William C. Griesbach

FINDINGS OF FACT AND CONCLUSIONS OF LAW

A seven-day trial to the Court addressed three questions in this CERCLA action. First, the Defendants attempted to prove that Appleton Coated Paper Company (ACPC), to which Plaintiff NCR is a successor, had arranged for the disposal of hazardous waste, pursuant to 42 U.S.C. § 9607. The Defendants also endeavored to show that NCR was itself liable for arranging to dispose of hazardous waste. Second, the Plaintiffs tried to demonstrate that insurance and other indemnity payments received by the Defendants should be offset from those Defendants' contribution claims. Finally, the Plaintiffs asserted that certain of the costs claimed by the Defendants are not recoverable under CERCLA. On the primary issue that occupied most of the attention of the parties, I conclude that ACPC did not arrange for the disposal of hazardous waste within the meaning of CERCLA. The offset and cost issues are also addressed. The following constitutes my findings of fact and conclusions of law on these issues.

This action arises out of the contamination of the Fox River with polychlorinated biphenyls (PCBs) during the period 1954-1971, which is known in this action as the production period. The history of the discharge of PCBs has been recounted in earlier decisions of this Court and elsewhere. For present purposes, the salient inquiry involves the recycling of "broke" produced by ACPC during the production period. All witnesses agreed that "broke"—an industry term—constitutes the trim, cuttings and waste inevitably created during the process of making or coating paper. Richard Wand, a former executive of the Bergstrom Paper Company (now Defendant Glatfelter), explained that the presence of a number of paper mills in the Fox Valley attracted recycling mills to the area as well. The paper mills and converters produced ample quantities of broke during the production of paper, and recycling mills like Bergstrom could take that wastepaper, recover the paper fibers from it and turn that recovered fiber into new paper. (Tr. 73.) Ultimately, the process of recycling NCR's paper resulted in the discharge of unusable substances, including toxic PCBs, into the Lower Fox River, and particularly into Little Lake Butte des Morts, which is known in this action as Operable Unit 1 (OU1).

I. Findings of Fact on Arranger Liability

A. ACPC's Production of Broke

The Appleton Coated Paper Company produced coated paper that NCR sold as business forms. As recounted in more detail elsewhere, NCR produced an emulsion that contained a solvent called Aroclor 1242, which contained polychlorinated biphenyls, or PCBs. That emulsion was a key component of NCR's carbonless copy paper (CCP). NCR sold the emulsion to ACPC, and ACPC coated it on paper and sold it back to NCR for sale to NCR's customers. Although the Defendants have persuasively argued that NCR understood the potential risks of PCBs, especially during the final years of the production period, they have not argued that anyone at ACPC knew, during the production period, that NCR's emulsion contained PCBs, or that PCBs were harmful.

An inevitable part of the paper coating process was the production of broke, which became the raw material that was recycled by recycling mills. Constituting the trimmings and unusable waste from the paper making or coating process, broke was a byproduct of the papermaking process rather than deliberate production effort. The goal of ACPC and papermakers generally was to produce "firsts" or "perfects"—the paper, envelopes, boxes, etc. that their operations were directed at producing and selling. At ACPC, for example, former employee Floyd Strelow, former CEO Dale Schumaker, and former assistant controller Robert Hietpas all testified that broke was something to be minimized and was detrimental to the company's bottom line. There was a "wide disparity in price between broke and the perfects," so in an ideal world ACPC would produce very little broke and would maximize production of the coated paper it sold. (Tr. 973.) If broke were treated as a separate "product," rather than a byproduct, it would have been a money-loser for ACPC because it was sold at a price below the value of the labor and materials that went into producing it. (Tr. 1007.) In short, it was undisputed that ACPC did not intentionally create broke and its employees hoped it would produce as little of it as possible.

Even so, broke was valuable. The Defendants' expert, Dr. Dolan, conceded that because ACPC recorded the broke as an asset on its balance sheet, that was a good indication that the broke was viewed as valuable. (Tr. 317:2-5.) In essence, although broke was an unprofitable byproduct of the paper-coating process, ACPC viewed the recovery and sale of broke as a way of mitigating production losses from its perfects rather than as the creation of an independent "product" for sale. Because broke was valuable to recycling mills, ACPC sold the broke rather than having it hauled away as trash or burning it. Floyd Strelow testified that at ACPC, employees had "fairly elaborate procedures for the handling of broke." (Tr. 956.) Employees would collect the broke and sort it based on a grade, after which it was placed in a box. Several witnesses testified that broke was graded based on how much fiber it was expected to yield during recycling: the higher the recoverable fiber yield, the more valuable the broke was to a recycling mill. (The fiber was the only point of recycling the broke in the first place.) Thus, sorting and grading the broke allowed ACPC to maximize the value it could obtain in the broke marketplace. In rough numbers, broke would fetch anywhere from $40 to $80 per ton, and Strelow testified that there was always a ready demand for ACPC's broke. When a box was full, the broke would be placed in a baler, a machine that would compact the broke and strap it into a bale of between 1,200 and 1,500 pounds. (Tr. 957.) When enough broke was produced, ACPC would load it onto a railcar and arrange a sale with a broker, who would instruct ACPC where to ship the broke.

Professor Breeden testified credibly that ACPC invested substantial costs in the recovery of broke. (Tr. 1177-1180.) Its employees had to gather the broke, grade it, sort it, store it, and bale it; the employees also had to spend time accounting for the sales of broke and deciding where to ship it. In addition to these labor costs, ACPC invested capital in a baler and also purchased other materials needed to sell the broke, such as twine and wrappings. It also had to commit valuable space to storing the broke.

B. Recycling of Broke

ACPC's broke, which of course contained the PCBs from NCR's emulsion, was shipped to a number of recycling mills along the Fox River. These mills competed with each other to obtain usable broke at the best prices, and these mills viewed broke as a valuable and even indispensable product, given that recycling it was the basis of their business. Dedric "Wally" Bergstrom IV, of the Bergstrom mill, testified that he was in brisk competition with the other mills to obtain broke, which was available in limited quantities. (Tr. 1302.) He stated that while he was in charge of procuring wastepaper for Bergstrom, he was on the phone constantly, talking to wastepaper dealers and brokers about obtaining broke. (Id.) Sometimes, during periods of scarcity, he would call dealers on weekends to ensure that the mill did not run out of broke.

