correlate with transportation. On the other hand, solar cell production depends on the manufacturing sector (not strongly fuel dependent), and this would be uncorrelated with transportation. If the subsystem cost components were completely correlated, the resulting distribution would look like the triangular figure in Fig. 4.25. If the components were completely uncorrelated the resulting distribution would look like the normal distribution in that figure. Since the SPS components are partially correlated, the distribution for the entire system starts out as a triangle and probably ends up something like the smooth distribution shown in Fig. 4.24 (dotted line). A distribution like this was obtained for the SPS, with 20% of the area to the left of the modal value and 80% to the right. This assumption was based on the earlier discussion that component projections for advanced technologies err more on the high side than on the low side. The resulting analysis could be performed using other distributions. This assumption is consistent with the capital cost range developed in Sec. 4.2.2. Probabilistic Cost Comparisons Between Coal and the SPS. The main problem in comparing the coal and SPS cost distributions is to establish the size of the overlap between the two technologies. Again, developing a cost distribution for the difference between these technologies is dependent on the degree of correlation of cost elements between coal and the SPS. The correlation was tested and found to be unimportant, because the two technologies stem from such radically different bases. The goal of this probabilistic analysis is to determine the probability that the cost of electricity from the SPS will equal that from coal, and the time at which this might occur. Figure 4.26 shows 1%, 10%, and 20% probability curves for the difference in generating cost between coal and the SPS. For example, the slope of the 20% curve indicates that there is a 20% probability that the SPS will cost the same as coal in the year 2018. Similarly, Fig. 4.25 Distributions of the Sum of Cost Elements
RkJQdWJsaXNoZXIy MTU5NjU0Mg==