National Academies Press: OpenBook

Ranking Hazardous-Waste Sites for Remedial Action (1994)

Chapter: 4 EPA'S PRIORITY SETTING

« Previous: 3 CLASSIFICATION OF PRIORITY-SETTING APPROACHES
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

4
EPA's PRIORITYSETTING

The U.S. Environmental Protection Agency (EPA) has primary responsibility for environmental management and regulation in the United States, and with it, the authority to identify the most serious abandoned hazardous waste sites for attention under the federal Superfund program. As part of this authority, EPA must determine criteria for inclusion on the National Priorities List (NPL) for Superfund sites and the pace at which sites continue along the administrative path from identification to remedial action and closure. The principal priority setting step occurs when a site, following preliminary assessment (PA) and site inspection (SI), is scored using the Hazard Ranking System (HRS) model. The score (ranging from 0 to 100) determines whether the proposed site is included on the NPL and remains under the continued auspices of the federal Superfund program. Other scoring and ranking systems are used by EPA in other phases of the Superfund program, although as shown later, they are considerably less formal and rigorous.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

In this chapter, the background and history of the HRS are presented, and the model's approach and structure are characterized and evaluated. The evolution of the HRS is traced in the context of the legislative mandate and management pressures that have guided and constrained EPA's administration of the Superfund program. The strengths and drawbacks of the HRS are discussed, with particular focus on changes that occurred with implementation of the revised HRS in December 1990.

BACKGROUND AND HISTORY

With the realization of the magnitude and potential impact of hazardous-waste contamination that occurred following the Love Canal incident in 1978, Congress passed the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) of 1980. That law granted EPA the authority to respond to current or potential releases of hazardous waste that could threaten "public health or welfare or the environment." It established the principle of strict, joint and several liability whereby all potentially responsible parties (PRPs) identified at a site are liable for the costs of addressing and removing the hazardous threat. A multi-billion dollar fund was established through taxes on petroleum and chemical feedstocks to pay the costs of response action and remediation in cases where viable PRPs were not present or in cases where immediate federal action was deemed necessary. This Superfund, administered by EPA, has since provided the name by which the entire CERCLA process and the sites themselves have become known.

The initial CERCLA legislation was debated and passed under highly charged conditions, and many of the involved parties, including EPA, were primarily concerned with establishing their position at the forefront of this new and powerful tool for harness-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

ing public concern and anger over environmental contamination (Landy et al., 1990). Many practical issues of implementation, such as methods for setting site priorities, were largely ignored in the development of the legislation. CERCLA did require EPA to establish "criteria for determining priorities among releases or threatened releases [of hazardous substances] through the United States for the purpose of taking remedial action." Furthermore, the "criteria and priorities . . . shall be based upon the relative risk or danger to public health or welfare or the environment" (emphasis added) (CERCLA, 1980, Section 105(8)(A)). These criteria were to take into account the following considerations as much as possible:

  • the population at risk;

  • hazard potential of substances at the facilities;

  • potential for contamination of drinking water supplies;

  • potential for direct human contact; and

  • potential for destruction of sensitive ecosystems

As highlighted above, the initial criteria and priorities were to consider public health, the environment, and public welfare. The HRS, however, is designed to focus solely upon human health and the environment, with socioeconomic impacts considered only in an indirect manner.

To determine which candidate sites would be included on the NPL, EPA contracted for the development of the original HRS model. The HRS model was developed by the MITRE Corporation to meet EPA's need for a multimedia environmental assessment model (Chang et al., 1981). At that time, multimedia assessment procedures were unavailable, and although pollutant transport and fate models had been developed for some of the individual pathways considered, those models were not connectable or comprehensive. Furthermore, methodologies for environmental and health risk assessment were just beginning to be devel-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

oped. The multimedia risk approach of the HRS model was thus very innovative for its time. Following scientific review and public comment, formal adoption of the HRS occurred with passage of the National Oil and Hazardous Substances Pollution Contingency Plan (40 CFR 300), which indicated that the original HRS would be "used to assess the relative threat associated with actual or potential releases of hazardous substances at sites" (Appendix A, 40 CFR 300).

Through the 1980s, dissatisfaction with the HRS, motivated in part by a desire to provide a more accurate representation of relative risks, particularly at large coal and other mining facilities, led to the push for modifications of the HRS. The requirement for modifications was included in the Superfund Amendments and Reauthorization Act (SARA) of 1986 which instructed EPA to amend the HRS to ensure, "to the maximum extent feasible, that the hazard ranking system accurately assesses the relative degree of risk to human health and the environment" (SARA, 1986, Section 105(C)(1)). It is noted in SARA that, given the need for expeditious site identification, the revised HRS is not required to be equivalent to a detailed risk assessment, but rather should be as accurate as possible using the screening level information usually available at the preliminary assessment (PA) and site inspection (SI) phases of the Superfund process. Further requirements of the mandated revisions included the need to consider potential and observed air contamination; effects through the human food chain; and better risk assessments for large-volume wastes, including the quantity, toxicity, and concentration of wastes and their potential for release to the environment. The target date given in 1986 for SARA-mandated revisions was October 1988; however, final promulgation of the revised HRS did not occur until December 1990 (Federal Register, 1990).

