Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 How Clean is Clean? The Need for Action THOMAS M. RELLMAN AND DEBORAH A. HAWKINS One of the major impediments to moving remedial cleanup actions under any of the various federal and state laws that apply is the resolution of the "How clean is clean?" issue. The statutory definition of what factors must be considered in determining the acceptable level of cleanup varies from law to law. These defini- tions, which were developed by legislative processes in Washington, D.C., and various state capitals, often are not clearly translatable to a determination of cleanup levels at remediation sites. Instead, they generate controversy and confusion. The practical result of this situation is that the cost of remediation increases as does the time required to implement solutions. Ultimately, fewer sites will be cleaned up. In this paper, ~ will discuss the following issues: the current legal/regulatory framework relative to cleanup levels; the cost and technical implications of various cleanup strat egies; the balance of today's cleanup costs versus future liabilities; and who ultimately pays. CURRENT IEGAI/REG=ATORY FRAMEWORK EPA has taken the position that a cleanup conducted under the Resource Conservation and Recovery Act's (RCRA) corrective action authority should be no different from a cleanup undertaken under the Superfund program. Both actions target the cleanup of historic contamination, and both have the same health-based cleanup goals. One significant difference is the fact that RCRA 98
HOW CLEAN IS CLEAN? 99 sites are generally associated with existing businesses; thus, oper- ating revenues can be used to pay for the cleanup. These costs in turn can be passed on to customers. This is not true for Superfund sites, for which available funds are more lirn~ted. Consistent with this position, EPA has announced its intent to merge RCRA and Superfund cleanup approaches and to im- plement a single programmatic response to the cleanup of historic contamination. Congress, however, has not cooperated. Notwith- standing EPA's official policy, differences in statutory language and approach make the choice of undertaking a cleanup under the authority of RCRA or Superfund an important one. The recent enactment of the Superfund Amendments and Reauthorization Act (SARA) has sharpened the differences be- tween RCRA and Superfund. The statutes adopt divergent ap- proaches on a number of critical issues, including cleanup stan- dards, use of alternative concentration limits (ACEs), cost-effec- tiveness considerations, and public participation. Under RCRA, the requirements can be more stringent as the criterion of cleanup to background is applied with an opportunity to modify cleanup levels using the ACL risk management approach. The Superfund procedure mandates more vigorous public involvement. In 1987 EPA issued its formal statement of policy on RCRA and Superfund in which it echoed the cleanup goal theme. When identifying those cleanups that should be included on Superfund's National Priority List and those that should be undertaken under RCRA's corrective action program, EPA acknowledged that juris- diction may lie under both statutes. The agency established a pre- sumption in favor of RCRA-authorized cleanups for RCRA facili- ties, but it emphasized that similar cleanup approaches would be followed, regardless of whether a cleanup proceeded under RCRA or Superfund. Agency spokesmen explained: "The Agency's goal is to develop RCRA corrective action requirements that remove inconsistencies between remedial actions performed under CER- CLA and corrective actions performed under RCRA." In practice, it may be difficult to achieve consistency between the two statutes because cleanup authority under RCRA and Superfund is not sirn~lar; in addition, each statute includes important elements that may have significant impact on the cleanup decisions. As a general rule, RCRA requires that cleanups protect "hu- man health and the environment." Because RCRA is much more of a hazardous waste management statute than a hazardous waste
