I have been writing a series of posts over the last month on the final report from the U.S. Environmental Protection Agency (EPA) covering the environmental impacts of hydraulic fracturing of oil and gas wells on drinking water resources. The report is entitled:  “Hydraulic Fracturing for Oil and Gas: Impacts from the Hydraulic Fracturing Water Cycle on Drinking Water Resources in the United States” (EPA-600-R-16-236Fa, December 13, 2016.) Today’s post concerns the potential impacts of chemical spills during the hydraulic fracturing process. This is my fourth post on the report overall and I will admit up front, it’s the worst post I have ever written. Not my fault. Unfortunately, there is just a total lack of quality data in the part of EPA’s report which covers chemical spills at oil and gas well sites being hydraulic fractured.

 

Hydraulic fracturing is not the only industry where mixing and transferring chemicals and fluids create spills. Spills happen in all sorts of industries and with all sorts of transportation. If you have worked in the environmental field for a while then you know chemical spills are still a problem throughout the United States, even though regulators and companies alike have worked aggressively to prevent them over the last forty odd years. I thought hazardous materials spill responders would all be out of a job by now. Hah, not so. The reason they’re not all collecting unemployment is mostly due to the human element involved in the process of mixing or transferring chemicals – in other words you and I screwing things up.

 

For their report, the EPA has sought to quantify the number and type of chemicals spilled at well sites which have undergone hydraulic fracturing and to determine the causes of the spills. EPA used data gathered by state governments and industry over a period of six years starting in 2006. EPA sorted through all this data and drew some conclusions.

 

For example, EPA learned through their data analysis that the combination of chemicals used in hydraulic fracturing is based on local conditions. Local conditions are such things as the type of rock formation being fractured and the type and quantity of fluid with which the chemicals are mixed.  The variability of chemical mixtures from site to site results in there being very little in common nationally in the types of chemicals spilled at oil and gas well sites. The chemical formulation of a fracking fluid in Pennsylvania can be completely different from the fracking fluid formula in West Texas or North Dakota.

 

The state and industry data also showed human error as the cause of the greatest number of spills at well sites, followed closely by equipment failure (such as hoses transporting chemicals) and failure of  containers storing chemicals. Out of approximately 36,000 spills reported in the state and industry data, EPA identified only 457 that were at or near the actual well site being fractured. Of these 457 spills, 151 were a result of chemical mixing operations at the well site. Of the 151 spills, 54 came from storage containers. Sixteen of the spills were actually caused by holes and cracks in the containers. In fact, storage containers were the number one source of spills which reached a drinking water resource. Storage container failures resulted in the most concentrated chemical spills and the largest volume of chemicals spilled, because the chemicals are generally being stored in a concentrated form. Once the chemicals leave storage and are mixed in tanks with fluids (generally water) for use in hydraulic fracturing then they become diluted and the impact of a spill is not as severe. One of the problems with the state and industry data collected by EPA is that there’s little actual information on the exact chemicals spilled at the well sites. There are just general descriptions, such as: oops we spilled some acid or a biocide. There’s better information on the volume of the spills though. The amount of chemicals spilled as a result of the mixing process ranged from 5 to 19,000 gallons.

 

Of the 457 spills which occurred at an actual well site, 225 were of produced water. Produced water consists of the natural groundwater in the oil and gas formation plus the remnants of the hydraulic fracturing fluid. Once hydraulic fracturing has taken place and pressure is released at the wellhead on the surface, then produced water comes back up the well bore. Usually it is collected in tanks, but obviously there have been spills. Produced water contains only very dilute amounts of the hydraulic fracturing fluid chemicals.

 

I know what you are thinking. You saw the big number of 36,000 spills. What the heck are all those spills? Sorry it is a bit of a mystery, because the data used by EPA is so bad. Other than the 457 spills mentioned above, EPA couldn’t determine where approximately 12,000 of the spills occurred. On top of that, 24,000 of the spills weren’t at the oil and gas well sites at all.

 

Did any of the identified 457 spills at the well sites cause contamination? You bet, 101 of the spills contaminated something, mostly soil (97 of the spills). Thirteen spills reached surface water and 9 of the spills contaminated both soil and surface water. There was no reporting on whether groundwater was contaminated because nobody is monitoring groundwater at hydraulic fracturing well sites.

 

What does this all mean? I don’t know. It’s the reason I called this post the worst I’ve ever written. The weak data on which the report is based makes it hard to say anything about the state of drinking water contamination caused by surface chemical spills at hydraulic fracturing well sites. Obviously there is a big gap in EPA’s analysis. But you know, it is hard to effectively quantify spills from incomplete and inconsistent data.

 

Probably the most interesting part of EPA’s report on chemical spills at hydraulic fracturing well sites is their analysis of the potential for chemicals used in hydraulic fracturing to migrate into water. Not all chemicals are equal in their ability to move through soils or water. EPA uses a set of chemical properties (for example the ability of the chemical to volatilize into the air and the ability of a chemical to dissolve in water) to determine if the chemical is one that is going to immediately contaminate a drinking water resource or if the chemical is going to hang around all gummed up in the soil and be a source of long term contamination as it slowly leaches drop by drop into groundwater and surface water. Of course nowadays, most spills that contaminate soils get cleaned up right away, so the second scenario involving long term contamination is unlikely in most cases.

 

Here’s what I thought was interesting about the chemical migration analysis. EPA states that of the 20 most frequently used chemicals in hydraulic fracturing nationwide, the majority are soluble (meaning they readily dissolve in water). The most soluble and mobile in water are methanol, isopropanol and ethylene glycol. There are a few petroleum based organic chemicals in the top 20 most frequently used chemicals in hydraulic fracturing that are not very soluble. But as EPA points out, there are other chemicals used in hydraulic fracturing fluid, like surfactants or alcohols, which can enhance the mobility of organic chemicals including those that are petroleum based. In other words, alcohol carries the organic chemicals along with it in the water. Fortunately chemicals in the environment are not necessarily forever. They are subject to breakdown by bacteria or by chemical transformation. So a chemical like methanol will eventually break down over a period of time to formaldehyde then to formic acid and finally to carbon dioxide.

 

Anyway, I’m going to say here, EPA’s report is less than a stellar account of the rate and fate of chemical spills at hydraulic fracturing sites. I suppose if you have bad data, you just have to kind of make do with what you have and at least try to say something about it. After all EPA has taken six years to produce a report to Congress. I guess they didn’t want to have to write in the report that their method of analyzing the scope and impact of chemical spills was a total bust. And I suppose regulators and industry can use the information, no matter how small a data set it is, to direct their attention to prevention efforts. Since most of the spills are from containers, they could concentrate attention on conducting inspections of storage areas. But does the report give a national picture of whether chemicals spilled during the hydraulic fracturing process are affecting drinking water resources? No, I don’t think so.