The Environmental Protection Agency’s (EPA) “Retrospective Case Study in Southwest Pennsylvania: Study of the Potential Impacts of Hydraulic Fracturing on Drinking Water Resources” is a work of nuance and intrigue. An exciting read. Well maybe not to everyone, but for me, it was science done at its best. The scientists used a multiple line theory of investigation to track down the cause of complaints of contaminated water in homeowner’s wells in southwest Pennsylvania. The complaints were similar. After hydraulic fracturing for gas in the Marcellus Shale had occurred nearby, homeowners found water from their wells turbid and odiferous. Their water tasted bad plus it left unsightly stains.

The EPA, as regular readers of this blog know, has been conducting “retrospective” studies of areas with significant complaints of groundwater impacts from hydraulic fracturing to determine if there was actually an impact to people’s drinking water wells and how it might have occurred. The study in southwest Pennsylvania was conducted in Washington County which is a short drive from Pittsburgh. I have visited Washington County and it is a pretty, rolling hill, rural sort of place that has a long history of coal mining and oil and gas development. The first oil well was drilled there sometime in the late 1800’s. Most people get their drinking water from the Monongahela River, but about a quarter of the residents are using groundwater. The groundwater there is pretty shallow, about 50 to 160 feet below the surface and the Marcellus Shale is pretty deep, about 5000 to roughly 7500 feet. The groundwater is generally pretty good, if you are going to drink groundwater. Most domestic wells are primarily producing a nice calcium bicarbonate type water and a lesser number of wells produce a sodium bicarbonate or calcium sulfate type. The problem with groundwater in Washington County is that it has a lot of naturally occurring iron and manganese. The groundwater also has naturally occurring methane gas. Remember methane is one of the components of natural gas but it can also be produced by bacteria as they decompose organic matter.

First, to determine if people’s drinking water wells had been contaminated by hydraulic fracturing fluid, the EPA analyzed for 133 organic compounds commonly used in hydraulic fracturing fluid. This doesn’t mean that companies are using 133 different chemicals every time they conduct hydraulic fracturing. Usually they use less than 10, but the chemical content of the fluid can vary depending on what the company is trying to achieve. So you’ve got to do a full analysis. I am happy to report that EPA found no organic contaminants they could trace to hydraulic fracturing in anybody’s water well.

EPA also analyzed the water for 102 other constituents including metals, stable isotopes, radionuclides, and your basic water quality constituents such as calcium, sodium, chloride and the like. Since there was no smoking gun of organic contamination from hydraulic fracturing in the wells, EPA looked at the results of stable isotope analysis, specifically the stable isotope signatures for the components of water in the drinking water wells – which we all know to be hydrogen and oxygen. The Marcellus Shale has water as well as gas. It is a brine type of water with lots of dissolved solids and other constituents such as sodium, chloride and bromide which make it briny. The brine is called formation or produced water and comes up oil and gas wells during hydrofracturing (remember this tidbit because we’ll come back to it later). Anyway the hydrogen and oxygen stable isotopes of the Marcellus brine and other deep rock formations are very different from shallow groundwater. If there was any movement of the formation water along with hydrofracturing fluid into shallow groundwater, it would show up in these analyses. Even a small amount of this briny fluid entering shallow groundwater would significantly change the stable isotope signature of hydrogen and oxygen in the drinking water wells. Again no bullet holes or dead bodies were found. Hydrogen and oxygen isotopes in people’s drinking water were the same as shallow groundwater in similar aquifers where no hydrofracturing has ever occurred.

Not to be deterred in their investigatory work, the EPA then looked at the stable isotopes of methane. As I mentioned earlier, methane is a component of natural gas and it is also formed by bacterial decomposition of organic material (these are commonly referred to as biogenic methane). The later is very common and occurs in wetlands and other places where you don’t have a lot of free oxygen, such as a groundwater well. Methane, which is composed of carbon and hydrogen, has very distinct stable isotope signatures for those components depending on whether it was produced at depth in an oil and gas formation or at the surface in a groundwater well. Unfortunately there were only two wells sampled that had sufficient methane in the water to analyze for stable isotopes. One sample showed clearly that it was biogenic methane; the other sample was less clear.

