Improvements to Modern Portfolio Theory based models applied to electricity systems
针对现代投资组合理论在电力系统规划中的不足，提出三项改进：以发电成本替代装机容量、考虑需求与发电的相关性、引入CVaR限制短缺风险，并在巴西案例中发现改进后的有效前沿显著不同，传统组合的欠产风险更高。
With the increase of the share of variable renewable energies in electricity systems, many studies were developed in order to determine their optimal technological and spatial mix. Modern Portfolio Theory (MPT) has been frequently applied in this context. However, some crucial aspects, important in energy planning, are not addressed by these analyses. We, therefore, propose several improvements and evaluate how each change in formulation impacts results. More specifically, we use generation cost instead of installed capacity as one of the objectives; we consider the correlation between demand and generation profiles; and we limit shortage risks via the inclusion of a CVaR measure. These modifications are presented in a formal model which is also applied to the case of Brazil. We found that, after including our proposed modifications, the resulting efficient frontier differs strongly from the one obtained in the original formulation. The main difference is that the new efficient frontier has a much shorter range of acceptable standard deviation values. Therefore, many of the portfolios obtained from the traditional formulation have a much higher probability of under-production, especially portfolios located at regions with standard deviation either too low or too high. Furthermore, we show that diversification plays an important role in smoothing output from portfolios of variable renewable sources.