Gould Elecs. Inc. v. Livingston Cnty. Rd. Comm'n

Full title: GOULD ELECTRONICS INC., Plaintiff, v. LIVINGSTON COUNTY ROAD COMMISSION…

Court: UNITED STATES DISTRICT COURT EASTERN DISTRICT OF MICHIGAN SOUTHERN DIVISION

Date published: Nov 19, 2020

Case No. 17-11130 (E.D. Mich. Nov. 19, 2020)

Opinion

Case No. 17-11130

11-19-2020

GOULD ELECTRONICS INC., Plaintiff, v. LIVINGSTON COUNTY ROAD COMMISSION, Defendant.

HON. MARK A. GOLDSMITH

OPINION & ORDER (1) CONTAINING FINDINGS OF FACT AND CONCLUSIONS OF LAW FOLLOWING BENCH TRIAL; (2) DENYING PLAINTIFF'S MOTION IN LIMINE TO BAR TESTIMONY FROM DEFENDANT'S EXPERT (Dkt. 92); (3) GRANTING IN PART AND DENYING IN PART AS MOOT PLAINTIFF'S MOTION IN LIMINE (Dkt. 109); (4) DENYING AS MOOT PLAINTIFF'S PRETRIAL OBJECTIONS TO PROPOSED TRIAL EXHIBITS (Dkt. 135); AND (5) DENYING AS MOOT DEFENDANT'S PRETRIAL OBJECTIONS TO PROPOSED TRIAL EXHIBITS (Dkts. 148 , 149, 150, 151, 152, 154) TABLE OF CONTENTS

I. INTRODUCTION ............................................................................................................. 4

II. BACKGROUND ............................................................................................................... 6

A. The Properties & Historical Operations ............................................................................ 6

B. Discovery of TCE ................................................................................................................ 8

C. Expert Opinions ................................................................................................................ 10

D. LCRC's Petitions to EGLE ............................................................................................... 12

III. DISCUSSION .................................................................................................................. 14

A. Evidentiary Matters .......................................................................................................... 14

1. Travers's Testimony ..................................................................................................... 15
2. Feenstra's Testimony .................................................................................................... 16
3. Browning's Testimony .................................................................................................. 17
4. MSG Reports ................................................................................................................ 21
5. EGLE Records .............................................................................................................. 25

B. Final Administrative Action .............................................................................................. 30

C. Gould's CERCLA Cost Recovery Claim ........................................................................... 32

1. Prima Facie Case ........................................................................................................... 33
2. Statutory Defenses ........................................................................................................ 39

a. Third-Party Defense .................................................................................................. 39

i. Sole Cause ............................................................................................................... 40

a) Historical Operations—RSF Facility ............................................................... 41
b) Floor Drain & PVC Pipe ................................................................................. 43
c) Historical Operations—LCRC Property .......................................................... 44
d) The Line of Occupation ................................................................................... 48
e) Travers's Theory .............................................................................................. 50
f) Rebuttal to Travers's Theory ........................................................................... 54
g) Feenstra's Theory ............................................................................................ 58
h) Rebuttal to Feenstra's Theory .......................................................................... 61
i) Grading & Excavation Activities .................................................................... 66
j) Conclusions Regarding Causation ................................................................... 67

ii. Due Care .............................................................................................................. 72

b. Contiguous Landowner Defense ............................................................................... 83

3. Divisibility .................................................................................................................... 84

D. LCRC's Contribution Counterclaims ............................................................................... 89

1. NREPA Contribution Counterclaim ............................................................................. 90
2. Gould's Liability ........................................................................................................... 91
3. Equitable Allocation ..................................................................................................... 95

a. Parties' Contributions ............................................................................................... 96
b. Parties' Degree of Care and Cooperation ............................................................... 100
c. Allocation of Response Costs ................................................................................. 103

IV. CONCLUSION ............................................................................................................. 107

I. INTRODUCTION

This is an environmental contamination case in which the parties dispute whether and how the costs of cleanup and remediation must be shared between them under the Comprehensive Environmental Response, Compensation, and Liability Act ("CERCLA"), 42 U.S.C. § 9601, et seq., and under Michigan's Natural Resources and Environmental Protection Act ("NREPA"), Mich. Comp. Laws § 324.101, et seq. The Court conducted a seven-day non-jury trial between July 13, 2020, and July 21, 2020, by way of videoconferencing. The parties have submitted post-trial briefs and proposed findings of fact and conclusions of law, as well as responses to the post-trial briefing.

Plaintiff Gould Electronics, Inc. ("Gould") contends that Defendant Livingston County Road Commission ("LCRC") is responsible for contributing to a plume of trichloroethylene ("TCE") traversing the parties' adjacent properties and migrating northeast toward Thompson Lake in Howell, Michigan. Pl. Proposed Findings of Fact & Conclusions of Law ("PFFCL") at 5-9 (Dkt. 253). LCRC, in turn, maintains that Gould is solely responsible for the contamination. Def. PFFCL at 10 (Dkt. 252). Gould advances a cost recovery claim under CERCLA § 107(a), 42 U.S.C. § 9607(a), and seeks to hold LCRC strictly liable for the entirety of its response costs. See 3d Am. Compl. (Dkt. 76).¹ LCRC maintains that it is shielded from liability under the third-party defense, CERCLA § 107(b)(3), 42 U.S.C. § 9607(b)(3), and under the contiguous landowner defense, CERCLA § 107(q), 42 U.S.C. § 9607(q). Def. PFFCL at 31-37. LCRC also argues that because the harm is divisible, it must be apportioned entirely to Gould. Id. Alternatively, LCRC advances contribution counterclaims under CERCLA § 113(f)(1), 42 U.S.C. § 9613(f), and under NREPA, Mich. Comp. Laws § 324.20129, seeking to equitably allocate all response costs, including its own, to Gould. Am. Counter-Compl. (Dkt. 118).

1 Gould filed an earlier case against LCRC on July 6, 2009. Gould Elecs., Inc. v. Livingston Cty. Rd. Comm'n, No. 09-cv-12633 (E.D. Mich.). On May 29, 2012, the parties stipulated to dismiss that action without prejudice, in accordance with an agreement executed on May 21, 2012. See Stip. Order of Dismissal, Ex. A to Am. Compl. (Dkt. 22-2). The order of dismissal provided that either party could revive the surviving claims by filing a complaint initiating a new action. Id. ¶ 2. Gould initiated the present action on April 11, 2017.

CERCLA is designed to promote timely cleanup of environmental hazards by requiring investigation and remediation of contamination—not just by those who actually degraded the environment through direct disposal of the contamination but also by landowners and certain other "responsible parties," regardless of whether they caused the toxic condition. The response costs incurred in conducting investigations and remediations may later be recovered, in whole or in part, from other responsible parties. For the reasons discussed below, the Court finds that Gould's corporate predecessor dumped vast quantities of waste chemicals, including TCE, onto the ground and into floor drains on its property over the course of fifteen years. These chemicals leached through the soils and groundwater and migrated onto neighboring properties, including LCRC's adjoining property. Although LCRC did not cause or contribute to the TCE contamination, it is liable as an owner of property where contamination is leaching through the soils and groundwater. And it fails to win the protection that CERCLA provides for innocent landowners because it was dilatory in investigating the source of the contamination and failed to fully cooperate with the state regulatory agency's efforts to investigate the site.

Under these circumstances, the Court finds in favor of Gould on its CERCLA cost recovery claim and holds that LCRC is strictly liable for Gould's response costs. However, the Court also finds in favor of LCRC on its CERCLA contribution counterclaim. Having considered the equities, the Court holds that Gould must bear the vast majority of the response costs, given that its corporate predecessor caused the toxic conditions. Accordingly, Gould's equitable share of its past and future response costs, as well as LCRC's past response costs, is fixed at 95%. Because LCRC engaged in delay tactics and was recalcitrant in cooperating with the state agency, it must bear 5% of the response costs. Finally, the Court dismisses LCRC's NREPA contribution counterclaim.

II. BACKGROUND

A. The Properties & Historical Operations

This case involves the determination of responsibility for costs associated with the investigation and remediation of TCE contaminating two adjacent parcels of land. One of the parcels, located at 325 North Roosevelt Street in Howell, Michigan, was owned and operated by non-party Gould Inc., one of Gould's corporate predecessors, from 1961 through 1976 (the "Gould Property"). 3d Am. Compl. ¶ 5; Rich Dep. at 11-12 (Dkt. 189-21).² Over this period, Gould Inc. operated a piston and connecting rod manufacturing facility (the Roosevelt Street Factory, or "RSF Facility") at the Gould Property. Rich Dep. at 12; Cronmiller Dep. at 35 (Dkt. 249); Trial Tr. V at 60 (Richardson) (Dkt. 262).³ The RSF Facility spans nearly the entire length of the northern boundary of the Gould Property, as depicted in the image below. See Figure 1 to Travers Report at PageID.32137 (Dkt. 168).

2 The parties stipulated to the admission of all depositions referenced in this Opinion as de bene esse depositions. Trial Tr. VI at 117-118 (Dkt. 263).

3 Gould "manages the assets and liabilities" arising from the discontinued operations of Gould Inc. and admits that it is responsible for addressing the site investigation and remediation work at the Gould Property. Trial Tr. III at 73 (Callahan) (Dkt. 260); Rich Dep. at 14, 19; Answer to Am. Counter-Compl. ¶ 1 (Dkt. 120).

Image materials not available for display.

LCRC currently owns the property immediately to the east of the Gould Property, located at 918 North Street in Howell, Michigan (the "LCRC Property"). Answer to 3d Am. Compl. ¶ 6 (Dkt. 79). LCRC operated at the LCRC Property between 1934 and 1990. Trial Tr. III at 157 (Craine) (Dkt. 260). LCRC is responsible for constructing and maintaining roads, bridges, storm systems, and dams. Id. at 139-140. Its operations at the LCRC Property were limited to vehicle repair and maintenance, equipment and material storage, and office work. Id. at 150, 153, 159, 161; Trial Tr. IV at 173 (Messner) (Dkt. 261). It did not operate any manufacturing facilities on the LCRC Property. Trial Tr. III at 140 (Craine).

The parties presented evidence regarding the historical operations at the Gould and LCRC Properties.⁴ Former employees of Gould Inc. testified that it was common practice at the RSF Facility to dispose of chemicals used in the manufacturing process by pouring them onto the ground on the Gould Property. Additionally, a large degreasing tank, which likely contained TCE, was located at the far eastern side of the RSF Facility. One former employee testified that the tank's contents were disposed of by permitting them to flow into a floor drain at the RSF Facility. LCRC maintains that this floor drain was connected to a PVC pipe that ultimately emptied into a storm sewer on the LCRC Property.

4 For simplicity, the Court provides a broad overview of the historical operations, the parties' investigations and response activities, and the expert testimony, without citation to the record. These matters are discussed in greater detail in later sections, with extensive record citations.

Meanwhile, records and testimony demonstrate that LCRC used TCE between 1985 and 1986 to perform a series of asphalt tests on-site. Former LCRC employees testified that they disposed of the waste TCE used to perform these tests in a drum that was later picked up for disposal by Safety Kleen, LCRC's vendor for TCE and other chemicals. These employees denied that TCE was ever disposed of by pouring it or spilling it onto the ground. Gould, however, disputes how many tests were performed and contends that LCRC has not definitively established what became of the waste TCE after Safety Kleen picked it up. Additionally, Gould maintains that LCRC could have caused the contamination on the Gould Property, as LCRC's operations allegedly extended beyond the property line and onto the Gould Property.

B. Discovery of TCE

In 1988, the Michigan National Bank ("MNB"), a former owner of the Gould Property, voluntarily undertook an excavation and other remediation activities to address petroleum contamination on the Gould Property. In 1993, the Michigan Department of Environmental Quality ("EGLE")⁵ designated the Gould Property a "facility," based on the presence of hazardous substances at the site, and identified Gould as a potentially responsible party under state environmental law.⁶ Gould thereafter assumed from MNB responsibility for the investigation of the Gould Property. In 1994, Gould retained the Mannik & Smith Group, Inc. ("MSG"), an environmental consulting firm, to investigate and clean up the contamination on the Gould Property.

5 The name of the agency was later changed to the Department of Environment, Great Lakes, and Energy. For simplicity, this Opinion refers to the agency throughout as EGLE, its current moniker.

6 A "facility" is defined, in relevant part, under CERCLA as "any site or area where a hazardous substance has been deposited, stored, disposed of, or placed, or otherwise come to be located . . . ." 42 U.S.C. § 9601(9). The NREPA definition is similar. See Mich. Comp. Laws § 324.20101(s). Ownership or operation of a facility can give rise to liability under CERCLA and NREPA, subject to limited defenses. See 42 U.S.C. § 9607(a); Mich. Comp. Laws § 324.20107a(1).

As its work progressed, Gould believed that it was nearing completion of its cleanup and would soon obtain EGLE's closure of the site.⁷ However, TCE was unexpectedly detected on the Gould Property in 1997 or 1998. As a result, EGLE required Gould to expand its investigation to determine the sources of the TCE. MSG performed a series of additional investigations and, in 2001, performed an excavation in the northeastern corner of the Gould Property during which 400 to 500 cubic yards of soil containing TCE were removed. Additionally, in 2016, MSG installed a pump and treat system to treat the groundwater contamination. Finally, MSG implemented a bioremediation plan, a process that involves injecting additives into the ground.

7 Although NREPA does not define "closure," Rebecca Taylor, an EGLE environmental analyst, testified that EGLE closes a site when it approves a "no further action report." Trial Tr. II at 182 (Taylor) (Dkt. 259). Under NREPA, parties may submit reports seeking a determination of no further action "[u]pon completion of remedial actions that satisfy the cleanup criteria" established by NREPA. Mich. Comp. Laws § 324.20114a(2).

The LCRC Property became a subject of agency interest and concern some time after discovery of TCE on the Gould Property. On August 9, 2007, EGLE formally designated the LCRC Property a "facility" and identified LCRC as a potentially responsible party, based on the presence of TCE contamination on the site. After receiving this notice, LCRC retained consultants, including Quantum Environmental ("Quantum"), to perform an investigation of the LCRC Property in coordination with EGLE. In connection with its investigation, Quantum completed soil and groundwater sampling and installed temporary monitoring wells. However, it did not prepare or implement an interim response plan to remediate or abate the TCE contamination, because LCRC maintained that it did not cause the TCE contamination.

C. Expert Opinions

The parties retained expert consultants to evaluate the data generated as a result of the investigations and to determine the sources of TCE contamination on the Gould and LCRC Properties. During trial, LCRC's expert, Constance Travers, and Gould's expert, Stanley Feenstra, presented competing theories regarding the origins and migration of the TCE.⁸ An overview of the competing theories follows, with elaboration set out later in this Opinion.

8 While Travers and Feenstra are the parties' primary experts, Gould and LCRC also presented testimony from the expert consultants who performed the investigations. Specifically, Gould presented testimony from John Browning, while LCRC presented testimony from Keith Gadway.

In Travers's view, releases of TCE in the northeastern corner of the Gould Property were the sole sources of TCE contamination, which has since migrated through the soil and groundwater onto the LCRC Property and other nearby properties. She observed that the highest concentrations of TCE in soil and groundwater are located in the northeastern corner of the Gould Property, with concentrations decreasing with distance from that area. She opined that groundwater near the northeastern corner of the Gould Property has two separate flow components, creating a "groundwater divide" running on a roughly west-east axis. Groundwater to the north of the divide flows northeasterly toward Thompson Lake, while groundwater to the south of the divide flows southeasterly across the LCRC Property. TCE from the northeastern corner of the Gould Property has dissolved in groundwater and migrated with the groundwater flow to the northeast, toward Thompson Lake, and to the southeast onto the LCRC Property. Because high concentrations of TCE were not detected in the soil or groundwater on the LCRC Property, Travers concluded that no disposals of TCE occurred on the LCRC Property.

Feenstra acknowledged that releases of TCE that took place on the Gould Property represent one source of the contamination. He further conceded that some of this contamination likely migrated onto the northwestern corner of the LCRC Property. However, he also maintained that concentrations of TCE found to the south of a salt barn located on the LCRC Property are indicative of additional releases of TCE on the LCRC Property originating from deposits onto its soils, rather than from off-site migration.⁹

9 The salt barn is a structure located on the northwestern corner of the LCRC Property, which LCRC used to store road salt. Trial Tr. III at 162-163 (Craine); Figure 1 to Travers Report. It was constructed—and the ground on which it was situated was paved—between 1979 and 1980, Trial Tr. VI at 43 (Travers) (Dkt. 263), to replace a much smaller salt shed that remains located immediately to the south of the salt barn, Trial Tr. III at 162-163 (Craine). The salt barn is referenced in the record by various names, including the salt shed and the salt storage building. All references to the "salt barn" in this Opinion refer to the salt barn constructed in 1979 and 1980.

Feenstra opined that the TCE concentrations detected south of the salt barn could not be attributed to TCE migration from the Gould Property for two reasons. First, because a subterranean layer of clay located between the Gould and LCRC Properties slopes downward to the northwest, TCE could not have migrated uphill from the Gould Property onto the LCRC Property. Instead, TCE released on the LCRC Property would have migrated down the slope of the clay layer onto the Gould Property. Second, Feenstra stated that the groundwater divide is located further south than Travers opined, such that the groundwater in the northeastern corner of the Gould Property flows north and could not have transported dissolved TCE southeast onto the LCRC Property. Thus, Feenstra concluded that a second source area exists to the south of the salt barn on the LCRC Property.

D. LCRC's Petitions to EGLE

Throughout the course of its investigations and the litigation, LCRC has consistently maintained that it is not liable for causing the contamination. Although LCRC submitted a number of response activity plans to EGLE, it never proposed any plans to remediate the contamination, instead denying that it was a source of the contamination.¹⁰ However, EGLE continually requested, over roughly ten years, that LCRC perform additional investigations in order to determine the origins of the contamination with greater certainty.

10 A response activity plan is a request for EGLE's approval of proposed response activities to address contamination. See Mich. Comp. Laws § 324.20114b(1).

In response to EGLE's continued requests for additional data, LCRC retained counsel to contact EGLE's upper management regarding the investigations. Trial Tr. IV at 47-48 (Craine). At LCRC's request, Timothy O'Brien, an associate director at EGLE, met with LCRC representatives in early 2017 to review the investigations of the LCRC Property. Trial Tr. IV at 144 (O'Brien). Subsequently, on May 11, 2017, LCRC representatives, including Travers, met with EGLE staff members including O'Brien and Susan Leeming, the division chief of EGLE's remediation and redevelopment division. Trial Tr. IV at 48-49 (Craine); Trial Tr. IV at 145-146 (O'Brien). During this meeting, Travers gave a presentation describing the history of LCRC's work on the site, the data and findings from the investigations, and her opinions regarding the source of the contamination. Trial Tr. IV at 50-51 (Craine). No representatives from Gould attended or were invited to this meeting. Id. at 54-55; Trial Tr. IV at 168 (O'Brien).

Following the meeting, EGLE staff members conferred and concluded that EGLE had no further regulatory interest in the origin of TCE contamination on the LCRC Property. Trial Tr. IV at 147-154 (O'Brien); Trial Tr. V at 13-14 (Leeming). This conclusion was set out in two letters EGLE issued in June 2017 (the "June 2017 Letters"). The first letter, issued on June 14, 2017, determined that "there is no indication that a release of chlorinated solvents to unsaturated site soils occurred, and no releases of chlorinated solvents in LCRC property site soils are demonstrated to be directly attributable to LCRC's historic operations." 6/14/17 Letter (Dkt. 201-12). The letter further stated that this conclusion "should not be construed as a liability determination for the chlorinated solvent contamination on the subject property." Id. EGLE also summarized certain "data gaps" in the evaluation of the origins of TCE on the LCRC Property and urged LCRC to "settle its differences" with Gould in order to cooperatively address the contamination and complete its evaluation and remediation of the releases. Id.

When O'Brien and Leeming became aware of the content of the first letter—which they had not reviewed before it was issued—they agreed that it did not accurately reflect EGLE's position, as it suggested that LCRC needed to perform additional work. Trial Tr. IV at 154-155 (O'Brien); Trial Tr. V at 14-16 (Leeming). On June 23, 2017, EGLE issued a second letter clarifying its position. Trial Tr. IV at 156 (O'Brien); Trial Tr. V at 17-18 (Leeming). This letter stated that EGLE was not requesting LCRC to perform any further sampling or to submit any additional reports and that EGLE had "no further regulatory interest in the origin of the TCE contamination on [the LCRC Property]." 6/23/17 Letter (Dkt. 201-13).

Rebecca Taylor, a senior environmental quality analyst and EGLE's project manager overseeing the investigation of the Gould and LCRC Properties, explained that a "no further regulatory interest" determination means that EGLE simply has no regulatory concerns regarding a property, as distinguished from a "no further action" determination resulting in EGLE's closure of a site following the completion of response activities. Trial Tr. II at 125-126, 181-182 (Taylor) (Dkt. 259). However, Leeming stated that EGLE's "no further regulatory interest" determination meant that EGLE would not go forward with any actions against LCRC. Trial Tr. V at 15 (Leeming).

III. DISCUSSION

The Court begins its analysis by resolving certain preliminary matters. It first evaluates evidentiary disputes raised by the parties in motions in limine and during trial. It then considers LCRC's contention that the present action constitutes an appeal from an administrative agency decision, which is subject to an arbitrary and capricious standard of review.

Regarding the merits of the action, the Court first evaluates whether Gould has established the elements of a prima facie cost recovery action. Next, it considers whether LCRC is shielded from liability under the statutory third-party and contiguous landowner defenses. The Court then turns to whether the environmental harm is divisible such that the response costs may be apportioned between the parties. Finally, it evaluates the equitable factors supporting allocation of the response costs between the parties.

Ultimately, because the Court finds that Gould, Inc. is solely responsible for causing the contamination, it holds that 95% of the response costs must be equitably allocated to Gould. But because LCRC failed to exercise due care with respect to the contamination or to fully cooperate with EGLE, it is allocated 5% of the response costs.

A. Evidentiary Matters

As an initial matter, the parties dispute the admissibility of certain evidence relevant to the Court's analysis of LCRC's defenses and counterclaims. Specifically, the parties dispute the admissibility of testimony by expert witnesses Travers, Feenstra, and Browning, as well as certain reports and documents prepared by MSG and EGLE. For the reasons explained below, the Court admits Travers's and Feenstra's expert testimony. While Browning is qualified to testify as an expert, the Court finds that his testimony premised on the work of another expert is inadmissible. Similarly, MSG reports prepared in expectation of litigation by an expert who did not testify are inadmissible as business records of MSG. Finally, correspondence from EGLE is admissible as public records of the agency, while the portions of EGLE requests for geological review containing expert opinions are inadmissible.

1. Travers's Testimony

Gould filed a motion seeking to bar introduction of Travers's testimony in its entirety. See generally Pl. Mot. in Limine to Bar Expert Test. (Dkt. 92). In its motion, Gould contends that Travers's opinions regarding the source of the TCE contamination are irrelevant because the parties' causation of the contamination is not an element of the prima facie case for a cost recovery claim. Id. at 3-4. The balance of Gould's motion objects to Travers's testimony regarding the parties' use and disposal of TCE—which, Gould contends, is premised on unfounded speculation and mischaracterizations of other witnesses' testimony. Id. at 5-12.

