EPRI Photovoltaic_White_Paper_1207

8/7/2019 EPRI Photovoltaic_White_Paper_1207

1/28

Solar PhotovoltaicsExpanding Electric Generation Options

An EPRI Technology Innovation White Paper

December 2007


2/28December 2007 Page 2

An EPRI Technology InnovationWhite Paper

Solar Photovoltaics: Expanding Electric Generation Options

Executive Summary

EPRI and others have demonstrated that a broad port-folio of cost-competitive supply technologies will beneeded to satisfy the worlds rising demands for energywhile meeting climate policy and other societal objec-tives. Solar energy is a particularly attractive renewableenergy option because it is well distributed and abun-dant over most of the earths surface.

Photovoltaic (PV) conversion of solar energy directlyto electricity is a proven power generation technology

whose present-day application is constrained primar-ily by its relatively high first cost. Nonetheless, remote,off-grid PV installations have been economical for morethan 20 years, and grid-connected PV deployment isgrowing extremely fastmost rapidly on the rooftops ofresidential and commercial buildings served by conven-tional electricity infrastructure, but larger-scale applica-tions, in the form of ground-mounted PV plants feedingpower directly to the grid, are also increasing modestly.At the end of 2006, installed PV capacity since 1991

totaled more than 7 GW (7,000 megawatts) worldwide.By 2011, annual PV deployment is projected to beabout 8 GWwith 90% targeted for grid-connectedapplicationsand global installed capacity at that timeis expected to be about 33 GW.

ContentsIntroduction 4Technology Overview 6State of Technology 7Breakthrough Possibilities 9System Economics 13

Case Study: Residential PV 15Market Status & Outlook 18Technology Prospectus 22Issues & Opportunities 24Collaborative R&D Responses 27Resources & Contacts 28This White Paper, prepared under the direction of Tom Key, was written by Brian Fies, Terry Peterson, and Chris Powicki, with graphic design by ElizabethHooper.

PV cost has declined for decades in response to steadyadvances in conversion effi ciency and manufacturing ex-perience. Evolutionary progress with todays technologieswill continue for some years to come, new technologiesare expected, and revolutionary breakthroughs are antici-pated. Political, social, economic, and technical driversvirtually guarantee continued worldwide growth in PVdeployment. Te remaining barriers are such that thequestion is no longer if PV will join todays mainstreamenergy technologies, but when.

Te majority of PV applications over the next 10 years

will continue to be in residential and commercial build-ings as their desirability grows in consumers eyes.Distribution utilities are first likely to be affected at theretail level: PV deployed on rooftops and integrated intobuilding designs and other systems will drive changes indistributed supply technology, delivery infrastructure,and consumer behavior in the next decade.

At the wholesale level, PV plants on land dedicated forpower production are projected to begin finding broadapplication after 2020, and they will likely compete

with all other generation options before mid-century.However, between now and 2030, PV generation willnot likely account for more than a few percent of totalelectricity needs.





Successful integration of PV into the distribution gridwill require modernization to enable the power con-trol and communications functions needed for energymanagement and market access. At the transmissionlevel, grid operations and markets will need to accountfor higher PV penetration. Te pace of technology costreductions and effi ciency gains will largely define thetiming of substantial PV deployment. Market entry gen-erally will occur fastest in regions and countries with thebest solar resources, the most favorable policy environ-ments and largest incentives, and the highest electricityprices. In all areas, the remote grid and residential and

commercial building applications are expected to be eco-nomic before central PV plants compete at the wholesalelevel. PV will play in more markets over time as public-private investments in research and development (R&D)and commercialization yield fruit.

Given the rates of PV technology advancement anddeployment, the following intriguing scenariosand thechallenges and opportunities they implyneed to beconsidered in evaluating and planning for the future roleof solar energy in the power industry:

In 10 to 15 years at a retail level, building-integratedand other innovative PV deployment options erodemarkets served by conventional electricity infrastruc-ture and precipitate novel consumer wants and needsthat make new demands on the distribution system.

By mid-century at a wholesale level, PV technolo-gies compete with fossil, nuclear, and other renewablegeneration options by offering daytime energy inde-pendent of fuel price volatility and with low operationsand maintenance costs.

Te time to prepare for eventualities like these is now,as they offer the potential for major change, risk, andreward. For utilities and other participants in retail mar-kets, new business models and technologies are neededto transform residential, commercial, and other distrib-uted PV applications into grid assets. At the wholesalelevel, the biggest business opportunities may lie in plan- Central-Station PV: 4-MW Solarpark Geiseltalsee (Credit: BP plc)

ning for large-scale PV deployment and in supportingthe development of future hardware and systems thatenable successful grid integration.

imely solutions to grid integration issues will workboth to accelerate overall PV use and to greatly increaseits specific value to the utility industry. Furthermore,the availability of cost-effective energy storage and/oradvanced power electronics and control technology toaddress intermittency issues would not only simplify gridoperation but also enable PV and several other renew-able generation options to make greater contributions in

addressing energy security, affordability, environmentalprotection, and climate change issues.

On behalf of the electricity enterprise, EPRI is commit-ted to assessing PV developments and markets, testingand evaluating promising PV technologies, and com-municating findings to enable informed and proactiveresponses by electric utilities and other stakeholders. Inaddition, EPRI is initiating work to directly address in-termittency issues and to develop technologies for effec-tively integrating distributed PV with building electrical

systems and grid communications and control systems.





IntroductionPhotovoltaic (PV) cells convert solar energy into electric-ity in a solid-state process that consumes no fuels duringoperation and emits no byproducts. Te PV effect wasnoticed in the laboratory over 150 years ago, but mod-ern PV began with the invention of the silicon solar cellby Bell Laboratory in 1954. Te U.S. space programdrove R&D in the 1960s, while the energy crisis in the1970s stimulated the first wave of private sector activ-ity, when many major oil companies made investments.Sunlight-to-electricity conversion effi ciencies improved,

costs dropped, and government support evolved fromR&D funding and early demonstrations to tax creditsand other financial incentives enabling development anddeployment of commercial products.

In the early 1980s, U.S. government support for PVwithered as petroleum prices fell, and all but a fewoil companies lost interest. Despite a bruising busi-

ness climate, worldwide installed PV capacity grew ata consistent and healthy average of over 20% per yearthrough the late 1990s driven in large part by off-gridapplications and steady improvement in PV perfor-mance and cost: Between 1980 and 2000, the effi ciencyof commercial modules nearly doubled while their costdropped by 80%. In recent years, favorable policies andsubsidies in nations such as Germany, Japan, Spain, andthe United States have stimulated the grid-connected PVmarket prodigiously, with annual growth averaging 41%for the past 5 years (Figure 1). otal cumulative installedPV capacity exceeded 5 GW (5,000 MW) in 2005 and

7 GW in 2006 and will approach 10 GW by the end of2007extraordinary expansion for an industry that justpassed the 1-GW milestone in 1999.

Te burgeoning PV market is also a maturing one.Profits are being taken worldwide through a value chainthat encompasses component suppliers, device manu-facturers, system integrators, vendors, and installers.