100 HAZARDOUS WASTESITEMANAGEM:ENT cleanup statute, it provides no additional guidance on cleanup methods. The statute includes no mention, for example, of whether cost-e~ectiveness can or should play a role in selecting among reme- dial alternatives, nor does it include any other guidance regarding the type of remedy that should be employed in cleanups. By not mentioning cost-effectiveness, the statute precludes consideration of it. In fact, the legislative history indicates it is not to be consid- ered. The use of ACEs determined by risk assessment is allowed, however. Even before its recent amendments, Superfund provided much more explicit guidance. Like RCRA, Superfund adopted a health- based standard, but it also specified that cleanups should be eval- uated and selected on cost-effectiveness grounds. It also mandated that fund balancing considerations were to be a part of the eval- uation whenever Superfund monies are being used for cleanup efforts. Now, under SARA, the Superfund scheme includes additional guidance and constraints on the selection of remedial actions. Specifically, Section 121 of SARA includes a strong bias in favor of permanent remedies and onsite remedies and requires that appli- cable or relevant and appropriate state and federal standards be applied. These requirements will ultimately be reconciled in the National Contingency Plan. Complementing the federal statutes are a number of state laws that drive cleanup activities. Virtually every state has its own form of the Superfund law, which requires the cleanup of sites not addressed by the federal Superfund. Several states are taking innovative approaches toward remediation of contaminated sites. New Jersey has, under its Environmental Cleanup Responsibility Act (ECRA), established a process by which the transfer of in- dustrial or commercial properties on which hazardous materials have been handled must be reviewed and approved by the New Jersey Department of Environmental Protection to ensure that any contamination has been cleaned up. This approach is being considered in a number of other state legislatures. Massachusetts approaches property cleanups using another method called Super- lien. Under Superlien laws, the state has the first lien on properties on which the state has expended money to clean up contamination. The banking community in Massachusetts has become very con- cerned about the potential consequences of making loans secured by contaminated property and has required environmental reviews
HOW CLEAN IS CLEAN? 101 of properties before making loans. Each of these laws ultimately drives a decision on whether a property is deemed to be contam- inated and, if so, what cleanup levels are appropriate. The "How clean is clean?" issue is usually resolved on a site-by-site basis using a combination of risk assessment techniques and applicable standards. The three criteria most often used by states to address the level-of-cleanup issue are: ~ cleanup to background (the level at which no industrial activity had taken place, allowing only for natural contamination, pH, radioactivity, and so on); cleanup to background, holding other responsible parties accountable for the contamination they caused on the property; and care. human health and environmental protection standard of COST VERSUS CIEANUP [EVEIS The two parameters most significantly affected by cleanup lev- els are the costs of the cleanup and the time required to accomplish the remediation. The overall impact of SARA on the Superfund process has caused EPA staff to project a Month increase in the time it takes to handle a Superfund cleanup-that is, from 58 months to 67 months. The remedial investigation/feasibility study (RI/FS) work plan is supposed to be developed within 6 months of the commencement of discussions with cooperative potentially responsible parties (PRPs). The RI/FS itself will take another 18 months. The health assessment should be available toward the end of the second year. Public and state comment will occur in the third year, after which the record of decision (ROD) is prepared. The remedial design will be available around the end of the third year, and consent decrees may be entered at any time there is a settlement. Thereafter, review and contracting will occupy most of the fourth year. Remedial action, which takes an average of 2.5 years, will bring EPA's estimate of total elapsed time to over 6.5 years that is, the remedy is even further away. This schedule illustrates that SARA has created a cleanup process with great potential for inflating costs. New cleanup stan- dards, health assessments, state and public participation, and
