Figure 1. Decline in Photovoltaic Module Prices, 1975-1998

In 1954, Bell Laboratories researchers announced the development of a silicon solar cell with a 4.5-percent energy efficiency,⁽¹⁾ sparking photovoltaic (PV) cell development that has progressed from space applications in the late 1950s to terrestrial applications today. Over this period, research and development have resulted in lower prices for solar cells and modules (Figure 1) and higher efficiency. U.S.-based photovoltaic manufacturers' development efforts have benefitted from a partnership with the Federal government. Similar partnerships at the State level have also been beneficial. Additionally, rising electricity prices and an increase in the cost of building new generation, transmission, and distribution capacity have had a positive impact on photovoltaic system economics and sales. Also during this period, photovoltaic system sales have expanded as a solution to remote distributed generation require- ments. In such markets, photovoltaic systems often prove to be cost effective when compared to the common distributed generation alternative, diesel generators, which may be high priced because of the cost of transporting fuel to remote regions.

Figure 2. World Photovoltaic Shipments, 1992-1999 Figure 3. Photovoltaic Shipments Market Share, 1992-1999

More recently, photovoltaic cell and module shipments have grown on an international scale. Data for 1999 show 201 peak megawatts (MWp) of worldwide shipments (Figure 2). Shipments from manufacturing capacity in the United States and Japan dominate the market, with about 30 percent of shipments from the United States and about 40 percent of shipments from Japan (Figure 3). This represents a marked change from 1995, when U.S.-based manufacturing capacity accounted for 45 percent of world shipments, with Japan at 26 percent. The increase in Japanese market share isdue to growth of the building-integrated photovoltaic (BIPV) applications market in Japan, which benefits from Ministry of International Trade and Industry (MITI) programs, subsidies, and net metering regulations.

The following analysis discusses the dynamics of the international photovoltaic (PV) market, addressing the activities of PV manufacturers and consumers that have shaped the international market and their impact on the U.S. domestic PV industry. It will explain three major features of recent PV manufacturing and shipment history.

Three Major Features

Figure 4. U.S. Photovoltaic Cell and Module Shipments, 1983-1999

History

The market for photovoltaic systems has developed in three stages, distinguished by the type of application and by the focus of State, Federal, and international market development initiatives.

Space Program

During the first stage (1950s through 1960s), PV development was motivated primarily by a need for electricity generation technology that would be suited for the space program. In 1958, Vanguard I became the first PV-powered satellite. The 0.1 watt (W), approximately 100 cm² (square centimeters), silicon cell system powered a 5 milliwatt backup transmitter for 8 years.⁽⁵⁾ It offered a relatively lightweight solution to power supply for satellites and spacecraft. The single-crystal silicon photovoltaic cells deployed in space in the late 1950s had cell efficiencies that ranged from 8 to 10 percent.⁽⁶⁾ By 1998, efficiencies of modules made from such cells had increased to between 14 percent and 16 percent.⁽⁷⁾

Oil Price Pressures

The second stage (1970s through mid-1980s) commenced with the Arab OPEC oil embargo of 1973, which resulted in a significant increase in oil prices. One response in the United States and other countries was to fund development of renewable and energy-efficient technologies that would relieve dependence on fossil fuels. Federal and State tax credits for both residential and commercial customers subsidized expansion of terrestrial applications markets during this period. In addition, in 1978, the Public Utilities Regulatory Policy Act (PURPA) provided another market development support by guaranteeing "qualifying facilities" access to the electricity utility grid and requiring utilities to purchase the electricity. In California, the Standard Offer Number 4 electricity purchase contract offered renewable electric "qualifying facilities" a very attractive purchase price, which was guaranteed for a period of 10 years. Qualifying facilities included renewable electric generators, such as photovoltaic systems. By the late 1980s, Federal tax credits had expired and other market mechanisms for new applicants were terminated. The result was a significant drop in the addition of new photovoltaic electric generation capacity.

Globalization of the Market

Japan has a subsidy program goal of increasing PV demand by 400 MW per year through 2010 and Germany has a goal of 100 MW per year through 2005. This increased demand is being met by domestic cell and module production and imports from the United States.

