|
30 ELR 10212 | Environmental Law Reporter | copyright © 2000 | All rights reserved Eight Basic Rules of Superfund Cost AllocationRichard Lane WhiteEditors' Summary: Can the complex process of allocating cleanup costs at Superfund sites be distilled into a set of practical and common-sense rules? This Dialogue, written by a co-author of an Article on Superfund allocation published in the February 1998 issue of ELR, submits that there are indeed eight fundamental rules to guide potentially responsible parties and regulators in moving from "Superfund soup" to allocations that are rational and equitable. The rules are derived from the statute, judicial interpretations, and EPA policies. By employing these rules, the author submits, what has been a highly controversial and mind-numbing procedure can be simplified, better understood, and consistently applied. Richard Lane White is a Vice President in the Cambridge, Massachusetts, office of PHB Hagler Bailly, Inc., an international management and economic consulting firm. A member of the firm's environmental and insurance coverage practice, Mr. White frequently writes on topics relating to Superfund and cost allocation. The views expressed here are those of the author, and not necessarily those of PHB Hagler Bailly, Inc. Allocating response costs among responsible parties at Superfund1 sites should be a simple, relatively straightforward issue—parties ought to be assigned the costs for which they are responsible. In reality, however, the allocation process is far from simple.2 All too often, the wastes that generators, transporters, and owners or operators contribute end up as a "Superfund soup" and it is virtually impossible to identify which costs are properly borne by individual parties.3 Almost 20 years after the passage of Superfund legislation,4 the allocation of response costs remains a highly contentious, hotly contested, and frequently debated issue.5 It remains contentious because there is a lack of agreement on the most basic issues, such as what is being allocated. And the reported decisions offer only limited guidance.6 And what is the objective of the allocation process—is what is allocated "harm" (however defined) or "response cost"? This Dialogue contends that despite the ongoing debate over the allocation objective there are several basic guiding principles, or rules, for the Superfund cost allocation process. In general, these rules, which are listed in Figure A, are not dependent on the methodology employed or even the objective of the process. They may be routinely applied regardless of actual methodology. The listing is by no means exhaustive, but does highlight several of the principles frequently encountered in the process. Several of the rules address how basic issues—wastes and costs, for example—fit into the allocation equation. Others can serve as benchmarks to evaluate the process in general, thereby shedding light on how the procedure should work independent of the methodology utilized. In addition, this Dialogue examines why the allocation standards are used and how they are actually employed. Basic Rules of Allocation Part 1: How to Allocate Waste Streams Quantification of waste streams is often the necessary, and unfortunately frequently contentious, first step in the allocation process. There are several basic rules that apply to the waste quantification process. One of the basic allocation steps in many instances is counting waste shipments and constructing a "waste-in" list. The initial question that arises is what waste to count and how to count it. The basic rule is that everything counts. [] CERCLA Is Not RCRA. What to count is a seemingly complicated issue. The Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) is not the Resource Conservation and Recovery Act (RCRA)7—what is considered a "hazardous substance" under CERCLA8 is not the same as a "hazardous waste" under RCRA.9 While everything that is a RCRA waste may be a CERCLA hazardous substance, the reverse is not true—the CERCLA statutory standard is much broader.10 Consequently, parties question whether, why, and how wastes either exempt or considered nonhazardous under RCRA should be addressed in the Superfund waste-in process. [] All Waste Has Superfund Liability. It is disposal, or more specifically activity related to the disposal of hazardous substances, that triggers CERCLA liability. And because under CERCLA the term "hazardous substance" has been defined so broadly,11 almost everything contains some quantity of a hazardous substance. Accordingly, a number of consumer products are subject to CERCLA because they contain hazardous substances, even though they are exempt from RCRA hazardous waste regulation or are considered to be nonhazardous waste.12 For example, the coins in one's pocket contain metals that are listed hazardous substances; if a penny is thrown into the the trash, it triggers liability if the refuse ends up at a Superfund site.13 Is that example so extreme as to be absurd? Perhaps, but what is equally unsettling is that parties continue to bicker over whether such an extreme example is in fact true. This issue was recently addressed by the Ninth Circuit in A & W Smelter & Refiners, Inc. v. Clinton.14 The court aptly described the peculiar implications of CERCLA: Read as the EPA suggests, CERCLA seems to give the agency carte blanche to hold liable anyone who disposes of just about anything. Drop an old nickel that actually contains nickel? A CERCLA violation. Throw out an old lemon? It's full of citric acid, another hazardous substance. It's not surprising that an agency would urge an interpretation which gives it such broad discretion. Perhaps more surprising is that CERCLA leaves us little choice but to agree.15 Furthermore, courts have repeatedly ruled that the quantity of hazardous substances is irrelevant when determining liability16; hence, the "molecule theory" of CERCLA liability.17 In effect, there is a bright-line standard for liability under [30 ELR 10214] CERCLA. Either a waste contains a hazardous substance, or it does not. It simply does not make sense for courts, allocation practitioners and individual responsible parties to argue over whether waste streams contain hazardous substances given that the term is so broadly defined so as to include nearly everything without respect to any quantification or threshold. As a practical matter, all waste contains hazardous substances—all waste counts. [] Waste—Not Merely Hazardous Substances—Should Be Counted. Should the waste itself, or alternatively its hazardous constituents be counted when compiling a waste-in list? Parties sometimes argue that the fraction of their waste that is not hazardous substances should not be subject to CERCLA liability, even though it is the waste itself that must be remediated. Where waste