
Cantilever 1 Cantilever 2 Cantilever 3 Cantilever 4
56.07 Hz −20
Welch’s PSD (m2/Hz)
10
95.36 Hz
688.55 Hz
140 Hz
−21
955.2 Hz
10
235.75 Hz −22
10
−23
10
−24
10
0
100
200
300
400
500
600
700
800
900
1000
Frequency (Hz)
Figure 2.11: Averaged power spectrum density of the silicon cantilevers, detailing the frequencies with larger amplitude.
The vibrometer support was fixed to the SEM door at one extremity and free at the other. It can therefore be interpreted as a clamped beam, and may have a vibration mode in the studied range. To verify this assumption, the finite element analysis shown that the expected vibration mode for this component is close to 120 Hz, within the frequency range of the experimental values (Figure 2.12). An experiment was proposed to validate this numerical result. It consisted in changing the vibrometer’s head placement on the support, as this component has a considerable mass (approxi- Figure 2.12: Finite element analysis for the vimately 300g). Positioning it closer to the sup- brometer’s support. port fixed end resulted in the frequencies located at 56.1 and 95.4 Hz to shift to 82.3 and 122.5 Hz, respectively, while other frequencies in the spectrum remained mostly unchanged. Following the same analysis, the microscope stage itself is also fixed to the SEM door and free in the other end. In this case, the finite element analysis was not possible, as no model for this component is available. Therefore, a different experiment is proposed, consisting in adding weights (metal blocks of approximately 185g) to the support and to the positioning table, alternately. When placing weights on the vibrometer support, frequency variations at the same peaks as before (56.1 and 95.4 Hz) were observed. It is possible to conclude that these are measurement noises and do not represent actual sample displacements. When placing the weight over the SEM positioning table, the peaks previously located at 140 and 235.7 Hz shifted in frequency. These
2.2. VIBRATION MEASUREMENT INSIDE A SEM
53
peaks were then attributed to mechanical vibrations of the positioning table. These values are within the frequency range of the positioning stage obtained in [Pluska et al., 2015] for a different SEM. To help identify the remaining frequency peak sources, an additional experiment was proposed. It aims to investigate the proposition found in literature that the SEM itself and its components could be the cause of some observed disturbances. For this purpose, measurements were acquired under two different conditions: • Condition 1: SEM under normal operation conditions, with a high vacuum (10−4 Pa) but no electron beam bombardment. • Condition 2: SEM powered off (electricity shut off), with a high vacuum (10−4 Pa). Enforcing Condition 2 ensures that all SEM elements (i.e. vacuum pumps, cooling fans, ...) were turned off, allowing to verify if those elements had significant contribution in the measured disturbance spectrum. As the specimen chamber stayed sealed, the vacuum was kept in similar conditions as the normal SEM operation levels. In Figure 2.13 the PSD for 688.5 and 955.2 Hz frequencies are compared in both conditions. It became evident these disturbances sources were related to the microscope. The frequency around 700 Hz is attributed to the turbo-molecular pump. Differently from the main rotary pump, this component is always active to ensure the high-vacuum level necessary for the SEM correct operation. In this particular case, this pump has a nominal rotation frequency of 42000 rpm (700 Hz), in agreement with the experimental values. It is worth to remark that the pumps are placed over passive damping systems, and connected to the SEM using components to minimize the vibrations. −20
10
Condition 1 Condition 2
Condition 1 Condition 2
−21
2
Welch’s PSD (m /Hz)
Welch’s PSD (m2/Hz)
10
−21
10
−22
10
−22
10
640
660
680
700
720
Frequency (Hz)
(a) Around 700 Hz
740
930
940
950
960
970
980
Frequency (Hz)
(b) Around 955 Hz
Figure 2.13: PSD around 700 (a) and 955 Hz (b) comparing the SEM in two conditions. It indicates that the SEM could be the possible origin for these disturbances. The 955.2 Hz frequency can also be related to the SEM operation, as shown in Figure 2.13(b). This peak presents some interesting characteristics. When performing multiple measurements, for the same cantilever, at different times and over a span of several days, the peak varies between 949 and 957 Hz, always with similar amplitudes. These characteristics are also present when considering the other silicon cantilevers. This effect was attributed to electronic interference over the vibrometer signal, as it presents itself in a well defined frequency, similarly to the effects
54
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
of 50 Hz noise and its harmonics. However, the nature of this phenomenon is not known precisely, except the fact that its origin is related to the SEM. The identification of disturbance sources in the SEM represents an important step towards improving the micro- and nano-manipulation in this environment. It allows designing improved systems that operate outside the disturbance ranges and offers an insight on how external and internal SEM components can affect samples in the vacuum chamber. However, as stated before, the major disturbance contribution are due to first mode vibrations. Figure 2.14 shows the PSD around Cantilever 1’s first mode, comparing between both test conditions. It is important to remark that, during normal operation (Condition 1), the PSD level for this peak is two orders of magnitude higher than any other peak presented in the spectrum. The cantilever vibration was largely reduced when the SEM was turned off, even though there were no evident disturbances acting close to the sample resonant mode. Other disturbance frequencies related to mechanical components, namely vibrations in the positioning stage and in the vibrometer support, also showed decreased amplitudes, with the average RMS displacement in Condition 2 of approximately 0.84 nm.
−17
10
Condition 1 Condition 2
Welch’s PSD (m2/Hz)
−18
10
−19
10
−20
10
−21
10
−22
10
1120
1140
1160
1180
1200
1220
1240
Frequency (Hz)
Figure 2.14: PSD of Cantilever 1, detailed for its first mode (1169.8 Hz), comparing both test conditions. The experiments shown above demonstrated that elements inside the vacuum chamber are prone to the effects of external mechanical disturbances. Furthermore, the microscope can contribute as a source of disturbance. This experimental identification also offers an approximated idea for each disturbance contribution to the overall vibration level. This information is valuable when considering systems to be operated in the SEM environment, trying to avoid those critical frequencies when possible. 2.2.2.2/ System under external acoustic excitation The existence of acoustic vibrations in the working environment is a recurring problem when dealing with micro- and nano-manipulation tasks. During previous experimental measurements, it became clear that acoustic noises can have a substantial impact on systems inside the
2.2. VIBRATION MEASUREMENT INSIDE A SEM
55
SEM. As the system operates in a vacuum environment, perturbations can only be transmitted through mechanical coupling. In the case where acoustic excitations carry enough energy to produce displacements in the SEM chamber, these can be transmitted to the positioning stage and, consequently, to samples and manipulators. Those vibrations can be generated by Signal SEM human interference and surrounding equip- Loudspeaker generator ment, are not necessary continuously affecting the system and can cover a wide range of frequencies and amplitudes. To better understand the coupling between external acoustic excitations and components located inside the SEM, a different experimental setup was proposed (Figure 2.15). Through the use of a signal generator and a loud-speaker positioned at 1 meter from the SEM, controlled acoustic noises with different frequencies and amplitudes were introduced as an external disturbance source, Figure 2.15: Photo of the experimental setup for while sample displacements inside the vacuum external acoustic excitation, with the SEM, the chamber were measured. The loudspeaker em- loudspeaker and signal generator. ployed for this test has an approximate constant response for frequencies between 1 kHz and 10 kHz (approximately 90 ± 4 dB for a 2.83Vrms signal at 1 meter), so sinusoidal excitations within this range should produce similar dB sound levels for a given voltage input. Additional information on its response characteristics can be found in Appendix A. The worst case scenario was tested, where the system was subjected to a constant sine wave disturbance. A set of sample displacement measurements was acquired before activating the loudspeaker, and the average RMS displacement obtained after ten measurements was 1.07 nm. Generating a sinusoidal signal of 1000 Hz through the loudspeaker (outside the range of any resonant frequency previously measured), produced small displacements of the sample. These vibrations were only noticeable through a careful analysis of the frequency spectrum, and for large noise amplitudes (70 dB or greater). In this case, the added disturbance had a small effect, and other mechanical disturbances previously measured were still dominant. For comparison, a 70 dB external acoustic excitation with a frequency not matching any resonant mode produced an average RMS vibration level of 1.15 nm, slightly above the previous measured vibration levels. Nonetheless, if the frequency applied matches a resonant mode (in this case, 1169.8 Hz), much lower excitation power can have a large impact over the vibration amplitude. For Cantilever 1, the system achieved a constant 21 nm peak-to-peak vibration for an approximated excitation level of 50 dB (normal conversation level) and more than 70 nm peak-to-peak for a 60 dB perturbation. Figure 2.16 shows the obtained result for this measure. In the graph, external excitation starts at 1 second. This meaningful increase in vibration amplitude for a relatively small acoustic excitation shows how easily external disturbances can affect components in the SEM chamber. The scenario described above, where an external acoustic disturbance frequency matches those of components inside the chamber for an extended period, is difficult to be found in real work environments. In reality, a laboratory is prone to different acoustic disturbances and pressure variations. They originate from the environment and may present themselves in a large frequency band, affecting one or multiple resonant modes of components in the specimen chamber. This test exemplifies how even low amplitude noises may have a large impact on the positioning accuracy
56
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
40 60 dB acoustic excitation 50 dB acoustic excitation
30
Displacement (nm)
20 10 0 −10 −20
Start of acoustic excitation
−30 −40 0
1
2
3
4
5
6
7
8
Time (s)
Figure 2.16: Displacements measured on Cantilever 1 under the influence of an external noise source of different powers, generated by the loudspeaker. A sinusoidal wave with frequency matching the sample’s first mode is applied as the disturbance signal. The black dashed line indicates when the external excitation starts.
inside the SEM. Even in a high vacuum environment, it is evident structures are still subject to acoustic disturbances through mechanical coupling.
2.2.3/
Conclusion
To analyze the disturbances acting inside a SEM without resorting to its electron gun or imaging methods, a new setup was proposed based on the use of a dedicated vibrometer. It was placed inside the SEM vacuum chamber and allowed to measure the displacements over a set of samples with sub-nanometric precision and with sampling rates up to tens of kilo-Hertz. This original structure allowed to obtain a unique insight on the disturbances in this environment. The work focused on mechanical and acoustic disturbances effects on the vacuum chamber. The analysis of collected data, through extensive experimental investigation, reveals that the scanning electron microscope itself contributes largely as a disturbance source. Figure 2.17 resumes the results found experimentally, showing the sources for each one of the most significant disturbances. The largest contribution, however, is due to the sample’s oscillations. Experiments indicate that this mode can be excited by the SEM during normal operation. Another effect commonly overlooked is the influence of external acoustic perturbations. Through the use of a loudspeaker, to control the acoustic disturbance power and frequency, it was possible to witness how even small noises can produce significant vibrations on elements inside the vacuum chamber. As acoustic disturbances are present in real operation environment and can cover a large band in the frequency spectrum, they constitute an authentic issue when high accuracies are required.
2.3. CONCLUSION
57
−19
10
Cantilever 1 Cantilever 2 Cantilever 3 Cantilever 4
Vibrometer support −20
Welch’s PSD (m2/Hz)
10
Positioning Table
SEM generated electronic noise
Secondary Vacuum pump
−21
10
−22
10
−23
10
−24
10
0
100
200
300
400
500
600
700
800
900
1000
Frequency (Hz)
Figure 2.17: Summary of the most important measured sources of disturbance.
2.3/
Conclusion
In this chapter, a literature survey indicates the presence of four possible sources of disturbance in the SEM: mechanical vibrations, acoustic noises, electromagnetic interferences and thermal related disturbances. When considering SEMs, studies concerning disturbances focus mainly on imaging and how those perturbations affect the electron beam column or the beam itself. However, a few results indicate that the column is not the only influenced component, and that other elements inside the SEM are also subject to those effects. The literature also indicates that, from those possible sources, mechanical and acoustic disturbances should present the largest contribution in producing undesired vibrations. The identification and quantification of those contributions in micro- and nano-devices is a challenge, where the local conditions (limited volume, vacuum environment) impose additional restrictions making precise position measurements of components inside the microscope not always possible. Through the use of a dedicated vibrometer, some of those difficulties have been overcome, allowing the measurement of disturbances acting over samples inside the SEM in real-time, independently of the imaging system. The performed measurements concerned real movements of components in the vacuum chamber, and not artifacts generated by the SEM image acquisition system, as those commonly caused by EMI. The samples consisted of silicon cantilevers of different lengths. During this experimental study, it became clear that the major vibration frequency was linked to the excitation of the sample vibration modes. Other mechanical components inside the SEM (positioning stage, vibrometer support) and the secondary vacuum-pump produced additional vibrations. Experimental tests demonstrate that the microscope is especially sensitive to acoustic disturbances. Figure 2.18 shows, in a simplified graph, how different sources of disturbance affect systems in relation to their scale. The curves are rough estimates, as they depend on multiple parameters (geometry, material, vibration modes, disturbance amplitudes and frequencies, ...). In this work, the studied cantilevers have lengths of a few millimeters. In the tests performed, the
58
CHAPTER 2. CHARACTERIZATION AND ANALYSIS OF DISTURBANCE ON SEM
frequency band of acoustic noises encompass the vibration mode of various components, while mechanical vibrations have a limited effect in high frequencies.
Disturbance influence
μm Acoustic
nm Mechanical Thermal pm nm
μm
mm
System’s size
Figure 2.18: Diagram relating the different sources of disturbances and their influence over systems of various sizes. Thermal noise affects a large frequency range, and should be the dominant effect in systems with high resonant frequencies. From the literature, it is possible to infer that it has a reduced influence on our system, where the expected induced displacement rests close to or under the resolution limits of the vibrometer used in the test. Another issue presented is related to heat accumulation, that becomes a critical factor in the vacuum that can change a system dynamic characteristics or even damage components, if no caution is taken. The reviewed literature indicates that thermal noise effects become more important as the dimensions of the samples are reduced. In this case, those effects are not considered as the tested cantilevers are sufficiently large. Mechanical vibrations, transmitted from the surroundings through the ground, act at much lower frequencies (generally from 0 up to 400 Hz, although some devices may achieve higher frequencies). SEMs are generally equipped with passive and active dampers, aiming for lower frequencies, what improves their performances in this range, yet those effects can still be observed. Although in this work the mechanical vibrations generated by the vacuum pump had a limited contribution, it could still be measured. Acoustic noises, however, act on a larger band (from 0 to 20 kHz normally) and are damped less effectively by the SEM. These variations in pressure can have various sources (human conversation, operating devices, air flow due to air conditioning, ...). In this study, acoustic disturbances of relatively low levels with frequencies matching the sample’s vibration modes can induce large displacements. These induced vibrations were sustained for a long time due to the low damping in the vacuum environment. In this study, acoustic excitation was the most important source of disturbance.
3 Micro-gripper dynamics identification in vacuum and at atmospheric pressure
The identification of micro-grippers and actuated cantilevers in air is widely spread in the literature, presenting several models, each one with its own specific characteristics depending on the studied system ([Carrozza et al., 2000, Eisinberg et al., 2001, Boudaoud et al., 2010, Rakotondrabe et al., 2010], to cite a few). However, the dynamic characterization of those structures in vacuum is not well documented. Studies about the pressure effects focus on simpler components, such as passive cantilevers and plates ([Epp et al., 2004, Zhang et al., 2005, Sandberg et al., 2005]), and only general information is available. In this chapter, the identification of an electrostatic micro-gripper actuation system, encompassing comb-drive and finger, is performed in air and vacuum, and its differences are analyzed. In the end, two models capable of representing the system’s most important dynamics in all operation range are presented.
3.1/
The FT-G30 micro-gripper
For this work, commercially available micro-grippers (model FT-G30, from FemtoTools GmbH.) were used. The micro-gripper consists of two fingers: a comb-drive mechanics actuates one of them, while the other is passive and capable of force measurements though a built-in capacitive sensor. Figure 3.1 shows a SEM image of one gripper and its components. The objective is to precisely characterize the gripper actuation system, including the comb-drive and the actuated finger, during its free movement (no contact interaction with samples or the sensing finger). The Figure 3.1: SEM image of the micro-gripper FTactuated finger dimensions are approximately G30 and its main components. 4000x120x50 µm, and the nominal distance between its fingers is 30µm, which defines its working range. An example of its dynamic response is presented in Figure 3.2 for a 30 V step input, acquired at atmospheric pressure. It reveals the behavior of a low damped oscillatory system, with overshoot close to 100% and settling time around 0.1 second. 59
60
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
11 10
Displacement (µm)
9 8 7 6 5 4 3 2 1 0 0
0.05
0.1
0.15
Time (s)
Figure 3.2: Example of the gripper’s step response in air, for a 30 V input. In [Boudaoud, 2012], the actuation mechanism of a similar electrostatic micro-gripper was modeled, considering the forces and moments on the comb-drive system. Its parameters were experimentally estimated for atmospheric pressure, in a gray-model identification process. The actuation system, detailed in Figure 3.3, is composed of a central shuttle (connecting the combdrive and the finger), a set of suspensions responsible for holding the shuttle, and a hinge serving as rotation point for the finger and connecting it to the base of the structure..
Shuttle Actuated finger
Suspensions Hinge 2 mm
Figure 3.3: Micro-gripper actuated finger and comb-drive system representation, showing its main components. Applying a voltage to the comb-drive mechanism produces an attraction force between the inter-digital structure. This force moves the shuttle, and consequently the finger. Therefore, their movements are related. Displacements of the shuttle are amplified by the geometric relation between the finger length and the distance between the hinge and the shuttle-finger connecting point. This way, small comb-drive movements can generate large tip displacements. The suspension mechanism holds the shuttle and gives some structural damping. It is also responsible to bring the system back to its initial position when no voltage is applied. The first resonant modes found for the studied devices in [Boudaoud, 2012] were located
3.1. THE FT-G30 MICRO-GRIPPER
61
between 1000 and 2000 Hz, and were attributed to the comb drive’s suspension mechanism. This work also shown that the finger’s first mode has a much higher frequency (close to 30000 Hz). Therefore, reasonable agreement with experimental data was possible considering a second-order model, with nonlinear, varying stiffness and damping, to model only the suspension behavior. This representation served as a base for the gripper model in air used during this work. When using the vibrometer, the measured displacements for the actuated finger tip depended on the laser beam incident point. Positioning the beam precisely at its extremity is not possible in reality due to the gripper geometry. Furthermore, the incident point in one experiment may not be the same as in a later moment, as relative movements between the sensor and gripper (i.e. during the vibrometer calibration process) will alter it. Figure 3.4 helps to illustrate these obstacles.
30µm
Laser beam
Capacitive sensor
Comb-drive To minimize their effect and ensure that all measurements correctly represent the Figure 3.4: Micro-gripper representation, indisame quantities (the actuated finger tip dis- cating that the laser beam measurement is not placement), a correction process is performed. performed at the finger’s tip and that variations A triangular wave, with slope small enough to on the laser incident point can occur. avoid inducing vibrations in the system, is applied to the gripper until its actuated finger touches the sensor finger. The contact between them can be easily noticed, as the displacement curve slope changes significantly. The obtained contact distance is then compared with the nominal distance between fingers (30µm). This process is repeated every time the working conditions are changed. Figure 3.5 shows an example of acquired data in function of the input voltage, detailing the contact between fingers. In this case, the measured displacement is approximately 25.5 µm, for an input voltage of approximately 68 Volts. The ratio between nominal and measured distance gives a correction factor that approximates the real grippers tip displacement.
26
Displacement (µm)
Displacement (µm)
25
20
15
Contact between fingers
25.5
25
24.5
65
10
66
67
68
69
70
Input voltage (V)
5
0 0
10
20
30
40
50
60
70
Input voltage (V)
Figure 3.5: Experimental data used for estimating the relation between measured and actual tip displacement. The zoom box illustrates the contact between fingers.
62
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
The electrostatic force Felec produced by this comb drive can be approximated by Equation 3.1: Felec = Kelec Vin2
(3.1)
where Kelec is the electrostatic gain of a comb-drive and Vin is the input voltage applied. The electrostatic gain Kelec can be estimated by Equation 3.2:
Kelec =
Na .ξ.hz 2.g
(3.2)
with Na = 992 the total number of fingers in the comb-drive, ξ = 8.85e−12 F/m the dielectric permittivity of the medium, hz = 50µm the comb drive fingers thickness and g = 6µm the gap distance between two comb-drive electrodes [Boudaoud et al., 2013]. As the electrostatic force is not applied to the finger extremity, a correction factor is introduced. It represents the lever effect, and is a ratio between two distances: from the hinge to the point of force application, and from the hinge to the finger tip.
Error (µm)
From Equation 3.1, the produced 0.2 force is expected to show a quadratic relation with the applied voltage. Comparing the 0 obtained displacement curves as functions of inputs voltages (after the correction factor is −0.2 applied) to the ideal function of Vin2 verifies this assumption. Figure 3.6 shows the er- −0.4 ror between the experimental curve and an −0.6 ideal quadratic response. This error is lower than 380 nanometers for all the free movement −0.8 range, and could be explained by fabrication process imperfections and unaccounted forces −1 0 10 20 30 40 50 60 70 Input voltage (V) present in this gripper structure. The error increases rapidly after 68 Volts with the contact Figure 3.6: Difference between experimentally between fingers. measured and theoretical displacements for the micro-gripper electrostatic actuator. These information over the actuation mechanisms on the FT-G30 gripper will be used in the next sections to help modeling and identifying its dynamic behavior in air and vacuum.
3.2/
Identification of actuated finger at atmospheric pressure
This section explains the experimental identification process employed to obtain a model capable of representing the most important dynamics of the micro-gripper actuated finger in air. The proposed model was derived from other authors works on comb-drives, modified to better accommodate particularities of the studied device, and validated experimentally. During the identification process, the influence of electronic input noises related to instruments in the experimental setup used to generate the driving voltage was observed.
3.2. IDENTIFICATION OF ACTUATED FINGER AT ATMOSPHERIC PRESSURE
3.2.1/
63
Experimental identification for the gripper in air
For the identification process in air, the vibrometer and the sample were placed on an antivibration table, and small voltage steps (∆V = 0.5 Volts) were applied at 23 different operation points (V0 = [5...60] Volts). Based on previous works [Boudaoud, 2012, Boudaoud et al., 2013], the gray-box model in air assumes the form shown by Equation 3.3: M x¨ + Cia x˙ + Kia x = Gai Kelec Vin2
(3.3)
where M represents the mass, C a the damping in air, K a the stiffness in air, Kelec the electrostatic gain of the comb drive actuator and Ga the input gain in air (ideally Ga = 1) for each one of the i operation points. The parameter M is considered the same for all the models. The parameter Ga , not present in previous works, is introduced here to compensate for variations in the electrostatic force, as shown in the previous section. The gray-box identification was performed manually for each one of the operation points. This was necessary as automatic methods available in the Matlab® software did not present satisfactory and consistent results estimating the parameters for this low-damped system. This was specially true when considering systems in vacuum, as will be seen later in this chapter, where the acquisition period necessary to capture the whole system dynamics was relatively long (up to 15 or more seconds for a single excitation). Therefore, the parametric identification was performed manually. To obtain the parameters, the experimental response curves, in time and frequency domains (via FFT), were studied. The system model described in Equation 3.3 was simulated considering different parameter values. Each one the them were adjusted individually and iteratively, so model and experimental responses were similar. Firstly, the stiffness Kia was chosen, with aid of the experimental data in frequency domain, to match the first vibration mode frequencies between model and experimental data, so their difference was considered acceptable (less than 1 Hz). Next, Gai was selected, considering the responses in time domain, to drive steady state responses to similar values. In this step, steady state errors lower than 12 nanometers were obtained when comparing experimental and simulation data. Finally, the damping Cia was chosen to closely capture the settling times presented by the system. Once all parameters for one of the operation points were completely identified, they were employed as an initial guesses for the model in a slightly different operation point. During this estimation phase, the stiffness K a and the input gain Ga presented a significant variation on their values over the studied range, while the damping C a remained almost unchanged, therefore being considered constant. Figure 3.7 shows the relation between parameters and the operation point Vin = V0 + ∆V. The variation for the estimated stiffness and input gain was relatively small at low voltages, becoming more important with the increase of the applied tension. From experimental data, up to 25% of variation in the input gain Gai and more than 18% of variation in the stiffness Kia were observed. This variation in stiffness is translated into a shift on the first mode frequency, ranging from 992 Hz to 1090 Hz. The obtained set of models describes the system behavior at specific operation points. Using this information, it was possible to obtain a single general model, capable of representing the actuated finger behavior over all the working range. For this purpose, the sets of parameters Kia and Gai were used to compute fitting functions K a (Vin ) and Ga (Vin ). These continuous functions, 3rd order polynomials computed by the linear least square method, are represented in Figure 3.7 together with the individual values computed for each operation point. Equation 3.4 gives the general model describing the actuation system in air, with stiffness and input gain defined as
64
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
1.8
1.25
1.2
3rd ord. polynomial fit Estimated value (from experiments)
1.75 1.7
Stiffness (N/m)
Input Gain
1.15
1.1
1.05
1
1.65 1.6 1.55 1.5
0.95
0.9
3rd ord. polynomial fit Estimated value (from experiments)
1.45
10
20
30
40
50
1.4
60
10
20
30
40
50
60
Input voltage (V)
Input voltage (V)
(a) Input gain Ga
(b) Stiffness K a −6
5
x 10
Estimated value (from experiments) 4.5
Damping (N.s/m)
4 3.5 3 2.5 2 1.5 1
10
20
30
40
50
60
Input Voltage (V)
(c) Damping C a
Figure 3.7: Variation of model parameters for different input voltages, in air.
polynomial functions of the input voltage:
M x¨ + C a x˙ + K a (Vin )x = Ga (Vin )Kelec Vin2
(3.4)
A validation process was then performed by comparing the general model with a new set of experimental data in frequency and time domains, this time considering large input steps. Figure 3.8 shows the comparison between fast Fourier transforms (FFT) for experimental data and simulated models. The results show a good agreement between simulation and experimental data, with first mode frequency estimation errors smaller than 8 Hz and a steady state displacement errors lower than 110 nm across all range. For input steps larger than 35 V, a second peak can be noticed in the experimental data. In air, its effects are small when compared with the first oscillation mode, and therefore was not included in the model. However, these effects will play an important role when placing the system in vacuum, as will be seen later in this chapter. The validation step demonstrated the accuracy of the proposed general model for the operation range, and therefore it will be used as a base for comparison with the same system in the vacuum.