Large energy and electronics corpora-tionsincluding some that dropped PVprograms a few years earlierare making

significant investments in PV R&D andcommercialization. Venture capital is flow-ing to aggressive newcomers, albeit pre-dominantly for later-stage developments.Meanwhile, the manufacturing cost of PVis steadily falling, its conversion effi ciencyis increasing, building-integrated andother innovative applications are growing,and advanced PV concepts and materialsoffering the potential for revolutionary

breakthroughs are being explored.

Given all these positive developments,many solar advocates are saying that thetime has come for PV to be taken seriouslyas a grid-connected supply option. Tishas been said beforemore than oncebut now the optimism appears more real-

Figure 1

Global PV Installat ions by Year, 1991-2006. Data for the years 1991-2000 are expressed as

worldwide totals, while other data specify where the PV was deployed. (Data source: Navigant

Consulting PV Service Practice)





Distributed PV: JapanJapanese policies have emphasized distributed rather than central-ized solar power production, as illustrated by residential develop-ments in Sapporo, a designated Solar City.

Sapporo established a goal of reducing per-capita carbon dioxideemissions in 2012 by 10% compared to 1990 levels. The city hasactive programs to increase public awareness, stimulate citizen initia-tives, provide incentives, and host clean energy installations. Distrib-uted PV is a major emphasis. Local schools are hosting five 10-kWdemonstration projects, and a suburban residential complex with 500

homes will be equipped with 1,500 kW of rooftop PV when com-pleted in 2008.

Many other Solar Cities have instituted similar goals and programs,including Copenhagen, Denmark; Barcelona, Spain; Qingdao, Chi-na; Adelaide, Australia; Freiburg, Germany; and Portland, Oregon.

Distributed PV in Sapporo (Credit: IEA-PVPS)

istic. In the industrialized world, securityand climate change concerns are focusingattention on solar energy as a globallyavailable, inexhaustible, and emissions-free resource. Interest in minimizing theadverse impacts of energy productionand use is being expressed politically, asoffi cials act to support clean energy, andindividually, as consumers deploy PVto increase self-suffi ciency and promoteenergy sustainability. Meanwhile, indeveloping and underdeveloped nations,

PV represents a leapfrog technologyfor delivering electric energy to improvequality of life and support economicdevelopment.

In brief, continuing improvements in PVcost-performance characteristicscom-bined with global socioeconomic consid-erations, energy security issues, climatechange factors, and popular demandvirtually guarantee a growing role for PV

in the carbon-constrained future. FromEPRIs perspective, whether PV willbecome a meaningful contributor to theglobal energy supply portfolio is not thequestion. When and how PV deploymentwill change the grid, as well as the retailand wholesale energy markets, are criticalquestions. And where these changes willlead the electricity industry and society isan issue best addressed before transforma-tive effects begin to happen.

Tis White Paperreviews the status of PVtechnology and markets, the potential forevolutionary and revolutionary technol-ogy advances, the issues and opportuni-ties facing the industry, and priorities forfuture PV research, development, dem-onstration, and deployment.





Technology Overview

PV technology relies on materials thatproduce electric currents when exposedto light, including semiconductors suchas silicon as well as other substances. Itis distinct from the other major solarelectric generation optionknown assolar thermal electricthat employslenses or mirrors to focus solar energy ona heat transfer medium and converts heatenergy into electricity via a conventional

electromechanical generator.

Teoretically, solar PV could satisfy glob-al electricity demand thousands of timesover, yet its potential remains unrealizedbecause its current performance rendersit more expensive than conventionalsources. Te performance of PV cells andmodules is judged by two fundamentalcriteria: effi ciencyand cost.

At noon on a clear day almost anywhereon Earth, light strikes surfaces directlyfacing the sun with a power density ofapproximately 1000 watts (1 kW) persquare meter. Effi ciencyis the percentageof the incoming power that a PV deviceconverts into electricity. Different PVmaterials absorb different colors of light,which correspond to different photonenergies. For example, blue photons are

nearly twice as energetic as red photons.Te lowest-energy photons that a semi-conductor can absorb define its energybandgap. Photons with energies lowerthan the bandgap pass through the PVcell, while higher-energy photons are ab-sorbed and some of their energy contrib-utes to the solar cells electrical output.

Utility-Scale PV: Germany

Due to the structure of its incentive policies, Germany is a worldleader in centralized PV deployment, with several megawatt-scaleplants in operation or development.

The 10-MW Bavaria Solarpark, dedicated in June 2005, includesground-mounted PV systems at three sites: the 6.3-MW SolarparkMhlhausen, the 1.9-MW Solarpark Gnching, and the 1.9-MWSolarpark Minihof. All together, the three projects comprise 57,600solar panels over 62 acres of land. Cumulatively, they make up thelargest PV plant in the world.

The Brstadt Plant in Brstadt is a 5-MW installation incorporatingbuilding-integrated and roof-mounted PV systems. It was completed inFebruary 2005. Solarpark Leipziger in Espenhain is a 5-MW systembuilt in August 2004. The facility has both stand-alone and grid-con-nected PV elements. The Solarpark Geiseltalsee/Merseburg employs25,000 mono- and polycrystalline modules to generate 4 MW ofelectricity.

Central-Station PV: 6.3-MW Solarpark Mhlhausen (Credit: SunPower)





Costdepends on perspective. When comparing technolo-gies, the PV industry often uses the cost of manufactur-ing a representative solar module as a fair baseline forcomparison. When considering PV applications, utilitiesand power producers typically compare the levelized costof electricity generated by PV to that of other generationtechnologies, while consumers often evaluate the in-stalled cost of a PV system and its projected output overtime in light of the retail cost of electricity. In this paper,units of cost will be defined when necessary to avoidambiguity.

An ideal PV technology would demonstrate both lowcost and high effi ciency. However, both conditions neednot be fully satisfied simultaneously for commercialapplication. For example, moderately low-effi ciency PVcan find vast markets if it is suffi ciently inexpensive,while high-cost, high-effi ciency cells are useful for cer-tain applications.

Cost and effi ciency are not the only parameters by whichPV is compared to other generation options. PV offersunmatched modularity and siting flexibilitythe same

technology can power hand-held calculators, rooftopinstallations, and central-station plants. When operating,PV produces no air emissions or greenhouse gases, noliquid or solid wastes, and no noise. Despite often-recurring notions to the contrary, land use concerns formost PV installations also are minimal. In most currentapplications, PV is installed on a buildings rooftop, aparking structure, or already disturbed land. Building-in-tegrated PV (BIPV) is incorporated directly into roofing,siding, or glass work. Only multi-megawatt PV installa-

tions demand large areas of dedicated land, amountingto 5 to 10 acres (2 to 4 hectares) per megawatt of capac-ity. Impacts on native plant and wildlife can be avoidedor minimized with proper project planning and manage-mentfor example, many sites suited for central-sta-tion installations consist of desert terrain of little use foralternative development.

Another frequent misperception relates to PVs energypayback perioda measure of how long a power gen-erator must operate before it produces more energythan was required to manufacture it. Researchers at theU.S. National Renewable Energy Laboratory (NREL)estimate energy payback periods of 4 years for systemsusing todays commercial crystalline silicon modules, 3years for existing thin-film modules, 2 years for futuremulti-crystalline modules, and 1 year for future thin-filmmodules. Assuming a module lifetime of 30 years, thismeans that 87% to 97% of a PV systems output will bepollution- and emissions-free electricity.

