Balancing safety, regulation, and practicality when working with higher-risk refrigerants.

Over the years I have touched on the subject of risk in this column from several different angles. Now seems like a good time for a deeper dive, particularly in relation to the approach to consumer safety taken by product manufacturers.

I want to make it clear at the outset that this is not about assessing whether an approach is right or wrong, or even whether it is appropriate or inappropriate. I am simply pointing out that there are, as is often the case, alternative viewpoints that are worth considering but that struggle to emerge from behind the screen of “we have always done it this way.”

The topical example is the use of flammable refrigerants in domestic and commercial air-conditioning systems. The present product safety standards approach is that the charge of the flammable substance must be limited to a quantity that could not result in an ignition in the event of a leak. From a risk assessment point of view, it seems that the probability of any sort of leak has been paired with the severity of the worst-case scenario; an explosion in the occupied space would undoubtedly cause severe damage to property and endanger the health and well-being of anybody who happened to be present at the time of ignition. At first sight this is reasonable. Probability and severity are the two elements of risk assessment and the combination of the two indicates whether the level of risk is acceptable or not.

The flaw in this approach is that “any sort of leak” covers a range of possible events with outcomes of various severity levels, but these are not all the same. Lumping them together can lead to false conclusions. For example, most leaks are very gradual, and the refrigerant charge is lost over an extended period of time, perhaps taking several days for the charge to drop enough to be noticeable. In this case, the probability of leakage is high, but the severity of the consequence is low (because the probability of ignition is low). On the other hand, a leak that releases the full charge in a matter of seconds is much less probable but the severity of the consequence is high (if the leaked refrigerant is ignited).

It is entirely reasonable for a product manufacturer to want to set a safety level for their goods to prevent injury or damage. However, in many other aspects of daily life, we are faced regularly with much higher levels of risk, and we seem to be comfortable with them. The chance of being struck by lightning in the U.S. is about 1 in 850,000 per year, and the probability of being killed by the lightning strike is about 10%, so there are around 40 deaths per year, according to the National Institutes of Health. Most members of the public would be happy with an air conditioner that was as risky as being struck by lightning. Road traffic fatalities, on the other hand, account for 1,000 times more deaths than lightning strikes with 44,450 fatalities in 2023, according to the National Safety Council. Would the public accept an air conditioner that was as safe as driving? I suspect the answer is no. However, the present safety standard approach is based on the risk being zero. I can’t think of any other situation in daily life that requires that level of protection.