Fracking: Chemicals, Cancer, and Relative Risks
The Colorado School of Public Health (CSPH) at the University of Colorado recently announced an article that will be published this month in the journal Science of the Total Environment. The article is based on a study of air pollution resulting from oil and gas development (including hydraulic fracturing or “fracking”) in Garfield County. According to the announcement, the article will reveal findings of benzene and other “potentially toxic petroleum hydrocarbons” at concentrations potentially hazardous to human health. But before this study, or any similar study, can be taken as a basis for alarm, several questions need to be answered: What are these “potentially toxic” chemicals? Where do they come from? And how dangerous are they really?
To answer these questions, it will be helpful to focus on just one chemical: in this case, benzene—the chemical most often associated oil and gas development. Focusing on benzene will also be helpful in evaluating the CSPH study, because, according to the announcement, benzene was the “the major contributor to lifetime excess cancer risk” found in the study.
One reason benzene is so often associated with oil and gas development is that it’s a natural hydrocarbon—like methane, propane, octane, and the hundreds of other chemicals in the mixtures we call “crude oil” and “natural gas.” Consequently, benzene is also found in gasoline, diesel fuel, and engine exhaust, which further increases the presence of benzene near oil and gas development. Lastly, because benzene has desirable chemical properties, it is also separated from crude oil for use in industrial applications, including—among many other things—use as an additive in fracking fluids.
Given the uses above, it should not be surprising that benzene can be found everywhere, not just near oil and gas development. According to the toxicology profile provided by the US Department of Health and Human Services, “Benzene is ubiquitous in the atmosphere.” Not only does it come from tailpipes, but also cigarettes, volcanoes, forest fires, and even camp fires. Government agencies, however, are usually not alarmed by the benzene levels found in our daily lives, because they recognize that the mere presence of a toxin (the fact that a laboratory can physically detect it) does not automatically pose a threat to public health: It is equally important to determine what concentrations can actually do harm.
At what level, then, does benzene become a problem? The truth is, we don’t know. With limited data, the EPA does the best it can to estimate relative risks at various levels of exposure. In the case of benzene, the EPA uses a 25-year-old study (published in 1987), in which workers were exposed to concentrations of benzene measured in parts per million (ppm)—concentrations literally thousands of times higher than the levels the EPA ultimately tries to estimate. The EPA then performs a linear extrapolation (i.e. draws a best-fit line through the data) to estimate a concentration of benzene that will result in 10 additional cancer cases (not to be confused with cancer deaths) per million people exposed. This is essentially the same as determining a level at which the cancer risk for an individual increases by one-thousandth of a percent (0.001%). When considering studies like the CSPH study, it can be more useful to think of the increased cancer risks on the individual level because the exposures in such studies are localized and rarely affect more than a million people—typically they affect a few hundred or less.
For benzene, the EPA estimates that a 0.001% increase in cancer risk corresponds to exposures of 0.4 parts per billion (ppb) in the air and 10 ppb in drinking water. For even greater caution, the EPA set the actual limit on drinking water, enforceable under the Safe Drinking Water Act, at 5 ppb.
While a 0.001% risk may seem small to begin with, there is one critical assumption that needs to be remembered when using these EPA estimates: the calculated risks assume a person will continue to be exposed to the same level of a toxin or carcinogen for their entire lifetime. This is especially unlikely in the case chemicals like benzene, because it is a biodegradable substance—and, in the case of the CSPH study, it is produced by temporary activities.
Also worth considering is the fact that a wide range of uncertainty results from the process used to generate the EPA estimates. Within the range of uncertainty, the EPA selects the most conservative (i.e. the most protective) estimate to establish as the official estimate. Thus, as explained in the EPA calculations, there is an “equal scientific plausibility” that the real levels of benzene corresponding to a 0.001% increase in cancer risk could actually be more than 3-times higher than the current estimates (1.4 ppb in the air and 35 ppb in the water).
These are important considerations when evaluating studies like the CSPH study, since these studies often express their findings in relation these EPA estimates. Without a proper understanding of what these estimates represent, they can give an exaggerated perception of the relative risks involved. Consider, for example, the EPA report that found benzene contamination in Pavillion, Wyoming. In press releases, it was announced that benzene was found at levels 49 times higher than the EPA limit. This, no doubt, caused considerable alarm for the public—but few realized that this represented a 0.02% increase in cancer risk, again, assuming a lifetime of exposure at that level. However, just six months later, the benzene level had fallen 40%. Adjusting for this rate of biodegradation, the total increased cancer risk would have been only 0.0005%. And while this level represents a very low risk, it’s also worth mentioning that this level was found in a deep monitoring well, specifically used to detect contamination—in other words, not a single person was ever actually exposed to this level of benzene. Unfortunately, none of these considerations, are quite as attention-grabbing as the statement: “Benzene found at levels 49-times above Safe Drinking Water Limit!”
Just as it is important to understand what EPA limits represent, it is also important to consider how the increased risks from a particular activity relate to increased risks from air pollution in general. The CSPH study calculated an increased cancer risk of 10 cases per million people living near oil and gas development. But, when compared to the average increased cancer risk nationwide—due to factors such as automobile emissions and industrial activity—the numbers are not quite as alarming. According to the EPA’s most recent National-Scale Air Toxics Assessment, the average increased cancer risk nationwide due to air pollution is 50 cases per million. The risk in Denver is even higher (almost 80 per million) simply because it is an urban environment. Garfield County, on the other hand, has a risk of only 20 per million. Thus, a person living near oil and gas development in Garfield County will experience a cancer risk of roughly 30 per million, far below the national average, and less than half the risk that results from living in an urban environment.
Benzene and other air pollutants should not be ignored when discussing oil and gas development. But it is important for the public to realize that the limits set by the EPA reflect concentrations that present very small—though perhaps not insignificant—risks, and that these risks are comparable to the risks associated with automobile emissions, urban living, and industrial activities in general.
It should also be remembered that, for the purposes of this post, benzene was used as an example because it is one of the most dangerous and most common chemicals associated with oil and gas development; however, the same considerations and relative risks apply the many other chemicals associated with oil and gas development—including xylenes, trimethylbenzenes, aliphatic hydrocarbons, and other compounds that will likely to receive attention in the CSPH study.