晶硅太阳能技术的现状与未来展望

Status and Future Prospects for Status and Future Prospects for Crystalline Silicon--Based PV Crystalline SiliconSiliconCrystalline Silicon-Based PVgTechnologiesDr. Tom Surek*Surek PV ConsultingSurek PV ConsultingDenver, Colorado, USAtom_surek@* Previously Manager of Photovoltaics Programat the National Renewable Energy Laboratoryat the National Renewable Energy LaboratoryTalk OutlineTalk OutlineSome basics and background •Some basics and backgroundPV market perspective•PV market perspective•Silicon PV value chain: status and research directions–Silicon feedstockIngots and wafers–Ingots and wafers–Solar cells–Modules•SummaryPhotovoltaics is Solar ElectricityPhotovoltaics (PV)• Direct conversion ofsunlight to electricityAdvantages•Modular (mW to many MW)•No (or few) moving partsNo(or few)moving parts•Noise and pollution free•Reliable; low operating costs•Abundant, indigenous resource(36,000 km2PV for 1000 GW)Solar Cell Structure Transparent Cover (e g glass)Antireflection coating adhesive Front contact Current(e.g.,glass)Back contact and covert t n-type semiconductor p-type semiconductorPower out (W) x 100%()Solar cell efficiency (%) = ———————————Area (m 2) x 1000 W/m 222semiconductorabsorbermaterial(s)?40absorber material(s)?Best confirmed efficiency under30%)20E f f i c i e n c y (Cu(In,Ga)Se 21.00.52.0 1.5eV CdTe 100510*******.50.5 1.0 2.0 2.5Theoretical efficiency >52%Theoretical efficiency > 52%semiconductorabsorbermaterial(s)?40absorber material(s)?Best confirmed efficiency under 30%)Black-body limitstandard conditions (AM1.5,T=25°C)20E f f i c i e n c y (Cu(In,Ga)Se 2InPSi GaAs 1.00.52.0 1.5eV CuInSe CdTe a-Si:H 1005101520AM1.5AM02Cu(In,Ga)(S,Se)2Cu 2S CuInS 2CuGaSe 2Ge CdS 25Bandgap(eV)1.50.5 1.0 2.0 2.5Four-junction device with bandgaps1.8 eV /1.4 eV /1.0 eV /0.7 eVTheoretical efficiency >52%Theoretical efficiency > 52%p42% concentrator25% crystalline silicon20% thin films20%thin filmsCommercial modulesCommercial modulesTypically only 50-65% of these valuesPV Technology OptionsgypCrystalline silicon~85%oftoday’s marketFlat platesThin films85% of todays marketNew technologiesConcentratorsSiliconSilicon shortage providednear term market opport nities Multijunctions near-term market opportunitiesfor the other technologies(III-Vs)PV Market PerspectivePV Market Perspective –from a Silicon Point-of-View•Markets have been (and are) driven by subsidies•There has been a significant change in the PV market paradigm –This will impact R&D and technology choices in the foreseeable future •PV “yesterday” –to September 2008–Rapid market growth and associated high expectations driven by nearly uncontrolled subsidized markets (especially Spain in 2008)–Large subsidies and shortages of products resulted in increased prices (and profits)along entire value chain from raw materials to systems(and profits) along entire value chain, from raw materials to systems –Silicon shortage drove investments in technologies without silicon–Huge venture investments in thin films and next-generation technologies Huge venture investments in thin films and next generation technologies (anything “nano”) –unreasonably short R&D-to-product cycles projected y q p g(,,,)–Turn-key equipment for new technologies (a-Si, CIGS, CdTe, etc.)–Some new entrants had time to bring production on line and optimize processes and products –this was key to First Solar’s success–It was a “seller’s market” –make modules and they will selly g–Global financial crisis –investors are increasingly more risk averse–Significant market uncertainties –huge cut in Spain, uncertainties in U.S., China, Japan, Italy, France, etc. –pending changes in German feed-intariff driving markets–Silicon in much larger supply –oversupply anticipated for next 3-5 years Sili i h l l l ti i t d f t35–Oversupply in all parts of value chain –prices dropping rapidly It is now and for foreseeable future a“buyer’s market”“bankability”key –It is now, and for foreseeable future, a buyer s market–bankability key –New start-ups face significant pricing pressures –must compete against lowest cost silicon producers and lowest cost thin films (e.g., First Solar)lowest-cost silicon producers and lowest-cost thin films(e.g.,First Solar)–Expect many delays, failures, and consolidations in industry, and decreasing venture investments for next phase(s)–R&D with near-term focus will predominate (aimed at cost reductions and efficiency improvements in silicon and thin films)–Long-term R&D will be more a government role (labs and universities)PV Markets: Yesterday, Today, and Tomorrow•PV “tomorrow” –to 2012 and beyond?