A new study accepted for publication in Environmental Science & Technology takes a close look at the amount of certain air pollutants in the Barnett Shale, a booming oil and gas region in North Texas. Using public monitoring data from 2010-2011, researchers from the University of Texas at Austin compared air pollution levels measured at a monitor surrounded by oil and gas operations to the levels that would be expected based on available emission estimates. The result brings to light that the emissions inventory from the Texas Commission on Environmental Quality (TCEQ) for the Barnett Shale does not add up to the observations.
There are numerous air pollutants that can be emitted by oil and natural gas development. Depending on the local composition of the produced gas, emissions can often include volatile organic compounds (VOC, such as propane, butane, pentane, etc.) that contribute to the formation of ground-level ozone (also known as smog), and toxic air pollutants like benzene and hexane that are directly hazardous to human health. Methane, the primary ingredient in natural gas and a greenhouse gas catching lots of attention these days, is another powerful pollutant associated with these operations. Unlike the pollutants listed above, methane directly affects the health of our climate rather than human health. Fortunately, available technologies designed to capture methane are also effective in reducing these other pollutants. However, methane controls alone may not ensure that local air quality concerns are addressed – these require special attention.
The paper, "Atmospheric hydrocarbon emissions and concentrations in the Barnett Shale natural gas production region," was written by UT’s Daniel Zavala-Araiza, Dave Sullivan and David Allen. Environmental Defense Fund sponsored this research; it is unrelated to another study that quantified methane emissions from U.S. natural gas production sites, which was co-sponsored by EDF and led by the same principal investigator, Dr. Allen. For me, this new paper offers two important insights on concerns over air pollution from the oil and gas industry.
TCEQ's emission estimates are too low
First, the best available estimate of VOC emissions from Barnett Shale operations appears to underestimate actual emissions. This finding is consistent with the growing number of projects helping us understand the amount of emissions and sources contributing to ambient air pollution levels.
In their paper, Zavala-Araiza et al. compared hourly air pollution data, captured using monitors in real time, to the levels predicted by their analysis of TCEQ’s Barnett Shale Special Inventory for 2009. They found that actual levels of VOC in the air are higher than the levels predicted by their inventory analysis — by 25 to 100 percent. The difference varies depending on which of the two monitors is used to represent background conditions (i.e., in the absence of natural gas development). The researchers found better agreement by increasing the inventory emissions from pneumatic devices, chemical injection pumps, and equipment leaks using the results from the separate UT production study. This is a significant adjustment, as it increases the total amount of VOC emitted in the region by a third. However, even with this adjustment the observed concentrations were still 60 percent higher than predicted when using one of the two background monitors (there was no bias when using the other background monitor).
Further work is needed to evaluate the completeness of the inventory used by the University of Texas team. Specifically, it is unclear how the team accounted for emissions that TCEQ says are not included in the Barnett Shale Special Inventory: 1) oil and gas sources in the same geographic area that produce from formations other than the Barnett, and 2) emissions from typically larger facilities that did not have to report to the special inventory because they report to a separate "point source" inventory.
A second key finding is that episodic emissions like those occurring from liquid unloadings, when a well is cleared of liquids inhibiting production, do not appear to have a significant influence on hourly air pollution readings; the hourly variability in pollution levels can be largely explained by meteorological effects. This was surprising, as the project initially set out to quantify the influence of such events. But great uncertainty underlies this finding if extrapolated to other areas without empirical data.
More air monitors needed
The Zavala et al. paper shows the kinds of insights that can be gained by investing in air monitoring. A clear recommendation in my mind is that the TCEQ should place more monitors in other major oil and gas producing regions of Texas, including the Eagle Ford. While the state of Texas has developed a commendable network of air monitors in the Barnett Shale, this is not true in other oil and gas plays. Installing monitoring networks early in the development of oil and gas operations would help isolate their air quality impact.