No abstract

No abstract

Until now virtually all of the development of terrestrial concentrator PV power plants has utilized silicon cell technology. If silicon concentrator technology achieves its predicted potential performance and cost [ 11, it would be very close to commercial viability as defined by DOE goals of 12 @/kWh. Also until now, GaAs cell technology has not been a viable alternative for these power plants, largely due to the cost of cells grown on GaAs substrates. The ability to grow high-efficiency GaAs solar cells on germanium substrates [2,3] presents the strong possibility of reducing their cost of energy produced.The cost of large-grain, optical-grade, polycrystalline germanium substrates is potentially 15 times less than single-crystal GaAs substrates [4], and cell conversion efficiencies approaching 35% are likely for tandem-junction GaAs cells. This analysis shows that for comparable 50 MW Fresnel lens plants, the 30-year, levelized cost of energy (COE) from GaAsIGe cells can be $0.118/kWh compared to $0.140lkWh for a 27.4% Si back-contact cell.

This paper presents the results of a recent study1 to assess the near-term cost of power in central station applications [ 11. Three PV technologies were evaluated: Fresnel-lens high-concentration photovoltaic (HCPV), central receiver HCPV, and flat-plate PV using thin-film copper indium diselenide (CIS) cell technology.Baseline assumptions included PV cell designs and performances projected for the 1995 timeframe, 25 and 100 MW/year cell manufacturing rates, 50 MW power plant size, and mature technology cost and performance estimates. The plant design characteristics are highlighted in Table 1. Potential sites were evaluated and selected for the PV power plants (Carrisa Plains, CA and Apalachicola, FL) and cell manufacturing plants (Dallas-Fort Worth, TX). Conceptual designs and cost estimates were developed for the plants and their components. Plant performance was modeled and the designs were optimized to minimize levelized energy costs. Cost estimates for both plant and energy delivered include effects of uncertainty in key parameters. Although the study did not involve detailed engineering, efforts were made to optimize all of the plant designs and minimize levelized energy costs. Cell and module fabrication processes were developed.Module prices were estimated for 25 and 100 MW annual production rates, with prices increasing 11 to 13 percent for the lower rate. The total capital requirement for the Fresnel lens and central receiver plants are close to k i n g equal, with the Fresnel lens 8 percent above their average and the central receiver 8 percent helow. The total c:ipital requirement for the flat plate plant is 23 percent lower than the average of the concentrator designs. However, the flat plate plant also generates 16 percent less annual energy at Ciurisa Plains. At the Apalachicola site, the average annual outputs of all of the plants are the same within +3 percent. 'Thus, meteorological conditions make the Apalachicola site more suited for flat plate plant designs.Overall, the flat plate design exhibited the lowest energy costs among the designs evaluated. Its levelized energy costs at the Carrisa Plains site were estimated to be 11.8 and 10.8 $/kWh (1990 $) for 25 and 100 MW/ year module production rates, respectively. This meets the 12 $/kWh DOE near-term goal. The energy cost of the Fresnel lens plant (at Carrisa Plains and a 100 MW/year cell production rate) was estimated to be 12.4glkWh and the corresponding central receiver energy cost was estimated to be 13.1 $/kWh, both of which are very close to the DOE goal. Fui-ther design optimization efforts are still warranted and can be expected to reduce plant capital costs. PLANT DESIGNSThe advanced back contact (ABC) concentrator silicon PV cells are based on the EPRI-sponsored work of Dr. R.M. Swanson at Sunpower and Stanford University. Cell peak efficiency of 27.4% is a projection of lot-average efficiencies expected to be attainable in the 1995 timeframe. The physical form of the cells is somewhat different for the two plants, but the cell tec...

Codes are used for pricing ali publications. The code is determined by the number of pages in the publication. Informationpertaining to the pricing codes =_ can be found in the current issue of the following publicationswhich are generallyavailable in most libraries: Energy Research Abstracts (ERA); Government Reports Announcements and Index (GRA and I); Scientificand TechnicalAbstract Reports (STAR)