3.2. IDENTIFICATION OF ACTUATED FINGER AT ATMOSPHERIC PRESSURE
0
0
Experimental data Model
10
−1
−2
−3
10
−4
10
Magnitude (µm)
Magnitude (µm)
Magnitude (µm)
−1
10
10
−2
10
−3
10
−4
10
−5
−5
500
1000
1500
2000
2500
3000
0
−5
10 500
1000
Frequency (Hz)
1500
2000
0
Experimental data Model
−3
10
−2
10
−3
10
2500
(d) 20V 0
1000
1500
2000
3000
0
−3
10
−4
−5
−2
10
−3
10
1500
2000
Frequency (Hz)
(g) 35V
2500
3000
0
−2
10
−3
10
−4
−5
1000
Experimental data Model
10
10
−5
10 500
3000
−1
10
10
2500
10
−4
10
2000
0
Experimental data Model
Magnitude (µm)
Magnitude (µm)
−2
1500
(f) 30V
10
10
1000
Frequency (Hz)
−1
10
500
(e) 25V 10
−1
2500
Frequency (Hz)
0
Experimental data Model
−3
10
−5
500
Frequency (Hz)
10
−2
10
10
0
3000
Experimental data Model
−4
10 2000
3000
10
−5
−5
2500
−1
10
10
2000
10
−4
−4
10
1500
10
Magnitude (µm)
Magnitude (µm)
−2
10
1500
1000
0
−1
1000
500
Frequency (Hz)
10
−1
10
500
0
(c) 15V
10
Experimental data Model
10
Magnitude (µm)
3000
(b) 10V 0
Magnitude (µm)
2500
Frequency (Hz)
(a) 5V
0
−3
10
10
10
0
−2
10
−4
10
10
Experimental data Model
10
−1
10
0
0
Experimental data Model
10
65
10 500
1000
1500
2000
2500
3000
0
500
1000
1500
2000
Frequency (Hz)
Frequency (Hz)
(h) 40V
(i) 45V
2500
3000
Figure 3.8: Comparison between responses for large step inputs (ranging from 5 to 45 V) in frequency domain. Experimental data is shown in blue, and simulated response based on the second order model is shown in red.
3.2.2/
Effects of electronic input disturbance over the electrostatic comb-drive in air
An unexpected phenomenon observed during the characterization stage was the influence of electronic disturbance in the system. These effects presented little visible impact in most of the experiment, becoming evident only when computing the frequency spectrum for a constant input signal. In this case, the electronic noise injected by the power system used to control the gripper (the dSPACE controller and the power amplifier) could be noticed, generating disturbances within the 50 Hz frequency and its harmonics, and resulted in finger tip displacements of a few nanometers. This disturbance signal can be modeled as a series of impulses, with 50 Hz frequency (Figure 3.9(a)). By simulating the system considering the general model in air and the proposed input disturbance signal, it was possible to achieve a close approximation for the results obtained experimentally. Figure 3.9(b) shows the results for the simulated and the real data in frequency domain. Both peaks amplitudes and frequencies are closely matched across the spectrum, including the decaying characteristics around the first vibration mode frequency. This undesired external disturbance serves as an example of the problems faced on real operation situations. This issue will be further analyzed when considering the system in a vacuum
66
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
0.12 −18
Experimental Simulation
10
0.1 −19
Welch’s PSD (m2/Hz)
Input tension (V)
10
0.08
0.06
0.04
−20
10
−21
10
−22
10
0.02 −23
10
0 0
0.05
0.1
0.15
800
900
1000
1100
1200
1300
1400
Frequency (Hz)
Time (s)
(a) Simulated input disturbance (time)
(b) Frequency response (simulated and experimental)
Figure 3.9: Influence of electronic input disturbance on the system in air.
environment.
3.3/
Identification of actuated finger in vacuum
This section describes measurements performed over the micro-gripper inside the SEM, at a high-vacuum in the range of 10−4 Pa, using the vibrometer set-up described in Section 2.2. The identification in vacuum was performed similarly as in air. However, some remarks should be made in consideration to the process. In vacuum, the damping is reduced as there is no fluidic drag to slow down the movement. Consequently, longer settling times and a larger sensitivity to excitations are expected. This effect is shown for the electronic input disturbance. In addition, the existence of non-linear stiffness in this system appears more clearly in vacuum.
3.3.1/
Nonlinear stiffness effects
The consequences of low pressure in the micro-gripper’s behavior are visible in Figure 3.10. On the left, the time-domain data indicates a large reduction in the damping on this environment, with a settling times increasing two orders of magnitude. On the right, its frequency spectrum reveals the presence of other effects on this system, where a second peak, firstly neglected in air due to its low amplitude, became more prominent even at lower voltages. This peak does not represent the second vibration mode of our system, but a geometric non-linearity [Villanueva et al., 2013, Miller, 2005]. This phenomenon, found in literature, can be represented by the Duffing equation (Equation 3.5): x¨ + ι x˙ + κx + βx3 = U
(3.5)
where the variable x represents the position, x˙ the velocity and x¨ the acceleration for this system, and the parameters ι, κ, β are constants for the damping, linear and nonlinear stiffness respectively. The input signal U is usually a sinusoidal signal. This equation describes a damped oscillator where its stiffness does not follow Hooke’s law. The coefficient κ describes the initial linear-elastic deformation, while β represents the stress-strain curve for large displacements. The non-linear term is responsible for deforming the main frequency peak and inducing
3.3. IDENTIFICATION OF ACTUATED FINGER IN VACUUM
67
1
10
18 16
0
10
14
Magnitude (µm)
Displacement (µm)
−1
12 10 8 6 4 2
10
−2
10
−3
10
−4
10
−5
10 0 −2 0
−6
1
2
3
4
5
6
7
8
10
9
0
Time (s)
500
1000
1500
2000
2500
Frequency (Hz)
(a) Time domain
(b) Frequency domain
Figure 3.10: Step response for the micro-gripper in vacuum, for a 40 V step input.
others modes. A positive value of β is named hardening, and a negative, softening non-linearity. Although there is no general analytic solution for Equation 3.5, approximations and numerical solutions can be determined. If the input signal U excites certain frequencies on this system, and if the damping ι is sufficiently small, the system may present multiple solutions. The effects of the non-linear stiffness can be seen in Figure 3.11(a), representing the deformation caused by a negative β in a system with sufficiently small damping coefficient. The red and blue dots represent bifurcation points, where the number of roots in Equation 3.5 changes in function of the frequency. At these points, jumps in magnitude can occur (dashed lines). Figure 3.11(b) shows the experimental frequency response around the first vibration mode for an input step applied to the gripper (red continuous line) compared with a linear second order response (blue dashed line). The experimental curve is bent, similarly to the expected response for a system with a non-linear stiffness, and the presence of bifurcation points can be hypothesized.
Magnitude
Magnitude (µm)
Experimental data Linear model approx.
Frequency (a)
0
10
−1
10
1072
1074
1076
1078
1080
Frequency (Hz)
(b)
Figure 3.11: Duffing equation frequency responses. (a) shows the bifurcations (blue and red dots) for a softening non-linearity (β < 0). (b) shows experimental response for the micro-gripper compared to a linear simulated response.
Although the nonlinear Duffing equation can approximate the phenomena acting over the
68
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
system, this nonlinear identification can be troublesome. The approach based on small displacements used in air is not adequate for this task, and the use of large input steps disfavors the linear parameters estimation. Even as the Duffing equation could be used to represent specific input steps (i.e. 0 to 30 V), the derivation of a single general model for all range is not straightforward. Through polynomial fitting, a good matching capable of correctly capturing the frequency peaks variation was not achieved, with errors as large as 100 Hz. For these reason, a linear representation capable of approximating both vibration peaks was favored. However, the knowledge over the system behavior will be used later in Section 4.3.2, when developing a controller for this system.
3.3.2/
Experimental identification for the gripper in vacuum
Based on the collected data over the system behavior in vacuum, a different model structure was proposed considering a linear 4th order system given by Equation 3.6.  2   M x¨1 + Civ1 x˙1 + Kiv1 x1 = Gv1  i Kelec Vin    2 M x¨2 + Civ2 x˙2 + Kiv2 x2 = Gv2   i Kelec Vin    y = x 1 + x 2
(3.6)
The system consists of two parallel second order models, with independent parameters for damping, stiffness and input gain. The parameters for this model are obtained in two consecutive steps, one for each second order system, where the first equation encompasses the first oscillation mode and the second, the peak generated by the nonlinear stiffness behavior. Using data acquired by applying small voltage steps to the system along its operation range, the parameters in vacuum Civ1 , Kiv1 and Gv1 i were estimated using the process described in the Section 3.3, while M was considered the same as identified in air. The obtained parameters showed similar trends as those obtained for the system at atmospheric pressure. The damping displayed a small variation across the studied operation points, and was considered as a constant value (Figure 3.12(a)), while stiffness Kiv1 and input gain Gv1 i presented significant variations. In the vacuum, the observed frequency range due to stiffness variations was located between 1070 and 1133 Hz. A similar process was performed for the parameters of the non-linear stiffness equation. However, as the phenomenon is only noticeable when considering large displacements, appropriate input signals should be applied. In this case, large input steps were used to excite the system, from 0 V to values between 5 to 50 V. Results showed small variation in the parameters Gv2 and C v2 , then considered constants. However, the stiffness K v2 presented important variations, following a slope tendency similar to K v1 . To encapsulate the obtained family of models into a single one, the varying parameters were also approximated by a polynomial function of the input Vin . The results are presented in Figures 3.12(b) and 3.12(c), together with the corresponding polynomial function, obtained through the linear least mean square method. Figure 3.12(d) shows the obtained experimental values for K v2 and its fit function, capable of capturing the slope tendency of K v1 , its expected behavior, even if only limited number of measurement points were used. The system of Equations 3.7 gives the final model for the micro-gripper’s actuated finger operating in vacuum, where the three parameters presenting significant variation across the operation range were considered as functions of the input voltage Vin : the stiffness K v1 and K v2 , and the input gain Gv1 .
3.3. IDENTIFICATION OF ACTUATED FINGER IN VACUUM
69
−8
5
x 10
1.3 Estimated value (from experiments)
3rd ord. polynomial fit Estimated value (from experiments)
4.5 1.25
3.5
Gain input
Damping (N.s/m)
4
3 2.5
1.2
1.15
2 1.1 1.5 1
10
20
30
40
50
1.05
60
10
20
30
40
50
60
Input voltage (V)
Input Voltage (V)
(a) Damping C v1
(b) Input gain Gv1 7.4
1.9 3rd ord. polynomial fit Estimated value (from experiments)
3rd ord. polynomial fit Estimated value (from experiments)
7.3
1.85
Stiffness (N/m)
Stiffness (N/m)
7.2
1.8
1.75
7.1 7 6.9 6.8
1.7 6.7 1.65
10
20
30
40
Input voltage (V)
(c) Stiffness K v1
50
60
6.6
10
20
30
40
50
60
Input voltage (V)
(d) Stiffness K v2
Figure 3.12: Variation of model parameters for different input voltages, in vacuum.
   M x¨1 + C v1 x˙1 + K v1 (Vin )x1 = Gv1 (Vin )Kelec Vin2     M x¨2 + C v2 x˙2 + K v2 (Vin )x2 = Gv2 Kelec Vin2      y = x 1 + x 2
(3.7)
The validation process was performed by comparing its results with experimental data obtained when applying large input steps. A comparison of experimental and simulated curves is shown in Figure 3.13. Overall, the frequency estimation error was smaller than 4 Hz for the first vibration mode and 5 Hz for the second peak, while steady state errors were lower than 400 nanometers across the measured operation points. The smaller error in frequency estimation is due to the sharper peak present on its frequency spectrum, facilitating the definition of its value in relation to air measurements. Conversely, the constant oscillation present in the time-domain data helpd to increase the errors for steady state estimations. The forth order linear model reveals to be sufficiently precise to estimate, in frequency and time domains, the oscillatory characteristics for this system. Furthermore, the development of linear control methods is straightforward, and will be explored in the next chapter. Still, the obtained results were less accurate than those for the system in air, as demonstrated by the higher steady state error. One of the difficulties presented in this process was related to the experimental acquisition of data. The reduced damping makes the system sensible to external disturbances, and
70
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
Experimental data Model
Magnitude (µm)
Magnitude (µm)
Experimental data Model
0
10
−2
10
−4
10
−6
−2
10
−4
10
−6
10
500
1000
1500
2000
2500
3000
0
500
1000
(a) 5V Experimental data Model
−4
10
1000
1500
2000
2500
−2
10
−4
10
Experimental data Model
500
−4
10
−6
1000
1500
2000
2500
1000
1500
2000
Frequency (Hz)
(g) 35V
2500
3000
−2
−4
10
3000
0
1000
1500
2000
(f) 30V Experimental data Model
0
−2
10
−4
10
0
500
Frequency (Hz)
2500
3000
Experimental data Model
0
10
−2
10
−4
10
−6
10
500
3000
10
Frequency (Hz)
−6
10
2500
Experimental data Model
0
(e) 25V 10
Magnitude (µm)
−2
10
2000
10
Magnitude (µm)
0
1500
−6
(d) 20V 10
1000
10
0
3000
500
(c) 15V
10
500
0
(b) 10V Experimental data Model
Frequency (Hz)
Magnitude (µm)
3000
−6
−6
0
2500
Frequency (Hz)
0
10
0
2000
10
Magnitude (µm)
Magnitude (µm)
−2
10
1500
Frequency (Hz)
Magnitude (µm)
0
−4
10
10
Frequency (Hz)
10
−2
10
−6
10
0
Experimental data Model
0
10
Magnitude (µm)
0
10
10
500
1000
1500
2000
2500
3000
0
500
1000
1500
2000
Frequency (Hz)
Frequency (Hz)
(h) 40V
(i) 45V
2500
3000
Figure 3.13: Comparison between responses for large step inputs (ranging from 5 to 45 V) in frequency domain, considering the gripper in vacuum. Experimental data is shown in blue, and simulated response based on the second order model is shown in red. the precise data acquisition becomes more critical and time consuming.
3.3.3/
Effects of electronic input disturbances over the electrostatic comb-drive in vacuum
While operating in air, electronic input disturbances were attenuated by the added environmental damping, only noticeable under certain conditions. This is not the case the case in vacuum, where its effects gain importance and become an issue in many cases. These disturbances are easily visible in the frequency spectrum, as shown in Figure 3.14(a). This curve illustrates the Welch’s PSD averaged from five measurements for a constant input of 10 V, in air and vacuum. This result indicates that the phenomenon has a much larger influence in at lower pressures, specially around the first vibration mode. However, while this effect was limited to a few nanometers in air, the low damping aggravate its impact over in vacuum. Its influence over the gripper’s behavior in this environment can be seen in Figure 3.14(b), where for a given operation point (in this case, around 53 V or 18.7µm), the comb-drive first mode matched one of the disturbance frequencies (in this case, 1050 Hz), producing large oscillations. In this example, vibrations up to 200 nanometers peak-to-peak
3.4. RESUME OF DIFFERENCES BETWEEN AIR AND VACUUM
−18
10
19
Gripper in vacuum Gripper in air
18.9
−19
10
Displacement (nm)
Welch’s PSD (m2/Hz)
71
−20
10
−21
10
18.8 18.7 18.6 18.5
−22
10
18.4 −23
10
500
600
700
800
900
1000
1100
Frequency (Hz)
(a) Frequency response
1200
1300
1400
18.3 0
2
4
6
8
10
Time (s)
(b) Time response (vacuum)
Figure 3.14: Effects of electronic input noise on the system.
were observed. These values are large in relation to the desired accuracies for this work, and imposed additional difficulties during the identification process. Although this problem was directly related to the hardware implementation of the setup, changes in the equipment were not possible at the time. This issue will be addressed in the next chapter, where the controllers should take this disturbance into account to effectively attenuate it .
3.4/
Resume of differences between air and vacuum
Through the results obtained during the identification phase, the influence of pressure variation on the behavior of the tested micro-gripper is undeniable. In this work, the differences lie on a large damping variation, shifts in the resonant frequencies and on a more noticeable non-linear effects. The experimentally identified parameters are given in Table 3.1. In [Chen et al., 2003], the relation between comb-drive parameters (thickness, electrodes separation, and others) was studied through numerical simulations, considering the flow and squeeze film generated when operating in a fluid (normally air). It concluded that, in this environment, the largest contribution to the comb-drive energy dissipation is due to the compression of fluid between fingers. Accordingly to its results, the quality factor of MEMS in air and vacuum can present variations up to 3 orders of magnitude, depending ov various construction parameters [Nguyen, 1995]. The effects of gas damping can be investigated by dividing it in three regions (also called flow regimes). In the first, the pressure influence is small enough so gas damping is negligible when compared to other sources. In this region, most of the energy dissipation is related and dependent on the surface-volume ratio, the material and surface characteristics, and the gas slips with respect to the surface. A second region is defined in the opposing extreme, where pressures are high enough so gas molecules often collide with each other. In this case, the gas viscous damping is the most important dissipation mechanism, inducing a temporary mass increase in the system. The third region represents an intermediate scenario, where molecules do not interact with each other, but are still present in enough amounts so when they collide with the system, moment is transferred between them [Ho et al., 1998]. These different regions can be formally classified with the aid of the Knudsen number Kn
72
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
Parameters
Environment Air Vacuum
M
K 00031
K 00032
G00031
G00032
C 00031
C 00032
3.7e−8 3.7e−8
[1.44 − 1.74] [1.674 − 1.875]
[6.69 − 6.91]
[0.95 − 1.233] [1.075 − 1.287]
4e−2
3e−6 2e−8
2e−6
Table 3.1: Summary of the identified parameter for the micro-gripper in different working environments.
[Mertens et al., 2003], defined by Equation 3.8: Kn =
λ 1 = w Dηw
(3.8)
where λ is the mean free path the gas molecule can move before colliding with another, w is the gas layer motion width, D the gas number density and η the collision cross section of the gas molecule. The three regions described above are classified and named as: free molecule regime (Kn > 10), transition regime (10 > Kn > 0.01) and viscous regime (Kn < 0.01). If Kn 001c 1, the gas obeys the Navier-Stokes equations of hydrodynamics, while a Kn 001d 1 indicates a system following the rarefied gas dynamics theory. The precise computation of the Knudsen number for complex shapes, like the combdrive structure, can become a laborious exercise. From experimental results found in literature [Mertens et al., 2003, Sandberg et al., 2005, Sumali et al., 2008] the free molecule regime is attained for different micro-structures at pressures between 101 and 102 Pa, values several orders of magnitude larger than the ones usually found in SEMs. Therefore, we assumed these systems as operating in the free molecular regime, where the environment influence on its damping is negligible. This large reduction in damping is traduced into an increase in the settling time, from 0.1 second in air to more than 10 seconds in vacuum. Figure 3.15(a) demonstrates this effect comparing both responses together, where the absence of air to brake the oscillation results in a sustained vibration for a longer period, while maintaining similar overshoot and steady state values for both environments. Due to the low damping, the non-linearities are more discernible in vacuum. During this work, the observed effects of a non-linear stiffness resulted in the intensification of a second oscillation peak, with frequency two times the comb-drive first vibration mode. In addition, the first mode appears to be bent, indicating the presence of a softening non-linearity. Figure 3.15(b) shows a comparison between step responses for both environments in the frequency domain. Finally, the variation in stiffness is studied. Figure 3.16 shows the values related to the first vibration mode obtained for both environments. In the literature, there are analytic models to describe the influence of pressure over the frequency mode of simple structures, such as cantilevers [Sandberg et al., 2005]. For a cantilever of thickness t and width w, these phenomena can be approximated by Equation 3.9 : fgas = fvac
πpwMgas 1+ 4RT ρt
!−1/2 (3.9)
3.4. RESUME OF DIFFERENCES BETWEEN AIR AND VACUUM
73
1
10 Vacuum Air
Peak vacuum
1
0
0.8
Magnitude (µm)
Displacement (µm)
10
0.6
0.4
0.2
Peak air −1
10
−2
10
−3
10
Experiment in vacuum 0
Experiment in air −4
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
10
1.8
Time (s)
0
500
1000
1500
2000
2500
Frequency (Hz)
(a) Time domain
(b) Frequency domain
Figure 3.15: Response comparison for micro-gripper step in air (blue) and vacuum (red).
Experiment − air Polynomial fit − air Experimental − vacuum Polynomial fit − vacuum
1.85
Stifness K (N/m)
1.8 1.75 1.7 1.65 1.6 1.55 1.5 1.45 1.4
10
20
30
40
50
60
Input Voltage (V)
Figure 3.16: Stiffness comparison for gripper in different environments, together with its corresponding polynomial fit functions.
with Mgas the molar mass of the gas, p the environment pressure, R the gas constant, T the absolute temperature, w and t the cantilever width and thickness, respectively, and ρ the material density. Experimental results found in the literature shown a reasonable agreement with the theory for single and multi-layered cantilevers. However, the precise frequency shift estimation for more complex geometries is still a challenge, relying on numerical simulations. The measured frequency shift due to pressure variations for samples with simple geometry (single and V-shaped cantilevers) was smaller than 4% in the studied references [Sandberg et al., 2005, Sumali et al., 2008, Epp et al., 2004, Zhang et al., 2005, Sader et al., 1995], while the obtained values for the comb-drive in this work ranged between 3.8% and 8.2%. This can be explained by the large contact surface between the comb-drive mechanism when compared with other cantilever structure (larger surface/volume ratio). Overall, the comparison between obtained models, in air and vacuum, highlights the difficulties in estimating the parameter variations due to pressure changes, where experimental methods are still the most spread method for obtaining reliable results.
74
3.5/
CHAPTER 3. MICRO-GRIPPER IDENTIFICATION
Conclusion
To obtain the dynamic model for the micro-gripper actuated finger, a gray-box model approach was used, where the parameters for a pre-defined dynamic equation were found. The identification process, performed in air and vacuum, showed how pressure influences its dynamic behavior, modifying its damping, stiffness and exposing non-linearities, present in vacuum and usually hidden when operating in air. To obtain models capable of capturing the most important characteristics of the system in each environment, parameters were identified at different operation points. These parameters were then employed to develop a general model for each pressure condition capable of representing the system in all its operation range, which were experimentally validated. The two general models encompass the most important system dynamics in both environments (main vibration modes, parameter variation in the working range, steady state responses) and can be used, together with information about disturbance obtained in the previous chapter, in the synthesis and simulation of controllers. These results emphasize the importance of considering different models and parameters when operating in a vacuum environments. Although the literature offers general clues about what to expect, their precise system behavior requires in dept analysis, often recurring to experimental observation. The data acquisition process in vacuum can be challenging, even when using strategies such as these applied in this chapter, and even more when methods base on image analysis are used. Due to the especially low damping shown in vacuum, external perturbation effects are much more noticeable and require additional attention. In this work, electronic perturbations injected by the circuit responsible for generating the input voltage required a greater effort during data acquisition and processing phases. This unexpected disturbance represents a real issue for the positioning of the studied micro-gripper inside a SEM, and will be addressed during the controller development.
4 Position control in the micro-world
The precise positioning of structures with micro and nano-metric accuracies is currently a challenge restricting the grasping, handling and characterization of fragile small samples. As presented in early chapters, the limiting issues on precise position control in the micro-world include difficulties in instrumentation, reducing our ability to measure dynamic behaviors and disturbances. This work employs different robust control strategies to explore the end-effector dynamic aspects, taking advantage of the available experimental setup. The positional information supplied by the vibrometer was used to correct, in real time, undesired gripper behaviors, leading to a significant improvement in its response characteristics. This chapter describes strategies for positioning stages and micro/nano-manipulators, highlighting their strengths and limitations. Next, the control problem inside the SEM is addressed, offering some clues about the specific challenge of operating in vacuum. Based on this information, two control strategies were selected and developed for the considered micro-gripper. Finally, simulation and experimental results are compared, validating these strategies and concluding about their strengths and weaknesses .
4.1/
Control strategies for nano-positioning
All systems, and more particularly micro-systems, are subject to uncertainties. This means the output of a real system cannot be exactly predicted, and may arise from two sources: inputs (i.e. noises and disturbances) and non-considered dynamics. A model should be able to represent the input-output relation in a system, so that a resulting controller is able to impose a desired performance in closed loop to the real system. As the model cannot be fully known, uncertainties will be present, thus resulting in a degradation of the real system performance. However, effects of disturbances and uncertainties can be taken in consideration up to some extent through effective modeling, analysis and design techniques. Robustness is achieved when the system response is insensitive to these effects. In the past years, different methods were developed and applied for researches to improve response characteristics of positioning systems. Figure 4.1 gives an overview of the different approaches found in the literature when considering control strategies applied for micro and nano-positioning of several devices (stages, cantilevers, grippers and piezotubes). This survey deals mostly with systems operating in air, and classifies the control methods in: proportional (P)-integral (I)-derivative (D), H∞ methods, sliding modes based approaches (SM), model predictive control (MPC), open-loop control (OL), observer based control and others. It is important to remark that two or more of the strategies/control structures mentioned above can be applied simultaneously in the same controller. 75
76
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
P/PI/PID
[Li et al., 2011] [Jasper et al., 2010] [Zhang et al., 2013a]
[Sitti et al., 2000]
[Tang et al., 2014] [Rakotondrabe, 2013] [Yi et al., 2009]
Observer based OL
[Rakotondrabe et al., 2010] [Rana et al., 2014] [Xu et al., 2012]
MPC
[Xu, 2013] [Ghafarirad et al., 2012] [Huang et al., 2009] [Xu, 2015]
SM
[Li et al., 2012] [Aphale et al., 2007, Pao et al., 2007, Boukhnifer et al., 2007] [Salapaka et al., 2002] [Liu et al., 2013]
H∞
[Grossard et al., 2011] [Fleming et al., 2010]
Others
[Kim et al., 2003] [Dong et al., 2001] [Nakabayashi et al., 2006] [Yu et al., 1999]
Not informed
1995
2000
2005
2010
[Polyakov et al., 2014]
2015
Figure 4.1: Summary on several published works considering nano-positioning control in air and vacuum, accordingly to the main strategy employed and year of publication.
• P/PI/PID: These error-based controllers are popular, with various applications in commercial devices. Their simplicity and intuitive parameter selection are major aspects making these methods attractive. In [Li et al., 2011], a PID controller was combined with an hysteresis model to control a micro-positioning stage, showing how the pre-compensation of this non-linearity can greatly improve the system response. In [Zhang et al., 2013a], two PID controllers were applied to regulate the positioning dynamics of a piezomotor. The controllers were switched depending on the desired operation mode (fine or coarse movement). • Observer based control: This category includes disturbance observers (DO) [Yi et al., 2009], unknown input observers (UIO) [Rakotondrabe, 2013] and extended state observer (ESO)[Tang et al., 2014]. These strategies rely on estimating external disturbances in realtime based on output measurements. By modifying the control signal using this estimation, an effect similar to a disturbance feed-forward controller can be obtained. This attenuation strategy is considered proactive, while methods based on the regulation of errors, such as PID, are considered reactive and present in general slower responses to perturbations. While DO and UIO rely on the knowledge over disturbances and plant dynamics, ESO makes less restrictive assumptions as it estimates states, disturbances and plant uncertainties simultaneously. This reduced dependency on plant models, and the potential to efficiently reject disturbances are attractive properties for practical control implementations. As these are only observers, a feedback controller is still required. • Open-loop: The use of open loop control is discussed in [Borovic et al., 2005], together with methods such as input shaping and inverse models to compensate hysteresis and other undesired effects [Rakotondrabe et al., 2010]. It was argued that, if a good model for the system is available, the control problem could be simplified though the use of open loop control methods, thus reducing the need for instrumentation.
