The Silicon Material Task of the Flat-Plate Solar Array Project was assigned the objective of developing the technology for low-cost processes for producing polysilicon suitable for terrestrial solar-cell applications. The Task program comprised sections for process developments for semiconductor-grade and solar-cell-grade products. To provide information for deciding upon process designs, extensive investigations of the effects of impurities on material properties and the performance of cells were conducted. The silane process of the Union Carbide Corporation was carried through several stages of technical and engineering development; a pilot plant was the culmination of this effort. The work to establish silane fluidized-bed technology for a low-cost process is continuing. The advantages of the use of dichlorosilane in a Siemens-type process were shown by Hemlock Semiconductor Corporation. The development of other processes is described.
The electrical power output of photovoltaic solar cell modules is dependent upon the operating temperature of the cells, and decreases at a rate of approximately 0.5% per /sup 0/C with increasing cell temperature. Because of this temperature sensitivity, it is important to understand the thermal characteristics of modules so that modules and their supporting structures can be designed to reduce cell temperature to the extent that it is cost-effective. Thermal testing of photovoltaic modules is described. The bulk of the testing has been the characterization of twenty-nine modules according to their nominal operating cell temperature (NOCT) and the effect on NOCT of changes in module design, various residential roof mounting configurations, and dirt accumulation. Other tests, often performed parallel with the NOCT measurements, evaluated the improvement in electrical performance by cooling the modules with water and by channeling the waste heat into a phase change material (wax). Electrical degradation resulting from the natural marriage of photovoltaic and solar water heating modules was also demonstrated. Cost effectiveness of each of these techniques are evaluated in light of the LSA cost goal of $0.50 per watt.