Ronald SCHMUCKER, et al., Plaintiffs-Appellants, v. JOHNSON CONTROLS, INC., and Tocon Holdings, LLC, Defendants-Appellees.
Between 1937 and 2006 Johnson Controls and a predecessor operated a manufacturing plant in Goshen, Indiana. The plant used chlorinated volatile organic compounds in its degreasing agents, some of which reached the groundwater. Chlorinated organics slowly break down by losing chlorine atoms. The version with three chlorine atoms, known as trichloroethylene or TCE, is a carcinogen. The end product with no chlorine atoms, ethene, is harmless. The breakdown process can take decades, and a plume of TCE remains in water under part of Goshen. Plaintiffs contend in this suit under 42 U.S.C. § 6972(a), part of the Resource Conservation and Recovery Act (RCRA or the Act), that Johnson Controls and Tocon Holdings (which bought the land in 2007) must do more to reduce the amount of TCE in the environment. For simplicity we refer to both defendants as Johnson Controls.
Johnson Controls, under the supervision of Indiana's Department of Environmental Management (IDEM or the Department) started remedial operations while the plant was still operating. Johnson Controls worried that houses above the plume might draw contaminated water from wells, so in 1992 it ensured that all of these houses were connected to Goshen's water mains, and Goshen ordered all private wells closed. Then it began a “pump and treat” procedure, which ran from 1994 to 2012. It pumped groundwater out of the affected areas, treated it to remove TCE and other contaminants, and injected the water back into the ground. This system was discontinued only when treatment did not further reduce levels of TCE in the water.
The possibility that contamination might reach the aquifer from which Goshen itself draws water was investigated. An expert found that Goshen's municipal water comes from a direction different from the contaminated plume. That study has been repeated several times and the finding confirmed. To date there is no sign of TCE in the public water supply. Nor has any TCE been found in Goshen's high school, which sits just outside the affected zone.
But TCE did appear in the air over the plume—TCE is a volatile organic compound, after all. Johnson Controls installed vapor mitigation systems in all houses whose air showed unsafe levels of TCE. Like radon, gaseous TCE can enter a house through cracks in the foundation or basement walls; outside it mixes with and is diluted to insignificance by the air, but inside a house TCE can reach dangerous levels. The sort of system used to prevent a buildup of radon works with TCE too. Foundations are inspected and patched to seal cracks. Fans create an area under the foundation with pressure less than the air inside the house. TCE and other gasses migrate to the outside of the house, where they are dispersed, rather than entering through cracks.
After these systems were installed, a process completed in 2011, not a single house registered TCE levels exceeding the safe threshold set by the Department, and many tests have found no detectable TCE. One expert concluded that if 10 residents in houses over the plume stayed indoors constantly for 70 years, the probability of even one excess cancer is less than 0.02%. And no one stays in an unventilated house that long. A different expert estimated that, under realistic assumptions, it would be a million years before Goshen saw a statistically significant risk of cancer caused by TCE. And long, long before then, all of the TCE will have changed to ethene naturally.
The facts we have just stated come from extensive findings made by the district court after a bench trial. 477 F. Supp. 3d 791 (N.D. Ind. 2020). The opinion contains a wealth of detail that we have omitted in order to concentrate on the basics. Plaintiffs (whose three homes are located above the plume) do not contend that any of the factual findings is clearly erroneous—but they nonetheless argue that the Act requires Johnson Controls to do more.
The Act permits citizen suits in two pertinent circumstances:
(1)(A) against any person ․ who is alleged to be in violation of any permit, standard, regulation, condition, requirement, prohibition, or order which has become effective pursuant to this chapter; or
(B) against any person ․ who has contributed or who is contributing to the past or present handling, storage, treatment, transportation, or disposal of any solid or hazardous waste which may present an imminent and substantial endangerment to health or the environment[.]
42 U.S.C. § 6972(a). Plaintiffs say that Johnson Controls is in violation of § 6972(a)(1)(A) because it has not complied with the Environmental Protection Agency's closure regulation, 40 C.F.R. § 265.111. They contend that Johnson Controls is in violation of § 6972(a)(1)(B) because the risk to which they are exposed is a form of “endangerment”.
Only the second of these contentions went to trial. The district judge granted summary judgment to Johnson Controls on the first after finding that it does not even arguably violate “any permit, standard, regulation, condition, requirement, prohibition, or order”. 2019 WL 718553 at *––––, 2019 U.S. Dist. LEXIS 26657 at *45–86 (N.D. Ind. Feb. 19, 2019).
That conclusion is sound. The regulation on which plaintiffs rely says that a hazardous site must be closed in a manner that:
(a) Minimizes the need for further maintenance, and
(b) Controls, minimizes or eliminates, to the extent necessary to protect human health and the environment, post-closure escape of hazardous waste, hazardous constituents, leachate, contaminated runoff, or hazardous waste decomposition products to the ground or surface waters or to the atmosphere, and
(c) Complies with the closure requirements of [other regulations promulgated under the Act].
40 C.F.R. § 265.111. The district judge thought the suit premature, because the Johnson Controls site is a work in progress. And he added that plaintiffs have not pointed to “any permit, standard, regulation, condition, requirement, prohibition, or order” that Johnson Controls has violated. It can't have violated § 265.111 because that regulation lacks content. It just says that polluters must clean up “to the extent necessary” and comply with other requirements. Section 265.111 does not impose any requirements of its own.
