Now What? Key Assumptions Underlying Centredale Manor Remedy Found Arbitrary and Capricious

The Centredale Manor site consists of an approximately nine-acre peninsula, formerly occupied by a chemical manufacturing operation (a predecessor to Emhart) and a barrel-reclamation facility (New England Container Company, or NECC), and a roughly two-mile stretch of the adjacent Woonasquatucket River, related ponds, and their floodplains. (NECC settled with the United States shortly before the Phase 1 trial. The third party defendants in Phase 3 are mostly former customers of NECC alleged to have sent drums to NECC for reconditioning.)

The peninsula (referred to as the source area), river sediments, and floodplain soils are contaminated, largely with dioxin (more specifically, 2,3,7,8-TCDD). After many years of study and an administrative record of thousands of pages, the EPA published its record of decision (ROD) in 2012. The EPA selected a remedy that includes installing an impermeable (Resource Conservation and Recovery Act (RCRA)) cap on the peninsula and removing soil and sediments to as low as 15 parts per trillion for dioxin, which some have argued is indistinguishable from background levels. The most highly contaminated material would be shipped offsite for treatment (i.e., incineration), while the remaining excavated or dredged material, containing lesser amounts of dioxin, would be placed in an upland confined disposal facility (CDF). The EPA has not found a location for the CDF, a point raised by Emhart but ultimately found unpersuasive by the judge.

In 2014, the EPA issued a unilateral administrative order (UAO) requiring Emhart to implement the selected remedy. Emhart refused to comply, arguing that it had “sufficient cause” for not complying based on its good-faith challenges to the selected remedy. The government sought penalties from Emhart for failure to comply with the order. The court rejected this request, stating the following: “given that Emhart has continuously objected to critical aspects of the remedial design that the Court has now found arbitrary [discussed below], and Emhart’s previous participation in the cleanup process, the Court finds that Emhart’s challenge to the UAO up to this point has been pursued in objective good faith.” The court also refused to grant an injunction requiring Emhart to comply with the UAO.

Throughout the administrative process and continuing into the litigation, Emhart claimed that numerous aspects of the remedy-selection process were flawed. (In its motion for reconsideration, the United States stated “Emhart tried to prove more than 60 individual instances where EPA decisions were wrong.”) Emhart’s post-trial brief setting forth its arguments was 277 pages long, plus appendices. The government’s post-trial brief was 352 pages long, followed by another 160 pages for Emhart’s reply. In the end, the court found three aspects of the remedy selection to be flawed, holding that the EPA: (1) did not sufficiently support its decision to classify the groundwater under the peninsula as a potential source of drinking water (which classification was tied to the decision to require a RCRA cap); (2) incorrectly assumed that no largemouth bass were consumed from one of the downgradient ponds, thus arbitrarily increasing the potential risk from fish consumption; and (3) failed to justify its assumption about the amount of fish consumed from the ponds at the site, which assumption was a departure from relevant guidance relied on by the EPA.

Before addressing the merits of Emhart’s challenges, the court had to deal with preliminary questions about what evidence and arguments it could consider. While judicial review under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) is generally limited to a review of evidence in the administrative record, the Centredale court heard expert testimony from both sides, given the highly technical nature of the subject matter involved. The focus of the court’s review was still the administrative record, but the extra-record evidence was used “as an aid to understanding the administrative record.” Regarding the arguments the court could consider, that is generally limited to arguments presented during the notice and comment period. However, the court held that the EPA was required to justify “key assumptions” that were part of its analysis, whether or not a party previously commented on or objected to these assumptions. As such, if Emhart raised an argument for the first time during the litigation, the court held that Emhart had waived the argument unless it related to a key assumption made by the EPA.

Two such “key assumptions” related to factors that influenced the baseline human health risk assessment (BHHRA) that the EPA conducted for the site. During the notice-and-comment period, Emhart questioned the sufficiency of the BHHRA, but did not raise all of the specific issues it raised in the litigation. The court nonetheless considered all of these BHHRA-based challenges, given that “EPA has emphasized that the BHHRA was the primary driver of EPA’s cleanup standards.” Examples of these key assumptions included fish-consumption rates and levels of exposure to media throughout the site. (The EPA argues, in its motion for reconsideration, that changing these key assumptions as Emhart suggests will not materially affect the risk assessment and thus, the selected remedy.)

