Enhanced representative days and system states modeling for energy storage investment

The variability of renewable energy sources and the chronological dependency of energy storage makes it necessary to incorporate short-term dynamics in renewable energy availability and de-mand into medium and long-term power system models, such as investment models. However, these short-term dynamics in longer term models can make the models computationally expensive thus methods that can reduce the temporal information but maintain chronology and variety of situations are greatly needed. This paper compares a unit commitment model considering ener-gy storage investment that uses the «System States» (SS) method of representing time to one that uses a «representative periods» (RP) method. It also proposes two new models: the «System States Reduced Frequency Matrix» (SS-RFM) model which treats short-term energy storage more approximately than the SS meth-od to reduce the number of constraints in the problem, and the «Representative Periods with Transition Matrix and Cluster Indices» (RP-TM&CI) model which guarantees some continuity between representative periods, e.g. days, and introduces long-term storage into a model originally designed only for short-term. While both system state models provide an excellent representa-tion of long-term storage, their representation of short-term stor-age is frequently unrealistic. The RP-TM&CI model, on the other hand, succeeds in approximating both short and long-term stor-age, which leads to more accurate storage investment results.