Domestic and International Supply

U.S.-based manufacturers had an early market lead based on inventing and patenting PV technology. This lead is being challenged by competition from countries such as Japan and Germany. This international competition, along with years of manufacturing experience and government research and development funding, has produced gains in photovoltaic module energy efficiency and cost reductions. New photovoltaic technologies that show promise for further energy efficiency gains and cost reductions are starting to emerge. However, single crystal silicon technology continues to dominate both U.S. and some international cell and module shipments (Figures 5 and 6). U.S. photovoltaic cell and module shipments are shown in Figure 7. The following section reviews manufacturing and research trends. It also discusses the impact that factors such as an educated labor force, Federal and State support of research and development (R&D), and availability of venture capital have on growth of manufacturing capacity in a country.

Figure 5. U.S. Shipments by Cell/Module Type, 1993-1998 Figure 6. World Shipments by Module Type, 1998

Figure 7. U.S. Photovoltaic Cell and Module Shipments by End Use, 1994-1998

U.S. and International Shipment and Capacity Trends

Manufacturing Strategies

Photovoltaic manufacturers have developed the following diverse strategies for competing in global markets:

Locating Near End-Use Markets. Manufacturers benefit from the end-user and system installer feedback they gain on product design and performance when selling photovoltaic systems locally. This can be integrated into improved system design, including balance of system improvements, which may result in cost reductions. Manufacturers hope this will support increased sales by providing end-users with desired features. Increased sales help reduce the cost per kW price of a PV module by spreading development and overhead costs over a higher kW sales volume.

The Spire Corporation/BP Solarex venture in Chicago is an example of the trend toward locating manufacturing capacity close to end-users. PV modules will be manufactured in Chicago and the modules, incorporated into solar systems, will be marketed to residential and commercial customers in the Midwest. The Spire agreement with the City of Chicago and Commonwealth Edison (ComEd), the local utility, will provide $8 million of PV systems. Funding from ComEd shareholders accounts for $6 million.⁽¹⁰⁾ The remaining $2 million will be funded from the City of Chicago's budget. Installing PV systems on schools is a priority. ComEd has first right of refusal on an additional $6 million of PV systems. Manufacturing plants built to service such markets are generally small, modular plants.

If proximity to the end-use market is beneficial, then U.S.-based manufacturers, who export most of their product, may be at a disadvantage when it comes to (1) designing and manufacturing photovoltaic products to meet most of their end-users' needs and (2) benefitting from the lower system costs per kW that may result from advances in product design and from increasedsales of systems that meet end-user design requirements. U.S.-based manufacturers compensate for their distance from many end-use markets with a willingness to place technically trained marketing representatives on site around the world. They also engineer cells and modules for long-term trouble-free operation, covering them with warranties of 20 to 25 years.

Production in Japan and Germany is growing, despite high labor costs in both countries compared with the United States. High labor costs are offset, however, by strong domestic markets, which enable emerging photovoltaic technology product development and cost reduction efforts to benefit from end-user feedback. Strong domestic markets also enable Japan and Germany to export lower cost systems.

Changing Plant Capacity. As mentioned above, there is a trend toward building smaller PV cell and module plants closer to end-user markets. These plants can be expanded as demand increases. This strategy is motivated by several factors.

Other manufacturers are taking the opposite approach, increasing plant size substantially. Large plants (e.g., over 20 MW) would be built to achieve economies of scale that will reduce the production cost of photovoltaic modules. For instance, as SMUD's residential grid-connected demand grows enough to support large capacity factories (40 MW and up), the wholesale price for a thin film module is expected to fall to $1/W from current costs of $4.50/W.

Price decreases are expected to occur in steps. When a higher capacity factory starts to produce modules, module prices will remain high until demand increases enough to take advantage of the economies of scale of the larger manufacturing plant. Breaking the $2/W manufacturing cost barrier for photovoltaic modules within the next 5 to 10 years will depend on high efficiency thin films (e.g., copper indium diselenide (CIS), cadmium telluride (CdTe)) and "next generation" production volume manufacturing facilities.⁽¹¹⁾ In Germany, Shell Renewables is following a strategy to build large facilities. They opened a 25-MW facility to manufacture cells in Gelsenkirchen, Germany in January 2000.⁽¹²⁾

