Optimal solar panel placement in microgrids

Abstract

Microgrids offer a flexible, modular, and scalable solution to facilitate the adoption of renewable energy sources, e.g., solar power, in distribution networks. Noticeably, the cost of daily operation of a microgrid is tightly coupled with its long-term planning decisions, such as the placement of renewable energy sources. This paper studies the optimal placement of solar panels in a grid-connected microgrid to minimize the energy cost during daily operation. Due to the coupling between long-term planning (i.e., solar panel placement involving integer variables) and short-term operation (i.e., supply-demand balance through optimal power flow (OPF)), the problem is naturally formulated as a two-time-scale stochastic integer programming problem. To solve the two-time-scale problem efficiently, a key contribution of this paper is to derive tight analytical bounds on the optimal value of the short-term OPF problem. The analytical expressions are useful in constructing a closed-form objective function for the long-term planning problem, which can then be efficiently solved using existing integer optimization approaches. Through numerical simulations on the IEEE 13-bus distribution test feeder, we demonstrate significant cost saving due to the optimal solar panel placement, and the accuracy and efficiency of our proposed approach.