PV has relatively minimal environmental impact, butit is not completely benign. Most notably, potentiallytoxic or flammable materials are handled during manu-facturing. Tese resemble the hazards encountered in thesemiconductor industry, and the PV industry has takenadvantage of many of the approaches employed there toreduce potential hazards.

State of TechnologyAt present, commercial PV technologies may be sortedinto three main categories. Crystalline silicon sliced fromingots was the first PV type invented and is the mostfamiliar and mature. raditional flat-plate crystallinePV devices are single-junction solar cells, meaning theyare made of two semiconductor layers joined at a p-njunction, across which flow electrons liberated by sun-light exposure. Multi-junction devices consist of two ormore single-junction cells, each having differing bandgapenergies, stacked together to absorb more wavelengths

of the solar spectrum. Te best conversion effi cienciesfor crystalline silicon cells achieved in the laboratoryto date are approximately 24% in non-concentratedsunlightroughly three-fourths of the correspondingtheoretical limit. Commercially available single-junctionmodules have effi ciencies ranging from about 10% fornon-branded products to 18% or 19% for premium-cellmodules from top manufacturers. However, crystalline





silicon PV remains expensive to manufacture and ap-pears to present limited prospects for breakthrough costreduction.

Tin-film PVis a less mature development but is captur-ing a growing share of PV markets. It offers more scale-able and more automated manufacturing and, throughprocesses such as vapor deposition that produce thinnersemiconductor layers and less waste, more effi cient useof materials. Also, thin films can be integrated with awide variety of useful substrates, such as plastic andglass, that make them suitable for BIPV applications that

serve double duty as both power sources and structuralcomponents. Te performance of the most advancedcommercially available thin-film PV technologies to-dayincluding amorphous silicon (a-Si), cadmium

telluride (Cde), and copper indium-gallium diselenide(CIGS)significantly trails that of crystalline silicon,both in absolute terms and relative to their theoreticallimits. In the laboratory, the effi ciency of small thin-filmcells has exceeded 13%, 16%, and 19% for a-Si, Cde,and CIGS, respectively. Based on the correspondinghistory with crystalline silicon PV, commercial thin-filmproducts are expected to improve markedly as manufac-turers gain production experience and the technologymatures.

Concentrator PV (CPV), in which sunlight is focused

onto a PV cell by mirrors or lenses to generate morepower per unit of cell surface area, was an early favoriteof electric utilities and was the centerpiece of EPRIssolar power technology development efforts for more

Figure 2

World-Record Conversion Efficiencies for Various PV Technologies(Source: U.S. Department of Energy National Center for Photovoltaics)





Figure 3

Grand Challenges: Narrowing the Gap Between Existing and Theoretical

PV Efficiency (Source: U.S. Department of Energy)

than 15 years from the 1970s to 1990s. CPVs mainattraction is that it can leverage modest cell productionvolumes to much larger-scale systems using relativelysimple and inexpensive optical concentration. Anotherattractive feature is that CPV systems can provide higherconversion effi ciencies than conventional flat-plate sys-temsmore than 30% (also roughly three-fourths of thetheoretical limit) for multijunction devices incorporat-ing epitaxial layers of Group III-V compounds, such asgallium-aluminum arsenide and gallium-indium antimo-nide, grown on crystalline substrates.

Despite these advantages, CPV has been slow to gain acommercial foothold because it is not well suited to thevery small installations that have been the mainstream ofthe PV market. oday, CPV is benefiting from growinginterest in larger central-station solar plants, especially inthe western United States, Spain, Australia, and SouthAfrica. Te future of very large-scale CPV depends onengineering development and commercial productionexperience.

In addition to these commercial and near-commercial

technologies, many emerging PV concepts are under de-velopment. Figure 2 shows record-setting PV effi ciencies,sorted by technology type, over time. Te most effi cientPV devices are currently multi-junction concentratorsystems, followed by crystalline silicon cells, thin-filmtechnologies, and then emerging technologies such asdye-sensitized and organic solar cells. Te ultimate theo-retical effi ciencies of conventional crystalline silicon andthin-film PV are constrained by fundamental physicallimits. Advanced concepts, often collectively called third-

generation PV, have the potential to transcend some ofthose limits and convert light to electricity with effi cien-cies four or five times those possible with earlier designs.Despite their long-term promise, they are the subject ofso-far modest R&D efforts by governments, universities,and industry around the world.

Breakthrough PossibilitiesTe entire PV industry is making steady, evolutionaryprogress in using increasingly automated processes toproduce ever-thinner cells in greater volumes with highereffi ciencies and lower costs. Advances in PV materi-als and manufacturing will continue to incrementallyimprove the technologys cost-competitiveness andgradually expand its markets without the need for game-changing scientific breakthroughs. If true breakthroughsare madethe type that exploit new understandingof physics and material science to shatter old limits of

effi ciency and costthe current solar power evolutioncould become a more rapidly paced revolution.

Defined in 1961, the Shockley-Queisser Limit estab-lished that a single-junction solar cell without sunlightconcentration (operating at one sun) has a maximumtheoretical effi ciency of about 31%. Similarly, a single-junction cell operating under a peak obtainable solarconcentration of about 50,000 suns has a maximumtheoretical effi ciency of about 41%. Multi-junctiondevices can exceed these maxima by capturing more of





the solar spectrum. Emerg-ing PV materials andconcepts show potentialto significantly exceed theShockley-Queisser Limitand approach the true ther-modynamic limits of 68%for PV under one sun and87% for PV under maxi-mum solar concentration(Figure 3). able 1 com-pares the theoretical limits

to the best actual resultsobtained to date, furtherillustrating how muchopportunity remains toimprove the performanceof all PV types.

When effi ciency and costare plotted on a graph, PVtechnologies tend to fall inthree clusters, often called

generations, as indicated in Figure 4. Tese genera-tions generally correspond to the age and maturity of thetechnologies as well. First-generation PV technologies,comprising more than 90% of todays commercial in-vestments, are the traditional, wafered crystalline silicondevices descended from the original Bell Labs solar cells.Second-generation technologies, just now emerging intothe PV business mainstream, employ low-cost, low-ener-gy-intensity manufacturing techniques such as thin-filmvapor deposition and electroplating. Tey are less fully

developed than first-generation devices, with potentiallylower cost but also lower effi ciencies constrained inprinciple by the Shockley-Queisser Limit and in practiceby their immaturity. Tird-generation technologies arepotentially low-cost devices that for the most part existonly in general concept, have seen limited R&D invest-ment, but can in principle exceed the Shockley-QueisserLimit by employing multi-junction layering or novel

materials and techniques. Leading concepts includemultiple exciton generation, optical frequency shifting,multiple energy level, and hot-carrier devices. Teseconcepts likely will be enabled by organic materials, car-bon nanotubes, and other nanofabrication technologies.

