As more renewable energy is added to power grids, large battery systems are becoming increasingly important.
They store electricity when there is extra power available and release it when demand is high.
This helps keep the electricity supply stable, reduces energy costs, and makes the power grid more reliable.
However, these large battery systems are expensive, and they gradually wear out as they are charged and discharged over time.
Understanding how and why batteries age is essential for building energy storage systems that last longer and cost less.
Researchers at the U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) have developed a powerful computer model that can predict how large battery systems will age after hundreds of charging cycles.
Their new approach uses high-performance computing to complete detailed battery simulations in just a few days instead of the weeks or even months that older methods often required.
A large battery system is made up of thousands of individual battery cells grouped into modules and battery packs.
Instead of studying only one battery cell at a time, the ORNL model follows how individual cells age and then combines this information to predict the performance of the entire battery system. This gives researchers a much more realistic picture of how large batteries behave in the real world.
The model can simulate more than 10,000 battery cells at the same time and works with several different types of lithium-ion batteries, which are currently the most common batteries used for large-scale energy storage.
The researchers focused on two common jobs that grid batteries perform. One is frequency regulation.
In this role, batteries provide very small bursts of electricity every few seconds to keep the electrical grid operating at a steady frequency. Although this happens very often, each charge and discharge is relatively small.
The second job is reducing electricity costs during periods of high demand. In this case, batteries store large amounts of energy and then release much more power over longer periods. These deeper charging and discharging cycles place greater stress on the batteries.
The simulations showed that batteries used mainly for reducing energy costs aged faster than those used mainly for frequency regulation.
The researchers also found that using batteries for a combination of both services may provide a better balance between short-term financial benefits and long-term battery life.
The computer model also revealed that not every battery cell ages at the same rate. Differences in battery design, electrical connections, and operating conditions can cause some cells to wear out faster than others. The researchers found that lower-voltage battery systems showed greater differences in aging than higher-voltage systems, highlighting how system design can influence battery lifespan.
The research team plans to make the model even more realistic by testing additional battery chemistries, different battery conditions, new system designs, and a wider range of operating temperatures. They also hope to study batteries used in other applications, including energy storage systems that support AI data centers.
By helping engineers understand battery aging much more quickly, this new modeling tool could reduce the need for lengthy and expensive real-world testing. In the future, it may help companies design battery storage systems that last longer, perform better, and lower the cost of delivering reliable electricity to homes and businesses.