4.1. CONTROL STRATEGIES FOR NANO-POSITIONING
77
• Model predictive control: The MPC controller relies on a constant on-line optimization process to find and simulate future plant outputs. In [Rana et al., 2014], a MPC based position controller for an AFM positioning stage was developed, aiming to reduce the piezoelectric actuator non-linearities and improve its scan speed. In the paper, it was shown how the MPC strategy can improve the stage’s performance, producing faster and more accurate responses when compared to traditional proportional-integral controllers, often used for controlling this kind of device. However, this process requires constant optimization and considerable computational power, and may not be applicable for systems with fast dynamics. • Sliding mode: This method is based on altering the dynamics of a system by applying a discontinuous, switching control signal. It presents naturally robust characteristics, with low sensitivity to parameter variations and disturbances. In [Huang et al., 2009] a mixed sliding mode/fuzzy controller for a long range nano-positioning mechanism was described. The sliding surface was used to modify the fuzzy controller gains, reducing the chattering, characteristic for this kind of controller, while presenting a small non-zero tracking error. In [Ghafarirad et al., 2012] the control of a multi-axial piezoelectric actuator was described. In this article, the effects of hysteresis present in this kind of actuator were precompensated through an experimentally estimated inverse model. To further reduce errors and non-modeled dynamic effects, a sliding mode structure coupled with a neural network was proposed. A discrete sliding mode coupled with a force observer and impedance filter, to control force and position of a micro-gripper under effect of external disturbances and non-linearities, was proposed in [Xu, 2013]. The disturbances and non-linearities were estimated as a lumped parameter, based on the system’s nominal model. The authors remark that, in cases of sufficiently large uncertainties and disturbances, additional control terms should be added to ensure the closed loop convergence to the sliding surface. One of the drawbacks for SM based controllers is the generated control signal, introducing chattering on the system and degrading its response. Furthermore, this signal can be harmful to actuators. Methods to reduce the chattering exist and can be implemented at the cost of complexity and added tracking error [Xu, 2015]. One of the alternative methods to reduce chattering is presented in [Xu et al., 2012], where a model predictive controller was included in the loop. • H∞ : This class of controller relies on translating design requirements into a small-gain robust control optimization problem to be solved. In [Salapaka et al., 2002], this method was applied to improve the bandwidth and precision of nano-positioning device actuated by piezoelectric stacks over standard proportional-integral approaches, while increasing the system robustness to non-linearities and uncertainties. In [Aphale et al., 2007], a similar control structure has been applied to reduce the resonant peak amplitude for a piezo-tube. Displacements were estimated based on external shunt measurements and did not require a dedicated displacement sensor. Other control structures were explored in [Pao et al., 2007, Li et al., 2012, Liu et al., 2013], with the addition of filters and feedforward elements to improve response times and disturbance attenuation characteristics. The challenge for this method lies on the selection of adequate weighting functions to build the optimization problem, what depends on knowledge about the plant, uncertainties and disturbance. • Others: Different structures can be found in the literature not belonging to any of the described categories. In [Fleming et al., 2010], a linear structure called Integral Resonant Control was used to damp multiple resonant modes of a scanning probe microscope, improving its bandwidth and stability. In [Grossard et al., 2011], a mixed High Authority Control/Low
78
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
Authority Control strategy was presented. It consisted of a two degrees-of-freedom nested controller, where the inner loop damping to the system while the outer loop focused on robustness and stability issues. • Not Informed: In some cases, no information about the controller was given (i.e. the micromanipulator manufacturer provided the controller). In general, these articles described practical micro- and nano-manipulation tasks, were their goals (biological material manipulation, characterization of nano-tubes) required high-precision. The presented survey offers a general idea about the development of control strategies for precise positioning devices. In recent years, researchers proposed different strategies, indicating that this particular problem is still open. The controller selection should take into consideration the specificities of precise positioning in SEMs. In the following subsection, a few examples of positioning task in this environment are briefly described.
4.1.1/
Examples of positioning tasks inside the SEM
Micro- and nano-manipulation inside a SEM was reported for the first time in the 1990’s in Japan with the handling of micro-objects (less than 100 µm). It consisted of two manipulators and a working table, producing several translational and rotational degrees of freedom, driven by ultra-sonic motors for coarse and piezoelectric actuators for fine motions [Mazerolle et al., 2005]. Since then, several research groups focused their work on improving the robotic and the automation aspect of manipulation tasks, trying to achieve higher positioning resolutions. However, only a limited success was obtained. While specific cases of automation were implemented, the large majority of tasks is still being performed by human operators. In addition, closed-loop methods for controlling robotic structures rely on SEM imaging and internal sensing capabilities methods, often disregarding fast dynamic effects. In [Yu et al., 1999], a custom build 3 DoF stage for manipulation and characterization of carbon nanotubes with approximately 50 nm thickness was proposed, using linear piezoelectric actuators and piezo-tubes to generate motion. It described the delicate operations required for separating individual samples from the substrate. These processes, made by an operator, were based on SEM image feedback while the stage was controlled by incremental steps. The authors noticed that the distance traveled by one-step displacements was not constant. Depending on the direction of movement (forward or backward), displacements between 4 and 12 nanometers per step could be produced, with occasional jumps of approximately 30 nm. The settling time for this construction in vacuum was estimated to be around 3 seconds. Similarly, the work of [Dong et al., 2002] developed a 10 DoF manipulator for carbon nanotubes with diameter varying between 30 to 50 nm. In this case, picomotors were used for coarse and piezoelectric actuators for fine motions, resulting in a system with nominal accuracy near one nanometer. The added degrees of freedom improved the manipulation capabilities of the setup. Both articles stated that, despite SEMs at the time being able to achieve image resolutions under 6 nm with slow scan speeds (improved signal/noise ratio), real-time manipulation requires faster acquisition speeds. This penalizes the image quality, resulting in effective achievable resolutions between 20 and 30 nanometers, further decreasing the positioning accuracy of tasks depending on image feedback. In [Saeidpourazar et al., 2008], the effects of non-linearities and uncertainties limiting the dynamics of micro-manipulators inside SEM were put in evidence when using only image feedback. In this work, a commercial nano-manipulator was adopted, possessing 3 DoF and nominal precisions close to 1 nanometer. However, it did not include any position or velocity sensor. Move-
4.1. CONTROL STRATEGIES FOR NANO-POSITIONING
79
ment steps were applied in open loop, and relied on external feedback (i.e. image from the SEM) to determine its position. To improve the system, a force sensor was attached to the manipulator extremity in order to capture the interaction forces between samples and the sensing tip. With this information, and by modeling the interaction forces between both components, it was possible to estimate their separating distance. By alternating the use of position information from image and force sensor (as proposed in [Zhou et al., 1998] for a micro-manipulator in air), the closed loop control was performed. To overcome the uncertainties and non-linearities existing in the manipulator, a modified sliding mode controller was implemented. This closed loop system was then simulated, producing a faster, non-oscillatory response for step reference tracking when compared with PID controllers, even under the effects of measurement noise and small errors in the system. In [Jasper et al., 2010], a piezo-actuated micro mobile robot to operate inside the SEM vacuum chamber was described. In this work, its position was controlled by performing line scans with the electron beam tracking a specific pattern in the robot. Using only line scans instead of full images improved the acquisition speed, capable of obtaining position data with frequencies up to 1 kHz, a large improvement from standard image acquisition and processing strategies, limited to values near 50 Hz in best case scenarios. This sensing strategy was used together with a proportional feedback gain, selected experimentally, to control the mobile micro-robot displacements with precision close to 20 nm. However, later studies shown carbon deposits around the scanned pattern after long exposures times, degrading its accuracy [Dahmen et al., 2013]. More recently, [Zhang et al., 2013a] described the development of a manipulation platform composed of two nano-manipulators, each one with 3 DoF, mounted and dismounted from a SEM through the specimen exchange chamber. This feature avoided breaking the main chamber vacuum, reducing contamination and simplifying the exchange of end-effectors. The manipulator displacements were measured through optical encoders, embedded in the actuators, with nominal resolution of 1 nanometer. During the system characterization, an encoder drift of approximately 1.2 nm/s was measured, together with an averaged precision of 98% over the traveled distance. PID controllers were developed for coarse and fine motions, and a combination of internal sensor and image servicing allowed to refine and obtain a reasonably accurate and fast positioning (under 1 second). Another example illustrating the importance of precise position control in a SEM environment is found in [Abrahamians et al., 2014]. It describes the stiffness characterization of thin (200 nanometers) suspended membranes through the use of a vibrating tip coupled to a tuning fork. To reduce failures, sensory information from the robotic system were used to update in real-time a virtual 3D model of the SEM’s interior, including positioning stage and manipulators, offering an interesting complement to the SEM image for position control and task planning. This allowed to improve the efficiency of precise positioning, specially in the Z axis. The authors commented that the probe approach phase needed to be done through nanometric steps to increase the success rate. Their remarks extended to future works when, considering even mode delicate samples, sub-nanometric accuracies may be required and the effects of disturbances, uncertainties in the manipulator motion and other effects must be considered. The available literature focusing on the positioning problem inside the SEM or in the vacuum environment is limited. The examples described above offer some insight about the specific challenges of micro- and nano-manipulation inside the scanning electron microscope, summarized next:
1.
The global image produced by the SEM is one of the main sources for position feedback information. Due to its low frame rate, it limits position control to static behaviors.
80
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
2.
The use of additional sensors to measure positions, velocities and forces are still limited. Furthermore, sensors and actuators may behave in unexpected ways in vacuum environment (i.e. due to the poor heat transfer).
3.
Disturbances and uncertainties will disrupt the operation of components, reducing its effective performance.
4.
Knowledge about the system behavior in vacuum is rarely explicitly considered on the applications.
4.1.2/
Conclusion
Through inspecting the literature related to position control of micro and nanomanipulators, positioners and stages, it was possible to identify certain challenges present on the development of high-performance controllers for the SEM environment. One of the most important issues is the difficulty to measure the dynamics of end-effectors position inside the SEM. A great amount of studies for positioning of end-effectors and other robotic structures were based on SEM imaging methods, and therefore limited to static and quasistatic position control. The use of image to supply feedback information for those devices, despite popular, faces a strong limitation due to its acquisition speed. By introducing other methods for position measurement and estimation, it was possible to improve overall response characteristics when positioning devices in SEM. However, as researchers shown, internal sensors can have their performance degraded in vacuum, and the characterization of components in this environment is necessary. Due to difficulties in obtaining accurate real-time positional information for those structures, disturbance effects were rarely considered and constrained to low frequencies/static domain. Additionally, system uncertainties and variations on their dynamics were often overlooked or roughly estimated. When considering control strategies for positioning, a wide range of different methods can be found in literature. Although most of the reviewed research deals with tasks at atmospheric pressure and the use of different sensing methods, it offers a rich panorama over the tools to be used for position control. By using the information obtained in previous chapters about disturbances acting in the SEM environment, together with a reasonably accurate dynamic model for the system, the design of control laws better adapted for this problem can be envisioned. By taking advantage of the existing experimental setup and its precise, fast, real-time data acquisition capabilities, advances in the end-effector dynamic positioning can be foreseen. It would be valuable to quantify the improvements brought by the use of specially conceived control laws to nano-manipulation.
4.2/
Robust position control strategies for a micro-gripper in vacuum
In the previous section, various control strategies explored for position control in the micro-world were briefly presented. An essential step in control design is to identify the strategies better suited for the problem. In this particular case, the controller should present robustness to uncertainties and parametric variations, while being able to attenuate effects of disturbances. Taking in consideration these goals, open-loop control and classic P/PI/PID controllers would require considerable efforts to achieve good performances and robustness characteristics
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
81
in all operation range. In addition, open-loop control is not capable of attenuating disturbances that, as seen in the previous chapters, are an important issue when operating in vacuum. Model predictive control offers an interesting alternative, and the optimization control problem can be formulated taking directly into consideration specific design objectives (i.e. control signal saturation, limitations on the system velocity, model uncertainties, ...). However, as it requires the online resolution of an optimization problem, the computational overhead may limit, at the current time, its practical implementation to systems with slower dynamics. Both sliding mode and H∞ methods can present good robustness characteristics in the face of disturbed and uncertain plants, and shown considerable performance improvements in the reviewed references over classic PID methods. Both methods, in their standard forms, were developed assuming reasonable well-known system models. The H∞ framework offers more tools when selecting desired response characteristics, an interesting feature for this application. Furthermore, the intrinsic chattering phenomenon present in the sliding mode approach requires the use of more complex solutions, such as fuzzy or adaptive-sliding mode [Gu et al., 2014]. An alternative control strategy is the use of observer structures to estimate undesired effects in the system. This category includes Disturbance Observers (DO), Unknown Input Observers (UIO) and Extended State Observers (ESO), among others. A perturbation in the system can be estimated in real-time from the inputs and outputs. This estimation can be then used to directly attenuate its effects, similarly to a disturbance feed-forward strategy. As these are only estimators, they need to be used together with feedback control methods. Based on this information, two strategies were selected: H∞ and Extended State Observers. They represent two different philosophies on the controller synthesis. The H∞ controller aims to obtain a robust controller, given that accurate information about the plant, its uncertainty range, the incident disturbance and its characteristics are available. This knowledge is then applied in the construction of an optimization problem, specific for this scenario. The resulting controller is obtained in the minimization of this problem. This strategy also allows a greater control over the requirements, the disturbances and the uncertainties to be taken into consideration, as these can be introduced separately into the problem formulation via the use of separated weighting function, as will be seen later in this chapter. The second proposed controller was based on the Extended State Observer. This estimator aims to identify, in real-time, the effects of disturbances acting upon a system, as well as to counter them. Contrary to the H∞ method, and to other perturbation observers such as DO and UIO, the ESO aims to reduce the need of accurate models, acknowledging that those are not often accessible or easy to obtain in many practical applications. This strategy is made possible by introducing a virtual state, called total disturbance, into the system approximate model. This total disturbance contains all external, unknown disturbance, together with plant uncertainties. The core idea consists of recognizing both disturbance and uncertainties as undesired quantities to be eliminated and, therefore, their separation into individual components, as often seen in other control strategies, is not necessary. By estimating this virtual state, it is possible to reduce disturbance effects and to improve the system robustness to uncertainties at the same time. Both control strategies present relevant characteristics to solve the robust position control problem. This work took advantage of the available setup to obtain displacement measurements with nanometric accuracy and fast sample rates inside the SEM vacuum chamber. The setup, applied for identifying the disturbances inside the SEM and micro-gripper dynamics in vacuum, was used to perform real-time, dynamic closed-loop control, where these control strategies could be tested. This section reviews the H∞ theory and the associated optimization problem formulation, and introduces the ESO structure and the observer gain selection, where high-gain strategy was
82
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
chosen. Before starting the controller development, a short review on linear fraction transformation (LFT) and the representations of uncertain systems is presented, summarizing how these tools are used in the context of robust control analysis and synthesis.
4.2.1/
Uncertain systems and the LFT representation
In practice, all models are uncertain representations of a real system, where uncertainty is the price to be paid in exchange of reduced complexity. These uncertainties can comprise unknown/neglected dynamics and parametric variations. An accurate model is capable of correctly mapping the plants input/output set, for all possible cases. If the selected model is too simple, this assumption may be not valid for all input/output pairs. However, a set of simple models may contain enough elements to model complex behaviors. The LFT representation is widely used to express uncertainties and to study its effects on closed loop systems. This structure is briefly introduced, and additional considerations can be found in Appendix B. The standard form for this representation is given by Figure 4.2.
d
v w
z
G0 y
u
M
K
Figure 4.2: General LFT representation, containing the nominal system G0 , the uncertainty block ∆ and the controller K. During robustness analysis, the nominal plant and controller are considered as a single system M.
This representation assumes a nominal plant model G0 , where multiple dynamic uncertainties affecting it can be lumped into a single perturbation block ∆. This block representation is usually refereed as the system’s uncertainty. Assuming its dynamics are not perfectly known, but norm-bounded, it is possible to represent the block ∆ by a transfer function, describing its limits that are assumed to be known. This transfer function follows a relation σ(δ( jω)) ≥ δ( jω), with σ the supreme of the singular value and δ a known function. Uncertainties are often used to represent non-modeled time delays, high-order dynamics, hysteresis and other non-linearities. When uncertainty comes from inaccurate description of parameters, another representation can be used, to model variations or uncertainties of some elements inside a possible bounded value range (i.e. a = a0 + r∆, where the parameter a is described by a nominal value a0 and a variation range r regulated by the uncertainty ∆ < 1). This can be used to represent the effects of wearing, deterioration or other alterations (i.e. due to temperature variation) present in a system.
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
83
If G0 is a matrix where each element is a transfer function between its inputs and outputs signals, the LFT becomes the closed loop transfer between w and z under the feedback effects of K (a controller) and ∆ (the system uncertainties). These connections are called Fl and Fu (lower and upper LFT), and defined in Appendix B. A common notation considers the lower feedback system as a nominal closed loop M, that is interconnected to the uncertain block ∆ by the feedback signals v and d, and the transfer map from external signals w to z can be written as: z = Fl (Fu (P, ∆), K)w = Fu (Fl (P, K), ∆)w
(4.1)
= Fu (M, ∆)w This structure, called the M −∆ representation, establishes a framework to perform stability and performance robustness analysis, through the use of another tool: the Structured Singular Value, discussed next.
4.2.2/
µ analysis - a tool for robust stability and performance evaluation
One of the most popular methods to analyze the robustness of uncertain time-invariant systems is based on the analysis of the singular value σ. This value, computed as the square root of the the eigenvalues of a matrix multiplied by its transpose conjugate, is a measure of the system magnification at each frequency and its maximum value can give some information about the stability and robustness. However, this metric revealed to be conservative, as the only assumption made about the uncertainty ∆ was its boundedness. To improve this result, the structured singular value (SSV), also denoted µ, was introduced. The analysis based on the SSV aims to refine the small-gain condition, that is based on σ, through the imposition of a structure to the perturbation [Doyle, 1982]. Before introducing this tool, some concepts are defined. Definition 1: Infinity norm The ∞-norm, also known as H∞ norm, maximum norm or uniform norm, of a vector v(·) is denoted v∞ and defined as the maximum of the absolute value of their components. v ∞ := sup v(·)
(4.2)
The ∞-norm of a linear, time-invariant, causal system G(s), is the peak value on a bode magnitude plot of G( jω), or the maximum of its singular value across the frequencies. G(s) ∞ := max σ (G( jω)) ω
(4.3)
The ∞-norm is an important metric for many control problems, as will be seen later. A second concept that is fundamental for control system investigation is the small-gain. Theorem 1: Small-gain Theorem If two systems G1 and G2 are stable, then the closed loop formed by them (Figure 4.3) is internally stable if: G1G2 ∞ < 1 and G2G1 ∞ < 1 (4.4) The theorem states that, if a system gain is lower than 1 in all directions (for all eigenvalues), for all frequencies, the energy in this loop will eventually be completely dissipated [Skogestad et al., 2007]. It is also possible to express the small-gain theorem by using LFTs.
84
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
Considering the M − ∆ structure (Figure 4.2), the internal stability can be evaluated by only considering the feedback loop. If the system M is stable, and the uncertainty block is bounded in norm ∆ ∞ < β, , with β > 0, the interconnection will be stable as soon as M ∞ < β−1 . The small-gain theorem can be used to infer about the stability of plants in a closed loop with a given controller K and, more inter+ estingly, when a nominal system is perturbed + 1 by an uncertainty matrix ∆. However, its application to practical problems is limited. The assumption that ∆ does not possess any particular structure (therefore being a full matrix) and only has a limited ∞-norm defines an excessively general problem. To obtain an im++ proved, less conservative estimation of stabil2 ity for more realistic problems, SSV can be applied by assuming that ∆ has a well-defined Figure 4.3: Feedback configuration for the small structure. gain theorem. The matrix ∆ is said structured if the uncertainties are organized accordingly to its nature, in a block diagonal form. It is always possible to find an interconnection matrix M so that ∆ will be structured [Doyle, 1985]. A general form for ∆ s is given in Equation 4.5:
G
G
∆ s (s) = diag[δ1 Ir1 , ..., δ s Irs , ∆1 , ..., ∆r ] : δi ∈ C, ∆ j ∈ Cm j ×m j
(4.5)
where δi represents the parametric uncertainties, Ir are identity matrices with adequate dimensions, and ∆ j describes dynamic uncertainties and can contain complex elements. The block diagonal matrix ∆ s is assumed bounded. The stability condition for a SSV is written as: det(I − M( jω)∆ s ( jω)) , 0, ∀ω ∈ R, ∀∆
(4.6)
where the nominal closed loop system M is assumed to be stable. The condition implies all uncertainties in the structure small enough so Equation 4.6 does not become singular at any frequency. Analog to this result, is possible to estimate the smallest uncertainty value, for a given structure, capable of destabilizing the closed loop system. This is the structured singular value µ, and can be used as a metric to describe the robustness of a system. Definition 2: Structured singular value The structured singular value µ∆ (M) of M ∈ Cn×n with respect to ∆ s is defined as: µ∆ (M) :=
1 min {σ(∆ s ) : det(I − M∆ s ) = 0}
(4.7)
∆ s ∈∆ s
where σ(∆ s ) is the singular value of ∆ s . If there is no value of ∆ s ∈ ∆ s so that (I − M∆ s ) is singular, then µ∆ (M) := 0. Definition 2 is, in essence, similar to the small-gain theorem. If the uncertainty ∆ s is bounded by ∆ s ∞ < β, ∀ ∆ s ∈ ∆ s , the perturbed system M is robustly stable, with respect to ∆ s , if and only if µ∆ (M) ≤ β−1 . Furthermore, if the disturbance matrix set is normalized, such as: B∆ s := {∆ : σ(∆) < 1, ∆ ∈ ∆ s }
(4.8)
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
85
the stability condition becomes µ∆ (M(s)) ≤ 1. In this case, the value of µ∆ (M) can provide an estimation on how much uncertainty ∆ s can be tolerated before destabilizing the closed loop. The search of µ∆ (M) constitutes a NP-complete problem, growing in complexity for large dimension, and the computation of its exact value can be prohibitive. Fortunately, there are different methods to estimate its upper and lower bounds µu∆ > µ∆ > µl∆ . These values can be computed with relative simplicity for cases where the structured uncertainty matrix is composed of purely complex or complex/real elements. In the cases where its elements are purely real values, the lower boundary estimation may present large imprecision. To improve this computation, adding a small magnitude complex component to the uncertainties is a common practice. More details in the boundary computation methods can be found in [Packard et al., 1993, Braatz et al., 1994]. Apart from stability, robustness in performance is also highly desired in practical applications. The resulting regulation and tracking characteristics between nominal and real system can vary largely, specially when external disturbances are considered. This variation can render a controller unacceptable for a task, despite still being stable. The robust performance test is developed to indicate how the controller can handle this degradation. To analyze closed-loop performances, the transfer from input to output aug becomes a function of ∆, and the complete f matrix M needs to be considered. Robust performances can be studied by introducing fictitious uncertainty blocks ∆ f across the inputs and outputs channels, thus generating an augmented uncertainty block ∆aug (Figure d v 4.4). By introducing this block, it is always possible to reformulate the robust performance problem in the stability framework [Doyle, 1985, Skogestad et al., 2007]. Therez w fore, the bounds for performance indicators per f per f µu∆ and µl∆ are obtained in the same Figure 4.4: LFT representation for the augmented structured disturbance ∆aug , applied in manner as for the stability analysis problem. the case of robust performance analysis.
M
4.2.3/
H∞ controller
Following the development of uncertainty representation and the small gain theorem in the late 1960’s, George Zames [Zames, 1981] proposed the solution of feedback gain selection as an optimization problem, aiming to minimize an operator norm. The H∞ was originated to solve a fundamental problem of sensitivity in a feedback loop, constraining signals in the loop to achieve a desired requirement, expressed in terms of a ∞-norm. Consider a linear, time invariant system P(s) with two sets of inputs and outputs, represented in Figure 4.5. The H∞ framework aims to minimize the relation between external input signals w (representing disturbances, reference signals, noises, ...) and output z containing tracking errors, control actions and other signals of interest. To achieve this goal, the measurement y is used to derive an input signal u. This mapping between y and u is given by a dynamic controller K, rendering the loop stable. This structure can be easily represented in the LFT framework.