Multiple exciton generation (MEG) cells overcome acentral limitation of existing PV technologythe one-to-one relationship between an absorbed photon and agenerated electron-hole pair. Tey exploit the phenom-enon of impact ionization, which converts one high-

energy photon into multiple electron-hole pairs and,thus increased current. Tis process has been observedfor decades in bulk semiconductor crystals, where itoccurs with relatively low effi ciency. According to recentexperimental reports, multiple electron-hole pairs canbe produced much more effi ciently in nano-sized (quan-tum dot) semiconductors. NREL found that quantumdots can produce as many as three electrons from each

Approximate Theoretic al

L imit E ffici ency

Approximate Be st Ex peri mental

Performance to Date

T hermodyna mic ( concentrator) 87 % n/a

T hermodyna mic (1 su n) 68 % n/a

Six -junction 58% n/a

Hot carrier 54% n/a

Triple-junction concentrator 64% 44% III-V al loys, monolithic s tack

Triple- junction (1 s un) 49% 15% Thin-film amorphous silic onalloys

Double-junction concentrator 56% 30% III-V al loys, monolithic s tack

Double-junction (1 sun) 43% 12% Thin-film amorphous silic onalloys

S hockley-Que isse r single-junction(46,200 s uns)

41 % 30 % C rysta lline si licon (5 0 0 suns)

S hockley-Queisse rsingle-junction (1 sun)

31% 24%20%12%

6%

C rystalline si liconThin multicrys talline silic onDye-sensitized cellOrganic c ell

Table 1

Theoretical PV Efficiency vs. Experimental PV PerformanceData show the potential magnitude of

future improvements in performance across device configurations.





photon of sunlight and could theoretically convert morethan 65% of the suns energy into electricity.

Optical frequency shiftinginvolves transforming the solarspectrum to maintain its overall power density withina much narrower range of photon energies, enablingincreased energy capture. Approaches include both upand down conversion, which involve creating a singlehigh-energy photon from two lower-energy photons orcreating two lower-energy photons from a single higher

energy photon, respectively; as well as thermophotonics,which employs an auxiliary device to combine thermaland optical energy inputs. A central feature is that thesolar spectrum is transformed by coatings or other ele-ments not part of the actual PV cell, providing a path-way toward substantially improving the effi ciency ofexisting PV technologies.

In multiple energy level solar cells, the mismatch betweenthe incident energy of the solar spectrum and a singlebandgap is accommodated by introducing additionalenergy levels such that photons of different energies maybe effi ciently absorbed. Multiple energy level solar cellscan be designed either with localized energy levels, calledquantum wells, or with continuous intra-bandgap mini-bands, also called intermediate bands. Both designs havea fundamental similarity in that they include multipleenergy barriers for light-excited electrons to surmount

when harvesting the photon energy.

Hot carrier solar cellsutilize selective energy contacts toextract light-generated hot carriers (electrons and holes)from semiconductor regions before their excess energyis converted to heat. Tis allows higher effi ciency de-vicesup to a thermodynamic limit of 66% at one-sunintensityby reducing the thermalization losses in

Figure 4

Efficiency vs. Cost Relations Schematically Define Three Generations of PV Technologies: First-Generation Crystalline Silicon, Second-Generation Thin Film,

and Third-Generation Concepts (Source: University of New South Wales)


12/28





System EconomicsAs noted previously, PV costs are looked at in differentways depending on perspective. PV industry experiencesince the 1970s shows a history of steady decline in theprice of modules driven by continuous innovation, as

summarized in Figure 5. Note that the scales of bothaxes are logarithmic. Te relationship between averagesales price and cumulative sales, which roughly mirrorsthe underlying declining-cost picture, is similar to thatseen with most manufactured goods. Te figure alsoshows the recent temporary stagnation in the price ofcrystalline silicon devices resulting from a combinationof a polysilicon feedstock shortage and subsidy-super-heated market demands in Europe and California.

Prices began easing in late 2006 and are likely to return

toward the historic trend as additional polysilicon capac-ity comes on line and thin-film module producers imple-ment aggressive expansion campaigns. ypical systemprices in 2007 range from about $6,000 to $10,000/kWof installed peak dc nameplate capacity, dependinginversely on size (the actual peak ac output of grid-con-nected systems is usually 70% to 80% of the dc rating).Going forward, Figure 5 projects a plausible future ofongoing evolutionary innovation and market expansionto a dramatic scale. Such a future is not guaranteed, of

course, but in light of PVs past 30 years of technologyand market progress it does appear reasonable.

As with every supply option, the price of the energyconversion system is only one component in calculat-ing the levelized cost of energy (LCOE). PV modulesaccount for about 50% of the typical up-front cost of aninstalled system, with the balance associated with items

such as power conditioners, wiring, support structures,and labor. So far, balance-of-system costs have declinedin parallel with module costs.

For all PV applications, resource availability is critical:Te better the solar resource, the greater the system out-put, and the lower the LCOE. As seen in Figure 6, theavailable sunlight on a 24-hour average varies only fromabout 125 to 300 W/m2 throughout most of the popu-lated parts of the Earth. Utility, energy, environmental,and climate policies also strongly influence PVs attrac-tiveness by directly or indirectly improving PVs LCOE

relative to other generation options.

Several aspects of PV economics are foreign to thenormal procedures used by utility planners and otherindustry practitioners focused on adding and operat-ing capacity to serve load. On the positive side, PV usesa fuel that is free, it imposes no environmental controlor waste management costs, and it has extremely lowO&M costsgenerally less than 2% of the total costof a PV system over its lifetime. On the negative side,it generates electricity when the sun shines rather than

when dispatched, and its output varies with unpredict-able weather conditions superimposed over daily andseasonal cycles. Tis means that the initial capital invest-ment for PV comprises almost the full cost of 20 to 30years worth of output, but an installations nameplatecapacity cannot be counted on at any particular instant.

Tese conditions present a conundrum to traditionalsystem planning and operations, where high premiumsare attached to predictability, reliability, and overall least

cost. However, with PVs LCOE several times greaterthan that of competing generation options, its lack ofdispatchability is not what has historically constrainedsupply-side PV applications. For consumers, the higherretail value of PV-delivered power has made justificationeasier, although the primary economic consideration forthese purchasers typically has been system affordabilityrather than cost-effectiveness.

the role of technology and other factors in shaping PVeconomics. Tis is one objective of the High Effi ciencyPV project recently begun by Electricit de France,EPRI, and others to explore and make more tangible thepotential of several third-generation PV concepts.





Because of the front-loaded nature of PV costs, theLCOE of a solar plant is dramatically impacted byfinancing terms: the cost of money. Tis means that theleast-cost approach to solar generation involves choosingthe right ownership modelin other words, arrangingfor the plant to be owned by the entity with the lowestfinancing costs and tax impacts.

Different ownership arrangements can easily produce25% differences in the LCOE. For smaller PV systems,often the least-cost approach means cash or mortgagefinancing obtained by a property owner; for large-scale

systems, it may indicate either independent powerproducer (IPP) or utility ownership, depending upongeopolitical site location. Incentive structures are also

important in influencing PV deployment. ypically,residents prefer incentives that reduce up-front costs,while commercial entities favor credits distributed over asystems lifetime.

Some solutions to the mismatch between PVs charac-teristics and standard industry planning and operationsare now appearing, including power purchase agree-ments (PPAs), feed-in tariffs, net-metering interconnec-tion policies, RPS requirements, and voluntary greenpower purchasing programs. A common characteristic ofthese solutions is that both the system performance risk

and the economic benefits belong to PV system own-ersconsumers or IPPsand not to traditional utilitiesor distribution companies. Moreover, these behind-the-

Figure 6

24-Hour Daily Average Worldwide Insolat ion, 1991-1993 (Source: M. Loster, U.C. Berkeley, www.ez2c.de/ml/solar_land_area/)





meter applications tend to reduce retail sales and, underconventional rate structures, revenues. Tis explains whytraditional industry players evidently need new businessmodels before they can embrace significant amounts ofPV generation in their portfolios.

