Tight power and energy coupling constraints of energy storage resources for unit commitment

Energy Storage Resources (ESRs) can help promote high penetrations of renewable generation and shift the peak load. However, the increasing number of ESRs and their features different from conventional generators bring computational challenges to operations of wholesale electricity markets. In order to improve the computational efficiency, this paper tightens the generic ESR formulation for unit commitment. To avoid the complexity caused by ESR operations in both discharge and charge directions, a novel decoupled analysis is conducted to analyze one direction at a time. For each direction, ESRs over two and three time periods are categorized into several types based on their parameters. For each type, our recent four-step systematic formulation tightening approach is used to construct the corresponding tight formulation. In order to consider more periods without analyzing all the drastically increased number of types, a series of major types are selected based on how many periods an ESR is able to discharge (charge) consecutively at the upper power limit. A related generic form of tight constraints over multiple periods is established. Moreover, validity and facet-defining proofs of our tight constraints have been provided. Numerical testing results illustrate the tightening process and demonstrate computational benefits of the tightened formulations.

Automated summary

Energy Storage Resources (ESRs) can promote renewable energy penetration and peak load shifting, but their increasing number and different features from conventional generators pose computational challenges in wholesale electricity market operations. To improve computational efficiency, this study tightens the generic ESR formulation for unit commitment through decoupled analysis and type-based formulation tightening. The effectiveness and validity of the tightened constraints are demonstrated through numerical testing, showcasing the computational benefits.