Separation of Cell Manufacturing and Module Fabrication Operations. Photovoltaic cell manufacturing processes require technically qualified labor to produce quality cells. Thus, cell manufacturing operations are located in countries where such labor is available (e.g., United States, Japan, Germany). Assembly of cells into modules does not require the same level of technicalexpertise; therefore, manufacturers often ship cells to countries with end-use markets for assembly into modules. The practice helps keep photovoltaic module costs as low as possible because many countries where photovoltaic modules are deployed also have large pools of low-cost labor qualified for module assembly and because cells are less expensive to ship than modules. For example, in South Africa the strategy is to provide low-cost module assembly to meet demand generated by the South African program to promote photovoltaics for rural electric applications. South Africa has two module assembly plants, several wholesalers, and about 40 distributor/systems integration companies.⁽¹³⁾

In-Country Corporate Presence. Photovoltaic manufacturers may establish a cell or module manufacturing presence in a country to obtain preferential treatment. For instance, a country may exempt the manufacturer with domestic operations from certain tariffs. Additionally, countries such as Germany provide investment incentives for manufacturers to build plants. The companies have employed these strategies in various ways. In the United States, photovoltaic manufacturing firms have formed alliances with utilities, as well as located the manufacturing plant near the end users. Examples include Tucson Electric/Global Solar (Arizona) and GPU, Incorporated (New Jersey, Pennsylvania), a subsidiary of GPU International, Incorporated, a worldwide developer of independent powerplants, which operates GPU Solar as a joint venture with AstroPower, Inc., a photovoltaic module manufacturer.

Export Strategies

Both government and corporate photovoltaic technology development programs are directing funding toward photovoltaic technology that can be produced more cost-effectively. There are four or five independent technology paths to low-cost PV, ranging from continuation of crystalline silicon technology to thin film alternatives.

Lower Cost of Single Crystal Silicon

One approach is to continue trying to push the cost of single crystal silicon lower. However, cost reductions are hindered because feedstock for single crystal silicon cells is the waste silicon from the electronics industry. Increasing demand for waste silicon is leading to shortages.

In addition, the single crystal silicon cell is thick compared to thin film alternatives. Use of more material increases product cost. On the positive side, single crystal silicon modules still command an energy conversion efficiency premium per square meter over alternative PV products. In addition, crystal silicon is a known material with years of proven performance in the field. Thus, single crystal silicon modules have an advantage over other PV flat-plate module technologies in applications where space is at a premium.

Another approach is amorphous silicon, which may be viewed as a transitional technology, since it has a lower energy efficiency than alternatives and since amorphous silicon modules must be aged prior to sale to ensure that their energy efficiency remains stable. Copper indium diselenide (CIS) is the leading material for amorphous silicon technology. The current problem with CIS is availability; Siemens Solar is manufacturing only pre-commercial market conditioning volumes.⁽¹⁷⁾ For the CIS market to develop, purchases in the 100 kW range are needed. To support such purchases, production in the one megawatt per year range is needed.

United States National Photovoltaics Program

The National Photovoltaics Program, funded by the U.S. Department of Energy, involves national laboratories, universities, and industry stakeholders in cooperative research and development of photovoltaic systems to attain higher module energy efficiencies, lower system costs, and longer system life. The long-term goal of the program is to make photovoltaic electricity available at an operating cost of $0.06/kWh. Current program goals were established by U.S.-based photovoltaic industry members to establish a "roadmap" for future industry development (Table 8).⁽¹⁸⁾ The roadmap's goal for shipments is 25 percent annual growth in shipments from manufacturing facilities based in the United States. This growth rate would result in at least 6 gigawatts-peak (GWp) installed worldwide by 2020 from manufacturing capacity based in the United States, including 3.2 GWp of domestic installations.⁽¹⁹⁾ The 3.2 GWp target assumes (1) a constant U.S. share of worldwide annual shipments of 40 percent and (2) installation of 30 percent of U.S. shipments in the United States in the year 2000, increasing to 50 percent by 2020. The expected application mix for the 3.2 GWp is the following:

• 50 percent alternating current (AC) distributed generation (remote, off-grid power for applications including cabins, village power, and communications)




• 33 percent direct current (DC) and AC value applications (consumer products such as cell phones, calculators, and camping equipment), and




• 17 percent AC grid (wholesale) generation (grid-connected systems including BIPV systems).⁽²⁰⁾

For FY2000, the Federal PV research and development program is funded at a level of $65.9 million (Table 9). The program is divided into three areas:

• Fundamental Research. Support industry and university research to characterize cell materials and devices; conduct research to understand defects in conventional crystalline silicon and thin film materials; and develop techniques to reduce efficiency-limiting defects in cell material; increase the efficiency of multijunction concentrating cells and large-area, monolithically interconnected thin films.