From the recyclers' end, the sole purpose in obtaining broke for recycling was to separate the valuable, reusable paper fiber from everything else that was in the broke. ACPC's broke, for example, was made up of fibers as well as components like ink, filler, clay, brightener and, unfortunately, NCR's toxic coating. When a recycling mill received a shipment of broke, it would send the broke into an agitator called a hydropulper or a broke beater, a machine something like a giant blender filled with water, which breaks the paper down to the fibers. From there, the useful fibers are separated out into an oatmeal-like sludge, and the rest of the components are screened away. (Tr. 61-62, 79-80, 350-51, 362, 1306-1308.) These other components end up in a slurry and are typically treated before being discharged. At Bergstrom, for example, wastewater treatment could remove 60 to 80 percent of pollutants before the sludge was discharged as effluent into Little Lake Butte des Morts, known in this action as Operable Unit 1 (OU1). (Tr. 63-64.) Charles Klass, an industry expert, testified that 80 percent was at the high-end of treatment efficiency during the production period. (Tr. 366.)

C. ACPC's Knowledge About Broke Recycling And Effluent

Defendants attempted to demonstrate ACPC's knowledge about the recycling process in essentially two ways. First, they tried to show that certain ACPC employees had specific knowledge about the process, either through education or work experience. Second, they presented evidence that it was common knowledge in the industry that significant quantities of broke waste product (the non-fibrous materials contained within broke, such as coatings, etc.) would end up in the Fox River or Little Lake Butte des Morts.

The Defendants' efforts to prove that specific individuals had knowledge that effluent from the broke recycling process would end up in the River were hamstrung by the passage of time. Because more than forty years has passed, most of the individuals involved in ACPC's broke during the production period have either died or have faded memories of that time period.

Ronald Jezerc, a manager of product development at ACPC during the production period, testified in person. He stated that he had no knowledge of what recycling mills did with their wastewater and did not know they were polluting the Fox River with effluent containing ACPC broke waste. On cross-examination, he conceded that he knew from experience at a mill in Maryland that the coatings removed from broke at that mill went to the mill's wastewater treatment facility. (Tr. 1350.) He further agreed with counsel that he knew that water was essential to the paper recycling process and that "water is utilized to carry off the byproducts, the stuff that's not fiber, and the undesirable wastes. Water is used to carry that off to wastewater treatment—a settling lagoon or some other facility." (Tr. 1351.) Jezerc further testified that he had been a member of TAPII, the Technical Association of the Pulp and Paper Industry, and eventually the president of his TAPII chapter. He had also obtained a masters degree in paper science from the Institute for Paper Chemistry in Appleton. Due to age and a serious accident, however, his memory of specific events and processes may have been clouded.

I am satisfied that if Jezerc had been asked in 1970 what he thought happened to the non-fibrous components of ACPC's broke, he very likely would have been able to provide a reasonably accurate description of the process, both from his work in Maryland (where he admitted he understood that the effluent went into the Potomac River after treatment) and from his general knowledge and education. Like many other witnesses, he was steeped in the paper industry, which was portrayed as a somewhat clubby and close-knit community with an extraordinary concentration in the Fox Valley. I am also satisfied, however, that he was truthful in stating that he had no idea what specifically happened to ACPC's broke while he was employed there. The reason he could testify in that fashion was that he was not involved in the process of selling ACPC broke or handling it at all. (Tr. 1333.) Similarly, he could truthfully testify that he had no knowledge of what any of the recycling mills did with their wastewater because he had never had occasion to even think about such concerns, given that he was not responsible for ACPC's broke. He testified that his job involved research for specialty papers, such as NCR paper, and thus was not involved in dealing with the byproducts of that paper. But, as noted earlier, had he been asked the question or given it any thought at the time, it would not have been difficult for someone with his experience and education to hazard a fairly accurate guess, particularly since he had an understanding of how the mill he had worked at in Maryland dealt with its wastewater.

The Defendants also attempted to show that other specific individuals had knowledge about the paper recycling process. For example, ACPC executive Thomas Busch drafted an informal, undated, handwritten memo titled "Potential Aroclor to Appleton Sewers 1954-69." He calculated that some 58,000 pounds of Aroclor had gotten into the sewer system; he further noted that there had been recycling of some 46 million pounds of paper waste. (Ex. 2084.)

The Defendants argue that this memo demonstrates unequivocally that Busch understood that the recycling of his company's paper inevitably led to the release of extensive amounts of Aroclor into the River. The memo, however, is not easily deciphered. At a minimum, it shows that Busch understood that Aroclor was the emulsion being coated onto the paper his company produced for NCR. It also shows that he understood that some amount of Aroclor was released into the Appleton sewers. But this is not surprising, as ACPC itself (his own company) sent its wastewater into the Appleton public sewer system. What's lacking is any suggestion that Busch knew the recyclers in OU1 flushed their wastewater (after treatment) into the River. Nor is there any suggestion that Busch understood that wastewater treatment was not 100% effective. Finally, because the undated notes must have been written after 1969, they cannot be relied upon to show that Busch or any ACPC employees knew about Aroclor releases during the production period (1954-71).

I cannot conclude that any specific individual at ACPC had knowledge about what happened to the non-fibrous components of broke after the fiber was repulped at a recycling mill. ACPC and its managers were members of numerous paper associations, and several of its employees may have attended conferences during which broke treatment and water disposal were discussed. But these individuals are no longer with us, and I am reluctant to make a finding that any given person must have had a particular kind of knowledge. Based on the evidence presented, I conclude that Ronald Jezerc is emblematic of a manager who had a generalized understanding of paper recycling and knowledge such that he could have known that broke waste products would end up in the Fox River. Clearly he would have known that the purpose of recycling broke was to remove the useful paper fibers and dispose of everything else. That is not disputed. He would probably also have known that disposing of the "everything else" meant the use of lots of water, and he could probably have imagined that its final destination would have been the River. But he also would have known that the water was treated before being released, and there is little evidence that he (or anyone else) had a thorough enough understanding of water treatment to know how much, if any, Aroclor might have remained in treated water. Like other employees, thinking about what happened to broke after it left the plant was simply not part of his job. In hindsight, given the horrible environmental and health problems that ensued, it is perhaps difficult to appreciate the indifference of ACPC employees to the final destination of the broke. But we have to remember that there is no evidence that ACPC employees knew that the PCBs in the NCR coating were toxic. With that in mind, it makes much more sense that ACPC employees would be wholly indifferent to the final destination of their broke—there was simply no reason to even consider it, because it was not deemed to be harmful, just as we do not worry about the final destination if we pour a half glass of orange juice down the drain. In sum, I do not find that the Defendants have shown by a preponderance of the evidence that any specific individuals had actual knowledge about where ACPC's coatings would go after its broke was recycled.