86
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
The loop transfer functions can offer further insight on how to define specific design requirements into the H∞ framework. For a given plant G0 , and considering a classical feedback loop structure, different relations between inputs and outputs are obtained:1 : • Loop transfer function: K(s)G0 (s); • Sensitivity 1 1+K(s)G0 (s)
function: = (1 + L)−1 =
L(s)
w
z
P y
u
=
K
=
S (s) e r
Figure 4.5: Basic structure for the H∞ problem, • Complementary sensitivity function: in the LFT representation. K(s)G0 (s) T (s) = 1+K(s)G = S (s)L(s) = yr 0 (s) • Control
K(s) 1+K(s)G0 (s)
sensitivity function u = K(s)S (s) = r
:
In Figure 4.6, the closed loop scheme is shown. The external reference signal r, the disturbance d and the noise n form the exogenous vector w while z1 , z2 and z3 are the signals forming the output vector z. The system G, together with weighting functions W1 , W2 , W3 (used to shape the desired response during the minimization process) and Wd (describing the disturbance characteristics) form the augmented system P, resulting in a loop structure equivalent to the one presented in Figure 4.5.
r
+
z1
z2
z3
W1
W2
W3
e
K
d n
u
+ +
G
y
Wd + +
P
Figure 4.6: Block diagram representing the augmented plant P, obtained from a nominal system G and the weight functions W. This structure is applied in the framework of H∞ synthesis to obtain the controller K minimizing a relation between inputs (r, d, n) and outputs in z. Sensitivity and complementary sensitivity are connected by the algebraic constraint T = 1 − S . It imposes a trade-off in the closed loop response, put in evidence by Equation 4.9. It describes the influence of each input over the output y. 1
In this case, the sensor dynamics was not considered
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
y = (I + L)−1 L r + (I + L)−1 Gd − (I + L)−1 L n {z } {z } {z } T
S
87
(4.9)
T
To obtain zero tracking error, S → 0, or equivalently T → I, the loop transfer L must have a large value. However, this also implies a direct transfer from noise. In addition, a large value of L is related to large control gains, what can be undesired, saturating the control signal and even damaging the actuators. This conflict can be partially solved by selecting different gains accordingly to the frequency ranges. By choosing L > 1 over low frequencies ensures that sufficiently slow reference transitions will be correctly tracked, while disturbances in these frequencies will have reduced effects. At high frequencies, where measurement noises are usually found, the gain L is made sufficiently small to disregards this effects. The selection of L, therefore, defines performance trade-offs to be made during controller synthesis. In the framework of H∞ control, desired performances can be imposed by the selection of appropriate weighting functions W1 , W2 and W3 . By selecting them as frequency dependent (W( jω)), complex requirements and specifications can be defined, and additional information (i.e. the disturbance dynamics) can be included in the problem. When multiple weighting functions are employed, the requirement conditions are written as:    W1 (s)S (s) ∞ < γopt     W2 (s)S (s)K(s) ∞ < γopt       W3 T (s) ∞ < γopt
(4.10)
where γopt is the minimal (optimal) value, representing its performance level. The solution for this problem is not unique except in the scalar case. If γopt ≤ 1, all the requirements are achieved. In practical design, the computation of an optimal controller may be too numerically demanding. Hence, it is common to accept sub-optimal stabilizing controllers, so that the norm value is γ → γopt , with γ > γopt . The optimization process in the H∞ methodology searches for the dynamic controller K(s) that stabilizes the system and, at the same time, minimizes the positive scalar γ value in the Inequation system 4.10. By representing the augmented plant P and partitioning it as shown in Equation 4.11:    x˙ = Ax + B1 w + B2 u     z = C1 x + D11 w + D12 u      y = C2 x + D21 w + D22 u
  #  A B1 B2  '   P P =  C1 D11 D12  = 11 12   P21 P22 C2 D21 D22
(4.11)
the system can be easily represented by a lower LFT, and the resolution a sub-optimal H∞ problem is summarized as: min Fl (P, K) ∞ , given that K stabilizes P (4.12) K
Several methods are available to solve Equation 4.12 and to find the controller K(s) that minimizes γ. The most popular are based on the DGKF algorithm [Doyle et al., 1989] (that solves simultaneously two algebraic Riccati equations) and based on linear matrices inequalities (LMIs) [Gahinet et al., 1994]. The DGKF algorithm, firstly limited by assumptions on the problem’s structure, was later expanded to encompass larger classes of systems
88
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
[Green et al., 1990, Stoorvogel, 1991, Scherer, 1992]. In other hand, the LMI based approach [Gahinet et al., 1994] presented a less restrictive algorithm for minimizing the cost function, at the expenses of larger computational effort. In [Liu et al., 2006], a demonstration for duality between this two methods shown, under certain assumptions.
4.2.4/
Extended state observer structure
In the H∞ approach, the controller is obtained based on information about the model, its range of uncertainty and parametric variations, and also over the characteristics of disturbances affecting the system (frequencies and amplitudes). This approach is, among others, heavily dependent on modeling and plant characterization. In the middle of the 1990’s, Jingqing Han proposed a new control structure, named Extended State Observer (ESO). He argued that modern control theory was too much focused on modeling, model-based analysis and synthesis [Gao et al., 2001]. Despite bringing important advances in the field, this paradigm had fundamental limitations, mostly when dealing with robustness issues, relying heavily on accurate plant representation. Han contemplated if effective control could be achieved only based on general information of the physical process and measurement data obtained in real time. The goal of perturbation observers is to identify, in real time, an unmeasured signal/uncertainty responsible for deviating the system from its desired/nominal performance and to compensate it. The core difference between ESO and other disturbance rejection methods, like Unknown Input Observer, Disturbance Observer and Perturbation Observer is that this method does not distinguish between external disturbances and model uncertainties, while others rely explicitly on the mathematical modeling of one or both elements [Schrijver et al., 2002, Radke et al., 2006]. This apparent small difference represented an important design methodology shift, focusing on system perturbations instead of modeling. Its structure addresses realistic disturbances and plant uncertainties while maintaining its simplicity. The main ideas shaping the ESO methodology are presented next. Considering a generic, nonlinear, single-input single-output system in state space representation:   x˙1        x˙2             x˙n      y
= x2 = x3 .. .
(4.13)
= a(x, d) + b(x, d)u = x1
where xi , i = 1, .., n represents the system states, u the control input, y the measured output, d an external, bounded disturbance and a(x, d) and b(x, d) are nonlinear functions describing the system dynamics. It is assumed the existence of a state feedback controller Φ(x) stabilizing the system origin x = 0 in closed loop. To implement this controller only using the available output measurements, state estimation is necessary. In most practical applications, values for a(x, d) and b(x, d) can be approximated through physical modeling and identification techniques, with varying degrees of accuracy. Supposing ˆ these functions can be approximated by nominal models, aˆ (x) and b(x). Next, an auxiliary variable is introduced, σ, and it is defined as:
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
ˆ σ = a(x, d) − aˆ (x) + [b(x, d) − b(x)]u
89
(4.14)
This variable represents the sum of modeling errors and external disturbance acting over the system, and is called total disturbance. By combining Equations 4.13 and 4.14, and introducing σ as a new virtual state, it is possible to obtain an equivalent, augmented representation of the original system:   x˙1        x˙2             x˙n        σ ˙     y
= x2 = x3 .. . ˆ = σ + aˆ (x) + b(x)u ˙ u, u˙ ) = φ(x, d, d, = x1
(4.15)
Instead of estimating the original states, it is possible to develop an observer for the aug˙ u, u˙ ) is bounded. An observer mented representation, assuming the virtual state dynamics φ(x, d, d, structure for the augmented system takes the general form of Equation 4.16, where hi are function of the estimation error.   x˙ˆ1        x˙ˆ2             xˆ˙n      σ ˆ˙
= xˆ2 + h1 (y − xˆ1 ) = xˆ3 + h2 (y − xˆ1 ) .. .
(4.16)
ˆ xˆ)u + hn (y − xˆ1 ) = σ ˆ + aˆ ( xˆ) + b( = hn+1 (y − xˆ1 )
The core idea of the ESO methodology is that, if the states can be closely tracked by the observer, then a controller can use the estimated states and total disturbance in the feedback. Building the control signal u as: 1
(−σ ˆ − Φ( xˆ)) ˆ xˆ) b( and substituting it in Equation 4.15, results in a Equation 4.18:   x˙1 = x2        x˙2 = x3      ..  .       x˙n = aˆ (x) − Φ( xˆ)      y = x 1 u=
(4.17)
(4.18)
assuming the estimation of σ is sufficiently close to its real value. This equation indicates the system will follow the dynamics defined by the nominal model aˆ (x), and a state feedback control Φ( xˆ) can be designed for this nominal system. Furthermore, if the observer is sufficiently fast, dynamic external disturbances can be estimated and attenuated. An interesting advantage of this method is that it requires only an approximated knowledge about the system dynamics, and only supposes the boundedness of external disturbances.
90
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
However, some remarks should be made about the selection of the function aˆ . The ESO method is often robust enough to accept aˆ = 0, what indicates that the whole plant dynamics is unknown and will be estimated by σ. ˆ Nonetheless, [Zheng et al., 2012] demonstrated how this function can have a significant impact over the its transient response. The article shown that, as aˆ approximated the real value, better results could be achieved (faster disturbance rejection and settling times, and less oscillatory responses). Therefore, any knowledge about these dynamics should be considered. Choosing the observer functions hi is the final design step. In the literature, several methods are proposed, but [Guo et al., 2011] points out its tuning process in various of those strategies remains largely experimental. The original work of Han proposed the use of a special function, combining linear and nonlinear elements [Gao et al., 2001, Tang et al., 2014]. Classic linear Luenberger observers have also been used in [Yang et al., 2009]. The use of sliding mode observers has been reported with good results in[Wang et al., 2003], as well as the development of hybrid tuning methods, which include switching gains and the use of optimized filters to reduce the influence of noise [Mado´nski et al., 2015]. Another method is funded on the use of high-gains [Freidovich et al., 2008, Guo et al., 2011, Khalil et al., 2014] for the function selection. In this case, hi are chosen as linear functions, selected as constant large gains to be multiplied by the observer estimation error. This class of observers presents performance recovery properties, indicating that a controller considering output feedback and estimated states will behave similarly to a state feedback one. This includes recovering the region of attraction on non-linear controllers. In addition, [Tian et al., 2009] highlighted the stability demonstration of the ESO with hi selected by the high-gain method can be achieved for problems with less restrictive assumptions. From the practical point of view, it can be easily parametrized, reducing the number of tuning variables and resulting in a straightforward design approach. Due to these characteristics, the observer gains hi will be chosen accordingly to the high-gain methodology.
4.2.4.1/ High-gain observer High-gain observers firstly appeared in the context linear feedback, to provide robust properties [Doyle et al., 1979]. The need for robust observers arose from the fact that state feedback methods, such as linear quadratic regulators, presented impressive robustness properties for gain and phase margins only for the full state-feedback case. However, in practice, only a limited number of states are available for measurement, and the use of observers and Kalman filters is necessary. In this case, no guaranteed robustness properties by the controller was held, and every design needed to be analyzed individually. This problem became known as the loop transfer recovery (LTR) [Athans, 1986]. Through the use of high gains, a trade-off is proposed, where closed-loop stability margins can be improved to the point of recovering the full-state feedback performance, at the cost of reduced noise rejection. Similar properties were later shown for classes of nonlinear systems. A demonstration of the separation principle for a class of nonlinear systems is shown in [Atassi et al., 1999], where it was possible to independently design the controller and a sufficiently fast high-gain observer. Furthermore, it also indicated, for adequately selected parameters, the convergence of the state feedback and output feedback trajectories. This difference could be made sufficiently small to recover certain closed loop characteristics for some nonlinear systems, including their region of attraction and asymptotic stability at the origin [Khalil et al., 2014]. The following example, based on a second order dynamic system, is taken to explain the methodology, and its generalization for high-order systems is straightforward. Assuming a general
4.2. ROBUST POSITION CONTROL STRATEGIES IN VACUUM
91
system given by Equation 4.19:    x˙1 = x2     x˙2 = f (x, u, w, d)      y = x 1
(4.19)
where d the vector of disturbance inputs, w the vector of known exogenous signals, u the control input and y the measured output. The function f is locally Lipschitz in (x, u) and continuous in (d, w). If a state feedback control u = Φ(x, w) stabilizes the closed loop system at x = 0 is available, its implementation will depend on the correct state estimation. An observer taking the form described by Equation 4.20 is proposed:     x˙ˆ1 = xˆ2 + h1 (y − xˆ1 )    x˙ˆ2 = fˆ( xˆ, u, w) + h2 (y − xˆ1 )
(4.20)
where fˆ(x, u, w) is the nominal model of f (x, u, w, d). If the function f its sufficiently known, it is possible to assume fˆ = f . In most cases, however, this is not valid. The estimation error for the observer is defined as the vector x˜i = xi − xˆi . The estimation of error dynamics x˙˜ can be obtained using Equations 4.19 and 4.20, resulting in Equation 4.21:     x˙˜1 = −h1 x˜1 + x˜2    x˙˜2 = −h2 x˜1 + δ(x, x˜, w, d)
(4.21)
where δ(x, x˜, w, d) is a function representing the error estimation in the system dynamics f − fˆ. To verify the observer stability, two cases must be studied. Considering the ideal case, where the function δ(·) is zero (representing a system with no external disturbances and with a perfectly known model), the asymptotic convergence of Equation 4.21 is achieved when its characteristic matrix, ' # −h1 1 −h2 0 is Hurwitz. This holds for any positive constant values of h1 and h2 . However, this is hardly the case, and δ(·) should be considered. In this situation, it would be interesting to minimize the influence of uncertainties over the estimation error. By examining Equation 4.21, the transfer function from δ(·) to x˜i can be represented by: ' # 1 1 T δ x˜ (s) = 2 s + h1 s + h2 s + h1
(4.22)
Equation 4.22 puts in evidence the impossibility to completely eliminate uncertainties and disturbance influences over the estimation error. Still, it can be reduced to a sufficiently small value over all frequency range by minimizing supω T δ x˜ ( jω) . Rewriting Equation 4.22 in the following form:
T δ x˜ (s) = 0012
√1 h2 √s h2
00132
+
√h1 √s h2 h2
+1
  
√s h2
√1 h2
+
√h1 h2
  
(4.23)
92
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
indicates this minimization objective can be met if the ratio √hh1 is a constant positive real number, 2 and the value of h2 tends to infinity. This can be implemented by selecting the observer gains as: hi =
αi εi
(4.24)
for positive constant values of α, and with a positive value of ε 001c 1. This assigns the observer eigenvalues to be 1/ε times the roots of the polynomial s2 +α1 s+α2 . Therefore, for a small enough value of ε, the observer dynamics can be made much faster than the system response under state feedback, and the influence of δ over it can be reduced. Further investigation on the observer characteristics can be performed in the time-domain. For this purpose, the estimation errors should be scaled as: ζ1 =
x˜1 , ζ2 = x˜2 ε
(4.25)
Combining the scaled variables with Equation 4.21 results in the following singularly perturbed equation: εζ˙1 = −α1 ζ1 + ζ2 (4.26) εζ˙2 = −α2 ζ1 + εδ(x, x˜, w, d) Equation 4.26 shows the effect of ε on the response: smaller its value is, smaller will be the effects of modeling errors and disturbances δ over the estimation, and faster the observer dynamics will be. However, Equation 4.25 also indicates the transient response can be large when the initial conditions x1 (0) , xˆ1 (0), in what is called peaking phenomenon [Esfandiari et al., 1989, Esfandiari et al., 1992]. Although the estimation error may decay fast and converge to its true value, the peak value tends to an impulsive-like behavior when ε → 0, which degrades the response and may even destabilize the closed loop system and damage actuators and other components. As this peak is an artifact introduced by the observer, it can be safely disregarded, what can be achieved by designing the control law and the function fˆ( xˆ, u, w) to be globally bounded in xˆ. In practice, this means limiting the operation range of our system by saturating the control u and/or the estimation xˆ outside a compact set of interest. Due to the fast high-gain observer dynamics, the estimation error should be of order O(ε) after this short transient period. By introducing saturation, the real plant state x is limited during the peaking phenomena, and the observer converges faster to the real value. In [Freidovich et al., 2008, Khalil, 2008], it was shown that, for sufficiently small values of ε, an output feedback controller using the states estimated by this method will closely approach the trajectories of a full state feedback controller, with errors of order O(ε), and will recover its stability and performance characteristics. This controller is based on the separation procedure, where it is designed as if the full states were available, and the observer is designed independently. The separation principle is well known in the context of linear systems, and similar forms for different classes of non-linear systems can be found. However, [Khalil et al., 2014] emphasized that in high-gain observers, trajectory recovery can be achieved by making the observer sufficiently fast. This feature allows designing state feedback controllers to meet design specifications and, by tuning the observer with decreasing values of ε, bringing output feedback trajectories closer to the state feedback ones. As stated in the start of the discussion about high-gain observers, its recovery properties and robustness come at the cost of an increased sensibility to measurement noises, due to the fact
4.3. CONTROLLER DESIGN
93
that this observer approximates a differentiator. In the presence of high frequency noises, this derivation effect can lead to a reduction in the signal/noise ratio. Assuming a measurement noise n is limited by µ = sup n , its effect on the state estimation error takes the form: µ x(t) − xˆ(t) ≤ c1 ε + c2 , ∀t ≥ T ε
(4.27)
for positive constants c1 , c2 , T [Khalil et al., 2014]. Appendix C details µ the procedure to obtain Equation 4.27. The c1 ' + c2 first part of this inequation right-side shows ' the direct influence of the gain ε in the state estimation accuracy. Yet, the second element indicates that, for small values of ε, the error will tend to infinity. This bound, represented in Figure 4.7, illustrates how the presence of measurement noise limits the observer k µ 12 a convergence velocity in practice. The trajec1 tories in a noisy closed loop system under this ' ca µ 2 √ feedback approach asymptotically the O( µ) √ √ neighborhood (ka µ, where ka = 2 c1 c2 ) Figure 4.7: Sketch of the noise effects on the ob√ as ε tends to an optimal value (ca µ, where server error. From [Khalil et al., 2014]. √ ca = c2 /c1 ). The high-gain observer presented here has a similar form to a standard ESO problem, and the interconnection between both methodologies can be made without large efforts. The resulting controller is called extended high-gain observer (EHGO).
4.3/
Controller design
When dealing with fragile components in the micro and nano-scales, requirements for the end-effector precision, accuracy, response times and robustness become more strict. During manipulation and characterization of such elements, it would be ideal to minimize the processing time and the chances of damaging components (samples and end-effectors). In this stage, the position control for the actuated micro-gripper finger is considered. By taking in consideration the open-loop dynamics of this device (obtained in Chapter 3), the goal requirements to be achieved by the closed loop systems are defined as follows: 1.
The system settling time for a reference step input in vacuum should be lower than 30 ms. This is approximately three times faster than its response in air, and more than three hundred times faster than the open loop response in the vacuum.
2.
The response overshoot to a reference step input should be zero, to avoid undesired collisions between the gripper and fragile samples.
3.
Steady state error to a reference step input should be smaller than 0.1%.
4.
An end-effector variation to perturbations of 1 nanometer (RMS) was set as goal challenge, once that it presents a genuine test when considering other works dealing with the precise positioning task at this scale.
94
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
In addition, as the system behavior changes for different operation points, the controllers should be robust enough to hold these characteristics for the whole operational range. These requirements consider the micro-gripper as a tool responsible for fine-position control, capable of quickly respond to events in the chamber. In the following subsections, H∞ and EHGO controllers are developed based on the above design requirements. These controllers were selected for this work based on their capabilities to deal with plant uncertainties and attenuate disturbances. While the H∞ strategy requires a good knowledge over those perturbations, the EHGO method relies on on-line estimation for these quantities based on reduced models and on-line measurements. A remark should be made about the input non-linearity present in the model. As the input electrostatic force is proportional to the square of Vin , the control signal is assumed to be U = Vin2 . This alteration is possible because the voltages applied to the gripper should always be positive, accordingly to the manufacturer specifications. The controllers then had their performances compared through simulation. The robustness levels achieved were investigated by considering an uncertain model for the gripper, taking in consideration the estimated parametric variations obtained during its identification process.
4.3.1/
Controller synthesis using H∞
As shown in Section 4.2.3, the optimization process to obtain an H∞ controller requires the construction of an augmented plant P, where design requirements are introduced by means of weighting functions W. The control design depends on their choice and may be not always evident when dealing with concurrent objectives. The process employed in this phase can be summarized as follows: (a) Define the weighting functions that will shape the desired response characteristics, and construct the augmented plant P, considering the nominal transfer function of the plant model G0 . (b) Perform the optimization process, and obtain a controller K and the sub-optimal γ value. If no controller is found and the optimization fails, redefine the weighting functions. (c) If the obtained γ < 1, the controller achieves all the requirements imposed by the weighting functions. In this case, the performance requirements can be modified to more restrictive ones (i.e. faster response, better disturbance rejection, ...). If γ > 1, investigate which requirements could not be achieved, and try to balance them if necessary by changing the weighting functions. (d) If a satisfactory controller is obtained for a nominal model, test its robustness for all the family of models G. If it does not achieve the desired performances, modify the weighting functions.
4.3.1.1/ Selection of the weighting functions In the multi-objective H∞ problem formulation, three functions can be used: W1 , W2 and W3 , each one defining a specific desired characteristic. However, using all of them may be not necessary. In a first instance, only W1 will be used, to introduce the first three design requirements in the optimization problem.
4.3. CONTROLLER DESIGN
95
These requirements can be expressed with the aid of a target closed loop function. The closed-loop response T d (Equation 4.28) represents a desired transfer function from the reference signal to the system output and will be used as a base to obtain the weighting function W1 . Td =
Kd ω2d s2 + 2ξd ωd + ω2d
(4.28)
Following the first three design requirements, and using a second-order transfer functions as target for the closed loop system, its parameter can be selected to achieve the desired response. In this case, Kd = 0.999 defines the allowed static error of 0.1%, ωd = 100rad/s the time response characteristics and ξd = 1.8 the system damping. The weighting function W1 is obtained through the relation 4.10, resulting in: W1 =
1 1 = S d 1 − Td
(4.29)
In this first iteration, when only the minimization of S W1 ∞ was considered, the resulting controller was able to achieved the second and third design requirements while obtaining a steady state error of 0.14% for the nominal plant, slight above the desired value. Analyzing the singular value diagram from a disturbance input d to the system output y revealed that the obtained controller could only reject low frequency disturbances. This means disturbances with frequencies close to the gripper’s oscillation modes would be still transmitted to the output. As the transference function from the disturbance signal d to the error signal e is given by GS , it was possible to improve the response by modifying W1 , taking this requirement into consideration. In an ideal case, the weighting function should only include the frequencies to attenuate, avoiding introducing unnecessary and conservative requirements. To include this design requirement, indicating how disturbances around the two vibration modes present in the system should be attenuated, a band-stop filter with narrow band approaching the characteristics of notch filters is used: Tf =
s2 + abs + ω2bs s2 + 2ξbs ωbs + ω2bs
(4.30)
where ωbs is the central frequency of the filter in rad/s, ξbs is the damping defining the peak width, and abs is a parameter regulating the filtering peak amplitude. Two band-stop filters were added to W1 , one for each resonant mode: W1 =
1 1 1 − T d T f1 T f2
(4.31)
To improve the performance, a function Wdist describing the input disturbances acting on the system was added. It was constructed to take into account the environmental disturbances described in Chapter 2, and the input electronic noise, generated in the hardware that produces the gripper actuation voltage and commented in Sections 3.2.2 and 3.3.3. To avoid using a function with elevated order, Wdist was designed as a high-pass filter to represent the influence of electronic noise close to the oscillation modes, combined with a band pass filter covering the environmental disturbances between 100 and 700 Hz. A new controller was obtained for the new augmented plant P, and the improvement in disturbance rejection can be seen in Figure 4.8. It shows the singular value from the disturbance signal d to the measured
96
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
output signal y. The graph indicates attenuation levels up to 50 dB for disturbances exciting the first mode in relation to the previous controller, with only a few decibels of improvement for the second oscillation mode. The improved response at high frequencies demands a small degradation in the attenuation capabilities at low frequencies, but still within acceptable values.
40 H∞ with W1
Singular Value (dB)
20
−1
H∞ with W1 . Tf
and Wdist
0 −20 −40 −60 −80 −100
0
10
1
10
2
10
3
10
Frequency (Hz)
Figure 4.8: Singular value plot from disturbance input d to tracking error e. Two different controllers were compared: one obtained through H∞ optimization when only reference tracking characteristics were considered (W1 , in blue), and another considering reference tracking and disturbance effects on the system (W1 T −1 f and disturbance characteristics Wdist , in red). While the obtained controller for the augmented system P considering the weighting functions W1 and Wdist presents good characteristics for step reference inputs and disturbance attenuation it also leads to larger control signals u. By adding the weighting function W2 , limit control actions can be imposed in the loop at different frequencies. This can, in practice, reduce saturation and high-frequency shattering. The function W2 was selected as: W2 = K p
s/ωz + 1 s/ω p + 1
(4.32)
where K p represents the allowed gain relation between input and output, and the parameters ωz = 300 and ω p = 900 were selected to shape W2 as a high-pass filter, penalizing high-frequency control. The gain K p , selected as 20/602 , indicates that displacements of 20 µm in the gripper should not surpass a 60 Volts input. The squared value was introduced to match the nonlinear input relation present in the electrostatic actuator. The resulting multi-objective optimization problem returned a γ value of 2.10. Figure 4.9 shows the comparison between desired requirements and obtained responses. The limitation in the control signal is achieved for the most of the frequencies, except for those close to the second oscillation peak present in the model. Although not all performance requirements were achieved, the resulting controller approached them closely, with a steady state error of 0.2% instead of the required 0.1%, while the other design requirements related to settling time and system damping were respected. Further modifications in the weighting functions, to improve the steady state error, resulted in a loss of performance over other characteristics. The obtained H∞ controller is of 14th order (same dimension as the augmented plant P),
4.3. CONTROLLER DESIGN
97
10
S
40
Singular Value (dB)
Singular Value (dB)
0
50 1/W1
−10 −20 −30 −40
30
20
10
−50
1/W2 0
−60 −70
−2
10
0
10
2
10
4
10
−10 −2 10
Frequency (Hz)
KS
0
10
2
10
4
10
6
10
Frequency (Hz)
(a)
(b)
Figure 4.9: Resulting sensitivity function S and control sensitivity function KS for the obtained H∞ controller (green line), compared with the target function (blue line).
and reached a balance between the achievable performances for this structure considering the different requirements. This design process is a practical example of the limits and difficulties that can be found when pushing performances to an extreme, where compromises should be made between desirable and achievable. It is important to highlight that the weighting functions could have being developed differently, considering different structures and parameters, what could result in radically different controllers. In [Bibel et al., 1992], it is stated that the relation between weighting function in the multi-objective H∞ problem is complex and often non-linear. This means a small variation in one of the weighting functions may lead to a drastically different controller after the optimization process. As results, performing fine adjustments of these functions to improve a certain aspect of the response may be not obvious.
4.3.2/
Controller synthesis using ESO
Contrasting with the H∞ control framework, the ESO design focus on disturbance rejection based on fewer information about plant and perturbations, where undesired effects are evaluated and compensated on-line. Therefore, the approximations made during the identification process, where a forth-order model with varying parameters is considered, can be relaxed. In this case, modeling the system as a second order Duffing equation with a unknown nonlinear stiffness can help to simplify the problem. Assuming the system has the form described by Equation 4.33, where the functions a(x, d) and b(x) hold the non-linear gripper behavior in vacuum, not perfectly known, and d is a bounded disturbance signal with bounded derivatives.    x˙1 = x2     x˙2 = a(x, d) + b(x)u       y = x1
(4.33)
Considering the gripper behavior, the nonlinear stiffness effects were included on the unknown dynamics of a(x, d), and b(x) includes the varying input gain for the system. These functions were approximated by defining its nominal models a0 and b0 as:
98
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
K0 C0 x1 − x2 M0 M0 Kelec0 b0 = M0 a0 = −
(4.34) (4.35)
where M0 , K0 , C0 and Kelec0 are nominal values for the systems mass, stiffness, damping and electrostatic gain, respectively. These parameters were precisely identified in Chapter 3 and their approximated values are summarized in Table 4.1. These values reflect the gripper dynamic behavior in air, intentionally selected to test the efficacy and robustness of the ESO in face of larger parametric uncertainties.