• Advanced Materials and Devices. Develop next generation thin film technologies through cost-shared efforts with industry and universities. This effort includes support of first-time manufacturing and scale-up of thin film amorphous silicon, CIS, CdTe, and thin silicon. Develop high efficiency crystalline silicon devices, emphasizing manufacturing methods that reduce cost.




• Technology Development. Develop manufacturing methods that result in lower cost, higher efficiency modules and in lower cost PV system components (e.g., batteries and inverters). This effort has included the Photovoltaic Manufacturing Technology (PVMaT) initiative, which addresses systems engineering and reliability issues through activities such as testing, developing domestic and international standards and codes, and analyzing factors affecting stability of encapsulated materials and performance of cells in modules. Technology development also includes: (1) developing advanced PV building concepts, tools, and modeling procedures; (2) motivating introduction of PV intobuilding systems through cost-shared projects (Partnerships for Technology Introduction) and support of the Million Solar Roofs Initiative; and (3) accelerating introduction of photovoltaic power as a rural electrification option for developing countries by developing prototype systems, advancing the concept of international equipment standards, and developing tools for analyzing distributed photovoltaic opportunities (International Clean Energy Initiative).

The Partnerships for Technology Introduction, Million Solar Roofs Initiative, and International Clean Energy Initiative elements of the Technology Development budget address market stimulation through funding of cost-shared projects, prototype systems, and activities to promote formation of Million Solar Roofs partnerships. None of the $1.5 million for the Million Solar Roofs Initiative is an end-use incentive.

Figure 8. Federal Photovoltaic R&D Budgets, United States, Japan, and Germany, 1981-1999

Japanese and German National Photovoltaic Development Programs

The Japanese and German development programs have provided competition for the United States over the years. For instance, during the 8-year period from 1981 to 1988, the German and Japanese Federal PV R&D budgets increased, while the U.S. Federal budget fell (Figure 8). Recent funding data show the willingness of the Japanese government to spend relatively large amounts on direct market stimulation for end uses to promote their building photovoltaic program. They are funding market stimulation at a rate over four times that spent by either the United States or German programs (Table 10). Data indicate that the Japanese PV promotional budget rose steadily from $53 million in 1995 to $132 million in 1998.⁽²¹⁾

Table 10. Research, Development, Demonstration, and Market Stimulation Budget Comparison, Fiscal Year 1998 (Million U.S. Dollars)
Program Area	United States	Japan	Germany
R & D	64.7	56.1	38.3
Demonstration	--	21.4	--
Market Stimulation	*	132.5	18.4
Total Budget	64.7	210.0	56.7
- = Not applicable.    * In FY 1998, about $30 million of the U.S. $64.7 million R&D budget was spent on a combination of market stimulation-related activities (market transformation, research initiatives, application-specific research, and manufacturing process research). These expenditures are included in the R&D budget for the United States because their objective is related more to R&D than to market stimulation. Market stimulation amounts shown for Japan and Germany reflect payment of subsidies to reduce the cost of photovoltaic systems.    Sources: International Energy Agency, Trends in Photovoltaic Applications in Selected IEA Countries Between 1992 and 1998 (IEA-PVPS 1-07:1999) (Paris, France, October 1999), p. 6. R&D budgets for Japan (Sunshine PV Program budget) and Germany (Federal Department of Education, Science, Research and Technology budget) from Jack L. Stone, National Renewable Energy Laboratory, National Center for Photovoltaics.

U.S. and International Demand

In 1999, worldwide shipments of PV cells and modules totaled 201 MW,⁽²²⁾ a 30-percent increase over 1998 worldwide shipments of 155 MW. U.S. manufacturers shipped just under 51 MWp of the total 1998 worldwide photovoltaic cell and module shipments. Factors motivating photovoltaic sales included Federal government and State tax incentives, utility rebate programs, "green" pricing programs, and donor agency programs to install photovoltaic systems in developing economies.

