When one thinks of a Superfund site, the image is of a large landfill, a former mining pit, or an industrial site like Love Canal; places where massive contamination has released into the air, ground, or water over many, many years and where it is not safe to live, animals and benthic organisms may not even exist, and where the full power of the federal government is often needed simply to restore environmental and public health to minimal levels. A private water well containing somewhat elevated levels of several naturally occurring substances (like manganese, arsenic, or sodium) is not typically among these images.

Yet, in both Dimock, Pennsylvania and Pavillion, Wyoming, EPA has used its authority under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) to investigate claims of drinking and ground water contamination. While EPA has broad authority under CERCLA to investigate all types of potentially contaminated sites, the Agency’s use of CERCLA at these two sites raises serious policy and legal questions.

Perhaps the most important evidence that EPA’s use of CERCLA may be problematic, comes from the Agency itself. It its Action Memorandum in Dimock [PDF], EPA states:

In taking this action, EPA is aware of and has considered the potential applicability of the natural gas exclusion under CERCLA, the Bentsen Amendment under the Resource Conservation and Recovery Act (RCRA), and the exclusion to the definition of ‘underground injection’ under the Safe Drinking Water Act (SDWA). EPA has concluded that this action is appropriate under CERCLA at this time.

In its August 2009, Site Inspection Analytical Results Report in Pavillion [PDF], the Agency writes:

While CERCLA allows for groundwater plumes, it is generally based on a point source or sources, exposure pathways and targets. Therefore a plume with no documented source or location presents a poor fit.

Curiously, the most recent Pavillion draft report issued several weeks ago does not mention CERCLA, or otherwise discuss the Agency’s use of its statutory authority. Nonetheless, the Pavillion study has proceeded pursuant to CERCLA authority.

EPA’s most recent decision to use CERCLA in Dimock is especially intriguing—particularly in the broader context of the Agency’s ongoing efforts to be involved at hydraulic fracturing operations. In its first attempt—the emergency enforcement action against Range Resources in Fort Worth—the Agency used its authority under the Safe Drinking Water Act (SDWA), and it has not gone particularly well. There, EPA issued a SDWA emergency order (42 U.S.C. § 300i), claiming federal intervention was required because the state had not adequately responded to the imminent and substantial endangerment associated with contamination in a drinking well near a Range Resource natural gas drilling site.

Numerous obstacles have plagued EPA following the emergency order—notwithstanding what appears to be a very weak, if not non-existent, evidentiary basis linking Range’s operations (including hydraulic fracturing) to the contamination. To begin, the SDWA emergency order was a novel approach, given the Agency’s lack of jurisdiction over oil and natural gas wells under the SDWA’s general authority. Shortly following the order, the Agency acknowledged that no individuals or properties were in any immediate danger, calling into serious question the imminence and substantiality of the purported “emergency.” And comments by the judges in both the District Court and the Fifth Circuit Court of Appeals have generally favored some type of pre-enforcement review, meaning the Agency will have to substantiate its claims, something it has not yet been able, or at least willing, to do. Most recently, a Texas District Court judge ruled that the well owner at the heart of the Range dispute essentially rigged a video of his tap water burning, calling it “deceptive” and “calculated to alarm the public into believing the water was burning.” These events have clouded EPA's authority and use of the SDWA.

Yet, for one key reason, the SDWA remains the better vehicle to investigate and mitigate “drinking water” contamination. The SDWA affords EPA authority over “contaminants,” which is broadly defined as “any physical, chemical, biological, or radiological substance or matter in water.” In the world of EPA’s various statutory authority over pollution, this is about as broad as it gets, and would include any hydrocarbon-related substance (i.e., methane).

CERCLA on the other hand, is much more narrow. CERCLA authority is confined to “hazardous substances,” which expressly precludes petroleum (including crude oil or any fraction thereof which is not listed), and natural gas and natural gas liquids. This means that EPA’s authority, investigative or otherwise, as a matter of law, is confined to non-petroleum hazardous substances. And as EPA acknowledges, CERCLA is most often used to investigate a known or suspected source of hazardous substances, like a plume, with a definable exposure pathway.

In Dimock and Pavillion, these CERCLA limitations are evident. For its part, EPA has strained to fit the Pavillion investigation into the typical CERCLA model. There is no delineated or suspected plume. The geology in the area is a mish-mash of various interrelated and overlapping formations, as opposed to more neatly layered strata. As a result, EPA has yet to, and is unlikely to be able to, identify any reliable exposure pathway (for hydraulic fracturing fluid or methane). Indeed, EPA Region 8 Administrator, Jim Martin, testified before Congress that EPA’s Pavillion study merely “hypothesized potential pathways” and that any scientifically supportable conclusion that comes from Pavillion will not be applicable to any other drilling operation anywhere else in the country. Hardly a ringing endorsement for the use of CERCLA.

