More Mercury Factoids: Compact Fluorescent Light Bulbs, Part 2

Note: part of a series. Other posts are here and here.

So, why aren’t you going to get mercury poisoning if you break a CFL bulb in your home? First, it’s going to release only a very tiny amount of mercury into the air for a fairly limited period of time (as was discussed in the previous post). Second, we’re comparing that tiny amount of mercury in air with a health-based level that is highly protective - the Reference Concentration (RfC).

From studies of exposed human populations, it’s been observed that the most sensitive effects of low level exposure to elemental mercury occur in the nervous system. The most sensitive neurobehavioral effects were observed at levels in air as low as 25 ug/m3 (that’s microgram per cubic meter of air) with exposure occurring over a period of several years, as confirmed in multiple studies of exposed workers. This lowest-observe-adverse-level (LOAEL) is further reduced with uncertainty factors to protect sensitive individuals and to address where we don’t know much about inhaling elemental mercury (such as effects to the fetus from exposure to pregnant women, or to young children). The value you then get as the RfC is 0.3 ug/m3, which is about 100-fold below the level at which the most sensitive adverse effects have been observed with long-term inhalation exposure to mercury.

In addition, the definition of the RfC further confirms it’s protective nature:

In general, the RfC is an estimate (with uncertainty spanning perhaps an order of magnitude) of a daily inhalation exposure of the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime. (see Note 1, below)

Recall that I had done some calculations of mercury concentrations in air that could result from the emissions from breaking a CFL bulb indoors. The modeled concentrations rose to 0.48 ug/m3 after a day, and drop below the RfC (i.e. below 0.3 ug/m3) after three days, and remain below it all times after that (see Note 2, below). After three days, the modeled concentrations in air continue to decrease. They won’t go to zero, because that’s the nature of a model with an exponential term in it. However, for the folks who believe they need to have zero exposure to be “safe”, the 5 milligram mass of mercury that’s been released will eventually volatilize completely, and at some point in the future, the mercury concentrations in air will become undetectable (which is a little different than “zero”; sorry about that). Things should remain that way until the next time you break a bulb, which should be fairly infrequently (for myself, I try to not break light bulbs because I don’t like taking the risk of stepping in the broken glass).

So what about those couple of days where exposures are over the RfC? There isn’t a published health effects level covering that situation According to the ATSDR as with EPA, the LOAEL for less serious effects (i.e., the most sensitive effects) in humans were neurobehavioral effects such as tremors and poor performance in neurological testing observed at exposures ranging from 14 to 25 ug/m3 over a period more than half a year to 41 years (sorry it can’t be more precise, but that’s the nature of observational data). There’s lot’s of information about neurological effects in humans with high-level short duration inhalation exposure, but no exposure information. However, the levels in air producing adverse neurological effects in lab animals such as rats or rabbits typically are around 1,000 ug/m3 in air or higher for a period of days or weeks. Taking these things into consideration, along with the underlying philosophy that occasional exposures above the RfC don’t imply adverse effects are going to occur, it doesn’t appear that a few days of slightly elevated exposure to mercury vapor from breaking a CFL bulb would produce the shakes. This isn’t really a reason to not buy CFL bulbs. (For more information about mercury-related health effects, here’s the ATSDR public health statement).

As mentioned before, the RfC includes uncertainty factors to address what we don’t know yet about developmental effects to a fetus. This is prudent – in the ATSDR profile, there is one study with laboratory animals in which pregnant rats were exposed to 50 ug/m3 in air for a few hours a day over seven days, where the offspring exhibited neurological impairment. This is different from the other developmental studies in animals, in which short term exposures of ranging from 500 to 1,800 ug/m3 in air were producing neurobehavioral effects in the offspring, which is a thousand-fold higher than the short-term levels potentially associated with breaking a CFL bulb. As before, it doesn’t seem that a few days with slightly elevated exposures constitutes “appreciable risk of deleterious effects”. Not enough to require hiring a cleanup contractor if a bulb breaks in your house, and not enough to deter a reasonable person from buying CFL bulbs.

Note 1: the RfC isn’t a fixed line – above it you’re risking your health, below it you’re safe. It doesn’t work that way. Levels below it are unlikely to be associated with adverse health effects, and are considered protective and of no concern to regulatory agencies. However, as the magnitude and frequency (both are important) of exposure to levels above the RfC increases, the probability of adverse effects occurring increases. However you can’t say categorically that all levels of exposure below the RfC are safe, and that all levels above the RfC are associated with adverse health effects. Welcome to the real world of toxicology.

Note 2: no, I don’t know why the Ellsworth American and Steven Milloy said that the Maine DEP finding was the mercury level was in excess of six times the "EPA standard" in Ms. Bridges home. As typically occurs with the media reporting environmental stories, they didn’t provide enough information. Maybe the inspector made the measurements at carpet-level, while the EPA’s study (and the model) evaluated general room air. Maybe it’s because the inspector was making a measurement with a real-time instrument, and the level represents a short-term peak concentration, while the EPA’s study (and the model) are calculations of time-weighted average concentrations. I won’t speculate about quality control, instrument calibration or operator experience because that’s pointless, but these are also factors that affect the reliability of the measurements.