The Defendants were slightly more successful in showing that some ACPC employees (albeit unnamed ones) could or would have had knowledge about where its broke waste products ended up. Charles Klass, an industry consultant with five decades of experience in the pulp and paper industry, testified credibly that it was common knowledge in the paper industry that recycled paper involved separating the fiber from the broke and disposing of the rest of the product in landfills and in the river. ACPC employees certainly knew that. Someone with only a "basic knowledge" of the paper industry would have understood that at least some of the non-fiber components (the coatings, inks, clay, etc.) of recycled broke would end up in the river. (Tr. 358.) He noted that ACPC, in particular, paid special attention to the yield its broke would provide because the yield of reusable fibers was what largely dictated the price it could command for its broke. The flip-side of yield is non-fibrous material. That is, anything in the broke that was not fiber would be suspended in a watery slurry. (Tr. 364-66.) That slurry would be treated by use of a clarifier—generally speaking, a process that collected wastewater in large quantities and used gravity to allow the solids to settle to the bottom. But it was common knowledge, he said, that no clarifying system is 100 percent efficient, and something on the order of 20 or more percent of the non-fibrous materials would remain in the water. That water was then released to the river. Klass believed that anyone in the paper industry would have at least a basic understanding of this process, in part because the quantities were so great. (Tr. 366.) For instance, even with high-yield broke and a highly efficient (80 percent) wastewater treatment process, 1.5 tons of broke waste product would end up in the river from every railcar of broke recycled. (Id.) Richard Wand testified that Bergstrom would go through several railcars of broke every day. (Tr. 76.)

Exhibit 5997 is an aerial photo of the Fox River and Little Lake Butte des Morts. Clearly visible in the lake is a large white plume of sludge resulting from effluent released by the Bergstrom Mill. Richard Wand, of Bergstrom, explained that although releases like the one in the photo were not everyday occurrences, they were fairly common and were visible to anyone boating on or driving past the Lake. (Tr. 68.) Moreover, in Wand's view the discharges were not a secret because they were explained in company brochures and tours of the mill, which were quite common. (Tr. 81-82.) He also testified credibly that it would be common knowledge in the paper industry that recovering usable fiber from wastepaper was a water-intensive process:

I've been in the business for 30 years, [and know] that all mills recycling use, one way or another, water in their recycling, and that to use wastepaper you have to break it apart and take out the contaminants. I believe most people in the paper industry have at least a rudimentary knowledge of that. It's that simple. Break apart the wastepaper, carry out the contaminants .. . .

The Plaintiffs' expert, Dr. Neil Shifrin, was an expert on wastewater treatment and ably explained the different kinds of primary and secondary water treatment facilities available. He testified that there was no way an ACPC employee during the production period would have been able to understand the process of treating its broke after it was recycled. His opinion rested on his argument, which was credible, that very few people had much understanding of the wastewater treatment process during the production period. For example, even as a newly-minted chemical engineer in 1971, Shifrin himself would not have understood exactly what happened to wastewater. (Tr. 1379.)

However, cross-examination revealed that Shifrin had employed an exacting standard to the question. In his view, an ACPC employee could not be said to have knowledge that a non-trivial release of broke waste would end up in the river unless that employee was an expert in wastewater treatment, either through education or training. Even Bergstrom employees viewing the large white plume visible in Exhibit 5997, he said, would not have had the requisite knowledge about the tons of broke slurry the plant had released.

Based on the testimony and exhibits provided, I am persuaded that the papermaking and recycling process was fairly straightforward and that there were likely managers at ACPC who knew, at least generally, that the chemicals contained in their paper could wind up in the River through the recycling process. The parties were able to explain the papermaking and recycling process to this Court in less than two weeks, and I find it likely that intelligent people who spent the entirety of their career in the business would have known (had they considered it, a key point) that the ultimate destination of broke waste was the River. This is not to say that they were experts in wastewater treatment; of that, there was no evidence. But there were vast quantities of broke being sent out for recycling, and industry people had a general appreciation that recycling was a water-intensive process. Moreover, the community was small and centered around the River and Little Lake Butte des Morts, where discharges were both visible and common. It is thus reasonable to conclude that ACPC employees responsible for broke recycling would have known that at least some portion of the broke waste product would end up in the River. The next question is whether such general knowledge is, under the circumstances, sufficient to support a finding of arranger liability.

II. Conclusions of Law on Arranger Liability

Section 9607(a)(3) applies to an entity that "arrange[s] for disposal ... of hazardous substances." The Supreme Court's 2009 decision in Burlington Northern and Santa Fe Ry. Co. v. United States is controlling. 556 U.S. 599 (2009). There, the Court observed some of the problems inherent in interpreting the term "arrange":

It is plain from the language of the statute that CERCLA liability would attach under § 9607(a)(3) if an entity were to enter into a transaction for the sole purpose of discarding a used and no longer useful hazardous substance. It is similarly clear that an entity could not be held liable as an arranger merely for selling a new and useful product if the purchaser of that product later, and unbeknownst to the seller,
disposed of the product in a way that led to contamination. Less clear is the liability attaching to the many permutations of "arrangements" that fall between these two extremes-cases in which the seller has some knowledge of the buyers' planned disposal or whose motives for the "sale" of a hazardous substance are less than clear. In such cases, courts have concluded that the determination whether an entity is an arranger requires a fact-intensive inquiry that looks beyond the parties' characterization of the transaction as a "disposal" or a "sale" and seeks to discern whether the arrangement was one Congress intended to fall within the scope of CERCLA's strict-liability provisions.

The Court concluded that under the plain meaning of the term, "an entity may qualify as an arranger under § 9607(a)(3) when it takes intentional steps to dispose of a hazardous substance." Id. at 610. In doing so, the Court noted that "in some instances an entity's knowledge that its product will be leaked, spilled, dumped, or otherwise discarded may provide evidence of the entity's intent to dispose of its hazardous wastes." Id. But the Court cautioned that "knowledge alone is insufficient to prove that an entity 'planned for' the disposal, particularly when the disposal occurs as a peripheral result of the legitimate sale of an unused, useful product." Id.

This Court addressed the nuances of the question in its Decision and Order denying summary judgment to the Defendants. (Dkt. No. 1080.) More recently, the Ninth Circuit case of Team Enterprises, LLC v. Western Investment Real Estate Trust addressed the question of knowledge and intent, using Burlington Northern as a framework. 647 F.3d 901 (9th Cir. 2011). There, a dry cleaner was deemed liable for disposing of a hazardous chemical used in the dry cleaning process—it simply dumped it down the drain. The dry cleaner sought contribution from the manufacturer of a machine that processed the chemical for reuse on the theory that it knew some of the chemical could not be reused and would inevitably be poured down the drain by dry cleaners. The Ninth Circuit rejected this effort on the grounds of intent:

Team [the dry cleaner] insists that intent can be inferred from Street's designing its product in such a way as to render disposal inevitable. According to Team, the Rescue 800 generated wastewater containing dissolved PCE, and Team allegedly had "no other choice than to dispose of the contaminated waste water" by pouring it down the drain. But the design of the Rescue 800 does not indicate that Street intended the disposal of PCE. At most, the design indicates that Street was indifferent to the possibility that Team would pour PCE down the drain. This is insufficient.