With respect to Gould's argument that causation of the contamination is irrelevant, this contention plainly lacks merit. While it is true that causation is not part of Gould's prima facie case, LCRC's degree of causal contribution is relevant not only to its third-party defense but also to any equitable allocation of damages. Consequently, causation is an integral component of the liability analysis.

As to Gould's objection to Travers's testimony describing the parties' use and disposal of TCE, this argument does not warrant exclusion of her testimony. The Federal Rules of Evidence contemplate that experts may testify regarding the facts and data underlying their opinions. Fed. R. Evid. 705 (providing that an expert may state an opinion without first testifying to the facts or data upon which the opinion is premised, but also providing that "the expert may be required to disclose those facts or data on cross-examination"). Experts' testimony regarding underlying facts does not serve to establish those matters as substantive fact, but rather to explain how they formed their opinions. See In re John Richards Homes Bldg. Co., L.L.C., 439 F.3d 248, 264 (6th Cir. 2006) (noting that "expert testimony may not be used to establish underlying facts not otherwise in evidence," but holding that an expert may rely upon those facts in forming his opinion); United States v. Sowards, 339 F.2d 401, 402 (10th Cir. 1964) ("As a general rule, an expert may testify as to hearsay matters, not to establish substantive facts, but for the sole purpose of giving information upon which the witness relied in reaching his conclusion . . . .").

Accordingly, Travers's testimony regarding the parties' use and disposal of TCE does not serve to establish those facts, but rather to explain the basis for her opinions. Additionally, to the extent that the underlying facts described by Travers are unsupported or premised on mischaracterizations of other witnesses' testimony, Gould had the opportunity to cross-examine Travers regarding those deficiencies and to highlight any discrepancies in its post-trial briefing. See John Richards, 439 F.3d at 264. Consequently, Gould's motion to bar Travers's testimony is denied.

2. Feenstra's Testimony

For the first time in its post-trial briefing, LCRC challenges the admissibility of Feenstra's expert testimony under Federal Rule of Evidence 702 and Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 597 (1993), on the ground that it represents unsupported "junk science." Def. PFFCL at 26-30. LCRC's objection to Feenstra's testimony is not well-taken. LCRC's argument is untimely, as motions to limit or exclude expert testimony were to be filed by November 5, 2019. See Case Management & Scheduling Order at 1 (Dkt. 40). Nevertheless, the Court will consider LCRC's criticisms in evaluating the persuasiveness of Feenstra's opinions.

3. Browning's Testimony

During trial, Gould offered testimony from John Browning, the executive vice president of MSG and the lead engineer directing and overseeing MSG's investigations of the Gould and LCRC Properties. Trial Tr. I at 26, 31-32 (Browning) (Dkt. 258). Browning testified regarding the subsurface geology of the site and the contouring maps depicting the elevation of the intermediate clay layer and the pattern of groundwater flow. See Trial Tr. I at 122 (Browning); Trial Tr. II at 32-42, 46-50, 60-61, 67-70, 75 (Browning). LCRC objected during trial to recognizing Browning as an expert in the field of hydrogeology, claiming that he is not qualified in this field. Trial Tr. I at 123; Trial Tr. II at 75-77. Additionally, LCRC argued that Browning impermissibly parroted the hearsay opinions of non-testifying expert Thomas Cok, the lead MSG hydrogeologist involved in the investigation of the Gould and LCRC Properties.¹¹ See Trial Tr. I at 95-103, 105, 149-150 (Browning); Trial Tr. II at 33-38, 46-47 (Browning).

11 Cok was variously referred to as a hydrologist, see Trial Tr. I at 149 (Browning), and a hydrogeologist, see id. at 105; Trial Tr. II at 37 (Browning). Neither party presented evidence explaining whether there is a distinction between a hydrologist and a hydrogeologist. According to the Oxford English Dictionary, hydrology is "[t]he science which treats of water, its properties and laws, [and] its distribution over the earth's surface," while hydrogeology is "[t]hat part of geology which treats of the relations of water on or below the surface of the earth." See https://www.oed.com/view/Entry/90003?redirectedFrom=hydrology#eid; https://www.oed.com/view/Entry/89987?redirectedFrom=hydrogeology#eid.

The admissibility of expert testimony is governed by Federal Rule of Evidence 702. Under Rule 702, "[a] witness who is qualified as an expert by knowledge, skill, experience, training, or education may testify in the form of an opinion or otherwise" if the following criteria are met:

(a) the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product of reliable principles and methods; and
(d) the expert has reliably applied the principles and methods to the facts of the case.

Fed R. Evid. 702.

Rule 702 places a special obligation on the trial court to act as a gatekeeper, ensuring that "any and all scientific testimony or evidence admitted is not only relevant, but reliable." Daubert, 509 U.S. at 589. In Kumho Tire Company v. Carmichael, 526 U.S. 137, 147 (1999), the Supreme Court clarified that the "gatekeeping obligation" is not limited to "scientific" expert testimony; it applies to all expert testimony. Rule 702 requires a district court to satisfy itself that the proposed expert testimony will assist the trier of fact, before permitting the trier of fact to assess such testimony. Id. at 148-149. The proponent of the expert must establish admissibility by a preponderance of the evidence. Nelson v. Tenn. Gas Pipeline, Co., 243 F.3d 244, 251 (6th Cir. 2001).

Browning testified that he is a licensed engineer and has master's degrees in geotechnical engineering, hydrogeology, and geologic bedrock. Trial Tr. I at 27-28 (Browning). According to his curriculum vitae, Browning is experienced in applying principles of geotechnical engineering, geology, hydrogeology, and hydrology to the investigation, evaluation, planning, and design of projects related to soil, sediment groundwater, and surface water environmental issues. Browning Curriculum Vitae (Dkt. 199-18). As a general matter, there is no question that Browning is well qualified as an expert to testify regarding matters related to geotechnical engineering, geology, hydrogeology, and hydrology.

However, as noted by LCRC, Browning's testimony regarding the intermediate clay layer and groundwater flow was premised on documents prepared by and conclusions drawn by non- testifying expert Thomas Cok. See Trial Tr. I at 95-103, 149-150 (Browning); Trial Tr. II at 33-38, 46-47 (Browning). Browning admitted that Cok drafted all reports submitted to EGLE and principally prepared all contour maps. Trial Tr. I at 149-150 (Browning). He also testified that he personally reviewed and approved Cok's work, and as part of that process frequently provided comments and suggestions for modifications. Id. at 98-99, 149-150. The precise degree of Browning's involvement in preparing these documents, however, was not explained. His testimony does not make clear to what extent the documents represent his own views, those of Cok, or some combination of the two. And for reasons that Gould never explained, Cok was not called at trial.

Browning's testimony is flawed because Gould never established the reliability of any opinions he expressed regarding the intermediate clay layer, the contouring maps, or the pattern of groundwater flow. Browning did not testify that these opinions were premised on his own independent analysis. Instead, any opinions appeared to be based in a significant way on Cok's analysis. While Browning "approved" Cok's work, Gould did not establish that the conclusions reached in Cok's work meet the reliability requirements set forth under Rule 702. Browning did not describe in any significant way the facts or data upon which Cok's conclusions rested or testify that those conclusions were the product of reliable principles and methods appropriately applied by Cok. Nor did he analyze the discretionary choices made by Cok in preparing records and in reaching conclusions. Because no foundation explaining the precise basis for Cok's views was established, it cannot be said that they are sufficiently reliable.

By offering Browning as a mere conduit for Cok's opinions, Gould runs afoul of the rule that experts should not simply parrot the opinions of other experts. While Rule 703 permits an expert to "consider remote statements that . . . may be inadmissible, he cannot properly act as a conduit by presenting an opinion that is not his own opinion but that of someone else." 3 Christopher B. Mueller & Laird C. Kirkpatrick, Federal Evidence § 7:16 (4th ed. 2013); see also Louis Vuitton Malletier v. Dooney & Bourke, Inc., 525 F. Supp. 2d 558, 664 (S.D.N.Y. 2007) (holding that "the expert witness must in the end be giving his own opinion. . . . [and] cannot simply be a conduit for the opinion of an unproduced expert" (emphasis in original)). "Simply stated, an expert's opinion must be based upon his own application of principles within his expertise to the facts of the case." Quiles v. Bradford-WhiteCorp., No. 10-CV-747, 2012 WL 1355262, at *7 (N.D.N.Y. Apr. 8, 2012). While an expert may rely on facts and data known to others in expressing "what he thinks," he may not simply parrot "what someone else thinks." 3 Mueller & Kirkpatrick, Federal Evidence § 7:16.

In our case, the lack of a foundation establishing reliability is particularly troubling because, realistically, MSG was hired to support Gould's position in its dealings with EGLE and LCRC—raising the specter that in formulating expert opinions, "any margin of discretion . . . could be expected to be exploited" for Gould's benefit. See Dura Auto. Sys. of Ind., Inc. v. CTS Corp., 285 F.3d 609, 615 (7th Cir. 2002); see also Daubert v. Merrell Dow Pharms., Inc., 43 F.3d 1311, 1317 (9th Cir. 1995) ("One very significant fact to be considered is whether the experts are proposing to testify about matters growing naturally and directly out of research they have conducted independent of the litigation, or whether they have developed their opinions expressly for purposes of testifying.").

Accordingly, Browning's testimony premised on work prepared by Cok—including Cok's conclusions relative to the direction of groundwater flow and the elevation and continuity of the intermediate clay layer—must be excluded.¹² However, for the reasons articulated below, even if the Court were to consider this testimony, it would not consider the testimony to be persuasive.

12 LCRC did not lodge individual objections to Browning's testimony regarding each document that Cok prepared. LCRC objected to Browning's testimony regarding MSG's 2016 Annual Report (Dkt. 241), Trial Tr. I at 83-84, but otherwise advanced a broad continuing objection to Browning's testimony regarding interpretations of data developed by others, Trial Tr. I at 122-123; Trial Tr. II at 76.

4. MSG Reports

During trial, Gould sought to introduce through Browning's testimony certain reports created by MSG. Pl. PFFCL at 33.¹³ LCRC objects to the admission of these reports because they set forth expert opinions that were prepared by Cok and not by Browning. Def. PFFCL at 1, 3-4. Consequently, LCRC maintains that these records are inadmissible on two grounds: (i) they constitute inadmissible hearsay under Rules 801 and 802, and (ii) they are inadmissible under Rules 702 and 705 because Cok was not called by Gould and available for cross-examination regarding the reports' reliability. Id. The Court addresses each of LCRC's objections in turn.

13 Specifically, Gould seeks to admit the following documents: (i) MSG Interim Response Work Plan for Hydraulic Control dated April 13, 2013 (Dkt. 209); (ii) MSG Modified Comprehensive Groundwater Monitoring Plan and Pump & Treat System Update dated February 2019 (Dkts. 210, 210-1); (iii) Interim Response Work Plan for ISCO/EAB Pilot Study dated July 29, 2013 (Dkts. 211, 211-1); (iv) MSG 2016 Annual Report (Dkt. 241); and (v) MSG 2008 Supplemental Investigation and Historical Data Summary dated May 28, 2009 (Dkt. 245).

In general, the Rules of Evidence permit "the admission of expert opinion testimony[,] not opinions contained in documents prepared out of court." Engebretsen v. Fairchild Aircraft Corp., 21 F.3d 721, 728 (6th Cir. 1994) (emphasis in original). Stated simply, experts testify—not their reports. This is primarily because expert reports created prior to trial, such as those created by MSG, are considered hearsay and are inadmissible absent agreement to the contrary. Bianco v. Globus Med., Inc., 30 F. Supp. 3d 565, 570 (E.D. Tex. 2014). Caselaw has held an expert's report to constitute inadmissible hearsay even where another expert, who assisted in the report's preparation, testified regarding the conclusions reached by the principal expert, as "one expert may not give the opinion of another expert who does not testify." Tokio Marine & Fire Ins. Co., Ltd. v. Norfolk & W. Ry. Co., 172 F.3d 44 (Table) (4th Cir. 1999).

Gould, however, contends that the reports are admissible under Rule 803(6), as records of MSG's regularly conducted business activities. Pl. PFFCL at 33. Under Federal Rule of Evidence 803(6), records of a regularly conducted business or organizational activity are not excluded as hearsay if the following conditions are met: (i) the record was created at or near the time by—or from information transmitted by—someone with knowledge; (ii) the record was kept in the course of a regularly conducted activity of a business, organization, occupation, or calling, whether or not for profit; and (iii) making the record was a regular practice of that activity. These conditions must be shown by the testimony of "the custodian or another qualified witness, or by certification that complies with Rule 902(11) or (12) or with a statute permitting certification . . . ." Fed. R. Evid. 803(6)(D). Records meeting these requirements are generally admissible if "the opponent does not show that the source of information or the method or circumstances of preparation indicate a lack of trustworthiness." Fed. R. Evid. 803(6)(E).

Records prepared in anticipation of litigation typically do not meet the requirements of Rule 803(6). Jordan v. Binns, 712 F.3d 1123, 1135 (7th Cir. 2013); 5 Jack B. Weinstein, et al., Weinstein's Federal Evidence § 803.08 (2d ed. 2018). Courts exclude these records because preparation for litigation is generally not a routinely conducted business activity. Jordan, 712 F.3d at 1135. Additionally, "documents prepared with an eye toward litigation raise serious trustworthiness concerns because there is a strong incentive to deceive." Id.

Here, MSG's records were created not as part of an ongoing, routine activity but rather in connection with a specific project that, at its outset, implicated significant liability considerations and the threat of litigation. MSG was retained by Gould after EGLE issued a facility-notification letter identifying Gould as a potentially responsible party. In 2007, LCRC was identified as a second potentially responsible party that might be accountable for the contamination. Unsurprisingly, litigation regarding the parties' liability for the contamination was initiated in 2009 and has continued over the past decade. Under such circumstances, there can be no doubt that MSG's investigations and reports were drafted with a strong incentive to uncover or create evidence favorable to Gould. Because these reports were not prepared as a matter of routine practice and demonstrate a lack of trustworthiness, they do not fall under the Rule 803(6) exception to the rule against hearsay.

Additionally, as LCRC contends, even if the MSG reports qualified as business records under Rule 803(6), they are nevertheless inadmissible under Rules 702 and 705 because their preparer did not testify and was unavailable for cross-examination regarding his qualifications and the substance of the reports. In Forward Communications Corp. v. United States, 608 F.2d 485, 627 (Ct. Cl. 1979), the court held that expert reports determined to be business records under Rule 803(6) can nevertheless be excluded if they fail to satisfy Rules 702 and 705. That is, although business records containing expert opinions may be admissible if they "are incident to or part of factual reports of contemporaneous events or transactions[,] . . . . reports which are prepared to state or to support expert opinions are not admissible without the preparer being present in court to testify as to his qualifications as an expert and to be cross-examined on the substance, pursuant to Rules 702 and 705." Id. at 629. Thus, an appraisal report containing a valuation of the company was held to be inadmissible in the absence of supporting testimony from the consultant who prepared the report. Id. at 626, 629.

Here, the MSG reports go far beyond contemporaneously recording factual data and instead contain what appear to be principally Cok's opinions regarding possible sources of TCE, the direction of groundwater flow, and depictions of the subsurface geology. Because the MSG reports set forth expert opinions, they are not admissible under Rules 702 and 705 without Cok being present in court to testify as to his qualifications and to be cross-examined regarding the substance of the reports. Browning's testimony regarding the documents is insufficient to serve this purpose, as his familiarity with the underlying data and with the techniques and principles applied by Cok in rendering these opinions was not established. The Court, therefore, is unable to assess Cok's qualifications or the reliability of the methodologies applied in rendering the opinions set forth in the MSG reports, while LCRC was unable to cross-examine Cok on these matters. Consequently, the MSG reports are inadmissible under Rules 702 and 705.

Gould argues that Forward Communications can be distinguished from our case, as it was in Johnson v. State Farm Fire & Cas. Co., No. CIV.A. 12-00534-N, 2013 WL 4607548, at *5 (S.D. Ala. Aug. 29, 2013), where the court considered an expert report, despite a claim that it was hearsay, because the expert who prepared the report was available for cross-examination during a deposition. Pl. Resp. at 3-4 (Dkt. 255). Likewise, Cok was deposed in the present action. Id. However, Johnson involved a challenge to the evidence considered by the court at the summary judgment stage, with the court specifically noting that courts "may consider a hearsay statement in passing on a motion for summary judgment if the statement could be reduced to admissible evidence at trial . . . ." 2013 WL 4607548, at *5 (internal marks omitted). The MSG reports, by contrast, were offered during trial as opposed to the summary judgment stage—therefore, they were required to comply with Rules 702 and 705. As concluded above, they did not. Thus, LCRC's ability to cross-examine Cok during a deposition does not excuse Gould's failure to present his testimony at trial.¹⁴

14 Neither party offered the Cok deposition as a de bene esse deposition.

Consequently, the Court excludes these records. However, as with Browning's testimony, even if the Court were to consider these records, they are determined unpersuasive for the reasons articulated below.

5. EGLE Records

Gould seeks to admit documents and records prepared by EGLE under Federal Rule of Evidence 803(8), the public records exception to the rule against hearsay. Pl. PFFCL at 36-37. Specifically, Gould seeks to admit letters sent by EGLE to LCRC (Dkts. 201-7, 201-10, and 207-11) and requests for geological review (Dkts. 198-5, 201-8, and 213-4).¹⁵ Pl. PFFCL at 38-39. LCRC opposes the admission of these records on the grounds that they set forth opinions as opposed to factual findings, and that they are inherently untrustworthy insofar as they are premised on evidence and data prepared by the parties' consultants. Def. PFFCL at 9-10.

15 The Court admitted the portions of the requests for geological review that were drafted by Taylor, but reserved ruling regarding the portions of the documents that were drafted by EGLE geologists. See Trial Tr. II at 118, 128, 147.

A public record is excluded from the rule against hearsay if it is a record or statement of a public office that sets out one of the following: "(i) the office's activities; (ii) a matter observed while under a legal duty to report, but not including, in a criminal case, a matter observed by law-enforcement personnel; or (iii) in a civil case or against the government in a criminal case, factual findings from a legally authorized investigation . . . ." Fed. R. Evid. 803(8)(A). The burden then shifts to the opponent to show that the "source of information or other circumstances indicate a lack of trustworthiness." Fed. R. Evid. 803(8)(B); Fed. R. Evid. 803(8) advisory committee note (stating that the Rule "assumes admissibility in the first instance but with ample provision for escape if significant negative factors are present"). Generally, this hearsay exception is premised on the notion "that circumstantial guarantees of trustworthiness are provided by the presumption that governmental officials will perform their duties faithfully." Zenith Radio Corp. v. Matsushita Elec. Indus. Co., Ltd., 505 F. Supp. 1125, 1145 (E.D. Pa. 1980), rev'd in part on other grounds sub nom. In re Japanese Elec. Prods. Antitrust Litig., 723 F.2d 238 (3d Cir. 1983).

"[P]ortions of investigatory reports otherwise admissible under Rule 803(8)(C) are not inadmissible merely because they state a conclusion or opinion." Beech Aircraft Corp. v. Rainey, 488 U.S. 153, 168-69 (1988); see also United States v. Midwest Fireworks Mfg. Co., Inc., 248 F.3d 563, 566 (6th Cir. 2001) ("Opinions, conclusions, and evaluations, as well as facts, fall within the Rule 803(8)(C) exception[,] and enjoy a presumption of admissibility." (internal marks omitted)). Thus, LCRC's view that the Rule 803(8) hearsay exception does not encompass public records containing opinions lacks merit.

A "legally authorized investigation" need not be part of a formal inquiry leading to a hearing. See Baker v. Elcona Homes, Inc., 588 F.2d 551, 558 (6th Cir. 1978) ("We do not believe that a formal hearing is a [s]ine qua non of admissibility under Rule 803(8)(C) when other indicia of trustworthiness are present."). In United States v. Summit Equip. & Supplies, Inc., 805 F. Supp. 1422, 1430 (N.D. Ohio 1992), the court held that reports, created at the direction of the EPA, that detailed the extent of contamination and summarized the response actions taken at a site were admissible in a CERCLA action under Rules 803(6) and 803(8). See also Roberts v. Heating Specialist, Inc., No. 3:12-CV-01820-SI, 2013 WL 1814894, at *3 (D. Ore. Apr. 29, 2013) (admitting a summary of the EPA's investigations of contamination at a site and collecting cases in which courts have admitted affidavits, letters, and other less formal documents that summarize agency investigations and assert agency conclusions).

Although public records falling within the hearsay exception must generally be premised on the firsthand knowledge of the official creating the record, the exception "applies equally to documents that summarize first-hand knowledge of others who had a duty to report to the declarant." Roberts, 2013 WL 1814894, at *4 (admitting EPA official's summary of evidence gathered by other EPA employees and contractors); see also United States v. Central Gulf Lines, Inc., 974 F.2d 621, 626 (5th Cir. 1992) (admitting reports created by independent surveyors hired by the government pursuant to statutory requirements). This proposition is further supported by a case cited by LCRC, which holds that "Rule 803(8)(B) permits the introduction of a public record only if it is made from matters within the personal knowledge of (1) a public official making the report (or his agent) or (2) someone with a duty to report the matter to a public official." Wetherill v. Univ. of Chicago, 518 F. Supp. 1387, 1390 (N.D. Ill. 1981).

The letters prepared by EGLE fall within the public records exception to the rule against hearsay. They represent official communications from EGLE pertaining to the investigations on the Gould and LCRC Properties. They were drafted pursuant to EGLE's legal authority under NREPA and CERCLA to take all necessary action to remedy a release of hazardous substance. As noted above, the fact that they contain opinions is irrelevant. And to the extent that EGLE's observations and conclusions are premised on data and reports provided by Gould and LCRC, these parties were under legal obligations to report their findings to the agency. Additionally, "the fact that some of the sources of information in a report may be biased, does not, in and of itself, render the report untrustworthy so long as the public official who prepared it considered other sources of information and was not biased himself." In re Complaint of Munyan, 143 F.R.D. 560, 565 (D.N.J. 1992). The EGLE officials who prepared these letters reviewed materials from both parties, and there has been no indication that they did so in a biased manner. Thus, the letters are admissible in their entireties (Dkts. 201-7, 201-10, and 207-11).¹⁶

16 Appended to the letter dated December 10, 2013 is a request for geological review. The Court's ruling admitting the EGLE letters applies only to the correspondence portion of this exhibit, 12/10/13 Letter at PageID.67010-67011 (Dkt. 201-10); the request for geological review portion of the exhibit, 9/27/13 Request for Geological Review at PageID.67012-67017 (Dkt. 201-10), is governed by the Court's ruling below.

With respect to EGLE's internal requests for geological review (Dkts. 198-5, 201-8, and 213-4), the portions of these documents on which the Court reserved ruling were prepared by EGLE's geologists and set forth the geologists' expert opinions regarding the parties' investigations and the likely origins of the TCE contamination. For example, in a review prepared in February 2012, the geologist adopted Gould's views relative to the groundwater divide, concluding that the direction of groundwater flow did not support LCRC's view that the groundwater contamination on the LCRC Property emanated from the Gould Property. 2/12/12 Request for Geological Review at 6-7 (Dkt. 201-8). But in another review prepared in January 2017, the same geologist evaluated the soil testing performed on the LCRC Property and concluded that there was no indication of a second source area for TCE in groundwater on the LCRC Property to the south of the salt barn. 1/30/17 Request for Geological Review at 3 (Dkt. 213-4).

