The Comprehensive Environmental Response, Compensation, and Liability Act ("CERCLA"),1 also known as the Superfund law, imposes strict, joint and several, and retroactive liability.  Over time, many companies and their legal counsel have concluded that fighting a CERCLA claim, particularly one asserted by the government, is in most instances a waste of time and resources, and that therefore litigation resources are better allocated to settlement.  While the CERCLA case law is indeed very plaintiff-friendly, the scope of liability is not as limitless as the government would like potential defendants to believe.  There are certain types of activity that may contribute to the contamination of a site, but nevertheless do not give rise to CERCLA liability.

This piece the second in a series of three installments focusing on little-known legal defenses to CERCLA allegations that a defendant "arranged for the disposal or treatment" of a hazardous substance at a site. 42 U.S.C. §9607(a)(3).  Part 1 discussed the legal basis for asserting that legitimate recycling transactions are not arrangements for disposal.  Part 3 will discuss case law holding that the emission of airborne hazardous substances is not an arrangement for disposal of those substances where they land.

This installment outlines the legal basis for the argument that CERCLA liability does not attach to most wastewater discharges subject to the Clean Water Act's National Pollutant Discharge Elimination System ("NPDES") program, including discharges that violate permit requirements.

Contaminated Sediment Sites

The subset of Superfund sites that encompass bodies of water, such as rivers and harbors with contaminated sediments, represents a relatively small portion of all Superfund sites, but a disproportionately high portion of the most costly sites to address.  Such sites are often larger in size than the average Superfund site and cleaning up submerged contamination presents more costly engineering challenges.  The U.S. Government Accountability Office reported in 2016 that sediment cleanup cost estimates at the more complex contaminated sediment sites ranged from $500 million to over $1 billion each.  And these staggering amounts do not even include the costs attributable to large natural resource damage claims that often apply to such sites based on decades of impacts to publicly-owned flora and fauna.

Parties have been found to have "arranged for disposal" at contaminated sediment sites in a variety of ways, including through spills into the waterway, waste dumping, allowing contaminated stormwater to run off their property into the waterway, and via their routine discharges of industrial wastewater.  Since 1972, section 402 of the Federal Water Pollution Control Act (aka the Clean Water Act)2 has required many industrial facilities and municipal sewage treatment plants to obtain NPDES permits to discharge their wastewater to surface water bodies via designated point sources, also referred to as outfalls. 

CERCLA's "Federally Permitted Release" Exclusion

As noted in Part 1, CERCLA was enacted against a backdrop of existing laws regulating the release of pollutants to the land, water and air, creating the possibility that CERCLA and one of the existing regulatory programs, such as the Clean Water Act, would have duplicative jurisdiction over some releases.  CERCLA's drafters addressed this potential overlap by embedding the concept of "federally permitted releases," see 42 U.S.C. §9601(10), in the new statute and by providing that "[r]ecovery by any person (including the United States or any State or Indian tribe) for response costs or damages resulting from a federally permitted release shall be pursuant to existing law in lieu of [CERCLA]."  42 U.S.C. §9607(j).

CERCLA enumerates eleven categories of federally permitted releases, encompassing releases regulated by the Clean Water Act, the Solid Waste Disposal Act, the Clean Air Act and the Atomic Energy Act.  There are three separate exclusions related to the NPDES program.  The exclusion of interest here provides:

"federally permitted release" means . . . (C) continuous or anticipated intermittent discharges from a point source, identified in a permit or permit application under section 402 of the Federal Water Pollution Control Act, which are caused by events occurring within the scope of relevant operating or treatment systems. . . .3

The scope and meaning of the federally permitted release exclusion for wastewater discharges from a facility's NPDES outfall may be CERCLA's most misunderstood and misapplied provision.  As explained below, the quoted provision encompasses, and thus exempts from CERCLA liability, many (if not most) current industrial wastewater discharges, even if such discharges violate or are unregulated by the facility's NPDES permit.

In 1980, CERCLA's drafters lifted?literally word for word?the federally permitted release exclusions for NPDES discharges from a provision added to the Clean Water Act in 1978 to serve a similar "traffic cop" role, allocating jurisdiction between the existing NPDES permit program and the Clean Water Act's new "Section 311" program imposing liability for non-NPDES discharges of oil and hazardous substances.  Compare 33 U.S.C. §1321(a)(2) with 42 U.S.C. §9601(10)(C).  This was no accident.  Both CERCLA's legislative history and U.S. EPA's administrative interpretations confirm that the NPDES-related exclusions are to have the same meaning under both statutes.

