Recent Results of the International Technology Roadmap for Photovoltaics (ITRPV) 8 th Edition, March 2017

Transcription

1 Recent Results of the International Technology Roadmap for Photovoltaics (ITRPV) 8 th Edition, March 2017 A. Metz, M. Fischer, J. Trube PV seminar at UNSW Sydney, March 23 rd, 2017 Source: VDMA Page 1 15 March 2017

2 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRPV Results Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page 2 15 March 2017

3 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRPV Results Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page 3 15 March 2017

4 ITRPV Methodology Working group today includes 40 contributors from Asia, Europe, and US SILICON CRYSTAL. WAFER CELL MODULE SYSTEM Participating companies Independent data collection / processing by VDMA Review of data Preparation of publication regional chairs Next ITRPV edition Chairs EU Chairs PRC Chairs TW Photovoltaic Equipment Chairs US Parameters in main areas are discussed Diagrams of median values VDMA Author Page 4 15 March 2017

5 µm silver per cell [g/cell] ITRPV 8 th Edition 2017 some statistics Review ITRPV predictions Silver amount per cell 0,45 0,4 0,35 0,3 0,25 0,2 0,15 0,1 0,05 Edition 8 th 7 th Contributors Figures Review ITRPV predictions Wafer thickness (multi) Edition 2. Edition 3. Edition 4. Edition 5. Edition 6. Edition 7. Edition 8. Edition Prediction quality since 2009: Silver consumption trend well predicted and realized (Silver availability depends on world market) Wafer thickness trend bad predicted and no progress (Poly-Si price depends on PV market development) Edition 2. Edition 3. Edition 4. Edition 5. Edition 6. Edition 7. Edition 8. Edition VDMA Page 5 15 March 2017

6 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRPV Results Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Page 6 15 March 2017

7 average module sales price [USD 2016/Wp] PV learning Curve Learning curve for module price as a function of cumutative shipments Shipments /avg. price at years end: 2016: 75 GWp / 0.37 US$/Wp o/a shipment: 308 GWp o/a installation: 300 GWp 300 GWp landmark was passed! 1 historic price data LR 22.5 % cumulative PV module shipments [MW] / GWp 1 LR 21.5% ( ) dramatic price drop due to market situation Comparable to 2011/2012, but faster VDMA Author Page 7 15 March 2017

8 Spot Pricing [USD/Wp] Module price (US$/Wp) Price considerations Learning curve for module price as a function of cumutative shipments 1,8 1,7 1,6 1,5 1,4 1,3 1,2 1,1 1,0 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 Module 25% Module 42% Cell Wafer 29% Poly Si 26% Cell 23% Wafer 23% Poly Si 12% Module 37% Cell 23% Wafer 16% Poly Si 24% share 01_2011 share 01_2016 share 01_2017 Module price break down [US$/Wp] 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0, US$ 0,395 0,32 0,462 0, US$ reduction 01/ /2016: 64 % reduction 01/ /2017: 36 % (reduction 01/ /2012: 40 %) Dramatic price drop during 2 nd half of 2016 Market driven drop 0,24 Module Cell Wafer Poly Si 0.37 US$ 0,135 0,138 0,13 0,086 0,058 0,072 0,087 01_ _ _2017 Silicon Multi Wafer Multi Cell Multi Module Poly-Si share increased again High pressure on module manufacturers VDMA Author Page 8 15 March 2017

9 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRPV Results Si / Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page 9 15 March 2017

10 IHS Markit data Silicon Materials: Poly Si Feedstock Technology Silicon feedstock technology World market share [%] Trend: Share of poly-si feedstock technology Poly Si price trend: E 2012: 20 US$/kg 8 FBR 02/ 2016: 14 US$/kg 02/ 2017: 16 US$/kg oversupply situation of 2016 relieved 6 5 Siemens 86% Siemens process will remain mainstream FBR shows potential for cost reduction FBR share will be increased moderately w/ new capacity (2016 values in line w/ IHS Markit) IHS 2016 Siemens FBR other Other technologies (umg, epi growth,..) Not yet mature but available VDMA Author Page March 2017

11 [kg] [µm] Wafer Processes: wafering technology (1) Kerf loss and TTV for slurry based and diamond wire sawing Trend: Kerf loss / TTV Wire sawing for slurry based and diamond wire sawing; crystal growth Throughput per tool: 2016 = Trend: throughput crystallization/ wafering Ingot mass in crystal growth crystal growth per tool (mc-si, mono-like, HPM) slurry based wire sawing relative troughput CCz[kg/h]/Cz(kg/h] diamond wire based Kerf loss for slurry-based wire sawing [µm] TTV for slurry-based wire sawing [µm] Kerf loss for diamond wire sawing [µm] TTV for diamond wire sawing [µm] diamond wire sawing advantage: enable faster kerf reduction No big change in thickness variation is expected Throughput increase in crystallization/wafering will continue Gen 6 mc-si Gen 7 mono-si Gen 8 VDMA Author Page March 2017