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

ROLE OF THE HRS IN THE SUPERFUND PROGRAM

The primary function of the HRS is to serve as the screening mechanism for determining which candidate sites are included on the Superfund NPL. The major steps in this process are summarized in Figure 4-1. A site where hazardous-waste problems are known or suspected is first placed on the CERCLA Information System (CERCLIS), which is the master list of hazardous-waste sites in the United States. A PA and SI are conducted to provide a screening evaluation of the site and to gather the necessary information for scoring the site with the HRS. The sites are scored under the auspices of a regional EPA office by designated contractors or state agencies and submitted to that office for review. The site is proposed for placement on the NPL if the final HRS score is greater than or equal to 28.50; if the score is below 28.50, the site is designated as "no further remediation action planned (NFRAP)" under the federal Superfund program. The selection of the 28.50 cutoff score was initially made in 1982; it was chosen to meet the CERCLA mandate (CERCLA, Section 105(8)(B)) that at least 400 of the approximately 700 CERCLIS sites first scored at that time would be included on the NPL. The cutoff number thus had no apparent significance in terms of an absolute level of environmental or human health risk. Sites proposed for the NPL as a result of their HRS score undergo a period of public comment, after which the final decision for inclusion on the NPL is made by EPA. Through February 1991, only 79 sites had been proposed but rejected for inclusion on the NPL, in most cases because their revised HRS score was below 28.50 or because the site was reclassified as a Resource Conservation and Recovery Act (RCRA) facility (EPA, 1991a). Other mechanisms are also available for placement of a site on the NPL. States are each allowed to nominate one high priority site irrespective of its HRS score. As of 1992,

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

FIGURE 4-1 The Superfund process and the role of the Hazard Ranking System

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

approximately two-thirds of the states had proposed sites in this manner (EPA, 1991b). In addition, a site can be placed on the NPL as a result of a health advisory from the Agency for Toxic Substances Disease Registry (ATSDR). Five sites have been proposed through this mechanism. With more than 1,200 sites now on the NPL, the HRS score has been the principal mechanism for determining whether candidate sites are nominated and included. This score thus serves a critical role in determining the priority and level of attention that a site will receive in the EPA Superfund program.

The steps shown in Figure 4-1 provide an idealized and highly simplified representation of the Superfund site-selection process and the role that the HRS plays in it. In actual practice, the process is more involved, as summarized in Figure 4-2. As shown near the bottom of the diagram, the HRS scoring must undergo review by EPA headquarters and a quality-assurance (QA) review by a contractor before the decision for nomination to the NPL. Furthermore, simplified screening versions of the HRS have evolved to allow sites to be prescored following the preliminary assessment (PA). The PA method is based upon the full HRS, but uses conservative default values for factors that are still unknown at the conclusion of the preliminary assessment phase of the analysis (EPA, 1991c). The PA method is designed to result in a score that is at least as high as the subsequent HRS score, and can therefore serve as a screening mechanism. The intent is to avoid, where possible, the expenditure of time and resources on sites where the potential for eventual inclusion on the NPL is low or nonexistent. The development of screening steps designed to eliminate false positives along the site-selection process has been largely motivated by the difficulties EPA has encountered in attempting to process the large number of CERCLIS sites under consideration for the NPL. As discussed later in this chapter, management considerations of this type, rather than environmental evaluation, have often been the driving factors in the evolution

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

FIGURE 4-2 More detailed steps in the HRS scoring process and National Priorities List development. Source: OTA, 1989.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

of the Superfund site-selection process and have subsequently affected the role of the HRS in this process.

In addition to its formal role in the NPL selection process, the HRS model has been used by others for various purposes. Some states use the HRS score to set priorities for sites under their jurisdiction that do not qualify for the NPL. Many states implemented their own site-scoring systems, which were often similar to the original HRS model. The HRS score is used by EPA regional offices as a starting point in their subsequent remedial investigation and feasibility study (RI/FS) priority process, discussed later in this chapter. A positive correlation between the HRS score and the pace of subsequent Superfund actions has been found by Hird (1990). Finally, the HRS score has been proposed as a general mechanism for quantifying risks from hazardous-waste sites and measuring the risk reduction achieved in subsequent remediation (Wilson, 1991; Butler and Jones, 1992).

MODEL STRUCTURE AND COMPONENTS

The HRS is a structured-value model in which various characteristics of the site, wastes, and surrounding environment are combined through use of a numerical algorithm to compute an overall score. As part of the calculations, separate scores are computed for each of four exposure pathways:

  • groundwater migration pathway (Sgw);

  • surface water migration pathway (Ssw);

  • soil exposure pathway (Ss); and

  • air migration pathway (Sa).