streams are relatively homogeneous, this is typically a non-issue. After all, whether the waste itself or some specific fraction of each waste stream is counted, the relative results are the same. This issue more frequently arises when waste types differ. Does one count a cubic yard of municipal solid waste (MSW), for example, or just the small fraction of hazardous substances in that waste? And then, how is a cubic yard of MSW compared to a cubic yard of industrial waste? The U.S. Environmental Protection Agency (EPA) has noted that "where hazardous and nonhazardous substances are mixed together, the mixture is considered hazardous and should be included in its entirety on waste-in and volumetric lists."18 In other words, the waste itself should be counted. A number of courts have also spoken on this issue. In Dartron Corp. v. Uniroyal Chemical Co.,19 for example, the court addressed the presumption that waste from an alleged waste oil generator contained hazardous substances: The [EPA] presumes to be hazardous wastes from the interior of a tank that held a petroleum product. The only way to rebut this presumption is to test the waste oil and show it does not contain contaminants. This presumption makes perfect sense, as it furthers the goals of CERCLA to achieve prompt cleanup of hazardous waste sites and to impose the cost of cleanup on those responsible for the contamination. If a person dumped waste oil onto a site that is later found to be contaminated with hazardous substances, that person cannot escape liability merely by arguing that there is no evidence the dumped waste oil actually contained a hazardous substance. Allowing such an argument would protect not only a simply negligent dumper of waste, but also a person who wilfully cast a blind eye to the possibility that the waste he was dumping contained dangerous contaminants.20 This issue involves more than simply ascertaining what to include in the waste-in list, however. It goes to the very core of the Superfund cost allocation debate. Is what is allocated the hazardous substances, or the waste itself? Again, EPA has spoken on the issue, in the MSW Settlement Policy: EPA's proposed methodology for calculating settlement offers to MSW G/Ts [generators and transporters] requires multiplying the known or estimated quantity of MSW contributed by the G/T by an estimated unit cost of remediating MSW at a representative MSW-only land-fill. This method provides a fair, reasonable and efficient means of completing settlements with MSW G/Ts that reflects a reasonable approximation of the cost of remediating MSW.21 EPA developed what was, in its view, the proper approach for allocating costs associated with the remediation of MSW at CERCLA sites. The cost share reflected the cost of remediating MSW, rather than some measure designed to address the fraction of hazardous constituents in the MSW. Because the remedy itself is what is being allocated, the waste-in list should count the waste, instead of the hazardous substances in the waste, and identify characteristics of the waste that can subsequentlybe used in the allocation process. When one considers alternative allocation methodologies, it becomes clear why it is appropriate to count the waste itself. For example, utilizing toxicity as an allocation factor involves an evaluation of the relative toxicity not of different constituents, but between waste streams. That MSW is less toxic than trichloroethylene (TCE) or other solvents is a given. But this is true not because the hazardous substances in MSW are "less" problematic than the hazardous substances in various solvents—often they are the same—but because the waste stream itself has less of them, may be less mobile, or does not leach as readily. If the molecules of hazardous substances were counted before toxicity is considered, the exercise would be largely redundant and would lead to nonsensical results. Likewise, cost causation, the allocation approach recently employed by EPA in its MSW Settlement Policy, estimates the remediation cost, i.e., the cost of cleaning up the waste. After all, the remedy employed remediates the waste, not the hazardous substances. If treatment costs for different hazardous substances were determined, very few of the actual remediation costs would be allocated because those costs are often associated with the mass of waste, instead of the fraction of hazardous constituents. [] "Contaminants of Concern" Are Not the Allocation Benchmark. When examining the various remedial alternatives parties often focus on the "contaminants of concern" (COC), which are often a subset of the hazardous substances found at a site. And while the COC may drive the remedy, they should not drive the allocation. All contributions count, regardless of whether they contain the COC. In developing the waste-in list and beyond that, the allocation itself, some allocators focus exclusively on the COC—as if those were the only hazardous substances brought to the site or the only [30 ELR 10215] materials that should be reflected in an allocation. For purposes of allocation, however, this focus is misguided. As the court in United States v. Atlas Minerals & Chemicals, Inc.22 noted: Third-Party Defendants urge the court to look instead to the administrative record to determine what contaminants influenced EPA's decisions concerning whether and to what extent remediation would be required at the Site. However, if the court followed this approach, the toxicity side of the equation would account only for those contaminants that were actually released. This approach would have the effect of rewarding polluters who disposed of hazardous substances that, by happenstance, were not released or detected during EPA's evaluation of the Site. Additionally, any hazardous substance present at the Site poses a threat to human health and the environment. CERCLA imposes liability if there is "a release, or a threatened release which causes the incurrence of response costs, of a hazardous substance. . . ." Whether a hazardous substance has been released at the time EPA undertakes its investigation is not determinative. The impermeable cap is designed in part to minimize the threat of future releases of hazardous substances. Thus, the "harm" that necessitates a clean-up includes the threat of future releases. Accordingly, the court accepts Third-Party Plaintiffs' argument that toxicity may be evaluated from an ex ante perspective.23 All waste counts. All hazardous substances count. As a practical matter, everything contains hazardous substances. When it comes to Superfund cost allocation, everything counts. Rule 2: Use Standard Protocols for Measuring and Converting Waste Streams Whatever waste shipments are to be counted must then be standardized. How does one compare a gallon of one waste with a pound