–Supply will likely continue to exceed demand –remain a “buyer’s market”S l ill lik l ti t d d d i“b’k t”–Market growth will likely accelerate in 2010 and beyond, especially with reduced module and system pricesreduced module and system prices–Government policies and well-conceived (and well-implemented) subsidy p g y p gprograms will be key for continued rapid market growth–“Grid parity” can be reached in 3-5 years in many locations (e.g., U.S.)–Silicon and low-cost thin films (e.g., First Solar) will dominate markets, with both having credible technical roadmaps to “grid parity”–Focused R&D in silicon and thin films will continue to predominate–Continued government role in long-term R&D (labs and universities)•PV “in the distant future”–Long-term, high-risk R&D is very important, but I don’t know how to predict a “breakthrough”–With silicon and the best thin films achieving grid parity, one can already With ili d th b t thi fil hi i id it l dsee a significant “PV future”>85% growthin 2008>40% per year growth>40%per year growthaverage 1998-2007Early projections:20-30% decrease >85%growth in 2009; first-ever>85% growth in 2008>40%per year growth >40% per year growthaverage 1998-2007Final (?) result:~30% increase>85%growth in 2009>85% growth in 2008>40%per year growth >40% per year growthaverage 1998-2007Supply/Demand 2009:f P l Mi t N i t C ltifrom Paula Mints, Navigant ConsultingU.S.Supply 2009 7861.3‐MWp U.S.Rest of W ld Demand 2009*5%6%World 11%Japan 16%Japan 7%Europe ROW 60%Germany Rest of 19%53%Europe 23%* 43% increase from 2008 shipments of 5491.8 MWpTechnology shipments:Thin films made gains in 2008 Technology shipments: Thin films made gains in 2008 and 2009, gaining two percentage points for 16% shareCdTe CIS/CIGS1%CIS/CIGS1% Poly c‐Si48%Ribbon Silicon 3%a‐Si5%8%Poly c‐Si45%Ribbona‐Si2%CdTe13%MonoSilicon2%Mono c‐Si35%c‐Si37%2008 5491.8‐MWp2009 7861.3‐MWp©2010 Navigant Consulting, Inc. PV Services ProgramPV Module Production Experience (or “Learning”) Curve1976“80% Learning Curve”:Module price decreases by 20% for every doubling ofcumulativeproduction cumulative production20052010X-Si: $1.60-2.50/W With supply > demandfor at least the next several years, only 80%Thin Film: $1.30-1.80/W pp y y ,y the lowest cost silicon producers and the lowest cost thin films(e.g., First Solar) will be competitive. Most new thin-film start-ups, especially turn-key producers, will have manufacturingclose to First Solar’s selling price It will require strategiccosts close to First Solar s selling price. It will require strategic markets and deep pockets to survive the needed developmentphase to improve processes and products. CPV producers willhave similar problems in these competitive markets.PV Module Production Experience (or “Learning”) Curve1976“80% Learning Curve”:Module price decreases by 20% for every doubling ofcumulativeproduction cumulative production20052010X-Si: $1.60-2.50/W With supply > demandfor at least the next several years, only “Bankability”80%Thin Film: $1.30-1.80/W pp y y ,y the lowest cost silicon producers and the lowest cost thin films(e.g., First Solar) will be competitive. Most new thin-film start-ups, especially turn-key producers, will have manufacturingclose to First Solar’s selling price It will require strategiccosts close to First Solar s selling price. It will require strategic markets and deep pockets to survive the needed developmentphase to improve processes and products. CPV producers willhave similar problems in these competitive markets.Silicon PV Value Chain: TodaySilicon feedstockIngot growth Solar cell Module Balance of Systems (BOS)plus wafering fabrication assembly Approximately 40-60% of total system cost of $$Approx. costs of ~$4.00 to $6.00/W manufacturing:$0.10-0.20/W plus $020030/W $0.25-0.80/W $0.35-0.70/W$0.40-0.60/W An important point:$0.20-0.30/W ~22%~22%~28%~28%An important point:every 1% in efficiency is worth ~$0.07-0.10/WTotal module manufacturing cost ~$1.50 to $2.50/W at the system levelResolving Resolving Polysilicon Polysilicon Shortages g y g •Large incumbent producers increasing capacities extensively with customer down payments and long term extensively with customer down payments and long-term contracts (ongoing)S l l i ith t i h i l •Several large companies with extensive chemical experience starting production (ongoing)•Many new start-ups (mostly in China) starting production with turn-key equipment for Siemens process (ongoing)•Potentially lower-cost fluidized-bed reactors are under development (ongoing)•Several approaches to upgraded metallurgical-grade (UMG) silicon