Optimal sizing and location of energy storage systems for transmission grids connected to wind farms

Abstract

Although modern renewable power sources such as solar and wind are increasing their share of the world’s power generation, they need to grow faster to replace a greater share of coal and gas power generation and thus, help prevent CO${}_2$ and other greenhouse gas emissions to reach critical levels. Renewable energy generation must be coupled with energy storage systems, which are unfortunately expensive investments. However, substantial cost savings may be possible if a system-wide solution is sought. This paper presents such an attempt for a transmission grid that has a mixture of renewable and non-renewable sources. The particular problem is to find the type, location and size of the storage systems in the grid, as well as the structure of the transmission network, to minimize total investment and system-wide operating costs of power generation, transmission and storage. A mixed integer linear programming formulation is devised for the problem, which can be very large because various operational decisions are made at short intervals. Hence, we develop a “divide-and-conquer” type solution approach based on time decomposition, wherein the problem is first solved in monthly time segments. Subsequently, optimal or near-optimal monthly generation schedules are merged to construct the greater portion of a grand schedule for the whole year. Although still considerably large, the model can be solved effectively after another set of heuristically developed restrictions on the transmission network structure. The formulation and solution method are implemented on a series of realistic instances for a modest-sized transmission grid adapted from Sardinia Island of Italy to demonstrate the effectiveness of the approach and the insight into related design decisions.