The overall score is determined as the root mean square average of the four pathway scores:

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

That score and each of the individual pathway scores range from 0 to 100, with higher scores reflecting higher degrees of threat. A schematic summary of the major components and calculations of the revised HRS model is presented in Figure 4-3. The algorithm is structured to include the effect of three factor categories:

  • likelihood of release or exposure;

  • characteristics of the wastes present at the site; and

  • characteristics of the target population or environment.

The score for each pathway is calculated as the product of its three factor category scores. The likelihood of release is determined based on the presence of an observed or potential release. Observed releases are verified with site monitoring data. The potential for release depends on pathway characteristics that either restrict or promote transport at or near the site.

The waste characteristics are chemical-specific and are intended to represent the properties of the chemical that indicate the likelihood of exposure and potential health hazard. The waste characteristics considered across all pathways include the toxicity, persistence or mobility, and hazardous-waste quantity. The bioaccumulation potential is considered in the surface water migration pathway for human food chain and environmental impacts.

The environmental and human health targets considered in the HRS vary across pathways. The groundwater migration pathway includes water supply wells, groundwater resources, and wellhead protection areas. The surface water migration pathway considers drinking water intakes, human food chain impacts, and sensitive environments. Calculations for the soil exposure pathway include potential health impacts to residents and workers on-site and

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

FIGURE 4-3B Individual pathway calculations of the HRS.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

nearby and on-site impacts on resources and sensitive terrestrial environments. The air migration pathway considers health impacts on nearby populations and environmental impacts on resources and sensitive systems.

Scientific Evaluation of Model Components

The HRS model includes a number of simple analytical and tabular functions for determining the values of individual factor scores. These functions have been developed from a combination of mechanistic factors, empirical relationships and subjective judgment. Given the empirical and subjective character of the model and the ambiguous nature of the overall computed score—intended to provide some index of risk, but not intended to be equivalent to a risk number per se—it is not feasible to perform a rigorous critique of the individual functions and factors that the HRS comprises. Still, it is desired that the relationships reflect good scientific judgment and that a consistent treatment be provided for different pathways and impacts. An evaluation of the basis and consistency of various components of the HRS is provided from this perspective, first with the HRS components common to all pathways and then by a derailed review of the individual pathways. This review considers the HRS as configured by the December 1990 revisions. Previous reviews have been made for the original HRS (e.g., Wu and Hilger, 1984); some of their comments have been addressed in the recent revisions, others remain pertinent.

Likelihood of Release Component

The likelihood of release component accounts for observed releases and the potential for a contaminant to be released. When

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

releases are observed for a particular pathway, the maximum score for that component is assigned directly, eliminating the need for further fate and transport considerations for that pathway. If no releases are observed, the potential for release to that pathway must be estimated.

The potential for contamination of the groundwater is based on the presence or absence of containment, the net precipitation, the depth to the aquifer, and the time it takes for the contaminant to reach the aquifer. The potential for contamination of surface water is based on the possibility of overland flow, which in turn depends on the presence or absence of containment, runoff characteristics of the site, and the distance to the surface water. If the site is in an area that is subject to flooding, then the likelihood of release is also dependent on the presence or absence of containment, and on the flood frequency. In addition, because contaminated groundwater may discharge to the surface water, consideration is given to the influence of containment, net precipitation, depth to the aquifer, and travel time for contaminated groundwater to discharge to the surface water. For the soil exposure pathway, only observed contamination is considered. For the air migration pathway, the potential release is considered for both gaseous and particulate emissions. For gaseous releases, the presence and effectiveness of gas containment measures, the type and source of the gas, and the gas migration potential are evaluated. For particulate releases, the presence and effectiveness of particulate containment measures, the particulate type and source, and the particulate migration potential are evaluated. Each of these considerations is based on site-specific information and does not consider the characteristics of the contaminants.

When a release has not been observed in a pathway and the potential for release is calculated instead, the maximum value for the likelihood of release component is taken as 90% of the value that would have been assigned had a release been observed. Since the HRS is used to rank sites before a full set of environmental data

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

has been collected, it is appropriate to allow a comparable, though slightly smaller, value for the likelihood of release component when the potential for release is high.

Waste-Characteristics Component

The quantity of hazardous substances and their characteristics of toxicity, persistence, and mobility are used to calculate waste-characteristic component scores for each of the four migration pathways. The toxicity factor for the hazardous substance and the hazardous-waste quantity are common to all the pathways, except for the surface water pathway, which treats ecosystem toxicity separately. The persistence and mobility factors are pathway-specific and are considered in the more detailed reviews that follow for the individual pathways.

