On the Selection of Intermediate Length Representative Periods for Capacity Expansion
As the decarbonization of power systems accelerates, there has been increasing interest in capacity expansion models for their role in guiding this transition. Representative period selection is an important component of capacity expansion modeling, enabling computational tractability of optimization while ensuring fidelity between the representative periods and the full year. However, little attention has been devoted to selecting representative periods longer than a single day. This prevents the capacity expansion model from directly simulating interday energy sharing, which is of key importance as energy generation becomes more variable and storage more important. To this end, we propose a novel method for selecting representative periods of any length. The method is validated using a capacity expansion model and production cost model based on California's decarbonization goals. We demonstrate that the representative period length has a substantial impact in the results of the capacity expansion investment plan.