Like the EGLE letters, these documents are public records setting forth factual findings made in the course of EGLE's investigations authorized under CERCLA and NREPA. Nevertheless, they are inadmissible because other circumstances surrounding their preparation indicate a lack of trustworthiness. While parties offering expert opinion testimony generally must lay a foundation meeting the requirements of Rule 702, in the case of public records and reports, this foundation is subject to challenge "under the trustworthiness rubric." Matsushita, 505 F. Supp. at 1149; see also Melville v. Am. Home Assur. Co., 584 F.2d 1306, 1316 (3d Cir. 1978) ("Before these objections [under Rules 702 and 705] may be recognized . . . the party challenging the validity of an official report admitted under 803(8)(C) must come forward with some evidence which would impugn its trustworthiness."). In evaluating the trustworthiness of a public record, courts may evaluate various criteria, including, among other factors, the finality of the public agency's findings and conclusions, the agency's expertise, the basis of its opinion or findings, and its helpfulness to the jury. Matsushita, 505 F. Supp. at 1147, 1149.

Here, the opinions of the EGLE geologists are not sufficiently trustworthy, as they represent preliminary, nonfinal findings and conclusions. Where agency reports are rendered at an early stage of an investigation, such that they "are not only subject to extensive reconsideration, but are highly susceptible to modification or reversal, they cannot be deemed trustworthy." Id. at 1147. The danger of relying on preliminary findings is underscored by the two seemingly contradictory opinions reflected in the requests for geological reviews discussed above. In 2012, the geologist concluded that the direction of groundwater flow did not support LCRC's position that the contamination on the LCRC Property emanated from the Gould Property. 2/12/12 Request for Geological Review at 6-7. But in 2017 the same geologist concluded that there was no indication of a second source area to the south of the salt barn on the LCRC Property. 1/30/17 Request for Geological Review at 3. Further, as discussed above, EGLE ultimately concluded that there was "no indication that a release of chlorinated solvents to unsaturated site soils occurred, and no releases of chlorinated solvents in LCRC property site soils are demonstrated to be directly attributable to LCRC's historic operations." 6/14/17 Letter. The opinions reflected in the requests for geological review clearly evolved as additional data and information became available.¹⁷ Consequently, the Court is unable to conclude that the opinions expressed in the reports reliably represent the geologists' final impressions and interpretations of the data.

17 That the documents represent preliminary conclusions is further underscored by the fact that they were internal documents reflecting communications between the EGLE project manager and the geologist, as opposed to official communications shared with other parties. See Trial Tr. II at 113 (Taylor).

Moreover, the opinions expressed in these reports do not sufficiently describe the basis for some of the conclusions reached. For example, in the 2012 report, the EGLE geologist stated, with no elaboration, that groundwater flow at the southern end of the Gould source area is predominantly to the north and northeast. 2/12/12 Request for Geological Review at 6-7. It is unclear on what data the geologist relied in reaching this conclusion. As discussed in the sections that follow, the direction of groundwater flow in this area is the subject of detailed debate between Gould's and LCRC's experts. Because the basis for this and other portions of the geologists' opinions are not adequately described, the Court is unable to determine that these opinions are sufficiently trustworthy. Consequently, the portions of the requests for geological review completed by EGLE's geologists are inadmissible.¹⁸

18 Even if these documents were admissible, the Court would assign them no weight for lack of trustworthiness.

B. Final Administrative Action

Turning to the next preliminary matter, LCRC contests the very nature of the present action, seeking to style it as an appeal of an administrative agency action as opposed to a CERCLA cost recovery and contribution action. Def. PFFCL at 22-26. According to LCRC, the Gould and LCRC Properties were the subjects of extensive investigations that culminated in EGLE's determination that it had no further regulatory interest in the origins of TCE on the LCRC Property. Id. Citing Sackett v. Envtl. Prot. Agency, 566 U.S. 120, 127 (2012), LCRC contends that EGLE's decision represents a final agency action, subject to reversal only if a court finds that the agency's action was arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law. Def. PFFCL at 24-25.

LCRC failed to raise this argument at any point in the litigation until, on the eve of trial, it filed a motion seeking leave to file a second motion for summary judgment. See generally Mot. for Leave to File 2d Mot. for Summ. J. (Dkt. 164); Suppl. Br. (Dkt. 165). In this motion, LCRC argued that Sackett limited the Court's review to an arbitrary and capricious standard. Suppl. Br. at 8. The Court denied the motion on the ground that the argument was belatedly raised, noting that it was premised on authority that has been extant for some eight years and on letters issued by EGLE in 2017. 6/12/20 Order at 3 (Dkt. 197).

LCRC did not reference its theory of an agency appeal in its answers to Gould's complaint or in its counter-complaints. See generally Answer to Am. Compl. (Dkt. 36); Am. Answer (Dkt. 39); Answer to 2d Am. Compl. (Dkt. 58); Answer to 3d Am. Compl.; Counter-Compl. (Dkt. 59); Am. Counter-Compl. In its motion for summary judgment, LCRC argued that EGLE's determination should be afforded deference, in accordance with Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837 (1984). See Gould Elecs., Inc. v. Livingston Cty. Rd. Comm'n, No. 17-11130, 2020 WL 806033, at *5 (E.D. Mich. Feb. 18, 2020). But LCRC did not argue in this motion that the litigation should be construed as an appeal from an agency action. Finally, the parties' joint final pretrial order is devoid of any contention regarding the appropriate standard of review. See Revised Joint Final Pretrial Order (Dkt. 244). The Court's local rules provide that the joint final pretrial order "supersed[es] the pleadings and govern[s] the course of trial unless modified by further order." E.D. Mich. LR 16.2(a). Because of its many failures to raise this issue on a timely basis, LCRC has waived this argument.

Even if LCRC had not waived this argument, it plainly lacks merit. This action was initiated as a CERCLA cost recovery action and not as an appeal of a final agency action under the Administrative Procedures Act, as was the case in Sackett. See 566 U.S. at 122. Nor do the June 2017 Letters constitute a final agency action. A final agency action must represent the "consummation" of the agency's decision-making process and must constitute a determination of rights or obligations "from which legal consequences will flow." Bennett v. Spear, 520 U.S. 154, 177-178 (1997).

No legal consequences flow from EGLE's letters with respect to LCRC's liability for the TCE contamination, as the June 14, 2017 letter expressly states that it "should not be construed as a liability determination for the chlorinated solvent contamination on the subject property." 6/14/17 Letter; see also Trial Tr. V at 15 (Leeming) ("It was supposed to let [LCRC] know that the State had no further regulatory interest in the site with respect to the TCE and also that we were not making a determination of liability for the TCE at the site."). EGLE made no factual findings and no determination regarding Gould's or LCRC's liability. Further, EGLE's determination that it required no further investigation by LCRC may have been motivated by the fact that remedial measures fully addressing the contamination were already underway. Therefore, the Court rejects LCRC's view that the Court's review is limited to a finding as to whether EGLE's determination was arbitrary and capricious.

C. Gould's CERCLA Cost Recovery Claim

The Court next turns to the merits of Gould's cost recovery claim. The Court first evaluates whether Gould has established the elements of a prima facie case. Second, the Court considers whether LCRC has established the statutory defenses asserted—the third-party defense and the contiguous landowner defense. Finally, the Court discusses the application of the divisibility doctrine and whether the response costs may be apportioned between the parties.

1. Prima Facie Case

CERCLA was enacted to serve two primary purposes: (i) to promote the prompt cleanup of hazardous waste sites and (ii) to place the cost of that cleanup on those responsible for creating or maintaining the hazardous condition. W.R. Grace & Co.-Conn. v. Zotos Int'l, Inc., 559 F.3d 85, 88 (2d Cir. 2009). Thus, "CERCLA is designed to encourage private parties to assume the financial responsibility of cleanup by allowing them to seek recovery from others." Key Tronic Corp. v. United States, 511 U.S. 809, 819 n.13 (1994). By requiring responsible parties to pay for cleanup efforts, CERCLA additionally ensures "that 'the taxpayers [are] not required to shoulder the financial burden of a nationwide cleanup.'" Lockheed Martin Corp. v. United States, 35 F. Supp. 3d 92, 122 (D.D.C. 2014) (quoting B.F. Goodrich Co. v. Murtha, 958 F.2d 1192, 1198 (2d Cir. 1992)).

In furtherance of these goals, § 107(a) of CERCLA creates a cause of action permitting a private party to recover from four categories of potentially responsible parties ("PRPs") response costs voluntarily incurred to clean up a contaminated site. Consol. Edison Co. of New York v. UGI Utilities, Inc., 423 F.3d 90, 97 (2d Cir. 2005). This remedy is available "to any person that has incurred necessary costs of response, and nowhere does the plain language of section 107(a) require that the party seeking necessary costs of response be innocent of wrongdoing." Id. at 100; see also United States v. Atl. Research Corp., 551 U.S. 128, 140 (2007).

To establish a prima facie case for cost recovery under CERCLA § 107(a), a plaintiff must demonstrate the following four elements:

(1) a polluting site is a "facility" within the statute's definition; (2) the facility released or threatened to release a hazardous substance; (3) the release caused the plaintiff to incur necessary costs of response; and (4) the defendant falls within one of four categories of potentially responsible parties.

Gould Elecs., Inc. v. Livingston Cty. Rd. Comm'n, No. 09-cv-12633, 2012 WL 5817937, at *7 (E.D. Mich. May 25, 2012) (quoting Village of Milford v. K-H Holding Corp., 390 F.3d 926, 933 (6th Cir. 2004)). Causation of the contamination is not an element of a prima facie case, United States v. Puerto Rico Indus. Dev. Co., 287 F. Supp. 3d 133, 144 (D.P.R. 2017), and "[l]iability under § 107(a) is generally joint and several on any defendant regardless of fault," Kalamazoo River Study Group v. Menasha Corp., 228 F.3d 648, 653 (6th Cir. 2000). Here, LCRC does not dispute that Gould has established the first and fourth elements. See Revised Joint Final Pretrial Order at 2-4, 12-13; Answer to 3d Am. Compl. ¶¶ 22, 28-29 (admitting that LCRC owns property that has been designated a facility).

With respect to the second element of a prima facie case, Gould has met its prima facie burden of establishing that a release has occurred on LCRC's property. As the Court has previously recognized, the term "release" is defined under CERCLA to mean "'any spilling, leaking, pumping, pouring, emitting, emptying, discharging, injecting, escaping, leaching, dumping, or disposing into the environment." Gould, 2012 WL 5817937, at *8 (quoting 42 U.S.C. § 9601(22)). Courts uniformly interpret the term broadly so as to not frustrate the legislative purposes of CERCLA, In re Joshua Hill, Inc., 294 F.3d 482, 489 (3d Cir. 2002) (collecting cases interpreting the term "release" broadly), such that the passive leaching of hazardous substances nevertheless constitutes a "release," Pakootas v. Teck Cominco Metals, Ltd., 452 F.3d 1066, 1075 (9th Cir. 2006) ("We hold that the leaching of hazardous substances from the slag at the Site is a CERCLA release.").

It is undisputed that TCE has come to be located in the soil and groundwater beneath the LCRC Property. See Answer to 3d Am. Compl. ¶ 13 ("Admits that TCE exists on the LCRC Property."). The record establishes that a plume of groundwater contaminated with TCE is migrating from both the Gould and LCRC Properties northward toward Thompson Lake. Trial Tr. V at 187-188 (Travers); Trial Tr. II at 89, 99 (Browning). Likewise, Feenstra admits that some of the contaminated groundwater from the northeastern corner of the Gould Property likely migrated onto the far northwestern corner of the LCRC Property. Trial Tr. VI at 160 (Feenstra) (Dkt. 263). Trial testimony from Browning also demonstrates that a "halo" of residual TCE soil contamination extends from the Gould Property onto the northwestern corner of the LCRC Property. Trial Tr. II at 45 (Browning). This evidence illustrates that a release of TCE has occurred on the LCRC Property because the presence and migration of TCE in the soils and groundwater under the LCRC Property indicates a "leaching" of that substance.

With respect to the third element of a prima facie case, Gould has also met its prima facie burden of showing that the release caused it to incur necessary costs of response. The Sixth Circuit has held that incorporated within this element are the requirements that "the response costs were necessary and consistent with the National Contingency Plan ('NCP')." Bob's Beverage, Inc. v. Acme, Inc., 264 F.3d 692, 696 (6th Cir. 2001)); see also Reg'l Airport Auth. of Louisville v. LFG, LLC, 460 F.3d 697, 709 (6th Cir. 2006) ("[P]roof of response costs incurred 'consistent with' the NCP is . . . an element of the prima facie private cost recovery action under CERCLA."). A response action is consistent with the NCP if the action is in substantial compliance with 40 C.F.R. § 300.700(c)(5)-(6), and results in a CERCLA-quality cleanup. ITT Indus., Inc. v. Borgwarner, Inc., 700 F. Supp. 2d 848, 880 (W.D. Mich. 2010) (citing 40 C.F.R. § 300.700(c)(3)(i)). An immaterial or insubstantial deviation, however, will not result in a determination that a cleanup is inconsistent with the NCP. Id. (citing 40 C.F.R. § 300.700(c)(4)). The NCP includes a tangled web of technical requirements relating to a party's response activities. See 40 C.F.R. § 300.700(5)-(6).

Gould asserts that the release on the LCRC Property has caused it to incur necessary response costs, including recoverable legal expenses, through September 20, 2019. Pl. PFFCL at 14. Invoices and testimony from John Callahan, Gould's chief administrative officer, establish that Gould has incurred $3,949,179 in environmental consulting fees, of which Gould claims $3,898,754 is recoverable. Trial Tr. III at 73-81 (Callahan); Consultant Summary & Invoices (Dkts. 207-5, 207-6). Invoices and testimony from Stephen Ormond, Gould's counsel, establish that Gould has incurred $354,543 in fees for legal work performed to gain access to neighboring properties to perform testing. Trial Tr. III at 104-122 (Ormond); Legal Work Summary & Invoices (Dkts. 207-7, 207-8); MNB Settlement Agreement (Dkt. 207-9). Accordingly, Gould has incurred a total of $4,253,297 in recoverable response costs through September 20, 2019.

These costs were incurred in connection with Gould's investigation and remediation efforts, which included investigations to determine the sources and extent of the TCE contamination, an excavation to remove soil, development and installation of a pump and treat system, and development and implementation of a bioremediation plan. Trial Tr. I at 128, 135-136, 157-158, 161, 171-172 (Browning). According to Browning and Taylor, the pump and treat system addresses the plume of contamination stemming from both the Gould and LCRC Properties. Id. at 130-131; Trial Tr. II at 148 (Taylor). Likewise, the bioremediation plan treats the entirety of the plume emanating from both the Gould and LCRC Properties, with a focus on the areas with the highest TCE concentrations. Trial Tr. I at 136-137 (Browning). Accordingly, Gould has established that it has incurred response costs as a result of the release on the LCRC Property, as well as the Gould Property.

Feenstra opined that Gould's response activities have been both reasonable and necessary. Trial Tr. II at 216 (Feenstra). In rendering this opinion, Feenstra evaluated the work performed as summarized in reports prepared by MSG, with a particular focus on the timing of its investigations, the scope of the different phases of work, and the types of techniques used to perform the work. Id. at 217-219.¹⁹ Based on his evaluation, Feenstra stated that the sequence of work performed by MSG was logical, appropriate, and consistent with what Feenstra has observed at similar sites. Trial Tr. III at 48-50 (Feenstra). According to Feenstra, a pump and treat system is a proven technology that has been used for many years, but it may only be implemented after the scope of the contamination has been fully delineated. Id. at 50. Because Gould did so, its installation of the pump and treat system to limit the migration of the contamination was appropriate. Id. LCRC does not rebut this testimony or otherwise challenge Feenstra's position that Gould's response activities were necessary. Based on this testimony, Gould has met its burden of demonstrating that its response activities were necessary.

19 In arriving at his opinion, Feenstra relied on MSG reports that this Court has determined to be inadmissible, including the 2008 Supplemental Investigation and Historical Data Summary. See Attachment B to Feenstra Report at PageID.77005-77011 (Dkt. 208-8) (listing sources consulted). However, these reports need not be admitted for the purpose of establishing what work took place, as Browning described the major phases of work undertaken by MSG. See Trial Tr. I at 87-108, 127-132, 157-172 (Browning). According to Feenstra, Browning's testimony describing the work was entirely consistent with Feenstra's summary of the work premised on the MSG reports. Trial Tr. III at 31-32 (Feenstra); see also Table 2 to Feenstra Report at PageID.76934-76935 (Dkt. 208-8) (summarizing the phases of work). Moreover, LCRC did not object to Feenstra's reliance on the MSG reports in rendering his opinion on the reasonableness of the work that took place.

As LCRC observes, Gould did not specifically address in its post-trial briefing whether its response activities were consistent with the NCP. See Def. Resp. at 6 (Dkt. 257). However, the evidence presented at trial is sufficient to support a finding of compliance with the NCP. Although Feenstra did not specifically address compliance with the NCP, he indicated that MSG's work was performed "quite well" and was "logical and appropriate." Trial Tr. II at 216 (Feenstra). Further, MSG employed industry-standard techniques and practices to accomplish its investigation and development of the appropriate remedial technologies. Id. Feenstra noted that the work was undertaken largely to satisfy state regulatory requirements, as well as some specific requests from EGLE. Id. MSG documented its activities and "kept the state regulator satisfied with regard to progress in dealing with the problem." Id.

EGLE reviewed and approved MSG's work plans for its investigations, the 2001 soil excavation in the northeastern corner of the Gould Property, the installation of the pump and treat system, and the implementation of the bioremediation plan. Trial Tr. II at 93, 99-106 (Taylor); Trial Tr. I at 80, 89, 128, 135-136, 171-172 (Browning). Following a public meeting during which Gould explained to local residents the work being undertaken relative to the pump and treat system, EGLE issued a permit authorizing Gould to treat the groundwater and to pump the treated groundwater back into Thompson Lake. Trial Tr. II at 102 (Taylor); MDEQ Permit Authorizing Discharge Under National Pollutant Discharge Elimination System (Dkt. 208-15). Browning described the process of obtaining approval of work plans from EGLE, stating that EGLE frequently offered comments, suggested revisions, and requested additional work. Trial Tr. I at 80-81, 90, 106 (Browning). Further, Browning stated that, as the lead engineer on the project, he was satisfied that MSG followed all protocols required by EGLE. Trial Tr. II at 62 (Browning).

In light of EGLE's close involvement in monitoring and approving the investigation and remedial work undertaken on behalf of Gould, as well as Feenstra's statements that all work was completed in accordance with industry standards, the Court finds that Gould has established by a preponderance of the evidence that its response activities were consistent with the NCP. Indeed, some courts have held that where, as here, a state environmental agency approves cleanup plans and monitors the remediation process, the NCP consistency requirement is satisfied. See NutraSweet Co. v. X-L Eng'g Co., 227 F.3d 776, 791 (7th Cir. 2000); Pfohl Bros. Landfill Site Steering Committee v. Browning-Ferris Indus. of New York, Inc., No. 95-CV-956A, 2004 WL 941816, at *22 (W.D.N.Y. Jan. 30, 2004). Moreover, LCRC has failed to identify any specific manner in which Gould's response activities were inconsistent with the NCP.

Thus, Gould has established all elements of a prima facie case for cost recovery under CERCLA § 107(a).

2. Statutory Defenses

Once a plaintiff establishes a prima facie case, the defendant is strictly liable for the presence of hazardous substances unless it successfully invokes one of the statutory affirmative defenses. Prisco v. A & D Carting Corp., 168 F.3d 593, 603 (2d Cir. 1999). "CERCLA's affirmative defenses shift the burden of proof on this question from the plaintiff to the defendant, who must show by a preponderance of the evidence that the release or threatened release was caused solely by an unrelated third party." Id. LCRC asserts that it is shielded from liability on Gould's cost recovery claim under two statutory affirmative defenses, the third-party defense under CERCLA § 107(b)(3), and the contiguous landowner defense under CERCLA § 107(q). Def. PFFCL at 31-36.

a. Third-Party Defense

Under the third-party defense, CERCLA § 107(b)(3), an otherwise liable defendant is exempt from CERCLA cost recovery liability if it can establish by a preponderance of the evidence "(1) that another party was the sole cause of the release of hazardous substances and the damages caused thereby; (2) that the other, responsible party did not cause the release in connection with a contractual, employment, or agency relationship with the defendant; and (3) that the defendant exercised due care [with respect to the hazardous substance concerned] and guarded against the foreseeable acts or omissions of the responsible party." PCS Nitrogen Inc. v. Ashley II of Charleston LLC, 714 F.3d 161, 179 (4th Cir. 2013) (internal marks omitted). This defense was intended to be "very narrowly applicable, for fear that it might be subject to abuse." Carson Harbor Vill., Ltd. v. Unocal Corp., 270 F.3d 863, 883 (9th Cir. 2001) (en banc).

Here, the second element is not at issue; the parties only dispute LCRC's ability to establish the first and third elements. Pl. Resp. at 10; Def. PFFCL at 31-33. In determining whether LCRC is protected from liability under the third-party defense, the Court begins by evaluating whether the TCE contamination was caused solely by the acts or omissions of a party other than LCRC. After considering the historical operations at the Gould and LCRC Properties, as well as the experts' competing theories of causation, the Court concludes that the TCE contamination is attributable entirely to releases of TCE that took place on the Gould Property and migrated onto neighboring properties, including the LCRC Property. Next, the Court evaluates whether LCRC exercised due care. Because LCRC was dilatory in beginning its investigations and was recalcitrant in complying with EGLE's requests for certain testing, the Court finds that it did not act with due care. Consequently, the third-party defense is unavailable to LCRC.

i. Sole Cause

Relying on evidence regarding historical operations at the Gould and LCRC Properties, as well as scientific data and opinions from Travers, LCRC contends that it never disposed of TCE on the LCRC Property and that Gould is solely responsible for the TCE contamination at issue. Def. PFFCL at 15-22, 31. Gould, in turn, contends that LCRC has not come forward with any evidence demonstrating that employees of Gould Inc. used TCE, let alone improperly disposed of TCE by pouring it onto the ground. Pl. Resp. at 5-6. Gould further contends that the scientific data and opinions from its own expert witness, Feenstra, support the conclusion that LCRC contributed to the contamination by disposing of TCE on its own property. Pl. PFFCL at 5-12.

a) Historical Operations—RSF Facility

LCRC maintains that the TCE contamination is solely attributable to Gould Inc.'s longstanding practice of disposing of chemicals by pouring them onto the ground at the Gould Property or by permitting them to flow into a floor drain within the RSF Facility. Def. PFFCL at 15-17. Former Gould Inc. employees Keith Richardson and Ron Galarneau testified that they disposed of waste chemicals by pouring them onto the ground of the Gould Property. Trial Tr. V at 69-74 (Richardson); Galarneau Dep. at 31, 41-43, 77-78 (Dkt. 166-4). Richardson, who was employed by Gould Inc. from 1972 through its closing in 1976, testified that he handled coolant fluid that was composed of kerosene, cutting oil, and "sweet oil." Trial Tr. V at 69, 73 (Richardson). He admitted that on a monthly basis, he disposed of approximately forty to forty-five gallons of these fluids by dumping them onto the ground to the south of the foundry—as directed by a supervisor. Id. at 70-73. Richardson also stated that he observed other employees dispose of fluids in the same manner. Id. at 73.