12 Wafer Processes: wafering technology (2) afering technology for mono Si orld market share [%] For mono-si Wafering technology for multi Si World market share For [%] mc-si slurry based electroplated diamonds resin bond diamonds other slurry based electroplated diamonds resin bond diamonds other diamond wire wafering now mainstream for mono-si Throughut 2x 3x faster than slurry based For mc-si change to diamond wire is ongoing main challenge: texturing VDMA Author Page March 2017

13 Wafer Processes: texturing of mc-si wafers Different texturing technologies for mc-si Trend: market share of mc-si texturing technologies World market share [%] standard acidic etching Wet Nano-texture technology Reactive Ion Etching (RIE) Acidic texturing is: mature and high throughput process changes in standard will apear Next step: wet nano texturing, esp. for diamond wire RIE share is expected to increase but no cost efficient alternative Wet processing remains mainstream in mc-si texturing VDMA Author Page March 2017

14 [µm] Wafer Product: Wafer thickness trend Wafer thickness multi Wafer thickness mono limit of cell thickness in future modul technology Still no progress in mc-si thickness reduction Waferthickness [µm] st 2nd 3rd 4th 5th 6th 7th 8th ITRPV Edition 180 µm mc-si preferred thickness since 2009 Thickness reduction expected to start for mono-si cost reduction potential diamond wire will support Mono-Si wafer: thickness reduction starts New module technologies should enable thickness reduction VDMA Author

15 IHS Markit data Wafer Product: market share of material types Different wafer types Trend: share of c-si material types World market share [%] 8 mono casted-si domination is not for ever: Trend of last years will continue 6 5 multi IHS 2016 p-type mc p-type HPmc p-type monolike p-type mono n-type mono Casted material is still dominating today with > 6 Mono share is expected to increase (driven by n-type) Author Casting technology: HP mc-si will replace standard mc-si no come back of mono-like expected Mono technology: n-type material share will increase n- + p-type market share today 35% (2016 values are in line w/ IHS Markit) p-type material is expected to stay dominant mainly due to progress in LID regeneration VDMA Page March 2017

16 Wafer market share of wafer dimensions (new) ifferent mono-si wafer sizes Different mc-si wafer sizes orld market share [%] Trend: mono-si World market share [%] Trend: mc-si * mm² * mm² * mm² * mm² * mm² Fast switch to new format: * mm² Transition to new format in 2017 New mainstream: x mm² Larger formats are upcoming Expected new mainstream: x mm² Larger formats may occur after 2020 VDMA Page March 2017

17 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRPV Results Si / Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page March 2017

18 Amount of silver per cell [mg/cell] Cell Materials: Silver (Ag) per cell Trend for remaining silver per cell (156x156mm²) Good prediction of Ag reduction continues Remaining Silver / Cell [mg] : 100mg 21 t / 19.6% 548 $/kg 1.1 $cent/ Wp* * avg. cell efficiency 19.6 % 4.8 Wp/cell nd 3rd 4th 5th 6th 7th 8th mg mg reached mg expected Ag accounts in 2016 for 8% of cell conversion cost Ag reduction is mandatory and continues delays substitution by Cu or other material Ag will stay main metallization in c-si technology No break through for lead free pastes so far Market introduction depends on performance VDMA Author Page March 2017

19 [Wafer/h] Cell Processes: cell production tool throughputs Trend: tool throughput increase + synchronization of frontend/backend ell production tool throughput chemical processes, progessive scenario chemical processes, evolutional scenario themal processes, progressive scenario thermal processes, evolutional scenario metallisation & classification processes, progressive scenario metallisation & classification processes, evolutional scenario Wet benches are leading today with > 7800 wf/h Throughput increase continues Challenge: increase throughput + Improve OEE Two throughput scenarios: Progressive = new high throughput tools Evolutionary = continuous improvement of existing tools (debottlenecking, upgrades ) VDMA Author Page March 2017