Value
M0
K0
C0
Kelec0
3.7e−8
1.6
3e−6
0.01
Table 4.1: Nominal parameters used for the micro-gripper dynamics, considered for the ESO.
A variable change was performed to introduce the tracking control task, where the signal x1 should follow a bounded reference signal r, with derivatives also bounded. Reformulating the problem as a function of the tracking error, using the relations e1 = x1 − r and e2 = x2 − r˙, allows to rewrite Equation 4.33 as:    e˙ 1 = e2     (4.36) e˙ 2 = a(e, r, r˙, r¨, d) + b(e, r)u      y e = e 1 The observer was designed based on this tracking error dynamic equation. An augmented equivalent system, with the addition of a virtual state σ considering the nominal model, is easily derivable, and is given by Equation 4.37:   e˙ 1 = e2        e˙ 2 = σ + a0 (e) + b0 u   ˙ u, u˙ , r, r˙, r¨)  σ ˙ = ϕ(e, d, d,      ye = e1
(4.37)
In this form, an observer can be obtained by the method described in Section 4.2.4.1, resulting in the system given by Equation 4.38:    e˙ˆ 1 = eˆ 2 + h1 (ye − eˆ 1 )    ˙ eˆ 2 = σ ˆ + a0 (ˆe) + b0 u + h2 (ye − eˆ 1 )      σ ˙ˆ = h3 (ye − eˆ 1 )
(4.38)
The observer gain functions hi , selected using the high-gain methodology, took the form of a large gain multiplied by the observer estimation error ye − eˆ 1 , given by Equation 4.24 and repeated here for convenience: hi =
αi εi
4.3. CONTROLLER DESIGN
99
The values for αi were selected so the polynomial s3 + α1 s2 + α2 s + α3 is Hurwitz. It was possible to parametrize this selection by choosing αi as the coefficients of a Hurwitz polynomial, such as (s + 1)3 . Consequently, a single parameter, ε 001c 1, remained to be tuned. It defined the convergence rate for the observer, and selected taking in consideration measurement noise effects. This parameter was chosen experimentally, by reducing its value iteratively. Selecting a sufficiently small value for ε renders a quick convergence for the observer, so that the estimated total disturbance σ ˆ approaches the real value of σ. If the convergence is made sufficiently fast, dynamic disturbances can be estimated and compensated, while keeping the tracking error small. The state feedback controller was designed independently from the observer, considering the nominal plant dynamics. It can be obtained through different methods (pole allocation, linear quadratic regulator, ...). For simplicity, this work used a static state feedback, considering pole placement, to obtain a system response with no overshoot and with settling time close to 1 millisecond, a value much faster than required in the specifications. The considered state feedback controller assumes the form Φ = [φ1 φ2 ]. The final feedback control is given by: u=
1 (−σ ˆ + Φˆe) b0
(4.39)
This control action was saturated between 0 and 70 Volts, a selected operation range for the gripper comprising positions between 0 and 30 µm, to limit possible undesired effects caused by peaking phenomena. In addition, a reference filter Fre f : Fre f =
1000 s + 1000
was included, to ensure a sufficiently smooth reference signal with bounded derivates. This filter slows the settling time to around 4 ms. Faster filters could still be applied, at the cost of some degradation in the initial transitory response. Figure 4.10 shows the final ESO structure. The observer, selected using the high-gain method, supplies estimations for position and velocity errors to the controller, as well as the estimated disturbances and model uncertainties σ. ˆ These values were used to construct the control signal u given by Equation 4.39.
4.3.3/
Robustness analysis
This section resumes the robustness properties for the controllers described above, comparing both performances in the face of uncertainties, external disturbances and measurement noises. The closed loop system was built considering the uncertain plant G(∆) as a structured uncertain model (based on the parametric variation range previously identified in Table 3.1) and the obtained, H∞ and EHGO controllers. It is important to remark that this polytopic uncertainty representation can generate parameter combinations that may not be obtainable by the real system, hence results may be conservative. However, this method still offered an interesting tool for comparing its robustness. The test were performed via simulation in the µ analysis framework, by rearranging the closed loop systems in the M − ∆ structure for the different controllers. Bounds for the real SSV value µ∆ (M) allowed estimating the controller’s stability and performance in the frequency spectrum.
100
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
Plant d u
Controller ˆ
+
G0
+
y
eˆ
Observer
e1
+
+
n
+
r
Fref
EHGO Figure 4.10: The block diagram for the ESO controller implementation.
The first result concerning the stability of these closed loop systems can be seen in Figure 4.11. Both controllers achieved robust stability for all the considered parametric range, as the upper bounds values are smaller than 1 for all frequencies. The critical frequencies for the H∞ controller were located around the two resonant frequencies. A maximum calculated upper bound value of 0.89 indicates that the closed loop system should hold for an uncertainty up to 112% of the considered range. In other hand, the controller based on ESO led to lower limit bounds (µu = 0.53). This indicates that the ESO control structure can stand much larger uncertainties (up to 188% of the present value).
1
1 Upper bound µ
u
0.9
Lower bound µl
0.8
0.8
0.7
Magnitude (Abs)
Magnitude (Abs)
0.9
0.6 0.5 0.4 0.3
0.6 0.5 0.4 0.3 0.2
0.1
0.1 0
10
1
10
2
10
Frequency (Hz)
(a) H∞ controller
3
10
u
Lower bound µl
0.7
0.2
0 −1 10
Upper bound µ
0 −1 10
0
10
1
10
2
10
3
10
Frequency (Hz)
(b) ESO controller
Figure 4.11: Robust stability index for the controllers. The curves show minimum and maximum bounds for the true value µ∆ . The next step in the analysis focused on the response characteristics considering the three first design requirements, testing if the transfer maps from inputs w to outputs z were insensitive per f to uncertainties. If the upper bounds µu is smaller than 1, or similarly T wz ∞ < 1, the system achieves robust performance for the modeled uncertainty. Considering the H∞ controller, it was possible to know beforehand that its reference
4.3. CONTROLLER DESIGN
101
tracking performance is not robust, as the H∞ norm of the nominal plant was greater than 1. Reexamining Figure 4.9(a) demonstrates this aspect, a consequence for the controller not being per f able to achieve the desired steady state error requirements. The resulting µ∆ (M) curve (Figure 4.12(a)) reflects this property at lower frequencies. At higher frequencies, a peak was present in the upper limit boundary close to 2000 Hz, indicating that signals in this frequency may be close to infringing some performance requirements. The same simulation was made for the ESO controller structure (Figure 4.12(b)), and revealed the reference tracking performance requirements were met with ease. Two peaks were present at high frequency range, but the upper boundary values were still considerably far from accepted unitary limit. This indicates the controller can ensure current performance requirements an even larger range in the parametric uncertainty.
1.8
0.7 Upper bound Lower bound
1.6
0.6
Upper bound Lower bound
Magnitude (abs)
Magnitude (abs)
1.4 1.2 1 0.8 0.6
0.5 0.4 0.3 0.2
0.4 0.1
0.2 0
1
10
2
10
Frequency (Hz)
(a) H∞ controller
3
10
4
10
0
1
10
2
10
3
10
4
10
Frequency (Hz)
(b) ESO controller
Figure 4.12: Robust performance index for the obtained controllers. The curves show minimum per f and maximum bounds for the true value µ∆ . To better interpret these results, the closed loop systems responses for a unitary reference step signal, in the time domain, are given in Figures 4.13(a) and 4.13(b). Twenty responses (blue continuous lines) were obtained for each controller, considering different plants obtained randomly from the family models set. The vertical black dashed lines indicates the time when a response enters its steady state, defined as 98% of the final value. The curve shows, for the H∞ controller, the defined settling time requirement of 30 milliseconds is achieved for almost all the samples model, with times varying between 19 to 33 ms. For the second controller, faster response times were obtained with settling times about 4 ms. This time is close to the selected reference filter Fre f response, indicating the controller is sufficiently fast to closely track such fast reference changes with a small dispersion among responses. Figures 4.13(c) and 4.13(d) show the respective control signals applied during the reference step. The ESO controller presents a sharp increase in the control signal in the initial moments, due to the peaking phenomenon. However, as the observer converges to the true value, this effect is eliminated and the control signal behaves normally. This occurs quickly (under 0.2 millisecond), with almost unnoticeable consequences over its displacement. The second design requirement was attained by both controllers in all range, as no overshoot was present in the responses. Finally, steady state errors 0.2 and 0.14% we obtained for the H∞ controller, where the latter value corresponds to the one obtained for the nominal plant during
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
1
1
0.9
0.9
0.8
0.8
Displacement (µm)
Displacement (µm)
102
0.7 0.6 0.5 0.4 0.3
0.7 0.6 0.5 0.4 0.3
0.2
0.2
0.1
0.1
0 0
0.01
0.02
0.03
0.04
0 0
0.05
0.002
Time (seconds)
0.006
0.008
0.01
(b) ESO - position
14
14
12
12
Control signal (Volts)
Control signal (Volts)
(a) H∞ - position
10
8
6
4
2
0 0
0.004
Time (seconds)
10
8
6
4
2
0.01
0.02
0.03
Time (seconds)
(c) H∞ - control signal
0.04
0.05
0 0
0.002
0.004
0.006
0.008
0.01
Time (seconds)
(d) ESO - control signal
Figure 4.13: Simulated response to unitary step reference change considering various plants within the parametric variation range. (a) and (b) show the responses for a unitary reference step. The dashed-dot lined indicate the settling time (t98% ). (c) and (d) indicate the respective control signals applied.
the controller synthesis. As commented before, the design objective of 0.1% could not be achieved for the H∞ controller without sacrificing other aspects of the response. For the EHGO controller, close to null error is obtained for all plants. Up to this point, the analysis concerned just signal tracking characteristics. The last design requirement, disturbance rejection, is addressed next. Recalling the results from Chapter 2, mechanical and acoustic vibrations operating in different frequencies can affect elements inside the microscope. If these undesired excitation frequencies match the vibration modes of the components in the SEM chamber, large displacements can be generated as result. Figure 4.14 shows the Bode diagram for the closed loop systems, from disturbance to the gripper tip position. The graphs show two curves for each case, one representing the nominal system (blue continuous line) and another the response obtained for the worst-case gain model (red dashed line). The worst-case gain is the parameter combination, within the defined variation range, producing the largest singular value. Perturbations acting in frequencies under 700 Hz and above 2 kHz presented a large attenuation for both controllers. The critical point occurs for perturbations in frequencies close to the system’s oscillation modes. For the H∞ controller, attenuation values between -15 and -10.8
4.3. CONTROLLER DESIGN
103
dB for the first and -36 to -20 dB for the second mode were obtained. This presented a large improvement in relation to the open loop response, where amplifications close to 40 dB for the first mode could be achieved.
0
−40
−10 −50
−30
Magnitude (dB)
Magnitude (dB)
−20
−40 −50 −60 −70 −80 −90 −100
−60
−70
−80
−90 Nominal Worst case 1
Nominal Worst case 2
10
10
Frequency (Hz)
(a) H∞ controller
3
10
4
10
−100
1
10
2
10
3
10
4
10
Frequency (Hz)
(b) ESO controller
Figure 4.14: Closed-loop Bode diagram, from the input disturbance d to the system output y, considering the nominal plant response (blue continuous line) and the worst case result (red, dashed line) achievable within the defined parametric variation range.
The ESO based control analysis indicated a larger disturbance attenuation capability, where values between 43.5 to 42 dB, and 48.1 to 49.8 dB, were achieved for the first and second vibration peaks, respectively. As this controller presents a proactive disturbance compensation structure, improved attenuation capabilities were already expected. Yet, it is worth to remark how its intrinsic perturbation rejection properties can be obtained without any a priori knowledge over the disturbances, a valuable characteristic for many practical applications. The influence of noise in the control signal is another aspect to be considered. The Bode curves, shown in Figure 4.15, demonstrate this limitation in the high-gain observers, as mentioned previously. As the observer gain increases, its dynamics become faster, improving its tracking and disturbance rejection capabilities. At the same time, the noise impact in the closed loop also increases, up to a limit where higher observer gains would degrade the estimation performance as the signal/noise ratio in the estimated states is reduced. Comparing both graphs, the larger impact of measurement noise in the control signal for the current observer parameters is evident, although it appeared to have small effects in the overall system response for simulations considering the expected noise levels delivered by the vibrometer.
4.3.4/
Conclusion
After the controller synthesis using the two selected methods, some important points should be highlighted. The H∞ controller depends directly on the choice of weighting functions and may show a large sensibility to it, where small variations in one of these functions can have large impact on the optimality index γ, resulting in large changes over the response. This is one of the main drawbacks of this method, and the fine-tuning process required to improve a response characteristics may be troublesome. In this specific work, the found γ value of 2.1 indicates that the design requirements are not fully met, and therefore a compromise should be made in its performance.
104
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
−10
0
−20 −20
−40
Magnitude (dB)
Magnitude (dB)
−30
−50 −60 −70 −80 −90
−110
1
10
−60
−80
Nominal Worst case
−100
Nominal Worst case
−100
−40
2
10
3
−120
10
Frequency (Hz)
(a) H∞ controller
2
10
3
10
4
10
Frequency (Hz)
(b) ESO controller
Figure 4.15: Closed-loop Bode diagram, from measurement noise n to control signal u, considering the nominal plant response (blue continuous line) and the worst case result (red, dashed line) achievable within the defined parametric variation range.
The ESO methodology is based on a different approach. Observer and feedback control can be designed separately. The observer estimates in real time the difference between the real plant and a nominal model. This difference, including effects of external disturbances and unmodeled plant dynamics, is compensated and results in a plant behaving as the nominal, undisturbed model. The feedback control, designed for this ideal case, gives the dynamic characteristics for the closed loop behavior. When considering high-gain observers, its convergence is made sufficiently fast through the selection of a single parameter ε. It may be selected so the system can estimate static and dynamic disturbances and behaviors. Through simulation analysis, the two controllers have shown to be robustly stable with respect to the parametric variations in the operational range. Considering performances, the H∞ controller, despite not being able to precisely achieve the initial requirements, approaches them sufficiently to be regarded as successful. It achieved adequate disturbance attenuation characteristics and a sufficiently small tracking error (less than 0.2 %). In contrast, the EHGO could achieve zero error in tracking and a faster settling time. The proactive disturbance attenuation strategy obtained improved responses in relation to the first controller, specially around the system natural vibration frequencies.
4.4/ Experimental validation and results As a final stage of controller development, an experimental validation was performed to ratify the simulations. For this purpose, several scenarios were devised to check stability, robustness and disturbance rejection characteristics. A scheme and the real setup for the experimental closed loop implementation are shown in Figures 4.16(a) and 4.16(b), respectively. The tests were performed inside the SEM vacuum chamber at pressures between 6e−4 and 2e−4 Pa. During the experiments, the micro-gripper FTG30 was fixed over the SEM’s stage. The vibrometer, also placed inside the SEM, measured displacements over the micro-gripper actuated finger (Figure 4.16(c)). The controllers, designed in continuous time, were implemented in Simulink® software
4.4. EXPERIMENTAL VALIDATION AND RESULTS
105
Supervisory system
Scanning electron microscope
FT-G30 Power amp.
dSPACE 1103
D/A
Vibrometer’s head
OUT
A/D
IN
Optical fiber I/O comm
Analog output
Vibrometer’s CPU
(a) Scheme detailing the interconnection between elements.
Control unit + power amp.
SEM
Vibrometer’s head
Supervisory PC
FT-G30 micro-gripper
SEM positioning stage
Vibrometer’s CPU
(b) Experimental setup: exterior
(c) Experimental setup: inside the SEM
Figure 4.16: Experimental setup for validation of closed-loop control strategies.
and uploaded to a dSPACE® card responsible for handling conversions between digital and analog signals. The generated control signal computed by the card were amplified to produce the voltage levels required by the comb-drive for its operation. Finally, a supervisory system, communicating with the control board, allowed the operator to activate the controllers, to define set points and test parameters, and to store various variables of interest. Before starting the experiments, dSPACE sampling frequency needed to be specified. During tests considering the H∞ controller, this frequency was set to 10 kHz, a constraint imposed by the available hardware to ensure the real time computation for this 14th order controller. The same hardware was able to achieve, for the ESO based controller, update frequencies up to 80 kHz, indicating a significant difference in their required computational power.
106
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
4.4.1/
Reference tracking and steady state error under environmental disturbances
To confirm the closed loop stability for the whole operation range, a series of step inputs with different amplitudes was used as reference signal. Due to measurement range limits imposed by the vibrometer, two sets of measurements were performed. One comprised steps between 50 nm and 2.5 µm (using the vibrometer analog output gain G1, with a measurement range of 2.6 µm and a resolution of 0.3 nm), and another with reference steps ranging from 500 nm to 25 µm (using vibrometer gain G3, range up to 40 µm and a resolution of 4 nm). Figure 4.17 displays the obtained responses and respective control signals, normalized and juxtaposed against each other. Besides verifying their stability, this analysis allowed to estimate the influence of parametric variations in each controller, so they could be compared with the simulated values (shown in Figure 4.13). The experimental results are well within the range of response obtained in simulation, with experimental settling times, approximated due to the present of measurement noise, between 28 and 33 ms for the H∞ controller and close to 4 ms for the ESO based controller.
1
Normalized displacement
Normalized displacement
1
0.8
0.6
0.4
0.2
0 0
0.01
0.02
0.03
0.04
0.05
0.8
0.6
0.4
0.2
0 0
0.06
0.002
Time (s)
0.008
0.01
(b) ESO - normalized displacement
1
1
0.8
0.8
Normalized Control
Normalized control
0.006
Time (s)
(a) H∞ - normalized displacement
0.6
0.4
0.6
0.4
0.2
0.2
0 0
0.004
0.01
0.02
0.03
0.04
Time (s)
(c) H∞ - normalized control
0.05
0.06
0 0
0.002
0.004
0.006
0.008
0.01
Time (s)
(d) ESO - normalized control
Figure 4.17: Experimental response to step reference change of different amplitudes (between 50 nanometers and 25 micrometers). The responses are normalized, to contrast differences due to parametric variations. A second test focused on the constant reference regulation, observing how well controllers sustain a desired fixed position when confronted with the perturbation levels present in our SEM room. The measurements were performed at different operation points (constant ref-
4.4. EXPERIMENTAL VALIDATION AND RESULTS
107
erences signals between 0 to 50 V), using the vibrometer analog output gain G1 to obtain measurements with sub-nanometric resolution. Each test set, considering different controllers and operation points, was performed five times, and RMS values for the displacements were computed and averaged. The results are summarized in Table 4.2, and compared against open loop responses in air (performed over an anti-vibration table) and in vacuum. Average RMS disturbance (in nm) Test condition Open loop (air) Open loop (vacuum) Closed loop (H∞ vacuum) Closed loop (ES O vacuum)
0V
10V
20V
30V
40V
50V
2.7 5.4 2.2 0.61
2.5 4.4 2.3 0.59
2.3 4.3 2.8 0.62
3.2 29.4 3.7 0.61
3.1 89.9 4.9 0.63
5.1 14.2 4.2 0.63
Table 4.2: Measured average RMS disturbance for the micro-gripper’s tip, for different constant input voltage values, considering four operation conditions.
From the obtained results, several interesting points arise. Firstly, it was possible to notice how open loop responses in vacuum presented larger RMS displacements values than when the same experiment was performed in air. Two factors may have contributed for this result: the increased sensitivity due to the lower damping in vacuum, and the improved performance of the anti-vibration table when compared to the SEM structure and dampers. Also, effects of input electronic disturbance (demonstrated in Section 3.3.3) can be easily noticed, resulting in the large displacements for 30, 40 and 50 Volts when in open loop in vacuum. In this context, responses to lower voltages were less affected by it. At lower voltages, it can be assumed that vibrations were mostly produced by other external sources (ground vibration, acoustic, ...), as seen Chapter 2. These values exemplify the vibration levels commonly found for this setup during real applications in SEM . The degradation due to electronic input disturbance at larger voltages is harsher in vacuum, where the reduced damping and the first harmonic excitation can generate displacements in the order of hundreds of nanometers. In this case, it becomes the most important source of disturbance in the system, while the same effect in air is limited to a few nanometers thanks to the damping provided by the air. It is important to notice that, as the cantilever stiffness varies with the pressure, this harmonic excitation will occur at different operation voltages. In this example, the frequency match in vacuum is located close to 40 V. A further increase in the input voltage in vacuum reduced the disturbance amplitude, as the micro-gripper’s resonant frequency shifted. The H∞ controller was able to attenuate disturbances significantly. At lower voltages, the RMS displacement in vacuum attained values compatibles to those obtained in air over the antivibration table, with RMS vibration levels between 35 to 60% lower. At higher voltages, when the electronic disturbance became dominant, the attained RMS level was slightly larger, yet displaying a great improvement when compared to the open loop response in vacuum. Overall, this controller appears to be able to operate within a 5 nm RMS displacement margin in the worst case scenario. The EHGO controller, in other hand, achieved an almost constant RMS value around 0.6 nm for all measured range, being less affected by the presence of input electronic noise. This result is impressive, as the controller can keep the tracking error near the sensor resolution level for the whole operation range, even in the face of unexpected, dynamic disturbances.
108
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
Control signal (V)
Displacement (nm)
A direct comparison between controller responses to a constant reference signal in the time domain is shown in Figure 4.18. The curves were obtained around 40 V, the operation point with the largest vibrations due to disturbances. The top graph shows the position variations around this operation point, and the bottom graph reveals the applied control signals. The control signal generated by the EHGO controller presented quicker transitions and was able to track closer the reference with lower amplitude oscillation. This also resulted in a less smooth control signal, with the presence of random peaks due to measurement noise. The noise influence was still too small to introduce significant degradation in the response, yet any increase in the observer gain should be made with caution. In other hand, the H∞ controller presents a much smoother, conservative control signal.
10 5 0 −5
H∞ ESO
−10 0
0.02
0.04
0.06
0.08
0.1
0.02
0.04
0.06
0.08
0.1
40 39.99 39.98 39.97 39.96 0
ESO H∞
Time (s) Figure 4.18: Experimental closed-loop response to a constant reference signal, and control signal u for each controller. No additional external disturbance is added.
4.4.2/
Disturbance rejection for external mechanical and acoustic perturbations
The previous tests considered only background perturbations that existed in the environment and the electronic noise introduced by the control hardware. To further explore the attenuation feature of the controllers, artificial disturbances were introduced. To test the effects of external acoustic excitations in closed-loop, a loudspeaker was placed in the room at 1 meter from the SEM chamber and directed at it. A signal generator was used to produce sinusoidal waves in different frequencies and amplitudes, as performed during the noise characterization phase in Section 2.2.2.2. The obtained curves are summarized in Figure 4.19. The graph compares the controllers under external excitation of a high-amplitude noise (approximately 70 dB), with frequency close to the first vibration mode. At the start, the controllers were disabled, allowing the finger to oscillate with increasing amplitudes. The operation point was selected around 3 V, to ensure control signals would not be negative. The EHGO controller, activated at t = 5.5 seconds, quickly attenuated the
4.4. EXPERIMENTAL VALIDATION AND RESULTS
109
80 H∞ ESO
60
Displacement (nm)
40 20 0 −20 −40 −60 −80 4.5
5
5.5
6
6.5
7
Time (s)
Figure 4.19: Experimental responses for external acoustic excitations, considering the two controllers. The disturbance signal frequency is 1080 Hz with an intensity of 70 dB. The controllers are disabled at the start, to better illustrate their efficacy.
acoustic effects, achieving steady state values similar to those obtained previously (0.64 nm RMS of vibration). The H∞ controller was activated near t = 6 seconds, and after a short transitory phase, steady state was achieved with a RMS displacement value of 2.4 nm. This demonstrates their capacity to quickly attenuate the effects of sustained acoustic excitations over the tested endeffector. The last experiment performed consisted of an impact test. A small rigid plastic half-sphere, with a radius of 2 centimeters and mass m = 10.8 grams, was dropped directly 2 cm besides the vacuum chamber and over the SEM table, so vibrations could be transmitted to the SEM interior more easily. The half-sphere was dropped from a height h0 of approximately 3 cm 3 centimeters (Figure 4.20), to mimic an impulsive disturbance exciting a wide frequency range. During tests, the half-sphere bounced back after the first impact, resulting in multiple consecutive impulsive excitations. Figure 4.20: Impact test example, showing the Even if the resulting impact force sphere and approximated drop height. could not be directly measured, its value was estimated from the impact velocity and the contact time between surfaces for the first impact by the relation F = (2mv)/ti . The half-sphere p velocity at the instant of impact was estimated using v = 2h0 g = 0.76 m/s. The impact time ti was approximated by recording the experiment on video at 60 frames per second. Through the study of these images, the moment of impact could be determined to occur between frames, ant therefore its duration was less than 1/60 of a second. If the impact time is taken as ti = 0.01 second,
110
CHAPTER 4. POSITION CONTROL IN THE MICRO-WORLD
an approximated force value of F = 1.6 N is obtained. The response obtained for the open loop test is shown in Figure 4.21(a). The first impact induced a displacement of over 50 nm, followed by a second impact and the sustained vibration due to the excitation of the low damped first mode. Repeating the experiment with the controllers in closed loop yields the curves shown in Figure 4.21(b). The ESO controller greatly reduced the disturbance effects, with a more noticeable degradation at the moment of first and subsequent impacts, when a maximum peak displacement of 11 nm was measured. Nevertheless, the vibrations were quickly attenuated (0.1 seconds). Regarding the H∞ controller, its response reveals the initial impact amplitude is only slightly attenuated in relation to the open loop case, with the subsequent vibrations suppressed in around 0.5 seconds. This test corroborates the excellent disturbance attenuation of the EHGO structure. The H∞ controller still presented a considerable reduction in the settling time, but its limited effectiveness for this special case of disturbance is clear. 40
60
H∞ ESO
30 40
Displacement (nm)
Displacement (nm)
20 20
0
−20
10 0 −10 −20
−40 −30 −60 0
1
2
3
4
5
Time (s)
(a) Open loop
6
7
8
−40 0
0.2
0.4
0.6
0.8
1
Time (s)
(b) Closed loop
Figure 4.21: Experimental impact test response, when a half-sphere of mass 10.8 grams is dropped near the microscope to mimic an impulsive disturbance on the system.
