Mapping Solar Photovoltaic Research in India

Bikramjit Sinha

Despite being the first country in the world to have a dedicated agency/department for renewable energy1, India lags far behind many nations in technological competency as well as utilization of solar energy2. Low level of solar energy utilization in India and also the world as a whole is because of the high unit cost of solar energy relative to other traditional fuels, even after incorporating associated positive externalities3. High cost of solar energy is mainly due to the low conversion efficiency and high cost of solar cell materials. Two main strategies that can help in reducing the unit cost and increasing the utilization of solar energy are: Improvement in technology to enhance conversion efficiency Search and develop low cost solar cell materials.

So far, solar energy has been harnessed mainly through two technological routes - solar photovoltaics (PV) and solar thermal. Of the two, solar PV is more advanced and developed than solar thermal; so the present study focuses on the former. As far as India is concerned, the above two strategies have been rightly identified in the 11th plan. Jawaharlal Nehru National Solar Mission (JNNSM) has identified these two areas for huge R&D investments to meet the ambitious target of 20,000 MW by the year 2022. After the solar mission, it has become all the more necessary to map the solar PV research in India including who is doing what, potential collaborators and areas of collaboration. The present study aims to analyze and understand the pattern of Indian solar PV research output during 2000 to 2009, including shift in research thrust, if any.

Methodology Indian solar PV publications during 2000-2009 were extracted from Scopus. For analysis, first author approach was applied. To compare the scientific impact, two indicators namely T4 - 93

citation per paper (CPP) and relative citation index (RCI) were used. RCI is relatively a new indicator developed by Thompson Reuters to calculate Science and Engineering indicators. Relative Citation Impact (RCI) = An institutes share of total national citations/ An institutes share of total national publications RCI = 1 indicates that the authors/institutes citation rate is equal to national citation rate; RCI>1 indicates that the authors/institutes citation rate is higher than the national citation rate and RCI<1 indicates that the authors/institutes citation rate is less than national citation rate. RCI is more robust than other indicators in the sense that it measures both the influence as well as visibility of research activity, irrespective of the level of evaluation, either country or institute or author. The Activity Index (AI) suggested by Frame4 and elaborated by Schubert and Braun5 has been used to analyze the agency-wise activities. Changes in research thrust of different solar PV materials have been studied using the Transformative Activity Index (TAI) . The Domestic Collaborative Index (DCI) and International Collaborative Index (ICI) has been used to analyze the nature of collaboration. DCI was calculated as DCI = [(Di/Dio)/Do/Doo)]*100, where Di = number of domestically co-authored papers in a particular block of year, Dio = total output in that particular block, Do = total number of domestically co-authored papers, Doo = total Indian output And ICI was calculated as ICI = [(Ii/Iio)/Io/Ioo)]*100, where Ii = number of internationally co-authored papers in a particular block of year, Iio = total output in that particular block, Io = total number of internationally co-authored papers, Ioo = total Indian output DCI/ICI=100 indicates that the collaborative effort corresponds to Indian average, DCI/ICI>100 reflects higher collaborative effort than Indian average and DCI/ICI<100 indicates less collaborative effort than the national average.

Pattern of Output During 2000-2009, India published 1,814 papers in solar PV, about 3.7% of the global output. Indian solar PV output have increased almost steadily at the rate of ~16% per year (global AAGR ~19%), with a slight fall in 2005. The rate of growth during the last 5 years of the decade was relatively higher than the first five years (Fig. 1). The correlation between the T4 - 94

yearly nu umber of pu ublications an the years was found to be weak (R2=0.8) in nd s ndicating lac ck of consis stency in pub blishing num mber of paper over the y rs years.

400 350 300

No.ofpapers

250 200 150 100 50 0

R=0.810

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 5 7

Fig 1. Pattern of s . solar PV pub blication in In ndia during 2000-2009

Commun nication Ch hannels Indian so olar PV researchers pu ublished thei findings in 481 journ ir i nals origina ating from 3 33 countries Only 185 papers (~1 s. 10%) were p published in Indian jou n urnals and t rest wer the re published in foreign journals. H d n Highest num mber (512) papers were published in America e an journals. The top 6 journals tog j gether attract 394 pape accounti for ~22% of the tot ted ers ing % tal publicatio (Table 1). The jour ons rnal, Solar E Energy Mat terials and S Solar Cells p published th he highest number of pa n apers.

