When trichloroisocyanuric acid burns, it does not simply release chlorine. The molecule breaks apart under fire conditions into 26 distinct chemical species, including bromine at concentrations four times the federal acute exposure threshold and isocyanic acid at levels 30 times that guideline, both measured 21 miles from the source [3]. This combustion profile was unknown until September 29, 2024, when a fire at the BioLab chemical facility in Conyers, Georgia, forced 17,000 evacuations and closed portions of Interstate 20 [3]. The March 2026 study published in Environmental Science & Technology Letters represents the first detailed characterization of what happens when a pool chemical warehouse burns [3].
What the compound became
The primary material stored in the burned building was trichloroisocyanuric acid, a chlorine-based pool sanitizer [3]. Georgia Tech researchers deployed a mobile laboratory within 48 hours of the fire and identified 26 different chemical species in the plume [3][4]. Bromine, not chlorine, dominated the immediate aftermath [3]. The finding contradicts the chemical identity suggested by the stored product's label and the initial public warnings, which focused largely on chlorine exposure [3].
Isocyanic acid concentrations reached 30 times above the Acute Exposure Guideline Level (AEGL-1) threshold in Atlanta, more than 20 miles from the facility [3]. Bromine measured four times AEGL-1 at the same distance [3]. Both compounds are respiratory irritants [3]. The measurements came from air samples collected near the fire site starting within two days of ignition, after the city of Conyers contacted Georgia Tech directly [4].
The study marks the first detailed examination of a pool chemical facility fire [3]. No prior combustion model existed for TCCA breakdown products. The chemical profile that exposed millions in metropolitan Atlanta was uncharacterized until the fire itself provided the data [3].
What authorities monitored for
The Environmental Protection Agency's initial monitoring list included chlorine, hydrochloric acid, and phosgene [3]. Bromine did not appear in public EPA monitoring summaries [3]. CTEH, a contractor hired by BioLab, identified bromine among chemicals that could be released during the fire [3], but that identification remained in private documentation rather than public health guidance.
The gap between what burned and what officials warned about persisted for weeks. Overnight shelter-in-place orders continued as the dominant compound in the plume, bromine, went unmentioned in federal public summaries [3][4]. Georgia Tech's mobile laboratory captured the chemical profile that EPA's monitoring framework missed.
The U.S. Chemical Safety Board is still investigating the fire and chemical release nearly two years later [3]. One year after the event, questions about resident exposure and long-term health effects remained unanswered [3].
The deployment gap
Conyers city government initiated contact with Georgia Tech within days of the fire [4]. The university's researchers began collecting data from close to the BioLab site within 48 hours [4]. That timeline contrasts with the federal investigation's pace: the Chemical Safety Board's inquiry, which carries statutory authority for industrial accident investigation, had not concluded 21 months after ignition [3].
The institutional speed difference is structural, not incidental. A city government calling a university research team produced the only comprehensive chemical characterization of the plume. Federal monitoring infrastructure, designed around anticipated compounds based on product labels and standard combustion assumptions, did not capture the dominant toxin.
BioLab manufactures chlorine and bromine-based pool and spa chemicals [3]. The facility halted manufacturing in Conyers following the fire [3]. Former employees have described ignored warnings about fire risk, and whistleblowers raised questions about the disaster response [3], though those accounts remain under investigation.
Storage without combustion profiles
The regulatory framework that permitted industrial-scale TCCA storage in residential proximity operated without combustion chemistry data. The study now provides that profile: 26 chemical species, bromine-dominant plume, isocyanic acid at 30 times acute thresholds [3]. The information arrived after 17,000 people evacuated and weeks of shelter orders, not before [3].
The question the study raises is whether storage and proximity decisions can be made without knowing what stored compounds become under fire conditions. The BioLab fire functioned as the experiment that generated the first dataset. Whether that sequence, storage, fire, characterization, represents acceptable regulatory practice or a gap that should close before the next facility is sited remains an open structural question. The combustion profile exists now. It did not exist when the facility was approved or when the chemicals were stored next to a residential population. The study documents what was released; it does not answer why that information was not required beforehand.
Regulatory structures now have the combustion data for TCCA. The timeline that produced it, ignition first, characterization second, community exposure throughout, defines the current standard for chemical storage oversight in proximate residential areas.
