UNCERTAINTY IN ENERGY COST COMPARISONS John K. Bragg Consultant to Argonne National Laboratory 100 West 57th Street, New York, New York 10019 The uncertainty in projecting energy system costs into the future is so great that a single number comparison of the life-cycle cost of different systems may be misleading as a basis for decision. However, it is also difficult to put comparisons on a probabilistic basis, particularly when the major cost elements of two competing systems are different in character. The present work illustrates some of the problems in a comparison of SPS and coal technologies for the generation of electricity. The major uncertainties in this comparison are the levelized cost of coal over the remote 2000-2030 A.D. time span, and the capital cost of the SPS system. The cause in the first case is primarily economic, in the second case technological. The methodology reported in this work primarily addresses the first question. Economic Uncertainty Projecting an economic quantity far into the future produces results that depend on hypothesis - "if the following things happen, then the result will be. . ." - without knowing the likelihood of the hypothesized set of circumstances. This problem is inescapable. The present work is not an attempt to resolve the basic problem. It does provide a framework for illustration of problems of uncertainty in energy comparison, in which the structure of the remote future is not arrived at by a conjectural leap, but rather by a year-to-year "random walk" whose parameters are easily recognized in the past. Specifically, the value of an economic quantity in the year n, Qn, is related to Qn-i by a factor fn which is assumed to be a random variable with an appropriate mean 1 + p and standard deviation a, both independent of n. Here p is the appropriate compound growth rate and a expresses year-to-year fluctuations about the mean. Historical values of such quantities may be determined from least-squares logarithmic fits of historical economic series; for example: The mean real growth rates might be called "scenario variables" and it cannot be assumed that they will be the same in the future. However the observed fluctuations may give some guidance to random behaviour in the future. It is possible to formulate the life-cycle cost, or equivalently the levelized cost, of the major elements of electric utility system costs (fixed costs, fuel
RkJQdWJsaXNoZXIy MTU5NjU0Mg==