102 HAZARDOUS WASTE SITE MANAGEMENT TABLE 5-1 Summary of Record of Decision Results Cleanup Options Average Cost Reprised Increase Multipliers Program Cost EPA remedy 1.00 $16 billion* Containment remedy 2.61 $39 billion Least-cost permanent solution 5.49 $81 billion *These estimates were derived from the EPA average cleanup cost estimate provided in the Superfund Section 301(a)~1~(c) study of future funding needs. In that study the average site cleanup cost was estimated at $8.84 million, and 1,800 sites were assumed to be listed on the National Priority List. This volume results in a total program cleanup cost of approximately $16 billion. The reprised program costs are estimated by applying the cost multiplier to design, construction, and operation and maintenance costs but not to remedial investigation/feasibility study costs, which should stay the same. Other new requirements of the statute all contribute to this poten- tial. EPA has estimated that the cleanup requirements in SARA would drive the cost of a Superfund cleanup from its present aver- age of about $8 million-$9 million per site to between $25 million and $30 million per site. States are responsible for paying 10 per- cent of the cleanup costs at fund-financed sites, and many state officials have expressed concern about the increased cost potential. In a study carried out in 1986, Putnam, Hayes, and Bartlett, Inc., examined all the RODs issued after January 1, 1985, to deter- mine the costs associated with various cleanup options. Thirty-five of these RODs were useful for the purpose of this study (Table 5-1~. The practical result of the increased cost per site would mean that either the Superfund tax would need to be adjusted to reflect the added cost of the more stringent cleanup requirements or fewer sites would be cleaned up. The following case studies address the cost issue on a smaller scale. Case Study 1 A relatively small electric equipment repair shop located in the southeastern United States had a polychIorobipheny! (PCB)
HOW CLEAN 15 CLEAN? TABLE 5-2 Costs of Different Cleanup Levele Remediation Level Volume (ppm PCB) (cu. yd) Cost ($000~) 50 2,260 1,000 10 3,750 1,750 1 8,290 3,500 103 contamination problem. The solution was deemed to be offsite disposal at an approved hazardous waste land disposal facility. The costs were a direct function of the amount of soil to be removed (Table 5-23. In addition to cost considerations, a judgment must be made on the wisdom of using limited hazardous waste disposal fa- cilities to dispose of a relatively low-risk waste-that is, PCB- contaminated soil. Case Study 2 A trichIoroethylene (TCE) ground water contamination prob- lem was discovered at a plant in the western part of the United States. Cleanup of the ground water contaminated with 20 ppm TCE was initiated by airstripping at a rate of 85 ppm. After 900 days of continuous pumping the TCE concentration in the aquifer had dropped to 1.3 ppm. After an additional 700 days the con- centration was 1 ppm. Thus, additional pumping had arrived at a point of decreasing benefit because with time the concentration was asymptotically approaching a nonzero value. The estimated costs to reach various cleanup levels are given in Table 5-3. The case study demonstrates the costs and the length of time that would be required if low parts-per-billion cleanup levels are required. It raises the practical question of who will be responsible for these kinds of abatement systems 20 to 50 years from now when the companies deemed responsible may no longer exist.
104 HAZARDOUS WASTE SITE A~4NAGEMENT TABLE 5-3 Estimated Costs for Several Cleanup Levele Total Cost ($000~) Cleanup Level (ppb) Time (years) 2221,300 2.5 3121,000 d.5 ~1,100100 20.0 -10,000 10 100.0 l~:Cl~OIOGY CONSIDE11ATIONS Basically, the universe of cleanup problems we face can be analyzed in terms of three major technology challenges: (1) con- centrated residues sludges and drums containing hazardous ma- terials are examples; (2) contaminated ground water typically having relatively low levels of organic and inorganic contaminants; and (3) contaminated soil with a wide variety of contaminants. At least two of these problems are currently capable of solu- tion. Adequate technology exists for concentrated residues because they are essentially the same hazardous wastes managed under the RCRA program. Ground water cleanup is In some respects merely a different form of water pollution control. Obviously, these char- acter~zations are an oversimplification, and yet, clearly much of the technology for destroying residues and cleaning ground water exists. The challenge ~ how to get that technology to the site needing remediation. The cleanup of contaminated soil remains a problem to be solved. For example, incinerating soil ~ extremely expensive, and the "burnt soil" product may be as hazardous as the original contaminated soil. Table ~4 shows various cleanup options for PC~contaminated soil and their costs. BALANCING CLEANUP COSTS VERSUS FUTURE LIABIIITIES In today's litigious society, more people are suing companies over environmental contamination-related issues (e.g., drinking water, property devaluation, illness, etc.~. For most large com- panies, the annual transactional costs alone are measured in the millions of dolBars.