were introduced to the market, but window-of-opportunity may now be closing •“Rule-of-thumb”: Every 1% (absolute) in module y ()efficiency gain is worth ~$20/kg in polysilicon pricePolysilicon Supply and Demand* y pp yWith ~33% CAGR for PV,With33%CAGR for PVproject excess capacityfor >5 yearsForecast price of~$30/kg in 2012* From presentation by RichardWinegarner of Sage Concepts atPhoton’s 7th Silicon Meeting, March 2009Crystalline Silicon ––IngotsCrystalline Silicon•Crystal growth developments: larger ingots (to 200kg CZ, 20 cm diameter), reduced consumables(gy gcosts (energy, crucibles, ambient gases, hotzones), melt replenishment and multiple ingots•PV-specific growth of CZ; starting low-cost FZ•Highest efficiencies for single crystals: FZ > CZHi h t ffi i i f i l t l FZ CZ25% laboratory cell (FZ)p16-23% production cells14-20% commercial modules25-year warranties•Cast multicrystalline (mc) ingots are fastestgrowing segment of PV industry:Commercial systems, up to 450 kg ingots-Commercial systems up to450kg ingots•Extensive development of thermal profiles,impurity distributions, crucibles, automation•Efficiencies ~10% lower than single crystals;need novel device architectures for mc-Si?•Opportunities for new crystal growth developments: low-cost CZ; largercast ingots; control grain sizes, defects, and impurities•Feedstock development; low cost, granular (fluidized bed)•Understanding defect/impurity interactions and passivation70 cm x 70 cm, 250 kgCrystalline Silicon Crystalline Silicon ––Waferingy gWafer Size and Thickness vsYieldWafer Size and Thickness vs. Yield0.60Wafering Yield -125 mm cell•terrestrial PV research:F t th ID (100025f /h )100150200250300350400450Thickness (microns)-Faster than ID saw (1000 vs. 25 wafers/hour)-Less surface damage-140-200 μm thickness, <200 μm kerf routineAs-cut 100m thin flexible •Ongoing advances needed:-Thinner wires; fixed diamond abrasive-Slurry recycling, water-based slurriesAs cut, 100 μm thin, flexible wafer (PV Crystalox Solar)-Thin wafer handling in processingCrystalline Silicon Crystalline Silicon ––Ribbons •First of new, terrestrial PVtechnologies to be commercialized•More than 20 innovative ribbon/ sheetgrowth approaches researched•Two leading techniques inproduction:Two leading techniques in production:-Edge-defined film-fed growth (EFG)-String RibbonEfficiencies aresimilarEfficiencies are similarto multicrystalline Si•Research issues:-Yield and throughput Granular silicon dropped here for continuous melt (growth rate, ribbon width)-Thin ribbons (~100 µm)-Thermal stress control M lt l i h t continuous meltreplenishment Mesa crucible -Melt replenishmentCrystalline Silicon Crystalline Silicon ––Sheets, Films, etc.•Silicon Film TM process(AstroPower, now GE, U.S.)-Melt/solidify granular Si onreusablesubstrate reusable substrate-High throughput (3 m/min)-Small grain size (~ thickness)-Low module efficiency (<8%)•RST technique (Solarforce, France)Many of these developments are y p terminating, but are constantly CSG Solar (Australia,replaced by new concepts and approaches to thin silicon Germany, China?)-1-2 μm thick filmapproaches to thin silicon •Sliver ®cells (OriginEnergy, Australia)Standard LowStandard Low--Cost Cell ProcessEfficiencies:15%-16.5% for multicystalline silicon15%-165%for multicystalline silicon17%-18.5% for single-crystal siliconOngoing research on improved and novel device structures and processes: Ongoing research on improved and novel device structures and processes:High High--Efficiency Silicon Solar Cells17.1%18.3%19.7%21.2%20.0%22.0%n c y HIT ll 14.7%14.8%15.6%16.5%16.8%120%14.0%16.0%18.0%C e l l E f f i c i e HIT cell Sanyo –23%10.0%12.0%Buried-contact cell/UNSWBP Solar –19%Point-contact cell SunPower –23%•Processing/lifetime relations•Gettering/passivation ofi iti /d f t (Si Himpurities/defects (Si 3N x :H deposition for polycrystalline Si)•Feedback to crystal growth•High-throughput, low-cost (e.g.,rapid thermal) processingS “f•Selecting “cheapest” wafer orHigh High--Efficiency Silicon Solar Cells17.1%18.3%19.7%21.2%20.0%22.0%n c y Suntech Pluto cell –19%HIT ll 14.7%14.8%15.6%16.5%16.8%120%14.0%16.0%18.0%C e l l E f f i c i e HIT cell Sanyo –23%10.0%12.0%Buried-contact cell/UNSWBP Solar –19%Point-contact cell SunPower –23%•Processing/lifetime relations•Gettering/passivation ofi iti /d f t (Si Himpurities/defects (Si 3N x :H deposition for polycrystalline Si)•Feedback to crystal growth•High-throughput, low-cost (e.g.,rapid thermal) processingS “f•Selecting “cheapest” wafer orImprovements through new processes (e.g., interconnects, back contacts), new materials (e.g., encapsulants, backsheets), aesthetics(e g encapsulants backsheets)aestheticsModule Packaging and Reliabilityg g yExample: X-SiBack sheet examplesyReliability Test ≠ Lifetime PredictionThese tests evolved after aThese tests evolved after acouple of decades ofoutdoor tests and fieldoutdoor tests and fieldfailure observations –resulted in accelerated testresulted in accelerated testproceduresTests are a necessary, butnot sufficient, conditionffi i di ifor ~25-year lifetime –required by system usersi d b tCost of qualification?Cost of qualification?