Industry Structure and Company Strategiesof Major Domestic and Foreign Wind and Solar Energy Manufacturers:Opportunities for Supply Chain Development in AppalachiaCo-Principal InvestigatorsGerald I. Susman, Ph.D.Smeal College of Business210 Business Building, University Park, PA 16802814-863-0448Amy K. Glasmeier, Ph.D.College of Earth and Mineral Sciences302 Walker Building, University Park, PA 16802814-865-7323Technical AdvisorsSusan K. Stewart, Ph.D.Research Associate, Energy Science and Power Systems, Applied Research LaboratoryDavid R. Riley, Ph.D.Associate Professor, Architectural Engineering, College of EngineeringThe Pennsylvania State UniversityUniversity Park, PA 16802ResearchersJared Freer, Graduate AssistantBarbara B. Kinne, Research AssistantMichael H. Patullo, Research ConsultantJenna P. Stites, Research AssistantCarmen Strand, Research AssistantMichael Waldhier, Graduate AssistantARC Project Number CO-15810-07October 1, 2007–February 27, 2009Final Report SubmittedNovember 20, 2009Project DirectorGerald I. Susman, Ph.D.Smeal College of BusinessThe Pennsylvania State University814-863-0448

TABLE OF CONTENTSData Dictionary .5Executive Summary .6Phase I. Industry Market Structure, Forecast, and Potential:Part 1: The Solar Industry.8Demand for Solar Energy .8Solar Energy Technologies .10Markets and Applications .10Industry Participants .12Photovoltaic (PV) Supply Chain.16Competitive Strategies .17Strategic Groups.21Future Directions .25Part 2: The Wind IndustryDemand for Wind Energy .27Policies that Stimulate Demand for Wind Energy .27Wind Energy Systems .30Markets and Applications .31Industry Participants .32Wind System Supply Chain .35Competitive Strategies .40Industry Evolution .48Future Directions .50Phase II: Solar and Wind Energy Industry Participation within the Appalachian RegionIntroduction .52NAICS Codes to Identify Potential Firmsand Employment in Solar and Wind Industries .52Part 1. Pattern of Manufacturing Activities and Potential Employment in Appalachia .54Part 2. Identifying and Surveying Firms Involved in the Solar and Wind Industries .58Analysis of Survey Results .68Part 3. Review of the Policy Landscape in the Appalachian Region.77Demand-Side Incentives .78Supply-Side Incentives .82Novel Policy Strategies .83Major Findings and Observations .85Bibliography .912

List of TablesTable 1Table 2Table 3Table 4Table 5Table 6Table 7Table 8Table 9Table 10Table 11Table 12Table 13Table 14Table 15Table 16Table 17Table 18Table 19Table 20Table 21Table 22Shipments of PV Cells and Modules by Application (peak kilowatts).11Shipments of PV Cells and Modules by Market and Type (peak kilowatts) .12Materials, Components, and Equipment Suppliers in Appalachian Counties .14Distributors/Installers in Appalachian Counties .15Plant Locations of Top Fifteen Cell Manufacturers .18Worldwide MW Additions, Plant Locations, and U.S. Installationsof Top Ten Wind Turbine Manufacturers in 2008.33Wind Turbine Suppliers and Locations .37Component and Equipment Suppliers in Appalachian Counties .39Product Line Range of the Global Top-Ten Turbine Manufacturers .41Size Distribution of Turbines from 1998–2007 .42R&D as a Percentage of Sales for Top-Ten Turbine Manufacturers .45NAICS 2007 Codes of Solar and Wind Industry Participants .62Founding Years of Responding Firms .69NAICS 2007 Codes of Survey Respondents .70Sectors Served by Founding Years of Responding Firms .71Position in the Supply Chain.71Products/Services Provided by Survey Respondents, by Sector .72Preparedness of Employees for Participation in the Solar or Wind Industry .75Renewable Portfolio Standards in Appalachian States .78Tax Incentives that Apply to Solar and/or Wind Installations, by State .79ARC State Grant and Loan Programs Applicable to Solar and/or Wind .80Profile of Typical Established and Emergent Firms .87List of FiguresFigure 1Figure 2Figure 3Figure 4Figure 5Figure 6Figure 7Figure 8Figure 9Figure 10Figure 11Figure 12PV Solar Supply-Chain .17Strategic Groups––PV Solar Industry.24Supply-Chains for Two Major Wind System Components .36Strategic Groups in the Wind Industry – 2004 .49Potential Renewable Energy Manufacturing Employment in ARC Counties(by State) .55Total Establishments with Renewable Manufacturing Potentialin ARC Counties (by State) .55Counties with Potential Solar Manufacturing Jobs over 500.56The Number of Firms and Components for Counties with PotentialSolar Job Totals over 500 .56Counties with Potential Wind Manufacturing Job Totals over 1,000.57Number of Firms and Components in Counties with PotentialWind Manufacturing Jobs over 1,000 .57Phase II––Construction of Firm Database .60Collection of Firms .613