Table 1: Top 6 most-preferred jo ournals of Ind dian solar PV publication V n Journ title nal Solar Energy Mate erials and Sola Cells, Neth ar herlands Solar Energy, Engl land Renew wable Energy England y, Proc. of SPIE th Internatio he onal Society for Optical l Engin neering, USA Energ Conversion and Manage gy n ement, Englan nd Thin S Solid Films, Switzerland S Total l Publicati ions 165 58 56 51 33 31 % share e 9.1 3.2 3.1 2.8 1.8 1.7 IF 2009 3.853 2.011 2.226 0.553 1.994 1.727

T - 95 T4

Based on 2009 IF of the journa in which Indian sola PV paper were publ n o als h ar rs lished, it wa as found tha Indian solar PV pape were pub at ers blished in journals with an average impact facto or (AIF) of 1.8. Of th 1.808 do f he ocuments, 378 (~21%) papers wer published in non-SC re d CI journals.

Thrust in solar PV r n research: A Analysis of a author keyw words Of the total 1808 publications only 120 had auth keyword The keywords wer t s, 02 hor ds. re standardi ized by exp panding the abbreviation and chem ns mical formu ulas. Of the 1998 uniqu ue keywords thus obtain ned, 1488 (7 75%) appear once and 232 appea red d ared twice. L Large numbe er only keywor indicates lack of con rds s ntinuity in re esearch and w wide dispari in researc ity ch of once-o focus. Th solar PV research in India appea to be still exploratory in nature. hus, V n ars l y

80 60 40 20 0

TAI

SILICON

DSSC

NONSILICON

20 40 60 80

100 120

Fig 2. Changes in TAI of different so PV mate olar erials

Thrust in solar PV research was studied with regar to differ n V w d rd rent solar PV material P ls. Observat tion of TAI values revea aled that rese earch in Si materials ha decreased while that o m as on non-Si m material has increased. Though ac ctivities in both c-Si an a-Si hav decreased b nd ve d, intensity of decrease in a-Si was comparativ e s vely higher than those in c-Si (Fig.2 Relativel 2). ly ublication ac ctivity in c-S is an indic Si cation of en nhanced research thrust a correlate and es higher pu well with the anticipa domina h ated ance of solar PV by c-Si.

T - 96 T4

OPSC

CdTe

GaAs

aSi a Si

cSi

CIGS

Table 2: Transformative Activity Index (TAI) of different solar PV materials e f Solar PV materia r al 2000-04 TP TAI T A: Broa categories ad SILIC CON 35 3 1.2 NON N-SILICON 48 4 0.9 Total l 83 8 B: Sub categories b Crystalline silicon (c-Si) n 7 1.1 Amo orphous Silico (a-Si) on 20 2 1.4 Cadm mium tellurid Solar cells (CdTe) de ( 15 1 1.6 Galli ium Arsenide solar cells (G e GaAs) 8 1.7 Copp indium ga per allium seleniu solar cell ( um (CIGS) 6 0.9 Dye-sensitized solar cells (DSS SC) 12 1 0.5 Orga and polym solar cell (OPSC) anic mer ls 8 0.6 Total l 76 7 2005-09 TP TAI I 73 152 225 17 36 22 10 19 76 40 220 0.9 1.0 Total

108 200 308 24 56 37 18 25 88 48 296

1.0 0.9 0.8 0.7 1.0 1.2 1.1

Among n non-Si mater rials, growth in DSSCs a OPSCs was outstand h and w ding. For ins stance, TAI o of DSSCs have increas from 0.5 during 2000-04 to 1.2 during 2005-09. Like h sed 5 2 ewise, TAI o of OPSCs have increase by 0.5 points (Table 2 h ed 2).

lar arch thrust: 2001-2008 8 Global sol PV resea

Ov verall, activit ties in Si ma aterials have decreased while those o non-Si ma w on aterials have inc creased. Am mong the Si m materials, the intensity of d decrease in ac ctivity in a-Si is relatively very higher. Re elatively high activity i non-Si ma her in aterials was mainly due to the unpre ecedented inc crease in DSS and OPSCs activity. SCs

Changes in TAI of glob solar PV m bal materials resear rch 100 80 60 40 20 0

TAI

NONSILICON

cSi

SILICON

20 40 60 80

Source: E Extracted from Sinha, 2011 S

T - 97 T4

OPSC

GaAs

CdTe

CIGS

aSi

DSSC

Agency-wise distribution of Output Knowledge of trend in research output and thrust at the agency level has important policy implications. For instance, it will be easier and more effective to define future R&D strategies at the agency level instead of defining them at the level of each sister institutes.