The toxicity of each substance is derived either from the reference dose data (RfD) for chronic toxins or from weight-of-evidence slope factors for carcinogens. As a fallback position, if no data on chronic or carcinogenic toxicity are available, toxicity factors are determined from a table of LD50 (acute toxicity) data for various exposure pathways. A major flaw in this approach is that the primary method of determining the toxicity factor value is independent of the exposure pathway. Human toxicity is dependent on the exposure pathway, and a toxicity database developed for removal actions and other purposes should allow consideration of the exposure pathway. Table 2-4 of the Federal Register Final Rule statement for the HRS model (reproduced here as Table 4-1a, b,c) provides ambiguous guidance for toxicity factor evaluation by assigning a value of 0 if RfD and slope data are not available, and also directing the use of the table of acute toxicity in such a case (Federal Register, 1990).

The toxicity-persistence-mobility component of the waste-characteristic score for each pathway is calculated for each hazardous

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

TABLE 4-1A Toxicity Factor Evaluation: Chronic Toxicity (Human)

Reference dose (RfD) (mg/kg-d)

Assigned value

RfD < 0.0005

10,000

0.0005 < RfD < 0.005

1,000

0.005 < RfD < 0.05

100

0.05 < RfD < 0.5

10

0.5 < RfD

1

RfD not available

0

 

Source: Federal Register, 1990.

substance found at the site. The hazardous substance with the highest score is used to calculate the final score for the waste-characteristic component. This procedure, which allows different hazardous substances to contribute to the component scores for different migration pathways, has two flaws. First, the HRS procedure does not give greater scores to sites that have a large number of hazardous substances. Although methods for determining the overall impact of multiple chemicals and chemical mixtures are only in their infancy (e.g., Arcos et al., 1989), it is likely that the cumulative effect of many chemicals at a site will be more harmful than the effect from a single substance. Second, the method does not allows the weighting of hazardous substances based on the amount present. The latter flaw could result in greater scores for sites having a small quantity of a hazardous substance that is slightly more toxic than another hazardous substance found at other sites in much larger quantities. This could be corrected by selecting the hazardous substance used to represent each pathway to be the one yielding the greatest waste characteristic score, rather than the substance with the highest toxicity or combined-factor value. Such a selection is desirable because that score is a result of the product of the toxicity or combined-factor value and the hazardous-waste quantity factor value.

The hazardous-waste quantity factor is determined by estimating the mass of the hazardous substance at the site. For hazardous wastes that are listed for reasons other than toxicity, the entire

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

TABLE 4-1B Toxicity Factor Evaluation: Carcinogenicity (Human)

Weight-of-evidencea/slope factor (mg/kg-day)-1

Assigned value

A

B

C

 

0.5 < SFb

5 < SF

50 < SF

10,000

0.5 < SF < 0.5

0.5 < SF < 5

5 < SF < 50

1,000

SF < 0.05

0.05 < SF < 0.5

0.5 < SF < 5

100

 

SF < 0.05

SF < 0.5

10

Slope factor not available

Slope factor not available

Slope factor not available

0

a A, B, and C refer to weight-of-evidence categories. Assign substances with a weight-of-evidence category of D (inadequate evidence of carcinogenicity) or E (evidence of lack of carcinogenicity) a value of 0 for carcinogenicity.

b Slope factor.

Source: Federal Register, 1990.

mass of the waste is estimated. Hazardous-waste streams are arbitrarily assumed to contain 0.02% of a hazardous constituent; waste in landfills is assumed to contain 2 × 10-5% hazardous constituents, while soil in a land treatment facility is assumed to contain 0.0086% hazardous material. No rationale is evident for these and other values selected for determining the hazardous-waste quantity.

The product of the toxicity factor and the hazardous-waste quantity factor, which is a "hazard-scaled" mass of the substance or waste present, is assigned a second scaling factor according to the log10 of the product. Although this may be appropriate, no rationale for the logarithmic relationship is provided, and this approach might result in an underweighting of sites for which the waste characteristic product is high.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

TABLE 4-1C Toxicity Factor Evaluation: Acute Toxicity (Human)a

Oral LD10 (mg/kg)

Dermal LD50 (mg/kg)

Dust or mist LC50 (mg/l)

Gas or vapor LC50 (ppm)

Assigned value

LD50< 5

LD50< 2

LC50< 0.2

LC50< 20

1,000

5 < LD50 < 50

2 < LD50 < 20

0.2 < LC50 < 2

20 < LC50 < 200

100

50 < LD50 < 500

20 < LD50 < 200

2 < LC50 < 20

200 < LC50 < 2,000

10

500 < LD50

200 < LD50

20 < LC50

2,000 < LC50

1

LD50 not available

LD50 not available

LC50 not available

LC50 not available

0

a LD50 refers to a toxicant dose that is lethal for 50 percent of the test subjects. LC50 refers to a toxicant concentration that is lethal for 50% of the test subjects.

Source: Federal Register, 1990.

Critique of the HRS Toxicology and Exposure Components

To predict the relative degree of human health hazard posed by chemicals of different toxicity associated with potential NPL sites, the HRS strategy employs toxicity factors computed from regulatory limit concentrations and screening concentrations. Chemical properties such as mobility and persistence are considered with toxicity to describe the waste characteristics. To some extent, these attributes of the HRS model mimic the "exposure assessment" and "dose-response assessment" portions of a quantitative health-risk assessment. However, as demonstrated in this sec-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

tion, the model appears to overemphasize toxicity considerations by assigning only relatively low levels of expected environmental human exposure. An internal inconsistency in the ranking of carcinogens versus noncarcinogens also appears. In most cases, and particularly with regard to potential acute exposures, carcinogens seem to be overweighted in severity of effect compared to noncarcinogens.

