Kernel density estimation based time-dependent approach for analyzing the impact of increasing renewable on generation system adequacy

Abstract

Integration of non-conventional renewables such as wind and solar to the power system may affect the system reliability, especially when the proportion of renewable power in the system is large. Therefore, with a significant level of renewable penetration, the intermittency and both diurnal and seasonal variations of renewable power generation should be deliberately modeled in order to accurately quantify the power system reliability. This paper presents a novel method based on Kernel Density Estimation (KDE) for modeling intermittency and both diurnal and seasonal variations of wind and solar power generation using historical renewable power generation data. The proposed KDE based renewable power models are used with non-sequential Monte Carlo simulation to evaluate the generation system adequacy. Several case studies are conducted on IEEE reliability test system to analyze the impact of increasing renewables on the generation system adequacy. The results show that the generation system adequacy tends to decay exponentially when the renewable integration is increased. It is shown that the reliability values obtained using the proposed approach are very close to those provided by the time-consuming sequential simulations. Importance of modeling seasonal variations of wind and solar is also investigated.