of another? The basic allocation rule is that waste shipments brought to the site are presumed to be full, standard protocols for measuring shipments in cubic yards, tons, gallons, drums, and other common shipment sizes should be employed, and parties should be treated consistently. [] All Containers Are Presumed Full. As a general matter, allocation is a comparison of the relative contributions of the various parties—determining absolute quantities is of little value.24 In most cases a party is assigned a share, representing its fraction of all wastes at the site. The key allocation question is not whether one party's containers are or are not full, but how they compare, with respect to fullness, to the other parties' containers. Nevertheless, individual parties frequently assert that "my containers weren't full." And they are often frustrated when the allocation process ignores their assertion and evaluates wastes as if containers were indeed full. The working presumption that a container of waste is full is an assumption that EPA employs,25 and allocations routinely default to this assumption absent other information. Obviously, if the transporter or owner/operator treats one party's waste shipments differently than the other party's waste shipments, there may be a basis for an adjustment. Such treatment may be demonstrated by invoice descriptions or pricing patterns. But absent a reliable basis for making such a distinction, it is generally considered reasonable to estimate waste shipments using container capacity. [] Use Standard Measures for Waste Conversion. Standard measures can and should be used to recalibrate waste shipments to a common standard. Drums, barrels, truckloads, tanker trucks, and other shipment descriptions frequently have default measurements. For example, absent other information, a drum has a default measure of 55 gallons. Many of these "default measurements" are provided in EPA materials or have been previously cited in other works. Figure B lists standard conversion factors found from EPA's waste-in guidance.26 For some shipments the recalibration process is straightforward. For example, gallons and cubic yards are both "volumetric" measures. As such, a specific number of gallons make up a cubic yard. "Volume to volume" conversions can be easily calculated using published conversion data.27 Many solid wastes, however, are measured in tonnage. Weight-to-volume conversions are not as straightforward because the conversion depends on the density of the material. For example, a cubic yard filled with dirt weighs more than a cubic yard filled with foam rubber. But by volume, both are the same—one cubic yard. When waste types of differing densities are disposed of at the same site, it is quite common to calibrate the shipments for their relative density and several calibration methods exist. For example, one can convert all shipments to weight rather than volume, selecting default values for different wastes. Another alternative commonly used at landfill sites is to adjust volumetric data for density under landfill conditions: a cubic yard of lead may take up a cubic yard of space in a landfill, whereas a cubic yard of trash may compact to a fraction of a cubic yard. An adjustment can be performed while the waste-in list is being developed, but frequently it is completed as part of the allocation process itself. For example, a cubic yard of trash is heavy—several hundred pounds. But a cubic yard of trash that has been compacted is much denser, on the order of 500 to 700 pounds per cubic [30 ELR 10216] yard. Whether it comes to the landfill "loose" or pre-compacted, however, waste continues to compact once in a landfill. Mixed solid wastes under landfill conditions typically have densities on the order of 1,000 to 1,200 pounds per cubic yard.28 If the cost allocation methodology (for example, cost causation) relies on the quantity of waste being remediated, adjustments for relative density and density under landfill conditions can be critical and necessary modifications to raw volumetric waste-in data. Here again, the maxim is that standard conversion factors are available and should be employed.29 Rule 3: Superfund Fairness Equals Consistency The objective of Superfund cost allocation is not so much precision or accuracy as consistency and fairness. The basic allocation rule is that the allocation should treat parties consistently and should carefully evaluate the use of data to eliminate or minimize the impact of bias toward "good records." When site data are incomplete and the allocator is left to supplement the common data with party-specific submissions, inconsistencies in data quality are frequently found. Some parties have little or no data, while others have complete records and may even maintain sampling results on their waste streams. "Good records" bias, which results when the allocator adjusts one party's contribution relative to another because of the completeness or lack thereof of data, can often lead to unfair treatment. Because allocation is a "zero-sum game," any action that benefits one party, by definition adversely affects all others. For example, should Party A be penalized by being left at "standard" simply because Party B's complete data result in a reduction that adversely affects Party A? Or, alternatively, should Party B be penalized because its complete data lead to an adverse adjustment in the allocation relative to the other parties that did not supply data? These are real questions that allocators routinely face and that parties confront as they are sought out to provide allocation-related data. If the objective is total accuracy, the allocator may have cause to use all of the available data. But, as in the container fullness issue, the goals of consistency and fairness argue for treating parties equally. If one believes that a mixed waste weighs 275 pounds per cubic yard, that weight may be assigned as a default value for all parties. What happens, however, when a party can demonstrate that its waste instead weighs 250 pounds per cubic yard, even though it was initially allocated the same rate per cubic yard as every other party? Should that entity be granted a discount relative to the other parties? Ought the default factor be adjusted downward from 275 to 250 pounds? Or should no adjustment be made? Absent a basis to conclude that the waste from one party is materially different from the waste contributed by others, allocators should treat them consistently. Most parties do not expect precision—the process is, after all, an approximation at best—but all parties do expect to be treated consistently30 absent a compelling reason to do otherwise.31 Part 2: The Waste Is the Basis for the Allocation In Part 1 of the allocation process the focus is on developing the waste-in list, identifying