Last week, researchers at University of Texas – Arlington (UTA) published a study claiming to have found abnormalities in well water samples taken from the Eagle Ford shale region in South Texas. While the study authors suggest these abnormalities are indicative of possible contamination from unconventional oil and gas development, the researchers fail to make such a connection.

In fact, the study directly states that the main abnormality in water quality, the ratio of bromide and chloride concentrations, is “unlikely to be attributed to contamination form UD [unconventional development] activities.”  Additionally, issues with methodology and misleading the public about the study findings cast serious doubt on the legitimacy of this report.

Methodology

To support their claims, the authors tested for various chemicals and substances in 77 private water wells across five counties within the Eagle Ford Shale. In addition to using an insignificant sample size of only 77 wells – something this same UTA research team received criticism for in the past – the depth of over half of the wells (42) studied was determined by “owner recollection and available documentation”, with the depths of the remaining wells left unknown.

Since the study compared concentrations of compounds found in the water against well depth to determine a possible source, not knowing the depths of 45 percent of the wells sampled undoubtedly interferes with the accuracy of the results.

Authors admit contamination “unlikely to be attributed” to O&G activities

From these samples, the authors categorized the wells into two distinct groups based on the different ratios of chloride and bromide found in each well – one group with normal chloride/bromide levels (Group A), which made up about 86 percent of the wells, and another with abnormal levels (Group B). As the study states:

“An analysis of bromide/chloride ratios revealed two distinct subpopulations: one that appears to fall on a mixing line exhibiting bromide/chloride ratios around 0.0034 (n=66, hereafter Group A), which is close to seawater composition (Davis et al., 1998; Hem, 1992) and connate formation water (Kreitler and Richter, 1986), yet consistent with historical measurements (Chowdhury et al., 2006); and one with surprisingly higher levels of bromide (N0.5 mg/L) exhibiting bromide/chloride ratios ranging between 0.022 and 0.101 (n=11, hereafter Group B) (Fig. 2).” (pg. 2)

Unsurprisingly, while the press release states that the high bromide concentrations in the Group B wells are “indicative of contamination events”, the authors directly point out that unconventional oil and gas development (UD) is not the cause of these events. According to the study:

“The geographical clustering of Group B samples in areas with similar frequency of UD activity as compared with Group A samples (Fig. 1), coupled with low levels of chloride that are not consistent with oil field brine contamination, suggest that the elevated bromide/chloride ratios defining the Group B samples are unlikely to be attributed to contamination from UD activities.” (p. 5; emphasis added)

Many wells studied weren’t drinking water wells

In addition to examining the chloride/bromide concentrations, the authors tested for a range of constituents in each well, including: pH, total dissolved solids (TDS), volatile organic compounds (VOCs), metals and salts. These compound concentrations were then compared between the two groups, as well as against the Primary and Secondary Maximum Contaminant Limits (MCLs) defined by the U.S. Environmental Protection Agency’s (EPA) Drinking Water Standards.

While it’s common practice to compare compound concentrations in drinking water resources against EPA drinking water standards, it’s very important to point out that the many of the wells studied in the report are not defined as drinking water resources. In fact, the authors mention multiple times that these well are used for purposes other than drinking water. As the study mentions,

“The wide range of well depths within Group A is likely a reflection of varying uses from residential (shallow wells generally <61.0m) to agricultural and industrial service wells (deeper wells generally 304.8m).” (p.4; emphasis added)

Confusingly however, neither the authors nor the press release make an effort highlight the fact that several of the wells studied are not used for drinking. More confusing still, is that the press release for the study actively misleads the public with the headline “UTA Study Finds Abnormalities in Drinking Water in Eagle Ford Shale Region.”