Galarneau, who worked at the RSF Facility from the mid-1960s through 1976, similarly admitted that he and other employees disposed of fluids by pouring them on the ground at the northeastern corner of the Gould Property. Galarneau Dep. at 31. He stated that employees also disposed of fluids on the ground to the south of the foundry, near the fence along the eastern boundary of the Gould Property, and near the railroad tracks to the north of the Gould Property. Id. at 41-43. Galarneau was unfamiliar with TCE, stating that he did not know the names of chemicals he used in his work or those used at the RSF Facility generally. Id. at 47, 63. However, he estimated that over the course of Gould Inc.'s operations at the RSF Facility, thousands of gallons of waste chemicals were collectively disposed of on the ground or into the floor drain—however, he did not further refine what quantities were disposed of in each manner. Id. at 61.

Richardson and Galarneau also testified that parts manufactured at the RSF Facility were degreased in a large tank located at the eastern end of the facility. Trial Tr. V at 64-65 (Richardson); Galarneau Dep. at 20-22. These parts were placed in baskets and submerged in the degreasing tank, after which they were raised and left to drip-dry over the tank. Trial Tr. V at 65-66 (Richardson); Galarneau Dep. at 24-25. The parts were then placed on racks, where they continued to drip-dry over the floor. Galarneau Dep. at 24-25. According to Galarneau, a spigot located near the bottom of the degreasing tank was used to drain the tank onto the floor, and fluids that collected on the floor would flow into a floor drain. Id. at 26-27. Richardson, however, had no recollection of whether the degreasing tank had a spigot. Trial Tr. V at 74 (Richardson). It is unclear from the evidence where the floor drains emptied, as will be discussed in greater detail below.

Galarneau stated that some of the fluids from the degreasing tank "could" have been emptied into the floor drain, while some "could" have been hauled away. Galarneau Dep. at 30. He never witnessed fluids from the degreasing tank being hauled away for disposal, but stated that a truck delivered fluids on a weekly basis. Id. at 30. Neither Richardson nor Galarneau knew what type of solvents were used in the degreasing tank. Trial Tr. V at 65 (Richardson); Galarneau Dep. at 33, 63. But based on his understanding of industrial processes during the time period in question, as well as his review of testimony from former employees of Gould Inc., Feenstra stated the degreasing tank was likely a vapor degreaser that used TCE. Trial Tr. VII at 8 (Feenstra) (Dkt. 264).

b) Floor Drain & PVC Pipe

According to LCRC, the floor drain at the RSF Facility was illegally connected to a PVC pipe that exited the facility and drained into a storm sewer on the LCRC Property, near the boundary line between the Gould and LCRC Properties. Def. PFFCL at 17. In support of its position, LCRC cites a series of photographs taken by Taylor during the 2001 soil excavation at the northeastern corner of the Gould Property. Id.

The first photograph depicts an interior view of the eastern wall of the RSF Facility, where a PVC pipe exits the building. Floor Drain Photo at PageID.34593 (Dkt. 169-4); Trial Tr. II at 176-177 (Taylor). The photograph is annotated with Taylor's observation that the PVC pipe originates from a "cleanout" for a suspected floor drain. Floor Drain Photo. A second photograph depicts an exterior view of the same PVC pipe exiting the eastern side of the RSF Facility and is annotated with Taylor's observation that the pipe was "illegally connected to [the] storm sewer." Photo No. 17 at PageID.34578 (Dkt. 169-3); Trial Tr. II at 172-173 (Taylor). In the photograph, the end of the pipe is broken off. Photo No. 17.

A third photograph shows a separate, broken-off PVC pipe buried near a fence separating the Gould and LCRC Properties. Photo No. 18 at PageID.34578 (Dkt. 169-3). This second PVC pipe was located across the excavation site from the pipe exiting the RSF Facility. Id.; Trial Tr. II at 174 (Taylor). Taylor noted on the photograph that this second PVC pipe was connected to a storm sewer located on the LCRC Property, near the boundary line between the Gould and LCRC Properties. Photo No. 18. Water from that storm sewer flowed to a catch basin beneath the LCRC Property and fed into a storm sewer line that emptied into Thompson Lake. Trial Tr. II at 177 (Taylor). In a fourth photograph of the storm sewer, Taylor noted that the PVC pipe that emptied into the storm sewer originated from the Gould Property and connected to a "suspected floor drain." Storm Sewer Photo at PageID.34595 (Dkt. 169-4); Trial Tr. II at 178-179 (Taylor).

The evidence, however, does not sufficiently establish that the PVC pipe was connected to the floor drain in the RSF Facility. While LCRC implies that the PVC pipe that exited the RSF Facility was connected to the PVC pipe that emptied into the storm sewer, Taylor could not definitively state that the two pipes were illegally connected in this manner, as portions of the pipes had been removed from the excavation site before she arrived. Trial Tr. II at 172-174. Moreover, Taylor testified that the purpose of the PVC pipe exiting the RSF Facility was unclear and posited that it could have been connected to a roof drain. Id. at 174-175. LCRC's conclusion that hazardous materials flowed from the floor drain inside the RSF Facility into the catch basin is, therefore, unsupported by concrete evidence. And as argued by Gould, it is uncertain whether the PVC pipe was installed when Gould Inc. operated the RSF Facility between 1961 and 1976, or whether it was installed after that time. See Pl. Resp. at 8. Accordingly, the Court rejects LCRC's theory that the floor drains in the RSF Facility fed into the storm sewer via the PCV pipe.

In sum, however, the evidence demonstrates that Gould Inc. had a regular practice of disposing of vast amounts of chemical waste by dumping it on the ground of the Gould Property, particularly in the northeastern corner of the Gould Property and along the fence near the boundary of the Gould and LCRC Properties. Additionally, as discussed below, the evidence establishes that the chemicals, more likely than not, included TCE.

c) Historical Operations—LCRC Property

In contrast to Gould Inc.'s operations, LCRC maintains that there is no evidence of improper disposal practices at the LCRC Property. Def. PFFCL at 20-22. LCRC's historical records reveal that LCRC's onsite use of TCE was limited to a maximum of seventeen asphalt tests performed between 1985 and 1986. Trial Tr. III at 178-179 (Craine). Craine inventoried and created a summary of LCRC's asphalt testing records for the years 1959 through 1993, which demonstrates that of the 312 total tests, a maximum of only seventeen tests could have been performed onsite; those seventeen tests took place between 1985 and 1986. Id. at 181-182; Summary of Asphalt Testing Records at PageID.32228-32230 (Dkt. 168).²⁰ The records for these seventeen tests do not indicate the location where the test was performed. Trial Tr. III at 178-179 (Craine). Travers independently reviewed these records and likewise determined that TCE was used in seventeen tests performed on the LCRC Property between 1985 and 1986. Trial Tr. V at 151 (Travers). Apart from this asphalt testing, Craine was unable to identify any other uses of TCE in LCRC's operations. Trial Tr. IV at 9 (Craine).

20 Although the Court reserved ruling on Gould's objection to this exhibit during trial, Gould did not discuss this exhibit in its PFFCL. Accordingly, Gould has waived this argument. See 7/21/20 Order for Post-Trial Submissions (Dkt. 248) ("Failure to address any evidentiary matter will be deemed a waiver of that party's position regarding that issue."). This exhibit is admitted under Rule 1006 as a summary of voluminous writings. LCRC offered for the Court's review the underlying records dating from 1959 through 1993, see Asphalt Records at PageID.32231-32547 (Dkt. 168), which are independently admissible under the hearsay exceptions for business records under Rule 803(6) and for ancient documents predating January 1, 1998, under Rule 803(16).

Asphalt testing was performed on the second floor of LCRC's main office building, Trial Tr. III at 158, 168-169 (Craine); Trial Tr. V at 54 (Hogan), which spans along the eastern boundary of the LCRC Property, see Figure 1 to Travers Report. Two former LCRC employees, Fred Marr and Eric Little, testified that TCE was used to perform asphalt tests on the LCRC Property for two years in the mid to late 1980s. Trial Tr. V at 23, 27 (Marr); Little Dep. at 11 (Dkt. 166-6). Marr, who was employed by LCRC from 1971 through 2014, estimated that he performed approximately twenty-five tests per year and that a single test required a quart of TCE. Trial Tr. V at 28 (Marr). Waste TCE would be collected in a five-gallon can, which was later poured into a large drum. Id. at 29; Little Dep. at 13-14. According to Little, who was employed by LCRC from 1978 through the date of his deposition in 2011, the drum was collected by Safety Kleen for disposal when it became full. Little Dep. at 10, 14. However, he had no knowledge regarding the ultimate fate of the recovery drum contents. Id. at 13-14. Marr denied that TCE was ever disposed of by pouring it onto the ground. Trial Tr. V at 29-30 (Marr). Although LCRC initially used TCE to conduct these tests, it began using a different solvent called Bio Act in the mid-1980s. Id. at 28; Little Dep. at 12. Francis Hogan, an LCRC engineering aide, confirmed that LCRC used Bio Act when he began performing asphalt testing in 1989. Trial Tr. V at 54 (Hogan).

No evidence was presented demonstrating that LCRC used TCE for any other purpose on the LCRC Property, such as for vehicle maintenance. George Messner, a former LCRC employee responsible for vehicle maintenance during the period from 1972 through 2012, denied using TCE as a solvent. Trial Tr. IV at 173-174 (Messner). In 1972, LCRC used diesel fuel to wash vehicle parts, and in 1976 or 1977, LCRC began using a parts washer provided by LCRC's vendor, Safety Kleen. Id. Billy Ross, the vice president of environmental health and safety at Safety Kleen, testified that by the year 1980, parts washers primarily used mineral spirits or petroleum naptha but did not use TCE. Ross Dep. at 18-22 (Dkt. 166-20).²¹ Messner never instructed or witnessed his employees disposing of chemical waste by pouring it on the ground. Trial Tr. IV at 175 (Messner).

21 Ross began his employment with Safety Kleen in approximately 1990 but was able to testify, based on his knowledge of the company's history, regarding the composition of parts washers from 1980 to the present. Ross Dep. at 6-7, 15-20.

Gould quibbles with the exact number of asphalt tests performed on the LCRC Property, arguing that Marr testified that he performed approximately twenty-five asphalt tests per year over the course of two years. Pl. Resp. at 15. Moreover, Gould highlights Little's testimony that LCRC directed Safety Kleen to deliver larger quantities of TCE because five-gallon containers were insufficient. Pl. PFFCL at 4 (citing Little Dep. at 10, 16). Even crediting Marr's estimate of the number of tests performed per year, LCRC would have performed a maximum of only fifty asphalt tests over the course of two years. Given that each test required approximately a quart of TCE, LCRC used a total of twelve to thirteen gallons of TCE to perform these tests. And although Little testified that LCRC began purchasing larger quantities of TCE from Safety Kleen, Gould has not demonstrated that these quantities were actually used or that LCRC improperly disposed of them.

Gould also disputes whether Safety-Kleen collected waste TCE for disposal. Pl. PFFCL at 4. First, Gould points out that neither Marr nor Little had knowledge regarding the ultimate disposal of the waste TCE after Safety Kleen collected it. This fact is immaterial, however, given that Marr and Little did not work for Safety Kleen and naturally would not have personal knowledge of its disposal practices.

Second, Gould highlights Ross's testimony that Safety Kleen's collection of waste TCE from LCRC would have been reflected in service documents, but that he came across no such documents in reviewing Safety Kleen's available records. Ross Dep. at 53-54. Based on his review of the available records, Ross testified that Safety Kleen did not provide disposal services for LCRC. Id. at 54. However, it appears that these records are incomplete. Ross admitted that the available records were limited to the years 1984 and 1985, id., but Messner testified that LCRC began using Safety Kleen as a vendor as early as 1976 or 1977, Trial Tr. IV at 173-174 (Messner). And Little testified that Safety Kleen collected LCRC's waste TCE, Little Dep. at 10, 14, which the records reveal would have been generated in 1985 and 1986. Although Ross stated that Safety Kleen's records were not maintained in any other locations, id., he did not offer testimony ruling out the possibility that those records were simply missing or otherwise unavailable. Therefore, Ross's conclusion that Safety Kleen did not provide disposal services to LCRC was premised on his review of incomplete records and, therefore, carries little weight. Ultimately, Gould has presented no evidence convincingly showing that TCE was dumped on the LCRC Property.

d) The Line of Occupation

Gould argues that from 1950 through the 1970s, LCRC's operations extended to a "line of occupation" approximately twenty feet west of the property line shared with the Gould Property, including a dirt roadway through the northeastern corner of the Gould Property. Pl. PFFCL at 3. Gould, therefore, claims that LCRC was operating on the area where TCE contamination was later discovered on the Gould Property, leaving open a question as to which party is responsible for the contamination on the Gould Property. Id. at 29.

In support of its position, Gould relies on Browning's interpretation of a series of historic aerial photographs dating from 1950, 1955, 1966, 1967, and 1970. See 1950 Aerial Photo (Dkt. 217-3); 1955 Aerial Photo (Dkt. 215-11); 1966 Aerial Photo (Dkt. 215-12); 1967 Aerial Photo (Dkt. 215-13); 1970 Aerial Photo (Dkt. 215-14). Based on his visual analysis of the photographs, Browning testified that there was a visual distinction between the operations on the Gould and LCRC Properties, and that LCRC's operations extended west of the property line to the line of occupation. Trial Tr. I at 35-36, 56, 62, 64, 66 (Browning). Additionally, a drawing prepared in connection with a 1966 survey by R.S. Scott Associates, denoted a "fence in place" west of the property line between the Gould and LCRC Properties. 1966 Arrow Head Products Survey at PageID.80255 (Dkt. 215-4). Browning stated that the fence was located approximately twenty feet west of the property line between the Gould and LCRC Properties. Trial Tr. I at 42-44 (Browning). Thus, Browning opined that operations on the LCRC Property extended to the fence line approximately twenty feet west of the property line. See Trial Tr. I at 46-47 (Browning); Annotated 1967 Aerial Photo (Dkt. 215-15).

Even assuming that Browning is qualified to testify regarding his interpretation of photographs and survey data, the Court finds this evidence to be unpersuasive. First, Browning fails to elaborate regarding his basis for concluding that the operations on the Gould and LCRC Properties can be distinguished simply by visually inspecting a series of aerial photographs. In the Court's view, the photographs do not support any such distinction.

Second, the 1966 survey indicating that a fence was located twenty feet west of the property line is rebutted by the testimony of Craine, Messner, and Marr, all of whom had actual knowledge regarding the location of the fence. Craine, who began working for LCRC in 1973, stated that the fence was present for as long as he has been familiar with the LCRC Property. Trial Tr. III at 164 (Craine). Craine stated that he commissioned a survey that indicated that the fence was located just east of the property lines between the Gould and LCRC Properties. Id. Messner, who began working for LCRC in 1972, stated that he did not recall the exact location of the fence but believed that it was located closer to the LCRC Property. Trial Tr. IV at 181 (Messner). Messner further stated that he and other LCRC employees never used the property immediately to the east of the RSF Facility. Id. ²² And Marr stated that he recalled a fence between the Gould and LCRC Properties being present when he began working for LCRC in 1971—and that the fence was never moved. Trial Tr. V at 31-32 (Marr).

22 Galarneau, who worked for Gould Inc. from the early 1960s until it closed operations in 1976, likewise did not believe that LCRC employees ever used the northwestern corner of the LCRC Property as a driveway or passageway. Galarneau Dep. at 72-73.

Even if LCRC did carry out its operations to the west of the property line shared with the Gould Property, Gould has not come forward with evidence demonstrating that LCRC employees disposed of TCE in that area. Indeed, the record reflects that LCRC's use of TCE was limited to a maximum of seventeen to fifty asphalt tests performed between 1985 and 1986. LCRC has established that no waste TCE was disposed of by pouring it onto the ground at the LCRC Property—rather, it was collected for disposal by Safety Kleen.

e) Travers's Theory

LCRC argues that the scientific data likewise demonstrate that the TCE contamination on the LCRC Property was caused solely by the disposal practices of Gould Inc. Def. PFFCL at 17-20. Travers described the general nature and behavior of TCE, which is a dense non-aqueous phase liquid ("DNAPL"). According to Travers, DNAPL TCE released on the surface of the ground will migrate down through the soil, leaving residual traces of TCE in the soil. Trial Tr. V at 156 (Travers). The DNAPL will continue to sink down into the water table and the groundwater, where it can pool on top of and migrate along less permeable clay layers. Id. at 157. TCE from the DNAPL will then dissolve into the groundwater and will move in the direction of the groundwater flow, creating a dissolved TCE plume. Id. DNAPL is present when TCE concentrations exceed the solubility limit of TCE in groundwater, approximately 1,100,000 micrograms per liter. Id. Samples containing concentrations at or higher than this number contain DNAPL, while samples containing concentrations lower than this number contain TCE dissolved in groundwater. Id. at 157-158.

With respect to TCE discovered in soil samples, Travers created a drawing reflecting the locations of all soil borings collected by both Gould and LCRC through 2016, as well as the TCE concentrations detected in those borings. Figure 15 to Travers Report at PageID.32163 (Dkt. 168). This figure also depicts a "halo," premised on a similar drawing created by MSG, delineating the physical range of soils impacted by the release of DNAPL TCE on the Gould Property. Id. The impacted area extends from the Gould Property onto the LCRC Property. Id.; Trial Tr. V at 168 (Travers). Travers observed that the highest TCE concentrations were detected in soil samples collected from the northeastern corner of the Gould Property, where employees of Gould Inc. admittedly disposed of waste fluids from operations at the RSF Facility. Trial Tr. V at 167, 169 (Travers). TCE concentrations decreased with distance from this location. Id. at 167. According to Travers, lower concentrations of TCE were detected in soil borings collected from under the salt barn, and were attributable to the impact from the contamination on the Gould Property. Id. at 170; Trial Tr. VI at 64-65 (Travers).

Travers emphasized that between 2012 and 2016, twenty-three soil borings were advanced in the area south of the salt barn on the LCRC Property, where Feenstra hypothesized that a second release of TCE took place by way of deposits onto LCRC soil via theorized LCRC dumping. Trial Tr. V at 164, 170 (Travers). TCE was not detected in any of these soil borings, indicating that no TCE was released in this area by way of the theorized dumping. Id. Browning and Keith Gadway, the director of Quantum, confirmed that TCE was not detected in any of these samples. Trial Tr. II at 15-16 (Browning); Trial Tr. V at 90, 92-93 (Gadway). Gadway further agreed with Travers's assessment that the absence of TCE in these soils indicates that no spill or dumping occurred at this location. Trial Tr. V at 88, 93 (Gadway). Taylor also agreed that no TCE was detected in the soils on the LCRC Property, other than the area under the salt barn. Trial Tr. II at 161-162 (Taylor).

In addition to analyzing TCE contamination within the soil, Travers analyzed the concentrations and migration of contamination within the groundwater. According to Travers, hydrogeologists determine the direction of groundwater flow by evaluating water elevation levels, measured in feet above sea level, as detected by monitoring wells. Trial Tr. V at 171-172. Generally, groundwater flows from areas of higher groundwater elevation to areas of lower groundwater elevation. Id. at 172. Hydrogeologists then draw contour maps based on the elevation data; groundwater flows perpendicular to the contours drawn on the maps. Id.

Travers evaluated the water levels measured in monitoring wells and created a series of contour maps depicting her interpretation of the groundwater flow patterns on the Gould and LCRC Properties from 2004 through 2011. See Figures 20-28 to Travers Report at PageID.32171-32179 (Dkt. 168). In Travers's view, the data demonstrate that groundwater in the northeastern corner of the Gould Property has two separate flow components, creating a groundwater divide—a line extending from west to east that passes over the northeastern corner of the Gould Property. Trial Tr. V at 176-177 (Travers). Groundwater to the north of the divide flows northeasterly, over the northwestern corner of the LCRC Property and then toward Thompson Lake. Id. at 176. Groundwater to the south of the divide flows southeasterly across the LCRC Property, to the south of the salt barn. Id. Travers observed that this pattern remained consistent over time, as depicted in a drawing consolidating all groundwater flow patterns derived from the water level measurements over the seven-year timeframe. Id. at 181; Figure 29 to Travers Report at PageID.32180 (Dkt. 168). Travers further noted that the northern and southern groundwater flow components either originate from or flow through areas on the Gould and LCRC Properties where TCE concentrations were indicative of the presence of DNAPL. Trial Tr. V at 182-183 (Travers).

Travers created a drawing, incorporated below, reflecting the location of the groundwater divide and her evaluation of the distribution and concentrations of TCE in the groundwater between 2003 and 2015. Figure 34 to Travers Report at PageID.32189 (Dkt. 168). The highest concentrations of TCE in the groundwater are located at or near the northeastern corner of the Gould Property, and concentrations decrease with distance from this area. Trial Tr. V at 184, 187 (Travers); Trial Tr. VI at 81 (Travers). One of the "hot spot" areas of highest concentration depicted on the drawing is located on the LCRC Property, while another is located on the railroad tracks just north of the Gould Property. See Figure 34 to Travers Report. Travers testified that DNAPL entered the soils in the northeastern corner of the Gould Property, sank into the groundwater, and spread out below the water table to create several hot spots where DNAPL TCE has come to be located—however, she did not specify whether this DNAPL migrated by way of a clay layer or some other impermeable surface. Trial Tr. V at 189 (Travers).²³ Travers also stated that TCE concentrations detected in groundwater samples from the area south of the salt barn were far below the solubility limit of TCE, indicating that DNAPL was not present. Trial Tr. VI at 63 (Travers).

23 Additionally, the Court notes Galarneau's testimony that Gould Inc. employees dumped waste chemicals near the railroad tracks. Galarneau Dep. at 41-43. One of the DNAPL TCE hot spots is near the railroad tracks.

Image materials not available for display.

Based on this data, as well as the groundwater flow data, Travers opined that DNAPL TCE from the source area in the northeastern corner of the Gould Property dissolves into the groundwater and flows to the northeast toward Thompson Lake and to the southeast toward the salt barn on the LCRC Property. Trial Tr. V at 183-184, 188 (Travers); Trial Tr. VI at 81 (Travers). Accordingly, Travers concluded that the northeastern corner of the Gould Property is the source of the entirety of the TCE plume. Trial Tr. V at 188-189 (Travers). Because there are no high-concentration areas along the southern part of the plume on the LCRC Property, and because no TCE was detected in extensive soil testing performed in the area south of the salt barn from 2012 through 2016, Travers opined that there was no second source area on the LCRC Property. Id. at 183-184.

f) Rebuttal to Travers's Theory

In an effort to undermine Travers's theory that the contamination originates from the Gould Property and flows onto the LCRC Property, Gould highlights a number of perceived flaws in Travers's interpretations of groundwater flow. Pl. PFFCL at 10-11. First, Feenstra observed that the southern component of Travers's groundwater flow pattern does not originate from or flow through areas of high TCE concentration on the Gould Property. Trial Tr. VI at 152-153 (Feenstra). Because the southern component flows through areas of relatively low levels of TCE contamination on the Gould Property, Feenstra opined that the higher levels of contamination observed on the LCRC Property could not have been transported from the Gould Property. Id.