A. Knowledge Alone is Not Enough

Although the factual parallels are not strong, in my view the Team Enterprises case demonstrates one key problem the Defendants face here. Above I have found it likely that some ACPC employees in control of the broke disposal operation knew, in a general sense, that some portion of the NCR emulsion could end up in the River. At best, however, that knowledge would have been in the nature of an afterthought—much like the manufacturer in Team Enterprises. The Team Enterprise case stands for the principle, already established in Burlington Northern, that knowledge alone is not a substitute for intent. Although some employees might have understood the process from beginning to end, there was no evidence that it was part of their job duties to consider such things. Nor was there any evidence that they actually did consider such things. Suppose someone has some scrap copper lying around, which is of no value to him but is valued by others. To him it is scrap, but to others it is useful and therefore valuable. When he sells the scrap to a dealer, he might have some general inkling that the copper could be used in wire or tubing or any number of other applications, but as far as he is concerned his purpose is simply to make a little money. And, although he wants to get the scrap off of his property, there is nothing harmful or toxic about the scrap that would give him extra motivation to have it disposed of or to cause him to think about its final destination. He is simply indifferent to the final destination of the copper.

Indifference is, at most, what occurred here. Even if some ACPC employees would have had a general understanding that broke waste product could end up in the River (had they ever considered the question), it was simply not a part of the calculus when they determined how and where to dispose of the broke. ACPC employees had no involvement in the recycling plants that released the broke waste into the River. In fact, they were even a further step removed because ACPC sold the broke through brokers who arranged for the sales rather than any specific recycling mills. Once they loaded the broke onto the train, their involvement and their interest in the final destination was at an end. Team Enterprises, 647 F.3d at 909 ("At most, the design indicates that Street was indifferent to the possibility that Team would pour PCE down the drain.") In sum, there was no evidence that any ACPC employees gave much thought to the final destination of their broke waste. Thus, although I have concluded that the Defendants showed a likelihood that some ACPC employees would have known that some broke waste would end up in the River, there was no evidence that this was a point that had ever been considered by any ACPC employee during the production period.

B. No Intent to "Dispose"

Other factors support my conclusion that ACPC's employees were not arrangers. For example, the fact that ACPC sold the broke for a profit distinguishes it from the typical "disposal" case. The Defendants spent considerable effort making the case that the broke was viewed as "waste" and thus not a "useful" product. Clearly, ACPC wanted to minimize broke production because its actual product (coated paper) was far more lucrative than scrap, which was a money loser. Although in a subjective sense the broke was "waste" to ACPC (because it had no use for it), ACPC's significant efforts to round up the broke and arrange for its sale (rather than disposal) are far more telling than whether the parties subjectively viewed the product as "waste" or "useful." The fact is that broke had elements of both a waste product and a useful product. ACPC disposed of the broke in the manner described above primarily because that was the only way ACPC could make money on it. Above all else, financial considerations were the driving force. ACPC treated the product as something of value and took care to maximize the value it would receive upon sale. It actually invested capital in broke and devoted economic resources to its collection and sale.

The recent General Electric case is an instructive "mixed motives" case. During the 1960's GE used large quantities of Aroclors in equipment it manufactured. During that period it contracted with another company that took the chemicals GE could not use and used them in paints. The paint company paid a small fee ("bargain prices") for the chemicals. Id. at 380. The district court and First Circuit agreed that GE could be found to be an arranger because the "waste" aspect of the chemical was obvious:

GE stored scrap Pyranol, a byproduct of its capacitor manufacturing operations, in second-hand 55- gallon drums—often labeled "scrap Pyranol," "waste Pyranol," "scrap oil," or otherwise, depending on the manner in which it was collected—which were then placed in its facilities' salvage areas. The parties' stipulated facts and other materials in the record show that GE pursued varied arrangements by which to deplete its scrap Pyranol stockpile, for example, by transferring scrap Pyranol to local landfills, selling it to local government entities which could use it as dust suppressant, giving it away to its employees for use as a weed killer, or discharging it into the Hudson River. Moreover, the unstable nature of the materials GE provided Fletcher, as well as the fact that a significant amount of GE's scrap Pyranol was contaminated with other substances or extraneous objects, betrays the fact that GE subjected these chemicals to minimal or non-existent quality control.

Several important distinctions stand out. First, the product is an obvious waste product one typically finds in arranger cases: 55-gallon drums of chemicals. This stands in sharp contrast to the seemingly innocuous bales of paper scraps at issue here. As noted further below, few people would suspect that paper scraps could be a "hazardous substance," whereas large drums of old chemicals pose an obvious environmental risk (even if the specific risk due to PCBs was not so obvious). Second, GE did not always sell the chemicals. It disposed of them however it could—to landfills, to the Hudson River, and it even gave the chemicals away to employees. In contrast, ACPC's broke was an important, if small, component of its business model, a consistent and predictable way to mitigate production losses. It was accounted for as an asset and there were procedures in place to grade and sort it. Third, GE did not subject the chemicals to quality control, in contrast to ACPC's broke. It was not contested that ACPC spent significant effort and expense to get the most out of its broke. By contrast, it appears that GE simply treated its unusable chemicals as "salvage" or "scrap" and happened to get lucky, on occasion, when it found someone who wanted to buy it.

Given the above, the First Circuit not surprisingly concluded that "GE viewed scrap Pyranol as waste material and that any profit it derived from selling scrap Pyranol to Fletcher was subordinate and incidental to the immediate benefit of being rid of an overstock of unusable chemicals." Id. at 385. By contrast, ACPC's benefit was primarily financial. In contrast to the ad hoc disposal arrangement in General Electric, here ACPC was operating in a well-established market for paper, where there was an established secondary market for paper scraps. There were brokers in place to arrange the transactions, and there was an entire industry designed around access to paper scraps for recycling. Selling broke was, in short, part of ACPC's business model. Although it was no doubt happy to be rid of the scraps, that is overshadowed by the largely financial aspect of the process. Had it simply wanted the broke off its property, ACPC could easily have simply contracted with a waste disposal company to haul the broke away, or it could have burned it or disposed of it in some other fashion much more easily. Instead, it invested in a process geared towards maximizing the value of the broke to recyclers, which in turn maximized the value ACPC received from the broke. In short, I am satisfied that broke bears more resemblance to a useful product than it does to waste.