three common principles useful in that process: First, countall the waste, because all waste has Superfund liability. And count the waste itself, not just the hazardous substances contained in it. Second, employ standard protocols for measuring and converting waste streams so that they can be evaluated in a consistent standardized manner. Furthermore, containers should be presumed to be full absent compelling information to the contrary. Finally, the goal of the development of a waste-in list should be to construct a uniform set of data across the parties, working carefully to avoid the pitfall of "good records" bias. For Part 2, the focus changes from the physical construction of the waste-in list to an evaluation of the basic relationships between the waste and the liable parties. Parties have liability at Superfund sites because of their waste-related activities.32 An examination of the linkage between the waste and liability of the different parties leads to several other basic allocation principles. Rule 4: The Waste Creates the Cost Being Allocated The basic allocation rule is that liability is based on waste, and the waste is the basis for the allocation. Therefore all of the site's remediation costs should be allocated based on the wastes contributed to the site. The purpose of Superfund cost allocation is to apportion the cost of the remedy among the various contributors, but as an initial matter the cost of the remedy should be apportioned to the various wastes, not to the various parties. Cost shares should not be allocated to parties based on who they are, but rather on what they sent to the site. A simple example can demonstrate this point: Assume a remediation site at which two drums of liquid wastes were disposed of. Each of the two drums will be incinerated. Three parties—two generators and one transporter—are involved. Party A, a municipality, is the generator of the first drum. Party B transported the drum. Party C, an industrial entity, self-transported the second. Rather than initially [30 ELR 10217] focusing on how to allocate costs to the three parties, allocate costs to each of the two drums of waste. In this example, each drum incurs the same remedy cost, the cost of incineration. Accordingly, initially the remediation cost should be allocated 50 percent to each drum of waste. Both drums are full, but the contents of the first drum are considered more toxic than those of the second. Although in this example the cost assignable to each drum is the same, that need not be the case. However, each drum can be directly linked to parties, so whatever the cost assignable to a drum, those parties should bear that cost. Also note that the contents of the drums are largely irrelevant in this example. Does it matter if one drum is considered five times more toxic than another? Certainly not from a cost standpoint, as long as the remedy is the same. In this example the allocation process, at least as it regards the individual drums, is straightforward. Wastes can and will be linked to parties, but the first step in the allocation process should be to apportion the costs of the remedy, i.e., the remediation or response costs, to the wastes themselves. Rule 5: Costs Assigned to a Waste Should Then Be Apportioned Once a waste shipment has received a "share" of the response cost based on the allocation process—any allocation process—that cost can be assigned to the various parties linked to it. The basic allocation rule is that the parties who, as a group, have liability under CERCLA for the costs related to that waste are the parties who should bear the costs of remediation. Continuing with the drum example, the cost of remediation may now be allocated. Party A and Party B should somehow share the costs related to the first drum, but they should not be assigned any costs related to the second drum. Party C should bear the costs related to the second drum. How to apportion costs as between the generator (Party A) and the transporter (Party B) remains a difficult allocation question. But the cost they collectively share, the cost related to the first drum, is easy to determine. The different parties—generators, transporters, owners, and operators—have liability not simply because of who they are, but because they have some relationship to a waste shipment. As a first step in the second phase of the allocation process, liability should be assigned to the waste shipment; as a second step, it should apportioned among the parties related to the shipment. Rule 6: Apportion "Orphan Shares" If a party that is initially assigned a share of the allocation is insolvent its share must be reapportioned or otherwise assigned to the remaining parties. The basic rule is that this "orphan share" should be assigned back to the other parties who share liability for the same waste. Continuing with the previous example, if the generator (Party A) and the transporter (Party B) are initially assigned equal responsibility for the cost of the first drum, the allocation is 25 percent of the total cost to Party A, 25 percent to B, and 50 percent to C. Assume it is later discovered that B is insolvent; its costs should be borne by the generator of the same waste shipment (A). If, however, a third drum that requires incineration is uncovered, and no party can be linked to the waste, those costs would then be subject to reapportionment among all of the parties, based on their relative shares of responsibility.33 What this means, of course, is that shares assignable to "orphan parties" are redistributed to other viable parties who share responsibility for those waste shipments. Truly "orphaned" wastes are then reassigned among all the identified solvent parties based on their relative share. Part 3: Allocate Response Costs Equitably Having examined the relationship between waste shipments and the various classes of liable parties, the focus now shifts to an examination of the allocation methodology itself. There are a number of frequently used allocation methods. The basic rules of cost allocation are less about which approach is right or wrong, and more about making sure that the methodology selected addresses the objectives of the cost allocation process. Rule 7: Allocate Response Costs Among Liable Parties As the statute says, courts are assigned the task of allocating "response costs among the liable parties using such equitable factors as the court determines are appropriate."34 For the most part allocators and parties focus on "equitable factors" and then proceed to question just what those are and how they should be employed. But before focusing on "how" to allocate, focus on "what" to allocate. The objective of cost allocation is to allocate response costs among the liable parties. Superfund doesn't impose penalties, per se; rather, it allocates the