The greater the toxicity of chemicals present at a site, the greater the potential for harm. Toxicity factors are thus appropriate for inclusion in the HRS. However, toxicity factors for different health endpoints are not directly comparable. Consider the numerical expression for the toxicity of carcinogens versus that of noncarcinogens. Carcinogens are considered to act by non-threshold mechanisms, while noncarcinogens are assumed to have thresholds for toxicity. Because both types of contaminants are found at waste sites, it is necessary to compare and weight them for the overall toxicity score.

Different procedures are used for assigning relative weights to carcinogens and noncarcinogens in the HRS. As shown in Table 4-1 (HRS Table 2-4), the reference dose (RfD) is used to score chronic toxicity for noncarcinogens and the slope factor is used to score carcinogens for the waste characteristics factor category. For noncarcinogens with the same score as "B" carcinogens, the RfD is equivalent to a lifetime cancer risk of 7.8 × 10-3. In the targets factor category, screening concentrations are used as triggers to place observed contaminant concentrations into the Level II or more serious Level I category.

Screening concentrations are concentrations that result in the RfD for noncarcinogens and a 1 × 10-6 risk for carcinogens. The equivalent cancer risk for a noncarcinogen at the RfD is thus nearly four orders of magnitude lower in the targets factor category than in the Waste Characteristics factor category (1 × 10-6 vs. 7.8 × 10-3). The HRS documentation provides no apparent reason for this inconsistency.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

The RfD for a noncarcinogen is the maximum acceptable dose to which a person could be exposed over a lifetime with no ill effects. Similarly, one in one million excess risk of cancer is considered to be a de minimus risk. The use of these gauges for ranking toxic effects of chemicals is most appropriate for doses that are close to the trigger levels. Doses much higher than the trigger levels would produce a different ranking order for hazard because of each chemical's unequal advance toward acute or lethal effects from a low or acceptable dose. In particular, for carcinogens, the dose-response curve is assumed to be linear with a constant slope. One thousand times a risk of 1 × 10-6 is quite far below the background incidence rate of cancer of 25%. One thousand times an RfD on the other hand would be a lethal dose for many noncarcinogens. Thus, evaluations of the toxicity of noncarcinogens should consider the concentration relative to the reference and the lethal dose.

To illustrate the potential distortion produced by the HRS at high dose levels, consider a comparison of cyanide (CN) and cadmium (Cd). Assume 1mg/kg-day doses are generated for CN at one site and for Cd at another; CN is then at 50 times its RfD and Cd is at 2000 times its RfD. This results in assigned toxicity-factor values of 100 and 1000, respectively, showing Cd to be worse than CN. At this dose, CN is at 0.2 times its LD50 and Cd at 0.004 times. From this perspective, CN is a much more serious threat than Cd, but its ranking indicates it is 10 times less serious. Although most sites would not produce doses of this magnitude and the ranking system is properly aimed at the possibility of long-term chronic exposure, sites yielding doses significantly above the RfD could be incorrectly ranked.

The toxic properties of chemicals also affect the HRS ranking through the use of regulatory limits and screening levels described in Section 2.5 of the Federal Register (1990) Final Rule. The concentration of a chemical detected in the groundwater or surface water at a site is compared to existing regulatory limits such as

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

Maximum Contaminant Levels (MCLs) or Food and Drug Administration Action Levels (Federal Register, 1990, Section 2.5.2). If these are not available, the contaminant concentration is compared to screening concentrations that produce a 1 × 10-6 risk for carcinogens or the RfD for noncarcinogens. Although the regulatory limits listed in the HRS in general reflect toxicity, they also incorporate other factors, such as practical quantification limits and cost of compliance. For carcinogens, this means that an MCL might reflect a risk greater than 1 × 10-6. An example is chloroform, for which the drinking water MCL is 100 µg/L, representing a cancer risk of 1.7 × 10-5 (based on a unit risk of 1.7 × 10-7 L/µg for chloroform in drinking water). The use of regulatory limits in the HRS can thus produce results quite different from a quantitative risk assessment, which considers only toxicity.

The amount of exposure to a contaminant, as well as the degree of its toxicity, determines the risk of an adverse health effect. Procedures for quantitative health risk assessment consider these aspects separately. The HRS combines toxicity with indices of mobility, persistence, and bioaccumulation. Those factors are multiplied in separate steps of the HRS and then combined in the overall waste characteristics category. Many other attributes of the HRS serve as surrogates for exposure, such as waste quantity, containment, and soil characteristics. The lack of separate subscores indicative of chemical toxicity and of exposure likelihood and magnitude makes it difficult to assess how the HRS conceptually compares to a quantitative risk assessment.

