Technical Memorandum on Assessment of Groundwater Sampling Results Completed by the U.S. Geological Survey by Tom Myers, Ph.D., September 30, 2012
The organic chemistry at MW01 has not changed substantially since the EPA sampled the well; some constituents have increased and some have decreased, as would be expected with organic contaminants discharging from a series of event, the hydraulic fracturing of natural gas wells. Because the water chemistry data at MW01 has essentially been replicated, the evidence supporting the hypothesis that natural gas drilling activities, including fracking, have contaminated the Wind River aquifer near Pavillion WY has been strengthened. The conclusions based on that analysis should be more widely accepted now that the water quality has been replicated.
The concentrations of gas, including methane and ethane, have increased and that of propane has remained relatively constant. The ratio of ethane and propane to methane and the isotopic signature of methane all indicate that the gas source is thermogenic, meaning a deep formation. An increasing concentration indicates the formation is likely the source because the concentration will increase as more of the formation contributes to gas at the monitoring well.
EPA monitoring well 2 was not sampled because it did not yield sufficient water. The EPA had been able to purge over a borehole’s volume of water, therefore they were clearly sampling formation water. There is no reason to consider that the current condition of MW02 negates the results of the EPA in 2011. The problems with MW02 however indicate other problems with the sampling of these wells. The USGS used standard purge techniques, not techniques designed to minimize losses of volatile organics to the atmosphere. Purging too fast or drawing the water level too low could cause the measurement to be biased too low.
This memorandum takes the USGS study an additional step by comparing the results released in the new study with the original EPA report (EPA 2011). It considers whether the new data refutes the original EPA study, either with the actual chemistry data collected or by showing problems with EPA monitoring well 2. … The USGS did not sample exactly the same constituents as did the EPA. The USGS sampled many constituents and their Table 7 lists many that had below detect (ND) levels, as did the EPA. … The concentrations of potassium (K) and the pH level are still much higher than the background levels in the formation, although K has decreased since the EPA sampling. EPA linked the presence of potassium to its use as a crosslinker and solvent during fracking, according to the Material Data Safety Sheets provided by the industry. Most of the fracking occurred several years ago, therefore the source is not a continuous release. A relatively conservative element such as potassium could move through the aquifer much more quickly than some of the organics.
Gasoline range organics and the various carbon-chain gases were found at concentrations that have increased significantly since the EPA study. Benzoic acid was found at concentrations similar to the EPA (2011). Diesel range organics and phenol remained present but at lower concentrations. The USGS found at least nine organic constituents that the EPA had either not found or not tested for. USGS found acrylonitrile at 21 ug/l in one of the replicate samples, not presented in Table 1. At least six constituents that had been detected by the EPA (2011) were not detected by the USGS. At least six constituents that EPA has found at various concentrations were not tested for by the USGS. The concentration of organics at Pavillion should vary for several reasons. Changes from one sampling event to the next do not represent a trend. A non-detect does not prove the constituent does not exist.
Organics attenuate by interactions with clay and silt sized particles so seasonal changes could be expected. This sampling occurred during late April, a time period during which recharge should be highest, since there is a mound in the shallow groundwater suggesting downward movement of water. Such vertical flow could dilute the formation water and cause seasonal changes not accounted for in spot samples as collected by the USGS. The concentration of methane and ethane increased substantially and that of propane remained relatively constant. The stable isotope ratios of carbon vs. hydrogen in methane are also almost exactly as found by the EPA. The gas in MW01 is thermogenic, and its concentration is increasing. An increasing concentration of thermogenic gas suggests its source is the formation rather than a leaky gas well. … If the formation is the source, the gas contamination will continue as long as the source releases gas.
The USGS did not sample MW02 because the well reportedly yielded only about 1 gallon per hour (Wright et al 2012). … There is no good explanation for the well’s failure to produce sufficient water for sampling, but its failure does not obviate the results found by the EPA for that well. The fact that the well produced substantial water from the sandstone twice indicates that the formation contained the constituents.
Bias Due to Volatilization
Most of the organic chemicals sampled for at the EPA monitoring wells will volatilize, meaning be lost to the air from the sample, under the correct conditions. In general those conditions are due to exposure to air which can be enhanced due to turbulence (Nielsen and Nielsen 2006). Sampling a well just after purging without allowing the well to recover without pumping can cause more volatilization and decrease the amount of constituent recovered in the sample (Herzog et al 1988). Too much purging or purging that causes too much drawdown can also increase volatilization because of the speed with which groundwater flows back into the well (McAlary and Barker 1987). Purging too rapidly or not sampling at the correct time after recovery can cause a bias in the resulting sample concentration. This could have occurred at both the USGS sampling of MW01 and in the EPA’s sampling of MW01 and MW02. Concentrations of organics, particularly VOCs, should be considered as potentially low compared to the background groundwater. [Emphasis added]
[Refer also to: Can Fracking Be Cleaned Up? The International Energy Agency says yes, but it will take tougher regulations that force producers…. “some fracking sites are relatively near to the level of drinking water, and the IEA suggests it might make sense to ban the procedure at such locations”