Plaintiffs recognize that the regulation “does not define what constitutes clean closure or set clean-up levels that are ‘protect[ive] of human health.’ ” Brief at 23. To furnish what the regulation omits, plaintiffs turn to a memorandum entitled Risk-Based Clean Closure that the EPA's staff issued on March 16, 1998. This does not get plaintiffs anywhere, however. The statutory list of enforceable requirements is “permit, standard, regulation, condition, requirement, prohibition, or order”. These are documents with legal force. Memoranda are not on the list, and for good reason. They have not been issued after notice-and-comment rulemaking or through administrative adjudication. All a memorandum can do is express a point of view—or perhaps imply a threat that people who ignore the agency's goals or desires may be sued. Staff-written memos are not themselves a source of legal obligations, any more than an agency's brief in a suit could create legal obligations.
Risk-Based Clean Closure is not even addressed to businesses. It is a memo from Elizabeth Cotsworth, then the Acting Director of the EPA's Office of Solid Waste, to the EPA's regional senior policy advisers. It tells regional officials what senior staff is looking for, and it wraps up: “I encourage you to use risk-based approaches to develop site-specific clean closure requirements and to continue in your efforts to eliminate duplication of effort among cleanup programs.” This is like the directives to U.S. Attorneys in a Department of Justice manual, or like the Antitrust Division's merger guidelines. The document provides helpful information about what headquarters wants subordinates to achieve, but it does not create enforceable rules.
And that's all plaintiffs have, for the purpose of § 6972(a)(1)(A). It should be clear from what we have said that the suit is not premature. Plaintiffs rely on a regulation, which they say Johnson Controls is violating. Their problem is that Johnson Controls does not even arguably violate § 265.111, which lacks substantive content. Plaintiffs lose their § 6972(a)(1)(A) claim on the merits.
Plaintiffs' other claim, based on § 6972(a)(1)(B), went to trial. The critical language of this subparagraph is: “present an imminent and substantial endangerment to health or the environment”. Plaintiffs emphasize that “endangerment” means risk rather than certainty; pollution can endanger people even if precautions curtail danger. So we have held. See, e.g., Liebhart v. SPX Corp., 917 F.3d 952 (7th Cir. 2019) (Liebhart I). The district judge did not doubt this. What he said, however, is that the risk from TCE is neither “imminent” nor “substantial”. Plaintiffs accuse him of applying a “heightened” imminent-and-substantial standard, but we don't see how. The judge used those words in their natural sense. Imminent means soon, and substantial means, well, substantial (or sizeable, if you prefer a synonym). The judge found that risk from TCE in Goshen is neither around the corner (imminent) nor big (substantial)—and plaintiffs have not challenged any of the judge's factual findings.
Take for example the finding that the plume of water containing TCE is not likely to impinge on the aquifer from which Goshen draws its drinking water. This could happen if Goshen substantially increases the draw in a way that affects the direction of underground flow. That's true, almost by definition, but the district judge did not err in concluding that such a result is remote. The experts who tested the feed area of Goshen's water supply constructed a model under which the draw could treble, and they found that even on that assumption the plume of contaminated water would not affect the aquifer. The district judge agreed with the expert analysis. That conclusion does not reflect a legal error about the meaning of “imminent”.
Or consider plaintiffs' argument that vapor mitigation systems may fail, or that measurements have missed ongoing problems. Again both of these concerns reflect possible outcomes, but the district judge thought the risks insubstantial. The judge observed that many thousands of similar systems have been installed throughout the nation to control radon, and that their record of success is outstanding. Each system includes two fans, so they work even if one fails (and occupants would notice the silence if both failed), and they are inspected twice a year. Most houses in which a vapor mitigation system has been installed have been tested at least three times for TCE, without a single measurement coming in at an unsafe level.
To plaintiffs' argument that more measurements might have turned up evidence of indoor TCE, the district judge replied, in essence: true (by definition), but low probability. A finding of zero excess TCE over hundreds of measurements implies this. Plaintiffs have not submitted a statistical analysis showing that the number of measurements taken so far is too few to produce confidence that the systems work reliably. Nor have plaintiffs conducted extra measurements in their own homes. (If they have, they have not submitted the results to the court, which they would have done had dangerous levels of TCE been detected.) And the district judge added that disputes about how much TCE should be considered dangerous don't matter much, because the level of danger is negligible by any standard. Again there's no legal error in deeming these circumstances to fall short of “substantial” risk. See Liebhart v. SPX Corp., 998 F.3d 772 (7th Cir. 2021) (Liebhart II).
We could go on, but enough has been said. The district judge wrote much, much more, and the opinion shows compellingly why homeowners' risk from TCE in Goshen is neither imminent nor substantial. Plaintiffs lost this case on the facts, not on the law. If vapor mitigation systems begin to fail, or the contaminated water migrates toward the aquifer, or conditions otherwise change for the worse, plaintiffs will be free to renew their litigation. A conclusion that hazards are not “imminent and substantial” today does not mean that they will be slight forever. But the district judge did not err in concluding on this record that the risks are too slight to compel more action than Johnson Controls is already undertaking with Indiana's supervision.
Easterbrook, Circuit Judge.