While Feenstra is correct that some of the depictions of the southern flow component do not stem from areas of high TCE concentration, many of them do. See Figure 29 to Travers Report. Each arrow representing the direction of groundwater flow in Figure 29 was premised on one synoptic event, or one "snapshot" of the water level elevations. Trial Tr. VI at 75 (Travers). Travers explained that she collected data from eight to ten synoptic events between 2004 and 2011 in order to assess the overall pattern of groundwater flow over time. Id.; Trial Tr. V at 180-181 (Travers). Based on Travers's drawings, the overall trend appears to be that the southern flow component either stems from or flows through an area of high TCE concentration. See Figure 29 to Travers Report.

Second, Feenstra opined that Travers's conclusion that the southern component flows southeast is flawed because for some portion of that pathway, the groundwater would be flowing from lower to higher elevations. Trial Tr. VI at 153-154 (Feenstra). Specifically, he said that six of eight measurements show higher groundwater elevations at monitoring well ("MW") 135 on the LCRC Property than the groundwater elevations at MW-112 on the Gould Property. Id. at 153-154, 156-160 (testifying regarding Figures 20-28 of Travers Report). Because groundwater flows from areas of higher elevation to areas of lower elevation, the groundwater cannot flow from west to east from MW-112 to MW-135. Id.

Travers acknowledged that, for some years, the water levels measured at MW-135 were higher than those measured at MW-112, while for other years this was not the case. Trial Tr. VI at 98-99 (Travers). But she explained that she accounted for this fact, as the contours depicted in her contour maps correctly form an S-curve "wrapping" around MW-112 and MW-135. Id.; see, e.g., Figure 23 to Travers Report at PageID.32174 (Dkt. 168). She also stated that the overall direction of groundwater flow cannot be determined based on data points from only two discrete wells; rather the entire flow pattern must be evaluated. Trial Tr. VI at 99, 101 (Travers). The Court credits Travers's testimony that her contour maps account for this trend and were accurately drawn.

Third, Gould criticizes Travers for averaging the water levels for two nearby monitoring wells (MW-103 and MW-121) in her contour maps. Pl. PFFCL at 16. Travers stated that she averaged the levels measured in these wells because they are positioned next to each other. Trial Tr. VI at 22-23 (Travers). Browning testified that averaging water levels in different wells is not an accepted practice because the wells are bored at different depths and averaging them could result in an inaccurate understanding of groundwater flow. Trial Tr. VII at 47-49 (Browning). Specifically, Browning maintains that three "zones" with varying flow regimes exist at different depths within the aquifer. Trial Tr. II at 40-41 (Browning). However, for reasons articulated below, the Court does not credit Browning's testimony that groundwater levels measured at different depths must be analyzed separately for purposes of determining the direction of groundwater flow.

Fourth, Gould contends for the first time in its post-trial briefing that Travers's contour maps inaccurately plotted the locations of certain wells, thereby undermining the reliability of the contour maps. Pl. PFFCL at 16-17. Specifically, Gould contends that MW-103 and MW-121 are thirty-five feet apart, as shown in an MSG drawing. Id. (citing Figure 2 to MSG 2008 Suppl. Investigation and Historical Data Summ. at PageID.81156 (Dkt. 245)). Even assuming that these wells are thirty-five feet apart, the significance of this distance has not been established. Nor has Gould explained the impact of averaging the water levels measured in two wells located in the southeastern portion of the Gould Property on determining the direction of groundwater flow in the northeastern portion of the Gould Property.

Gould also contends that Travers inaccurately plotted the locations of MW-4r, MW-104r, MW-109, and MW-123. Id. at 17. While the location of MW-4r as plotted by Travers differs from its location as plotted by MSG, it does not appear from the Court's review that the locations of MW-104r, MW-109, or MW-123 diverge significantly, if at all. Compare Figure 2 to MSG 2008 Suppl. Investigation and Historical Data Summ. with Figure 23 to Travers Report. But given the utter lack of trial testimony regarding this subject, it has not been established that the MSG drawing represents an accurate depiction of the locations of these wells, while Travers's depiction is inaccurate. Nor has Gould established through any evidence what impact such discrepancies would have on Travers's overall theory.

Finally, Gould emphasizes Feenstra's testimony observing the abrupt termination of the southern lobe of the groundwater plume as depicted in Figure 34 to the Travers Report. Pl. PFFCL at 11-12. Feenstra testified that this abrupt termination of the southern lobe is inconsistent with Travers's theory that the contamination has migrated southeast in a dissolved groundwater plume. Trial Tr. VI at 161-162 (Feenstra). If the contamination had migrated in that manner, Feenstra would expect that the concentration levels would decrease gradually over a greater distance than depicted in Travers's drawing. Id. Thus, Feenstra opined that the abrupt termination of the southern lobe suggests that a deposit of TCE occurred in the area to the south of the salt barn. Id. at 163.

Because LCRC did not recall Travers as a witness to rebut this point, the Court credits Feenstra's testimony that concentrations of dissolved TCE gradually decrease over a great distance. However, Feenstra did not explain why an abrupt termination of the plume is consistent with his theory that TCE was dumped on the LCRC Property. Thus, while this point raises a question with respect to Travers's theory, it does not lend support to Feenstra's theory. And as discussed below, Feenstra's opinions ultimately suffer from more troubling weaknesses.

g) Feenstra's Theory

In arguing that an independent release of DNAPL TCE occurred on the LCRC Property, Gould relies on the opinions and theories developed by Feenstra. Pl. PFFCL at 8-12. Feenstra described the process by which DNAPL TCE migrates through the subsurface in a manner consistent with Travers's description above. Trial Tr. VI at 122-124 (Feenstra). He emphasized that DNAPL pathways are difficult to locate and that DNAPL is rarely encountered directly at contaminated sites. Id. at 124-125. Feenstra described two methods by which the presence of DNAPL TCE may be determined. Id. at 127. First, if the concentration of TCE detected in a soil sample exceeds the solubility level of TCE in soil, DNAPL TCE is present in that sample and was historically released in that location. Id. at 127-128. Second, if the concentration of TCE detected in groundwater is approximately 10% of the solubility limit of TCE in water, it may be inferred that DNAPL TCE is "in close proximity." Id. at 129-130. Feenstra did not define "close proximity," instead stating that determination would be site-specific. Id. at 132. Feenstra stated that he considers TCE concentrations in the range of 10% to 20% of the solubility limit to be "in the DNAPL zone"—another term that Feenstra did not define. See id.

According to Feenstra, a release of TCE took place on the LCRC Property in the area south of the salt barn. Id. at 120-121, 137. In that area, TCE concentrations exceeded 10% and approached 20% of the solubility of TCE, indicating the presence of DNAPL TCE that originated on the LCRC Property. Id. at 136-137. Feenstra additionally evaluated levels of ethene, a TCE byproduct produced during degradation of TCE. Id. at 143. The groundwater ethene levels in the area south of the salt barn ranged from approximately 10% to 20% of TCE solubility, signifying that a TCE DNAPL zone is present on the LCRC Property. Id. at 146-147.

Feenstra acknowledged that releases of DNAPL TCE have taken place in the northeastern corner of the Gould Property. Id. at 121. He also conceded that contaminated groundwater from the northeastern corner of the Gould Property likely migrated northeast onto the far northwestern corner of the LCRC Property. Id. at 160. Nevertheless, he opined that the TCE and ethenes detected within the southern lobe of the TCE plume on the LCRC Property cannot be attributed to migration from the Gould Property for two reasons. First, the intermediate clay layer slopes downward from the area south of the salt barn toward the northeastern corner of the Gould Property. Id. at 120-121, 133-134, 141. Accordingly, DNAPL could not have migrated uphill from the Gould Property to the LCRC Property along this path. Id. at 120-121. Instead, DNAPL TCE disposed of to the south of the salt barn on the LCRC Property would have migrated through the soils and groundwater, pooled on top of the nonpermeable intermediate clay layer, and continued to migrate downslope to the northwest toward the Gould Property. Id. Second, according to MSG's contouring of the groundwater flow, the groundwater divide is located farther south than Travers opined, such that groundwater in the northeastern corner of the Gould Property and the northwestern corner of the LCRC Property would flow to the northeast, rather than into the area south of the salt barn. Id. at 138-139; 150-151.

Feenstra's opinions are premised on MSG's contouring of the elevation of the intermediate clay layer, see id. at 133-134, 141, and on MSG's contouring of the groundwater flow, see id. at 150-151. Browning testified that an intermediate clay layer within the aquifer is continuous in the area between the Gould and LCRC Properties. Trial Tr. II at 29 (Browning). This conclusion was premised on MSG's drawings of cross-sections of the subsurface geology, which were prepared by collecting samples through soil borings and monitoring wells, i.e., by taking samples by drilling and placing "straws in the ground." Id. at 32-33 (describing Modified Figure 3A from MSG 2015 Annual Report at PageID.33768 (Dkt. 168-23)); see also Trial Tr. I at 93-94 (Browning). From the data collected, MSG interpolated the geologic features occurring between these discrete sampling points, including whether the intermediate clay layer was continuous or contained gaps. Trial Tr. II at 36-38 (Browning). Based on these studies, Browning opined that the intermediate clay layer is continuous and slopes downward from the LCRC Property to the Gould Property. Trial Tr. I at 116-122 (Browning); Trial Tr. II at 75 (Browning); Figure 3C to MSG 2008 Suppl. Investigation and Historical Data Summ. at PageID.81162 (Dkt. 245). Accordingly, DNAPL TCE would migrate downslope to the northwest from the LCRC Property to the Gould Property and could not migrate uphill. Trial Tr. I at 122 (Browning).

With respect to groundwater flow, Browning testified that a groundwater divide extending from west to east is located just north of the former sign shop (also called the cold storage building) on the LCRC Property. Trial Tr. II at 48 (Browning). The precise location of the groundwater divide is depicted in various drawings created by MSG. See Figure 4A to MSG 2008 Suppl. Investigation and Historical Data Summ. at PageID.81174 (Dkt. 245). Below is a drawing depicting MSG's opinion regarding the location of the groundwater divide, in relation to Travers's theory. May 2008 Groundwater Flow (Dkt. 216-9). Groundwater north of this divide generally flows northward, while groundwater to the south of the divide generally flows southward. Trial Tr. II at 48 (Browning). Therefore, because the TCE plume is primarily located to the north of the divide, the contaminated groundwater comprising the plume would flow northward. Id.

Image materials not available for display.

h) Rebuttal to Feenstra's Theory

As Travers's theory was subject to criticisms by Gould and its experts, so too is Feenstra's theory subject to criticism by LCRC and Travers. First, in connection with his theory that there must have been releases of TCE on the LCRC Property, Feenstra does not define what it means for DNAPL TCE to be in "close proximity." Def. PFFCL at 27-29. LCRC further maintains that no industry-wide standard recognizes that the presence of DNAPL TCE can be inferred based on TCE concentrations within a percentage of the solubility limit. Id. Gadway testified that while he was aware of a theory regarding concentration levels at which it may be inferred that there is a DNAPL source in close proximity, he is unaware of any accepted industry standard to this effect. Trial Tr. V at 94 (Gadway). Similarly, Travers testified that she was aware of general EPA guidance stating that TCE concentrations above a certain level indicate the likely presence of DNAPL and suggest that further investigation is warranted. Trial Tr. VI at 84-85 (Travers). However, she was unaware of any science establishing that DNAPL TCE would be present within a given distance. Id.

The Court agrees with LCRC's criticism that Feenstra failed to adequately define the terms "close proximity" and "DNAPL zone." Does "close proximity" mean 5 feet, 10 feet, or 100 feet? Without any definition by Feenstra, it would be sheer speculation to conclude that TCE was deposited on the LCRC site by way of dumping onto soils. And without any clarification from Feenstra regarding the proximity of DNAPL, it is possible to conclude that DNAPL TCE located in wells on the Gould Property could account for the presence of lower levels of TCE in the groundwater near the salt barn. See Trial Tr. VI at 85-86 (Travers). Further, no industry standard recognizes that the proximity of DNAPL may be inferred from TCE concentrations at a percentage of the TCE solubility limit. Thus, the concentrations of TCE detected to the south of the salt barn do not establish the presence of a "DNAPL zone." But even accepting Feenstra's premise, his theory that there is a DNAPL zone on the LCRC Property appears to be in tension with his theory that DNAPL from the LCRC Property migrated along the intermediate clay layer onto the Gould Property.

Moreover, in connection with Feenstra's theory that DNAPL TCE must have migrated northwest from the LCRC Property to the Gould Property via a continuous, sloping clay layer, Travers disputes whether the intermediate clay layer is continuous between the LCRC Property and the Gould Property. Id. at 50-51. Travers notes that MSG's data was drawn from sampling performed at discrete points. Id. at 67-68. Indeed, these samples were collected from borings and monitoring wells located approximately fifty feet apart. See Figure 2 to 2008 Suppl. Investigation and Historical Data Summ. at PageID.81156 (Dkt. 245); Figure 3C-VC to 2008 Suppl. Investigation and Historical Data Summ. at PageID.81163 (Dkt. 245). However, MSG has presented no evidence supporting its interpolation of the geological features—such as the continuity of the intermediate clay layer—between these discrete points. Trial Tr. VI at 67-68.

Browning did not satisfactorily explain the basis for MSG's conclusion that the intermediate clay layer is continuous between the Gould and LCRC Properties, when this assessment was premised on discrete data points. Browning stated that he and others at MSG are professionals qualified to make interpolations between data points. See Trial Tr. II at 37 (Browning). And although Browning stated that MSG evaluated additional data in determining that a clay layer was continuous between two points, he was unable to identify what that information was. See id. at 38. Thus, it cannot be said with any degree of confidence that DNAPL could have migrated from the LCRC Property to the Gould Property along a continuous clay layer.

In addition, Travers testified that MSG's contour maps depicting its interpretation of groundwater flow and the groundwater divide do not incorporate all water level data. Trial Tr. V at 194-196 (Travers). Browning admitted that the contours were premised only on data from shallow wells and stated that MSG created other contour maps depicting flow regimes at different depths—that is, within three different "zones" within the unconfined aquifer. Trial Tr. II at 40-42 (Browning); Trial Tr. VII at 45-46 (Browning). Travers further indicated her disagreement with MSG's decision to separately analyze groundwater flow within different zones, as she maintains that TCE contaminates the entirety of the single, unconfined aquifer. Trial Tr. VI at 30 (Travers). Travers noted that the portion of the aquifer above the basal clay layer is an unconfined aquifer that is hydraulically connected, given that no portion is fully enclosed or separated by clay. Id. at 69-70 (testifying regarding Figure 4A to Feenstra Report at PageID.76939 (Dkt. 208-8)). Because the sand in this aquifer is connected, Travers stated there was no basis to separate the aquifer into individual zones S1a, S1b, and S1c, as MSG has indicated. Id.

Browning offered minimal testimony describing the rationale or principles forming the basis for MSG's contour maps. While he stated that MSG created multiple contour maps depicting groundwater flow patterns within different zones, he failed to explain in an intelligible manner why flow patterns were measured and contoured at different depths. Browning stated that the three zones represent "different areas of flow impact" that are separated by discrete, non-continuous clay lenses. Trial Tr. II at 41 (Browning). Yet in the next breath, Browning admitted that the groundwater within the aquifer was connected and that the groundwater generally flows in the same direction at different depths, though there "could be" some variations. Id.

Further, Browning offered no explanation for why he primarily evaluated the water level data for the shallow wells in determining the direction of groundwater flow, and whether the flow regimes at different depths would impact the groundwater divide or MSG's opinion regarding the overall direction of groundwater flow. And while Gould offered a report containing these maps into evidence, see Figures 4B and 4C to MSG 2008 Suppl. Investigation and Historical Data Summ. at PageID.81175-81176 (Dkt. 245), it failed to offer any other evidence interpreting the impact of the different contour maps on the overall direction of groundwater flow.

In the Court's view, Gould has not offered sufficiently persuasive evidence justifying MSG's opinion that the aquifer has three distinct zones. Likewise, there is no sufficient justification supporting MSG's decisions to separately analyze the groundwater flow patterns within different zones and to create separate contour maps for each zone. Given the dubious accuracy of MSG's contour maps, the Court finds that they are not persuasive.

Additionally, Travers observed that MSG's contour maps included contours that were unsupported by any data. Trial Tr. V at 198 (Travers). For example, between MW-115 (which has a recorded water level of 907.97 feet above sea level) and MW-116 (which has a recorded water level of 907.54 feet above sea level), MSG plotted two contours at 908 feet above sea level. Id. (testifying regarding Figure 4A to MSG 2008 Suppl. Investigation and Historical Data Summ.). However, Travers stated that no data supported the assumption that the water levels between MW-115 and MW-116 reached or exceeded 908 feet above sea level. Id.

Again, Browning's response to this argument is unsatisfying. He stated that MSG created a three-dimensional map of the data points and manually interpreted the data. Trial Tr. VII at 45 (Browning). He admitted "less confidence" on MSG's part with respect to the dashed, eastern portions of the two contours at 908 feet above sea level. Id. In creating the contour maps, MSG additionally relied on an overview map created by EGLE. Id. However, this overview map was not described or offered into evidence. Based on this testimony, Gould has not persuasively rebutted Travers's criticism that MSG's placement of the groundwater divide is premised on some degree of speculation.

Finally, Gould acknowledges that soil in the area to the south of the salt barn has been sampled extensively from 2012 through 2016, and that no TCE was detected in any of these samples. Pl. PFFCL at 19. Nevertheless, it contends that those results are inaccurate because the soil samples consisted primarily of non-native fill material. Id. at 21. According to Feenstra, TCE will be detected in soil only if the same soil was present at the time of the release. Trial Tr. VI at 147-148 (Feenstra). Thus, he opined that soil borings through non-native soils, such as fill, do not reliably establish TCE entry locations in the vadose zone (i.e., the soil above the water table). Id. at 147-149. Gadway and Travers testified that in 1979 and 1980, LCRC graded the soil and added fill when constructing the salt barn. See Trial Tr. V at 109 (Gadway); Trial Tr. VI at 10, 43-44 (Travers). Therefore, in Feenstra's and Browning's view, the fact that TCE was not detected in soil borings advanced in the area south of the salt barn does not foreclose the possibility that there was a TCE entry area at this location. Trial Tr. VI at 147-149 (Feenstra); Trial Tr. II at 15-16 (Browning).

Gadway and Travers soundly rebut this argument. Both Gadway and Travers stated that most soil samples collected from the area south of the salt barn contained native soils in addition to fill. Trial Tr. V at 109, 111 (Gadway); Trial Tr. VI at 104 (Travers).²⁴ Gadway further testified that the soil boring logs demonstrate that there was not a widespread introduction of non-native soils in this area. Trial Tr. V at 111 (Gadway). In any event, fill would have been introduced in 1979 and 1980, before LCRC's documented use of TCE occurred in 1985 and 1986. TCE disposed of on the ground in 1985 and 1986 would have left traces in the fill material that was already present.

24 Gould highlights certain discrepancies between soil boring logs relied upon by Travers in rendering her expert report and those relied upon by Feenstra. Pl. PFFCL at 20. According to Gould, the logs Travers relied upon did not include areas of fill reflected in the logs Feenstra relied upon. Id. But the soil boring logs relied upon by Feenstra reflect the presence of native soils in every sample. See Boring Logs at PageID.42709-42716 (Dkt. 172-14).

i) Grading & Excavation Activities

Raising a separate causation theory, Gould contends that LCRC is not entitled to invoke the third-party defense because it caused a release by redistributing TCE-contaminated soils across a broader area of the northwestern portion of its property during grading and construction projects. Pl. PFFCL at 29. Specifically, in 1979 and 1980, LCRC constructed a new salt barn that required the movement of soils in order to flatten the area in preparation for construction. Trial Tr. VI at 34, 43-44 (Travers).

As acknowledged in other cases, excavation activities involving the redistribution and spreading of contaminated soils may result in a new release of contaminants contained in the soil, thereby foreclosing application of the third-party defense. PCS Nitrogen, 791 F. Supp. 2d at 494; United States v. Honeywell Int'l, Inc., 542 F. Supp. 2d 1188, 1200 (E.D. Cal. 2008). However, the evidence before the Court does not clearly demonstrate the physical scope of LCRC's grading activities or whether the grading activities introduced the contamination into a broader area of previously uncontaminated soils. As argued by LCRC, no evidence was presented regarding the extent of the grading activities. Def. Resp. at 10. In fact, Travers emphasized that she was unaware of what soils were moved, from where they were taken, or generally how LCRC graded the area. Trial Tr. VI at 10, 34, 43-44, 104 (Travers). Consequently, there is no evidence supporting Gould's argument that LCRC caused a new release by spreading the TCE contamination via its grading efforts.

j) Conclusions Regarding Causation

In sum, the majority of the scientific data supports the view that the contamination is attributable entirely to releases of TCE that took place on the Gould Property and migrated onto neighboring properties, including the LCRC Property. As acknowledged by Gould's own expert, releases of DNAPL TCE occurred in the northeastern corner of the Gould Property. Trial Tr. VI at 121 (Feenstra). And Browning and Travers agree that TCE soil contamination from the Gould Property extends onto and impacts soil in the northwestern corner of the LCRC Property. Trial Tr. II at 45 (Browning); Trial Tr. V at 170 (Travers); Trial Tr. VI at 64-65 (Travers). Feenstra concedes that some of the contaminated groundwater from the northeastern corner of the Gould Property likely migrated onto the far northwestern corner of the LCRC Property. Trial Tr. VI at 160 (Feenstra). In spite of extensive investigations on the LCRC Property, Gould has ultimately come forward with no data persuasively showing that TCE was improperly disposed of on the LCRC Property.

The data also demonstrate that TCE concentrations in the soil and groundwater were greatest in the northeastern corner of the Gould Property and decreased with distance from this location. See Figure 15 to Travers Report; Figure 34 to Travers Report. MSG created a map depicting its view of the groundwater plume, including the distribution and concentrations of TCE across the Gould and LCRC Properties. Figure 5B to 2019 Modified Comprehensive Groundwater Monitoring Plan at PageID.77725 (Dkt. 210). According to this drawing, the highest concentrations of TCE occur in the northeastern corner of the Gould Property and extend southeast across the LCRC Property toward the salt barn. Id. This figure differs from Travers's representation of the groundwater plume, as it depicts the highest concentrations spread across a larger area of both the Gould and LCRC Properties. Compare id. with Figure 34 to Travers Report.

Notably, the figures' legends use different criteria to define the "highest levels" of concentration. In Travers's drawing, the highest concentrations of TCE are in excess of one million micrograms per liter and are located on or very near the northeastern corner of the Gould Property. Figure 34 to Travers Report. In MSG's drawing, the highest concentrations of TCE are in excess of only 10,000 micrograms per liter. Figure 5B to 2019 Modified Comprehensive Groundwater Monitoring Plan. By using a far lower threshold to define the highest levels of TCE concentration, MSG's drawing represents that the core of the TCE plume is spread across a much broader area. Travers's depiction, by contrast, offers a more detailed and nuanced view of the areas of greatest concentration. Rather than showing a large area of relatively moderate concentration levels, Travers's drawing shows the precise areas where concentrations approach the solubility limit of TCE. In the Court's view, therefore, Travers's drawing is the more accurate representation of the groundwater plume.

Both Travers's and Feenstra's theories suffer from certain ambiguities or inconsistencies discussed above. However, the Court finds the flaws in Feenstra's theories to be more problematic than those in Travers's. The Court finds Travers's interpretation of groundwater flow and position of the groundwater divide to be more persuasive. For the reasons discussed above, the Court credits Travers's testimony that her contour maps were accurately drawn and accounted for variances in water levels in MW-112 and MW-135. And although Gould criticizes Travers's contouring maps because she averaged groundwater levels measured in different wells and inaccurately plotted the locations of certain wells, it fails to elaborate on whether these alleged flaws undermine her opinions and in what way.

