What Does a Negawatt Really Cost? Evidence from Utility Conservation Programs *

Abstract

We use information reported by ten utilities about their electricity conservation programs to calculate the life-cycle cost per kWh saved - the cost of a "negawatt " -- associated with these programs. These computations indicate that the cost associated with utilities "purchasing" negawatthours is substantially higher than implied by standard sources such as Amory Lovins (Rocky Mountain Institute) and EPRI. The costs calculated for residential programs, in particular, are much higher than conservation advocates have suggested. However, 80% of the expected savings from these programs are attributed to commercial and industrial customers rather than residential customers. We find substantial variation in costs between utilities for similar programs as well as significant intra-utility variation in the cost associated with various sub-programs. We proceed to examine whether or not there are any systematic biases in the reporting of costs and energy savings by the utilities in our sample. In many cases, utilities fail to report all relevant costs, rely on engineering projections of savings rather than applying methods to measure savings based on actual experience, and fail to make appropriate adjustments for free riders. Further biases may result firorn adopting measure lives that are too long. As a result, on average the cost of a negawatthour computed from utility reports significantly underestimates the true societal cost of conservation achieved this way. Mile it is difficult to compute the underestimate with any precision, the evidence that we have suggests that computations based on utility expectations could be underestimating the actual societal cost by a factor of two or more on average. Better utility cost accounting procedures and the application of more sophisticated methods to estimate actual energy savings achieved are clearly necessary before large sums of money can be expended wisely on these programs.