Meanwhile, installed PV capacity is growing, and ap-proaches for encouraging additional deployment arebeing evaluated. For example, in 2005, the WesternGovernors Associations Clean and Diversified EnergyAdvisory Committee formed a Solar ask Force toanalyze the feasibility of adding large amounts of solargeneration to the western U.S. power grid and the policymeans to encourage that to happen. Te ask Forcefound that 8 GW of solar power generation, combin-ing all types, could reasonably be deployed by 2015 if apackage of Federal and State incentives was implemented

to cut the owners effective cost and, therefore, theLCOE, nearly in half, as shown in Figure 7. Te key rec-ommended Federal actions were to extend the recentlyimplemented 30% investment tax credit (IC) for solarsystems from its present 2-year term to 10 years and toallow utility companies to access this credit.

Case Study: Residential PVTe effects of different parameters on PV LCOE arebest illustrated by example. A grid-connected behind-the-meter system is considered here because that is theapplication experiencing the highest growth at presentand it is the one where PV will likely have its greatestimpact over the next 20 years.

able 2 characterizes a typical 5-kWdc

rooftop instal-lation in Fresno, California, based on published datafor PV systems installed in the State during 2007 and

assuming a 20-year simple payback. (Tis is not usuallyjudged a suitable economic metric for businesses but isoften invoked by consumers when weighing investmentdecisions.) o recoup an up-front cash investment, theprojected output over 20 years has an equivalent retailLCOE of 32/kWh, not considering any of the avail-able State or Federal incentives. Subtracting the currentFederal investment tax credit and the estimated Califor-nia Solar Initiative (CSI) Expected Performance-BasedBuydown (EPBB) payment from the capital cost reduces

this figure to 24/kWh. If mortgage-based financing isemployed over a 20-year period, the after-tax equivalentretail LCOE is 37/kWh for the first year of this systemsoutput and 45/kWh (before tax deductions) over theloans lifetime.

Te comparison point for these LCOE figures is not thebusbar generation cost, often quoted as in the range of3 to 5/kWh, but the retail electricity rate for residents,which currently ranges from 12 to 36/kWh in Califor-nia, depending upon usage. Even with State and Federal

incentives and tax-deductible financing, PV installationscurrently supply only nominally competitive electricity.Nonetheless, consumer demand has spurred installa-tion of about 25,000 PV systems totaling over 200 MWin California in the past few years. Future retail powerprices are uncertain, but they seem likely to increasefor a variety of reasons. Accordingly, a PV system thatappears nominally competitive at todays rates may well

Figure 7

Cumulative Effects of Proposed Federal and State Incentives on Solar LCOE

(Source: Western Governors Association CDEAC Solar Task Force Report,

http://www.westgov.org/wga/initiatives/cdeac/solar.htm)





seem to be a comparative bargain by the end of its life-time.

Te example shown in able 2 offers other importantinsights on PV economics. First, the simple-paybackequivalent LCOE for fixed financing and incentive as-

sumptions is essentially inversely proportional to localinsolation. Tus, at todays system price, ignoring State-to-State incentive differences, the cost of PV electricityvaries with location based on resource availability, asshown for some selected sites in Figure 8. Setting asideFederal and State incentives and assuming an up-frontcash investment, equivalent retail LCOE values varyfrom a low in El Paso, exas, of 90% of the Fresno

baseline to a high in Seattle, Washington, of 167%. Tisvariation between States is comparable to that caused bythe effects of subsidies, which means that PV ownershipin a relatively low-resource State with strong incentives,such as New Jersey or New York, may actually be moreattractive than in a higher-resource State with weaker

incentives, such as Utah or Colorado. Furthermore, thisvariation is less than that in regional U.S. electricityrates, which range over a factor of 4. Te implicationhere is that a PV bargain in one region can appear veryexpensive in another.Second, the simple-payback LCOE for fixed financingand incentive assumptions is essentially inversely pro-

5-kW dc (3.75 kW ac ) Re s idential PV S ys tem in F resno C A

Annual E nergy Output kW h 7,3 35 E PBB c alculator*T otal C a pital Ex pe nditure $4 7,000. 00 B a s ed on 200 7 C A inst allation e xper ien ce

Annual O &M $18.34 Es timate at 0. 25 cents/kW h

T otal 20-year C ost $47,366. 75

20-yr sim ple payback r etailL C O E $0. 32 T otal cos t/total output in kW h for 20 years

R educe c osts vi a Federal & St ate Incentives

$2, 0 00- Ca p ped 30% F ed. IT C $2, 0 00.0 0

CS I E PB B ( "$2 .33/W " this s ite) $9, 8 03.0 0 E P B B c a lculator*

Net C ost $35,563. 75

20-yr sim ple payback r etailL C O E $0. 24 Net cost/total output in kW h for 20 yea rs

C onsider T ax-Deductible F inancing

E ffective tax rate 25%

Annual interest rate 7%

Loan lifetime (yea rs) 20

P resent V alue $35,197. 00

Monthly P ayment $274.41 Loan + O&M

L C OE f or 20- yea r 7% loa n $0. 45 Annua l pa yments /annual output in kW h

F irs t-year a fter-tax retailL C OE $0. 37

(1st

-year payments - tax rate * 1st

-yearinteres t)/annual output in kW h

*

Table 2

Cost of a Typical Residential PV System in Fresno, California These costs are based on published experience and available incentives for 2007.





portional to the initial capital expenditure. Accordingly,future behind-the-meter PV deployments will offer moreattractive economics: Extrapolating from the projectionof evolutionary technological progress in Figure 5, the20-year cash payback retail LCOE for a 5-kW

dcsystem

in Fresno decreases from 32/kWh today to 18/kWhif purchased in 2015 and 7/kWh if purchased in 2030.Tese may indeed appear to be attractive costs in thosetimes.

By the same token, BIPV systems, which can take creditnot only for electric outputs but also for replacing one

or more structural building components, will likely havebecome cost-effective as well. Tis may launch a muchmore vigorous wave of these installations than seen to-day, where they are limited primarily to cases involving abuilding owners desire to project a green image.

Tird, the relationship between capital costs and LCOEhas implications for future central-station PV deploy-ment. At present, busbar-connected PV systems are notcost-effective by a wide margin because of todays highPV module price. Multi-megawatt PV systems are onlybeing deployed where policies place a special premiumon solar generation and, even then, solar thermal electrictechnology is capturing the bigger installations and alarger market share due to more favorable economics andperformance characteristics.

However, the evolutionary technological progress shown

in Figure 5 implies that the gap between the LCOE ofcentral-station PV and other generation options will nar-row over time, to the point where large-scale PV projectsmay also become cost-competitive generators in manylocations by about 2030.

Figure 8

Relative Cost of PV Electricity in the United States Due to Resource Variability This figure does not account for the effects of State subsidies and average

temperature differences.





Market Status & Outlook

PV markets have grown rapidly andchanged dramatically in recent yearsdriven by technological progress and asupportive policy environment.