List of MapsMap 1Map 2Map 3Map 4Potential Participants in Solar and/or Wind Industry .64Participants in Solar and/or Wind Industry .65Manufacturers in Solar and/or Wind Industry .66Service Providers (Including Installers) in Solar and/or Wind Industry.67List of AppendicesAppendix IAppendix IIProfiles of Solar Energy Industry Companies by Strategic Groups .99Acquisitions, Partnerships, and Framework Agreements between TurbineManufacturers, Suppliers and Buyers .104Appendix III NAICS Codes for Manufacturing Firms with Technical Potential to EnterSolar PV and/or Wind Turbine Markets .109Appendix IV Summary State Potential Employment and Establishment Data by RenewableResource .110Appendix V Survey Instrument .112Appendix VI Database of Potential Participants in Solar and/or Wind Energy Industry .1174

DATA AFeed-in TariffFERCGWIPOIPPISOITCKerf LosskWkWhLikert PSRTOSCADAVEETCWAppalachian Regional CommissionAmerican Wind Energy AssociationBuilding Integrated PhotovoltaicComputer Numerical ControlledConcentrated PhotovoltaicConcentrated Solar PowerConcentrated Power SystemsDouble Fed Induction GeneratorsEuropean Economic CommunityEnergy Policy ActEnergy PhotovoltaicEthyl Vinyl AcetateThe price per unit of electricity that a utility or supplier has to pay forrenewable electricity from private generators. The government regulatesthe tariff rate. []Federal Energy Regulatory CommissionGigawatt [1 gigawatt 1,000 megawatts 1 billion watts]Intellectual Property OwnerIndependent Power ProducersIndependent System OperatorsInvestment Tax CreditMaterial loss associated with any type of cutting and sectioningKilowatt [1 kilowatt 1,000 watts]Kilowatt Hour [1 kWh 1,000 watts of electricity used for one hour]Psychometric scale used in attitude/opinion researchMegawatts [1 megawatt 1,000 kilowatts 1 million watts]Megawatt Hour [1 MWh 1,000 kilowatts of electricity used for one hour]North American Industrial Classification SystemNational Renewable Energy LaboratoryNew York State Energy Research and Development AuthorityOrganization for Economic Co-operation and DevelopmentPermanent MagnetPower Purchase AgreementProduction Tax CreditPhotovoltaicRenewable Energy CreditRenewable Energy StandardsReturn on InvestmentRenewable Portfolio StandardsRegional Transportation OrganizationSupervisory Control And Data AcquisitionVolumetric Ethanol Excise TaxWatt [standard unit of power, or energy unit per time]5

EXECUTIVE SUMMARYThis report presents results from a two-phase study of the status of the solar and wind industriesin the U.S., with special focus on product and service suppliers in the thirteen Appalachianstates, and the challenges these firms and their state governments face in preparing for andcompeting in these two rapidly emerging worldwide industries.Phase I concerns the structure of the solar and wind energy industries, and focuses on sets offirms that follow similar competitive strategies. A brief overview of the dimensions of thesecompetitive strategies is provided, including choice of market, geographical region, breadth ofproduct line, and vertical integration. Other factors that influence competitive strategy are costdynamics, differentiation, technology choice, and technology leadership. The evolution of theseindustries depends on demand that is stimulated by government mandates, feed-in tariffs, taxincentives, rebates, price of conventional energy and carbon offsets. It also depends on supplythat is influenced by production capacity, availability of raw materials, process innovation, rateof learning, and economies of scale.Some of these demand and supply factors affect all firms in these industries equally, while othersaffect strategic groups differently, and thus their current and future market share andprofitability. The performance of firms in these industries also depends on strategic choices (e.g.,preemptive moves, plant location, and rate of market expansion). This report provides anoverview of current and projected structure in these two industries, and speculates on thechallenges that firms within these industries face now and in the future.Phase II, Part 1 reviews a previously sponsored ARC study that focuses on the spatial location ofestablishments and employment in the component elements of the solar and wind industries. Thestudy adopted a commonly used methodology that relates NAICS codes that are associated withmanufacturing solar and wind components to establishments in the targeted region. The studyidentified two distinct geographical patterns: concentration and dispersal; that is, although mostjobs and plants are concentrated in a few predominantly urban counties, a significant number ofcounties have at least one plant within them.At the state level, potential solar and wind employment and plants are found in relatively fewstates. Pennsylvania, Tennessee, North Carolina, South Carolina, Georgia and New Yorkembrace the lion’s share of employment and plants in the Appalachian region. Pennsylvaniaalone accounts for 30% of total employment in the two industries. At the county level, a state’splants and potential employment in the wind or solar industry is concentrated in very few,generally urban counties. In eight of the 13 ARC states, 30% or more of solar-relatedemployment is concentrated in one county. Additionally, single plants rather than small numbersof jobs are found in many counties.Part 2 presents findings from a survey of firms that operate facilities in the ARC region. Thesefirms were identified by using relevant NAICS codes and by consulting industry associationwebsites and published reports that list firms reputed to be in the solar or wind industry. Thesurvey documented the characteristics of the firms (e.g., age, size, corporate structure), theirtypical means of market entry, their awareness of their competitive context, the extent of their6

