Estimating Cost Structures in the U.S. Nuclear Power Industry

Abstract

With growing concerns about global warming and surging oil prices, nuclear power is now back on the U.S. energy policy agenda. This paper provides firm level analysis of the production technology and cost structures in the U.S. nuclear power generation industry. The paper applies an econometric approach into a dual restricted variable cost function within a "temporal equilibrium" framework. A panel data set of 32 nuclear power generations for major U.S. investor owned utilities over the period 1986-2002 is used. The major finding indicates that most electric utilities in the nuclear electricity generation industry over utilized capital in production. The estimated results show evidence of scale economies in the production of the electricity industry. The paper presents supporting evidence that nuclear technology is cheaper than other alternative technologies and fuels, and it can offer possible cost savings to other alternative technologies.

On the performance of the United States nuclear power sector: A Bayesian approach

“…Further, traditional Bayesian SFA produces estimated RTS much smaller than either of the BETEL models, with a modal RTS of 0.58. Our BETEL estimates (both I and II) align closely with those of Rungsuriyawiboon (2008), who found a short run RTS of 0.895 and a long run RTS of 0.776, albeit for an earlier time period than we are analyzing here. 16 It is clear that the vast difference in modal/average RTS between our BETEL estimates (both BETEL I and II) and standard Bayesian estimation suggest that, between the allowed asymmetry of the noise term and the robust estimation that the new approach we adopt, has important implications for what is observed regarding optimal plant size.…”

“…Thus, a finding of very little average technical change is not surprising. The finding of mainly decreasing returns to scale is in line with the work of Rungsuriyawiboon (2008), albeit for a different time period. Furthermore, our new approach delivers estimates that are quite different than a standard Bayesian setting, suggesting that the BETEL estimation approach is indeed imparting its impact on the empirical findings.…”

“…It is difficult to compare these estimated input elasticities to the extant literature since so few papers have focused exclusively on nuclear power generation. Perhaps the closest is Rungsuriyawiboon (2008). Rungsuriyawiboon (2008) estimates price and output elasticities (found through a cost function) with the results suggesting that fuel, labor and capital had statistically different price elasticities (with capital being the most price responsive in the long run) and fuel having the largest output elasticity.…”

“…Perhaps the closest is Rungsuriyawiboon (2008). Rungsuriyawiboon (2008) estimates price and output elasticities (found through a cost function) with the results suggesting that fuel, labor and capital had statistically different price elasticities (with capital being the most price responsive in the long run) and fuel having the largest output elasticity. This last result of Rungsuriyawiboon ( 2008) trues with our finding here that fuel appears to have the most sensitivity.…”

“…Lastly, we note that a full accounting of the productive aspects of the nuclear power generation sector in the US have not been extensively studied. To our knowledge, only Krautmann & Solow (1988) and Rungsuriyawiboon (2008) have focused on technical change and scale economies. However, neither of these papers has modeled if the plants operate efficiently.…”