Groundwater Pathway

A groundwater migration pathway subscore is calculated for each aquifer at the site, and the highest of theses is used. Net precipitation factor values are derived from maps or on-site data for

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

annual precipitation, less evapotranspiration. Unfortunately, the method does not take into account runoff, which can significantly reduce the flow rate of water to the subsurface, even for soils with a relatively high infiltration rate. When the map of net precipitation factor values (Fig. 3-2 on page 84 of the final HRS rule) is compared to net percolation depths for soils of Hydrological Groups A, B and C (Brown et al., 1977), it is evident that even though the shapes of some of the zones are similar in some parts of the country, they differ significantly in others. For example, areas of greatest percolation to the groundwater are in eastern Tennessee and Kentucky, not in northern Alabama and Georgia and eastern North Carolina as shown. These discrepancies could result in the erroneous scoring of some sites.

Although one might argue that it is sufficient for the present purpose to use net precipitation without accounting for runoff, such detail is included in other parts of the calculations. Specifically, soil hydrological groups and rainfall runoff values are used in Section 4.1.2.1.2.1.2 of the final HRS rule for the surface water component of the scoring. It would be better to use the same level of detail for calculations throughout the scoring.

The model's value representing likelihood of release to the groundwater includes factors on the depth and travel time to the aquifer. Although they are structured in such a way that the travel time is based on the thickness of the restricting layer and not the entire depth to the aquifer, inclusion of both factors introduces some degree of double counting, depending on the relative thickness of the most restrictive zone. Furthermore, the value of the hydraulic conductivity of 10-8 cm/sec assigned to clay and low permeable till (compact unfractured till) in Table 3-6 of the HRS final rule is inappropriate. Such materials often have much greater permeabilities (Freeze and Cherry, 1979; Griffin et al., 1985), and the use of the low conductivity values will result in very low factor scores. Furthermore, at some sites with hazardous wastes, channels for leakage have been created by drilling activities such as for

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

wells and mine shafts. Natural faults allow liquids to flow rapidly through zones that would otherwise be classified as having low permeability.

The mobility of each hazardous substance in groundwater is governed by a number of factors including the water solubility of the substance and its soil sorption coefficient. The model's procedure appropriately results in greater values for karst environments, and higher values if the substance is present as a liquid. However, the partition coefficient, Kd, can only adequately represent partitioning between water-soluble substances and the solid matrix. It is not appropriate to use it when a nonaqueous phase liquid (NAPL) is in direct contact with the soil solids. Thus, it is inappropriate to adjust the groundwater mobility factor values downward from 1 for liquid wastes when the Kd value is > 10 ml/g. The value should remain at unity whenever an NAPL is present. To do otherwise will underestimate the potential for mobility and result in an underscoring of sites. The mechanisms that facilitate movement (or, conversely, cause retardation) are quite different for inorganic and organic species. Thus combining Kd and water solubility in an apparent attempt to handle inorganic and organic species in the same table is not justified. Kd is widely used for organic species and calculated as follows: Kd = (Koc) foc, where foc is the fraction of organic carbon and Koc is the partition coefficient between the water and the organic carbon in the soil (Roy and Griffin, 1989). In general, is not necessary to include the fraction of clay since there is negligible adsorption of organics on the clay fraction if water is present, even for NAPLs.

Solubility only partially controls the movement of inorganic species. Partitioning of cations is regulated by their speciation (valence state, type and degree of organic ligand formation, pH, and oxidation state), the charge density of the medium, and the presence of competing ions. In most of the situations where cations are present the potential concentrations at the receptor points would be well below the solubility of the compound in water. The groundwater mobility factor value does change in the

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

appropriate direction as the Kd changes, but the suggested dependence on solubility is not technically correct and could lead to erroneous scores, particularly when comparing sites with primarily organic contaminants to those with primarily inorganic contaminants.

Surface Water Pathway

The surface water migration pathway considers overland flow and release caused by flooding. The runoff factor value of the overland flow component is derived from the U.S. Department of Agriculture (USDA) curve number (SCS, 1972) using the two-year, 24-hour rainfall frequency. This approach is appropriate but the resulting values are not proportional to the amounts of runoff that are predicted for the soil hydrological groups. The HRS factor values for soil groups A, B, C, and D for the 3.5-inch rainfall can be normalized to be 1, 1.33, 1.66, and 2 respectively. However, actual average runoff amounts generated by the USDA method have ratios of 1, 2, 3.6, and 4.2 respectively. Thus, the HRS procedure undervalues the amount of runoff for the less permeable soils in groups C and D, relative to those in Group A, by a factor of over 2. It is possible that this was done deliberately to account for increased dilution resulting from greater runoff, but this was not documented and is not appropriate since much of the transport of contaminants will be associated with erosion, which increases as runoff increases.

