Measurement of technology progress and capital cost for nuclear, coal-fired, and gas-fired power plants using the learning curve

Abstract

This paper treats the problem of measuring aggregate technology progress and capital cost of gas-fired, coal-fired, and nuclear power plants using the classical learning curve. Regression analysis on disclosed information demonstrates the strength of the technique. Composite learning performance for gas-fired power plants ranges between 85.0% and 88.8% and for coal between 91.7% and 92.8% for the plateau region of the United States. Composite learning performance for nuclear power plants ranges between 78.3% and 81.3% for the United States. Changes in the learning progress for coal-fired and nuclear power plants are revealed beginning in 1973. Technique and interpretation of the data are provided. Adoption of the method as an aggregate predictor of technology progress and cost is encouraged.