20 Cell Processes: in line monitoring in cell production (new) Trend: in-line process control backend/test Inline process control World market share [%] Trend: Inline process in-line control process control frontend World market share [%] automatc optical inspection (AOI) after front/back silver & back Aluminum print electroluminescence (EL) imaging infrared (IR) imaging for hotspot detection AOI is widely used in printing 2016: Inspection at cell testing: EL use will expand (currently 5% only) IR inspection is not widely used optical quality control after antireflectioncoation (AR) coating incomming wafer inspection sheet resistance measurement after diffusion SiNx quality control has constant share 5 Incoming wafer inspection will exceed after 2021 Emitter sheet rho in-line control will increase VDMA Author Page March 2017

21 Cell processes: c-si metallization technologies Front side metallization technologies Different front side metalization technologies World market share [%] Rear side metallization technologies Different back side metalization technologies World market share [%] screen printing stencil printing direct plating on Si plating on seed layer sctreen printing plating PVD (evaporation/sputtering) Screen printing remains main stream metallization technology Plating is expected for rear and front side For rear side PVD methods may appear VDMA Author Page March 2017

22 [µm] Cell processes: finger width / number of bus bars / bifaciality Front side metallization paramters Trend: Finger width / alignment precision 8 60 Busbar technology Trend: number of bus bars (BB) World market share [%] busbars 4 busbars 5 busbars busbarless 10 Bifacial cell technology Trend: market share of bifacial cells World market share [%] 0 Finger width Alignment precision Front side grid finger width reduction continues 2016: < 50µm reached! Enables Ag reduction, requires increase of number of busbars 4BB are mainstream 3 BB will disappear Alignment precision will improve to sig. Selective emitters + Bifacial cells require good alignment Bifacial cells will increase market share monofacial cells monofacial c-si bifacial c-si VDMA Author Page March 2017

23 Recombination current [fa/cm 2 ] Cell processes: recombination current densities Recombination current densities Trend: J0 bulk, J0 front, J0 rear J0 bulk will be further reduced by optimizing crystallization p-type mc-si: fa/cm² further reductions will appear: p-type mc-si: fa/cm² p-type mono-si: fa/cm² n-type mono-si: 25 = fa/cm² 0 J0 bulk p-type multi J0 bulk p-type mono J0 front p-type material J0 rear p-type material J0 bulk n-type mono SHJ or back contact J0 front/rear n-type mono SHJ or back contact J0 front / J0 rear Further reductions by > 5 to < 50 fa/cm² p-type improvements are limited at the front side (i.e. need of improved diffusion / new pastes) Wide use of rear side passivation concepts p- type: reducing recombination losses is on a good way n-type: overcomes p-type bulk material limitations VDMA Author Page March 2017

24 Ohms / square Cell processes: emitter formation for low J0 front Emitter sheet resitance for phosphorous doping (p-type cells) Trend: emitter sheet resistance Different phosphorous emitter technologies for p-type cells Trend: emitter formation technologies World market share [%] Essential parameter for J0front Ω/ are standard today Increase to 135 Ω/ is predicted Challenge for tools and front pastes homogenous emitter by gas phase diffusion selective emitter by laser doping selective emitter by etch back homogenous emitter by ion implantation selective emitter by ion implantation Mainstream: homogenous gas-phase diffusion selective doping: etch back or laser doping Ion implant stays niche VDMA Author Page March 2017

25 Cell processes: technology for low J0 rear Different rear side passivation technologies Trend: rear side passivation technologies World market share [%] ITRPV prediction for J0 rear were good fa/cm² PECVD AlOx + capping layer ALD AlOx + capping layer PECVD SiONx Rear side passivation is mandatory for PERC PECVD AlOx will stay mainstream ALD will hold up to 10 % SiONx will disappear BSF cannot deliver required low J0 rear PERC takes over competing technologies in PERC PECVD Al2O3 + capping Al2O3 ALD + capping PECVD SiONx/SiNy etc. VDMA Author Page March 2017

26 IHS Markit data stabilized cell efficiency Cell Products: cell technologies / cell efficiency trends Different cell technology World market share [%] Trend: market share of cell concepts other PERC : PERC 15% (in line w/ IHS Markit) Si-tandem Trend: stabilized cell efficiencies; p-type PERC outperforms Average stabilized efficiency values for Si solar cells (156x156mm²) 27% 26% 25% 24% 23% 22% 21% PERC BSF IHS 2016 BSF PERC/PERL/PERT Si-herterojunction (SHJ) back contact cells Si-based tandem PERC is gaining market share ( 2017) BSF share is shrinking Back contact + HJ: slow increasing share Si tandem: under development 19% 18% 17% BSF cells p-type mc-si PERC/PERT cells p-type mc-si PERC, PERT or PERL cells n-type mono-si back contact cells n-type mono-si BSF cells p-type mono-si PERC/PERT cells p-type mono-si Silicon heterojunction (SHJ) cells n-type mono-si p-type mono PERC will reach n-type performance mc-si PERC is about to outperform mono BSF n-type IBC + HJ for highest efficiency applications stabilized >21% p-type mono PERC is in production VDMA Author Page March 2017