awareness of and need for specific resources, including skilled labor, and the extent of theirinvolvement in international markets.Of the 363 firms in ARC designated counties we contacted in the survey, 72 (20%) reported thatthey were involved in the solar or wind industry. The survey revealed a limited number ofmanufacturing firms that were potential suppliers to the solar or wind industry. Manufacturersare generally older, established firms with a small percentage of their domestic or internationalsales derived from the solar or wind industry. Few reported that the solar or wind industry wasthe primary function of the business. We found many more service providers that had thecapacity to serve as installers or distributors of renewable energy products, primarily toresidential customers. The barriers to market entry are lower for them than for theirmanufacturing counterparts. The workforce of established firms requires specialized training thatis transferable between making renewable and non-renewable products. The workforce ofemergent firms focuses mainly on installing, servicing, or selling renewable energy products, andrequires more general skills development.In Phase II, Part 3, we examine the policy environment in the region on a state-by-state basis andthe programs available to stimulate and support the development of renewable energy industriesin the region. The states are predominantly emphasizing conventional economic developmentpractices, including tax abatements, location incentives and grant and loan programs. Seven ofthe 13 Appalachian states have a renewable energy portfolio standard or goal. New York andPennsylvania have aggressive policies that encourage experimentation and demonstrate a varietyof innovative industry-state collaborative approaches to solar and wind energy development.Although nationwide policies that would promote wind and solar industry development havebeen proposed, nothing has developed thus far. States that do have rapid growth in solar or windinstallations and/or manufacturing have introduced a set of mutually reinforcing policies thatlower the initial capital outlay for solar or wind installations (e.g., feed-in tariffs, rebates, lowinterest loans, sales or property tax abatement), have a renewable portfolio standard (usually witha solar or wind set-aside), and/or have energy costs above the national average, thus shorteningthe payback period for these investments.The highly decentralized policy environment that is characteristic of the U.S. has impededgrowth of the renewable energy industry. States have myriad policies that are varied and subjectto change. Appalachia, more than other regions of the nation, is unlikely to emerge as a leader inthe global renewable energy industry due to insufficient incentives and the general lack of asupportive policy framework that would encourage industry development. The presence of somany firms that could contribute to the wind and solar industry supply chain in Appalachiameans that there is a great deal of potential for development of the industry, given the right mixof policies and incentives. In some states, wind energy is at grid parity with conventional energysources. However, in states with low energy prices, energy efficiency improvements are a morecost-effective way to reduce energy costs and avoid carbon emissions. Thus, it may make senseto focus policy, incentives, and resources on improving energy efficiency in the near term, whilecontinuing to encourage a policy environment that is more conducive for development of thedomestic wind and solar industry.7

