Analysis

Assumptions

The bounding extremes of the geographic distribution of wind installations with regards to resource quality are: (i) a normal random distribution around the mean and (ii) a perfect declining-returns-to-scale curve. The latter conforms to a marginal optimization perspective that the locations with the best resource quality would be utilized first and, once saturated, installations would cascade down to lower quality locations. Put differently, for a given technology and process, CF would be expected to decline reflecting also on other related metrics, e.g. EROEI. In practice, the actual distribution is dependent on multiple factors. On the global scale, countries with varying support for renewables immediately create differentiation potential. Neither Denmark, nor Germany have the best wind resources in the world (see also Fig. \ref{957922}), yet they were the leaders in installed capacity for decades. Other factors at play on a regional scale include: 
Given that the confluence of these factors is difficult to model explicitly, statistical analysis methods are appropriate in analyzing deployment patterns and forecasting the future trends of wind resource quality. In this paper, we consider the on-shore and off-shore wind projects as two separate resource categories, with their own learning and resource potential