cost of remediation, i.e., the cost of site cleanup. The allocation process should be designed in such a way that it develops a methodology for relating different wastes and the costs they create to the costs of the remediation itself. As such, the goal or objective of the cost allocation process should be to allocate response costs, not some measure of harm. [] "Harm" and "Response Costs" Are Not the Same. Some allocators focus on attempting to allocate "harm" among the responsible parties. The terms "harm" and "response costs" are seemingly used interchangeably in the cost allocation process, but they are not the same. Cost causation measures cost; harm measures harm, usually in the form of toxicity. Allocators attempting to develop an allocation based on harm (a term routinely discussed in terms of toxicity) generally focus on the relative toxicity or hazardousness of waste shipments. Sometimes, allocators go so far as to assert that toxicity and cost causation are the same thing. In most cases, however, cost causation and toxicity weighting analyses are not the same, and this fact can be easily demonstrated. The previous example demonstrates why. Two drums cost the same to remediate. But the contents of one are five times more toxic or hazardous than those of the second. For purposes of cost allocation toxicity in and of itself, in this [30 ELR 10218] example at least, is meaningless. Consider the myriad wastes that make their way to a landfill that requires remediation. These wastes generally cover the hazard spectrum, from basically innocuous to indisputably hazardous. If one were to score these wastes for their toxicity or hazardousness a number of differently scored waste streams would be identified; for example, some industrial waste streams would be "worse" than others. No real debate here; some industrial wastes are more hazardous than others, so a focus on toxicity would properly reflect those differences. But toxicity is not the same as cost causation, and while there might be a variety of wastes, including a number of different industrial wastes, deposited at the site the process employed to remedy those wastes might well be the same regardless of hazardousness. In a case in which one industrial waste of relatively low toxicity is deposited, along with another of relatively high toxicity, in equal amounts, a toxicity allocation would assign the higher score to the relatively more hazardous waste stream. If, however, in either case the remedy that would have been used to remediate these wastes is the same (for example, RCRA Subtitle C capping of the site), then in fact each waste contributes equally to the actual response cost, and relative differences in toxicity or hazardousness are meaningless. Those differences merely serve to distort the allocation. A more interesting question arises when one party sends RCRA Subtitle D nonhazardous waste35 containing hazardous substances, while another contributes RCRA Subtitle C hazardous waste, with the remedy required to address both waste streams being the more stringent RCRA C one. But here again, if the objective is to allocate response costs, an examination of the response costs necessary for each waste stream is of more value than an understanding of relative toxicity. Rule 8: Allocations Should Survive the "Sanity Check" Another basic allocation rule is that regardless of the allocation approach utilized, the allocation itself should be able to survive critical examination and a basic "sanity check," by satisfying simple equity criteria.36 [] A Simple Allocation Test. The typical Superfund site does not lend itself to simple cost categorization; industrial wastes are not typically confined to one cell, all municipal wastes to another cell, and all liquid wastes to a lagoon. Unfortunately, they're mixed together as "Superfund stew." And the allocation process is supposed to sort out this mess and assign costs equitably amongst the parties. Regardless of the approach that is employed, there is a simple test that can be performed to examine the fairness of the allocation. Take the costs shares that have been assigned to all of the different wastes (and to their parties), and segregate them by waste and remedy types. Are all similar waste types treated equitably? Are all wastes that require the same remedy treated equitably? Frequently an allocator will fail this test and not equitably allocate response costs if it loses sight of the fact that the objective is to allocate components of the remedy. For example, at a co-disposal landfill the debate is frequently over what "but-for" remedy would have existed in the absence of either the MSW or the industrial waste. Even though the site itself is Superfund stew, pull it apart. Instead, focus upon the industrial parties and their industrial solid wastes. Are the industrial solid wastes treated equitably? After all, all of these wastes would have the same "but for" remedy. Each would arguably be remediated with the use of a RCRA Subtitle C cap. Within that group, the subset of industrial solid wastes, does the allocation equitably allocate response costs? Regardless of methodology, an allocation can be poked and prodded to see how it responds to basic equity concerns. Are all parties treated consistently? Does a waste get a particular share of cost regardless of who contributed it, or how many parties generated it? Are relationships between waste, based on toxicity, cost causation, or other factors, maintained such that the allocated shares reflectthose relationships? Any equitable allocation, under any methodology, should be able to survive the sanity check of post-allocation review.37 [] Focus on the Waste, Not on the Party Who Sends It. This rule seems obvious, yet it is frequently ignored. For example, municipalities send MSW to landfills. So do many other generators. If office trash from a business is MSW, and is both regulated and remediated the same way as MSW from a residence, aren't they the same? Of course they are. But sometimes allocators get lost by concentrating on parties rather than what those parties contributed, and end up assigning more to one party simply because of its status.38 At times allocation approaches stray to the point at which the allocation to parties is based more on who they are than on what they sent. But the parties themselves do not cause the cost of the remedy—their wastes do. A drum of solvent has a remedy cost associated with it, and a relative toxicity associated with it. It makes no difference whether Party A generated the drum, or whether Party B did so, or whether the parties are municipalities or large industrial entities. An allocation, done properly, should be blind to the origination of the waste. [] Quantity Is Quantity Regardless of How It Gets There. Equity dictates that parties be treated