A popular concept in renewable-energy-project siting is the use of contaminated properties for potential projects. EPA’s initiative, “RE-Powering America’s Land,” reflects the federal government’s encouragement of this concept. But the use of contaminated land presents its own risks to developers, industry, property owners, and state and local governments. Under the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (“CERCLA”), owners and operators of contaminated land may be held jointly and severally liable for cleanup costs related to the contamination at the site. Accordingly, developing renewable energy projects on contaminated sites has unique risks—including potentially large and difficult-to-quantify cleanup costs.
With this in mind, a brief overview of the law of CERCLA joint and several liability is appropriate—and timely. May 4 marked the three-year anniversary of the U.S. Supreme Court’s decision in Burlington Northern & Santa Fe Ry. Co. v. United States, 556 U.S. 599 (2009). This decision, which held that apportionment was proper whenever there is a “reasonable basis for determining the contribution of each cause to a single harm[,]” sent shockwaves across the legal landscape and immediately provoked impassioned responses from the legal community. Some commentators broadly declared that joint and several liability under CERCLA was “dead”; still others bemoaned the Court’s affirmation of “bad math”  and predicted the demise of the role of science in apportionment analyses. But the three years after Burlington Northern have demonstrated that joint and several liability under CERCLA is alive and well. An examination of lower court opinions since Burlington Northern reveals that many courts have not drastically changed analysis of joint and several liability since the Supreme Court’s apportionment holding, and defending against joint and several liability after Burlington Northern may only be marginally easier, if different at all.
Much of the controversy surrounding the apportionment holding stems from the Court’s endorsement, as reasonable, of what many commentators consider a rather crude and inaccurate methodology. Based on this methodology, the Supreme Court affirmed the district court’s finding that joint and several liability did not apply, because apportionment of the harm was possible. Of additional note is the fact that the trial court in Burlington Northern undertook its analysis sua sponte—and the Supreme Court did not object—in what Justice Ginsburg described as a “heroic labor” where “[n]either party offered helpful arguments to apportion liability.”
The district court’s apportionment methodology in Burlington Northern was relatively basic. It involved three variables: (1) the percentage of surface area that the defendant Railroads owned, in relation to the entire contaminated site; (2) the length of time that the Railroads leased its property to an agricultural chemical distribution business; and (3) the percentage of contamination that this particular parcel of property was responsible for, in relation to the entire contaminated site. Percentage of surface area owned was 19%; length of time owned was 45%; and percentage of contamination contributed was 66%, or two-thirds. Multiplying these three variables together and adding a (somewhat arbitrary) 50% margin of error, the district court concluded that the Railroads’ liability was 9%.
But district courts have not treated Burlington Northern as the sea change some commenters suggested it would be. For example, the district court in Appleton Papers Inc. v. George A. Whiting Essay Co. was among the first courts to address Burlington Northern’s apportionment holding, and its application of Burlington Northern was relatively circumscribed. While acknowledging that Burlington Northern was a “watershed” apportionment case regarding “evidentiary” issues,  it downplayed the reach of the decision: “The Court’s holding on apportionment was actually quite limited . . . [and it] was merely holding that the lower court could rightly consider such things as the physical surface areas of the damaged land, the length of time over which the pollution occurred, and the areas where the pollution was released.” This sentiment did not change in subsequent months, as many lower courts limited Burlington Northern to its facts, and in one case, characterized it as “simply reiterat[ing] the law.”
Some courts have rejected simple analogies to Burlington Northern—that is, when the defendant seeks only to mirror Burlington Northern’s apportionment methodology by providing evidence of space, time, and contamination without providing evidence that supports the underlying assumptions that make these variables and methodology relevant. For example, some courts have required evidence that releases of contamination was steady over the time at issue or some evidence of how contamination migrated or dispersed over time before even considering a theory of divisibility based on percentage-of-time owned, which is similar to the methodology used in Burlington Northern. Even when a defendant makes relatively painstaking efforts to avoid joint and several liability by presenting multiple alternative methods for apportionment, the presumption of joint and several liability is difficult to overcome. Apportionment still carries a “heavy burden.”
Renewable energy project developers should be aware of the risks involved in locating a project on a brownfield site—first to manage the risk of incurring liability for cleanup costs under CERCLA and second to be cognizant the risks that other stakeholders must face. Although the EPA appears actively interested and engaged in the redevelopment of brownfields for renewable energy projects, developers should exercise particular caution when considering these sites. Many brownfield cleanups remain complicated and expensive. Even if the project developer is not liable for cleanup costs under CERCLA, it may be nonetheless involved substantively in the cleanup process. While the project developer might not be directly dealing with these issues of joint and several liability, other critical stakeholders may be. Notwithstanding the early reactions to the Supreme Court’s decision in Burlington Northern, the complex and often high-stakes cost apportionment issues involved in brownfields redevelopment remain little changed, and joint and several liability remains difficult to overcome.
 Pub. L. 96–510, 94 Stat. 2767 (codified as amended at 42 U.S.C. §§ 9601–75).
 Burlington Northern & Santa Fe Ry. Co. v. United States, 556 U.S. 599, 614 (2009).
