Bridging the Gap Between Battery Economics and Chemistry

Sponsors: This research is supported by University of Washington’s Clean Energy Institute (CEI)

Institutions: University of Washington

Investigators: Prof. Miguel A. Ortega-Vazquez and Prof. Daniel Schwartz

Researcher(s): Mushfiqur R. Sarker and Matthew Murbach

Project Description

Optimizing battery storage has typically been segregated into research focused on the battery’s chemistry and material properties, and research focused on the battery’s integration, operation, and economic performance. This gap is notorious in the research community and in commercial utilization of batteries; especially for power grid applications where the operational decision-making tools use over-simplified models to schedule the operation of batteries. Furthermore, battery degradation receives little attention and is limited simply by enforcing conservative bounds on battery operation. However, by embedding chemistry-dependent characteristics into a decision-making optimization tool, the performance of these storage devices can be maximized attaining additional revenues when providing grid services and while ensuring the degradation is economically justified. This integration of technologies increases the value proposition and ultimately lowers the barriers of using battery storage for grid applications.

Figure 1 shows the incorporation of degradation prediction (a) into an economic decision-making process (b) in order to generate a strategy (c). As this strategy is executed, the current state (d) is updated and the cycle continues. In this exploratory proposal, the framework for connecting (a) and (b) will specifically be explored.

The outcome of this exploration step will be the development of procedures for testing and data-driven analysis of lithium-ion batteries under realistic operating conditions. The research will use a large number of commercially available lithium-ion battery cells, and subject them to different cycling protocols in order to create accurate curves of battery usage against degradation.

The operating characteristics studied will include temperatures, charging/discharging rates, and depth-of-discharge. For instance, a battery may be discharged at its maximum C-rate in order to increase its potential revenue from providing power grid services; however, such usage of the battery has an impact on the battery’s lifetime and is only viable if the revenue is larger than the degradation cost. These proxy curves will be directly embedded into a computationally tractable optimization model. Such an optimization model will schedule the charging/discharging strategy of the batteries, based on explicit tradeoffs between the economics of providing grid services and the associated degradation costs. The major goal in this exploration step is to explicitly characterize battery degradation characteristics in the decision making process, to maximize battery performance at an economically justifiable cost.  This has major immediate applications along with continued research opportunities.

This exploratory step will begin building the underlying framework required to incorporate sophisticated methods to estimating battery state, including a physics-based modeling. By beginning with empirical curves, this project will determine a priority list of factors to consider when incorporating degradation into a decision-making tool for grid scale storage applications.

This project is a joint collaboration between the Electrical Engineering department’s strengths in grid optimization led by Prof. Miguel Ortega-Vazquez and the Chemical Engineering department’s strength in battery health characterization led by Prof. Daniel Schwartz.