fairly. All else being [30 ELR 10219] equal, a waste from one party should receive the same allocation as the same quantity of the same waste from another party. Accordingly, whether an industrial party or a municipality delivers MSW to a Superfund site, each ton of MSW should receive the same allocated share. But the problem that can arise is more basic than simple cross-party equity for identical wastes. Consider the following example. A small business owner hauls five loads of waste a landfill, and is invoiced for its disposal. The invoice is directed to the owner in his individual capacity. He then contributes five more loads of waste to the landfill. This time, the company, rather than the owner, is invoiced. In total, the business owner has generated 10 loads of waste. An adjacent business borrows the owner's truck and delivers 10 loads of waste to the same landfill; that business is invoiced for the 10 loads. For purposes of the example, assume that the waste from both sources is very similar, if not identical, so that they are of equivalent hazardousness and would cost the same per unit to remedy. Obviously, any correct allocation would assign both businesses the same share of the overall allocation, whatever that share might be. The more interesting question arises when the allocator does not know that the business owner and his company are the same entity, and therefore believes there are three parties, not merely two. If there really are 3 parties, then there are certainly reasons why the remediation cost associated with 5 loads from 1 party, plus 5 from another, might not equal the cost for 10 loads generated by a single party (due to economies of scale). Otherwise, however, one would expect that 10 loads of waste—whether 10 loads from 1 party or 1 load from each of 10 parties, would receive the same score in an allocation as 10 loads from another party. This "sanity check" becomes a basic benchmark to evaluate whether an allocation has been performed correctly. Sometimes, allocations fail this test. More often than not, it is because parties receive the equivalent of a baseline score, or weight to the allocation, just for "being there" or largely for who they are. Such cases can lead to nonsensical results; e.g., a party contributes a large volume of waste, but the same quantity of the same waste, contributed by several parties, receives a disproportionately higher share of the overall allocation. That result would flunk the basic equity test. The "I'm Twice as Bad, But You Sent 10 Times More" Problem. Under whatever allocation approach is chosen the quantity of waste is frequently an important factor. Whether the approach is a toxicity weighting, some measure of cost causation, or another method, the process involves relating how one waste stream compares to another, frequently on a unit basis. For example, if it is assumed that the waste from one party is "twice as bad" as the waste from another—for example, assume that it costs twice as much to remediate one type of waste—quantity still matters. If the first party sends one ton, and the second party sends twice that amount the quantity has directly offset the cost relationship, as shown in Figure C. Likewise, if a toxicity factor or some other measure of relative toxicity is employed, quantity still matters. If 1 waste stream is 3 times more toxic than another, but the less toxic stream has 30 times more quantity, then the large volume, less toxic stream should still receive most of the allocation. This is true not simply because it contributes most of the volume, but rather, because under the measure used for the allocation, toxicity, it is clear that the larger volume waste stream is the more toxic contribution to the site.39 This result can be seen in Figure D. Although Party B's waste is cheaper to remediate, and relatively less toxic, it has sent so much waste that its quantity has dictated the allocation result. An equitable allocation approach should "catch" that. Too often the objective of Superfund cost allocation is left unclear, and parties seemingly use the terms "harm" and "response costs" interchangeably. The goal of the process should not be subject to any serious debate; it is, after all, spelled out in the statute. Beyond the central objective, parties too often take opposing viewpoints as to the method of allocation, arguing for volume on one hand, or toxicity on the other. Both volume and toxicity have a role to play together as they relate to cost. But neither factor has much of an independent or isolated role in an equitable allocation of response costs. The allocation rules set forth in this Dialogue do not focus on the appropriate allocation method. Indeed, most of them are much more basic and apply to any allocation methodology. They address issues such as what wastes to count, how to count them, and how to evaluate whether the process is equitable. Not all of the rules apply in every case; likewise, in some instances modifications are appropriate. And there are undoubtedly more basic rules than just these. The purpose of this Dialogue was not to enunciate a definitive set of basic allocation rules and assert that everyone should follow all of them, all the time. The goal was, instead, to identify basic allocation principles that routinely are employed, albeit with some modifications, and discuss why and how they are employed. The allocation process is filled with confusion and conflicting information. Getting a grasp on basic allocation issues should help to narrow the range of uncertainty. FIGURE A EIGHT BASIC RULES OF SUPERFUND COST ALLOCATION PART 1: HOW TO QUANTIFY WASTE STREAMS Rule 1 ALL OF THE WASTE GETS COUNTED * All waste has Superfund liability. * Waste, not merely hazardous substances, is being allocated. * Don't limit the focus to contaminants of concern or some other subset of hazardous substances. Rule 2 USE STANDARD PROTOCOLS FOR MEASURING AND CONVERTING WASTE STREAMS * All containers are full. Use standard measures for waste conversion. Rule 3 TREAT PARTIES CONSISTENTLY AND AVOID "GOOD RECORDS BIAS" PART 2: THE WASTE IS THE BASIS FOR THE ALLOCATION Rule 4 THE WASTE CREATES THE COST BEING ALLOCATED Rule 5 COSTS ASSIGNED TO A WASTE SHOULD THEN BE APPORTIONED TO THE LIABLE PARTIES RELATED TO THAT WASTE Rule 6 APPORTION "ORPHAN SHARES" AMONG OTHER PARTIES RESPONSIBLE FOR THE SAME WASTES PART 3: MAKE SURE THE COST ALLOCATION METHODOLOGY ALLOCATES RESPONSE COSTS EQUITABLY Rule 7 THE PURPOSE OF SUPERFUND COST ALLOCATION IS TO ALLOCATE RESPONSE COSTS AMONG LIABLE PARTIES * Harm and resoonse costs are not the same. Rule 8 AN EQUITABLE ALLOCATION SHOULD SURVIVE THE "SANITY CHECK" * A simple allocation test exists. * The allocation should focus on the waste, not on the party who sends it. * Quantity is quantity regardless of how it gets to the site. * The "I'm twice as bad, but you sent 10 times more" problem. FIGURE B EPA'S STANDARD CONVERSION FACTORS FOR WASTE-IN LISTS AND VOLUMETRIC RANKINGS