HOW CLEAN IS CLEAN? TABLE 5-4 Cleanup Options and Costs for PCB-Contaminated Soil (100 ppm PCB) Treatment Estimated Cost/cu. yd Landfill--no pretreatment Fly ash/cement stabilization Fixation onsite with inorganic polymer/cement mixture Chemical destruction onsite In situ ~ritrif~cation (glassifying the soil) maxtrix with complete destruction of PCBs Incineration of soil onsite (PCB destruction) $200-3400 $60-$80 3180 $100 $200-$250 3200-6300 105 The quandary we face is that the lower the cleanup stan- dards that is, the more stringent they arc the higher the cost per site and the longer each cleanup will take, with the result that fewer sites will be cleaned. On the other hand, lower cleanup standards may also result in lower future liabilities for responsible parties with respect to the site being remediated. Clearly, when responsible parties are evaluating what constitutes an adequate cleanup level, consideration should be given to the impact of the cleanup on future liabilities. WHO PAYS? The final unport ant question is, who pays? The answer is, we all do. An examination of the magnitude of the monies being expended today on cleanups may be instructive. Over the last few years, based on the original Sup erfund, EPA's spending rate has been about $20 million per month. It is now about $30 million per month, about one-half of which is spent at waste sites. About 50 percent of waste site money goes for RI/FS; the remainder goes for cleanup. Thus, $5 million to $10 million per month is being spent by EPA on cleanup efforts. SARA will ultimately boost
106 HAZARDOUS WASTE SITE MANAGEMENT that spending rate to $100 million per month. EPA has indicated that some 13 cleanups have been completed and about 300 are be- yond the RI/FS stage. Companies with significant involvement in national priority site listing, which essentially includes most of in- dustrial America, are spending $15 million to $30 million per year for cleanups. In addition, a number of states have passed rn~ni- Superfund legislation to generate the necessary matching funds and to undertake the cleanup of sites that are not on the National Priority List. New Jersey and New York have significant funds available for cleanup. As with the federal Superfund, these state programs are generally funded by industry, and their costs are ultimately passed on to customers in the form of higher prices for American goods and services. Yet American industry pays an ad- ditional price by being further disadvantaged in relation to foreign competition in a world economy. The point is that we should all feel responsibility for making sure that the limited resources avail- able to address environmental contamination problems are spent in the most effective manner. Some form of overall risk assessment should be used to determine what action represents the greatest risk reduction potential per dollar spent. EPA could then evaluate its performance against a meaningful yardstick. CONCH REMARKS AND RECOM~NDATIONS In the arena of hazardous waste site cleanup, the price we pay for inaction is elevated risk to the impacted population. The decisionmaking process is often prolonged by the desires of certain interested parties to achieve the ultimate solution, despite the ne- cessity of working from an imperfect data base and using complex yet unverified modeling systems. The attempt to distinguish be- tween risk rates of 10-5 to 10-7 is somewhat analogous to trying to distinguish between the third and fourth decimal place using a slide rule it is in the error bracket. Clearly, a balancing of issues, including public health, environmental protection, and economy, must take place with a premium on cleanup action. We often lose sight of the tremendous disjunction between the "How clean is clean?" issue and the real world. Through increased political and agency pressure, we are tightening cleanup requirements to degrees of stringency that push the cost of cleanup beyond what is possible, thus inducing further delays. We spend
HO W CLEAN IS CLEAN? 107 too much time on site-specific data development and modeling- especially considering the complex "chemical soup" that exists at most sites, which is too complicated for our current techniques to model accurately. These issues may be some of the underlying causes of the lack of achievement in the Superfund program to date. We need more objective future-use considerations in our reme- diation planning. We need to take site-specific actions consistent with good engineering practices and the circumstances of the site. We must also develop a decisionmaking system that can be ap- plied on a mass production basis. The premium must be on getting cleanup activities under way with incentives built in for those who are willing to go forward. The solutions may not always be perfect, but let us opt for some imperfection versus paralysis. REFE1lENCE Putnam, Hayes, and Bartlett, Inc. 1986. Cost Implications of Changes in Superfund Cleanup Standards. Study conducted for U.S. Environmental Protection Agency, Washington, D.C. PROVOCATEUR'S COMMENTS flyby Page I thought that Tom HelIman's paper was thoughtful and inter- esting. He said in the paper that we have a confusing set of criteria that counteract each other and lead to inefficiencies. He also said there are diminishing returns that waste cleanup. Nonetheless, he pointed out that we have to do some sort of balancing because we have mixtures of goals and mixtures of costs. As a practical result we end up having a few big sites treated to a large extent, and we lose efficiency in the course of that. We might do better with less treatment and more sites and quicker treatment. This is a theme that makes sense to me. To move from this summary to something a little more provoc- ative, ~ will say that what we really need to do is think about what is driving the present system and what we will need to change in order to drive it in the direction we would like to see. One of the things that is driving the system is liability. The liability