Table 3: Activity Index (AI) of different agencies in solar PV research Agency Universities/deemed universities & colleges (AI) Engineering universities/colleges including IITs (ENGU) Department of Science & Technology (DST) Council of Scientific & Industrial Research (CSIR) Department of Atomic Energy (DAE) Department of Space (DoSp) Defense Research & Development Organization (DRDO) Industry Agricultural Universities (AGRU) Indian Council of Agricultural Research (ICAR) Others Total TP 705 564 122 112 55 53 40 38 30 20 69 1808 2000-04 TP AI 257 1.1 127 0.7 59 1.4 42 1.1 34 1.8 25 1.4 20 1.5 8 0.6 6 0.6 11 1.6 28 617 2005-09 TP AI 448 1.0 437 1.2 63 0.8 70 0.9 21 0.6 28 0.8 20 0.8 30 1.2 24 1.2 9 0.7 41 1191

As observed in the case of Indian science in general, academic institutes were the major contributors (~39%) to the solar PV output. The next highest contribution (31%) was made by the engineering universities (including IITs and engineering colleges) followed by DST (7%) and CSIR (6%). DAE and DoSp contributed 3% each while DRDO, industry and agricultural universities contributed 2% each (Table 3). The AI of all the agencies have decreased in the latter half of the study period except those of engineering universities, industries and agricultural universities (Fig. 3). Being a high-tech area, the ideal picture should be an increase in AI of related agencies such as DST, CSIR, DoSp and DRDO.

T4 - 98

100 0 50 0

TAI

DoSp

DRDO

ENGU

50 0 100 0 150 0

Fig 3. Agency-w changes in solar PV A during 20 wise AI 000-2009

Engineer ring universi ities, DST, DoSp and i industry are emphasizin on Si wh academic e ng hile institutes DAE and DRDO are doing more research in non-Si mat s, n terials (Tabl 4). CSIR is le maintaini a balanc between Si and non-S materials, evident fro equal TA in both th ing ce S Si om AI he forms of materials. Only academ institute engineer f mic es, ring universi ities, DST a CSIR ar and re carrying out research in all the different type of solar ce materials considered in the presen h es ell nt The major t thrust of ac cademic ins stitutes and CSIR is o DSSC while that o on w of study. T engineeri universit and DST is on a-Si which indic ing ties T cates agency y-wise differe ences in sola ar PV resea arch emphasi is.
Table 4: Agency-wise thrust in solar PV mater A rials Agenc cy Univer rsities/deemed universities & college (A d s AI) Engine eering univers sities/colleges including II s ITs (ENGU U) Depart tment of Scien & Techno nce ology (DST) Council of Scientifi & Industria Research ic al (CSIR) ) Depart tment of Atom Energy (D mic DAE) Depart tment of Spac (DoSp) ce Defens Research & Developme Organizati se ent ion (DRDO O) Industr ry Others s Total Si TP 36 24 29 12 1 3 1 1 0 107 TAI 0.6 1.3 2.3 1.0 0.7 1.7 0.3 1.4 non-Si TP TAI 124 1.2 30 7 24 3 2 9 1 1 201 0.9 0 0.3 0 1.0 1.1 0.6 0 1.4 0.8 0 Total 160 54 36 36 4 5 10 2 1 308

T - 99 T4

Industry

AGRU

ICAR

AI

CSIR

DAE

DST

Highly productive institutes

The output of 1808 documents came from 934 institutes. On the basis of first author, altogether 388 institutes contributed to the total output, of which only 100 (26%) institutes have CPP >=national CPP. In terms of RCI values also, a similar picture was observed. Some of the institutes having outstanding and very high RCI values in decreasing order are NIISTCSIR (17.3), Raman Research Institute (11.4) and Saha Institute of Nuclear Physics (10.9).