So EPA conducted a more definitive stable isotope test. This test was for what is called: dissolved inorganic carbon. Methane is considered organic carbon. There is another carbon that is associated with calcite in the rock formation. Calcite is what makes up limestone. It is calcium carbonate. As it moves along in an aquifer, the carbon in carbonate is oxidized, which results in the formation of carbon dioxide. When the carbon dioxide enters a water well (or other oxygen depleted location) the bacteria love it and use it to make biogenic methane, resulting in enriched inorganic carbon. Dissolved inorganic carbon produced this way has a very distinct isotopic signature compared to the inorganic carbon you find in the brine from the Marcellus Shale. The isotopic signature for dissolved inorganic carbon in the two wells tested was the same as other shallow groundwater in similar rock formations where no hydraulic fracturing has taken place and was totally different from the signature of dissolved inorganic carbon from the Marcellus. Evidence of contamination of people’s drinking water wells from hydrofracturing failed to appear.

Strontium isotope analysis of drinking water also failed to show impact from hydrofracturing. The radionuclide radium was then looked at because the Marcellus brine has some pretty extraordinary levels of both radium 226 and 228. Again there was nothing unusual in the drinking water wells to show impacts to groundwater from hydrofracturing.

The interesting find EPA made was simply using good old fashioned water quality data. EPA had several sets of historical water quality data from the area that had been collected decades before hydraulic fracturing even occurred in Washington County. EPA looked at such mundane water quality constituents as chloride, bromide, sodium and calcium and found that two springs in the southern part of Washington County exhibited some strange behavior during the collection of the 3 sets of samples EPA took over a year and a half period. The springs initially had elevated chloride and calcium levels exceeding both historical groundwater data and other groundwater data from wells EPA had sampled. Even more mysterious, those spurious high numbers decreased during the next two sampling rounds. It looked like the type of data you get when sampling groundwater after a spill. EPA looked around for what might be causing such an anomaly. It didn’t take long to find that both springs were close by and downgradient from an impoundment and reserve pit which had been used during the construction and hydrofracturing of a gas well. Remember the tidbit I told you to about where the briny formation water comes back up the hole during hydrofracturing and well production. The surface impoundment was used to store this water. So finally, here was a culprit. There was an extremely limited effect, since it was local and didn’t impact anyone’s personal water well. But obviously the use of a surface impoundment in this particular case caused leaching into groundwater and impacts to the two springs.

So if EPA couldn’t find any suspects from all of this incredible sleuthing, what did they say caused all of the homeowner’s water quality problems after hydraulic fracturing of the Marcellus Shale in southwest Pennsylvania? Here the EPA kind of failed me. They went into the territory of speculation. It is a good speculation though. It’s based on scientific literature and a good knowledge of groundwater reactions. Being water scientists, the EPA investigators looked at the water complaints, saw that there was a lot of naturally occurring iron and manganese in the groundwater and used their noggins to deduce that staining, taste and odor issues were probably a result of high levels of iron and manganese in the water.  In fact, when they tested the water for iron and manganese, they ran both a filtered and an unfiltered sample. The unfiltered sample was much higher for both of these constituents in the water, which leads them to believe the increased turbidity was the result of a lot of undissolved iron and manganese particulates entering the water wells. What was causing the sudden turbidity? EPA made note in their report that other studies have shown the drilling process for an oil and gas well produces a lot of vibrations and that has resulted in mobilization of iron and other rock constituents, like manganese, into water wells resulting in a lot of turbidity in those wells. So there you go, no contamination from hydraulic fracturing in southwest Pennsylvania, but possibly just poor water quality as the result of the drilling of the gas well. Little consolation for the homeowners who had to put up with foul taste and having their laundry stained, but thank goodness they aren’t drinking something really foul from the hydrofracturing fluid.