Data-Driven Sizing of Co-Located Storage for Uncertain Renewable Energy

Abstract

We propose a high-level stochastic steady-state model to analyze the value of co-located energy storage systems for wind power producers that participate in an electricity market through forward contracts and use storage to unlock access to capacity payments. In particular, we try to find optimal storage and contract sizing, as well as stationary operating policies for profit maximization in the long-run. We propose a stylized model calibrated to actual wind power production that allows us to obtain limiting distributions of battery storage levels, assess the value of storage size and perform a sensitivity analysis on key parameters such as contract prices, capacity payments and storage efficiency. We develop the case with contracts of constant price, outline how this model can be extended to a variable-price setting and discuss potential challenges in that avenue.