27 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRVP Results Si / Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page March 2017

28 [years] Module Materials: front cover material Different front cover materials World market share [%] 8 6 Trend: market share of front cover material Trend: lifetime of AR coating Expected lifetime of AR-coating on module front glass non-structured & non-coated front glass AR-coated front glass deeply structured front glass AR coated glass is mainstream 0 AR coating lifetime will increase to 25 years VDMA Author Page March 2017

29 Module Processes: interconnection technology Different techologies for cell interconnection World market share [%] Trend: cell interconnection technology Trend: cell connection material Different cell interconnect materials World market share [%] lead-containing soldering lead-free soldering conductive adhesive Expanding market share: lead free soldering + conductive adhesives Cu-ribbon Cu-wires structured foils shingled/overlapping cell Ribbons/wires will remain most widely used cell connection material VDMA Author Page March 2017

30 Module Power [Wp] Module Products: module power outlook Trend of Cell to Module power ratio Trend: cell to module power ratio (CTM) Throughput increase [%] 2016 = 104% 390 Module Power for 60-cell (156x156mm²) module Trend: module power of 60 cell (156x156mm²) 103% % 101% 99% 98% % % 95% acidic textured multi-si alcaline textured mono-si CTM will increase to > Acidic texturing has higher CTM 250 BSF p-type mc-si PERC/PERT p-type mc-si PERC, PERT or PERL n-type mono-si back contact cells n-type mono-si BSF p-type mono-si PERC/PERT p-type mono-si Silicon heterojunction (SHJ) n-type mono-si 60 cell modules 2017: Mono p-type PERC: 300 W are standard Multi p-type PERC: 285 W are common VDMA Author Page March 2017

31 Module Products: framed modules and J-Boxes Fameless c-si modules World market share [%] Trend: share of frameless c-si modules "smart" Junction-Box technology Trend: share of smart J-Boxes World market share [%] Different frame 2016materials World market share [%] framed frameless 8 standard J-Box without additional function microinverter (DC/AC) DC/DC converter 6 5 Aluminum Plastic other Al-frames will stay mainstream frameless for niche markets Standard J-Box remains mainstream Smart J-Boxes for niche applications VDMA Author Page March 2017

32 Module Products: module size Trend: share of cell dimensions Different cell dimensions in c-si modules World market share [%] Trend: Different share module sizes of module (full cell) size (full cell) World market share [%] full cell half cell quarter cell 60-cell 72-cell 96-cell other Full cell will remain main stream half cell implementation started! quarter cells currently a niche Big is beautiful: 72 cell module share will increase 60 cell modules mainstream until 2020 VDMA Author Page March 2017

33 warranty [years] degradation [%] Module Products: module reliability (new) Trend: warranty conditions and degradation for c-si modules Waranty requirements & degradation for c-si PV modules ,5% 3, Product warranty will remain 10 years Performance warranty 2024+: 30 years 25 2,5% , 1,5% 1, 0,5% 0, degradation: Initial / linear/year 2016: 3.0 % / 0.7% 2017: 2.5 % / 0.68% 2019+: 2.0 % / 0.68% 2021+: 2.0 % / 0.6 Performance waranty [years] Product waranty [years] Initial degardation after 1st year of operation [%] Degradation per year during performance waranty [%] VDMA Author Page March 2017

34 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRVP Results Si / Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page March 2017

35 Systems Balance of system (BOS) for power plants Cost elemements of PV System in US and Europe Trend: BOS in Europe and US For Systems > 100 kw Trend: BOS in Asia Cost elemements of PV System in Asia For Systems > 100 kw 1 0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 12% 12% 13% 8% 55% 87% 11% 11% 12% 7% 45% 75% 11% 11% 64% 58% 9% 11% 9% 8% 7% 8% 6% 8% 5% 5% 36% 33% 31% 29% % 13% 6% 8% 59% 94% 15% 13% 6% 7% 53% 84% 81% 77% 15% 15% 15% 12% 11% 12% 5% 11% 5% 11% 7% 5% 6% 5% 6% 5% 45% 43% 38% 0,1 0 Module Inverter Wiring Mounting Ground Module Inverter Wiring Mounting Ground Module costs still significant Costs in Asia are assumed to be significant lower VDMA Author Page March 2017

