RACA Journal I October 2026 www. refrigerationandaircon. co. za 51
R290 AND FLAMMABILITY
BEYOND FLAMMABILITY: WHAT R290 SAFETY DEBATES REVEAL ABOUT AIR MOVEMENT AND INDOOR RISK technical
Macroventor | Magnific. com
By Eamonn Ryan
Discussions around the transition to R290 in air conditioning and heat pumps are usually framed as a debate about flammability.
Yet beneath that debate lies a fundamental HVAC engineering question: How leaked refrigerant disperses within occupied spaces, how air movement influences that dispersion, and how ventilation affects the resulting level of risk.
At a recent Green Cooling Summit 2026 hosted by GIZ, Daniel Colbourne of Re-Phridge and Zhuolun Chen of UNEP emphasised one technical point: refrigerant charge limits expressed per cubic metre of room volume. At first glance, charge limits appear to be little more than compliance requirements. In reality, they are based on assumptions about how leaked refrigerant disperses through occupied spaces and whether concentrations can approach flammable levels.
Current international safety standards define two key constraints: a maximum charge per system, and a charge per unit room volume. For R290 and other A3 refrigerants, the latter typically becomes the determining factor. The underlying assumption is straightforward: in the event of a leak, the refrigerant will disperse into the room; if the charge is too high relative to the volume( and implicitly, the mixing and air movement within it), concentrations may reach flammable levels.
However, the data presented in the session challenge some of the more simplistic narratives about this process. One set of measurements described involved releasing around 400 grams of R290 from an air conditioner over roughly two minutes. Despite what the presenter called a“ massive leak rate”, the resulting concentrations measured near the floor remained well below the lower flammability limit. The results suggest that under the test conditions, stratification, air mixing and natural or mechanical ventilation limited the build-up of flammable concentrations, even during a relatively rapid release.
The probability of an R290 leak leading to ignition is substantially lower than the risk posed by more common causes such as electrical faults or fuel-gas incidents.
A second analysis explored what happens when refrigerant charge increases, but room size increases proportionally, keeping the grams-per-cubic-metre ratio constant. Counterintuitively, the modelling showed that when room volume increases in proportion to refrigerant charge, the peak concentration following a leak can actually decrease because the larger volume provides greater dilution. From an air quality perspective, this reinforces a point that practitioners in ventilation and indoor environmental quality know well: context matters more than absolute quantities. The same charge that would be problematic in a small, poorly mixed space may be benign in a larger, better-ventilated one.
To provide context, the speakers compared the estimated ignition risk associated with an R290 air-conditioner leak with broader household fire statistics. Their comparison suggested that the probability of an R290 leak leading to ignition is substantially lower than the risk posed by more common causes such as electrical faults or fuel-gas incidents.
RACA Journal I October 2026 www. refrigerationandaircon. co. za 51