4.4.3/
Conclusion
The validation process presented in this section confirmed the expected results from simulation. Both controllers behaved as designed, were stable in the grippers operational range and improved, to different degrees, the achievable performance in relation to the open loop. The disturbance attenuation characteristics presented by the ESO are a strong argument in favor or this method, achieving accuracies in the sub-nanometric range (close to sensor’s resolution values), while the H∞ controller obtained values ranging from 2.2 to 4.9 nanometers RMS when considering natural background vibrations and electronic input perturbation, the later playing an important role in the experimental validation process. Although it could not be removed within the current hardware implementation, future works should consider the use of new devices to reduce/eliminate it, an essential step for real nano-positioning applications. Regardless, the remarkable ESO properties allowed countering its effects. The H∞ controller was also able to greatly reduced its impact on the gripper’s position. However, the existence of this disturbance needed to be explicitly considered when constructing the weighting functions. When additional external disturbances were applied to the system (acoustic and mechanical impact), both controllers were capable of attenuating the first vibration mode excitations,
4.5. CONCLUSION
111
greatly improving the system performance and settling times. However, the vibration amplitudes produced by this test varied significantly, with the ESO based controller presented 4 times less peak-to-peak displacements than the H∞ controller.
4.5/
Conclusion
This chapter described the issue of precision control of nano-positioning devices inside the SEM. In literature, various methods were proposed to control micro- and nano-positioning devices. However, few was said when considering control strategies for similar systems in vacuum. The challenges restricting the precise micro- and nano-manipulation in SEM are strongly linked to the obstacles in obtaining fast and accurate positional information over components located in the vacuum chamber. Difficulties in instrumenting and measuring devices and components of interest (due to the operation environment and limited work volume), and the strict requirements in resolution and dynamic range demanded by sensing devices (the reduced dimension of objects, unfavorable signal/noise ratios) contribute to it. Thanks to the available setup, consisting on a fast acquisition, sub-nanometric resolution vibrometer placed inside the SEM’s specimen chamber, the implementation of real-time, dynamic control laws in this environment and the improvements brought by it to precise position control could be demonstrated. Two control strategies were carried out in this work. The H∞ controller was computed based on an optimization process, where different weighting functions were selected to achieve desired closed loop characteristics, attenuation specific disturbance frequencies and limiting the control signal. The ESO controller, developed more recently in the late 1990’s and still under study, is based on the on-line reconstruction and compensation of external disturbances and plant uncertainties. Both methods are regarded as robust and able to improve the overall system responses and disturbance attenuation characteristics. The H∞ synthesis required reasonable precise knowledge over the system and its disturbances, and the selection of weighting functions can be a complex task when requirements are too strict and conflicting. In the ESO framework, the use of high-gain method for the observer reduced the total number of parameters to be selected, in this studied case, to four: the feedback proportional and derivative gains, the reference filter pole Fre f , and the observer gain ε. The need of only an approximated model is an interesting feature for practical applications, as it can greatly simplify the identification and controller synthesis processes. Overall, the EHGO controller delivered better reference tracking and disturbance rejection characteristics, as simulations indicated this framework as presenting larger robustness margins. Using real-time measurements, this control structure consistently achieved nanometric resolution, even under the effects of disturbances. This consists of an important and promising result to be further explored, that may lead to more accurate, faster, and more accessible nano-manipulation.
5 Conclusions and perspectives
This work aimed to explore one of the current challenges in nanotechnology: the correct and safe manipulation of small and fragile structures. It was seen that, when approaching these structures to perform tasks such as characterization and grasping, precise position control of the end-effector is essential, specially when dealing with nano-objects. Various examples of manipulation in SEMs, scattered in the literature, indicated that the precision and accuracy limits of currently available methods may be not sufficient to ensure the repeatability and success rates required for characterizing and assembling structures in this scale, restricting their applications. Failing to achieve the required positional precision can result in broken components, unreliable measurements and large variations on the tasks outcome. This work main contributions towards improving the quality of nano-manipulations in the SEM environment are summarized next, together with perspectives and ideas for future developments.
5.1/
Conclusions
In this thesis, the difficulty to measure and account for system dynamics on the vacuum chamber was pointed as one of the major limitation on SEM manipulation. This work proposed the use of a dedicated vibrometer, placed inside the vacuum chamber, to directly measure displacements of samples and end-effectors. This allowed to capture data about the position dynamics of components in real time. The sensor characterization process indicated that it could capture displacements in the sub-nanometric range with acquisition frequencies of tens of kHertz, being therefore an effective tool for this investigation. The study of mechanical disturbances in SEM revealed that, despite operating in a vacuum chamber placed over an active damping system and other efforts to reduce perturbations in the room, components were still affected by surrounding vibrations. The identified vibrations included those produced by the secondary vacuum pump and those induced over the SEM positioning stage. Although these vibrations resulted in small displacements, they could become important if their frequencies match the oscillation modes of components in the chamber (samples, actuators, endeffectors, ...). During experimental tests, perturbations from the environment achieved RMS values close to 2.55 nm for a silicon cantilever with 5 mm length, and up to 5 nm RMS for the considered FTG-30 micro-gripper, significant values when nanometric precisions are desired. This effect became clear when adding a source of acoustic perturbation, transmitted to the interior of the SEM via mechanical coupling. When these excitations matched the first mode of a sample, large displacements could be produced. From experiments, vibrations amplitudes over 50 nm could be easily produced by the external acoustic noise source, and even low intensity noises (50 dB) were sufficient to significantly degrade the response, with oscillations close to 20 nm peak-to-peak. 113
114
CHAPTER 5. CONCLUSIONS AND PERSPECTIVES
The analysis of pressure influence was performed by taking an electrostatic micro-gripper as a case-of-study. Its actuated finger was characterized both in air and vacuum. During experiments, damping presented the most noticeable variation, where its settling time for a voltage step input changed from 0.1 second in air to more than 10 seconds in the vacuum. This indicates the air drag was the responsible for dissipating the major part of the energy in this micro-system and corresponded to a near 100 times increase in the quality factor. Also, changes in the vibration frequencies were observed due to pressure influence, where the first mode presented up to 8.2% variation on its frequency between both environments. The full characterization of the actuated finger revealed the non-linear nature of its behavior, with parameters varying according to its operation point, and the presence of a non-linear stiffness became more evident in the vacuum. Through the acquired data, general models for the gripper operating in air and vacuum were derived and validated. It is important to stress that some of the effects shown in vacuum may be difficult to predict accurately, requiring extensive analysis and simulations, while an experimental study in vacuum, as presented here, can be laborious, time consuming, and require special instrumentation and components. After considering the disturbance effects in SEM and the micro-gripper characteristics in vacuum obtained during the previous stages of this work, the necessity of improved position control in this environment inspired the development of two strategies: one based on the H∞ methodology, and other based on the Extended State Observer approach. While the first took advantage of the available information over the system dynamics, its parametric variations and existing disturbances, the second presented a structure less dependent on models and based on realtime estimation of plants disturbances and uncertainties. Both controllers succeeded in obtaining improved response for reference step signals, with settling times of 30 ms for the H∞ and 4 ms for the ESO and both presenting zero overshoot. More interesting, the disturbance attenuation properties achieved on both controllers allowed to greatly improve its performance when compared with open-loop. The designed ESO controller presented encouraging results, capable of achieving a RMS displacement level under 0.64 nm for all operation range, even under the effects of constant acoustic disturbance and electronic input noises. These results demonstrated the importance of taking into consideration the system dynamics when high-accuracy positioning tasks are required, and how overlooking disturbances and its effects can produce large degradation in the final position. The performed test revealed that vibrations levels up to 15 nm peak-to-peak were not uncommon for the case of study considered, achieving even higher values (100 nanometers or above) under specific acoustic excitations. These vibrations attenuation is essential for tasks requiring high precision and accuracies. From the control point of view, the ESO framework offers an interesting alternative to speed up the development of positioning controllers, which is interesting for practical applications, given that reliable and fast sensor information is available. Its robustness to plant uncertainty and disturbance attenuation characteristics could be further explored for tasks in micro and nano-scale, where both uncertainty and perturbations are present and largely influential in the outcome.
5.2/ Future perspectives The experimental setup employed for this thesis served to demonstrate how real-time information over the system dynamics can improve the precise positioning task in the SEM environment. However, the actual implementation, serving as a proof of concept, lacked in flexibility and some problems were encountered. To improve these limitations, a few modifications are necessary.
5.2. FUTURE PERSPECTIVES
115
The sensor used had a considerable size (9 cm x 9 cm x 4 cm for the sensing head) and required to be positioned over a rigid, large support, limiting the operations that could be realized in the SEM. The miniaturization of the vibrometer head is essential to facilitate its positioning inside the chamber, minimizing the required space and its interference with other components. Commercial devices with compact sensor heads (few millimeters in diameter and less than 2 cm in length) are available, as well as the possibility to developing custom miniaturized devices for this application. The sensor fixed support could be changed for a robotic structure, increasing the setup flexibility. A robotic arm (Figure 5.1) was designed specifically for the SEM used during this thesis and possesses 6 DoF (3 translations in the arm and 3 rotations in the wrist). It is fixed at the vacuum chamber’s ceiling and can be extended and retracted, reducing its interference during other operations. This setup could be used as a flexible measurement device to characterize the vibrations and dynamic behaviors of various components in SEM.
Robotic arm
Electron emitter
Focus ion beam (FIB) emitter
Gas injector
SEM stage
Figure 5.1: CAD representation of a serial 6 DoF robot for SEM developed at the FEMTO-ST laboratory, showing how it is positioned inside the SEM chamber. In addition, modifications should be considered to eliminate or at least reduce the electronic input disturbance. As seen in the electrostatic micro-gripper identification phase, these disturbances could be dominant for an important part of the operation range. Tests should be performed with other processing boards and signal converters to find one suited for delivering noise and disturbances compatible with this application. The results over disturbance influence can be explored to other tasks in SEM such as fabrication via FIB/e-beam lithography, where vibrations in the stage can reduce their attainable resolution. It is evident how SEM room preparation require special care, and the development of remote supervision strategies to eliminate non-essential interferences in the environment can be contemplated. In this situation, the SEM controlling computer, pumps and all robotic interface components (i.e. joysticks, ...) can be installed farther from the microscope, in a separated room. In the scope of the NanoRobust project, the results obtained are expected to be integrated in a prototype for complete autonomous SEM manipulation cell. The works developed during this thesis would be used at the approach phase and characterization in real-world problems, and would serve as a final validation for this strategy. Combining the precise positioning strategy for the approach phases and force control during component handling could improve manipulation and characterization tasks, while reducing the risks of damaging components. A possible complete setup to be explored includes the use of end-effectors to counter vibrations and displacement errors
116
CHAPTER 5. CONCLUSIONS AND PERSPECTIVES
for the robotic arm or other positioning elements. In this case, coarse positioning of a robotic structure may be obtained via standard methods (i.e. visual servoing), while fine position control of the end-effector compensating any undesired effects present at the extremity. Figure 5.2 exemplifies a real problem to be tackled: the characterization and manipulation of a nano-helix. So far, only traction tests were made with aid of AFM tips, where the pads at each extremity were glued to testing probes and displacements were measured via image processing techniques. The use of grippers with force-sensing capabilities could simplify the whole manipulation processing, while allowing to perform compression test with the aid of real time positioning data.
Figure 5.2: Nano-helix between pads, to facilitate manipulation tasks without damaging the sample. Fabricated and kindly offered by the Laboratory of Photonics and Nanostructures (LPN) in Marcoussis, France.
Personal publications
[Boudaoud et al., 2014a] Boudaoud, M., De Faria, M. G., Haddab, Y., Haliyo, S., Le Gorrec, Y.,
Lutz, P., et R´egnier, S. (2014a). Robust microscale grasping using a self scheduled dynamic controller. In World Congress, volume 19, pages 7492–7498. [Boudaoud et al., 2015] Boudaoud, M., De Faria, M. G., Haddab, Y., Haliyo, S., Le Gorrec, Y.,
Lutz, P., et R´egnier, S. (2015). Robust microscale grasping through a multimodel design: synthesis and real time implementation. Control Engineering Practice, 39:12–22. [Boudaoud et al., 2014b] Boudaoud, M., De Faria, M. G., Le Gorrec, Y., Haddab, Y., et Lutz, P.
(2014b). An output feedback lpv control strategy of a nonlinear electrostatic microgripper through a singular implicit modeling. Control Engineering Practice, 28:97–111. [de Faria et al., 2015a] de Faria, M. G., Haddab, Y., Le Gorrec, Y., et Lutz, P. (2015a). Extended
high-gain obserer for robust position control of a micro-gripper in air and vacuum. In IEEE Conference on Automation Science and Engineering (CASE). [de Faria et al., 2015b] de Faria, M. G., Haddab, Y., Le Gorrec, Y., et Lutz, P. (2015b). Influence
of mechanical noise inside a scanning electron microscope. Review of Scientific Instruments, 86(4):045105.
117
References
[Aur, ] Auriga 60 Modular Crossbeam workstation - Installation requirements manual. Carl
Zeiss Microscopy. [kle, ] Kleindiek MM3A-EM nano prober. http://www.nanotechnik.com/mm3a-em.html. (ac-
cessed September 16, 2015). [pi, ] P-911KNMV
UHV-Compatible Miniature Piezo http://www.physikinstrumente.com/product-detail-page/p-911knmv-700850.html. September 16, 2015).
Hexapod. (accessed
[sma, ] SmarAct 15D manipulator. http://www.smaract.de/index.php/manipulators. (accessed
September 16, 2015). [zyv, ] Zyvex nProber. http://www.zyvex.com/Documents/nProber.pdf. (accessed September 16,
2015). [SP1, 2012] (2012). Plane-mirror Miniature Interferometer Series SP120/2000 User’s Guide.
SIOS Meßtechnik Gmbh, 4 edition. [Sma, 2013] (2013). MCS - Mudular Control System Programmer’s Guide. Smaract GmbH. [Abrahamians et al., 2014] Abrahamians, J.-O., Sauvet, B., Polesel-Maris, J., Braive, R., et Reg-
nier, S. (2014). A nanorobotic system for in situ stiffness measurements on membranes. Robotics, IEEE Transactions on, 30(1):119–124. [Acosta et al., 2011] Acosta, J. C., Hwang, G., Polesel-Maris, J., et R´egnier, S. (2011). A tuning
fork based wide range mechanical characterization tool with nanorobotic manipulators inside a scanning electron microscope. Review of Scientific Instruments, 82(3):035116. [Agnus et al., 2005] Agnus, J., Nectoux, P., et Chaillet, N. (2005). Overview of microgrippers
and design of a micromanipulation station based on a mmoc microgripper. In Computational Intelligence in Robotics and Automation, 2005. CIRA 2005. Proceedings. 2005 IEEE International Symposium on, pages 117–123. IEEE. [Akita et al., 2002] Akita, S., et Nakayama, Y. (2002). Manipulation of nanomaterial by carbon
nanotube nanotweezers in scanning probe microscope. Japanese journal of applied physics, 41(6S):4242. [Aoki et al., 2008] Aoki, K., Guimard, D., Nishioka, M., Nomura, M., Iwamoto, S., et Arakawa, Y.
(2008). Coupling of quantum-dot light emission with a three-dimensional photonic-crystal nanocavity. Nature Photonics, 2(11):688–692. [Aphale et al., 2007] Aphale, S., Fleming, A., et Moheimani, S. (2007). High speed nano-scale
positioning using a piezoelectric tube actuator with active shunt control. Micro & Nano Letters, 2(1):9–12. 119
120
REFERENCES
[Arai et al., 1996] Arai, F., Andou, D., et Fukuda, T. (1996). Adhesion forces reduction for
micro manipulation based on micro physics. In Micro Electro Mechanical Systems, 1996, MEMS’96, Proceedings. An Investigation of Micro Structures, Sensors, Actuators, Machines and Systems. IEEE, The Ninth Annual International Workshop on, pages 354–359. IEEE. A separation principle for the stabilization of a class of nonlinear systems. IEEE Transactions on Automatic Control, 44(9):1672–1687.
[Atassi et al., 1999] Atassi, A. N., et Khalil, H. K. (1999).
[Aten et al., 2014] Aten, Q. T., Jensen, B. D., Burnett, S. H., et Howell, L. L. (2014). A self-
reconfiguring metamorphic nanoinjector for injection into mouse zygotes. Review of Scientific Instruments, 85(5):055005. [Athans, 1986] Athans, M. (1986). A tutorial on the LQG/LTR method. Citeseer. [Babic et al., 2003] Babic, B., Furer, J., Sahoo, S., Farhangfar, S., et Sch¨onenberger, C. (2003).
Intrinsic thermal vibrations of suspended doubly clamped single-wall carbon nanotubes. Nano Letters, 3(11):1577–1580. [Beyeler et al., 2009] Beyeler, F., Muntwyler, S., et Nelson, B. J. (2009). A six-axis mems force–
torque sensor with micro-newton and nano-newtonmeter resolution. Microelectromechanical Systems, Journal of, 18(2):433–441. [Beyeler et al., 2007] Beyeler, F., Neild, A., Oberti, S., Bell, D. J., Sun, Y., Dual, J., et Nelson, B. J.
(2007). Monolithically fabricated microgripper with integrated force sensor for manipulating microobjects and biological cells aligned in an ultrasonic field. Microelectromechanical Systems, Journal of, 16(1):7–15. [Bibel et al., 1992] Bibel, J. E., et Malyevac, D. S. (1992). Guidelines for the selection of weight-
ing functions for h-infinity control. Technical Report, DTIC Document. [Borovic et al., 2005] Borovic, B., Liu, A., Popa, D., Cai, H., et Lewis, F. (2005). Open-loop
versus closed-loop control of mems devices: choices and issues. Journal of Micromechanics and Microengineering, 15(10):1917. Caracterisation dynamique des bruits a lechelle nanometrique et commande robuste LPV de systemes de micromanipulation. PhD thesis, Universite de Franche-Comte, Besancon, France.
[Boudaoud, 2012] Boudaoud, M. (2012).
[Boudaoud et al., 2010] Boudaoud, M., Haddab, Y., et Le Gorrec, Y. (2010). Modelling of a
mems-based microgripper: application to dexterous micromanipulation. In Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on, pages 5634–5639. IEEE. [Boudaoud et al., 2013] Boudaoud, M., Haddab, Y., et Le Gorrec, Y. (2013). Modeling and op-
timal force control of a nonlinear electrostatic microgripper. Mechatronics, IEEE/ASME Transactions on, 18(3):1130–1139. [Boukhnifer et al., 2007] Boukhnifer, M., et Ferreira, A. (2007). H∞ loop shaping bilateral con-
troller for a two-fingered tele-micromanipulation system. Control Systems Technology, IEEE Transactions on, 15(5):891–905. [Bouscaud et al., 2012] Bouscaud, D., Pesci, R., Berveiller, S., et Patoor, E. (2012). Estimation of
the electron beam-induced specimen heating and the emitted x-rays spatial resolution by Kossel microdiffraction in a scanning electron microscope. Ultramicroscopy, 115:115–119.
REFERENCES
121
[Braatz et al., 1994] Braatz, R. P., Young, P. M., Doyle, J. C., et Morari, M. (1994). Computa-
tional complexity of µ calculation. Automatic Control, IEEE Transactions on, 39(5):1000– 1002. [Busch et al., 2004] Busch, K., L¨olkes, S., Wehrspohn, R. B., et F¨oll, H. (2004). Photonic crystals.
New York: Willey, pages 241–247. [Carrozza et al., 2000] Carrozza, M. C., Eisinberg, A., Menciassi, A., Campolo, D., Micera, S.,
et Dario, P. (2000). Towards a force-controlled microgripper for assembling biomedical microdevices. Journal of Micromechanics and Microengineering, 10(2):271. [Chen et al., 2009] Chen, B. K., Zhang, Y., et Sun, Y. (2009). Active release of microobjects
using a mems microgripper to overcome adhesion forces. Microelectromechanical Systems, Journal of, 18(3):652–659. [Chen et al., 2003] Chen, C.-S., et Kuo, W.-J. (2003). Squeeze and viscous dampings in micro
electrostatic comb drives. Sensors and Actuators A: Physical, 107(2):193–203. [Chu et al., 1994] Chu, P. B., et Pister, K. S. (1994). Analysis of closed-loop control of parallel-
plate electrostatic microgrippers. In Robotics and Automation, 1994. Proceedings., 1994 IEEE International Conference on, pages 820–825. IEEE. [Chu, 1994] Chu, W.-H. M. (1994). Microfabricated tweezers with a large gripping force and
a large range of motion. PhD thesis, Case Western Reserve University. [Ciubotariu, 2015] Ciubotariu, A. (2015). Pmn-pt based integrated actuator for 6 dof micro-
gripper. Technical Report, FEMTO-ST Institute. [Cl´evy et al., 2006] Cl´evy, C., et Chaillet, N. (2006). Micromanipulation and micro-assembly
systems. In IEEE/RAS International Advanced Robotics Programm, IARP’06., pages sur–CD. [Cui, 2009] Cui, Z. (2009). Nanofabrication: principles, capabilities and limits. Springer Sci-
ence & Business Media. [Dahmen et al., 2013] Dahmen, C., et Tiemerding, T. (2013). Fast and robust position determi-
nation in the scanning electron microscope. In Australasian Conference on Robotics and Automation. [de Jonge et al., 2003] de Jonge, N., Lamy, Y., et Kaiser, M. (2003). Controlled mounting of
individual multiwalled carbon nanotubes on support tips. Nano Letters, 3(12):1621–1624. [Dechev et al., 2004] Dechev, N., Cleghorn, W. L., et Mills, J. K. (2004). Microassembly of 3-d
microstructures using a compliant, passive microgripper. Microelectromechanical Systems, Journal of, 13(2):176–189. [Dong et al., 2001] Dong, L., Arai, F., et Fukuda, T. (2001). Three-dimensional nanoassembly
of multi-walled carbon nanotubes through nanorobotic manipulations by using electronbeam-induced deposition. In Nanotechnology, 2001. IEEE-NANO 2001. Proceedings of the 2001 1st IEEE Conference on, pages 93–98. IEEE. [Dong et al., 2002] Dong, L., Arai, F., et Fukuda, T. (2002). Three-dimensional nanorobotic
manipulations of carbon nanotubes. Journal of Robotics and Mechatronics, 14(3):245–252. [Doyle, 1982] Doyle, J. (1982). Analysis of feedback systems with structured uncertainties. In
IEE Proceedings D (Control Theory and Applications), volume 129, pages 242–250. IET.
122
REFERENCES
[Doyle, 1985] Doyle, J. (1985). Structured uncertainty in control system design. In 1985 24th
IEEE Conference on Decision and Control, number 24, pages 260–265. [Doyle et al., 1991] Doyle, J., Packard, A., et Zhou, K. (1991). Review of lfts, lmis, and µ. In
Decision and Control, 1991., Proceedings of the 30th IEEE Conference on, pages 1227–1232. IEEE. [Doyle et al., 1989] Doyle, J. C., Glover, K., Khargonekar, P. P., Francis, B., et others (1989). State-
space solutions to standard h 2 and h∞ control problems. Automatic Control, IEEE Transactions on, 34(8):831–847. [Doyle et al., 1979] Doyle, J. C., et Stein, G. (1979). Robustness with observers. Technical
Report, DTIC Document. [Driesen et al., 2005] Driesen, W., Varidel, T., R´egnier, S., et Breguet, J.-M. (2005). Micro manip-
ulation by adhesion with two collaborating mobile micro robots. Journal of Micromechanics and Microengineering, 15(10):S259. [Egerton et al., 2004] Egerton, R., Li, P., et Malac, M. (2004). Radiation damage in the TEM
and SEM. Micron, 35(6):399–409. [Eisinberg et al., 2001] Eisinberg, A., Menciassi, A., Micera, S., Campolo, D., Carrozza, M., et
Dario, P. (2001). Pi force control of a microgripper for assembling biomedical microdevices. IEE Proceedings-Circuits, Devices and Systems, 148(6):348–352. [Epp et al., 2004] Epp, D. S., Ozdoganlar, O. B., et Sumali, H. (2004). Dynamic measurement of
gas damping effects in MEMS. In Current Proceedings, SEM International Congress X. [Esfandiari et al., 1989] Esfandiari, F., et Khalil, H. (1989). Observer-based control of uncer-
tain linear systems: Recovering state feedback robustness under matching condition. In American Control Conference, 1989, pages 931–936. IEEE. [Esfandiari et al., 1992] Esfandiari, F., et Khalil, H. K. (1992). Output feedback stabilization of
fully linearizable systems. International Journal of control, 56(5):1007–1037. [Fahlbusch et al., 1999] Fahlbusch, S., Fatikow, S., Seyfried, J., et Buerkle, A. (1999). Flexible
microrobotic system miniman: design, actuation principle and control. In Advanced Intelligent Mechatronics, 1999. Proceedings. 1999 IEEE/ASME International Conference on, pages 156–161. IEEE. [Fahlbusch et al., 2005] Fahlbusch, S., Mazerolle, S., Breguet, J.-M., Steinecker, A., Agnus, J.,
P´erez, R., et Michler, J. (2005). Nanomanipulation in a scanning electron microscope. Journal of materials processing technology, 167(2):371–382. [Falvo et al., 1997] Falvo, M., Clary, G., Taylor, R., Chi, V., Brooks, F., Washburn, S., et Su-
perfine, R. (1997). Bending and buckling of carbon nanotubes under large strain. Nature, 389(6651):582–584. [Fatikow, 2007] Fatikow, S. (2007). Automated nanohandling by microrobots. Springer Pub-
lishing Company, Incorporated. [Fatikow et al., 2000] Fatikow, S., Seyfried, J., Buerkle, A., Schmoeckel, F., et others (2000). A
flexible microrobot-based microassembly station. Journal of Intelligent and Robotic Systems, 27(1-2):135–169.
REFERENCES
123
[Fleming et al., 2010] Fleming, A. J., Aphale, S. S., et Moheimani, S. R. (2010). A new method
for robust damping and tracking control of scanning probe microscope positioning stages. Nanotechnology, IEEE Transactions on, 9(4):438–448. [Franklin et al., 2012] Franklin, A. D., Luisier, M., Han, S.-J., Tulevski, G., Breslin, C. M.,
Gignac, L., Lundstrom, M. S., et Haensch, W. (2012). Sub-10 nm carbon nanotube transistor. Nano letters, 12(2):758–762. Performance recovery of feedback-linearization-based designs. Automatic Control, IEEE Transactions on, 53(10):2324–2334.
[Freidovich et al., 2008] Freidovich, L. B., et Khalil, H. K. (2008).
[Friedt et al., 2007] Friedt, J.-M., et Carry, E. (2007). Introduction to the quartz tuning fork.