Although EPA never adopted any regulations implementing CERCLA's federally permitted release definition, the agency did adopt regulations implementing the Clean Water Act version of the NPDES provisions in 1979.  See 40 C.F.R. §117.12.  With respect to the "continuous or anticipated intermittent discharge" exclusion, those regulations provide:

A discharge is a "continuous or anticipated intermittent discharge from a point source, identified in a permit or permit application under section 402 of this Act, and caused by events occurring within the scope of the relevant operating or treatment systems," whether or not the discharge is in compliance with the permit, if:

(1)    The hazardous substance is discharged from a point source for which a valid permit exists or for which a permit application has been submitted; and

(2)    The discharge of the hazardous substance results from:

(i)     The contamination of noncontact cooling water or storm water, provided that such cooling water or storm water is not contaminated by an on-site spill of a hazardous substance; or

(ii)    A continuous or anticipated intermittent discharge of process waste water, and the discharge originates within the manufacturing or treatment systems; or

(iii)   An upset or failure of a treatment system or of a process producing a continuous or anticipated intermittent discharge where the upset or failure results from a control problem, an operator error, a system failure or malfunction, an equipment or system startup or shutdown, an equipment wash, or a production schedule change, provided that such upset or failure is not caused by an on-site spill of a hazardous substance.4

This regulatory language was already on the books when CERCLA's federally permitted release provision was drafted, and the record is clear that Congress was aware of that interpretation and intended for it to apply to CERCLA as well,5 a fact that U.S. EPA has itself acknowledged.6

Looking at the practical effect of U.S. EPA's implementing language on a CERCLA defendant, note first that an excluded discharge of hazardous substances need not comply with the facility's NPDES permit terms.  Indeed, there is no requirement that the hazardous substances are even regulated by the permit.  Accordingly, the received wisdom that only releases in compliance with an NPDES permit can qualify as "federally permitted releases"[7] is incorrect. 

Note next that the discharge need only satisfy two criteria; (i) it must have occurred at an outfall covered by an NPDES permit or permit application, and (ii) it must have resulted from any one of three types of operational events.  With the clearly stated exception of accidental spills, the three categories of covered causes encompass most of the wastewater management events that will be encountered at an industrial facility.  Thus, in most instances, hazardous substances released via a facility's NPDES outfalls should not give rise to liability if the receiving water is someday addressed under CERCLA.

Practical Considerations

Due to the very nature and purpose of CERCLA's federally permitted release provision, the defense will not reduce a company's potential civil penalty exposure under the Clean Water Act for violations of its NPDES permit.  The current per-day maximum civil penalty, which adjusts annually for inflation, is $55,800 per day of violation, but actual penalties are rarely imposed at the maximum rate.  Moreover, a discharger's NPDES liability is limited to discharges that occurred within a five-year period preceding the filing of a civil action.

In addition, the NPDES federally permitted release defense only extends to wastewater discharges from a facility subject to the NPDES program, which did not take effect until 1972.  To the extent that CERCLA liability is based on discharges that occurred before the facility submitted its first NPDES permit application, the exclusion will not apply.  Given CERCLA's unlimited retroactive reach, sediment requiring cleanup at many Superfund sites was contaminated by discharges that occurred well before the NPDES program existed.  For example, the manufacture and use of polychlorinated biphenyls ("PBCs"), one of the most common contaminants at contaminated sediment sites, began in the U.S. in 1929.  Heavy metals, like lead, were widely used in manufacturing processes even earlier. Thus, the exclusion's inherent temporal limitation to post-1972 discharges will effectively take the defense off the table, in whole or in part, at some Superfund sites.


The NPDES federally permitted release defense will only be applicable to some industrial wastewater discharges (i.e., post-1972) and only at certain Superfund sites (i.e., contaminated sediment sites).  However, these types of sites generally pose the greatest monetary exposure.  Moreover, CERCLA's joint and several liability scheme often shifts burdens attributable to the long-ago releases of now defunct companies onto viable defendants with the most recent and easily proven connections to a site.  In the right circumstances, this defense might "flip the script" by providing complete CERCLA liability protection to a defendant whose corporate presence at a site occurred entirely during the NPDES era.


1 42 U.S.C. §9601, et seq.

2 33 U.S.C. §1342.

3 42 U.S.C. §9601(10)(C).

4 40 C.F.R. §117.12(d) (emphasis added).

5 See 98 S. Rep. 848, at 47 (1980) ("The first three parts of the [federally permitted release definition] carry the [1978 Clean Water Act] amendments forward without change.  Thus, the explanation at that time and the implementing regulations of the Environmental Protection Agency are continued").

6 In a 1994 policy statement, U.S. EPA confirmed that these Clean Water Act regulations also define the scope of the analogous CERCLA exclusions.  See U.S. EPA, Policy Statement on Scope of Discharge Authorization and Shield Associated with NPDES Permits at 4 (July 1, 1994), available at

7 See, e.g., United States v. Washington State Dept. of Transportation, 716 F.Supp.2d 1009, 1016 (W.D. Wash. 2010) (denying summary judgment on defense because "there is a dispute as to whether WSDOT is in compliance with the permits.")

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.