Managing the Intermittency of Wind Energy Generation in Greece

Abstract

This paper undertakes an in-depth exploration of a primary strategy aimed at mitigating the volumetric risk in wind power generation within Greece, which originates from the variability of wind speeds. The proposed strategy hinges on portfolio theory, serving as the foundation for creating diverse generation portfolios, thus facilitating the strategic distribution of available capacity across space. Optimization techniques and quadratic programming are harnessed to derive the minimum variance portfolio along with alternative optimal allocation plans. In parallel, Factor Analysis techniques are deployed to discern prevalent sources of variability that exert influence on wind generation dynamics within Greece. Our assessment encompasses an appraisal of the diversification efficacy inherent in an array of distinct portfolios. This analysis is conducted through the application of Principal Component Analysis (PCA) techniques. Furthermore, we expound upon key spatiotemporal attributes characterizing the Greek landscape, elucidating their potential contributions to the site-selection process.

Efficient Risk Management for Distributed Clean Energy: Principal Component based Weather Derivatives

“…The data and sources of information used in this study are described as follows 7 : For solar generation, data from the following nine areas are used for each: 1.Hokkaido; 2.Tohoku; 3.Tokyo; 4.Chubu; 5.Hokuriku; 6.Kansai; 7.Chugoku; 8.Shikoku; 9.Kyushu (see Figure 1). On the other hand, for solar radiation, data for one major city in each area is used (1.Sapporo; 2.Sendai; 3.Tokyo; 4.Nagoya; 5.Toyama; 6.Osaka; 7.Hiroshima; 8.Takamatsu; 9.Fukuoka).…”

“…In general, PCA has been employed in the analysis of distributed renewable generation due to its effectiveness in compressing dimensionality and enhancing the understanding of spatial data features [4]. For example, Burke et al [5] found that PCA is effective in limiting the number of random statistical variables needed to represent distributed wind power; Fonseca Junior et al [6] proposed a highly accurate model using PCA to predict solar power generation at local sites in Japan; Christodoulou et al [7] conducted a PC risk analysis of wind power in the context of site selection for wind power investment. As these studies show, PCA exhibits a natural affinity for the analysis of regionally distributed renewable energy generation and holds significant potential for modeling geographic time series data through dimensionality compression.…”

“…Whether using area derivatives or PC derivatives, if 𝑛 different instruments are listed in the market, we assume that all instruments are used to hedge to maximize the hedging effect 7. Data from the following nine areas are used for each: 1.Hokkaido; 2.Tohoku; 3.Tokyo; 4.Chubu; 5.Hokuriku; 6.Kansai; 7.Chugoku; 8.Shikoku; 9.Kyushu.…”