The flood frequency factor values are equal for annual and 10-year flood plains. A small adjustment, perhaps setting the value for the 10-year flood plain 10% lower than that for the annual flood plain, would be consistent with similar distributions assigned elsewhere in the procedure. Persistence factor values for substances in surface water are determined as the greater of the values determined either by the half-life or the logarithm of the

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

octanol-water partition coefficient, log Kow. Although mobility is dependent on the Kow in the environment, and Kow may be related to the relative partitioning of chemicals into the fat in fish (as it is correctly used in calculating the bioaccumulation factor), it is not necessarily related to the rate at which hazardous substances are metabolized. Thus, its inclusion for the purpose of estimating persistence may be inappropriate, e.g., for poorly degradable water-soluble heterocyclics. The bioaccumulation factor values may be arguable whether calculated from actual bioconcentration data or from log of Kow, but they are not related to water solubility. Thus, this portion of Table 4-15 of the final HRS rule might give misleading scores.

Soil Pathway

The soil exposure pathway score is based on the exposure of workers, residents and nearby populations. It also includes a ranking for the sensitivity of terrestrial environments. The scores are based on the size of the contaminated area and the value of the hazardous-waste quantity factor for the selected contaminant. The only pathway considered for soil exposure is direct ingestion by residents, site workers, or other individuals who may visit the site. Inhalation of associated gases or particles are addressed in the air pathway. Dermal exposure is not addressed explicitly. Because exposure is assumed to occur when individuals visit the site, there are no considerations of fate and mobility contributing to the soil pathway score.

Air Pathway

The air migration pathway score is based on the exposure of

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

individuals or the population within one mile to observed or potential gaseous or particulate releases for the selected hazardous substance. Factor values used in the score for potential releases include containment considerations, source-type considerations, and the vapor pressure and Henry's Law constants for the substance (Table 6-4 of the final HRS rule). The assigned values appear to be internally consistent, and also consistent with our understanding of the importance of each, with the exception of the apparently low factor value assigned for evidence of biogas release from landfills. It is known (Wood and Potter, 1987; Smith et al., 1989) that landfill gases are effective in transporting hazardous substances, and thus this value appears low as compared with others in Table 6-4 of the final HRS rule. The value for potential release from surface impoundments also appears low, since these are direct sources of air emissions.

Particulate migration is based on observed releases or the site-specific mobility factor used to calculate the likelihood of release. Mobility is also a component of the waste characteristic factor value and is determined from site-specific information or ranges provided. Thus, particulate mobility is included in the scoring twice for substances with lower vapor pressures. The maps provided to determine the particulate migration potential factor values and the particulate mobility factor values are different from each other and also differ from maps delineating areas of actual (Plaster, 1985) or potential (Donahue et al., 1977) wind erosion for the continental United States.

Summary Evaluation of Pathway Calculations

The fate and mobility factors of hazardous substances are used to determine potential exposure for three of the four pathways. In

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

most instances, the judgmental values used to evaluate containment appear to be appropriate. Double counting for factors influencing mobility is evident in the groundwater and air migration pathways. Since the score is based on just one substance for each pathway, it is critical that the most appropriate substance be selected. The selection should include consideration of the quantity of the hazardous substances in addition to the toxicity and mobility factors that are now the only controlling factors. Another problem with the scoring methodology is that the toxicity factor is not weighted for the means by which the individual is exposed.

There are differences in the details of the mechanisms controlling mobility in the different pathways. Such differences might or might not have a significant impact on the score, but a consistent level of mechanistic detail should have been used throughout. In a few instances, it is apparent that the judgmental values assigned to site-specific conditions may not be completely reflective of the relative hazards. Examples of this are consideration of the presence of NAPL and the mobility attributable to landfill gas, which are likely undervalued. Although the scoring procedure is generally logical in terms of the direction of effects of different input factors, some of the observed flaws in the fate and transport components of the scores could result in the scores of sites, particularly those with no observed releases, being inappropriately ranked.

Ecological Factors in the HRS

An important aspect of the recent HRS revisions was EPA's desire to "improve the evaluation of sensitive environments by addressing a broader range of sensitive ecosystems and to afford a higher weight for sensitive environmental factors" (Caldwell and Ortiz, 1989). Although sensitive environments were included in

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

the original HRS, a site that had only environmental impacts in the absence of human health effects, even if it had major impacts on an endangered species or a national park, could not score high enough for placement on the NPL.

The current HRS includes impacts on sensitive environments in the surface water, soil, and air migration pathways. Sensitive environments in the surface water pathway are rated from 5 points for state-designated areas for protection or maintenance of aquatic life, to 100 points for critical habitats and other formal federally designated areas. Soil impacts are included for sensitive terrestrial environments, with ratings ranging from 25 for state lands designated for wildlife management to 100 for federally designated critical habitats and endangered species areas. A range of sensitive environments is considered for the air pathway, with special emphasis on large wetland areas.