As recently as 15 years ago, PV was usedalmost exclusively to generate powerin high-cost, off-grid applications notserved by conventional electricity supplyand delivery infrastructure. Now, it is far

more commonly used in grid-connectedapplications, with literally millions ofkilowatt-scale distributed installationson the rooftops of homes and businessesoffsetting the retail cost of grid electricityand a far smaller number of megawatt-scale central-station projects supplyingpower to the grid.

Figure 9 displays this dramatic trans-

formation: Te number of new remote,off-grid applications continues to in-crease worldwide, but their growth rateand capacity have been enormouslyoutstripped by those of grid-connectedPV since 2000. In 2006, off-grid ap-plicationsthough still maintaining aworldwide compound annual growthrate (CAGR) of about 16%accountedfor approximately one-eighth as manyshipped megawatts as grid-connected

applications, which had a CAGR of ap-proximately 50%. Tis trend is projectedto continue.

Te shift in PV applications is paralleledby a shift in markets. wo primary end-use markets for PV systems exist in theUnited States, and a third is emerging.

Concentrator PV: Arizona, Spain & Australia

Since 1995, Arizona Public Service (APS) and Amonix have beeninstalling Amonix CPV systems, with more than 600 kW presentlyoperating on a daily basis. They have reported total installed costsof $6/W, which is comparable to installed costs for large flat-platePV systems. As the technology matures, APS expects the cost of the

Amonix CPV system to drop to perhaps $3/W or less.

Amonix recently announced a joint venture with Guascor, which hasbuilt a 10-MW/yr assembly plant in Spain. This market is made pos-sible by Spains new feed-in tariffs and long-term contracts that attract

investors wanting to maximize the kilowatt-hours generated for theirinvested project dollars.

Solar Systems Pty Ltd. of Australia has spent more than 15 yearsdeveloping several generation of CPV prototypes leading up to a154-MW project being built in northwestern Victoria. A proprietaryHeliostat CPV system will use fields of sun-tracking mirrors to focussunlight on high-efficiency solar cells mounted 40 meters high. Thefirst stage of the project is due to be completed in 2010, with full com-missioning in 2013.

Concentrator PV in Arizona (Credit: Amonix)





Buyers of grid-independent systems represented the larg-est market segment prior to 2000. Tis segment contin-ues to grow steadily, with existing and new PV systemspowering off-grid homes, ranches, farms, and facilities,as well as communications installations, environmentalmeasurement and monitoring sites, security systems,signs, and other equipment located in remote locations.Buyers of consumer electronics, outdoor lighting sys-tems, and other products that incorporate small PV cellsmay also be considered participants in this segment, withthe manufacturers of these devices serving as middlemenbetween PV producers and consumers.

Buyers of grid-connected, behind-the-meter PV systemsnow represent the largest and fastest growing market seg-ment. As in off-grid applications, these systems are mostcommonly deployed in rooftop installations and less

Figure 9

Proportion of Grid-Connected vs. Off-Grid PV Applications Worldwide, 1999-2011 (Source: Navigant Consulting PV

Service Practice)

frequently as ground-mounted arrays, while building-in-tegrated applications are beginning to appear.

Residential and commercial consumers, and, to a lesserextent, industrial consumers participate in both the off-grid and grid-connected markets. Government agencieshave historically been significant PV consumers. TeU.S. Department of Defense, for example, has beeninstalling both grid-independent and grid-connectedPV systems at military facilities since at least 1976,with annual capacity additions usually totaling a fewmegawatts per year, much of it in the form of PV-bat-

tery-engine hybrid systems that range in size to over 100kW. Additional agencies at the Federal, State, and lowerlevels have also been deploying PV systems to servespecific off-grid applications, to support R&D, and,more recently, to comply with new procurement policies

mandating increased reli-ance on renewable energysources.

Buyers of grid-connected,supply-side PV systems

represent the third majorPV market. In the UnitedStates, large-scale invest-ments by utilities andIPPs have been few andfar between. Te mainexamples to date havebeen deployments of ahandful of megawattsmade patently for dem-

onstration purposes, suchas the 4-MW SacramentoMunicipal Utility Districtinstallation at RanchoSeco, California, and uc-son Electric Companys6-MW field in Springer-ville, Arizona. However,





State RPS mandates, such as Californiasrequirement that 20% of electricity comefrom renewable sources by 2017, havespawned several announcements of utilityPPAs with prospective builders of PVplants, indicating that capacity in thedozens of megawatts may come on line inthe next few years.

Markets similar to those in the UnitedStates exist internationally, with theirrelative magnitude and importance vary-

ing in individual countries depending ondomestic circumstances. For example, PVprograms in Japan have encouraged resi-dents and businesses to install many tensof thousands of grid-connected rooftopsystems with capacities of a few kilo-watts. Incentive structures in Germanyhave stimulated similar residential-scaleinstallations, as well as some deploymentsof large-scale PV parks with nameplatecapacities of several megawatts. In the

developing world, off-grid PV is com-monly employed to deliver the benefits ofelectricity to remote areas.

Wherever PV has flourished, it has doneso with government support: Policiessuch as those adopted in Japan, Ger-many, Spain, and California over thepast decade or so have nurtured PV andbuilt these regions into world leaders.

Subsidies, rebates, and tax credits helpmake PV more affordable for consumersand close the cost gap between PV andalternative generation technologies. Netmetering provisions, feed-in tariffs, andother policies provide consumers andpower producers with additional incen-tives to site and own grid-connected PV

Building-Integrated PV: New York City

Reconstruction of New Yorks Stillwell Avenue subway station provid-ed an opportunity to integrate amorphous silicon thin-film PV into asemi-opaque roof canopy that, upon its completion in 2005, was oneof the largest BIPV structures in the world.

The canopy roof provides the station with electricity as well as shade.Some 2,800 thin-film modules covering 76,000 square feet (7,060m2) generate approximately 210 kW while permitting 20% to 25%of daylight to pass through.During summer, the system

provides approximately two-thirds of the stations powerneeds (not related to poweringthe trains). Its annual output isabout 250,000 kWh.

Planning and design tookmore than four years, and thestations design process wasdone in conjunction with an educational component that included alarge-scale industry workshop involving several major PV companies.The station was designed to invoke the historic architecture of nearby

Coney Island and provide passengers with a grand sense of arrival,elegance, and civic pride.

Building-Integrated PV in New York City (Credit: Schott AG)





systems and to supply power to local grids under a vari-ety of credit and compensation plans. Interconnectionpolicies, building codes, and safety standards encourageand facilitate these applications, rather than present bar-riers. Air and climate policies are increasing the cost offossil generation, and Renewable Portfolio Standards arerequiring electricity providers to procure a growing por-tion of their supply from clean and renewable resources.

Figure 10 displays the tangible effect that these favorablepolicies have had on overall PV market expansion. Tefigure shows that in the decade preceding the regional

subsidies in Japan, Germany, and California, average an-

nual shipment growth was a respectable 15%, primarilyin off-grid applications. But during the period 1999 to2006, when those three regions had implemented sub-stantial subsidies for grid-tied systems, that growth ratespurted to 41%, representing continued strong growthin the off-grid market plus extremely rapid growth ingrid-connected applications.