In Dimock, EPA is pinning its hopes (as it must) solely on finding hazardous substances in unacceptably high amounts, and then linking those substances to fracking fluid. To support its case, EPA has focused on three substances—manganese, sodium, and arsenic—in four wells for which the Pennsylvania Department of Environmental Protection (DEP) collected data. EPA’s toxicologist reviewed the DEP data, and concluded [PDF] that four of eight wells had elevated levels of sodium and manganese, and one well had concentrations of arsenic sufficient to cause a “long-term” cancer risk. In the follow-up Action Memorandum, EPA’s On-Scene Coordinator (OSC) took liberties with the toxicology report, concluding that “historic drilling activities “ and “spills and other releases” from these drilling activities “is reason to believe that a release of hazardous substances has occurred.” The Action Memo, however, does not provide supporting evidence for these statements.

Sodium (salt), manganese, and arsenic are ubiquitous, naturally occurring elements. EPA is precluded under Section 104(a)(3) from using CERCLA to respond to a release or threat of release of a naturally occurring substance through naturally occurring processes. Therefore, EPA’s case must turn on establishing a causal link between the drilling operations and the elevated levels of these naturally occurring substances in the four wells. Given the generally innocuous nature of sodium and manganese, and the rather low levels found in the Dimock wells, EPA is likely to zero in on the arsenic in well number 8. Aside from the fact some forms of arsenic occur naturally, there is no evidence that the operator ever used arsenic in its drilling or fracking fluid or that it is even a common component of fracking fluid. Even if EPA finds that drilling operations either contained arsenic or caused the arsenic to migrate into the one drinking well, such an isolated incident doesn’t seem worthy of EPA’s CERCLA authority.

Therefore, one must ask why EPA has chosen to proceed under CERCLA at both sites (and particularly in Dimock)? Unfortunately, the decision may have more to do with politics than practicality. The SDWA is the most logical choice for the Agency to curb true imminent and substantial endangerments to drinking water. However, the Agency’s SDWA authority is currently in flux, and the court’s may end up requiring a more substantiated causal nexus prior to enforcement than EPA is comfortable with. Until the Range Resources litigation is resolved, EPA is likely to shy away from the SDWA.

RCRA is arguably the second best option. RCRA Section 7003 grants EPA broad authority to mitigate imminent and substantial endangerment for hazardous and solid wastes, including oil and gas exploration and production (E&P) Subtitle D wastes. Unlike CERCLA, however, RCRA does not ban pre-enforcement judicial review, and EPA is confined to seeking only injunctive relief (i.e., no civil penalties, no cost recovery, and no punitive damages). In other words, RCRA would not be as punitive.

For the reasons highlighted above, CERCLA is the most cumbersome authority available to EPA. Given the inherent statutory hurdles, the fact EPA is using CERCLA tends to show the Agency is more focused on finding a smoking gun linking hydraulic fracturing to ground or drinking water contamination, than actually protecting the public health and environment in Dimock or Pavillion. This seemingly politically-motivated approach is concerning.

Industry is rightly concerned about EPA’s broad CERCLA enforcement authority. Most notably, the Agency can issue a Section 106 Unilateral Administrative Order (UAO) compelling remedial or removal actions where an imminent and substantial endangerment may exist, without pre-enforcement judicial review, and subject to substantial penalties, including cost recovery and punitive damages. The use of CERCLA is also problematic from a public relation perspective, as the public is generally familiar with the negative connotations of “Superfund,” and a challenge to a CERCLA order—no matter how meritorious—will inevitably be used as fodder by anti-drilling activists.

It remains to be seen how this will all shake out, or whether EPA will succeed in linking hydraulic fracturing to the water contamination in Dimock, Pavillion, or elsewhere. In the meantime, it is clear that EPA is committed to using all available statutory tools at its disposal to regulate or enforce against hydraulic fracturing operations. The recent reliance on CERCLA is particularly troubling and signals that EPA is ratcheting up its enforcement efforts. Some have suggested that companies should be conducting baseline monitoring of groundwater and drinking water conditions prior to drilling. While there are drawbacks to such monitoring, including access issues, pre-drilling data disproving a causal connection between alleged water contamination and a drilling operation may be the best tool in the industry’s arsenal to counter a very active EPA.