C. ACPC did not Know the Broke Contained Hazardous Substances

The Burlington Northern court concluded that under the plain meaning of the term "arrange," "an entity may qualify as an arranger under § 9607(a)(3) when it takes intentional steps to dispose of a hazardous substance." 556 U.S. at 610. It is seems doubtful that a defendant can ever be found to be an arranger if he did not know the substance in question is hazardous. Clearly liability may attach if the arranger did not know about the specifics of the hazard in question (e.g., that PCBs are harmful) if, for example, the product is obviously or inherently dangerous, such as large drums of oil or other chemicals. See United States v. Gen'l Electric, supra. But here there is not even any indication that ACPC employees knew much, if anything, about the nature of the NCR coating at all. The ACPC employees were not sending 55-gallon drums of chemicals away to be disposed of, but seemingly harmless bales of paper scraps. The idea that they were intentionally disposing of a "hazardous substance" is thus even further attenuated.

Even if a would-be arranger is not required to appreciate the particular hazard of the substance being disposed of, certainly a party's knowledge about the nature of the product is relevant to the inquiry as to whether he is "disposing" of anything at all within the meaning of CERCLA. It also speaks to intent. If one knows that a given chemical is hazardous, and if that chemical is of no use to him, it would be far easier to conclude that he was "disposing" of that chemical, even if he did receive some compensation for it. Gen'l Electric, 670 F.3d at 385-86. The substance being disposed of is not only useless, but potentially harmful, and thus it is likely that the overriding motivation is to dispose of the substance. That is, of course, the typical arranger case. By contrast, if one is able to sell a product that is ostensibly safe, his desire to get rid of that product is not so pressing, particularly if the product is of value. One simply does not "dispose of" otherwise valuable things unless there is a good reason. Given that there is no evidence any ACPC employees knew that Aroclor contained toxic PCBs during the production period—or even that they suspected their paper scraps could be harmful in any way—it is difficult to conclude that they had the requisite intent to dispose of the broke. To the extent they wanted to get rid of the broke, it certainly wasn't because they viewed it as potentially harmful. At worst, it was messy scrap paper that was taking up space.

This line of analysis meshes well with the underlying purpose of arranger liability under CERCLA. "Within the CERCLA scheme, arranger liability was intended to deter and, if necessary, to sanction parties seeking to evade liability by 'contracting away' responsibility." Gen'l Elec. Co., 670 F.3d at 382. Here, when the dangerousness of the product is unknown to the would-be arranger, it is difficult to find that the disposer was trying to evade liability for that danger. Moreover, given the absence of knowledge, there is no conduct to deter by imposing arranger liability because the disposal was essentially an innocent act: people who do not even suspect that their product is harmful are not in a position to be deterred. Thus, when the Burlington Northern court notes that "if an entity were to enter into a transaction for the sole purpose of discarding a used and no longer useful hazardous substance," it is implicit or assumed in such a statement that the entity knew or at least suspected that the substance was harmful. 556 U.S. at 610. The "purpose" of the transaction is to "discard" the harmful substance. As the First Circuit observed, "following Burlington Northern, a discernible element of intent to dispose of a hazardous substance is necessary for an entity to be sanctioned pursuant to § 9607(a)(3)." Gen'l Electric, 670 F.3d at 383.

Once again, the contrast with General Electric is telling. As noted above, there is no indication that GE knew that PCBs were toxic at the time (the mid-1960's) it arranged for the disposal of its chemicals. But it did know that the chemicals were a hazardous substance. Large drums of industrial chemicals are inherently hazardous—the court described the chemicals as "unstable" and noted that the drums contained other chemicals, as well as "contaminants" and oil. An internal memo from GE noted that the plant was being overrun by drums of its waste chemicals. Given the "potential health hazards" of their chemicals, the memo asked for legal advice about disposal of the chemicals in landfills. Id. at 386 n.8. Thus, it was fairly obvious that the drums of chemicals being disposed of by GE were hazardous, even if the specific nature of the PCB problem was not then appreciated.

By contrast, there is nothing "inherently" hazardous about the bales of paper scraps ACPC sold to brokers. No one was at all concerned about health or environmental issues at the time, and thus the desire to rid itself of a hazardous product is wholly absent in this case. Even if ignorance about the hazardous nature of a product is not dispositive of arranger liability, I conclude that it is at least strong evidence suggestive of the disposer's lack of the requisite intent.

In sum, I conclude ACPC was not an arranger because it: (1) lacked knowledge that broke, a byproduct of its manufacturing process, could be hazardous; (2) invested money and labor in treating, sorting and selling its broke; (3) always sold the broke, rather than sometimes sending it to a landfill or otherwise disposing of it; (4) sold the broke through brokers in a well-established secondary market; and (5) treated the broke, for accounting and other purposes, as an asset that was integral to its business model. For these reasons, the broke had far more characteristics of a useful product than of a waste product, and ACPC was thus indifferent, at best, about what might happen to the broke waste products after the broke was recycled. Although ACPC obtained some benefit by removing the broke from its facility, that was outweighed by the financial benefit it obtained by consistently treating broke as a valuable product with a known and predictable market.

D. NCR's Liability

Defendants also argued that NCR itself (in addition to ACPC) is liable as an arranger because it knew that the paper ACPC coated would be repulped and that the emulsion would need to be removed as a part of that process. This argument was a late addition to the trial, but I allowed Defendants to offer evidence on their theory given the lack of prejudice I found to NCR.

Even accepting that NCR itself might have had some appreciation for the papermaking and recycling process, this is a case of "knowledge alone." Burlington Northern, 559 U.S. at 612 ("knowledge alone is insufficient to prove that an entity 'planned for' the disposal, particularly when the disposal occurs as a peripheral result of the legitimate sale of an unused, useful product.") In sending its emulsion to ACPC for it to be coated on paper, NCR was not "arranging" for the disposal of its emulsion. Where ACPC was at least getting rid of its product, in some sense, NCR was merely selling a useful product to be applied by another company and then sold back to NCR. There was almost no aspect of "disposal" at all.

The other aspect of Defendants' argument that NCR is itself liable as an arranger involves their theory that NCR and ACPC were engaged in a toll manufacturing arrangement, whereby NCR controlled the entire process from beginning to end. But even if NCR exercised control over the manufacturing process of its carbonless copy paper (for example, by imposing strict production standards and enforcing its intellectual property rights), there was no evidence that it had anything whatsoever to do with the disposal of ACPC's broke. Selling the broke was something that was strictly within ACPC's purview. Moreover, as noted above, there was no evidence that NCR had any purpose in selling its emulsion to ACPC other than to produce a commercially viable product. Broke was simply not part of the equation. If ACPC is not itself an arranger, the arranger liability case against NCR is even weaker. In short, I found no credible evidence that NCR was itself an arranger.

III. Costs

In addition to the arranger liability issues set forth above, the trial addressed some aspects of costs that were not decided at the summary judgment stage.

