LISTEN: A new Georgia Tech study found the 2024 BioLab fire plume contained 26 chemical species, with bromine dominating the earliest plume measurements. GPB's Pamela Kirkland reports.
A new study found the chemical plume from the September 2024 BioLab fire in Conyers, Ga., released bromine, not chlorine, as its dominant compound in the immediate aftermath, a finding that stands in stark contrast to early public warnings about the fire.
The fire prompted 17,000 evacuations, closed portions of I-20, and led to overnight shelter-in-place orders for weeks.
Nearly two years later, the U.S. Chemical Safety Board is still investigating the fire and chemical release.
The Georgia Tech paper containing the study was published in the March 2026 issue of Environmental Science & Technology Letters and identified 26 different chemical species in the air following the Sept. 29, 2024, fire at the BioLab facility in Conyers. The authors wrote that the chemically complex plume "exposed millions in metropolitan Atlanta to numerous toxic compounds" and represented the first detailed study of a pool chemical facility fire.
The study showed that bromine concentrations 21 miles away in Atlanta measured four times the federal government's Acute Exposure Guideline Level (AEGL-1), a benchmark used to evaluate short-term exposure to hazardous chemicals. The study also found elevated concentrations of isocyanic acid, known as HNCO, that reached levels 30 times above that threshold.
Public warnings immediately after the fire largely focused on chlorine. BioLab manufactures chlorine and bromine-based pool and spa chemicals. One of the primary products stored at that burned-down building was trichloroisocyanuric acid, or TCCA, a chlorine-based sanitizer.
Both chlorine and bromine are respiratory irritants that can cause eye and throat irritation, coughing, and difficulty breathing, according to federal health and safety data. Each can be dangerous, but researchers know less about the effects of bromine, making the Georgia Tech study a vital addition to the ongoing BioLab investigation.
In an exclusive interview with GPB, study author Greg Huey, a professor in the School of Earth and Atmospheric Sciences at Georgia Tech, said his team expected more chlorine when they began measuring the plume from campus in Midtown Atlanta the next day.
"We expected to see chlorine, maybe a little bit of bromine," Huey said. "And the largest signals we were clearly seeing in our mass spectrometer were not from chlorine. They were coming from bromine and also other organic compounds."
While bromine concentrations in Atlanta were high, Huey says levels closer to the source in Conyers were likely much higher.
"We’re 21 miles away," Huey said. "You would expect things to be a hundred times the concentration around, say for example, Conyers City Hall, than they were here."
While the paper was not published until March 2026, the research began almost immediately. Huey said the city of Conyers contacted his team within days of the fire and helped deploy a Georgia Tech mobile laboratory near the site, allowing researchers to begin collecting more data from closer to the site of the BioLab fire and resulting plume within 48 hours. GPB first spoke with Huey about the initial findings in December 2024 and again in March 2025.
The findings raise questions about the gap between the compounds that were tested for and the chemicals of concern. The Georgia Tech researchers describe the study as the first comprehensive chemical characterization of a pool chemical facility fire and say it revealed "complexity far beyond that emphasized in public health communications."
Documents obtained by GPB show CTEH, a contractor hired by BioLab, identified bromine among the chemicals that could be released during the fire, alongside chlorine and hydrochloric acid. However, bromine did not appear among the chemicals listed in public summaries of U.S. Environmental Protection Agency monitoring.
The initial chemicals of concern that the EPA were monitoring for were chlorine, hydrochloric acid and phosgene, according to early readouts after the fire. Last October, the EPA told GPB via email its early plume modeling was based on information provided by BioLab.
Huey was careful to note that his Georgia Tech team's work was designed to answer scientific questions about what was in the plume, not to assess regulatory compliance or public health outcomes. The instruments used could detect only certain compounds, meaning other substances may have been present but went undetected.
Huey said in a situation like this, where large quantities of chemicals like oxidizers begin reacting, the results can form compounds that are difficult to identify and study.
"You have the potential to make a lot of compounds that are very hard to know what they are, know what the toxicity is, know what the health impacts are," Huey said. "So, I think you introduce a much higher level of uncertainty with chemicals of this nature."
The study also found the plume evolved over the more than the two weeks it impacted the metro Atlanta region. While early measurements were dominated by bromine, chlorine became more prominent weeks later.
"We were still seeing significant levels of chlorine in the parking lot at Conyers," Huey said. "Chlorine and bromine ... are very short-lived during the day. So we were surprised to still be seeing this much coming out after several weeks."
The research was supported by a rapid-response grant from the National Science Foundation in December 2024 which allowed the team to spend more than a year analyzing data from the plume. Huey’s team, including graduate researchers Christine Harper and Katherine Paredero, began collecting data on Sept. 30, 2024, and spent more than a year analyzing the data before publishing the findings in March.
In May 2025, BioLab said it would no longer manufacture at the Conyers site but it would continue to be used as a distribution hub.
Asked about the study, EPD spokesperson Sara Lips said the agency was not aware the paper had been published and that it had not yet been reviewed by agency staff. The EPA did not provide comment before this story was published. BioLab did not respond to a request for comment. Huey said none of the entities reached out to him or his team about the research.
"I don't know of any other chemical incident like that where you happen to have an atmospheric chemist 20 miles downwind with a really nice mass spectrometer to say, 'What's really in this plume?'" Huey said. "So it was a unique opportunity."
For more on the 2024 BioLab fire and GPB’s investigation, listen to Manufacturing Danger: The BioLab Story wherever you get your podcasts.