By contrast, Feenstra's opinion depends in part on the unsupported premise that a release of DNAPL TCE can be inferred when groundwater concentrations reach 10% to 20% of the solubility limit of TCE. His theory is further premised on MSG's conclusion that the intermediate clay layer continuously slopes downward to the northwest from the LCRC Property toward the Gould Property, as well as on MSG's interpretations of groundwater flow. As described above, MSG's conclusion relative to the essentially uninterrupted continuity of the intermediate clay layer is speculative at best. And its contour maps depicting the groundwater flow and the position of the groundwater divide are derived from incomplete data from only shallow wells—with no clear explanation why other wells were excluded in developing MSG's interpretation of groundwater flow, or how the data from other wells would impact the maps.

Even so, contour maps created by both MSG and Travers are consistent with Travers's theory that groundwater flows in an easterly direction from the Gould Property onto the LCRC Property. See Figure 4A to MSG 2008 Suppl. Investigation and Historical Data Summ.; Figures 20-28 to Travers Report. As a general matter, the groundwater levels reflected in these exhibits are higher in monitoring wells to the west and decrease in monitoring wells to the north and east, suggesting that groundwater flows from the west to the north and east. See Figure 4A to MSG 2008 Suppl. Investigation and Historical Data Summ.; Figures 20-28 to Travers Report. Therefore, as described by Travers, dissolved TCE in the groundwater has migrated from the source location on the Gould Property onto the LCRC Property in the direction of the groundwater flow. See Trial Tr. V at 183-184, 188 (Travers); Trial Tr. VI at 81 (Travers).

The conclusion that the TCE contamination was caused solely by releases on the Gould Property is also consistent with the limited evidence concerning operations at the RSF Facility versus those on the LCRC Property. To be sure, the record is replete with gaps in the evidence regarding the historical operations at the Gould and LCRC Properties. Gould, in particular, came forward with no records regarding the practices and operations of Gould Inc. James Cronmiller, Gould's former director of environmental affairs, stated that he has no knowledge regarding what might have become of the operating records from the RSF Facility and admitted to making no effort to determine what became of them. Id. at 30-32. Thus, Gould has offered no explanation accounting for its inability to locate such records.

Nevertheless, the limited evidence demonstrates that Gould Inc. had a sustained practice of disposing of its waste chemicals by dumping them on the ground, including in the northeastern corner of the Gould Property. See Trial Tr. V at 69-74 (Richardson); Galarneau Dep. at 31, 41-43, 77-78. Gould Inc.'s disposal was prolific, as Galarneau estimated that thousands of gallons of waste fluids were disposed in this manner, Galarneau Dep. at 61, while Richardson testified that he personally disposed of forty to forty-five gallons of coolant fluids at a time, Trial Tr. V at 70- 73 (Richardson). Gould is correct that none of the testimony directly links Gould Inc. to TCE usage in particular. But decades later, extraordinarily high concentrations of TCE were detected in the northeastern corner of the Gould Property—one of the very areas where employees poured fluids on the ground. These waste disposal practices, when coupled with the scientific evidence, establish by a preponderance of the evidence that Gould Inc. was responsible for releasing TCE on the Gould Property.

By contrast, there is no evidence demonstrating that there were any deposits of TCE onto soils on the LCRC Property. LCRC has presented evidence demonstrating that its use of TCE was limited to a series of asphalt tests performed between 1985 and 1986. See Trial Tr. III at 178-179 (Craine); Trial Tr. V at 23, 27 (Marr); Little Dep. at 11. These tests were performed on the second floor of a building located along the eastern boundary of the LCRC Property—on the opposite side of the property from the TCE plume. See Trial Tr. III at 158, 168-169 (Craine); Trial Tr. V at 54 (Hogan). According to LCRC's employees, waste TCE was collected for disposal by Safety Kleen and was never poured onto the ground. Little Dep. at 10, 14; Trial Tr. V at 29-30 (Marr). This evidence is sufficient to establish that no disposal of TCE occurred on the LCRC Property, and Gould has come forward with no evidence rebutting this conclusion. Thus, the evidence demonstrates that Gould Inc., and not LCRC, generated the TCE contamination, which migrated onto neighboring properties, including the LCRC Property.

Gould contends that LCRC is unable to establish the third-party defense, as it fails to rule out the possibility that prior and subsequent owners and operators of the Gould Property, as opposed to Gould Inc., were culpable for the contamination. Pl. PFFCL at 21. In essence, Gould seeks to require LCRC to definitively prove each and every party that contributed to the contamination. Id. But this is not the standard.

Caselaw analyzing the third-party defense focuses on the innocence of the party asserting it, as opposed to requiring that it definitively prove all parties who were the sources of the contamination. See Lincoln Props., Ltd. v. Higgins, 823 F. Supp. 1528, 1542 (E.D. Cal. 1992) (holding that the county established that any releases were caused solely by third parties when there was no evidence of conduct by the county that contributed to the releases). As stated above, LCRC has proved by a preponderance of the evidence that the contamination at issue was caused by Gould Inc.'s disposal practices and not by LCRC. Because LCRC has established its own innocence, the possibility that third parties other than Gould Inc. may have also contributed to the contamination is immaterial for purposes of the third-party defense.

Moreover, the Court agrees with LCRC's assessment that Gould seeks to alter the burden of proof to require that LCRC's proofs be absolute. See Def. Resp. at 7-8. The standard requires LCRC to establish the elements of its defense only by a preponderance of the evidence, requiring that its evidence "make the scales tip slightly" in its favor. See Gjinaj v. Ashcroft, 119 F. App'x 764, 775 (6th Cir. 2005). It has done so with respect to the "sole cause" element of the defense.

ii. Due Care

In addition to establishing that environmental harm was caused solely by a third party, a party asserting the third-party defense must establish that it exercised due care with respect to the hazardous substance. To establish the due care element, a PRP must show "that he took all precautions with respect to the particular waste that a similarly situated reasonable and prudent person would have taken in light of all relevant facts and circumstances." Foster v. United States, 922 F. Supp. 642, 657 (D.D.C. 1996) (quoting H. Rep. No. 1016, 96th Cong., 2d Sess., pt. I, at 34 (1980), reprinted in 1980 U.S.C.C.A.N. 6119, 6137). "[D]ue care 'would include those steps necessary to protect the public from a health or environmental threat.'" United States v. A & N Cleaners & Launderers, Inc., 854 F. Supp. 229, 238 (S.D.N.Y. 1994).

LCRC first contends that it was under no obligation to undertake any investigation or remediation under a policy statement promulgated by the Environmental Protection Agency ("EPA"). Def. PFFCL at 32. The policy statement provides that the EPA will not take enforcement action against a property owner when "hazardous substances have come to be located on or in a property solely as the result of subsurface migration in an aquifer from a source or sources outside the property . . . ." 5/24/95 EPA Policy Statement, Ex. A to Def. PFFCL, at 3 (Dkt. 252-1). With respect to the due care element of the third-party defense, the policy statement provides that:

Not only is groundwater contamination difficult to detect, but once identified, it is often difficult to mitigate or address without extensive studies and pump and treat remediation. Based on EPA's technical experience and the Agency's interpretation of CERCLA, EPA has concluded that the failure by such an owner to take affirmative actions, such as conducting groundwater investigations or installing groundwater remediation systems, is not, in the absence of exceptional circumstances, a failure to exercise "due care" or "take precautions" within the meaning of Section 107(b)(3).

Id. at 7.

But LCRC has not cited, nor is the Court aware of, any authority supporting the notion that this policy applies to private cost recovery matters. Rather, the policy governs EPA enforcement actions. Further, the policy expressly states that its application is within the discretion of the EPA and that it "does not constitute rulemaking by the Agency and is not intended and cannot be relied on to create a right or a benefit, substantive or procedural, enforceable at law or in equity, by any person." Id. at 11. Accordingly, LCRC is not relieved of its obligation to exercise due care based on the above policy statement.

LCRC alternatively maintains that it has fulfilled its due care obligations, as it has engaged in extensive soil and groundwater investigations and analyses in coordination with EGLE. Def. PFFCL at 32. In 1988, LCRC became aware of MNB's investigations and response activities on the Gould Property, including an excavation immediately adjacent to the LCRC Property.²⁵ Trial Tr. IV at 17-18 (Craine). As a result of the MNB's investigations, Craine admitted that he became aware in March 1991 of the nature and former use of the Gould Property and, given its proximity to the LCRC Property, of the possibility of TCE contamination on the LCRC Property. Id. at 94-95.

25 MNB excavated an area near the property line between the Gould and LCRC Properties, referred to as the "pond," which involved petroleum contamination and not TCE. See Trial Tr. II at 93-98 (Taylor).

On March 11, 1991, Craine submitted a letter to the Board of County Road Commissioners recommending that LCRC approve its environmental consultant's proposal to perform a cleanup of underground fuel storage tanks on the LCRC Property. 3/11/91 Letter (Dkt. 196-4). The letter further noted that hydrogeological assessments would be necessary to monitor TCE contamination on the LCRC Property, which was believed to have migrated from the Gould Property. Id. Cleanup efforts took place between 1991 and 1995, Trial Tr. IV at 69 (Craine), and involved the removal of the storage tanks and surrounding soil, as well as soil and groundwater testing, 4/21/93 Letter (Dkt. 196-8). It is unclear from the record where on the LCRC Property this excavation took place. Nor does the record reflect whether LCRC's consultants performed further investigations of TCE, or the results of those investigations. In 1993, LCRC's consultant recommended to EGLE that no further testing for solvents be conducted because LCRC's waste solvents were collected for disposal and because the soil samples did not detect any additional waste oil contamination from the storage tanks. Id. Craine stated that EGLE closed the site investigation in 1995. Trial Tr. IV at 69 (Craine).

The record does not reflect that LCRC undertook further investigations of TCE until EGLE made its first contact with LCRC in 2007. On August 9, 2007, LCRC received formal notice from EGLE designating the LCRC Property a "facility" based on the presence of contamination at the site. Id. at 24; 8/9/07 EGLE Notice (Dkt. 202-13). This notice stated that due care obligations can apply to both liable and non-liable owners and operators of a site. 8/9/07 EGLE Notice ("Owners and operators of this site who are not liable for contamination may have due care and other obligations under Part 201 [of NREPA]." (footnote omitted)); see also Trial Tr. II at 108-109 (Taylor) (agreeing that due care obligations arise regardless of liability for contamination upon notice that a property is considered a facility).

According to Craine, after receiving the 2007 notice, LCRC submitted a FOIA request for EGLE's records regarding the site. Trial Tr. IV at 24 (Craine). Additionally, Craine searched for records suggesting possible TCE usage on the LCRC Property, as described above. Trial Tr. III at 178-179 (Craine). LCRC retained several expert consultants to review EGLE's files for the site, which dated back to the 1990s, and to investigate the source of TCE on the LCRC Property. Trial Tr. IV at 24-30 (Craine). Though Craine did not specify the type of investigations performed by LCRC's consultants, the record indicates that, until 2010, they primarily reviewed the scientific data generated by Gould and its consultants. See Response Activity Plan for Chlorinated Solvents at 17-27, 33 (Dkt. 172) (summarizing LCRC's investigations commencing in October 2010 and noting that LCRC's consultants performed "extensive analysis of the 20 years of prior work conducted by Gould's consultants").

In 2010, LCRC retained Quantum to evaluate the soil and groundwater in the northwestern corner of the LCRC Property adjacent to the Gould Property. Trial Tr. V at 79 (Gadway). Beginning in October 2010, Quantum completed soil and groundwater sampling and installed temporary monitoring wells in order to investigate the extent of TCE contamination near the salt barn. Id. at 79-80; Response Activity Plan for Chlorinated Solvents at 20. Between 2010 and 2016, Quantum advanced over twenty-nine soil borings on the LCRC Property, including six borings beneath the salt barn and twenty-three borings in the area south of the salt barn. Trial Tr. V at 82 (Gadway); Response Activity Plan for Chlorinated Solvents at 18-19. LCRC tailored these investigations to address comments and feedback received from EGLE. Response Activity Plan for Chlorinated Solvents at 17, 23, 25. According to Gadway, Quantum did not advise LCRC to undertake any remediation of the TCE contamination, nor did LCRC ask Quantum to prepare or implement an interim response plan to remediate the TCE contamination. Trial Tr. V at 102-105 (Gadway). In LCRC's view, it had no obligation to take steps to mitigate the contamination because it was not responsible for causing the TCE contamination. Trial Tr. IV at 102 (Craine).

In 2012, LCRC submitted a no further action ("NFA") request to EGLE, which Taylor stated was deficient because the extent of the contamination had not been fully delineated. Trial Tr. II at 125-126 (Taylor); Trial Tr. IV at 32 (Craine). Craine admitted that LCRC "stumbled a little bit," and that an NFA request was not "the right pathway." Trial Tr. IV at 32 (Craine). Accordingly, LCRC withdrew its NFA request on March 12, 2013, in order or pursue a response activity plan ("RAP"). 3/12/13 Withdrawal Letter (Dkt. 201-9). LCRC subsequently submitted a number of RAPs to EGLE, Trial Tr. II at 180 (Taylor), the most recent of which was submitted in September 2016, see generally 2016 RAP (Dkts. 170 through 172-17).

The 2016 RAP contains thousands of pages of data, technical reports, and analyses of recent soil and groundwater investigations. Id. It primarily responds to EGLE's 2012 determination that LCRC had not conducted sufficient site investigation with regard to known and suspected TCE releases; it also asserts that the Gould Property is the sole source of the TCE contamination. Report on Observed Trichloroethylene Contamination at 1 (Dkt. 170-3). While the 2016 RAP indicates LCRC's agreement to install an additional monitoring well at EGLE's request, it argued that doing so would be of limited value to the investigations. Response Activity Plan for Chlorinated Solvents at 33-34. LCRC further recommended that it take no further action with respect to TCE contamination. Id.

Taylor provides additional context regarding LCRC's response activities and cooperation with EGLE. According to Taylor, LCRC's initial response after receiving the August 2007 notice was to deny having used TCE on its property. Trial Tr. II at 109 (Taylor). In a letter dated February 7, 2008, LCRC's consultant asserted that there was no evidence that LCRC used TCE on its property, and that the TCE contamination originated solely from the Gould Property. 2/7/08 Letter at 1, 11 (Dkt. 198-6). The letter also set forth a work plan proposing that LCRC would collect samples from existing monitoring wells, advance soil borings, and analyze the storm sewer. Id. at 16. In assessing the proposed work, Taylor opined that the proposed sampling was "limited" and that the proposal was vague, leaving out many details typically included in a work plan. Trial Tr. II at 114-115 (Taylor); 2/11/08 Request for Geological Review (Dkt. 198-5).

In a letter dated January 12, 2010, EGLE acknowledged receiving correspondence from LCRC in which LCRC again asserted that it did not cause a release of TCE. 1/12/10 Letter (Dkt. 201-7). Notwithstanding LCRC's position, EGLE requested that LCRC prepare a work plan by April 2010. Id. In a letter dated February 8, 2012, EGLE indicated that it had approved a limited scope work plan submitted by LCRC in April 2010 but had received no indication that any work had been completed. 2/8/12 Letter at 2 (Dkt. 207-11).²⁶ EGLE noted that LCRC had not conducted sufficient investigations supporting its position that the TCE present on the LCRC Property was caused by releases on the Gould Property. Id. The letter also reminded LCRC of its due care obligations and noted that EGLE had not received reports from LCRC demonstrating that it had performed significant site evaluation or mitigation work since it received the 2007 notice. Id. Finally, the letter reiterated a list of investigations EGLE required to be performed. Id. at 5-6.

26 It is unclear from the record whether this work plan was submitted as part of a RAP or some other type of submission.

In a letter dated January 8, 2015, EGLE noted that LCRC had completed a portion of the work requested in its letter of February 8, 2012, including soil borings, soil sampling, and aquifer sampling. 1/8/15 Letter (Dkt. 212-13). The letter also reflected that LCRC had met with EGLE representatives in June 2014 and agreed to install another monitoring well and to perform additional soil borings; however, EGLE noted that it had not received a proposal for such work since that time. Id. Thus, EGLE concluded that LCRC had not performed response activities in compliance with its due care obligations under NREPA. Id. Taylor testified that the purpose of the letter was to communicate to LCRC the need for additional work determining the nature and extent of the TCE plume, as LCRC appeared reluctant to address the TCE contamination. Trial Tr. II at 133-134 (Taylor).

Additional correspondence from EGLE demonstrates a continued reluctance on LCRC's part to conduct response activities. For example, on August 10, 2015, EGLE issued a letter denying LCRC's amended work plan for evaluating releases of hazardous substances. 8/10/15 Letter (Dkt. 213-1). In July 2015, EGLE conditionally approved a work plan submitted by LCRC, contingent on LCRC completing additional work, including performing certain soil and groundwater evaluations and installing a monitoring well. Id. LCRC responded with an amended work plan objecting to EGLE's requests. Id. EGLE rejected LCRC's amended proposal. Id.

On September 7, 2016, EGLE denied two remedial investigation RAPs submitted by LCRC because the plans did not propose to complete meaningful response activities and because they contained insufficient information to support the conclusion that LCRC was not responsible for releasing TCE into the site soils. 9/7/16 Letter (Dkt. 213-2). Further, Taylor testified that although LCRC always cooperated with EGLE's entry onto the LCRC Property, LCRC never submitted a work plan proposing response activity such as removal or groundwater treatment. Trial Tr. II at 147, 169-170 (Taylor).

In an abrupt reversal of course, EGLE issued the June 2017 Letters indicating its determination that it had no further regulatory interest in the LCRC Property, and that LCRC was not required to perform any further sampling or to submit any additional reports. In explaining this determination, O'Brien observed that EGLE had required LCRC to perform extensive testing over the course of many years. Trial Tr. IV at 151, 159-160 (O'Brien). He further criticized the approach of requiring additional testing to gain absolute certainty regarding the sources of contamination. Id.

EGLE's position in 2017, however, does not discredit its earlier requests that LCRC perform additional work. EGLE noted LCRC's recalcitrance in performing the requested response activities in 2012 through 2015. See 2/8/12 Letter at 2, 5-6; 1/8/15 Letter; 8/10/15 Letter. The fact that O'Brien testified, based on his later evaluation of data available in 2017, that further investigation was not warranted does not bear on the necessity of the response activities requested by EGLE in 2012 through 2015. O'Brien did not testify that the scope of investigation required by EGLE with respect to the LCRC Property was inappropriate. Moreover, by 2016, Gould had installed a pump and treat system and had begun bioremediation efforts—treatments that addressed the entire scope of the contamination. See Trial Tr. I at 130-131, 136-137 (Browning); Trial Tr. II at 148 (Taylor). EGLE's change of course in 2017 may well be explained by the fact that the contamination was already being fully addressed, rendering further action by LCRC unnecessary. Thus, LCRC's earlier reluctance in performing necessary investigations is not excused.

It is well established that a failure to take timely precautions with respect to contamination demonstrates a lack of due care. See, e.g., A & N Cleaners, 854 F. Supp. at 243 (holding that immediately after being put on notice of the state's investigations into the sources of contamination on the defendants' property, the defendants should have made inquiries of environmental officials regarding the results of those investigations). For example, in Franklin Cty. Convention Facilities Auth. v. Am. Premier Underwriters, Inc., 240 F.3d 534, 548 (6th Cir. 2001), the plaintiff was not responsible for causing the contamination at the site, but was determined not to have exercised due care in spite of its response efforts. Within a year of discovering the contamination on its property, the plaintiff contacted an environmental consultant, developed a remediation plan that was approved by the Ohio EPA, and began remediating the site by removing the contaminated soil. Id. at 539-540. Upon discovering during an excavation that the contamination had migrated into a sewer line, the consultant erected a barrier to prevent further migration. Id. at 540. But the Sixth Circuit determined that the plaintiff failed to exercise due care, as it failed to act expeditiously in preventing migration of the contamination—despite being on notice of the threat of migration. Id. at 548.

Additionally, courts have recognized that a failure to comply with agency orders can demonstrate a lack of due care. For example, in United States v. Domenic Lombardi Realty, Inc., 290 F. Supp. 2d 198, 207, 211-212 (D.R.I. 2003), the court held that the defendant failed to exercise due care where it failed to remediate its property or notify tenants or visitors of the contamination, as ordered by governmental agencies. And in Interfaith Comm. Org. v. Honeywell Int'l, Inc., 263 F. Supp. 2d 796 (D.N.J. 2003), the court found that Honeywell failed to exercise due care where it engaged in avoidance tactics and, instead of responding to the pollution, "continually took the path of further testing, further debate and negotiation."

Here, the evidence demonstrates that LCRC failed to take timely measures to investigate the TCE contamination. LCRC admittedly became aware in 1991 that TCE contaminated the LCRC Property and received formal notice of the contamination in August 2007. See Trial Tr. IV at 94-95 (Craine). Although LCRC hired consultants to perform a cleanup from 1991 through 1995, the record reflects that these efforts were primarily focused on the removal of fuel storage tanks as opposed to investigation of the TCE. LCRC's inquiry into the source of the TCE contamination apparently did not begin in earnest until it retained consultants to review Gould's investigations in 2007, some sixteen years after it became aware that its own property was contaminated. The fact that EGLE did not take action with respect to the LCRC Property until 2007 is immaterial to LCRC's obligation to take timely precautions. A & N Cleaners, 854 F. Supp. at 244 ("It is no defense to insist that . . . the Government should have notified the Berkman Defendants that there was a problem on the Property, since Congress has seen fit to shift the public responsibility of locating contamination onto the shoulders of individual property owners."). LCRC did not begin performing its own investigations until 2010, three years after receiving formal notice from EGLE. See Trial Tr. V at 79-80 (Gadway). Between 2007 and 2010, LCRC appears to have focused its efforts on disputing whether it caused the releases on the LCRC Property. See 2/7/08 Letter; 1/12/10 Letter.

More troubling is the fact that even after LCRC began performing its own investigations in 2010, its activities were insufficient under the circumstances, as it consistently resisted complying with EGLE's requests to undertake certain investigations. In 2012, EGLE stated that it had received no indication that any work had been completed, suggesting that LCRC failed to keep EGLE apprised of the results of its investigations. See 2/8/12 Letter at 2. Although EGLE later acknowledged in January 2015 that LCRC had performed some of the work requested, it concluded in the same letter that LCRC's response activities were insufficient to meet its due care obligations under NREPA. See 1/8/15 Letter. The evidence demonstrates that LCRC persisted in its reluctance to cooperate with EGLE's requests for additional investigations through August 2015 and the submission of its most recent RAP in September 2016. See 8/10/15 Letter; Response Activity Plan for Chlorinated Solvents at 33-34.

Although LCRC has undertaken some investigation of the contamination, its efforts were aimed at exonerating itself from liability as opposed to mitigating the spread of the contamination. Denying responsibility for the TCE contamination has been a constant theme throughout LCRC's submissions to EGLE, including its most recent RAP. See Report on Observed Trichloroethylene Contamination at 1. As indicated above, the 2016 RAP stated that the Gould Property was the sole source of TCE contamination and proposed that LCRC undertake no further action or investigations with respect to the TCE contamination. See id.; Response Activity Plan for Chlorinated Solvents at 33-34. Taken as a whole, this evidence demonstrates a fervid resistance and lack of cooperation on LCRC's part to investigate the TCE contamination as directed by EGLE.