Of Emhart’s numerous challenges to BHHRA assumptions, the judge found two persuasive. First, the EPA assumed that no largemouth bass were present in Allendale Pond, despite knowing that the sampling on which this assumption was based was not representative. The sampling had been done after a breach in the Allendale Dam, which likely resulted in largemouth bass being washed out of Allendale Pond into the downstream Lyman Mill Pond. The absence of largemouth bass affected the BHHRA. Contaminants concentrate more in fish organs than they do in the flesh, so consumption of whole bodies of certain species, such as eels and white suckers, will lead to higher contaminant exposure. (Emhart unsuccessfully challenged the EPA’s assumption that people consume the whole bodies of white suckers.) However, risk assessors generally assume that people eat only the fillets of largemouth bass, which contain lower contaminant concentrations comparatively. Therefore, the failure to include any consumption of largemouth bass from Allendale Pond, which meant greater consumption of whole-body species, led to an arbitrary increase in projected risk.

The court also found fault in the EPA’s estimate of the reasonable maximum consumption rate of fish from the site. In making this estimate, the EPA considered data from the Maine Angler Survey, a study of fish-consumption rates in Maine from 1992. The survey estimated the rate of fish consumption to be approximately 14 grams per day. However, for the Centredale site, the EPA assumed that the entire amount of fish consumption came from fish caught at the site. This assumption failed to account for fish caught at other locations or fish purchased for consumption, contrary to the data from the Maine Angler Survey. Given the EPA’s lack of a justification in the record for this departure, the court found the fish-consumption rate arbitrary.

The court also took exception to the EPA’s classification of groundwater beneath the source area as a potential source of drinking water. For years, Rhode Island classified the groundwater as unsuitable for use as a drinking-water source due to contamination not only from source-area activities but also from other sites along the river upgradient from the peninsula. The EPA’s initial feasibility study reached a similar conclusion. The EPA changed its position in the ROD, changing the classification to Class II, which applies to current and potential sources of drinking water. This change in classification led to stricter cleanup goals for the groundwater and was part of the justification for requiring a RCRA cap at the source area.

In reviewing Emhart’s claim regarding the groundwater classification, the court cited the national contingency plan (NCP), which directs the EPA to restore usable ground waters to beneficial uses where practicable. The court also considered the EPA’s claim that its revised classification was consistent with EPA guidance. The court held that the EPA had to comply with the NCP, regardless of whether its decision was consistent with guidance. The court also held that the EPA did not sufficiently collect information or justify its decision to classify groundwater as a potential source of drinking water. The court acknowledged that the EPA has broad discretion in how it classifies source area groundwater, but if it continued to classify it as drinking water, it must demonstrate that restoration is practicable and will be effective.

The court denied a number of other challenges by Emhart, but in some cases retained jurisdiction. For example, Emhart argued that the EPA did not collect sufficient soil and sediment data to adequately estimate the amount of material that will require removal. This inevitably leads to considerable uncertainty in the cost estimates. Similarly, Emhart asserted that the amount of excavated material that will require incineration may have been substantially underestimated, leading to a lower cost estimate than is more reasonably likely to occur. The court was not persuaded by these arguments, finding that the EPA collected enough data to adequately characterize the site, and any uncertainty can be addressed by additional sampling and other efforts during the remedial design phase. The court went on to say that “if information uncovered during the remedial design reveals that the cost will differ significantly from the cost outlined in the ROD, EPA has the responsibility to update the administrative record as necessary.” The EPA could issue an explanation of significant differences, a ROD amendment, or a revised Feasibility Study. The court stated that “while CERCLA and the NCP leave it up to EPA to determine whether such actions are necessary in the first instance, the Court retains jurisdiction over this matter in the event that Emhart, in light of new evidence, seeks to challenge EPA’s decision.” Given the defiant tone EPA took in its motion for reconsideration (“no further administrative process is required by CERCLA or the NCP”), the likelihood that EPA would consider any of these significant steps seems unlikely, at least in the short term.

That is, unless Administrator Pruitt takes a fresh look at this matter, as the remedy cost is well in excess of his $50 million threshold.

James P. Ray is a partner in the Hartford, Connecticut, office of Robinson & Cole LLP.

^{Copyright © 2017, American Bar Association. All rights reserved. This information or any portion thereof may not be copied or disseminated in any form or by any means or downloaded or stored in an electronic database or retrieval system without the express written consent of the American Bar Association. The views expressed in this article are those of the author(s) and do not necessarily reflect the positions or policies of the American Bar Association, the Section of Litigation, this committee, or the employer(s) of the author(s).}