Table 5: Top 10 productive institutes in solar PV and impact of their research output Institute Indian Institute of Technology-New Delhi Indian Association for the Cultivation of SciencesKolkata Jai Narayan Vyaas University-Jodhpur National Physical Laboratory (CSIR)-New Delhi Jadavpur University-Kolkata Shivaji University-Kolhapur Indian Institute of Technology -Bombay Banaras Hindu University-Varanasi Anna University-Chennai Indian Institute of Science-Bangalore TP 160 89 63 48 48 47 43 42 39 36 TC 791 500 167 110 117 404 163 139 221 192 CPP 4.9 5.6 2.7 2.3 2.4 8.6 3.8 3.3 5.7 5.3 RCI 1.1 1.2 0.6 0.5 0.5 1.9 0.8 0.7 1.2 1.2

Taking 2% share of the national output as cut-off, the most productive institutes along with their total publication (TP), total citations (TC), CPP and RCI values are listed in table 5. These 10 institutes together contributed 615 papers and attracted altogether 2804 citations. With a total of 160 papers, IIT-Delhi heads the most prolific institutes. CPP of 5 out of the 10 most prolific institutes was higher than the national CPP. Further, performance of the institutes judged from their RCI values revealed that only 5 institutes had RCI >= 1. These indicate that all highly productive institutes do not attract more citations. Based on first author, 1147 contributed to the total output of 1808 papers, of which only 248 authors had CPP higher than the national average. Based on RCI values, it is evident that impact of the research done by only 25% authors was higher than the national level.

T4 - 100

Citation Analysis of Indian solar PV research The total 1808 papers have received 8379 citations during 2000-2009 with an average rate of citation as ~5. Of the 1808 documents, 775 (~43%) papers did not receive any citation and the remaining 57% papers received one or more citations (Table 6). Of the 1033 cited papers, about 38% received more than 5 citations each.
Table 6: Impact of Indian solar PV output

Times cited 0 1 2 3-5 5-10 10-20 21-50 51-100 101-200 >201 Total

No. of papers 775 202 144 293 189 139 47 14 3 2 1808

Total citations 0 202 288 1135 1436 2006 1398 1047 416 451 8379

Four out of the 5 highly cited papers (citations >100) were written in joint collaboration with foreign institutes except the one from Shivaji University. National Institute of Interdisciplinary Science and Technology (CSIR), Shivaji University, Mysore University, Indian Institute of Science, Tata Institute of Fundamental Reseacrh and Bhaba Atomic Research Centre make their presence in the top cited papers. Absence of centres of excellence such as IITs, NITs, and nationally reputed universities in the list of top-cited papers is clearly visible. This otherwise indicates the emergence of new centres of research in the country.

Collaboration pattern in Indian solar PV research Collaboration pattern was studied at the institute level. Here, a paper written by multiple authors but belonging to the same institute is considered as a non-collaborative paper and all such papers are termed as domestic single-institute publications. Out of 1808 papers, 933 (52%) were domestic single-institute publications and the remaining 875 were written in either domestic or international collaboration. Most of the collaborative papers have been produced through bilateral collaboration (Fig. 4).

T4 - 101

4instt. 4

instt. 5i

>5 5instt.

3instt.

1in nstt. 2instt.

Fig 4. Distr ribution of so PV outp by numbe of collabor olar put er rating institu utes

There see to be a balance betw ems ween domes and inter stic rnational col llaboration in Indian sola n ar PV resea arch: out of t 875 colla the aborative pa apers, 440 were written in domestic collaboratio w on and almo equal num ost mber of pap (435) w pers were written i internatio in onal collabor ration. Withi in the collab borative pap pers, both do omestic as w as intern well national coll laboration have increase ed almost in parallel (Fi 5). n ig.

250 200 DomesticSingle D einstpap Domesticcollab D bpap collaborativ vepap Internationalcollabpap

No.ofpapers

150 100 50 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Fig 5. Trend of collaboratio in solar P research i India g f on PV in

Domestic and internat tional collabo orative profile e

Number of papers w written in dom mestic and i international collaboratio and the c l on correspondin ng DCI and ICI values in two bloc of years is given in table 7. D d cks s n During 2000-04, domest tic collabora ative effort was relativ vely higher than interna ational effor while du rt uring 2005-0 09 T4 - 102 4

international collaborative effort was relatively higher than domestic effort. So, can we say globalization of Indian solar PV research is happening slowly?