“80% Learning Curve”:Module price decreases by20% for every doubling ofcumulative production200580%PV Module Production Experience (or “Learning”) Curve1976“80% Learning Curve”:Module price decreases by 20% for every doubling of cumulative production Crystalline SiliconTechnologiescumulative production20102015g 35% per year growth Evolutionary technology advancesare needed for ~80% learning curve:Lower cost (high-quality)feedstock 2005Lower cost (high-quality) feedstockImproved ingot growth processesThinner wafers (to 150 μm)Improved cell/module processing 80%90%80%Improved cell/module processing Improved performance –high efficiencyDisruptive advances are needed for~70%learning curve:70%70% learning curve:Much thinner wafers (<100 μm)Much improved performance (>25%)Innovative processing and processesp g pSilicon PV Value Chain: Today“SOON”Silicon feedstockIngot growth Solar cell Module Balance of Systems (BOS)plus wafering fabricationassemblyApproximately 40-60% of total system cost of $$Approx. costs of ~$4.00 to $6.00/W manufacturing:$0.10-0.20/Wplus $020030/W $0.25-0.80/W$0.35-0.70/W $0.40-0.60/WAn important point:0.20-0.30020*********~$2.00 to $3.00/W$0.20-0.30/W~22%~22%~28%~28%An important point:every 1% in efficiencyis worth ~$0.07-0.10/W0.20-0.300.20-0.500.30-0.50Total module manufacturing cost ~$1.50 to $2.50/Wat the system level<$1.00 to $1.50/WSolar PV Industry –Long--Term Outlook0.30Solar PV Industry Long Term Outlook Electricity from solar Demand0.25, ($/k W h )‐Si (2x)CdTeyNew technology0.150.20t , s o l a r P V p r i c e US -Average price of 0.10G r i d c o s c-Si (ave)c-Si (hi/lo)c-Si (hi CE)4%Periodic over-supply0.05200920102011201220132014201520162017201820192020c S (a e)c S (/o)c S (C )Periodic over-supply N t h i l b kth h id t hi l PV t d ti ()❑“Grid parity” is a conditional number; no single number is adequate ❑No technical breakthroughs are required to achieve solar PV cost reduction curve(s)❑Oversupply is inevitable and will be acute over the near term (credit driven)❑New technology could accelerate solar PV cost reductionBroad cost convergence over the next 6plus years Stephen O‘Rourke, Deutsche BankBroad cost convergence over the next 6 plus yearsSolar Industry–2010 Solar Industry 2010There will be farp fewer companies in 2011+ and beyond Why? beyond–PV Module Production Experience (or “Learning”) Curve1976“80% Learning Curve”:Module price decreases by 20% for every doubling of cumulativeproduction cumulative production“Bankability”20052010X-Si: $1.60-2.50/W With supply > demandfor at least the next several years, only Bankability80%Thin Film: $1.30-1.80/W pp y y ,y the lowest cost silicon producers and the lowest cost thin films (e.g., First Solar) will be competitive. Most new thin-film start-ups, especially turn-key producers, will have manufacturing close to First Solar’s selling price It will require strategic costs close to First Solar s selling price. It will require strategic markets and deep pockets to survive the needed development phase to improve processes and products. CPV producers willhave similar problems in these competitive markets.SummaryP C Pros•Abundance of material Cons•Indirect bandgap limits light capture •Well understood materialsystem because of IC industry Pro en reliable PV technolog •Energy intensive manufacturing processes: silicon feedstock,wafers and cells (~3years EPBT)•Proven, reliable PV technology with up to 30 years in field •Equipment for production is wafers, and cells (~3 years EPBT)•Feedstock availability constrained growth from 2004-2008Equipment for production is readily available from multiple vendorsg•Wide range of device structures, substrates, and production •Low barriers to entry for new companies and opportunities along entire value chain technologies•Module cannot be monolithically along entire value chain•Evolutionary improvements will result in “grid parity”–processed•Hazards similar to semiconductor industry (e.g.,gases,liquids,waste)result in grid parity SOONindustry (e.g., gases, liquids, waste)。