LCRC's ultimate lack of responsibility for causing the contamination is no defense to failing to exercise due care. As indicated in EGLE's 2007 notice, due care obligations arise irrespective of whether a party is liable for the contamination. See 8/9/07 EGLE Notice; see also Kalamazoo River, 228 F.3d at 657 ("CERCLA's scheme of strict liability . . . serves to encourage parties to clean up the site quickly and then litigate later to sort out the specifics of who should pay."); Kelley v. Envtl. Prot. Agency, 15 F.3d 1100, 1106 (D.C. Cir. 1994) (noting that a private party ordered by the EPA under 42 U.S.C. § 9606(a) to clean up a site must comply whether or not it is liable). In this manner, CERCLA's statutory scheme requires "parties to shoot first (clean up) and ask questions (determine who bears the ultimate liability) later." Kelley, 15 F.3d at 1106.

Given LCRC's failure to diligently initiate and pursue adequate response activities, as well as its demonstrated failure to cooperate fully with requests for investigations made by EGLE, it has failed to establish by a preponderance of the evidence that it exercised due care with respect to the TCE contamination. As a result, LCRC has not prevailed on its invocation of the third-party defense.

b. Contiguous Landowner Defense

LCRC also contends that it is shielded from liability under the contiguous landowner defense. Def. PFFCL at 33-36. Under CERCLA § 107(q), a property owner is not liable under § 107(a) if he "owns real property that is contiguous to . . . and that is or may be contaminated by a release or threatened release of a hazardous substance from, real property that is not owned by that person," if certain rigorous requirements are met. One requirement a defendant must establish to invoke this defense is that, at the time at which the defendant acquired the contaminated property, it (i) "conducted all appropriate inquiry . . . with respect to the property," and (ii) "did not know or have reason to know that the property was or could be contaminated by a release or threatened release of one or more hazardous substances from other real property not owned or operated by the person." 42 U.S.C. § 9607(q)(1)(A)(viii).

The Court previously held that LCRC was unable to establish the contiguous landowner defense because LCRC had actual knowledge of the TCE contamination at the time it repurchased the LCRC Property. Gould, 2020 WL 806033, at *9. The Court reasoned that LCRC sold the LCRC Property to Livingston County in 2002. Id. In 2011, LCRC repurchased the LCRC Property back from Livingston County, after it received formal notice from EGLE in 2007 that the property was contaminated with TCE. Id.

LCRC contends, as it did at the summary judgment stage, that this element of the contiguous landowner defense is inapplicable because LCRC owned the LCRC Property before the TCE contamination occurred. Def. PFFCL at 36. Additionally, citing Craine's testimony, LCRC argues that the real estate transactions involving the LCRC Property were inter-county transactions made strictly for budgetary purposes. Id. (citing Trial Tr. IV at 126-128 (Craine)). But LCRC fails to explain the legal significance of these facts. Nor has it presented any authority supporting its view that such real estate transactions can be ignored. Accordingly, as determined at the summary judgment stage, LCRC is foreclosed from invoking the contiguous landowner defense.

3. Divisibility

LCRC next contends that, in accordance with the divisibility doctrine described in Burlington Northern & Santa Fe Railway Company, 556 U.S. 599 (2009), the Court should apportion the entirety of the liability to Gould. Def. PFFCL at 36-38. Because Gould is responsible for causing the entirety of the harm, LCRC maintains that the parties' contributions to the harm are divisible and seeks to fix its share of the damages at zero. Id.

In a pretrial motion in limine and in its post-trial briefing, Gould contends that LCRC waived the divisibility defense. Pl. Mot. in Limine at 3-4 (Dkt. 109); Pl. Resp. at 12-13. Specifically, Gould maintains that LCRC failed to plead the defense in its answer or in the joint final pretrial order submitted in the earlier action, which was dismissed by the parties' stipulation. Pl. Mot. in Limine at 3-4; Pl. Resp. at 12-13. Although LCRC contends that the divisibility doctrine is not an affirmative defense that must be pleaded, it cites no persuasive authority supporting this proposition. See Def. Resp. to Mot. in Limine at 3-5 (Dkt. 113).²⁷ To the contrary, courts have referred to the divisibility doctrine as an affirmative defense. See, e.g., United States v. Alcan Aluminum Corp., 964 F.2d 252, 268 (3d Cir. 1992) ("Similarly, section 881 sets forth the affirmative defense based upon the divisibility of harm rule in section 433A"); New York v. Next Millennium Realty, LLC, 160 F. Supp. 3d 485, 511 (E.D.N.Y. 2016) (noting that "the divisibility of harm doctrine is a common law tort defense to joint and several liability under CERCLA § 9607(a)"); Akzo Coatings, Inc. v. Aigner Corp., 881 F. Supp. 12020, 1210 (N.D. Ind. 1994) ("Section 881 of the Restatement sets forth the affirmative defense based upon Rule § 433A."). Accordingly, LCRC was required to plead this defense. See Fed. R. Civ. P. 8(c).

27 Contrary to LCRC's view, Burlington does not hold that a court may sua sponte consider the issue of divisibility—in that case, the parties raised the divisibility defense before the district court. See Burlington, 556 U.S. at 615; see also United States v. Atchison, Topeka & Santa Fe Ry. Co., No. CV-F-92-5068, 2003 WL 25518047, at *82 (E.D. Cal. July 15, 2003).

LCRC has not pleaded the divisibility defense in its operative answer. Answer to 3d Am. Compl. at 8-9. While LCRC sought to amend its answer and counter-complaint to plead new claims and defenses, including the divisibility defense, see Proposed 2d Am. Answer, Ex. B to Def. Mot. to Am., at 10 (Dkt. 44-2), the Court denied the motion, 6/27/19 Op. & Order at 4, 6 (Dkt. 56); 9/27/19 Op. & Order at 3 (Dkt. 73).²⁸ The Court reasoned that the new claims and defenses went beyond the scope of the legal positions advanced in the earlier action, by which the parties agreed to be bound. 6/27/19 Op. & Order at 4, 6; 9/27/19 Op. & Order at 3. Having twice rejected LCRC's attempts to newly introduce the divisibility defense, the Court declines to reconsider those rulings. Gould's motion in limine on this issue is granted in part, as LCRC has waived the divisibility defense.²⁹

28 The Court likewise rejected Gould's attempts in an amended complaint to newly plead an NREPA cost recovery claim and to reference contaminants other than TCE, as those matters exceeded the scope of the prior action. 6/27/19 Op. & Order at 5, 6; 9/27/19 Op. & Order at 3-4.

29 The motion raised other evidentiary issues that did not materialize at trial. Therefore, the balance of the motion is denied as moot.

Even if LCRC had not waived the divisibility defense, it has not met its burden of establishing a reasonable basis for apportioning the harm. Under the divisibility doctrine, a defendant may avoid joint and several liability under CERCLA if it establishes that a harm attributable to itself and another PRP is divisible. Burlington, 556 U.S. at 614. That is, "apportionment is proper when there is a reasonable basis for determining the contribution of each cause to a single harm." Id. (internal marks omitted). Equitable considerations play no role in this analysis, as apportionment is proper only when the evidence supports the divisibility of damages. Id. at 615 n.9. The party asserting the defense bears the burden of proving that a reasonable basis for apportionment exists. Id. at 614.

The basis for apportioning the harm must be tied to the circumstances giving rise to a party's liability—whether as an owner, operator, arranger, or transporter. This principle is articulated in United States v. Twp. of Brighton, 153 F.3d 307 (6th Cir. 1998)—a CERCLA decision that generated three opinions. Judge Boggs's lead opinion reversed the district court for failing to consider fully the ways in which the contamination attributable to the Township, as an operator at a landfill owned by others, might be divisible or otherwise susceptible to reasonable apportionment. Id. at 319-320. In a concurring opinion, Judge Moore emphasized the necessity of considering the specific circumstances giving rise to a PRP's liability—emphasizing that a generator's liability is based on the contamination it actually contributed to a site, whereas the liability of an owner or operator is not. Id. at 329 (Moore, J., concurring). With respect to previous owners and operators, she reasoned that the "act" giving rise to liability "is not their involvement in the disposal of hazardous waste, but simply their ownership or operation of the facility at the time of the disposal." Id. at 330. Thus, she rejected the proposition that an owner or operator can escape liability under the divisibility doctrine when it asserts that "it lacked knowledge of or involvement in the disposal of hazardous substances at its facility at the time of its ownership or operation of the facility." Id. Although the majority opinion rejected Judge Moore's ultimate conclusion that previous owners and operators are limited to apportionment based on temporal divisibility—i.e., comparing harm created during their period of ownership or operation to periods when they did not own or operate at the site—it credited Judge Moore's point that the "distinction between operator liability and other forms of liability is very important to consider when determining divisibility." Id. at 320 (majority opinion). In other words, the different ways in which liability is established matters for purposes of apportionment. For an owner, this means that apportionment is based on contamination attributable to its parcel, and not the owner's activities (or lack of activities) at the site.

Tying apportionment to the circumstances giving rise to liability in this manner is necessary to give effect to CERCLA's liability scheme, which imposes liability on persons who did not actually cause contamination through their own disposal of hazardous substances. As courts have recognized, "Congress clearly intended that the landowner be considered to have 'caused' part of the harm." United States v. Northernaire Plating Co., 670 F. Supp. 742, 748 (W.D. Mich. 1987) aff'd sub nom. United States v. R.W. Meyer, Inc., 889 F.2d 1497 (6th Cir. 1989). Permitting owners to escape liability by demonstrating a lack of personal causation would frustrate this intent. Id.; see also Weyerhaeuser Co. v. Koppers Co., Inc., 771 F. Supp. 1420, 1425-26 (D. Md. 1991) (declining to hold a landowner harmless based on a lack of causation, as doing so would undermine CERCLA's design imposing strict liability on both the owner and the operator of a facility without reference to causation). Further, because owner liability is premised on the "act" of owning a facility as opposed to dumping hazardous substances, owners could always avoid liability unless they were also a PRP on some other basis (e.g., if the owner was also an operator that generated the contamination). Thus, upholding a zero-share apportionment based on a lack of causation would effectively eviscerate CERCLA liability for any PRP that was not personally involved in dumping hazardous substances.³⁰

30 Legal scholars have likewise noted the anomalies that would arise if the divisibility doctrine permitted parties to avoid liability based on their lack of personal involvement in dumping hazardous substances. See, e.g., Steve C. Gold, Dis-Jointed? Several Approaches to Divisibility After Burlington Northern, 11 Vt. J. Envtl. L. 307, 356-357 (2009) (suggesting that a single harm cannot be apportioned among different classes of PRPs); James R. MacAyeal, The Comprehensive Environmental Response, Compensation & Liability Act: The Correct Paradigm of Strict Liability & The Problem of Individual Causation, 18 UCLA J. Envtl. L. & Pol'y 217, 330 (2000) (observing that the divisibility analysis "breaks down" in the case of owner liability).

Burlington is consistent with the view that an owner's liability should be apportioned based on the contamination attributable to its parcel rather than its activities. In Burlington, the contamination was caused by Brown & Bryant, Inc. ("B&B"), an agricultural distribution business that operated on its own parcel and on land leased from two railways. 556 U.S. at 602-603. As a result of B&B's "sloppy" operation and spills that occurred upon delivery of pesticides and other chemicals, the soil and groundwater on B&B's parcel and the parcel it leased from the railways became contaminated. Id. at 604. Although the railways did not cause the contamination, they were found liable as owners of a portion of the facility. Id. at 605. Citing Northernaire and Weyerhauser, the district court rejected the notion that apportionment based on the railways' lack of causation could offer a complete defense (i.e., a zero-share apportionment), as the railways' liability was premised on their ownership of the facility, rather than their "actual involvement in the disposal of hazardous waste." United States v. Atchison, Topeka & Santa Fe Ry. Co., No. CV-F-92-5068, 2003 WL 25518047, at *44, 88 (E.D. Cal. July 15, 2003). In considering the divisibility defense on appeal, the Supreme Court, in apparent agreement with the district court that the railways were not entitled to a zero-share apportionment, affirmed the district court's apportionment of 9% of the response costs to the railways. Burlington, 556 U.S. at 618-619. The 9% apportionment was premised on comparing the railways' parcel and the larger site, using factors such as the relative land areas and time when contamination took place. Id. at 616-617. Thus, the railways were subject to an apportionment based on factors attributable to their property but not based on the railways' lack of involvement in the disposal. Id.

Here, like the railways' liability in Burlington, LCRC's liability stems from its status as an owner of a facility. See Answer to 3d Am. Compl. ¶ 28 (admitting that LCRC owns property that has been designated a facility). And like the railways in Burlington, it seeks a zero-share apportionment based on its lack of causation or contribution to the contamination. This is an apportionment theory completely untethered to the basis of its liability and at variance with Burlington, Brighton, and the other authorities cited above. As such, it is not a reasonable basis for apportionment. Therefore, even if LCRC had not waived the apportionment defense, it has not established it.³¹

31 In accordance with Burlington and Brighton, LCRC might have tried to prove apportionment based on a comparison of the volume or magnitude of contamination at its property relative to the Gould Property. But it did not present that theory to the Court.

D. LCRC's Contribution Counterclaims

Having been found jointly and severally liable under CERCLA § 107(a), LCRC is liable for any of Gould's response costs that are consistent with the NCP. See Kerr-McGee Chem. Corp. v. Lefton Iron & Metal Co., 14 F.3d 321, 326 (7th Cir. 1994). However, a defendant found jointly and severally liable for cost recovery under CERCLA § 107(a) can "'blunt any inequitable distribution of costs by filing a § 113(f) counterclaim.'" Lockheed, 35 F. Supp. 3d at 122 (quoting Atl. Research Corp., 551 U.S. at 140). LCRC has asserted contribution counterclaims under both CERCLA § 113(f)(1), and under NREPA, Mich. Comp. Laws § 324.20129. LCRC maintains that there is no evidence that LCRC improperly disposed of TCE, whereas Gould Inc. consistently disposed of waste solvents on the ground of the Gould Property. Def. PFFCL at 39. Thus, LCRC argues that it should be allocated a zero-share of Gould's response costs and should be awarded the full costs of its own investigations. Id.

In evaluating the merits of these claims, the Court first turns to the viability of LCRC's contribution claim under NREPA and concludes that this claim must be dismissed. Next, the Court determines that LCRC has established the elements supporting a finding that Gould is liable under CERCLA. Finally, considering the equitable factors, the Court concludes that Gould must be allocated 95% of the response costs because its corporate predecessor caused the entirety of the contamination. However, because LCRC failed to fully cooperate with EGLE, it must bear 5% of the response costs.

1. NREPA Contribution Counterclaim

At the outset, the Court notes that NREPA provides that "[a] person may seek contribution from any other person who is liable under section 20126 during or following a civil action brought under this part." Mich. Comp. Laws § 324.20129(3) (emphasis added). CERCLA § 113(f)(1) similarly provides that a suit for contribution may be brought "during or following any civil action under section 9606 of this title or under section 9607(a) of this title." Although there is no state law on the matter with respect to NREPA, the Supreme Court has held with respect to CERCLA that a private party who has not been sued under § 107(a) may not obtain contribution from other liable parties under § 113(f). Atl. Research, 551 U.S. at 138-139; Cooper Indus., Inc. v. Aviall Servs., Inc., 543 U.S. 157, 160-161, 168 (2004).

NREPA is patterned after CERCLA and is, therefore, construed in accordance with CERCLA. City of Detroit v. Simon, 247 F.3d 619, 630 (6th Cir. 2001); ITT Indus., 700 F. Supp. 2d at 894; Freeport-McMoran Res. Partners Ltd. P'ship v. B-B Paint Corp., 56 F. Supp. 2d 823, 838 n.7 (E.D. Mich. 1999). As such, a party that has not been sued under NREPA may not assert a contribution claim under Michigan Compiled Laws § 324.20129. Because LCRC has not been sued under NREPA, its NREPA contribution counterclaim must be dismissed.

2. Gould's Liability

Under CERCLA § 113(f)(1), an entity found liable under § 107(a) "may seek contribution from any other person who is liable or potentially liable under section [107(a)] . . . ." 42 U.S.C. § 9613(f)(1). In resolving contribution claims, courts are to "allocate response costs among liable parties using such equitable factors as the court determines are appropriate." Id. Initially, a plaintiff bringing a contribution claim must prove that the defendant is liable under CERCLA. Kalamazoo River, 228 F.3d at 656-657. The liability standard for contribution claims under § 113(f)(1) is the same as for cost recovery claims under § 107(a). Id. at 653.

Courts are divided regarding what elements a party must establish to prevail on a contribution counterclaim. Some courts have required a party to establish the four elements of a prima facie case under § 107(a). See, e.g., ITT Indus., 700 F. Supp. 2d at 888-889; Castaic Lake Water Agency v. Whittaker Corp., 272 F. Supp. 2d 1053, 1073 (C.D. Cal. 2003); G & H Landfill PRP Group v. Am. Premier Underwriters Inc., No. 96CV72947, 1999 WL 33432164, at *4-5 (E.D. Mich. Jan. 29, 1999). Other courts, however, have held that a party asserting a contribution counterclaim need only prove that the counter-defendant is a PRP and that the equities favor allocation of the response costs. See, e.g., Bd. of Cty. Comm'rs of Cty. of La Plata v. Brown Group Retail, Inc., 768 F. Supp. 2d 1092, 1120 (D. Col. 2011); Ashley II of Charleston, LLC v. PCS Nitrogen, Inc., 791 F. Supp. 2d 431, 490 (D.S.C. 2011) ("A party making a claim under CERCLA § 113(f) bears the burden of proving: 1) that the defendant is a responsible party under § 107(a) of CERCLA; and 2) the defendant's equitable share of costs."); Anspec Co., Inc. v. Johnson Controls, Inc., 788 F. Supp. 951, 958 (E.D. Mich. 1992) ("To establish a claim for contribution under § 9613(f), Hoover Group must prove at the outset that the plaintiffs are within one of the four categories of covered persons under § 9607(a) of CERCLA."). Because LCRC is able to establish the elements of a prima facie claim for cost recovery, it is able to demonstrate Gould's liability under either of these standards.

As stated above, a party seeking to establish a prima facie case must establish the following four factors:

(1) a polluting site is a "facility" within the statute's definition; (2) the facility released or threatened to release a hazardous substance; (3) the release caused the plaintiff to incur necessary costs of response; and (4) the defendant falls within one of four categories of potentially responsible parties.

Gould, 2012 WL 5817937, at *7.

Here, the first element is met because Gould has admitted that EGLE issued a notice designating the Gould Property a "facility" based on the presence of environmental contaminants on the site. Cronmiller Dep. at 33-34; Answer to Am. Counter-Compl. ¶ 26 (Dkt. 120).

The second element is also established. In its answer to LCRC's counter-complaint, Gould admits that a release took place on the Gould Property. Answer to Am. Counter-Compl. ¶ 90. And as discussed above with respect to the LCRC Property, the evidence demonstrates that a release has taken place on the Gould Property because TCE is leaching through the soil and groundwater. The evidence illustrates that a halo of residual TCE soil contamination extends from the Gould Property onto the northwestern corner of the LCRC Property. Trial Tr. II at 45 (Browning). Moreover, a plume of groundwater contaminated with TCE is migrating from both the Gould and LCRC Properties northward toward Thompson Lake. Trial Tr. II at 89, 99 (Browning); Trial Tr. V at 187-188 (Travers). Consequently, a release has taken place on the Gould Property.

With respect to the third element, LCRC has established that it incurred necessary costs of response as a result of the releases on the Gould Property. The Court concluded above that the TCE contamination present on the LCRC Property is attributable to migration from the releases that occurred on the Gould Property. Thus, costs incurred by LCRC in connection with investigating the TCE contamination are tied to the releases that took place on the Gould Property. LCRC offered testimony from its managing director, Steven Wasylk, regarding the costs incurred by LCRC in investigating the contamination. Wasylk stated that LCRC has paid Quantum $657,711.49, and has paid Stratus Consulting and Abt Associates, Inc., Travers's consulting firms, $578,938.95, for a combined total of $1,236,650.44. Trial Tr. VI at 110-112 (Wasylk). LCRC also produced summaries documenting these payments. Summaries of Quantum Invoices at PageID.47664-47665 (Dkt. 175); Summary of Stratus/Abt Invoices at PageID.47640-47641 (Dkt. 175).³² However, Wasylk indicated that these summaries—which add up to numbers that are somewhat less than the figures he referenced in his testimony—do not account for additional payments made to Quantum after February 29, 2020 and to Stratus Consulting and Abt Associates, Inc. after March 19, 2020. Trial Tr. VI at 110-112 (Wasylk).

32 Though LCRC filed copies of the underlying invoices on the docket, it did not move for admission of these records. See Stratus Invoices (Dkt. 175-1); Abt Invoices (Dkt. 175-2); Quantum Invoices (Dkt. 175-12). The testimony of Wasylk, however, is sufficient to establish the payments assertedly made.

These response costs were necessary, given that the scope of work performed by LCRC is attributable, at least in part, to EGLE's directives as summarized above. See ITT Indus., 700 F. Supp. 2d at 884-885 (holding that response costs incurred in performing work required by state agency were necessary and recoverable). For example, EGLE requested in January 2010 that LCRC prepare a work plan by April 2010. 1/12/10 Letter. EGLE later itemized a list of work it required to be performed, including investigating the storm sewer, delineating the extent of the contamination, performing soil borings and groundwater analyses, and installing monitoring wells. 2/8/12 Letter.

Further, the Court noted above that LCRC's response costs were limited to investigation and monitoring. Within the Sixth Circuit, it is well-established that "preliminary" or "initial" investigative costs may be recovered even if the plaintiff did not comply with the NCP. Krygoski Constr. Co. v. City of Menominee, 431 F. Supp. 2d 755, 762 (W.D. Mich. 2006); see also Village of Milford v. K-H Holding Corp., 390 F.3d 926, 934 (6th Cir. 2004) (holding that consistency with the NCP is not required for recovery of monitoring and investigation costs and that investigatory and removal costs are not subject to the NCP's public comment requirements). Accordingly, LCRC need not independently prove that its response activities were consistent with the NCP.

Finally, Gould is a PRP. CERCLA defines the following four categories of PRPs who may be held liable under CERCLA:

(1) the current owner/operator of a facility from which there has been a release; (2) a person "who at the time of disposal or treatment of any hazardous substances owned or operated any facility at which such hazardous substances were disposed of"; (3) generators of hazardous waste; and (4) arrangers for the disposal of hazardous waste.

Bob's Beverage, 264 F.3d at 697 (quoting 42 U.S.C. § 9607(a)(1)-(4)). The Court concluded above that Gould Inc. was an owner and operator at the time TCE was disposed of on the Gould Property. Gould assumed responsibility for the obligations and liabilities of Gould Inc. arising out of the ownership of the Gould Property. Answer to Am. Counter-Compl. ¶ 1 (admitting that Gould is responsible for "certain obligations" of Gould, Inc.); Trial Tr. III at 73 (Callahan) (admitting that Gould is responsible for addressing the site investigation and clean-up work at the Gould Property); Rich Dep. at 19 (admitting his understanding that a liability arising from Gould Inc.'s operations at the RSF would reside with Gould). Accordingly, Gould is a PRP because its predecessor, Gould Inc., owned and operated the RSF Facility at the time TCE was disposed of on the Gould Property. LCRC has, consequently, established a prima facie case against Gould.

