Summary: Data-Based Aging Model for Scale BESS

The study addresses the challenge of developing lithium-ion battery aging models for large-scale Battery Energy Storage Systems (BESS), which is hindered by the lack of real-world operational data. The goal is to analyze operational data from a BESS to detect anomalies, characterize stress factors, and replicate real-world cycling conditions in lab tests. These insights will contribute to the development of a data-driven State of Health (SOH) model for predictive maintenance.

Anomaly Detection

The research identifies two types of anomalies:
• Fleet Anomaly: An underperforming battery rack within a bay, indicating local deviations in performance.
• Temporal Anomaly: Stress periods affecting certain racks, potentially due to specific operational conditions.
A deeper analysis of local deviations reveals unexpected behavior in the State of Charge (SoC) trends.

Cycling Protocol and Lab-Based SOH Assessment

To better understand battery degradation, a cycling protocol is developed based on typical BESS operational conditions, including voltage control and reactive balancing. Two cells are tested in the lab under controlled conditions of charge/discharge rate (C-rate) and Depth of Discharge (DOD) at a fixed temperature. Key findings include:
• Higher C-rates accelerate battery degradation.
• Reactive balancing and voltage control contribute to early stress and faster degradation.
• Lower C-rate charging (under specific conditions) results in better capacity retention.

Stress Factor Analysis and Statistical Replication

By tracking the evolution of stress factors over multiple Equivalent Full Cycles (EFCs), the study identifies gradual performance degradation trends. Histograms of stress factor distributions indicate shifts in local distributions, which serve as early indicators of battery performance changes. These findings reinforce the need for adaptive aging models that incorporate temporal factors.

Next Steps

The research aims to develop a data-driven SOH model, integrating real-world BESS data with lab results to enable real-time monitoring and predictive maintenance. By replicating real-world battery cycles in the lab, the study seeks to bridge the gap between operational data and aging model accuracy.

This approach provides a scalable and standardized framework for evaluating BESS aging, improving reliability, and extending battery lifespan through predictive maintenance strategies.