Over 80 percent of 1998 shipments by U.S. manufacturers went to the following end uses: remote and grid interactive electricity generation (45 percent); communications (16 percent); transportation, e.g., power on boats, in cars, in recreational vehicles, and transportation support systems (13 percent); and water pumping (9 percent). Key market niches encompassed by these end uses include building integrated photovoltaics promoted by utilities and national climate change or green power initiatives; other village, rural, or distributed generation applications in both developed and emerging economies; water pumping and irrigation systems, communications, and consumer products.

The following sections characterize these markets and discuss factors that influence demand.

U.S. Demand

The U.S. market is characterized by several niches that accounted for 15 MWp of cell and module shipments from manufacturing facilities in the United States in 1998. The domestic U.S. market includes the following segments, defined by application:⁽²³⁾

Building Integrated Photovoltaics (BIPV). These are PV arrays mounted on building roofs or facades. For residential buildings, analyses have assumed BIPV capacities of up to 4 kWp per residence. Systems may consist of conventional PV modules or PV shingles. This market segment includes hybrid power systems, combining diesel generator set, battery, and photovoltaic generation capacity for off-grid remote cabins.

Non-BIPV Electricity Generation (grid interactive and remote). This includes distributed generation (e.g., standalone PV systems or hybrid systems including diesel generators, battery storage, and other renewable technologies), water pumping and power for irrigation systems, and power for cathodic protection. The U.S. Coast Guard has installed over 20,000 PV-powered navigational aids (e.g., warning buoys and shore markers) since 1984.⁽²⁴⁾

Communications. PV systems provide power for remote telecommunications repeaters, fiber-optic amplifiers, rural telephones, and highway call boxes. Photovoltaic modules provide power for remote data acquisition for both land-based and offshore operations in the oil and gas industries.

Transportation. Examples include power on boats, in cars, in recreational vehicles, and for transportation support systems such as message boards or warning signals on streets and highways.

Consumer Electronics. A few examples are calculators; watches; portable and landscaping lights; portable, lightweight PV modules for recreational use; and battery chargers.

Market growth in each segment is affected by countervailing factors. The primary factor thwarting growth is the installed cost per kilowatt of the photovoltaic system, which often causes the cost of electricity (e.g., cents per kilowatthour) from such systems to be higher than the cost of electricity produced by fossil-fired or hydropower generation alternatives. National and international research efforts focus on ways to reduce the cost of photovoltaic systems.

Cost-Effective Markets

Near-term market growth is occurring where the end-use is in a remote location or the measurable cost ofelectricity from alternative generation technologies is high enough for photovoltaic systems to be cost-effective. U.S. distributors have identified markets where photovoltaic power is cost-effective now, without subsidies. Examples include the following: (1) rural telephones and highway call boxes, (2) remote data acquisition for both land-based and offshore operations in the oil and gas industries, (3) message boards or warning signals on streets and highways, and (4) off-grid remote cabins, as part of a hybrid power system including batteries.⁽²⁵⁾

The current installed cost of photovoltaic systems ranges from $0.20 to $0.50 per kilowatthour, depending on factors such as the volume purchased and the level of solar insolation. Therefore, the electric price of the next best alternative must be no lower than this range for PV to be cost-effective. High electric prices tend to be found where there is no cost-effective access to the electric grid (e.g., remote applications markets, including distributed generation, telecommunications, navigational aids, and cathodic protection). Diesel generator sets are the alternative to photovoltaic electricity in some of these markets. In remote applications, diesel generator sets may be at a disadvantage to PV because these systems bear high costs of hauling fuel to the site, storing fuel, and maintaining equipment.

In the longer term, it will take a combination of wholesale system price below $3.00/W and large volume dealers for PV to be cost-effective in the residential grid-connected market. PV installed system costs must fall to a range where they are competitive with current retail electric rates of $0.08 to $0.12/kWh in the residential market and $0.06 to $0.07/kWh in the commercial market.⁽²⁶⁾

Photovoltaic "Green" Power