A. Fox River Group (FRG) Damages

1. FRG's NRD Costs

Plaintiffs challenge the Defendants' ability to recover certain costs the Fox River Group (the Group) incurred , as part of an agreement with the State of Wisconsin, to investigate the nature of any natural resources damages (NRDs). They argue that these NRD costs, which amount to nearly $9 million, are not recoverable because the information the Group ultimately obtained was not relied upon by the authorities, who did not respond to or comment on the NRD work. Plaintiffs' expert Robert Rock testified that there was no evidence that the NRD work was necessary for the effort to remediate the pollution. In his view, "it does not appear that . . . this information ended up actually being utilized in the final natural resources work here." (Tr. 1219:7-9.) From his review, Rock concluded that the work done was part of an effort to persuade the government to pursue a lower-cost alternative remediation strategy, but from what he could tell the government had already made its mind up to reject the FRG approach. (Tr. 1219.)

Although Plaintiffs continue to dispute my conclusion, I have found that Defendants may recover overpayments for natural resources damages. (Dkt. No. 1191.) What constitutes proper NRD costs is an issue of first impression, however. Plaintiffs cite regulations governing best procedures for such assessments, but I am not satisfied that a PRP's effort to recover overpayments for NRD costs is an invitation to grade the efficacy of an NRD assessment, so long as the assessment was a bona fide effort. In other words, the question should be whether the assessment was a reasonable effort and expense undertaken at the time and under the circumstances, rather than an ex post facto review of those procedures some fifteen years later. The regulations cited by Plaintiffs, 43 C.F.R. § 11.10, are not mandatory, and thus the fact that the NRD assessment did not strictly comply with the regulation should not be fatal to the recovery of those expenses. Moreover, the fact that the government ultimately rejected the undertaking does not mean it was not a legitimate expense or an unworthy effort. By Plaintiffs' own admission, the NRD effort was an attempt to lower the cost of remediation which, if successful, would have redounded to the benefit of all PRPs. In sum, there is no evidence that the NRD assessment expenses were "unnecessary" in the sense that they were some sort of frivolity undertaken with no chance of influencing the relevant decisionmakers. Instead, the evidence shows that at the time the expenses were incurred, they were legitimate and reasonable. Accordingly, I conclude that Defendants are entitled to recover them.

2. Fox River Group SMU 56/57A Landfill Work

Plaintiffs argued that the Defendants' analysis included nearly $1.8 million in costs that the FRG never actually incurred. Defendants have now conceded that there was an accounting error. Accordingly, that amount will be deducted from the Defendants' recovery.

3. FRG Assessments

Defendants argue that this is the time to make a "final allocation" of Fox River Group assessments, as provided in the parties' 1999 Allocation Agreement. As already suggested in this Court's denial of leave to amend (Dkt. No. 1294), however, claims under the parties' contract are not a part of this CERCLA action. The discovery and trial were directed at answering the novel and fact-intensive issues that arise under that statute, not at litigating an allocation under a private contract. Accordingly, I will not make an allocation under that agreement.

B. Georgia Pacific Damages

1. Cell 12A Landfill

Plaintiffs argue that the costs related to GP's landfill in Cell 12A should be disallowed because GP had already been reimbursed by the FRG for those costs in the form of a credit. But Defendants' expert, Arthur Vogel, an environmental lawyer with Quarles & Brady, testified without contradiction that the $2 million amount was not a credit to GP in the traditional sense. It merely reflected that GP had contributed its own landfill to the FRG effort and the FRG treated that contribution as a credit, of sorts. "It was treated by the group as an additional cash equivalent payment to the Fox River Group activities." (Tr. 714:8-10.) He further testified that the FRG did not reimburse or otherwise pay GP for the landfill. (Tr. 714:24.)

Plaintiffs argue that Vogel's testimony should be disregarded because it is contrary to a pretrial stipulation. (Dkt. No. 1320, Sec. C ¶ 68.) I am satisfied, however, that Vogel's testimony adequately explains a situation that likely proved confusing to all sides, namely, the use of the term "credit." It also makes more sense that the FRG would have given GP a "credit" for its contribution in the sense Vogel described rather than paying it cash for the contribution. Accordingly, I will not reduce GP's reimbursement by the nearly $2 million Plaintiffs seek.

2. Vinland Landfill

Beginning in 2003, GP investigated whether its pre-existing landfill in the Town of Vinland could be re-permitted such that it could accept PCB waste. In doing so, it incurred some $183,000 in legal fees and payments to the Town. In Plaintiffs' expert's opinion, these expenses were not necessary and should not be subject to contribution, primarily because the landfill was never used to dispose of PCB waste.

As above, however, the fact that the landfill was ultimately not used for PCB waste does not, in itself, make the expenditures unreasonable or unrecoverable. GP's expert was a chemical engineer with thirty-five years' experience in the waste disposal industry. He testified that it was common, particularly with a large site like the Fox River, for PRPs to evaluate multiple sites for possible disposal. (Tr. 650.) Such efforts necessarily involve expenditures for permitting, such as attorney's fees.

But it is clear that the landfill was proposed to be used at least primarily for OU1 waste from Glatfelter and WTM. (Tr. 641-644.) Vinland is a town a few miles south of OU1 (Little Lake Butte des Morts), and Glatfelter and WTM were looking for an area to deposit OU1 waste. Under other circumstances I might agree with GP that its costs are recoverable, but here, having found that Plaintiffs are not liable for OU1 costs, it would not make sense to award GP costs that were primarily incurred with respect to OU1 waste.

3. NRD Commissions

Plaintiffs also challenge payments of some $41,000 GP seeks for commissions it paid for land transfers under a 2002 Natural Resources Damages settlement. Although GP had documentation for three properties, it lacked such documentation for three others. It argues that it is not surprising that records might not exist for transactions occurring a decade ago. Although that might be true, I am satisfied that Plaintiffs are correct that more is required than GP has been able to provide. Absent any evidence that the commissions were actually paid, I will decline to allow GP to be reimbursed for these payments.

4. PricewaterhouseCoopers Payments

Plaintiffs also contest GP's payments to the PricewaterhouseCoopers accounting firm totaling some $96,000. Plaintiffs' expert found these payments "just an area of confusion." (Tr. 1217:5.) He noted that GP's explanation had changed: one time the payments were on behalf of the government, and another time they were for land valuation services. The invoices, however, state that they are for natural resource damage assessment relating to alleged contamination of the Fox River and possible settlement of Fort James [now GP] liability. (Ex. 451.) I am satisfied that these are recoverable costs and that Plaintiffs' expert's confusion about them is not enough to justify withholding them.