PHASE I. INDUSTRY MARKET STRUCTURE, FORECAST, AND POTENTIALPART 1: THE SOLAR INDUSTRYDemand for Solar EnergySolar cell production grew 85% in 2008; 7.9 gigawatts (GW) were added worldwide. The topfive photovoltaic (PV)-producing countries are China, Germany, Japan, Taiwan, and the U.S.1Cumulative PV power installed worldwide jumped from 9 GW in 2007 to almost 15 GW in20082. Worldwide photovoltaic (PV) installations reached a record high of 5.95 GW in 2008.Spain led the world in new solar installations in 2008 (2,011 MW); Germany (1,500), UnitedStates (342), Korea (274), Italy (258), and Japan (230) follow in that order3. Cumulativeinstallations in the U.S. in 2008 totaled 9,183 MW, an increase of 16% from 2007. Californiaaccounted for the lion’s share of installations (178.6 MW). New Jersey was second ininstallations (22.5 MW). Colorado (21.6 MW) Nevada (14.9 MW) and Hawaii (11.6 MW) arethird, fourth, and fifth, respectively4.Germany has one of the highest electricity rates (cost/kWh) in the world, and is heavilydependent on imported oil. That explains why political support in the German parliament wassufficient to enact a strong feed-in tariff in 1999 (revised in 2004 and 2008). The feed-in tariffrequires utilities and other power providers to buy renewable energy at above market rates for upto 20 years. Thus, they pay owners of solar panels more for the energy they generate (viarebates) than the owners pay utilities or independent power producers (IPPs) for conventionalenergy. This subsidy is scheduled to decrease each year in order to encourage the industry topass on lower costs to the end users. The feed-in tariff in Germany is 0.50-0.60 USD/kWh. It islower in other EEC countries, but still substantial. Spain’s generous feed-in tariff prompted ahuge increase in solar installations during 2008. In September, the government significantlyreduced payments under the feed-in tariff and capped annual PV installation from 2009 through2010, aiming at a target of 3,000 MW by the end of 20105.One might expect demand to be highest where solar is most efficient; for example, where thehours/days of solar radiation per year (i.e., insolation) are highest, but this fact matters far lessthan government subsidies. Solar insolation in Los Angeles, California is 5.62 KWh/m2/daywhen a solar array is providing peak output; it is 2.63 kWh/m2/day in Hamburg, Germany6. Thisclearly suggests that Germany’s current leadership in cumulative solar installations is less relatedto insolation than it is to subsidies.1PHOTON International. (2009, March). Little smiles on long faces. PHOTON International, p. 170.Li, Y. (2009, June 20). Solar Power Experiences Strongest Year of Growth Yet. Worldwatch Institute.3LaPedas, M. (2009, March 24). U.S. lags in top 10 solar markets. EE Times.4Solar Energy Industries Association. (2009, March). US solar industry year in (2008, September 29). Spain Makes Changes to Solar Tariff. Retrieved Atmospheric Science Data Center. (2009). NASA surface meteorology and solar energy data set. Retrievedfrom s01#s0128

In the U.S., investors in solar energy are allowed a 30% investment tax credit (ITC)7. Also mostforms of renewable energy are eligible for accelerated depreciation over five years. TheEmergency Economic Stabilization Act (October 2008) extended the ITC for eight years,removed the 2,000 cap on residential installations and allowed participation of utilities. TheAmerican Reconstruction and Recovery Act (February 2009) provides commercial businesseswith a cash payment to cover 30% of the cost of installing solar equipment. The Act also createda fund to provide up to 60 billion in loan guarantees for renewable energy and transmissionprojects8.Twenty-nine U.S. states and the District of Columbia have renewable portfolio standards (RPSs),which mandate that a certain percentage of renewable energy be available by a specific date9.Five states have renewable portfolio goals (Virginia, Vermont, North Dakota, South Dakota, andUtah). Liberal RPSs include those in California (20% by 2010), New Jersey (20% by 2020 ofwhich 1,500 MW must be solar) and New York (24% by 2013). Some states also offer cashincentives or rebates for solar investments. For example, the California Solar Initiative offerscash incentives up to 2.50 per watt (based on system performance) for installations on existinghomes in areas served by specified public utilities. California has set a goal to create 3,000 MWof new solar-produced electricity by 2017 at a cost of 3.3 billion. Solar must be offered as anoption on all new homes in 2011. Other states offer a mix of grants, loans and rebates to supportthe goals of their RPS.Several states also allow renewable energy credits (RECs) to be traded like commodities. RECowners can claim to have purchased renewable energy equal to 1 MWh of electricity that wasgenerated from an eligible renewable energy source. Buyers of RECs (e.g., utilities) raise thecost of producing conventional electricity (to comply with a state’s RPS) and subsidizeproducers of electricity generated from renewable sources (REC sellers).Subsidies are expected to remain in place in most countries until “grid parity” is reachedsometime around 2012. Grid-parity (when the price/kWh for solar is the same as grid-basedprice/kWh for oil, gas, coal) already exists in California and New Jersey, and in 15% of OECDcountries (for peak load rates). Solar demand in the U.S. is highest in areas with the highestprice/kWh, e.g., New England ( 0.14), Mid-Atlantic ( 0.11), and California ( 0.12)10. Mostnorth and south central states and southeastern states have low average cost per kWh ( 0.06 0.07), so there are fewer subsidies and less use of solar energy.7Database of State Incentives for Renewables & Efficiency (DSIRE). (n.d.). Retrieved August 25, 2008, fromhttp://www.dsireusa.org8Solar Energy Industries Association. (2009, March). US solar industry year in review.9Database of State Incentives for Renewables & Efficiency (DSIRE). (n.d.). Retrieved August 25, 2008, fromhttp://www.dsireusa.org10Think Energy Management. (n.d.). Electricity Costs. Retrieved from