FIGURE C THE RELATIONSHIP BETWEEN COST AND QUANTITY
FIGURE D THE RELATIONSHIP BETWEEN TOXICITY AND QUANTITY
2. See Richard Lane White & John C. Butler III, Applying Cost Causation Principles in Superfund Allocation Cases, 28 ELR 10067 (Feb. 1998). 3. See Browning-Ferris Indus. of Ill., Inc. v. Ter Maat, 13 F. Supp. 2d 756, 777, 29 ELR 20142, 20150 (N.D. Ill 1998): The landfill at issue in this case was in existence and actively operated from the late 1960s to the late 1980s. For twenty years, a variety of waste was brought to, and deposited in, the landfill. Over 3.5 million cubic yards of waste were landfilled during that time. All during this time, as the volume and variety of waste increased, what was there began to decompose. The decomposition process resulted in the waste returning to its base constituencies. This further led to a complex and ill-defined mass of substances. All of this waste also contained a certain amount of water which mixed with the base elements from the decomposition process to create what the experts call leachate but which may fairly be described as toxic soup. This "soup" eventually intermixed from various locations throughout the landfill and worked its way into the surrounding soil and finally into the groundwater. And like grandma's special soup, the final product was the result of a unique blend of ingredients, some known and some unknown. Now the parties ask, and the law requires, that this court analyze the "soup" and determine who put in which ingredients and how much. Like someone trying to figure out what makes Coke taste the way it does, this is no easy task. Unraveling a twenty-year process involving millions of cubic yards of waste and complex ecological, biological and geological forces is likely to yield less than precise results. 4. For a recap of relevant decisions during Superfund's first decade, see Kyle E. McSlarrow et al., A Decade of Superfund Litigation: CERCLA Case Law From 1981-1991, in SUPERFUND DESKBOOK 487 (Environmental Law Institute 1992). 5. See, e.g., John Copeland Nagle, CERCLA, Causation, and Responsibility, 78 MINN. L. REV. 1493 (1994). 6. Ter Maat, 13 F. Supp. 2d at 777, 29 ELR at 20150: Additionally, as to the many issues raised in this case, most of which ultimately bear on the allocation issue, there is little guidance in the way of case law, particularly at the appellate level. This is most likely due to the fact that many of these cases settle. The high rate of settlement is probably attributable to the inherently unpredictable results a trial will bring. Further, most of the cases performing allocations rightfully caution that each case is unique and should provide little, if any, guidance in future allocation matters. Given the dearth of cases on many of the issues, combined with the less than precise language of CERCLA itself, the court is plowing new ground on many issues. The bottom line is the court is left with an underdeveloped legal framework in which to resolve the many issues raised by the parties. 7. Resource Conservation and Recovery Act, 42 U.S.C. §§ 6901-6992k, ELR STAT. RCRA §§ 1001-11011. 8. See, e.g., City of New York v. Exxon Corp., 744 F. Supp. 474, 21 ELR 20248 (S.D.N.Y. 1990); Amoco Oil Co. v. Borden, Inc., 889 F.2d 664, 20 ELR 20281 (5th Cir. 1989); Louisiana-Pacific Corp. v. ASARCO, Inc., 735 F. Supp. 358, 20 ELR 21452 (W.D. Wash. 1990). 9. See United States v. A & F Materials, Inc., 582 F. Supp. 842, 14 ELR 20432 (S.D. Ill. 1984). See also United States v. Wade, 577 F. Supp. 1326, 14 ELR 20096 (E.D. Pa. 1983); United States v. Carolawn Co., 14 ELR 20696 (D.S.C. June 15, 1984). 10. See United States v. Atlas Minerals & Chems., Inc., No. CIV.A. 91-5118, 1995 WL 510304, at *97 (E.D. Pa. Aug. 22, 1995). The Atlas Minerals court addressed the issue by noting that: CERCLA's definition of "hazardous substances" includes not only those wastes defined as "hazardous waste" under RCRA but also a medley of other substances. CERCLA broadly defines "hazardous substance" as any substance that is regulated under virtually any federal environmental statute. Regulations promulgated under CERCLA list over 700 hazardous substances. . . . Third-Party Plaintiffs aptly remark that although municipal waste, which typically contains many hazardous substances, is excluded from the RCRA's hazardous waste provision, it is not exempt from CERCLA. Courts have imposed CERCLA liability based on disposal of household waste, despite arguments [to the contrary]. . . . To draw a bright line between hazardous and nonhazardous waste based on RCRA's definition for purposes of allocation ignores the fact that CERCLA casts a much wider net than does RCRA. For purposes of allocating costs in this case, it is more acceptable to recognize that degrees of toxicity exist along a continuum within the broad category of CERCLA hazardous substances. Id. (citing 42 U.S.C. § 9601(14), ELR STAT. CERCLA § 101(14); 40 C.F.R. § 302.4 (1996). See also B.F. Goodrich Co. v. Murtha, 958 F.2d 1192, 1202, 22 ELR 20683, 20687 (2d Cir. 1992): RCRA is preventative; CERCLA is curative. It does not follow that because the environmental risk posed by household waste is deemed insufficient to justify the most stringent regulations governing its day-to-day handling that the environmental harm caused when that risk is realized is insufficient to require holding liable those responsible for that harm. See New Jersey Dep't of Envtl. Protection v. Gloucester Envtl. Management Servs., 821 F. Supp. 999, 23 ELR 21420 (D.N.J. 1993). 11. See, e.g., United States v. Nicolet, Inc., 712 F. Supp. 1205, 20 ELR 20864 (E.D. Pa. 1989); United States v. Conservation Chem. Co., 619 F. Supp. 162, 16 ELR 20193 (W.D. Mo. 1985). 12. See, e.g., discussion of liability for municipal solid waste in Atlas Minerals, 1995 WL 510304, at *97 ("CERCLA advances this goal [of encouraging maximum care and responsibility in the handling of hazardous waste] by imposing joint and several liability on parties that dispose of even de minimis quantities of hazardous substances."). 13. See, e.g., B.F. Goodrich Co. v. Murtha, 754 F. Supp. 960, 965-66, 21 ELR 20777, 20779 (D. Conn. 1991), aff'd, 958 F.2d 1192, 22 ELR 20683 (2d Cir. 1992). 14. 146 F.3d 1107, 28 ELR 21341 (9th Cir. 1998). 15. Id. at 1110, 28 ELR at 21342. The court also noted that "nothing in the law suggests that quantities of hazardous substance below its reportable level render it no longer hazardous. The Second, Third and Fifth Circuits have faced this very question and all agree that CERCLA's definition of hazardous substance has no minimum level requirement." Id. (citations omitted). 