Modeling framework reveals grid battery aging effects

Energy storage is becoming critical to grid resilience and electricity affordability because battery systems can help balance supply and demand and stabilize power.

Long battery lifetimes are key to unlocking those benefits. But that requires understanding how project design and operation affect stress on large-scale battery systems over hundreds of cycles - before spending the time and money to deploy them.

Researchers at the Department of Energy's Oak Ridge National Laboratory are tackling this problem by coupling high performance computing (HPC) with a physics-based modeling framework to rapidly evaluate how different operating strategies affect battery aging over time. The result is a powerful computational tool for designing megawatt-hour battery storage systems so they will last longer and reduce costs.

A full battery pack - what most people think of as a battery - is made up of modules filled with many battery cells. The ORNL modeling framework starts with cell-level aging simulation, then scales up to module- and pack-level performance. This multiscale approach enables optimization of electrical architecture, control strategies and operating schedules.

HPC speeds insight into battery wear

ORNL's approach relies on high performance computing for detailed analysis of battery degradation after 500-1,000 operating cycles, delivering results in days instead of weeks, said Srikanth Allu, ORNL computational research scientist. The flexible, reusable framework can simulate more than 10,000 cells at once and accommodate multiple chemistries of lithium-ion batteries, the technology most widely used for grid-scale storage.

Previous approaches for predicting degradation in lithium-ion systems often relied on cell-level studies or simplified assumptions. These did not fully reflect how battery aging is influenced by system-level designs and real-world operating conditions driven by different uses of batteries in the grid.

The ORNL model simulated two common grid energy storage services: reducing energy costs and stabilizing the frequency of changes in electric current flow.

To regulate frequency, batteries receive signals from operators to provide rapid, short bursts of power every few seconds. Momentary swings in supply or demand can destabilize frequency, but batteries quickly provide balancing. This service involves frequent, shallow cycling.

In contrast, when used to reduce energy costs during high-demand periods, batteries draw on stored energy with deeper, more energy-intensive cycles. Researchers found that this operating profile wears down the battery faster.

Battery systems can be used for a single grid service or a combination. ORNL simulations showed that using them for multiple services can balance near-term savings with long-term battery lifetime and replacement costs.

"We found that these two applications drive different degradation pathways within the battery, highlighting how operating conditions influence aging at the material level," said Allu. "The simulations provide new insight into why batteries may degrade at different rates and point to opportunities to design battery systems that better balance performance, lifetime, and economic return."

Simulations show what's outside the battery matters too

The model also showed that aging varies among battery cells and across entire battery packs. Because of both inherent design differences and varied operation for different tasks, low-voltage systems showed more aging variation than high-voltage systems. The results highlight the importance of electrical architecture and interconnections in determining how degradation evolves across large-scale battery systems.

Allu said the next step in the research would be to extend the framework to additional battery formulations and more realistic operating conditions, including varied designs, different states of health and a wider range of temperatures. In the future, the approach could also be adapted to study battery performance across a broader range of grid and energy storage applications, such as battery support for AI data centers.

"Simulations at this scale can significantly reduce the need for costly, time-intensive, full-system testing to evaluate battery aging," Allu said.

The ORNL team used the data from this research as the basis for a foundational model that further speeds up aging analysis to hours instead of weeks or months. This effort supports the DOE's Genesis Mission, a national initiative to build the world's most powerful scientific platform to accelerate discovery science, strengthen national security, and drive energy innovation.

Other researchers who contributed to the project include Michael Starke, head of ORNL's Electrical Systems Integration Section, and former postdoctoral researcher Surya Mitra Ayalasomayajula. The team received a Best Paper Award at the Institute of Electrical and Electronics Engineers (IEEE) Electrical Energy Storage Application and Technologies Conference earlier this year for their work describing the project. The research was funded by DOE's Office of Electricity.

By combining leadership-class high performance computing with expertise in battery materials and grid systems, ORNL can model battery behavior from individual cells to full systems, at a range of scale and detail not typically achievable elsewhere.

UT-Battelle manages ORNL for the Department of Energy's Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science. - S. Heather Duncan