Over the past fifteen years, wind power has evolved at an extraordinary pace. The global installed capacity in 1990 was approximately 2.0 GW; today it has risen to nearly 48 GW, with an average annual growth of about 16% for the last 5 years. The bulk of this development has occurred in Europe, with Germany, Spain, and Denmark, respectively, leading the way. Commensurate with this growth, has been the upward trend in the average size of wind turbines (WTs). This trend varies somewhat from country to country, but taking Germany as an example, the average rated power for WTs has gone from about 150 kW in 1990 to about 1600kW in 2003, and the largest machines today have blade spans in excess of 120m, and rated power up to 5MW.
Reducing the cost of wind-generated electricity, to increase competitiveness with conventional power generation technologies, has been the principal reason for these developments. Driving this cost reduction are essentially two factors: scale economies, and learning effects.
Although both of these play a very unique role in the cost reduction trend of wind energy, they are often left undistinguished because to the economist and policy makers that use the cost-reduction trend parameters in their technology forecasts it is often of no relevance what is causing the cost reduction, or rather, how the reduction is distributed across these effects. However, it is conceivable that through the superposition of these factors, possible scale diseconomies could remain concealed. Thus, while the aggregate may suggest an overall cost reduction trend, the full potential of cost reduction could be constrained. There is empirical and theoretical that this is indeed the case.
An analytical means to capture this cost reduction rate is the so called “experience curve”. The experience curve concept is based on the observation that for every doubling of installed capacity the price per unit of energy reduces by a constant proportion. This has been successfully tested in many different technologies. The analysis in this dissertation seeks to break open the technological black box of cost reduction by analytically disaggregating learning and scale effects in experience curves.