Chance-constrained optimal capacity design for a renewable-only islanded microgrid

Microgrids offer a promising opportunity for achieving greater use of renewable generation. In this paper, we consider optimal capacity design for an islanded microgrid supplied by a wind turbine, solar panel and battery system. The objective is to reduce plant cost while ensuring energy sufficiency, taking into account stochasticity of renewable generation and load. An affine control policy is designed to dispatch battery power under uncertain renewable in-feed and load. The policy is integrated into a stochastic chance-constrained optimization problem, which is solved using a probabilistically robust method. In order to address conservativeness inherent in the robust method, we develop two approaches to set reshaping that reduce the volume of the robust set, thereby enabling less conservative designs.