In recent years, prospects for the future deployment of battery energy storage resources in the electric sector have increased markedly, as battery costs have declined, and their technical performance has continued to improve. However, battery energy storage technologies have complex cost, value, and performance characteristics that make them challenging to model in long-term power system capacity expansion models.
Phase 1 of this project explored the potential value of incorporating five features that are not commonly represented in existing long-term capacity planning models: (i) battery energy storage system (BESS) degradation; (ii) grid (network) modeling; (iii) ancillary services; (iv) sub-hourly temporal resolution; and, (v) uncertainty. Simulation results identified that BESS degradation is the feature with the largest impact on planning outputs in models with energy storage.
Phase 2 of this study, which is described in this report, takes the next logical step and focuses on experimenting with different degradation models for Lithium-ion battery energy storage systems in a capacity planning model. The objective of Phase 2 is to answer the question: What is the most economically efficient (that is results in the lowest overall system costs when degradation is taken into account) and computationally efficient approach to battery degradation that can be incorporated into long-term power system resource planning models? To answer this question, five methods to make a capacity planning optimization problem “degradation-aware” were developed and simulation experiments were completed that incorporated one method at a time. These simulations were done using a “maquette” version of a large power system located in the Southeastern United States.
Experimental results show that selecting a proper degradation approach can materially reduce total system costs relative to a result that assumed a fixed battery lifetime. Models that keep the battery’s state of charge at lower levels can extend the battery’s operational life and improve economic efficiency (that is the added benefit of extending the battery’s life exceeds the added cost of controlling the
battery’s use pattern). Other models that limited the total amount of energy discharged from the battery or simulated energy capacity fade are not as economically efficient. Many of these degradation simulation models can be solved in a computationally efficient manner and do not add a significant computational burden to existing battery optimization models.
Capacity expansion models used by electric companies to develop future long-term resource plans continue to evolve to be able to better assess the potential value and future deployment of energy storage and other emerging power system technologies. One challenge in developing these new and improved models is that doing so often requires tradeoffs between economic efficiency and computational complexity. The simulation results in this analysis pave the way for incorporating “degradation-aware” yet computationally light models for battery storage systems in power system capacity planning problems.
Authors Adam Diamant and Miles Evans