3. Equitable Allocation

In resolving a contribution claim, courts typically look to a set of six equitable factors, called the "Gore factors," when allocating response costs between responsible parties. The Gore factors were originally part of an amendment to the 1980 House Superfund Bill proposed by then-Congressman Albert Gore as a moderate approach to joint and several liability. Envtl. Transp. Sys., Inc. v. ENSCO, Inc., 969 F.2d 503, 508 (7th Cir. 1992). The Gore factors include:

(1) the ability of the parties to demonstrate that their contribution to a discharge, release or disposal of a hazardous waste can be distinguished; (2) the amount of the hazardous waste involved; (3) the degree of toxicity of the hazardous waste involved; (4) the degree of involvement by the parties in the generation, transportation, treatment, storage, or disposal of the hazardous waste; (5) the degree of care exercised by the parties with respect to the hazardous waste concerned, taking into account the characteristics of such hazardous waste; and (6) the degree of cooperation by the parties with Federal, State, or local officials to prevent any harm to the public health or the environment.

Kerr-McGee, 14 F.3d at 326 n.4. Courts have significant discretion in choosing which factors to consider when determining equitable allocation of liability, PCS Nitrogen, 714 F.3d at 186, and "a court may consider several factors, a few factors, or only one determining factor . . . depending on the totality of circumstances presented to the court," Envtl. Transp. Sys., 969 F.2d at 509.

Many courts have commented on the difficulties of allocating liability where incomplete records prevent a precise determination of each party's causative contribution to the contamination at issue. Under such circumstances, courts have gone forward with making as reasonable an assessment as possible based on the available evidence, noting that "[c]ourts are not required to make meticulous findings as to the precise causative contribution each of the parties have made to a hazardous site, as in many cases such a finding would be literally impossible." Kalamazoo River Group v. Rockwell Intern., 107 F. Supp. 2d 817, 822 (W.D. Mich. 2000); see also United States v. R.W. Meyer, Inc., 932 F.2d 568, 573-574 (6th Cir. 1991) ("I do not believe Congress intended to require meticulous findings of the precise causative contribution each of several hundred parties made to a hazardous site. In many cases, this would be literally impossible."). Although parties are not held to a standard of scientific certainty, they must nevertheless prove their case by a preponderance of the evidence. Kalamazoo River, 107 F. Supp. 2d at 822.

a. Parties' Contributions

The first, second, and fourth Gore factors may be condensed into the same inquiry, i.e., the parties' relative contributions to the release of hazardous substances. Lockheed, 35 F. Supp. 3d at 132. In many cases, this represents "'the dominant factor in determining each party's equitable share of liability . . .'" Id. (quoting United States v. Davis, 31 F. Supp. 2d 45, 63 (D.R.I. 1998)). This case is no exception to that principle.

As discussed above, the evidence establishes that the TCE contamination was caused solely by releases on the Gould Property. The admittedly limited evidence regarding the parties' historical operations demonstrates that Gould Inc. had a sustained practice of disposing of its waste chemicals by dumping them on the ground on the Gould Property. Former employees Richardson and Galarneau testified that they disposed of waste fluids by pouring them on the ground on the Gould Property. Trial Tr. V at 69-74 (Richardson); Galarneau Dep. at 31, 41-43, 77-78.

Richardson and Galarneau also testified that a large degreasing tank was located on the eastern side of the RSF Facility, near the area of the 2001 soil excavation. Trial Tr. V at 64-66 (Richardson); Galarneau Dep. at 20-27. Gould's own expert stated that the tank was likely a vapor degreaser containing TCE. Trial Tr. VII at 8 (Feenstra). Parts that were soaked in the degreasing tank were left to drip dry first over the tank and then were moved to racks where they continued to drip dry over the floor. Galarneau Dep. at 24-27. According to Galarneau, the tank was periodically drained, with the waste fluids flowing into a floor drain. Id. at 26-27. And while Galarneau observed a truck deliver fluids used in the tank on a weekly basis, he never observed the truck collect fluids for disposal. Id. at 30.

Although the evidence does not establish a precise volume of waste disposed of by Gould Inc., there can be no doubt that its practices resulted in the disposal of enormous quantities of chemicals on the Gould Property or within the RSF Facility. Galarneau estimated that thousands of gallons of waste fluids were disposed of on the ground or in the floor drain, Galarneau Dep. at 61, while Richardson testified that he personally disposed of forty to forty-five gallons of coolant at a time, Trial Tr. V at 70-73 (Richardson). Galarneau estimated the dimensions of the degreasing tank to be approximately four feet wide by eight feet long, and up to his chest in depth. Galarneau Dep. at 20-22. Richardson testified that the tank was approximately thirty feet long, eight feet wide, and six feet in depth. Trial Tr. V at 65 (Richardson). Assuming that Galarneau's lesser dimensions are more accurate, the degreasing tank held a capacity of approximately 128 cubic feet, or 957 gallons.

Travers similarly estimated that the degreasing tank held 1,000 gallons and, assuming the tank was filled halfway and refilled weekly as indicated by Galarneau, she estimated that Gould Inc. used roughly 416,000 gallons of TCE over the course of its sixteen years of operation. Trial Tr. V at 136-137 (Travers). This calculation is premised on approximate dimensions of the tank and on the accuracy of Galarneau's testimony that a truck delivered fluids for the tank on a weekly basis. While this evidence does not allow for a precise calculation of the amount of TCE used by Gould Inc., it does provide a reasonable estimate that Gould Inc. likely generated thousands of gallons of waste fluids over the course of its operations.

While none of the evidence definitively establishes that Gould Inc. used TCE in particular, it is not uncommon for companies to dispose of waste without knowing its contents. See Kalamazoo River Group v. Rockwell Int'l, 107 F. Supp. 2d 817, 839 (W.D. Mich. 2000). In such circumstances and where the parties lack direct evidence regarding the amount of hazardous wastes that were dumped, courts may rely on circumstantial evidence to accomplish the remedial purpose of CERCLA. Id. Here, extraordinarily high concentrations of TCE were detected in the northeastern corner of the Gould Property—the very area where Galarneau testified that he and other employees poured fluids on the ground and the approximate location of the degreasing tank. When viewed in the context of the scientific data, the Court can easily conclude by a preponderance of the evidence that Gould Inc.'s irresponsible waste disposal practices account for the presence of TCE on the Gould Property.

Meanwhile, the evidence demonstrates only minimal use of TCE on the LCRC Property. As determined above, LCRC's use of TCE was limited to a series of seventeen to fifty asphalt tests performed between 1985 and 1986. See Trial Tr. III at 178-179 (Craine); Trial Tr. V at 23, 27-28 (Marr); Little Dep. at 11. Each test required roughly one quart of TCE. Trial Tr. V at 28 (Marr); Little Dep. at 18. Accordingly, even if LCRC performed as many of fifty asphalt tests over the course of two years, it would have used a total of only twelve to thirteen gallons of TCE. While Gould contends that LCRC's use was much higher, as Little testified that Safety Kleen began to deliver larger quantities of TCE, Gould has not established that LCRC used these quantities or that LCRC improperly disposed of the waste solvent.

Additionally, there is no evidence suggesting that LCRC disposed of its waste TCE by pouring it on the ground or that it otherwise caused a release of TCE. LCRC employees performed the asphalt tests on the second floor of a building located on the opposite side of LCRC's Property from the TCE plume. See Trial Tr. III at 158, 168-169 (Craine); Trial Tr. V at 54 (Hogan). Testimony from LCRC's former employees established that waste TCE was collected for disposal by Safety Kleen and was never poured onto the ground. Little Dep. at 10, 14; Trial Tr. V at 29-30 (Marr). This evidence is sufficient to establish that no improper disposal of TCE occurred on the LCRC Property, and Gould has come forward with no evidence rebutting this conclusion.

So too does the scientific evidence support the conclusion that Gould is responsible for the entirety of the contamination. The highest concentrations of TCE in soil and groundwater are located at or near the northeastern corner of the Gould Property. See Figure 15 to Travers Report; Figure 34 to Travers Report. And it is undisputed by Gould's experts that a halo of contaminated soils extends from the Gould Property onto the LCRC Property. See Trial Tr. II at 45-46 (Browning); Trial Tr. V at 168 (Travers). The Court also credited Travers's theory that TCE contamination in the groundwater has migrated in the direction of the groundwater flow onto the LCRC Property. Trial Tr. V at 188-189 (Travers).

Accordingly, the Court concludes that Gould is solely responsible for contributing to the TCE contamination present on the Gould and LCRC Properties, as well as downgradient properties.

b. Parties' Degree of Care and Cooperation

The fifth and sixth Gore factors evaluate the parties' degree of care and level of cooperation with federal or state officials in preventing further harm by the contamination. "Because non-cooperating parties can undermine CERCLA's goal of promoting quick and efficient cleanups, '[t]he degree of cooperation with government officials to prevent any harm to the public health or the environment is very important in the contribution analysis.'" Lockheed, 35 F. Supp. 3d at 132.

The Court has already determined that LCRC failed to exercise due care and was recalcitrant in cooperating with EGLE's requests for additional investigations on the LCRC Property. LCRC performed some of the investigations requested by EGLE and permitted EGLE access to the LCRC Property whenever necessary. However, even after becoming aware in 1991 that the LCRC Property was contaminated with TCE, the record does not reflect that LCRC undertook any significant investigations until it received formal notice from EGLE in 2007. After receiving formal notice, LCRC was slow to perform any response activity and resisted some of EGLE's requests. Moreover, its investigations were largely driven by a desire to avoid liability. A party's failure to assist with investigation and remediation of contamination in favor of devoting its resources to avoiding liability can weigh against that party at the equitable allocation stage. See Brown Group, 768 F. Supp. 2d at 1121-1122.

Gould, on the other hand, has exercised due care in responding to the TCE contamination. In 1988, MNB undertook an excavation and other remediation activities on the Gould Property. See ASTI 1989 Investigation & Remediation Activities Report at 1 (Dkt. 170-4). At this time, the investigations on the Gould Property involved petroleum contamination and not TCE. See Trial Tr. II at 93-98 (Taylor). In January 1993, EGLE designated the Gould Property a "facility" due to the presence of environmental contaminants on the site, and it identified Gould as a potentially responsible party. Cronmiller Dep. at 33-34; MNB Settlement Agreement; Answer to Am. Counter-Compl. ¶ 26. At that time, Gould assumed responsibility for MNB's investigation of the Gould Property. Trial Tr. I at 32 (Browning); MNB Settlement Agreement. In 1994, Gould retained MSG to investigate the Gould Property and to ensure that all necessary remediation had been completed by MNB. Trial Tr. I at 31-32 (Browning).

According to Browning, Gould was close to closure of the site investigation until, in 1997 or 1998, the investigation unexpectedly revealed the presence of TCE on the Gould Property. Trial Tr. II at 20 (Browning); Trial Tr. I at 154 (Browning). As a result, EGLE required Gould to expand its investigation to determine the sources of the TCE. Trial Tr. I at 157-158 (Browning). MSG, on behalf of Gould, performed additional investigation on the Gould Property and, in 2001, performed an excavation in the northeastern corner of the Gould Property during which 400 to 500 cubic yards of soil containing TCE were removed. Id. at 158, 161. However, the excavation did not remove all of the TCE contamination, as high concentrations of TCE remained in the groundwater beneath the excavated site. Id. at 168. To remediate the groundwater contamination, MSG recommended in 2005 that Gould install a pump and treat system. Id. at 170. In 2013, MSG developed a work plan for the pump and treat system, which EGLE ultimately approved. Id. at 128, 171-172. Because the work plan was subject to EGLE's oversight and input, the pump and treat system was not installed until 2016. Id. Additionally, MSG sought and obtained EGLE's approval of a work plan to treat groundwater in the areas with the highest TCE concentrations through bioremediation, a process that involves injecting additives into the ground. Id. at 135-136.

Some testimony in the record suggested that Gould failed to act with reasonable care because it did not install a pump and treat system until 2016, eleven years after MSG recommended doing so and twenty years after TCE was first detected on the Gould Property. Gadway testified that such a lengthy delay was not customary and that there was no scientific basis accounting for the delay. Trial Tr. V at 96, 98-99 (Gadway). He further testified that installing a pump and treat system earlier would have reduced the likelihood of TCE reaching Thompson Lake, and that he advises clients that they must begin the remediation process as quickly as possible. Id. at 99-100.

Browning explained that between 2001 and 2016, MSG was planning and undertaking investigative efforts in preparation for steps toward remediation. Trial Tr. I at 171-172 (Browning). Browning explained that in 2005, Gould, MSG, and EGLE believed that additional investigation was necessary to gain a full understanding of the source and extent of the contamination before a treatment system could be implemented. Trial Tr. II at 80-81 (Browning). Further, the fifteen-year delay between 2001 and 2016 was attributable to the process of gaining EGLE approval of MSG's investigations, developing a work plan for the pump and treat system, and obtaining EGLE's approval of the proposed plan. Id. at 81; Trial Tr. I at 172 (Browning).

Feenstra agreed with Browning's assessment that a pump and treat system may only be implemented after the extent of the contamination has been delineated. Trial Tr. III at 50 (Feenstra). He also stated that the length of time MSG took to identify the sources of contamination, delineate the extent of the groundwater plume, and develop a remediation plan was consistent with his experience at other contaminated sites. Id. at 33. He elaborated that it can take ten to twenty years to satisfy regulatory requirements for long-term remedial projects. Id. In reviewing the course of the work performed, Feenstra did not believe there was any indication of delay in the part of Gould or MSG. Id. at 37-38.

In light of Browning's and Feenstra's testimony, Gould has persuasively rebutted LCRC's argument that it unnecessarily delayed its remediation efforts. The Court accepts their explanations that a thorough investigation defining the extent of the contamination must take place before an effective plan for remediation can be developed and implemented. The Court also credits their explanation that the process of obtaining EGLE's approval of investigative work and remediation can prolong the time necessary to interdict the contamination. As a result, Gould has exercised an appropriate degree of care with respect to the TCE and has cooperated with EGLE.

c. Allocation of Response Costs

Notwithstanding the commendable degree of care and cooperation exercised by Gould, it is inescapable that, based on the evidence submitted, Gould Inc. was the sole cause of the TCE contamination migrating onto the LCRC Property and other downgradient properties. Given its predecessor's role in causing the harm, the fact that Gould is spearheading the response effort is only proper. Additionally, Gould Inc.'s contamination caused LCRC to become embroiled in the aftermath and to incur its own significant response costs. Thus, because Gould Inc. caused the TCE releases, the equities justify allocating the lion's share of the response costs to Gould.

LCRC, however, is not entirely blameless, as it failed to exercise reasonable care in investigating the contamination and failed to fully cooperate with EGLE. Courts frequently increase the allocated share of response costs as a penalty for failing to cooperate with regulatory agencies. See, e.g., ASARCO LLC v. Atl. Richfield Co., LLC, 975 F.3d 859, 870 (9th Cir. 2020) (affirming the district court's allocation of an additional $1 million against a party that "repeatedly evaded responsibility for any environmental contamination at the Site, flagrantly misled the EPA regarding its releases at the Site, and made ongoing misrepresentations throughout the course of the litigation"); Agere Sys., Inc. v. Advanced Envtl. Tech. Corp., 602 F.3d 204, 235 (6th Cir. 2010) (affirming the district court's allocation of an additional 8.7% of response costs against a party for working with a known polluter and another 8.7% for not cooperating with the EPA); United States v. Consolidation Coal Co., 345 F.3d 409, 414-415 (6th Cir. 2003) (doubling a party's initial 3% allocation of the response costs because of the party's failure to cooperate or participate in the remediation); Am. Int'l Specialty Lines Ins. Co. v. United States, No. CV09-01734, 2013 WL 135405, at *5-6 (C.D. Cal. Jan. 9, 2013) (holding that the plaintiff's cooperation with state regulatory authorities over a twenty-year period, without assistance from the federal government, necessitated a 5% increase in the government's equitable share and a corresponding 5% decrease in the plaintiff's equitable share).

A party is not insulated from shouldering an equitable share of the response costs simply because it did not cause or contribute to the contamination. In Valbruna Slater Steel Corp. v. Joslyn Mfg. Co., 298 F. Supp. 3d 1194, 1204 (N.D. Ind. 2018), the court allocated 75% of the response costs to the defendant and 25% of the response costs to the plaintiff, although the defendant was responsible for causing the entirety of the contamination. The court noted that the plaintiff was aware of the contamination at the time it purchased the property at a discounted price and, consequently, assumed the risk of the cleanup. Id. at 1201.

Here, LCRC acted in a manner that frustrated EGLE's efforts to determine the source of the TCE plume and to prevent its further migration toward Thompson Lake. As noted above, a party's failure to cooperate with regulatory authorities' response efforts undermines CERCLA's goals of promoting prompt cleanups and preventing harm to the public health and the environment. Thus, it is appropriate to assign some portion of the costs to LCRC.

Based on these equitable considerations, the Court finds that Gould is responsible for 95% of the response costs, while LCRC is responsible for 5% of the response costs. LCRC's share is in line with the shares allocated in other cases to PRPs that failed to fully cooperate with regulatory agencies. The allocation applies to Gould's past and future response costs, as well as to LCRC's past response costs.³³ Because LCRC has not claimed that it will incur any future response costs, the Court does not make any ruling in that regard.

33 Future response costs may be awarded where there is a likelihood that a party will incur future costs recoverable under CERCLA. GenCorp, Inc. v. Olin Corp., 390 F.3d 433, 451 (6th Cir. 2004); see also 42 U.S.C. § 9613(g)(2) ("In any such action . . . the court shall enter a declaratory judgment on liability for response costs or damages that will be binding on any subsequent action or actions to recover further response costs or damages."). Here, Gould has adduced evidence that, over the next ten years, it is likely to incur $2,933,480 in continuing remediation costs. MSG Ten-Year Cost Projection (Dkt. 207-10). These costs include the operation and maintenance of the pump and treat system, improvement of the remediation system, and long-term monitoring of the system. Trial Tr. I at 145 (Browning).

Gould disputes LCRC's ability to recover its response costs, arguing that these costs were incurred for litigation purposes and to avoid liability rather than to mitigate the contamination. Pl. PFFCL at 24 n.10. Response costs are recoverable under CERCLA only if they are necessary. Carson Harbor, 270 F.3d at 871. "It is generally agreed that this standard requires that an actual and real threat to human health or the environment exist before initiating a response action." Id. A party's litigation-related costs, including investigative activities incurred solely for litigation purposes, are generally not compensable under CERCLA. ITT Indus., 700 F. Supp. 2d at 883.

In determining whether response costs are recoverable, courts generally decline to consider a party's subjective motivations for completing response activities. See Carson Harbor, 270 F.3d at 872 ("In determining whether response costs are 'necessary,' we focus not on whether a party has a business or other motive in cleaning up the property, but on whether there is a threat to human health or the environment and whether the response action is addressed to that threat."). In ITT Industries, the court rejected the defendants' argument that the plaintiff could not recover its investigation response costs admittedly incurred in an effort to avoid liability by establishing that its predecessor was not a historical source of TCE contamination. 700 F. Supp. 2d at 883. The court held that the plaintiff's goal of avoiding liability did not render its costs unnecessary or unreasonable, given that the risks of being determined a source of the contamination justified a thorough investigation and that the work was required by the state regulatory agency. Id. at 884. Thus, the court concluded that these costs were not unnecessary or unreasonable simply because they also served the purpose in avoiding liability. Id.

A different result was reached in Champion Labs., Inc. v. Metex Corp., No., 02-5284, 2009 WL 2496888, at *22 (D.N.J. Aug. 13, 2009), where the court held that the plaintiff could not recover its past costs because none of the expenditures furthered a cleanup of the site or sought to contain the migration of hazardous substances onto the site. For ten years, the plaintiff made no effort to clean up the contamination but instead unsuccessfully sought a no further action determination from the state agency to relieve it of any further investigatory or remedial responsibilities. Id. Thus, the plaintiff's sampling and investigation was not "conducted for purposes of site investigation ultimately leading to the selection of a remedy," but rather to persuade the state agency that the defendant was at fault and to obtain a no further action determination. Id. at *8, 22. Ultimately, the state agency did not grant the NFA request but instead determined that the plaintiff failed to undertake certain requests for installation of monitoring wells or cleanup activities. Id. at *8.

In the present case, LCRC's response costs are more akin to those at issue in ITT Industries than to those in Champion. As in ITT Industries, even though LCRC incurred its costs in connection with a successful defensive strategy establishing that it was not a source of the TCE contamination, LCRC's investigation was relevant and responsive to EGLE's interest in determining the sources of the contaminants. See 700 F. Supp. 2d at 884. Furthermore, the results of LCRC's investigation have been informative of the overall effort to address the TCE plume. See id. Consequently, because LCRC's response costs were necessary and reasonable, they may be recovered in accordance with the allocation described above.

Having determined that Gould incurred past response costs in the total amount of $4,253,297.00, of which LCRC is responsible for a 5% share, the Court finds that Gould is entitled to an award of $212,664.85 against LCRC for Gould's past response costs. And having determined that LCRC incurred past response costs in the total amount of $1,236,650.44, of which Gould is responsible for a 95% share, the Court finds that LCRC is entitled to an award of $1,174,817.92 against Gould for LCRC's past response costs. Additionally, LCRC will be responsible for 5% of Gould's future response costs.³⁴

34 The parties submitted no argument or authority on the issue of what date should be used to separate past from future response costs. Because the parties were able to seek response costs up to the start of trial, an appropriate dividing line would be the trial's start date of July 13, 2020. Five percent of any costs incurred on or after that date will be subject to recovery by Gould against LCRC as future response costs.

IV. CONCLUSION

The Court finds in favor of Gould on its CERCLA cost recovery claim and holds that LCRC is liable for Gould's response costs. However, the Court also finds in favor of LCRC on its CERCLA contribution counterclaim. Having considered the equities, the Court holds that Gould's equitable share of its past and future response costs, as well as LCRC's past response costs, is fixed at 95%. LCRC's equitable share of these response costs is fixed at 5%. Accordingly, with respect to past response costs, LCRC is liable to Gould in the amount of $212,664.85, while Gould is liable to LCRC in the amount of $1,174,817.92. LCRC will also be responsible for 5% of Gould's future response costs. Finally, the Court dismisses LCRC's NREPA contribution counterclaim.

Gould filed a motion in limine seeking to exclude the testimony of LCRC's expert Constance Travers, which the Court denies (Dkt. 92). Gould also filed a motion in limine seeking to exclude other evidence, which is granted in part and denied in part as moot (Dkt. 109). Both parties additionally filed pre-trial objections to proposed trial exhibits (Dkts. 135, 148, 149, 150, 151, 152, 154). The Court addressed certain portions of these objections during a hearing held on July 7, 2020, but deferred ruling on some of the objections until a specific exhibit was offered during trial. See 7/9/20 Order Regarding Motions in Limine and Objections to Exhibits (Dkt. 240). Because none of the objections on which the Court deferred ruling arose during trial, the parties' objections are denied as moot.

A separate judgment will be issued.

SO ORDERED. Dated: November 19, 2020

Detroit, Michigan

s/Mark A. Goldsmith

MARK A. GOLDSMITH

United States District Judge