Experiences in Japan and Germany further illustratethe strong influence of policies on market evolution. Inthe mid 1990s, the Japanese government implementednet-metering policies and a generous rebate program to

encourage the grid-connected PV applications offering

Figure 10

Effects of Policies on PV Market Growth Before incentives took effect for grid-tied systems in Japan, Germany, and California, annual PV module shipments

grew at 15% on average, mainly in off-grid applications. Subsidies boosted average annual growth to 41% overall. (Data source: Navigant Consulting PV

Service Practice)





consumers maximum advantage in light of high-pricedretail power. Over time, rebates on capital expenditures

declined and customer participation increased gradually,system prices fell substantially, and the net cost to thecustomer remained about the same, as shown in Figure11. Despite a dearth of incentives after 2004, Japans PVmarket continues to grow, albeit more slowly.

In contrast to Japans capital cost rebate program, Ger-manys production-based incentive programs are basedon the actual energy produced over a 20-year period andare paid through a national utility feed-in tariff, similarto a long-term electricity sales agreement. Because the

value of the feed-in tariff (roughly 50/kWh) is highenough to repay PV system construction costs, it hasstimulated entrepreneurial groups to form joint venturesthat have financed multi-megawatt PV farms.

Industry observers agree that the extremely rapid cur-rent worldwide growth in PV deployment likely cannotbe sustained over the next few years without continued

Figure 11

Japanese PV Program Rebates and Participation, 1994-2004 (Source: California Energy Commission)

Technology ProspectusTe latter part of this century will see energy supply,delivery, and use systems that could be as profoundly dif-ferent from todays as todays are from those of 50, 75,and 100 years ago. PVs role in that future is impossible

to predict because the technology has only recently be-gun appearing in significant commercial quantities andhas yet to find a routine place in mainstream settings.Early experience, however, indicates that its uniquemodularity and other characteristics have the potentialfor innovations that may transform the energy industry.PV is as different from other power generation technolo-gies as transistors are from vacuum tubes. Consider how

public support through tax incentives, rebates, and othersubsidies. However, as suggested previously (Figure

5), PV prices may decrease by the 2015 to 2025 time-frame to a point that such financial support is no longerneeded for further market growth.





mance impacts of such speculative technologies to thesame realm as those currently envisioned. Terefore, themain benefit of additional breakthroughs would likely bein accelerating the time to market for super-high-effi ciency PV products.

Figure 12 compares schematically the range of potentialPV costs in evolutionary and revolutionary scenarioswith the retail cost of conventional power in the UnitedStates, extrapolating from current cost estimates. Tebands indicate the range of projected costs dependingon resource availability, on whether PVs continuing

evolution proceeds in more or less accelerated fashion,and on whether or not aggressive R&D investments aremade to pursue PVs revolutionary technology potential.Tese cost estimates are obviously speculative, and theirlikely error margin increases dramatically beyond a fewyears time. Nevertheless, the depicted timeframes for theimplied innovations and the relative scales of their costimpacts are plausible. So too are the following projec-

the cellular phones that have leap-frogged landlines in developing coun-tries, transformed personal and busi-ness communications in the developedworld, and created consumer wantsand needs being met by new industryplayers and a seemingly continuousstream of innovations.Assuming the continued evolutionaryprogress projected in Figure 5, PV de-ployment in distributed grid-connected

applications is clearly heading in thedirection of causing landscape changesin the retail electricity sector. Majorscientific breakthroughs may acceleratethe growth of these PV applicationsat the retail level, shake up wholesaleelectricity markets, and enable unfore-seen innovations spanning the energysector. Tese effects would not happenovernight, but PV will almost certainlychange the way significant quantifies of electricity are

supplied, delivered, and used.Among the third-generation PV concepts, it is far tooearly to predict the fate of any individual approachbecause they have not been experimentally verified.Nevertheless, scientific breakthroughs and successfulcommercialization of any emerging third-generation PVconcepts could produce a three-fold to five-fold increasein commercial module effi ciency and a dramatic im-provement in PVs deployment potential. Although im-

possible to forecast confidently, breakthroughs enablingmarket-competitive busbar electricity from PV technol-ogy are clearly possible by 2025 or 2030.

Given the relatively short history of PV science, it is alsoreasonable to anticipate the discovery of additional, yet-unimagined PV technologies in coming years. However,the laws of thermodynamics appear to limit the perfor-

Figure 12

Projections of Technological Progress This schematic representation shows cost relationships between

conventional power sources and likely evolutionary PV development as well as the possible impacts

of revolutionary developments. Achieving the benefits of the leading edge depicted for revolutionary

innovations will require an aggressive, sustained R&D effort.


24/28





forecasted substantial improvements incost-performance characteristics couldposition central-station PV deploymentas a cost-competitive non-emitting gen-eration alternative, shifting the dispatchorder for current units and eroding themarket share of competing options fornew capacity additions. Te biggest busi-ness opportunities may lie in planning forthis scenario and, possibly, in securing astakehold by supporting the developmentof future proprietary hardware.

Te cost structure of PV technologiesdoes not contain a single dominant factorwhose elimination or substantial reduc-tion would render their capital-equip-ment investments competitive with moreconventional generation options on awidespread basis. Rather, there is a com-plex of factors, ranging in both degree ofleverage and time-scale of effectiveness,which must all be addressed to achieve

the goal of cost-competitive, globallysignificant PV power generation. For PVto proliferate and achieve its potential,incentives must remain in place, remain-ing policy and market barriers to deploy-ment must be cleared, and major R&Dinvestments must be made to reduce thecosts of crystalline silicon, thin film, andCPV technologies and advance third-gen-eration concepts.

Market-transforming policy measureshave the most immediate effect by en-couraging otherwise uneconomical earlydeployments and thereby acceleratinglearning curve advances where costsare reduced as production volume isincreased. However, the cost reductions

Tracking PV: Portugal

In March 2007, GE Energy Services, PowerLight, and Cataventocommissioned an 11-MW solar power plant in Serpa, Portugal. Thestations 52,000 modules cover 150 acres (60 hectares) and employthe SunPower single-axis tracking system to keep the PV panelspointing toward the sun, increasing their daily electricity output by upto 35%. The project cost approximately $150 million.

Portugal relies heavily on imported fossil fuels and has implementedaggressive incentives for renewable energy installations. A key com-ponent of Portugals Energy Efficiency and Endogenous Energies

(E4) program is a feed-in tariff of $0.317 to $0.444/kWh for bothground-mounted and rooftop solar power systems with a 15-yearpower purchase guarantee. Adopted in 2001, the E4 program isexpected to provide 4,400 MW of renewable energy by 2010, 150MW of it in the form of PV.

Central-Station Tracking PV in Portugal (Credit: SunPower)





associated with manufacturing effi ciencies and improvedeconomies of scale are often incremental. Tus, policyincentives offer relatively little leverage on PV costs.Tey also entail the risk of retarding price declines byoverstimulating market growth and pushing demandahead of supply, as has happened with crystalline siliconPV over the past couple of years.

Improving the effi ciency of PV cells and modules is themost effective route to increased cost-competitiveness.More power can be generated per unit of module area,and the required scale of nearly every system component

can be reduced. Tus, even incremental effi ciency gainsaffect total cost in a more-than-linear fashion. Improvedeffi ciency also offers the promise of major cost reduc-tions because breakthroughs would allow several timesmore energy to be harvested per unit of module area.

