NCR Corporation recently avoided joint and several liability under CERCLA for its contribution to PCB contamination in the Lower Fox River. In United States v. NCR Corp., No. 10-C-913, 2015 WL 2350063 (E.D. Wisc. May 15, 2015), a Wisconsin District Court held that recent 7th Circuit guidance in the wake of Burlington Northern & Santa Fe Railway, Co., v. United States, 556 U.S.599 (2009) required a volumetric apportionment of NCR’s liability.

The case is the latest installment in a long-running CERCLA saga over PCB contamination in the Lower Fox River. The case made its first trip to the 7th Circuit when NCR decided it had done more than its fair share of the cleanup and ceased complying with a unilateral administrative order issued by the EPA. The EPA brought an action to enforce its order and obtained a preliminary injunction forcing NCR to continue the remediation. NCR appealed the preliminary injunction, arguing that the harm was divisible.

On the limited record before it, the 7th Circuit in United States v. NCR Corp., 688 F.3d 833 (7th Cir. 2012) believed that remediation of the river would be required where PCBs in the river exceeded 1.0 ppm and that even without contributions from other potentially responsible parties (“PRPs”), NCR’s PCB releases would have caused the remediation. The 7th Circuit acknowledged that remediation costs themselves were not necessarily equal to the harm caused by the pollution; nonetheless, NCR should be jointly and severally liable because its contamination alone would have triggered remediation.

After the 7th Circuit affirmed the preliminary injunction, the case was tried on the merits and a fuller record was developed. After trial, the District Court found the harm to be indivisible and again held NCR jointly and severally liable for the entire cleanup. United States v. NCR, Corp., 960 F. Supp. 2d 793 (E.D. Wisc. 2013). NCR appealed again, and the 7th Circuit vacated the District Court’s divisibility ruling. United States v. P.H. Glatfelter Co., 768 F.3d 662 (7th Cir. 2014). The 7th Circuit chalked up its prior decision to a “sparse record,” which led it to believe that the contamination of the Lower Fox River was “binary” in nature (i.e., PCB contamination above 1.0 ppm was harmful, but PCB contamination below 1.0 ppm was not).

Evidence developed at trial undercut this “binary” theory and showed that the harm was “continuous” in nature. Rather than requiring remediation for any section of the river where PCBs exceeded 1.0 ppm, the EPA’s remediation goal was to achieve a surface-weighted average PCB concentration of 0.25 ppm throughout the river. In some areas of the river, active remediation would be used to achieve the remedial goal, while in others, the goal would be achieved through natural attenuation. Some areas with concentrations below 1.0 ppm PCB (and indeed some areas with concentrations below 0.25 ppm) still posed a threat to human health and the environment. Thus, in Glatfelter, the 7th Circuit reversed course and held that its prior binary divisibility analysis was incorrect because the facts now showed that increased concentrations of PCBs (above or below 1.0 ppm) equated to increased harm.

On remand, the District Court viewed Glatfelter to require divisibility of harm based on the amount of PCBs released by each PRP. United States v. NCR Corp., No. 10-C-913, 2015 WL 2350063 (E.D. Wisc. May 15, 2015). The District Court found that the actual toxicity of the release was a better measure of the harm than remedial costs caused by each PRP. Thus, the District Court abandoned its complex remediation-based approach to divisibility because the 7th Circuit had “open[ed] the door to a simple volumetric approach to divisibility.” Id. at *3.

After finding the harm to be theoretically capable of apportionment based on volume because each additional quantity of PCB roughly correlated to an additional amount of harm, the court turned to the question of whether NCR had presented evidence that could allow the court to actually apportion the harm. Finding that Burlington Northern and Glatfelter required reasonableness rather than scientific precision in the apportionment analysis, the court used ballpark estimates based on averages to determine that NCR should only be liable for a portion of the cleanup. Id.at 8–9.

This case may signal a relaxation of the hesitancy to apportion harm under CERCLA and could allow PRPs to more frequently use estimates of the volume of their individual contributions to comingled contamination to avoid joint and severability under CERCLA — at least where volume can be reasonably linked to actual toxicity.