But it’s not just the headline that is misleading.  The press release notes the authors’ findings of “dissolved hydrocarbon gasses” in water samples near oil and gas development, which many would assume shows a contamination of drinking water wells near these sites. According to the release:

“There were also multiple cases of water effervescence indicative of dissolved hydrocarbon gasses and a few volatile organic compounds in areas near oil and natural gas extraction sites.” (emphasis added)

However, the press release fails to mention that those wells showing “water effervescence” were converted oil wells used to provide water for nearby oil and gas development – not drinking water or residential wells. As the study mentions:

“In particular, three water wells had TOC [total organic carbon] values above 130 mg/L, which might be reflective of the fact that these wells were once shallow conventional oil wells converted into water wells to service nearby unconventional oil and gas activity.” (p. 5; emphasis added)

It continues,

“No VOCs were detected in any of these three samples; however, the high TOC values could be attributed to dissolved gases as the water from each of these wells was effervescent and consequently exhibited a phase separation at the surface indicative of a large amount of dissolved gases being present.” (p. 5; emphasis added)

Compounds found to be naturally occurring, from other sources

With this in mind, several compounds were nonetheless compared with EPA defined MCLs. These compounds include total dissolved solids, which, while above the secondary MCL of the EPA, were not out of the ordinary as the results were consistent with previous reports. Some compounds were found to be in irregular concentrations though, such as fluoride and nitrate.  According to the report, the median level of nitrate from water samples in Group B was found to “20-fold higher” than levels in Group A, and correlated negatively with well depth. The authors point out that such a result is likely to have been caused by agricultural activities in the region, not oil and gas development. As the study mentions,

“In addition to UD activities, the Eagle Ford Shale region also facilitates a variety of agricultural practices including, but not limited to, cattle, poultry, and grass production. It is quite possible that these processes contribute nitrate to shallow groundwater from the surface; as overall, 26 of the 42 nitrate detections were observed from water wells with a depth <61.0M (6 of the 42 nitrate detections also came from wells of unknown depth).”

Fluoride, too, was found not to be associated with fracking in the region, but instead the results suggest is it naturally occurring from the volcanic rocks in the formation or from other processes. According to the study,

“These findings suggest that the loading of fluoride in Group A may be attributable to the leakage of brines and possible rock-water interactions (Uddameri et al., 2014), whereas the loading of fluoride in Group B may be attributed to other processes.” (p. 4)

The authors continue, explaining that if “rogue hydrofluoric” acid from fracking were the cause of this increased fluoride concentration, there would be other signs, but these were not present:

“However, if rogue hydrofluoric acid from UD operations were contributing to the elevated levels of fluoride, one would expect to observe a drop in pH that would be represented by an inverse correlation between fluoride concentrations and pH. This relationship was not observed in the dataset as a whole (r=0.06, p=0.611). In fact fluoride concentrations correlated positively with pH values within the Group B samples (r=0.60, p=0.051), further indicating that fluoride levels detected in Group B Sample were likely not attributed to contamination from shale-drilling-related acidization.” (p. 4, emphasis added)

The authors elaborate further, explaining that the results show the fluoride is most likely natural occurring. As the study mentions:

“The lack of covariation between fluoride and a majority of all of the other analytes in Group B, except for inorganic carbon (r=0.673, p=0.023), suggest that the fluoride detected in Group B samples may be attributed to other naturally occurring hydrogeochemical processes that have not been previously documented in the southern Texas region.”

Conclusion

Overall, this study plainly lays out that fracking is not the source of the “abnormalities” found in these South Texas water wells. The results point to a number of other sources including natural and agricultural processes.

More concerning however, is that while this study fails to connect unconventional oil and gas activities with the water quality irregularities examined, the authors nonetheless make a concerted effort to connect the two throughout the study and the press release. For example, while one author, Kevin Schug, states that they don’t know the exact source of the abnormalities, in the same breath he mentions that working with the oil and gas industry is necessary to see if development activities are having an impact. As the press release states,

“’The next step in this research is to further characterize what these gases and chemicals are attributed to. Hopefully closer collaboration with industry to gain more intimate knowledge of the chemicals used during shale energy extraction will allow us to better assess whether or not unconventional oil and gas development is having a significant impact on groundwater quality in the south Texas region,’Schug added.”

Additionally, the authors’ failure to acknowledge the complete inaccuracy in the headline of the press release –characterizing the wells as “drinking water” when in reality the study mentions that several are used for agriculture and industrial services — is confusing at the very least. Couple this with the study suffering some of the same methodological issues that have plagued previous studies from the same UTA team (examples: here, here, here) and this appears to be yet another example of the UTA researchers trying to blame fracking with little justification.