American Journal of Physics, 75(5):415–422. [Fujioka et al., 1980] Fujioka, H., Nakamae, K., et Ura, K. (1980). Function testing of bipolar
ics and lsis with the stroboscopic scanning electron microscope. Solid-State Circuits, IEEE Journal of, 15(2):177–183. [Fukuda et al., 2002] Fukuda, T., Arai, F., et Dong, L. (2002). Fabrication and property analysis
of mwnt junctions through nanorobotic manipulations. International Journal of Nonlinear Sciences and Numerical Simulation, 3(3-4):753–758. [Fukuda et al., 2003] Fukuda, T., Arai, F., et Dong, L. (2003). Assembly of nanodevices with car-
bon nanotubes through nanorobotic manipulations. Proceedings of the IEEE, 91(11):1803– 1818. [Fukuda et al., 2013] Fukuda, T., Arai, F., et Nakajima, M. (2013). Micro-nanorobotic manipu-
lation systems and their applications. Springer Science & Business Media. [Gahinet et al., 1994] Gahinet, P., et Apkarian, P. (1994). A linear matrix inequality approach
to h∞ control. International journal of robust and nonlinear control, 4(4):421–448. [Gao et al., 2001] Gao, Z., Huang, Y., et Han, J. (2001). An alternative paradigm for control
system design. In Decision and Control, 2001. Proceedings of the 40th IEEE Conference on, volume 5, pages 4578–4585. IEEE. [Gauthier et al., 2011] Gauthier, M., et R´egnier, S. (2011). Robotic micro-assembly. John Wiley
& Sons. [Ghafarirad et al., 2012] Ghafarirad, H., Rezaei, S., Zareinejad, M., Hamdi, M., et Ansari, R. J.
(2012). Robust control with unknown dynamic estimation for multi-axial piezoelectric actuators with coupled dynamics. Comptes Rendus Mecanique, 340(9):646–660. [Gjerde et al., 2006] Gjerde, K., Kjelstrup-Hansen, J., Clausen, C. H., Teo, K. B., Milne, W. I.,
Rubahn, H.-G., et Bøggild, P. (2006). Carbon nanotube forests: a non-stick workbench for nanomanipulation. Nanotechnology, 17(19):4917. [Goldfarb et al., 1999] Goldfarb, M., et Celanovic, N. (1999). A flexure-based gripper for small-
scale manipulation. Robotica, 17(02):181–187. [Gong et al., 2013] Gong, Z., Chen, B. K., Liu, J., et Sun, Y. (2013). Automated nanoprobing
under scanning electron microscopy. In Robotics and Automation (ICRA), 2013 IEEE International Conference on, pages 1433–1438. IEEE.
124
REFERENCES
[Gorman et al., 2006] Gorman, J. J., Kim, Y.-S., et Dagalakis, N. G. (2006). Control of mems
nanopositioners with nano-scale resolution. In ASME 2006 International Mechanical Engineering Congress and Exposition, pages 151–159. American Society of Mechanical Engineers. [Gorman et al., 2007] Gorman, J. J., Kim, Y.-S., Vladar, A. E., et Dagalakis, N. G. (2007). Design
of an on-chip microscale nanoassembly system. International Journal of Nanomanufacturing, 1(6):710–721. [Green et al., 1990] Green, M., Glover, K., Limebeer, D., et Doyle, J. (1990). Aj-spectral factor-
ization approach to h∞. SIAM Journal on Control and Optimization, 28(6):1350–1371. [Grossard et al., 2011] Grossard, M., Boukallel, M., Chaillet, N., et Rotinat-Libersa, C. (2011).
Modeling and robust control strategy for a control-optimized piezoelectric microgripper. Mechatronics, IEEE/ASME Transactions on, 16(4):674–683. [Gu et al., 2014] Gu, G.-Y., Zhu, L.-M., Su, C.-Y., Ding, H., et Fatikow, S. (2014). Modeling
and control of piezo-actuated nanopositioning stages: a survey. IEEE Transactions on Automation Science and Engineering. [Guo et al., 2011] Guo, B.-Z., et Zhao, Z.-l. (2011). On the convergence of an extended state
observer for nonlinear systems with uncertainty. Systems & Control Letters, 60(6):420–430. [Haddab et al., 2000] Haddab, Y., Chaillet, N., et Bourjault, A. (2000). A microgripper using
smart piezoelectric actuators. In Intelligent Robots and Systems, 2000.(IROS 2000). Proceedings. 2000 IEEE/RSJ International Conference on, volume 1, pages 659–664. IEEE. [Hatamura et al., 1990] Hatamura, Y., et Morishita, H. (1990). Direct coupling system between
nanometer world and human world. In Micro Electro Mechanical Systems, 1990. Proceedings, An Investigation of Micro Structures, Sensors, Actuators, Machines and Robots. IEEE, pages 203–208. IEEE. [Hertel et al., 1998] Hertel, T., Martel, R., et Avouris, P. (1998). Manipulation of individual
carbon nanotubes and their interaction with surfaces. The Journal of Physical Chemistry B, 102(6):910–915. [Ho et al., 1998] Ho, C.-M., et Tai, Y.-C. (1998). Micro-electro-mechanical-systems (mems) and
fluid flows. Annual Review of Fluid Mechanics, 30(1):579–612. [Horsley et al., 1998] Horsley, D., Cohn, M. B., Singh, A., Horowitz, R., Pisano, A. P., et others
(1998). Design and fabrication of an angular microactuator for magnetic disk drives. Microelectromechanical Systems, Journal of, 7(2):141–148. [Huang et al., 2009] Huang, S.-J., et Wang, S.-S. (2009). Mechatronics and control of a long-
range nanometer positioning servomechanism. Mechatronics, 19(1):14–28. [Hwang et al., 2009] Hwang, G., Acosta, J. C., Vela, E., Haliyo, S., et R´egnier, S. (2009).
Graphene as thin film infrared optoelectronic sensor. In Optomechatronic Technologies, 2009. ISOT 2009. International Symposium on, pages 169–174. IEEE. [Ishikawa et al., 1993] Ishikawa, H., Dobashi, H., Kodama, T., Furuhashi, T., et Uchikawa, Y.
(1993). Investigation of micro mechanical vibration of piezoelectric actuators using a stroboscopic sem. Journal of electron microscopy, 42(1):35–40.
REFERENCES
125
[Ivan et al., 2013] Ivan, I. A., Ciubotariu, D. A., Cl´evy, C., Lutz, P., et Chaillet, N. (2013). Duo-
bimorph actuator made of pmn-pt [011]: 3d modeling, development and characterization. In Advanced Intelligent Mechatronics (AIM), 2013 IEEE/ASME International Conference on, pages 140–145. IEEE. [Iwaya et al., 2012] Iwaya, K., Shimizu, R., Teramura, A., Sasaki, S., Itagaki, T., et Hitosugi, T.
(2012). Design of an effective vibration isolation system for measurements sensitive to low-frequency vibrations. Journal of Vacuum Science & Technology A, 30(6):063201. [Jasper, 2011] Jasper, D. (2011). SEM-based motion control for automated robotic nanohan-
dling. PhD thesis, Carl von Ossitezky Universit¨at, Oldenburg. [Jasper et al., 2010] Jasper, D., Diederichs, C., Edeler, C., et Fatikow, S. (2010). High-speed
nanorobot position control inside a scanning electron microscope. In Electrical Engineering/Electronics Computer Telecommunications and Information Technology (ECTI-CON), 2010 International Conference on, pages 513–517. IEEE. [Johnson, 1928] Johnson, J. B. (1928). Thermal agitation of electricity in conductors. Physical
review, 32(1):97. [Jung et al., 2012] Jung, K. O., Kim, S. J., et Kim, D. H. (2012). An approach to reducing the
distortion caused by vibration in scanning electron microscope images. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 676:5–17. [Khalil, 2008] Khalil, H. K. (2008). High-gain observers in nonlinear feedback control. In Con-
trol, Automation and Systems, 2008. ICCAS 2008. International Conference on, pages xlvii–lvii. IEEE. [Khalil et al., 2014] Khalil, H. K., et Praly, L. (2014). High-gain observers in nonlinear feedback
control. International Journal of Robust and Nonlinear Control, 24(6):993–1015. [Kim et al., 2003] Kim, K., Lim, S., Lee, I., An, K., Bae, D., Choi, S., Yoo, J., et Lee, Y. (2003). In
situ manipulation and characterizations using nanomanipulators inside a field emissionscanning electron microscope. Review of scientific instruments, 74(9):4021–4025. [Komati et al., 2014] Komati, B., Agnus, J., Cl´evy, C., et Lutz, P. (2014). Prototyping of a highly
performant and integrated piezoresistive force sensor for microscale applications. Journal of Micromechanics and Microengineering, 24(3):035018. [Korayem et al., 2012] Korayem, M., et Esmaeilzadehha, S. (2012). Virtual reality interface for
nano-manipulation based on enhanced images. The International Journal of Advanced Manufacturing Technology, 63(9-12):1153–1166. [Koyano et al., 1996] Koyano, K., et Sato, T. (1996). Micro-object handling system with con-
centrated visual fields and new handling skills. In Photonics East’96, pages 130–140. International Society for Optics and Photonics. [Li et al., 2011] Li, Y., et Xu, Q. (2011). A totally decoupled piezo-driven xyz flexure parallel
micropositioning stage for micro/nanomanipulation. Automation Science and Engineering, IEEE Transactions on, 8(2):265–279. [Li et al., 2012] Li, Y., et Xu, Q. (2012). Design and robust repetitive control of a new parallel-
kinematic xy piezostage for micro/nanomanipulation. Mechatronics, IEEE/ASME Transactions on, 17(6):1120–1132.
126
REFERENCES
[Lim et al., 2005] Lim, S. C., Kim, K. S., Jeong, S. Y., Cho, S., Yoo, J.-E., Lee, Y. H., et others
(2005). Nanomanipulator-assisted fabrication and characterization of carbon nanotubes inside scanning electron microscope. Micron, 36(5):471–476. [Liu et al., 2006] Liu, K.-Z., et He, R. (2006). A simple derivation of are solutions to the stan-
dard h∞ control problem based on lmi solution. Systems & control letters, 55(6):487–493. [Liu et al., 2013] Liu, L., Bai, Y.-G., Zhang, D.-L., et Wu, Z.-G. (2013). Ultra-precision mea-
surement and control of angle motion in piezo-based platforms using strain gauge sensors and a robust composite controller. Sensors, 13(7):9070–9084. ´ [Madonski et al., 2015] Mado´nski, R., et Herman, P. (2015). Survey on methods of increasing
the efficiency of extended state disturbance observers. ISA transactions, 56:18–27. [Matteson et al., 2002] Matteson, T., Schwarz, S., Houge, E., Kempshall, B., et Giannuzzi, L.
(2002). Electron backscattering diffraction investigation of focused ion beam surfaces. Journal of Electronic Materials, 31(1):33–39. [Mazerolle et al., 2005] Mazerolle, S., Breguet, J.-M., Steinecker, A., Agnus, J., P´erez, R., Mich-
ler, J., et others (2005). Nanomanipulation in a scanning electron microscope. Journal of Materials Processing Technology, 167(2):371–382. [Mertens et al., 2003] Mertens, J., Finot, E., Thundat, T., Fabre, A., Nadal, M.-H., Eyraud, V., et
Bourillot, E. (2003). Effects of temperature and pressure on microcantilever resonance response. Ultramicroscopy, 97(1):119–126. [Meyer et al., 2008] Meyer, E., et Braun, H.-G. (2008). Micro-and nanomanipulation inside the
sem. In Journal of Physics: Conference Series, volume 126, page 012074. IOP Publishing. [Miller, 2005] Miller, L. A. (2005). Structural dynamics and resonance in plants with nonlin-
ear stiffness. Journal of theoretical biology, 234(4):511–524. [Mokrane Boudaoud, 2014] Mokrane Boudaoud, S. R. (2014). An overview of gripping force
measurement at the micro and nano-scales using two-fingered microrobotic systems. Int J Adv Robot Syst, 11:45. [Muller et al., 2006] Muller, D., Kirkland, E., Thomas, M., Grazul, J., Fitting, L., et Weyland,
M. (2006). Room design for high-performance electron microscopy. Ultramicroscopy, 106(11):1033–1040. [Nakabayashi et al., 2007] Nakabayashi, D., Silva, P., et Ugarte, D. (2007). Inexpensive two-tip
nanomanipulator for a SEM. Applied Surface Science, 254(1):405–411. [Nakabayashi et al., 2006] Nakabayashi, D., Silva, P. C., Gonz´alez, J. C., Rodrigues, V., et Ugarte,
D. (2006). Low-cost nanomanipulator for in situ experiments in a sem. Microscopy and Microanalysis, 12(04):311–316. [Nakazato et al., 2009] Nakazato, Y., Yuasa, T., Sekine, G., Miyazawa, H., Jin, M., Takeuchi, S.,
Ariga, Y., et Murakawa, M. (2009). Micromanipulation system using scanning electron microscope. Microsystem technologies, 15(6):859–864. [Newbury et al., 2003] Newbury, D. E., Joy, D. C., Echlin, P., Fiori, C. E., et Goldstein, J. I. (2003).
Scanning electron microscopy and X-ray microanalysis. Springer.
REFERENCES
127
[Nguyen, 1995] Nguyen, C. T. (1995). Micromechanical resonators for oscillators and filters.
In Ultrasonics Symposium, 1995. Proceedings., 1995 IEEE, volume 1, pages 489–499. IEEE. [Nordstr¨om Andersen et al., 2009] Nordstr¨om Andersen, K., Petersen, D. H., Carlson, K.,
¨ Horsewell, A., Eichhorn, V., Fatikow, S., et Bøggild, P. (2009). Mølhave, K., Sardan Sukas, O., Multimodal electrothermal silicon microgrippers for nanotube manipulation. IEEE Transactions on Nanotechnology, 8(1):76–85. [Otsuka et al., 2012] Otsuka, H., Nagasaki, Y., et Kataoka, K. (2012). Pegylated nanoparticles
for biological and pharmaceutical applications. Advanced drug delivery reviews, 64:246– 255. [Packard et al., 1993] Packard, A., et Doyle, J. (1993). The complex structured singular value.
Automatica, 29(1):71–109. [Pao et al., 2007] Pao, L. Y., Butterworth, J., Abramovitch, D. Y., et others (2007). Combined
feedforward/feedback control of atomic force microscopes. In American Control Conference, 2007. ACC’07, pages 3509–3515. IEEE. [Park et al., 2010] Park, D. S.-W., Nallani, A. K., Cha, D., Lee, G.-S., Kim, M. J., Skidmore, G.,
Lee, J.-B., et Lee, J.-S. (2010). A sub-micron metallic electrothermal gripper. Microsystem technologies, 16(3):367–373. [Peralta et al., 2013] Peralta, E., Soong, K., England, R., Colby, E., Wu, Z., Montazeri, B.,
McGuinness, C., McNeur, J., Leedle, K., Walz, D., et others (2013). Demonstration of electron acceleration in a laser-driven dielectric microstructure. Nature, 503(7474):91–94. [P´erez et al., 2005] P´erez, R., Agnus, J., Cl´evy, C., Hubert, A., et Chaillet, N. (2005). Modeling,
fabrication, and validation of a high-performance 2-dof piezoactuator for micromanipulation. Mechatronics, IEEE/ASME Transactions on, 10(2):161–171. [Pluska et al., 2009] Pluska, M., Czerwinski, A., Ratajczak, J., Katcki, J., Oskwarek, L., et Rak, R.
(2009). Separation of image-distortion sources and magnetic-field measurement in scanning electron microscope (sem). Micron, 40(1):46–50. [Pluska et al., 2015] Pluska, M., Czerwinski, A., Wzorek, M., Juchniewicz, M., et Katcki, J.
(2015). Identification and reduction of acoustic-noise influence on focused ion beam (fib). Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 348:106–110. [Polyakov et al., 2014] Polyakov, B., Vlassov, S., Dorogin, L. M., Butikova, J., Antsov, M., Oras,
S., L˜ohmus, R., et Kink, I. (2014). Manipulation of nanoparticles of different shapes inside a scanning electron microscope. Beilstein journal of nanotechnology, 5(1):133–140. [Postma et al., 2000] Postma, H. W., Sellmeijer, A., et Dekker, C. (2000). Manipulation and
imaging of individual single-walled carbon nanotubes with an atomic force microscope. Advanced Materials, 12(17):1299–1302. [Radke et al., 2006] Radke, A., et Gao, Z. (2006). A survey of state and disturbance observers
for practitioners. In American Control Conference, 2006, pages 6–pp. IEEE. Combining self-sensing with an unkowninput-observer to estimate the displacement, the force and the state in piezoelectric cantilevered actuators. In American Control Conference (ACC), 2013, pages 4516–4523. IEEE.
[Rakotondrabe, 2013] Rakotondrabe, M. (2013).
128
REFERENCES
[Rakotondrabe et al., 2010] Rakotondrabe, M., Cl´evy, C., et Lutz, P. (2010). Complete open loop
control of hysteretic, creeped, and oscillating piezoelectric cantilevers. Automation Science and Engineering, IEEE Transactions on, 7(3):440–450. [Rana et al., 2014] Rana, M. S., Pota, H. R., et Petersen, I. R. (2014). The design of model
predictive control for an afm and its impact on piezo nonlinearities. European Journal of Control, 20(4):188–198. [R´egnier, 2006] R´egnier, S. (2006). La manipulation aux e´ chelles microscopiques. Habilitation
a` diriger des recherches, Universit´e Pierre et Marie Curie (Paris 6, France). [Rong et al., 2006] Rong, W., Ding, W., Maedler, L., Ruoff, R. S., et Friedlander, S. K. (2006). Me-
chanical properties of nanoparticle chain aggregates by combined afm and sem: Isolated aggregates and networks. Nano letters, 6(12):2646–2655. [Sader et al., 1995] Sader, J. E., Larson, I., Mulvaney, P., et White, L. R. (1995). Method for
the calibration of atomic force microscope cantilevers. Review of Scientific Instruments, 66(7):3789–3798. [Saeidpourazar et al., 2008] Saeidpourazar, R., et Jalili, N. (2008). Towards fused vision and
force robust feedback control of nanorobotic-based manipulation and grasping. Mechatronics, 18(10):566–577. [Salapaka et al., 2002] Salapaka, S., Sebastian, A., Cleveland, J. P., et Salapaka, M. V. (2002).
High bandwidth nano-positioner: A robust control approach. Review of scientific instruments, 73(9):3232–3241. [Salim et al., 1997] Salim, R., Wurmus, H., Harnisch, A., et H¨ulsenberg, D. (1997). Microgrip-
pers created in microstructurable glass. Microsystem technologies, 4(1):32–34. [Sandberg et al., 2005] Sandberg, R., Svendsen, W., Mølhave, K., et Boisen, A. (2005). Temper-
ature and pressure dependence of resonance in multi-layer microcantilevers. Journal of Micromechanics and Microengineering, 15(8):1454. [Sardan et al., 2008] Sardan, O., Eichhorn, V., Petersen, D., Fatikow, S., Sigmund, O., et Bøggild,
P. (2008). Rapid prototyping of nanotube-based devices using topology-optimized microgrippers. Nanotechnology, 19(49):495503. [Scherer, 1992] Scherer, C. (1992). H∞-optimization without assumptions on finite or infinite
zeros. SIAM Journal on Control and Optimization, 30(1):143–166. [Schmoeckel et al., 2001] Schmoeckel, F., et Worn, H. (2001). Remotely controllable mobile mi-
crorobots acting as nano positioners and intelligent tweezers in scanning electron microscopes (sems). In Robotics and Automation, 2001. Proceedings 2001 ICRA. IEEE International Conference on, volume 4, pages 3909–3913. IEEE. [Schrijver et al., 2002] Schrijver, E., et Van Dijk, J. (2002). Disturbance observers for rigid
mechanical systems: equivalence, stability, and design. Journal of Dynamic Systems, Measurement, and Control, 124(4):539–548. [Semonin et al., 2011] Semonin, O. E., Luther, J. M., Choi, S., Chen, H.-Y., Gao, J., Nozik, A. J.,
et Beard, M. C. (2011). Peak external photocurrent quantum efficiency exceeding 100% via meg in a quantum dot solar cell. Science, 334(6062):1530–1533.
REFERENCES
129
[Sharma et al., 2005] Sharma, G., Mavroidis, C., et Ferreira, A. (2005). Virtual reality and hap-
tics in nano-and bionanotechnology. Handbook of theoretical and computational nanotechnology, 10:1–33. [Sievers, 2011] Sievers, T. (2011). Global sensor feedback for automatic nanohandling in-
side a scanning electron microscope. In Intelligent Production Machines and Systems-2nd I* PROMS Virtual International Conference 3-14 July 2006, page 289. Elsevier. [Sitti, 2001] Sitti, M. (2001). Survey of nanomanipulation systems. In Nanotechnology, 2001.
IEEE-NANO 2001. Proceedings of the 2001 1st IEEE Conference on, pages 75–80. IEEE. [Sitti et al., 2001] Sitti, M., Aruk, B., Shintani, H., et Hashimoto, H. (2001). Development of
a scaled teleoperation system for nano scale interaction and manipulation. In Robotics and Automation, 2001. Proceedings 2001 ICRA. IEEE International Conference on, volume 1, pages 860–867. IEEE. [Sitti et al., 2000] Sitti, M., et Hashimoto, H. (2000). Two-dimensional fine particle positioning
under an optical microscope using a piezoresistive cantilever as a manipulator. journal of Micromechatronics, 1(1):25–48. [Skogestad et al., 2007] Skogestad, S., et Postlethwaite, I. (2007). Multivariable feedback con-
trol: analysis and design, volume 2. Wiley New York. [Stievater et al., 2002] Stievater, T., Rabinovich, W., Newman, H., Mahon, R., Goetz, P., Ebel, J., et
McGee, D. (2002). Measurement of thermal-mechanical noise in microelectromechanical systems. Applied physics letters, 81(10):1779–1781. [Stoorvogel, 1991] Stoorvogel, A. A. (1991). The singular h∞ control problem with dynamic
measurement feedback. SIAM journal on control and optimization, 29(1):160–184. [Suga et al., 2009] Suga, H., Naitoh, Y., Tanaka, M., Horikawa, M., Kobori, H., et Shimizu, T.
(2009). Nanomanipulation of single nanoparticle using a carbon nanotube probe in a scanning electron microscope. Applied Physics Express, 2(5):055004. [Sumali et al., 2008] Sumali, H., et Carne, T. G. (2008). Air-drag damping on micro-cantilever
beams. In XXVI International Modal Analysis Conference (IMAC): Conference & Exposition on Structural Dynamics. Society of Experimental Mechanics. [Tamadazte et al., 2012] Tamadazte, B., Paindavoine, M., Agnus, J., P´etrini, V., et Le-Fort Piat, N.
(2012). Four dof piezoelectric microgripper equipped with a smart cmos camera. Microelectromechanical Systems, Journal of, 21(2):256–258. [Tang et al., 2014] Tang, H., et Li, Y. (2014). Development and active disturbance rejection
control of a compliant micro-/nanopositioning piezostage with dual mode. Industrial Electronics, IEEE Transactions on, 61(3):1475–1492. [Tian et al., 2009] Tian, G., et Gao, Z. (2009). From poncelet’s invariance principle to active
disturbance rejection. In American Control Conference, 2009. ACC’09., pages 2451–2457. IEEE. [Toda et al., 2010] Toda, M., Ono, T., Liu, F., et Voiculescu, I. (2010). Evaluation of bimaterial
cantilever beam for heat sensing at atmospheric pressure. Review of Scientific Instruments, 81(5).
130
REFERENCES
[Tsai et al., 2005] Tsai, Y.-C., Lei, S. H., et Sudin, H. (2005). Design and analysis of planar
compliant microgripper based on kinematic approach. Journal of Micromechanics and Microengineering, 15(1):143. [Villanueva et al., 2013] Villanueva, L., Karabalin, R., Matheny, M., Chi, D., Sader, J., et Roukes,
M. (2013). Nonlinearity in nanomechanical cantilevers. Physical Review B, 87(2):024304. [Vladar, 2003] Vladar, A. (2003). Scanning electron microscopy in real world environment. [Wang et al., 2003] Wang, W., et Gao, Z. (2003). A comparison study of advanced state ob-
server design techniques. In American Control Conference, 2003. Proceedings of the 2003, volume 6, pages 4754–4759. IEEE. [Wei et al., 2010] Wei, X., Chen, Q., Peng, L., Cui, R., et Li, Y. (2010). In situ measurements
on individual thin carbon nanotubes using nanomanipulators inside a scanning electron microscope. Ultramicroscopy, 110(3):182–189. [Welch, 1967] Welch, P. D. (1967). The use of fast fourier transform for the estimation of
power spectra: A method based on time averaging over short, modified periodograms. IEEE Transactions on audio and electroacoustics, 15(2):70–73. [Wong et al., 2007] Wong, C.-L., et Wong, W.-K. (2007). In-plane motion characterization of
mems resonators using stroboscopic scanning electron microscopy. Sensors and Actuators A: Physical, 138(1):167–178. [Wong et al., 1997] Wong, E. W., Sheehan, P. E., et Lieber, C. M. (1997). Nanobeam mechanics:
elasticity, strength, and toughness of nanorods and nanotubes. Science, 277(5334):1971– 1975. [Wu et al., 2013] Wu, Z.-S., Parvez, K., Feng, X., et M¨ullen, K. (2013). Graphene-based in-plane
micro-supercapacitors with high power and energy densities. Nature communications, 4. [Xie et al., 2009] Xie, H., et R´egnier, S. (2009). Three-dimensional automated micromanipu-
lation using a nanotip gripper with multi-feedback. Journal of Micromechanics and Microengineering, 19(7):075009. [Xu, 2013] Xu, Q. (2013). Precision position/force interaction control of a piezoelectric multi-
morph microgripper for microassembly. Automation Science and Engineering, IEEE Transactions on, 10(3):503–514. [Xu, 2015] Xu, Q. (2015). Robust impedance control of a compliant microgripper for high-
speed position/force regulation. Industrial Electronics, IEEE Transactions on, 62(2):1201– 1209. [Xu et al., 2012] Xu, Q., et Li, Y. (2012). Micro-/nanopositioning using model predictive out-
put integral discrete sliding mode control. Industrial Electronics, IEEE Transactions on, 59(2):1161–1170. [Yamamoto et al., 2003] Yamamoto, Y., Konishi, R., Negishi, Y., et Kawakami, T. (2003). Proto-
typing ubiquitous micro-manipulation system. In Advanced Intelligent Mechatronics, 2003. AIM 2003. Proceedings. 2003 IEEE/ASME International Conference on, volume 2, pages 709– 714. IEEE.