For the near and intermediate term, continued policyincentives and aggressive R&Despecially in manufac-turingare critical to continue the evolution of exist-ing PV technologies, modules, and integrated systemsfor distributed generation. In an era of enduring public

awareness of energy and environmental issues, it seemslikely that third-generation PV concepts can sustain acritical mass of public R&D support over the next quar-ter century, the likely time required to advance those of-fering the greatest commercial potential. Already, smart,adaptive businesses as well as entrepreneurs are beingattracted to the opportunities presented by investing inPV technology to respond to and anticipate tighteningRPS and emissions requirements.

Favorable long-term outcomes may be effectively acceler-ated by a focused R&D program dedicated to identify-ing the easiest routes to terawatt-scale deployment ofPV devices offering effi ciencies greater than 50%. Tecomplexity of the tasks required to bring the technol-ogy from concept to commercial realityin terms ofhighly effi cient and cost-competitive devicesmakes thedevelopment timeframe unavoidably long, measured by

the decade rather than year or month. Coordinated ef-forts by public and private stakeholders will be requiredto share in the costs and technology risks associated withmeeting societys long-term needs for clean generation.

echnical issues beyond module cost and performancealso must be addressed. EPRI, in collaboration with

many others, has been studying the potential impacts ofPV and other intermittent generation sources on gridoperations over the past 30 years. Results indicate thattodays distribution networks, which were not designedto accommodate numerous small-scale, intermittentgenerating units, can operate with moderate levels ofPV penetration (nominally at least 10% to 15% of localloads) without negative effects. Te full range of gridimpacts at much higher penetration levels is not yetwell understood, but PV-ready and PV-enabled systemsclearly will require automated controls and sophisticatedcommunications hardware that are not in place andmay not even exist today. Besides new hardware, thesegrids will need novel operational strategies, new analyti-cal tools, and forecasting methods like those developedfor wind power. Field data from experimental trials andlarge-scale applications will be needed to assess deploy-ment options and determine the business potential of

Grid-Connected PV in California





aggregating distributed and central-station PV withwind, hydro, DSM, plug-in hybrids, and other technolo-gies. Delivering output from central-station PV systemsdeployed in remote areas to serve loads elsewhere will ne-cessitate transmission system expansion, along with equi-table rules for cost allocation. Lower-cost, higher-perfor-mance storage systems will help transform intermittentsources such as PV into more valuable grid assets.

Collaborative R&D Responses

EPRI recently formed a Solar Electric Interest Group

(SEIG) to bring utility, industry, technology, and otherexperts together to explore common interests, issues, op-portunities, and challenges relating to solar power. TeSEIG supports peer interaction and discussion regardingcurrent PV technologies, applications, and market devel-opments, as well as roles for PV in meeting RPS require-ments and other business needs. SEIG participants alsobenefit from early access to results from solar energyR&D being performed by EPRI and others.

A couple projects funded through EPRIs echnologyInnovation Program are pursuing novel near-term PVapplications. In an ongoing demonstration, PV is beingused to power advanced wireless sensors deployed for on-line monitoring of aging substation components. Speci-fications for a nanotechnology-enabled lithium-ion bat-tery energy storage system are currently being developedto support 2008 demonstration of a dispatchable rooftopPV system offering both consumer and grid benefits.

Te Renewable Systems Interconnection (RSI) study, a

comprehensive project launched by the U.S. DOE inearly 2007 in conjunction with EPRI and other stake-holders, addresses technical and analytical issues associ-ated with supporting widespread PV deployment at thedistribution level and, eventually, on the transmissionsystem. Initial results appear consistent with earlierstudies indicating minimal grid effects for penetration ofintermittent PV systems at 15% or 20% of total regional

capacity and for much higher penetration levels if PVsystems are served by storage technology. RSI work alsoaddresses the state of the art and R&D needs relating tooperational and planning tools, storage and interconnec-tion technologies, and business models. Findings are tobe published in a series of 2008 reports.

o assist the industry in understanding, monitoring,and accelerating development of third-generation PVtechnology, EPRI launched the High-Effi ciency PVResearch Project in early 2007 in collaboration withElectricit de France (EDF) and others. Initial effort fo-

cuses on modeling and experiments to evaluate advancedmaterials and structures for their capacity to achieveconversion effi ciencies exceeding 40%. Tis work willbuild on worldwide R&D activities, including the $21.7million DOE program announced in November 2007 tosupport basic research and pre-commercial developmentfor third-generation concepts, with the goal being toproduce prototype devices and manufacturing processesby 2015. Te EPRI-EDF project will evaluate progressunder this program and others to support creation of atechnical roadmap defining the further R&D required to

bring a preferred PV device concept to market.

Clearly, PV technology is headed for increasing relevanceto, and significant impacts on, the electric sector. Utili-ties and other businesses that attentively track its prog-ress are most likely to be successful in addressing thechallenges it will create and benefiting from the opportu-nities it will bring. Substantial and sustained R&D com-mitments to advance all aspects of the technologyfromPV modules and auxiliary hardware to their interconnec-

tion with grid operations, power markets, and businesssystemsare yet needed to establish PV as a globally sig-nificant contributor to the worlds portfolio of low- andnon-emitting generation options. EPRI is committed tobeing on the forefront of PV development and applica-tion and to highlighting emerging issues and transferringkey findings to utilities and other stakeholders.


28/28

The Electric Power Research Institute (EPRI)

The Electric Power Research Institute (EPRI), with major locations

in Palo Alto, California; Charlotte, North Carolina; and Knoxville,

Tennessee, was established in 1973 as an independent, nonprofit

center for public interest energy and environmental research.

EPRI brings together members, participants, the Institutesscientists and engineers, and other leading experts to work

collaboratively on solutions to the challenges of electric power.

These solutions span nearly every area of electric ity generation,

delivery, and use, including health, safety, and environment.

EPRIs members represent over 90% of the electric ity generated

in the United States. International participation represents nearly

15% of EPRIs total research, development, and demonstration

program.

TogetherShaping the Future of Electricity

2007 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric PowerResearch Institute, EPRI, and TOGETHER...SHAPING THE FUTURE OF ELECTRICITY areregistered service marks of the Electric Power Research Institute, Inc.

Printed on recycled paper in the United States of America.

EPRI Contact

om Key, echnical Leader, Renewables & [email protected]

EPRI ResourcesSolar Electric Interest Group (www.epri.com/SEIG)Recent Publications

High-Effi ciency Photovoltaic Research Project(1015473), 2007

Renewable Energy echnical Assessment GuideAG-RE: 2006 (1012722), 2007

Program on echnology Innovation: Strategizing a New

Approach to EPRIs Solar Electric Research(1014603), 2006

Solar Electric Interest Group (1014470), 2006Other Resources

U.S. National Renewable Energy Laboratory Photo-voltaics Research (www.nrel.gov/pv)

U.S. Department of Energy Solar Energy echnolo-

gies Program (www.eere.energy.gov/solar)

Solar Electric Power Association(www.solarelectricpower.org)

Solar Energy Industries Association (www.seia.org)

European Photovoltaic Industry Association(www.epia.org)

Electric Power Research Institute

3420 Hillview Avenue, Palo Alto, California 94304 PO Box 10412, Palo Alto, California 94303 USA800.313.3774 650.855.2121 [email protected] www.epri.com

1016279 December 2007

EPRI Photovoltaic_White_Paper_1207

Documents

Transcript of EPRI Photovoltaic_White_Paper_1207