Table 7: Collaborative activities in Indian solar PV research during 2000-09

Years 2000-2004 2005-2009 Total

Domestic 134 306 440

International 121 314 435

Total Indian output 617 1191 1808

DCI 89 106

ICI 82 110

Highly collaborative institutes

To find most collaborative institutes, normal counting was used. Taking 2% share of the total collaborative papers as cut-off, the top collaborating Indian institutes were selected. AI of the top 10 collaborating institutes indicate that half of them are producing more international collaborative papers and the other half publishing more domestic collaborative papers (Table 8). So, can we say highly collaborative institutes were more into domestic collaboration (IIT, Delhi) and less collaborative institutes were inclined more towards international collaboration (IACS, Kolkata).

Table 8: Collaboration nature of highly collaborative institutes Collaboration Domestic International TP AI TP AI 57 1.1 26 0.9 38 1.3 7 0.4 23 0.9 18 1.3 28 26 33 19 14 7 16 1.1 1.0 1.4 0.8 0.7 0.3 0.8 13 13 3 16 18 25 14 0.9 1.0 0.2 1.3 1.6 2.2 1.3

Institute Indian Institute of Technology-New Delhi Jai Narain Vyas University-Jodhpur Alagappa University-Karaikudi CSIR-Central Electrochemical Research InstituteKaraikudi Indian Institute of Science-Bangalore Defence Laboratory-Jodhpur Jadavpur University-Kolkata Anna University-Chennai Indian Association for the Cultivation of Science-Kolkata CSIR-National Physical Laboratory-New Delhi,
TCP: total collaborative papers

TCP 83 45 41 41 39 36 35 32 32 30

T4 - 103

Collaboration with foreign institutes

In solar PV research, India collaborated with 391 foreign institutes belonging to 54 countries. Some of the most preferred foreign institutes are Hanyang University, South Korea (18), Sungkyunkwan University, South Korea (18), and Ecole Polytechnique, France (17). At the country level, India published more collaborative papers with USA (104), followed by South Korea (78), Japan (47), and others (Fig. 6). Higher number of collaborative papers with some countries may be an outcome of the increased bilateral/multilateral cooperation for taking up joint R&D programmes thereby establishing institutional linkages.

Switzerlan d Malaysia

States 120
100 80 60 40 20 0

South Korea Japan

China

Germany

Netherlan ds Taiwan Mexico Italy

United Kingdom France

Fig 6. Indias leading collaborative countries in solar PV research

Academia-Industry collaboration

In Indian solar PV research, collaborative publication between academia and industry still seems elusive. Out of total 1808 solar PV papers, industry collaborated in only 112 papers (Fig. 7). In these 112 academia-industry joint papers, as many as 88 industries (59 Indian and 29 foreign) belonging to 15 countries were involved. It is surprising to note the virtual absence of leading Indian solar PV cells manufactures such as IndoSolar Ltd, Tata BP Solar India Ltd, Solar Semiconductor Pvt Ltd, and similar others in the publication arena. Low level of industry participation may be one of the explanations for a less developed Indian solar PV sector. T4 - 104

30

No.ofpapers
Key References 1. 2. 3. 4. pp. 98-110. 5.

25 20 15 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Fig 7. Industry involvement in Indian solar PV research

Concluding Remarks With the formal launching of the JNNSM in 2010 India conveys the message to the rest of the world about its intention to exploit the ultimate sources of energy in a massive way. To achieve the mission objectives, it will be necessary to carry out a lot of R&D activities in solar energy related areas. This, in turn demands knowledge of the present status of research in the country as well as the research thrusts and the present study does exactly that. Though the study has provided some valuable insights into the status and trend of solar PV research in India, there is a necessity to prepare a comprehensive profile of solar PV R&D capacity and capability in India. This will be useful in making wise decisions about future solar PV R&D strategies in the country, more so in view of large-scale ambitious projects like the JNNSM. Lastly, being an applied area, industry participation in research may be encouraged by creating suitable policy instruments.

Parthsarathi A., A champion of new technologies. Nature, 2003, 422:17-18. REN21, Renewables 2010 Global Status Report (Paris: REN21 Secretariat), 2010, pp. 80. Borenstein S., The Market Value and Cost of Solar Photovoltaic Electricity Production. CSEM WP 176, University of California Energy Institute, 2008, 38 pp. Singh, D. and Jennings, P., The outlook for crystalline silicon technology over the next decade. In Renewable Energy for Sustainable Development (ed. Jennings, et al.), American Institute of Physics, 2007, Goncalves LM, Bermudez VZ, Ribeiro HA and Mendes AM, Dye-sensitized solar cells: A safe bet for the future. Energy Environ. Sci., 2008, 1: 655-667.

T4 - 105