108 HAZARDOUS WASTE SITE MANAGEMENT issue is certainly affecting the way industry is doing things. It is also affecting the way regulators are doing things, and this is not as well known. But ~ think it is import ant. Liability from a regulator's point of view has to do with how much trouble he can get into from making a decision. When you think of liability in that sense, it explains some of the delay that we are seeing. If you put off a decision and have another study, wait a few years, get a contractor's report, all of this puts distance between you and the liability you wright fee] for your decisions. ~ think this is part of the problem. Another part of the problem is that if you ever sign off on something, saying, I This is clean enough, let's stop work," you might get into trouble. If you drag on the process and insist on more and more, then you put off the day of reckoning and you put off the liability. Therefore, liability affects not only what industry does but also the way regulators work, and the combination of the two can lead to some of the things Tom has observed especially the phenomenon of a small number of large-scale cleanup sites. A second thing driving the system and leading to inefficien- cies is the old-fashioned way of looking at uncertainty and making decisions on the basis of it. In contrast to our management of toxic wastes, there has been a shift in economics, and the de- cision sciences quite generally, toward the ideas of de Finetti, Ramsey, Savage, and others. These decision theorists are more explicitly judgmental and subjectivist than traditional scientists and statisticians. The modern decision theorists do not believe in probabilities being "really out there. They do not think of proba- bilities being in the dice or in the toxic chemical; they believe that probabilities are "really in here, in the judgment, in the mind of the assessor. This newer perspective changes the way one looks at decisionmaking. It means that one makes probabilistic evalua- tions of scientific uncertainty, including both systematic error and measurement error. Probabilistic assessment of systematic error is often lacking in current risk assessments. We often substitute analyses of measurement error for analyses of the assessment of systematic error. Instead of saying a mode} is useful in the sense that it yields information for a decision in a particular way, we say it is either valid or not valid in a very brittle kind of way. This either/or approach has gotten us into trouble when we deal with decisionmaking under uncertainty. A third thing driving the system is our difficulty in dealing with criteria other than efficiency. To an economist the criterion of
HOW CLEAN IS CLEAN? 109 economic efficiency is precise. It can be translated under varying kinds of assumptions into criteria having to do with cost-benefit analysis, risk-benefit analysis, cost-effectiveness, the minimization of costs not just for the particular project but in the design of the institutional device to make it work, and the minimization of costs over an entire decision process. All of this is a well- trodden field. ~ do not think it has been applied very well in the case of the Superfund program, but actually the consideration of efficiency is the easy part. The hard part is that other criteria are also important, criteria that have to do with distributional considerations such as the protection of victims and restitution for those who are harmed. Another criterion has to do with corrective justice, the holding to account of a perpetrator of bad actions. In the case of Superfund, this is usually translated into money that the polluter should pay. Distributional criteria are important and help explain why we have had so much trouble implementing straightforward cost-benefit analysis, which tends to neglect them. A concern for distributional criteria also helps to explain another anomaly: we seem to be spending enormous amounts of money on the remediation of hazardous waste-sites, whereas we seem to be much more accepting of hazardous materials in other media, such as air or surface water. If you think of what it takes to identify a source, it may be easier in the arena of ground water protection than it is in the air, and that may help explain why there is greater political salience for one rather than the other. These three things our current approaches to liability, un- certainty, and distributional criteria- can be added to the obvious fourth factor, the traditional conflicts between the potential gain- ers and losers from any collective decision. The four factors help explain not only what is driving the system but also why it is so hard to get to a satisfactory destination.
