Fall 2026

Modeling and Emissions Quantification to Understand Potential Risks from Battery Energy Storage Systems

By Christopher DesAutels, M.S.

Use of a battery energy storage system (BESS) has become increasingly important within a growing and diverse electrical grid.  Green energy systems, such as solar and wind, produce variable levels of electricity, relying on base load power sources (e.g., nuclear or coal-fired power plants) to meet constant demand.  Thus, the ability of a BESS to store and release power during periods of higher demand minimizes the need to maintain separate base load power generation systems.  For performance-critical industry sectors, such as computing and data centers, battery storage is one method to maintain operation during periods of power interruption and minimize on-site generators and electricity production.

Thermal runaway involves an exothermic chain reaction within a battery and the release of gases, which can spread to adjacent battery cells within a BESS.”

The storage of large quantities of power comes with potential risks.  On April 19, 2019, firefighters responded to a call reporting smoke in a remote desert area in Surprise, Arizona.  Batteries within a BESS enclosure had entered thermal runaway and were venting gas that posed potential health risks, as well as the threat of explosion.  A subsequent ignition and explosion occurred that resulted in injury to multiple emergency responders (UL, 2020).  Thermal runaway involves an exothermic chain reaction within a battery and the release of gases, which can spread to adjacent battery cells within a BESS.  Some of the released gases, such as hydrogen fluoride, represent health risks for the community and first responders, while others, such as methane and hydrogen, can ignite and result in explosions and subsequent fires.  The potential for these types of emergency BESS failures has led to the development of procedures for evaluating any resulting risks, such as battery and BESS emissions testing and air dispersion modeling evaluations to estimate the risk of explosion and chemical exposure due to emitted airborne gases.

A BESS consists of multiple racks of battery modules, with each module containing dozens or even hundreds of individual battery cells.  At utility-scale installations, hundreds of BESS enclosures can be arranged in a grid pattern and provide gigawatts of stored power.  Underwriters Laboratories (UL) has developed test procedures, known as UL 9540A, for BESS battery cells, modules, and enclosures to measure gases released during thermal runaway and fire events.  These test procedures are applied to commercial BESS products that are available for use at utility installations.  UL 9540A specifies the method for measuring gas emissions from thermal runaway of an individual battery cell and tests the extent of thermal runaway within a larger module and BESS enclosure.  In addition, UL 9540A measures gas releases from full-scale BESS enclosure fires and evaluates the potential for these fires to spread to adjacent enclosures.

Translation of these measured gas release rates into estimates of airborne concentrations can be done using air dispersion models.  The Electric Power Research Institute (EPRI) has identified a number of air dispersion models that could be used when modeling BESS releases, including SCICHEM, PHAST, and TRACE (EPRI, 2020).  Air dispersion models estimate the air concentrations that result from a given emission rate and set of meteorological conditions.  Additionally, air dispersion models can account for how high a fire-generated plume will rise above ground due to the buoyancy of hot air, where a higher plume will contribute to reduced ground level air concentrations.  A conservative modeling analysis will need to evaluate the impacts of a high-end gas release rate combined with a low-end plume rise, all occurring at a time when meteorological conditions minimize dispersion of the emitted plume.  Ultimately, the modeling analysis can help to establish maximum transport distances for gas concentrations above relevant health standards.

Through the use of BESS emissions testing and air dispersion modeling, the comparative risks from BESS installations differing in their battery technology (e.g., lithium-ion, lead-acid, and solid-state batteries), engineering design, and size can be characterized and planning can be done to develop effective emergency responses.  Modeling identifies the regions of a designed plant that have the highest risk and can provide plant designers and first responders with information for planning a detailed emergency response.  Consequently, while the capacity of worldwide BESS installations has grown exponentially from less than 5 gigawatt hours (GWh) in 2018 to over 600 GWh in 2025, the number of failure incidents (shown in Figure based on EPRI compiled data) has not increased commensurately (EPRI, 2024) – representing a significant decrease in the incident rate per installed unit of power.  Continued development of battery technology and BESS safety design, along with quantification and modeling of potential risks to enable risk planning, will be crucial to supporting a diverse power grid going forward.

Number of Global Grid-scale BESS Failures