 See, e.g., John M. Barkett, Burlington Northern: The Super Quake and Its Aftershocks, Chemical Waste Litig. Rep.: Interim Rep. (May 15, 2009).
 See Evansville Greenway & Remediation Trust v. S. Ind. Gas & Elec. Co., 661 F. Supp. 2d 989, 1012 (S.D. Ind. 2009) (“The full import of Burlington Northern is hotly debated, in this case and elsewhere.”).
 See, e.g., Robert C. Cook, Superfund: Government Threat of 100 Percent Liability ‘Pretty Much Dead’ After Burlington Decision, 41 Bureau Nat’l Affairs Env’t Rep. 216 (Jan. 29, 2010).
 See, e.g., Nicholas J. Houpt, Bad Math in CERCLA Apportionment: The Untold Tale of Burlington Northern, Columbia J. Envtl. L. Field Reps. (Oct. 17, 2010). See also Walter Mugdan, The Burlington Court’s Flawed Arithmetic, 40 Envtl. L. Rep. News & Analysis 10637 (July 2010).
 See, e.g., Mark R. Misiorowski & Joel D. Eagle, The Diminishing Role of Science in CERCLA After Burlington Northern & Santa Fe, 41 Bureau Nat’l Affairs Env’t Rep.1205 (May 22, 2009).
 See, e.g., Michael K. Foy, From Chem-Dyne to Burlington Northern: Apportioning Cleanup Costs in the New Era of Joint and Several CERCLA Liability, 51 Santa Clara L. Rev. 625, 625–50 (2011).
 See Burlington Northern & Santa Fe Ry. Co. v. United States, 556 U.S. 599, 606, 619 (2009).
 Id. at 622 (Ginsburg, J., dissenting).
 Id. at 615 (Stevens, J., writing for the majority).
 No. 08-C-16, 2009 U.S. Dist. LEXIS 111648 (E.D. Wis. Nov. 18, 2009).
 Id. at *9 (emphasis in original).
 Id. at *8–9.
 United States v. Iron Mountain Mines, Inc., No. 91-0768-JAM-JFM, 2010 U.S. Dist. LEXIS 44331, at *8 (May 6, 2010). See also, e.g., Ashley II of Charleston, LLC v. PCS Nitrogen, Inc. (PCS Nitrogen II), No. 2:05–cv–2782–MBS, 2011 WL 2119256, at *44 (D.S.C. May 27, 2011) (“[T]he Supreme Court’s approval of a fifty percent margin of error in the Burlington Northern case was fact-specific and did not indicate that a fifty percent margin of error will always be appropriate in apportionment calculations.”).
 See, e.g., Pakootas v. Teck Cominco Metals, Ltd., No. CV-04-256-LRS, 2012 U.S. Dist. LEXIS 47889, at *47–49 (E.D. Wash. Apr. 4, 2012); United Alloys, Inc. v. Baker, No. CV 93–4722 CBM (Ex), 2011 WL 2749641, at *21–22 (C.D. Cal. Jul. 14, 2011); United States v. NCR Corp., No. 10–C–910, 2011 WL 2634262, at *7 (E.D. Wis. Jul. 5, 2011); Bd. of Cnty. Comm’rs v. Brown Groups Retail, Inc., No. 08-cv-00855-LTB-KMT, 2011 WL 816792, at *22–24 (D. Colo. Mar. 3, 2011); ITT Industries, Inc. v. BorgWarner, Inc., 700 F. Supp. 2d 848, 877–81 (W.D. Mich. 2010); United States v. Saporito, 684 F. Supp. 2d 1043, 1061–62 (N.D. Ill. 2010); 3000 E. Imperial, LLC v. RobertShaw Controls Co., No. CV 08-3985 PA (Ex), 2010 U.S. Dist. LEXIS 138661, at *24–32 (C.D. Cal. Dec. 29, 2010); Ashley II of Charleston, LLC v. PCS Nitrogen, Inc. (PCS Nitrogen I), No. 2:05-cv-2782-MBS, 2010 U.S. Dist. LEXIS 104772, at *116–33 (D.S.C. Sept. 30, 2010).
 See, e.g., 3000 E. Imperial, No. CV 08-3985 PA (Ex), 2010 U.S. Dist. LEXIS 138661, at *31–32; PCS Nitrogen I, No. 2:05-cv-2782-MBS, 2010 U.S. Dist. LEXIS 104772, at *126–27.
 PCS Nitrogen II, No. 2:05–cv–2782–MBS, 2011 WL 2119256, at *43–48 (D.S.C. May 27, 2011) (rejecting five different methods for apportionment).
 Ashland Inc. v. GAR Electroforming, 729 F. Supp. 2d 526, 548 (D.R.I. 2010) (comparing the difference in the burden of proof between claims for section 107 apportionment and claims for section 113 apportionment). Burlington Northern did not shift the burden of proof to plaintiff—joint and several liability remains the default rule. See ITT Corp. v. Borg-Warner Inc., No. 1:05-CV-674, 2009 U.S. Dist. LEXIS 75637, at *10 (W.D. Mich. Aug. 25, 2009) (“The ultimate burden of proving divisibility is on the party invoking the doctrine.”).