5. Consent Decree Payment

GP paid $7 million pursuant to a 2010 consent decree. Such payments are presumptively consistent with the National Contingency Plan and thus recoverable under CERCLA. The twist here, however, is that GP made the $7 million payment as part of a settlement with the government after this Court had already ruled, in 2009, that the Plaintiffs were wholly liable for the cleanup effort. With that background, Plaintiffs note, GP was essentially spending Plaintiffs' money to settle its liability with the government because GP knew it would just turn around and seek reimbursement from Plaintiffs for that money. In other words, GP obtained a significant benefit from the government settlement (contribution protection), but it did so on NCR's dime. Allowing it to recoup the $7 million from NCR would thus not be equitable.

Although the Plaintiffs' objection has some ostensible appeal, I conclude that there is nothing inequitable about awarding GP its consent decree payment, as in any other case. Importantly, although by the time the decree was lodged I had ruled against Plaintiffs in their contribution action, I had not ruled in GP's favor on its own contribution counterclaims. As noted at some length in that decision, it was not a simple matter of rubber-stamping the counterclaims as though the result in GP's favor followed automatically from my earlier decision. (Dkt. # 1080.) Thus, when GP and the government entered into their settlement, GP had no assurance that it would have any entitlement to recoup the funds it paid as a part of that settlement. Accordingly, I find no reason to deny GP the ability to obtain compensation for its $7 million consent decree payment.

6. Interest

GP claims it is entitled to prejudgment interest of some $2.8 million, as calculated by its expert, Robert Zoch, from the date GP filed its counterclaim. Plaintiffs argue that no prejudgment interest is awardable because GP never made a demand for a "specified amount." 42 U.S.C. § 9607(a)(4) ("interest shall accrue from the later of (i) the date payment of a specified amount is demanded in writing, or (ii) the date of the expenditure concerned.")

Although it is true that GP did not demand a "specified amount," that is seldom possible in the CERCLA universe. Interpreting this statute, the Fifth Circuit "set the bar for meeting the written demand requirement at a low level." Halliburton Energy Services, Inc. v. NL Industries, 553 F. Supp.2d 733, 768 (S.D. Tex. 2008) (citing In re Bell Petroleum Servs., Inc., 3 F.3d 889, 908 (5th Cir.1993)). The Fifth Circuit concluded that the complaint, which did not specify an amount of damages, was sufficient to meet the requirement of making a demand in writing. The same holds true here. GP filed a counterclaim putting Plaintiffs on notice of their demand for compensation, and there could be little doubt as to the magnitude of the counterclaim's potential. Accordingly, I conclude GP is entitled to the prejudgment interest it seeks.

7. Glatfelter and WTM Interest

In 2003 Glatfelter and WTM entered into a consent decree to perform OU1 cleanup work, and in accordance with that decree they deposited more than $80 million into an escrow account. Over time, that account generated significant amounts of interest (more than $4 million), which WTM and Glatfelter now want to obtain. Given my conclusion that Plaintiffs are not arrangers, however, Glatfelter's and WTM's OU1 payments are not recoverable.

IV. Trial Objections

Finally, I must rule on a number of objections made to some of the exhibits offered at trial. As a preliminary matter, it is worth pointing out that none of the exhibits described below were dispositive of my conclusion that ACPC was not an arranger. My conclusion was reached primarily on the basis of the testimony and facts, most of which were largely undisputed. For completeness, however, I will briefly address the specific objections Plaintiffs have raised.

A. Authenticity Objections

The proponent of an item of evidence has the burden of proving that it is authentic, in the words of Rule 901(a) of the Federal Rules of Evidence, that "the item is what the proponent claims it is." However, "Rule 901 does not erect a particularly high hurdle." Thanongsinh v. Board of Educ., 462 F.3d 762, 779 (7th Cir. 2006) (quoting United States v. Dhinsa, 243 F.3d 635, 658 (2d Cir. 2001)). The Rule's requirements are satisfied if evidence has been introduced from which a reasonable factfinder could find that the document is authentic. Id. An ancient document or data compilation is generally considered authentic if it "(A) is in a condition that creates no suspicion about its authenticity; (B) was in a place where, if authentic, it would likely be; and (C) is at least 20 years old when offered." Fed. R. Evid. 901(b)(8).

1. The "Heinritz Letter" (Ex. 2159 and 2208)

Defendants sought to introduce a one-page typed letter purportedly written by Fred Heinritz, of ACPC, to a Mr. A.E. Burroughs of Wiggins Teape in England. The letter explains which recycling mills were repulping ACPC's broke and speculates that these mills would be removing the coating "in a rather simple manner." Defendants offered the exhibit to demonstrate that ACPC knew its broke would be treated in some way so as to remove the coating. From this, one could draw the inference that ACPC appreciated the fact that the NCR coating it applied to the paper was removed and would be flushed out into the waterways.

Plaintiffs object on numerous grounds, but primarily on authenticity. They note that Mr. Heinritz has denied writing the letter. It is not on ACPC stationary and is unsigned. Moreover, it uses the British spelling of the word "favourably," which Heinritz would not have used. Defendants have explained that the letter is most likely a re-typed version that Wiggins Teape made for its own file. That would explain the lack of signature and the other unusual issues.

Although Plaintiffs have raised some legitimate objections to the document, I am satisfied that it is authentic. In essence, the document is so benign and perfunctory that it defies reason to imagine it could be anything but an authentic letter, even if it is a transcription. The document does not suggest that Heinritz had any kind of detailed understanding of the process for removing NCR coating, much less that he had any inkling of the final destination of that coating. Had anyone endeavored to manufacture a document, surely it would not have been a document so harmless, one that merely conveys the fairly obvious fact that a paper executive knew that in order to reuse his company's product, the coating his company applied to it would need to be removed. Thus, although I find the document admissible, I gave it little weight in reaching my conclusion.

2. California Tobacco Documents

A number of documents, designated with the GPFOX Bates identifier, came from an online database maintained by the University of California, San Francisco (UCSF) as part of what it calls a "Legacy Tobacco Documents Library." Although such a location appeared to be an unusual place to find documents about paper production in the 1960's, this is explained by the fact that Wiggins Teape was once owned by the British-American Tobacco Company. In 2002, UCSF acquired the documents from a depository in England. The documents have all the markings one would expect from ancient corporate records—including letterheads, prior Bates stamps, etc.—and, as with the Heinritz letter, it would defy reason to find the documents were not authentic. These documents were obtained by UCSF long before this litigation ensued and, in any event, they are not the sort of material one would expect to find if anyone wanted to manufacture evidence. Although I found them material in ruling against the Plaintiffs in their contribution claim at the summary judgment stage (where there had been no objection), there were no smoking guns or anything even remotely suggestive of fabrication. And, as suggested at the outset, none of these documents was particularly persuasive in suggesting arranger liability.

