Consideration of Extreme Wind Geographical Correlation Scenarios in Reliability Assessment Studies Using Sequential Monte Carlo Simulations

Abstract

Electrical generation based on the use of renewable energies is emerging in modern grids. In that way, one of the most popular solutions as well in transmission as in distribution grids is certainly coming from wind energy. However, wind resources on a given location randomly fluctuate with time and have thus a major impact on the capacity of the electrical system to continuously face the load. In order to evaluate this impact and to consequently adapt required reinforcements, Monte Carlo simulations are often used. Those approach can be either sequential or not. Nowadays, load shifting solutions (storage, demand side management…) are practically set in order to adapt consumption to time varying generation without involving too consequent investments. In that way, sequential approach is currently preferred when it comes to long-term planning evaluation and adapted time series models are developed to characterize wind generation on a given site. The consideration of the geographical correlation between those models has been recently investigated in some references. This paper proposes to complete those contributions by evaluating the impact of wind geographical correlation on classical reliability indices such as the Loss of Load Expectation (LOLE) or the Expected Energy not Served (EENS).