36 LCOE [USD/kWh] Assumed system price [USD/KWp] Systems Levelized Cost of Electricity (LCoE) Calculated LCOE values for different insolation levels Trend: LCoE progress a minimum approach 0,12 0,10 0,08 0,06 0, ,077 0,051 0, ,8 0,073 0, ,8 0, ,7 0,063 0,043 0, ,9 0,059 0, , System prices 2016: 970 $ / kwp 2027: <680 $ / kwp LCoE 2016: $ct/kwh (GER avg. 7.7 $ct**) 2027: $ct/kwh are ralistic System live times of 25 years are assumed 0,037 0,036 0,033 0,032 Next steps to further reduce LCoE: 0,030 0,02 0, extended service live to 30 years (supported by performance warranty trend) 0,00 0 further efficiency improvements 1000 kwh/kwp 1500 kwh/kwp 2000 kwh/kwp assumed system price + cost down measures LCoE depends strongly on local conditions ~5.7 US$ct/kWh lowest auction bidder in GER 2016** (avg. 7.7 $ct) ~2.42 US$ct/kWh possible near Abu Dhabi* today * ** VDMA Author Page March 2017

37 Outline 1. ITRPV Introduction 2. PV Learning Curve and Cost Considerations 3. ITRVP Results Si / Wafer - Materials, Processes, Products - Cell - Materials, Processes, Products - Module - Materials, Processes, Products - Systems 4. Summary and Outlook VDMA Author Page March 2017

38 average module sales price [USD 2016/Wp] average module sales price [USD 2016/Wp] Outlook: in detail view at PV learning curve earning curve for module price as a function of cumulative shipments : LR= 22.5% : LR= 39. Learning curve 10 3 for module price 10 4 as a function 10of 5 cumutative 10 shipments ITRPV finding : Wp learning ~ 7% (continually) per piece learning ~26% (market influenced) Learning was and will always be a combination of: historic price data LR 22.5% LR 39. ( ) historic price data LR 22.5% LR 39. ( ) Wp learning only ( ) LR 6.8% - Wp learning only per 0 piece 10 1 learning 10 2 only 10 3 ( ) LR 26.2% cumulative - per PV piece module learning shipments [MW] cumulative PV module shipments [MW] 10 1 VDMA Page March efficiency increase + continues cost reduction per piece = cost reduction of PV generated electricity But how will PV proceed in future? Approach: logistic growth exemplified by nature:

39 plant hight [cm] Spreading of PV like a plants life: An Experiment PV experiment: Investigation of the growth of a Tomato plant* watching milestones in a tomato plant s live 100 growth of a Tomato developing fruits G N t = 1 + e k(c t) being in flower singulation start fertilizing Tombolino 20 shooting days after seeding * Plant grown in Thalheim April July 2016 Parameter set: G = 1.12 m k = 0.07; c = 67 d VDMA Page 23 10/26/2016

40 plant hight [cm] Spreading of PV like a plants life: An Experiment PV experiment: Investigation of the growth of a Tomato plant* PV is starting growth of a Tomato (Tombolino) developing fruits Aproximation being in flower G N t = 1 + e k(c t) Parameter set: G = 1.12 m k = 0.07; c = 67 d shooting 10 singulation start fertilizing Tombolino PV is at the beginning ITRPV industry outlook: future PV Installation and future PV production requirements days after seeding * Plant grown in Thalheim April July 2016 VDMA Author: ITRPV 2016 Page March 2017

41 Annual Market [GWp] Global Installations [GWp] PV market trend until 2050: logistic growth nstallation forecast: Scenario 3 (high) Scenario 3 high : 9.2 TWp/ 14.3 PWh (< 10 % primary energy) Approach: 3 scenarios for 190 different countries in 4 regions Asia / America / Africa / EU Europe Asia Americas Africa Annual Market Shipments ITRPV finding: - Shipments until 2016 slightly above all scenarios - Annual PV market: 335 GWp/a to 800 GWp/a Replacement rate = key to overcome down cycles Evolutionary technology development works for all scenarios VDMA Page March 2017

42 Summary Silicon PV will remain a fast developing technology Further reductions of c-si PV manufacturing cost are possible without sacrificing quality and reliability cell efficiency improvements will support the cost reduction Silicon PV will significantly contribute to future power supply We are just at the beginning of PV-market development VDMA Page March 2017

43 Thank you for your attention! Contact us: Full version of 7 th edition available at: Source: VDMA Page March 2017