16. See, e.g., Louisiana-Pacific Corp. v. ASARCO, Inc., 735 F. Supp. 358, 20 ELR 21452 (W.D. Wash. 1990). 17. There is no threshold requirement for hazardous substances in order to establish liability. "Twice we have said that quantity 'is not a factor' when determining CERCLA liability because had Congress wanted to distinguish liability on the basis of quantity, it would have so provided." B.F. Goodrich v. Betkoski, 99 F.3d 505, 517, 27 ELR 20329, 20329 (2d Cir. 1996). "Quantity or concentration is not a factor either; when Congress wanted to draw distinctions based on concentration or quantity, it expressly provided as much." B.F. Good-rich v. Murtha, 958 F.2d 1192, 1200, 22 ELR 20683, 20686 (2d Cir. 1992). See also United States v. Alcan Aluminum Corp., 990 F.2d 711, 720, 23 ELR 20706, 20710 (2d Cir. 1993) ("The statute on its face applies to 'any' hazardous substance, and it does not impose quantitative requirements."). The absence of threshold quantity requirements in CERCLA leads logically to the conclusion that the Act's "hazardous substance" definition includes even minimal amounts. "Therefore the presence of any detectable amount of these substances, without regard to concentration, is sufficient to satisfy the requirements of the statute." HRW Sys., Inc. v. Washington Gas Light Co., 823 F. Supp. 318, 340, 23 ELR 21586, 21595 (D. Md. 1993). 18. U.S. EPA, GUIDANCE ON PREPARING WASTE-IN LISTS AND VOLUMETRIC RANKINGS FOR RELEASE TO POTENTIALLY RESPONSIBLE PARTIES (PRPS) UNDER CERCLA, OSWER Directive No. 9835.16 (Feb. 20, 1991) [hereinafter GUIDANCE ON PREPARING WASTE-IN LISTS], attach. 3 at 2. 19. 917 F. Supp. 1173, 26 ELR 21056 (N.D. Ohio 1996). 20. Id. at 1184, 26 ELR at 21061 (emphasis in original; citations omitted). 21. EPA Proposal for Municipal and MSW Liability Relief at CERCLA Co-Disposal Sites, 62 Fed. Reg. 37321, 37233 (July 11, 1997) [hereinafter EPA Proposal for MSW Relief]. See also EPA Policy for Municipality and Municipal Solid Waste CERCLA Settlements at NPL Co-Disposal Sites, 63 Fed. Reg. 8197 (Feb. 18, 1998) [hereinafter MSW Settlement Policy]. 22. No. CIV.A. 91-5118, 1995 WL 510304 (E.D. Pa. Aug. 22, 1995). 23. Id. at *101 (emphasis included; citing 42 U.S.C. § 9607(a)(4), ELR STAT. CERCLA § 107(a)(4)). 24. One exception to this would be a municipal party seeking to settle under EPA's MSW Settlement Policy, which provides a cash-out formula based on tonnage. See MSW Settlement Policy, supra note 21, at 8197; EPA Proposal for MSW Relief, supra note 21, at 37321. 25. GUIDANCE ON PREPARING WASTE-IN LISTS, supra note 18, attach. 3 at 1: A 55-gallon drum or any other container of hazardous substances for disposal was full when it was shipped and when it was disposed. Unless a shipping or disposal record unambiguously indicates otherwise, either because the recorded volume is less than that of the full container volume, or the price is less than that normally charged for a full container, the burden of proof is on the PRP to show that a container was less than completely full. 26. Id. at attach. 2 (supplemented with data obtained from the MSW Settlement Policy, supra note 21). 27. See, e.g., A HANDBOOK OF UNIVERSAL CONVERSION FACTORS (S. Gerolde ed., 1977). 28. See, e.g., MSW Settlement Policy, supra note 21, which references 1,200 pounds per cubic yard for MSW under landfill conditions. See also Figure B, infra. 29. One source that can be used to evaluate density measurements is PERRY'S CHEMICAL ENGINEERS' HANDBOOK (Robert H. Perry et al. eds., 7th ed. 1997). 30. Consistent treatment is, after all, the cornerstone of an objective environmental program. See, e.g., G. Nelson Smith III, Lawmaker as Lawbreaker: Enforcement Actions Against Municipalities for Fuiling to Comply With the Clean Water Act, 41 CLEV. ST. L. REV. 685 (1993). 31. As the Atlas Minerals court noted: It is inescapable that allocation will be difficult in this case. To begin with, this is an abstruse area of law. Also, the court's findings regarding the parties' volumetric contributions are only approximate because of gaps in the evidence attributable mostly to the passage of time. Such uncertainty makes it somewhat confounding for the court to devise a principled scheme for allocating costs. Therefore, any method the court employs to establish the parties' shares will be reasonably fair at best, and imprecise at worst. 32. Primarily disposal and/or handling of wastes, but it can also include other activities such as actions that affect the waste contributed by other parties. 33. For discussions of orphan share allocation, see Richard Lane White & Allen Kezsbom, The Debate Over Orphan Share Allocation, 34 Chem. Waste Litig. Rep. 391-409 (1997); Kenneth T. Wise et al., Economists, Orphans, and Superfund Allocation: Clearing Up the Controversy, 34 Chem. Waste Litig. Rep. 763-66 (1997); Richard Lane White & Allen Kezsbom, No Need to Redefine Orphan Shares, 34 Chem. Waste Litig. Rep. 975-80 (1997). 34. 42 U.S.C. § 9613(f)(1), ELR STAT. CERCLA § 113(f)(1). 35. See John H. Turner, Off to a Good Start: The RCRA Subtitle D Program for Municipal Solid Waste Landfills, 15 TEMP. ENVTL. L. & TECH. J. 1 (1996) (discussion of the federal Subtitle D program). 36. See Ridgway M. Hall et al., Superfund Response Cost Allocations: The Law, the Science, and the Practice, 49 BUS. LAW. 1512 (1994). 37. For an example of an equitable allocation, see Ellman v. Woo, 22 ELR 20875 (E.D. Pa. 1991). 38. This issue was addressed in United States v. Atlas Minerals & Chems., Inc., No. CIV.A. 91-5118, 1995 WL 510304 (E.D. Pa. Aug. 22, 1995). The court dealt with several issues, including classification of parties (as opposed to wastes), waste segregation as it affects the allocation process, and waste segregation/commingling. It noted that: Third-Party Defendants' [allocation] model characterizes parties, not waste streams, as either hazardous or nonhazardous. Thus, if a party disposed of several different types of waste, and at least one waste stream is considered to be "hazardous" under Third-Party Defendants' conception of the term, the party lands in the hazardous group and is allocated its portion of 90% of the costs based on its total volume of both hazardous and nonhazardous waste. 1995 WL 510304, at *97. The court rejected the argument, concluding that it would "confine its analysis to an evaluation of objective factors leading to the Site's remediation and its cost, [rather than] the parties' culpability regarding their waste disposal practices." Id. at *98. 39. The approach to evaluating toxicity in this example—determining a factor for a waste stream and then multiplying the quantity of waste by that factor—is consistent with the approach that was used for this same analysis in Atlas Minerals, 1995 WL 510304, at *102, 104, 113 (app. E-1). 30 ELR 10212 | Environmental Law Reporter | copyright © 2000 | All rights reserved |