Separating Economies of Scale and Learning Effects in Technology Cost Improvements

Healey, S. (2015). Separating Economies of Scale and Learning Effects in Technology Cost Improvements. IIASA Interim Report. IIASA, Laxenburg, Austria: IR-15-009

Preview

Text IR_15_009_Healey.pdf - Published Version Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

A contributing factor to the range of results of energy-economy models on feasibility and costs for obtaining specified policy goals such as GHG reductions is the mechanism that translates technological progress into cost reductions per technology. In many models this mechanism is represented via a simple learning curve model, where costs decline as a function of experience, usually represented by cumulative capacity. However, many other factors influence technology-specific cost dynamics, with one major confounding variable being unit level economies of scale- declines in average unit costs arising from the building of larger capacity plants or units. Thus, this paper considers ‘de-scaled’ learning rates as an alternative to the conventional representation of learning phenomena. De-scaling involves removing unit scale’s influence on cost for a given technology, thereby creating a variable that is the residual of cost’s remaining determining factors, which then is estimated and interpreted as a more appropriate variable to capture learning effects proper. The influence of scale economies and remaining learning effects on technology costs are estimated econometrically with the analysis complemented by simpler analytical methods as well as incorporating results from the engineering literature to represent uncertainties. This paper finds de-scaling substantially reduces the learning rate for a number of energy supply technologies. De-scaled learning rates expressed over cumulative units installed is concluded to be a superior measure of learning effects over traditional formulations of specific costs versus cumulative capacity that confound economies of scale with learning effects, potentially misguiding policies via its resulting overestimation of the potential cost lowering impacts of demand-pull technology deployment incentives.

Actions (login required)

View Item