Solar Energy TechnologiesThe most common solar energy technology is based on the photovoltaic effect that wasdiscovered by A.E. Becquerel in 1830. This effect occurs in solar cells that are comprised of twolayers of semiconducting material: P and N-. The boundary between P and N- acts as a diode:electrons can move from N- to P but not the other way. The voltage difference can be used as apower source. The P and N- layers are created by doping silicon or similar materials with boronand phosphorous, respectively. Silicon-based PV cells (mono- and multi-crystalline) made up87% market share in 2007. Thin-film-based PV cells make up about 13% of market share11.The conversion efficiency (ratio of sunlight to energy produced) of mono-crystalline cells ishigher than for multi-crystalline cells12. However, multi-crystalline cells are cheaper to make.The conversion efficiency of thin-film is lower than for silicon-based cells, but they have otheradvantages. For example, they use much less material – the cell's active area is usually only 1 to10 micrometers thick, whereas silicon-based cells typically are 100 to 300 micrometers thick.Also, thin-film cells usually can be produced with an automated, continuous production process.Finally, thin-film material can be deposited on flexible substrates (e.g., metal, plastic, glass) thatenhance their utility, (e.g., integrated into roofs and windows). Major thin-film producers are:Uni-Solar and EPV Solar (uses amorphous silicon); First Solar (uses cadmium telluride (CdTe));and Heliovolt, Nanosolar, Miasole (uses copper indium gallium selenide (CIGS)). Thin-filmmarket share is expected to grow to 20% in the next four years13. Finally, concentrator PV orCPV uses Fresnel lenses to concentrate diffuse sunlight onto a smaller, but highly focused cell ormodule area (Amonix, Concentrix).Solar thermal is another solar energy technology. Concentrating solar power (CSP) uses mirrorsto heat fluids and thereby create steam that drives turbines. CSP projects are large-scale andexpensive, thus mainly utilities or large power producers use this technology. Differentcompanies use different types of solar thermal technologies, e.g., parabolic troughs (Ausra andSchott Solar), dish-Stirling engines (Stirling Engine Systems), Distributed Power Towers (LuzII). Solar thermal technology can also be used to heat hot water tanks and swimming pools(Heliodyne, Thermomax).Markets and ApplicationsBy far, the most common application of solar energy in the U.S. is electricity generation forprivate and public buildings (94% - See Table 1)14. This percentage includes crystalline-based,thin-film silicon, and concentrator silicon cells and modules15. The remaining 6% of applicationsinclude government and industrial, (e.g., street lights, roadside call boxes, telecommunications,water pumps, and health). Also included are space applications, (e.g., satellites).11PHOTON International. (2008c, March). The Q factor, Sharp and the market. PHOTON International, p. (n.d.). Solar cell technologies. Retrieved from d, T., Grama, S., Wesoff, E., & Bhargava, A. (2007, August). The future of thin film solar. GreentechMedia and Prometheus Institute , 1 (1).14Energy Information Administation. (n.d.). Annual photovoltaic module/cell manufacturers survey, Form EIA-63B.15Although these figures are for cells and modules, they are a reasonable proxy for installed systems in the U.S. EIAdata show that 130,757 kilowatts (39%) were exported.1210

Table 1. Shipments of PV Cells and Modules by Application (peak kW)Application (end-use)ElectricityGrid ConnectedOff-GridCommunicationsConsumer GoodsTransportationWater PumpingOEMs16HealthTotalTotal (2007)Percent of Total 75.902. Energy Information Administration, Form EIA-63B, “Annual Module/Cell Manufacturing Survey”On-grid/Off-grid. Utilities and independent power providers (IPPs) generate solar energy from acentralized location (large PV power plant or CSP plant) to customers in relatively populatedareas via state or regional grids. In such cases, solar energy generally supplements conventionalenergy, and is especially useful during peak loads when conventional energy is most expensive.Off-grid applications are

In eight of the 13 ARC states, 30% or more of solar-related employment is concentrated in one county. Additionally, single plants rather than small numbers of jobs are found in many counties. Part 2 presents findings from a survey of firms that operate facilities in the ARC region. These