cess/failure mapping, where and Sm(^ S) are the sample distribution functions corresponding to F^S) and F(^ S) for n successes and m failures. Of the 700 trials, 374 were successes and 326 failures. Only three of the eighteen parameters appear to have any appreciable influence on success or failure, silicon energy intensity (dm^n = 0.653), solar cell thickness (dm_n =0.390), and cell efficiency ^m,n = 0.136) with silicon energy intensity being most important by a large margin. In fact, the importance of cell thickness, which ranks second in our analysis, is due primarily to the linkage with silicon energy intensity; as might be expected, these two parameters showed a high negative correlation (-0.41) for successful outcomes, i.e., a low energy intensity allows thicker cells and vice versa. Separation of the parameter distributions for all other parameters was below the 90% level of significance when testing the hypothesis that Ff^S) = F(F S). The importance of silicon energy intensity to the ER distribution was noted by Herendeen et al. and comes as no surprise. However, our analysis indicates that the uncertainty in this parameter is sufficiently large to dominate their analysis and limit the conclusion that can be drawn from it as to the promise of the SPS system. To the extent that the energy ratio is a good measure of future promise, it appears that a critical research issue is to obtain refined estimates of the energy costs and technology associated with the manufacture of the solar cells. Cirillo et al. (5) have arrived at a similar conclusion based on their independent and extensive recent analysis of the net energy balance of SPS systems. Woodcock (6) has estimated the current energy requirement of solar cells at about 25% below the mean estimate of Herendeen et al., with the potential for very substantial reductions as continuous manufacturing processes are developed. If Woodcock’s estimates turn out to be accurate, the SPS will be a much more attractive option for the future than the analysis of Herendeen et al. would seem to imply. Acknowledgments — The Editor wishes to thank R. A. Herendeen and Gordon Woodcock for reviewing this paper. REFERENCES 1. J. Grey, Solar Power Satellite: A Plea for Rationality, Science 203, 1979. 2. R. Herendeen, T. Kary, and J. Rebitzer, Energy Analysis of the Solar Power Satellite, Science 205, 451, 1979. 3. R. Herendeen, T. Kary, and J. Rebitzer, Energy Analysis of the Solar Power Satellite, ERG Doc. No. 265, Office of the Vice Chancellor for Research, Univ, of Illinois, Urbana, 1978. 4. R.C. Spear and G.M. Hornberger, Eutrophication in Peel Inlet: II. Identification of Critical Uncertainties Via Generalized Sensitivity Analysis, Water Research 14, 43, 1980. 5. R.R. Cirillo, B.S. Cho, M.R. Monarch, and E.P. Levine, Comparative Analysis of Net Energy Balance of Satellite Power Systems (SPS) and Other Energy Systems, DOE/ER-0056. U.S. Dept, of Energy, Washington, 1980. 6. G.R. Woodcock. SPS Cost Considerations, J. Energy 2, 196, 1978.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==