REFERENCES
131
[Yang et al., 2009] Yang, X., et Huang, Y. (2009). Capabilities of extended state observer for
estimating uncertainties. In American Control Conference, 2009. ACC’09., pages 3700–3705. IEEE. Disturbance-observer-based hysteresis compensation for piezoelectric actuators. Mechatronics, IEEE/ASME Transactions on, 14(4):456–464.
[Yi et al., 2009] Yi, J., Chang, S., et Shen, Y. (2009).
[Yu et al., 1999] Yu, M., Dyer, M. J., Skidmore, G. D., Rohrs, H. W., Lu, X., Ausman, K. D.,
Von Ehr, J. R., et Ruoff, R. S. (1999). Three-dimensional manipulation of carbon nanotubes under a scanning electron microscope. Nanotechnology, 10(3):244. [Zadegan et al., 2012] Zadegan, R. M., et Norton, M. L. (2012). Structural dna nanotechnology:
From design to applications. Int. J. Mol. Sci, 13:7149–7162. [Zames, 1981] Zames, G. (1981). Feedback and optimal sensitivity: Model reference trans-
formations, multiplicative seminorms, and approximate inverses. Automatic Control, IEEE Transactions on, 26(2):301–320. [Zhang et al., 2005] Zhang, W., et Turner, K. L. (2005). Pressure-dependent damping charac-
teristics of microsilicon beam resonators for different resonant modes. In Sensors, 2005 IEEE, pages 4–pp. IEEE. [Zhang et al., 2013a] Zhang, Y., Zhang, Y., Ru, C., Chen, B., et Sun, Y. (2013a). A load-lock-
compatible nanomanipulation system for scanning electron microscope. ASME Transactions on Mechatronics, 18(1). [Zhang et al., 2012] Zhang, Y. L., Li, J., To, S., Zhang, Y., Ye, X., You, L., et Sun, Y. (2012).
Automated nanomanipulation for nanodevice construction. Nanotechnology, 23(6):065304. [Zhang et al., 2013b] Zhang, Y. L., Zhang, Y., Ru, C., Li, J., et Sun, Y. (2013b). Nanorobotic ma-
nipulation of 1d nanomaterials in scanning electron microscopes. In Nanorobotics, pages 155–165. Springer. [Zheng et al., 2012] Zheng, Q., Gao, L. Q., et Gao, Z. (2012). On validation of extended state
observer through analysis and experimentation. Journal of Dynamic Systems, Measurement, and Control, 134(2):024505. [Zhou et al., 1998] Zhou, Y., Nelson, B. J., et Vikramaditya, B. (1998). Fusing force and vision
feedback for micromanipulation. In Robotics and Automation, 1998. Proceedings. 1998 IEEE International Conference on, volume 2, pages 1220–1225. IEEE. [Zhu et al., 2015] Zhu, S., Cheng, H.-H., Blakey, I., Stokes, N., Ostrikov, K. K., et Cortie, M.
(2015). Plasmonic ‘top-hat’nano-star arrays by electron beam lithography. Microelectronic Engineering, 139:13–18. Introduction to micro/nanofabrication. In Springer Handbook of Nanotechnology, pages 197–238. Springer.
[Ziaie et al., 2007] Ziaie, B., Baldi, A., et Atashbar, M. (2007).
List of Figures
1.1
Size comparison for various objects, in logarithmic scale. . . . . . . . . . . . . .
5
1.2
Examples of micro- and nano-objects produces with different fabrication methods.
7
1.3
Structures produced by electron beam lithography. The smaller ones were 500 nm wide and 50 nm tall [Zhu et al., 2015]. . . . . . . . . . . . . . . . . . . . . . . .
8
Characterization of a nano-structure in the SEM. (a) Sample before starting the measurements. (b) Membrane damaged by the measuring probe [Abrahamians et al., 2014]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8
Representation of the characterization/assembly workflow of micro and nanostructures in SEM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
Carl Zeiss Auriga 60 scanning electron microscope, used during this work, and its main components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13
Interaction between the focused electron beam with the surface of a sample produces backscattered electrons and secondary electrons. . . . . . . . . . . . . . .
14
Image of the mineral skeleton of a Radiolarian protozoa, acquired with optical and scanning electron microscopes [Newbury et al., 2003]. . . . . . . . . . . . . . .
15
SEM images of a micro-gripper extremity. . . . . . . . . . . . . . . . . . . . . .
16
1.10 Examples of commercially available nano-manipulators. . . . . . . . . . . . . .
17
1.11 Comparison between position measurements given by the internal sensor and by an external laser vibrometer for a step in the desired reference position. . . . . .
18
1.12 Piezoelectric nano-manipulators from the MINIMAN family. Multiple robots were used in [Fahlbusch et al., 1999] to increase the flexibility for this manipulation station. In addition, the MINIMAN-II robot is mobile and could move freely over the platform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
1.13 Manipulation of a CNT over a surface using a needle-type end-effector [Hertel et al., 1998]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19
1.14 Cutting of a ultra thin (50 nm) polystyrene film using a sharp needle. [Meyer et al., 2008] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20
1.15 Two AFM tips with different stiffness were used to hold and perform traction test on a nanoparticle chain aggregate [Rong et al., 2006]. . . . . . . . . . . . . . . .
20
1.16 A passive gripper used to assemble 3D structures inside the SEM [Dechev et al., 2004]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
21
1.17 Thermo-electric micro-gripper. By heating different parts of each finger, open and close movements could be produced [Nordstr¨om Andersen et al., 2009]. . . . . .
22
1.4
1.5 1.6 1.7 1.8 1.9
133
134
LIST OF FIGURES
1.18 Piezoelectric actuated gripper. The arms are composed of dual bi-morph structures, giving the end effector 6 DoF [Ciubotariu, 2015]. . . . . . . . . . . . . . .
23
1.19 Electrostatic comb-drive actuated gripper, located at the extremity of the circuit board (lower left corner of the image) [Beyeler et al., 2007]. . . . . . . . . . . .
23
1.20 6 DoF force sensor [Beyeler et al., 2009]. . . . . . . . . . . . . . . . . . . . . .
24
1.21 Tunning fork and application in characterization inside a SEM. . . . . . . . . . .
25
1.22 Electric characterization surface for nano-objects. It is composed of two gold covered pads, separated by 40 nm. Objects were carefully placed over the gap, and tested for its electric properties. [Suga et al., 2009]. . . . . . . . . . . . . . .
25
1.23 Proposed schematics for the first nano-manipulation station inside a SEM [Hatamura et al., 1990]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
1.24 SEM nanorobotic station. (a) Scheme for the station, which included 1 moving stage, 2 manipulators and an optical microscope to give lateral vision over the system. (b) Actual implementation, where all the components should be arranged to fit the vacuum chamber tight space [Koyano et al., 1996]. . . . . . . . . . . .
27
1.25 Robotic manipulator developed in [Yu et al., 1999], to test CNTs in a free space, without the substrate interference. . . . . . . . . . . . . . . . . . . . . . . . . .
27
1.26 CNT nano-tweezer fabricated on an AFM tip. Fingers could be controlled by applying a small voltage [Akita et al., 2002]. . . . . . . . . . . . . . . . . . . . .
28
1.27 SEM robotic station composed of 4 independent manipulators, for a total of 16 DoF. [Fukuda et al., 2003]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
28
1.28 Basic CNT transistor, build manually by placing the nanowire over pre-patterned electrodes [Kim et al., 2003]. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
1.29 3D photonic crystal, built by stacking multiple planar structures. Scale bar of 5 µm [Aoki et al., 2008]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
29
1.30 Robotic probing station. Combining vision servoing, the testing probes are automatically positioned over the desired spots. Contact between probes and substrate were detected by image inspection [Gong et al., 2013]. . . . . . . . . . . . . . .
31
1.31 Automatic selection and isolation of single carbon nanotubes, signaled by the red arrow in the image. Needle tips were used to sewer undesired connections between two electrodes [Zhang et al., 2012]. . . . . . . . . . . . . . . . . . . . . . . . .
31
1.32 Virtual manipulator model and interaction between virtual and real components. [Abrahamians et al., 2014]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
32
1.33 Desing of a microscale nanoassembly system to be fully contained on a chip, for operation in small environments, such as the SEM [Gorman et al., 2007]. . . . .
33
2.1
Effect of a short mechanical impulse to the SEM electron optical column (500000x amplification). From [Vladar, 2003]. . . . . . . . . . . . . . . . . . . . . . . . .
38
2.2
Effect of induced EMI disturbances on images [Pluska et al., 2009]. . . . . . . .
40
2.3
Drift quantification of optical encoders on long usage inside the vacuum chamber [Zhang et al., 2013a]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
42
LIST OF FIGURES
2.4
135
Deflection (a) and first resonance frequency (b) as a function of temperature for mono (silicon nitrite, Cantilever A) and bi-material (silicon nitrite with goal coating of different thickness, Cantilevers B,C and D) samples. [Mertens et al., 2003].
43
Normalized resonance frequencies of the first five flexural modes (M1–M5) of (a) the uncoated and (b) the gold-coated cantilever as functions of temperature [Sandberg et al., 2005]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
44
Carl Zeiss SEM Auriga 60, from the EQUIPEX ROBOTEX project, where the experiments were performed. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
47
(a) Scheme of the experimental setup proposed and (b) implementation inside the SEM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
47
Vibrometer PSD results, considering different reflection levels and analog output gains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
Acquired signals for different output analog gains, at a constant reflection level of 50%. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
50
2.10 Silicon cantilevers in the sample holder. . . . . . . . . . . . . . . . . . . . . . .
50
2.11 Averaged power spectrum density of the silicon cantilevers, detailing the frequencies with larger amplitude. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
52
2.12 Finite element analysis for the vibrometer’s support. . . . . . . . . . . . . . . . .
52
2.13 PSD around 700 (a) and 955 Hz (b) comparing the SEM in two conditions. It indicates that the SEM could be the possible origin for these disturbances. . . . .
53
2.14 PSD of Cantilever 1, detailed for its first mode (1169.8 Hz), comparing both test conditions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
54
2.15 Photo of the experimental setup for external acoustic excitation, with the SEM, the loudspeaker and signal generator. . . . . . . . . . . . . . . . . . . . . . . . . . .
55
2.16 Displacements measured on Cantilever 1 under the influence of an external noise source of different powers, generated by the loudspeaker. A sinusoidal wave with frequency matching the sample’s first mode is applied as the disturbance signal. The black dashed line indicates when the external excitation starts. . . . . . . . .
56
2.17 Summary of the most important measured sources of disturbance. . . . . . . . .
57
2.18 Diagram relating the different sources of disturbances and their influence over systems of various sizes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
58
3.1
SEM image of the micro-gripper FT-G30 and its main components. . . . . . . .
59
3.2
Example of the gripper’s step response in air, for a 30 V input. . . . . . . . . . .
60
3.3
Micro-gripper actuated finger and comb-drive system representation, showing its main components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
60
2.5
2.6 2.7 2.8 2.9
3.4
Micro-gripper representation, indicating that the laser beam measurement is not performed at the finger’s tip and that variations on the laser incident point can occur. 61
3.5
Experimental data used for estimating the relation between measured and actual tip displacement. The zoom box illustrates the contact between fingers. . . . . .
61
136
LIST OF FIGURES
3.6
Difference between experimentally measured and theoretical displacements for the micro-gripper electrostatic actuator. . . . . . . . . . . . . . . . . . . . . . . . .
62
3.7
Variation of model parameters for different input voltages, in air. . . . . . . . . .
64
3.8
Comparison between responses for large step inputs (ranging from 5 to 45 V) in frequency domain. Experimental data is shown in blue, and simulated response based on the second order model is shown in red. . . . . . . . . . . . . . . . . .
65
Influence of electronic input disturbance on the system in air. . . . . . . . . . . .
66
3.10 Step response for the micro-gripper in vacuum, for a 40 V step input. . . . . . . .
67
3.11 Duffing equation frequency responses. (a) shows the bifurcations (blue and red dots) for a softening non-linearity (β < 0). (b) shows experimental response for the micro-gripper compared to a linear simulated response. . . . . . . . . . . . .
67
3.12 Variation of model parameters for different input voltages, in vacuum. . . . . . .
69
3.13 Comparison between responses for large step inputs (ranging from 5 to 45 V) in frequency domain, considering the gripper in vacuum. Experimental data is shown in blue, and simulated response based on the second order model is shown in red.
70
3.14 Effects of electronic input noise on the system. . . . . . . . . . . . . . . . . . .
71
3.15 Response comparison for micro-gripper step in air (blue) and vacuum (red). . . .
73
3.16 Stiffness comparison for gripper in different environments, together with its corresponding polynomial fit functions. . . . . . . . . . . . . . . . . . . . . . . . . .
73
3.9
4.1
Summary on several published works considering nano-positioning control in air and vacuum, accordingly to the main strategy employed and year of publication. .
76
General LFT representation, containing the nominal system G0 , the uncertainty block ∆ and the controller K. During robustness analysis, the nominal plant and controller are considered as a single system M. . . . . . . . . . . . . . . . . . .
82
4.3
Feedback configuration for the small gain theorem. . . . . . . . . . . . . . . . .
84
4.4
LFT representation for the augmented structured disturbance ∆aug , applied in the case of robust performance analysis. . . . . . . . . . . . . . . . . . . . . . . . .
85
4.5
Basic structure for the H∞ problem, in the LFT representation. . . . . . . . . . .
86
4.6
Block diagram representing the augmented plant P, obtained from a nominal system G and the weight functions W. This structure is applied in the framework of H∞ synthesis to obtain the controller K minimizing a relation between inputs (r, d, n) and outputs in z. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
86
4.7
Sketch of the noise effects on the observer error. From [Khalil et al., 2014]. . . .
93
4.8
Singular value plot from disturbance input d to tracking error e. Two different controllers were compared: one obtained through H∞ optimization when only reference tracking characteristics were considered (W1 , in blue), and another considering reference tracking and disturbance effects on the system (W1 T −1 f and disturbance characteristics Wdist , in red). . . . . . . . . . . . . . . . . . . . . . . . . .
96
Resulting sensitivity function S and control sensitivity function KS for the obtained H∞ controller (green line), compared with the target function (blue line). .
97
4.2
4.9
LIST OF FIGURES
137
4.10 The block diagram for the ESO controller implementation. . . . . . . . . . . . .
100
4.11 Robust stability index for the controllers. The curves show minimum and maximum bounds for the true value µ∆ . . . . . . . . . . . . . . . . . . . . . . . . . .
100
4.12 Robust performance index for the obtained controllers. The curves show minimum per f and maximum bounds for the true value µ∆ . . . . . . . . . . . . . . . . . . . .
101
4.13 Simulated response to unitary step reference change considering various plants within the parametric variation range. (a) and (b) show the responses for a unitary reference step. The dashed-dot lined indicate the settling time (t98% ). (c) and (d) indicate the respective control signals applied. . . . . . . . . . . . . . . . . . . .
102
4.14 Closed-loop Bode diagram, from the input disturbance d to the system output y, considering the nominal plant response (blue continuous line) and the worst case result (red, dashed line) achievable within the defined parametric variation range.
103
4.15 Closed-loop Bode diagram, from measurement noise n to control signal u, considering the nominal plant response (blue continuous line) and the worst case result (red, dashed line) achievable within the defined parametric variation range. . . . .
104
4.16 Experimental setup for validation of closed-loop control strategies. . . . . . . . .
105
4.17 Experimental response to step reference change of different amplitudes (between 50 nanometers and 25 micrometers). The responses are normalized, to contrast differences due to parametric variations. . . . . . . . . . . . . . . . . . . . . . .
106
4.18 Experimental closed-loop response to a constant reference signal, and control signal u for each controller. No additional external disturbance is added. . . . . . .
108
4.19 Experimental responses for external acoustic excitations, considering the two controllers. The disturbance signal frequency is 1080 Hz with an intensity of 70 dB. The controllers are disabled at the start, to better illustrate their efficacy. . . . . .
109
4.20 Impact test example, showing the sphere and approximated drop height. . . . . .
109
4.21 Experimental impact test response, when a half-sphere of mass 10.8 grams is dropped near the microscope to mimic an impulsive disturbance on the system. .
110
5.1
CAD representation of a serial 6 DoF robot for SEM developed at the FEMTO-ST laboratory, showing how it is positioned inside the SEM chamber. . . . . . . . .
115
Nano-helix between pads, to facilitate manipulation tasks without damaging the sample. Fabricated and kindly offered by the Laboratory of Photonics and Nanostructures (LPN) in Marcoussis, France. . . . . . . . . . . . . . . . . . . . . . . .
116
B.1 Two possible representations for the LFT framework. . . . . . . . . . . . . . . .
147
5.2
List of Tables
1.1
Summary of issues presented during nano-manipulation tasks inside the SEM. . .
34
2.1
The vibrometer offers several analog output gain options, with different operation ranges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
48
Measurement noise RMS value and standard deviation for different output analog gains and reflection levels, and maximal measurement resolution for each gain. .
49
Nominal dimensions (length l, width w and thickness t) and first resonance mode measured at atmospheric pressure for the different silicon cantilevers. . . . . . .
51
Summary of the identified parameter for the micro-gripper in different working environments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
72
4.1
Nominal parameters used for the micro-gripper dynamics, considered for the ESO.
98
4.2
Measured average RMS disturbance for the micro-gripper’s tip, for different constant input voltage values, considering four operation conditions. . . . . . . . . .
107
2.2 2.3
3.1
139
List of Definitions
1
Definition: Infinity norm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
83
2
Definition: Structured singular value . . . . . . . . . . . . . . . . . . . . . . . .
84
141
I Appendix
143
A Atohm SD28CR08F datasheet CLASSIC SERIES
ATOHM®
SD28CR08F
100 W(rms) 300 W 90 dB/2.83V/1M 8 ohms 5.8 ohms 660 Hz 0.011mH 6.6cm² 28mm 3.2mm 2.5mm 11000 gauss +/- 0.35mm 1700 Hz/12dB
Conditions de mesure de la réponse en fréquence : bruit rose 2.83Vrms, distance 1 m, support de 23*150cm.
Puissance admissible (avec filtrage) Puissance admissible en crête Sensibilité Impédance nominale Résistance au CC Fréquence de résonance QTS QES QMS Cms Mms Vas BL Le Surface émissive Diamètre bobine Hauteur bobine Hauteur entrefer Densité de flux X-max linéaire Excursion maximale en crête Fréquence de coupure conseillée
Paramètres faible amplitude obtenus avec une précision maximale par la mesure simultanée du déplacement de l’équipage mobile et du courant traversant la bobine. (Analyseur Klippel et capteur micrométrique laser sur banc haute stabilité).
Le SD 28 CR08F est un tweeter hautes performances conçu pour couvrir une large bande de fréquence. Son dôme de 28mm en soie traitée est fixé à une bobine mobile support aluminium. L’entrefer est ferrofluidé afin d’augmenter la puissance admissible. Le noyau est doté d’une bague de cuivre dont l’avantage est de réduire sensiblement la distorsion par harmonique de rang 3. La face arrière du dôme est chargée par une cavité amortie permettant d’obtenir un excellent comportement aux fréquences les plus basses. La réponse en fréquence est caractérisée par une excellente linéarité conjuguée à une faible directivité et par une extension supérieure à 30kHz.
Notre souci de la performance nous conduit à faire évoluer fréquemment nos produits. Les présentes caractéristiques sont susceptibles d’être modifiées sans avertissement préalable. ATOHM ®est une marque déposée de Welcohm Technology (France). WWW.ATOHM.COM tel : +33 3 81 47 91 01 fax : +33 3 81 47 90 53
145
B The Linear Fraction Transformation
A common framework for modeling and analysis of uncertainty in systems is the use of Linear Fraction Transformations (LFT), a notation that allows representing uncertainties and its effects on systems flexibly. Supposing a matrix P, complex, that can be partitioned as [Doyle et al., 1991]: ' # P11 P12 P= ∈C (B.1) P21 P22
z1
w1
P
u y2
u2
y1
u1
l
P
w2
(a) Lower LFT
z2
(b) Upper LFT
Figure B.1: Two possible representations for the LFT framework. Let ∆u and ∆l be two other complex matrices of appropriate dimensions. Assuming the matrices are connected as shown in Figure B.1, the maps connecting its input wi to the output zi are given by: ' # ' #' # z1 P11 P12 w1 = , y1 P21 P22 u1
(B.2)
u1 =∆l y1 for the scheme on the left image and: ' # ' #' # y2 P11 P12 u2 = , z2 P21 P22 w2 u2 =∆u y2 for the scheme shown on the right image. 147
(B.3)
148
APPENDIX B. THE LINEAR FRACTION TRANSFORMATION
Considering the signals ui and yi as internal signals, the relation between input and output can be described by Equations B.4 and B.5:
Fl (P, ∆l ) :=P11 + P12 ∆l (I − P22 ∆l )−1 P21
(B.4)
Fu (P, ∆u ) :=P22 + P21 ∆u (I − P11 ∆u ) P12
(B.5)
−1
where the Fl is the lower and Fu is the upper LFT. Those LFTs represent the transfer of signal w to z, for a system under effects of a perturbation proportional to y. An interpretation of the LFT is that the system has a nominal mapping connecting w to z (P11 in the lower form and P22 in the upper form) while the other elements in P show the interaction of how perturbations affect this nominal map.
C Noise impact on high-gain observer
One of the limitations in the use of high-gain observers is that, as its structure approximates a differentiation, high-frequency measurement noises can lead to great deterioration in the signal/noise ratio. This demonstration was extracted from [Khalil et al., 2014]. Considering the system described in Equation C.1:    x˙1     x˙2      y
= x2 = f (x, Φ( xˆ, w), w, d) = x1 + n
(C.1)
with w a vector of known external signals, d a vector of unknown external disturbances, Φ a feedback control stabilizing the system around its origin, n the measurement noise and xˆ the estimated values of the real states x. The noise n and its derivatives are assumed to be bounded. A high-gain observer for the system above will have the form:  α    x˙ˆ1 = xˆ2 + ε1 (x1 + n − xˆ1 )    x˙ˆ2 = fˆ( xˆ, Φ( xˆ, w), w) + α22 (x1 + n − xˆ1 ) ε
(C.2)
with fˆ(·) the estimated value of the real system dynamics f (·), αi > 0 and 0 < ε 001c 1 the observer parameters. in:
Computing the estimation error x˜i = xi − xˆi considering the output with noise will result  α α    x˙˜1 = − ε1 x˜1 + x˜2 − ε1 n (C.3)    x˙˜2 = − α22 x˜1 + δ(x, x˜, w, d) − α22 n ε ε
with δ(x, x˜, w, d) = f (x, Φ( xˆ, w), w, d) − fˆ( xˆ, Φ( xˆ, w), w). Using the scaled estimation error transformation ζ1 = x˜ε1 , ζ2 = x˜2 allows to obtain the singularly perturbed equation C.4  α   εζ˙1 = −α1 ζ1 + ζ2 − ε1 n (C.4)   εζ˙2 = −α2 ζ1 + εδ(x, x˜, w, d) − α2 n ε In this case, the estimation error in closed loop satisfies an inequality of the form: µ x(t) − xˆ(t) ≤ c1 ε + c2 , ∀t ≥ T ε
(C.5)
with µ = sup n . The noise amplitude imposes a limit on the practical value of ε, limiting the speed for observer convergence. 149
Document generated with LATEX and: the LATEX style for PhD Thesis created by S. Galland — http://www.multiagent.fr/ThesisStyle the tex-upmethodology package suite — http://www.arakhne.org/tex-upmethodology/
Abstract: This work studies the nano-positioning problem inside the scanning electron microscope (SEM). To acquire fast and accurate positional information, a dedicated setup was implemented consisting of a vibrometer placed inside the SEM. This approach differs from methods based on image processing, as it allows to capture realtime data on the dynamic behavior of structures. In a first study, the mechanical disturbances acting inside the microscope’s vacuum chamber were characterized and its sources were identified. This demonstrated how external mechanical vibrations and acoustic noises can largely influence the components inside the SEM through mechanical coupling, limiting the effective positioning precision of manipulators. Next, a commercial micro-gripper was studied, both in air and in vacuum, and the differences between its response were highlighted. This allowed to obtain two dynamic models for this end-effector, one for each environment. Two control laws were proposed (H∞ control and Extended State Observer based control) for the system, to obtain a realtime, precise positioning in the vacuum environment and to attenuate the effects of the external mechanical disturbances. Results were demonstrated through simulation and experimental validation. Keywords:
Nano-manipulation and characterization, scanning electron microscope (SEM), robust position control, micro-gripper, disturbance characterization in SEM, H∞ control, extended state observer.
R´esum´e : Cette th`ese e´ tudie le probl`eme de nano-positionnement a` l’int´erieur d’un microscope e´ lectronique a` balayage (MEB). Pour obtenir des informations de position avec rapidit´e et pr´ecision, une installation d´edi´ee compos´ee d’un vibrom`etre plac´e a` l’int´erieur du MEB a e´ t´e mise en œuvre. Cette approche diff`ere de m´ethodes bas´ees sur le traitement d’images, car elle permet de saisir des donn´ees en temps r´eel sur le comportement dynamique des structures e´ tudi´ees. Dans une premi`ere e´ tude, les perturbations m´ecaniques agissant a` l’int´erieur de la chambre a` vide du microscope ont e´ t´e caract´eris´ees et leurs sources ont e´ t´e identifi´ees. Cela a d´emontr´e comment les vibrations m´ecaniques externes et les bruits acoustiques peuvent influer largement sur les composants a` l’int´erieur du MEB par couplage m´ecanique, limitant ainsi la pr´ecision des manipulateurs. Dans un deuxi`eme temps, une micro-pince du commerce a e´ t´e e´ tudi´ee. Une diff´erence entre ses comportements dans l’air et dans le vide a e´ t´e mise en e´ vidence, ce qui a permis d’obtenir deux mod`eles dynamiques pour cet organe terminal, un pour chaque environnement. Deux lois de commande ont e´ t´e propos´ees (commande H∞ et commande bas´ee sur un observateur d’´etat e´ tendu), afin d’obtenir en temps r´eel un positionnement pr´ecis dans le vide, et d’att´enuer les effets des perturbations m´ecaniques externes. Les r´esultats ont e´ t´e valid´es en simulation et exp´erimentalement. Mots-cl´es :
Nano-manipulation et caract´erisation, microscope e´ lectronique a` balayage (MEB), commande robuste de position, micro-pince, caract´erisation des perturbations dans le MEB, commande H∞ , observateur d’´etat e´ tendu.