3. Exhibit 3067

This exhibit, a document produced from a third-party subpoena to Monsanto, is the report and minutes of a 1970 meeting between officials of Wiggins Teape, NCR and Monsanto, in which they discuss the transition from Aroclor as the solvent in NCR paper to another chemical. Among the reasons for the transition was the increasing awareness of the adverse health effects of the PCBs in the solvent. The last page of the exhibit is a crude diagram showing the path of Aroclor from start to finish, including the sewer and UK rivers.

Plaintiffs object on the grounds that the former Monsanto employee who authenticated the records had left the employment of Monsanto several years earlier and lacked personal knowledge of the documents' authenticity. I am satisfied, however, that the documents are authentic for the reasons noted earlier: they have all the indicia of being bona fide corporate documents from long ago, and they are not the kind of thing one would expect to be fabricated, particularly given that the party that produced the document is not a party in this action. Moreover, given the age of the document (forty years), it would be difficult to find a corporate employee with personal knowledge of the document's authenticity.

4. The Hoover Report (Ex. 2001 and 2949)

This is a twelve-page report authored by NCR's T.E. Hoover in 1972 entitled "The Status of Polychlorinated Biphenyl Uses at NCR." Plaintiffs object on the ground that Hoover, in a deposition in another case, denied writing the report. But the report comes from API's own files. Either a document is authentic or it is not. If not authentic, there must be some reasonable explanation for the document's existence, e.g., that the document is a forgery. Here, as the Defendants note, it is not credible that someone would create a fake report and then somehow sneak it into the Plaintiffs' own files. The document bears all the indicia of reliability and, as with the other documents, there is nothing about it suggestive of an effort to deceive. In short, as with the other documents, if anyone had a desire to manufacture evidence, surely they would have created something more compelling than a report that is actually largely favorable to the Plaintiffs' own position.

B. Hearsay Objections

Plaintiffs also raise a number of objections based on hearsay. Specifically, they assert that several documents, including some of those described above, contain hearsay statements within the document, which is itself hearsay. But, as noted at trial, the statements have not been offered—and I have not relied upon them—for the truth of the matters asserted therein. That is, even if a statement within a hearsay document might otherwise itself constitute hearsay within hearsay, that is not the case if the statement is offered, for example, merely to show that it was made (as opposed to that it was true). Here, the key issues involve what NCR knew or suspected about problems relating to PCBs. If NCR representatives were in a given meeting in which these problems were discussed, that suffices to place NCR on notice about those problems regardless of whether the underlying statements made in the meeting were actually true. Accordingly, I do not find that hearsay comes into play here.

V. Insurance Setoff Issues

The final question to be resolved at trial was the extent (if any) of setoff allowed to the Plaintiffs due to the large insurance (and other) settlements some of the Defendants have received. In other words, to the extent these Defendants might recover in contribution from Plaintiffs, that recovery would be limited by any payments the Defendants had received that were "for" the expenses they are now seeking to recover from Plaintiffs.

Georgia-Pacific and the Plaintiffs settled GP's insurance setoff issues during the trial. GP, being far downstream of OU1, did not have cleanup expenses relating to OU1 liability, and so the arranger issue addressed above did not affect its insurance settlements and the issue of setoff. Two of the other key Defendants, however—Glatfelter and WTM—were OU1 defendants, and thus the lion's share of the contribution they seek arose out of the OU1 consent decree and other OU1-related expenses. My conclusion that the Plaintiffs are not liable for OU1 would seem to render moot the question of setoffs because there will be no recovery from which anything can be set off. Accordingly, I will not address setoff issues with respect to these two Defendants.

That leaves Defendant CBC Coating. CBC Coating is a smaller mill located in OU2 (Appleton). It contributed $665,225 to the Fox River Group custodial account, and it seeks this amount, with interest, from Plaintiffs, for a total of some $1.16 million. It received insurance settlements totaling $1,432,000. Part of these proceeds resulted from a settlement with the insurer known as Carrier X, and the remainder resulted from two settlements with insurers it had sued in Illinois state court.

CBC argues that there should be no offset because it has not been made whole for its losses relating to the Site. Although its settlements exceed what it paid into the FRG account (including interest), it has incurred substantial attorney's fees. Moreover, some aspects of the settlements cannot be considered recoverable costs because they were not "for" the past cleanup costs CBC now seeks.

I agree with CBC that the majority of its settlement proceeds cannot rightly be considered "for" cleanup costs, and thus I agree that in most ways it is not seeking to recover again for payments it has already been reimbursed for. In fact, NCR's expert, Professor Abraham, agrees with this point as well. More than one million dollars of its recovery, he believes, is fairly attributable to CBC's Fox River defense costs, not cleanup costs. Moreover, half of the Carrier X settlement is not fairly attributable to cleanup costs.

But Abraham also noted that $150,000 of the settlement with CNA was to be applied to erode policy limits, meaning that it appears it was designed as an indemnity payment rather than as payment for attorney's fees or something else. And, as noted above, half of the Carrier X settlement (for a total of $83,000) should be deemed a payment for past cleanup expenses. Thus, although Abraham agreed that most of the payments were not "for" the same costs now being sought, he believed $233,500 should be offset.

I am satisfied that Professor Abraham's analysis is a conservative and reasonable approach to a difficult problem. It recognizes that the majority of insurance settlement payments are likely attributable to defense costs, which are not recoverable here, meaning that CBC is entitled to keep those funds. The other adjustments are a reasonable recognition that some of the payments were, in fact, attributable to the costs now being sought in contribution. CBC's position is that it has not yet been made whole on a "net" basis, which includes the significant attorney's fees it has paid (and will pay). Thus, there is no "double recovery." But although that may be true in some sense, the fact remains that it received large settlement payments with its insurers, and at least part of those settlements can be chalked up to funds that it seeks to recover from Plaintiffs. Accordingly, I believe an offset of $233,500 should be allowed.

VI. Remaining Issues

The trial, and this decision, do not quite end everything. As noted above, there may be some minimal insurance offset issues with respect to WTM and / or Glatfelter to the extent they had expenses outside of OU1. If so, it is hoped that a settlement can be reached on those points. In addition, the Defendants have state law counterclaims that are the subject of motion practice. Absent settlement, the counterclaims will be subject to the schedule set forth below.

The motion to file a reply brief [1355] is DENIED as moot. The motion for judgment on partial findings [1341] is DENIED. The motion in limine [1311] is DENIED. The motion to vacate the order staying briefing on Defendant's counterclaims [1244] is GRANTED. The initial brief has already been filed. A response brief is due 30 days from the date of this order. A reply brief, if any, is due 21 days following the filing of the response brief.

________________

William C. Griesbach

United States District Judge