An important issue in understanding and comparing the different priority-setting and ranking models for hazardous-waste sites is the relative weights applied to the ecological versus human health targets. Are the weights clear and identifiable in the development and presentation of the model, and is an adequate (or for that matter, any) rationale presented for their selection? The weights for environmental impact in the HRS remain lower than those assigned to human health, but are intended to be high enough that sites which seriously threaten an important sensitive environment can score sufficiently high for placement on the NPL. However, the precise weighting in the HRS is difficult to determine because of the various multiplicative and additive steps in the algorithm.

The environmental versus human health weights in the HRS were established using a Delphi method within the EPA work-group (Caldwell and Ortiz, 1989). Although public comment on the weights was solicited in the preamble to the proposed rule, it is not clear how broadly representative the weights were intended to be, and there are no provisions for adjusting the weights to ac-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

count for differences in local or regional preferences and values that may be relevant at a site. As discussed in Chapter 9, the committee believes that it would be desirable for a national model to have adjustable value weights to reflect local preferences, perhaps on a state-by-state basis.

Socioeconomic Factors in the HRS

Potential economic benefits from remediating hazardous-waste sites include reduction of community damage, appreciation of property, increased productivity of land, creation of jobs, and reduced expenditures of health care. Social benefits include enhancement of existing communities, especially disadvantaged communities. Although there have been few empirical studies of those benefits, despite their potential importance, recent studies have shown some economic impacts. For example, McClelland et al. (1990) reported that housing values in a Los Angeles neighborhood increased by approximately $5,000 per unit after a landfill was closed. Skaburskis (1989) pointed to a 15% decrease in sale prices near a landfill, but showed no impact beyond 0.25 mile from a site that he described as noncontroversial.

Social impacts of siting and remediating hazardous-waste repositories are also likely to be important, but are difficult to document. Edelstein (1988) interviewed residents of communities with landfills and hazardous-waste sites, attended public meetings, and read meeting transcripts. He concluded that stigmatizing land uses gradually changed people's self-image, image of their family, and images of their community, environment, and government. Adults and children became depressed and pessimistic and felt betrayed by their government. It was also reported that many moved and others wished that their homes would burn down, so

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

that they could afford to move. The implication of this research is that remediation can ameliorate such problems.

Economic effects that are generally evaluated include those involving site and off-site land use and property values, employees, and neighboring populations. For example, remediation of contamination at a site might increase the values of nearby residential, farm, industrial, and commercial properties; encourage nearby land-use investment and development; and positively affect the future livelihood of nearby citizens.

Although the original directive to EPA for establishing criteria and priorities for site remediation specifically indicated inclusion of consideration of public welfare, only health and environmental impacts are directly considered in the HRS. Public welfare is considered in an indirect manner, as it is affected by human health and environmental resource impacts. In this respect, the HRS does consider a reasonably broad and representative set of human and environmental resource targets. The human targets include resident, student, and worker populations, including the individuals nearest the site. However, transient populations are not considered, and no consideration is given to the age, sex, or socioeconomic status of the target group or individual. The environmental resource targets (in addition to sensitive environments) include commercial farming, food preparation, recreational areas, and drinking water supplies.

No attempt is made in the HRS to incorporate direct estimates of economic impact (such as property value losses near the site) or social effects (such as disruption of existing communities or distributional effects). The human population and resource components of the HRS do provide some measure of these, and more quantitative, detailed estimates would probably be difficult to make, certainly at the early assessment stage of the HRS. Still, EPA has made the establishment of close ties and effective communication with local communities an important part of their

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

Superfund strategy, and some further consideration, such as designation of "sensitive communities" (in parallel with "sensitive environments"), could be appropriate.

Protocol for Model Development and Use

As discussed in Chapters 2 and 3, the quality and acceptability of a hazard-ranking model is affected by the procedures and protocols adhered to in the development and application of the model. Included in this are the procedures for model testing and validation, the determination of the sensitivity of the model results to uncertainties in model inputs and formulation, and the provision of effective quality control mechanisms to ensure proper and consistent application of the model to site scoring. Additional issues in model development and use include the level of field testing and peer and public review of the model prior to its release for general use, the degree of user-friendliness, its transparency, and the quality of the documentation that guides the data collection and scoring steps.

In the development of the original and the revised HRS model, EPA, in conjunction with the MITRE Corporation, undertook extensive field testing programs (Chang et al., 1981; Caldwell and Ortiz, 1989; Zaragoza, 1990). These exercises helped to work out a number of early computational errors and perceived inconsistencies in the model. That effort, combined with extensive opportunities for peer review and public comment provided by EPA, has led to a model formulation that many have accepted as reasonable, though not necessarily ideal (Wu and Hilger, 1984; EPA, 1988; Wilson, 1991; Haness and Warwick, 1991).

The quality of the documentation that describes the HRS model is considered generally adequate. It is written in a straightforward manner and decomposes the scoring steps into multiple indepen-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

dent tasks. However, there has been little guidance for sampling and collecting the data that are input to the HRS, a gap which has allowed a wide range in data collection procedures and corresponding levels of effort to collect data. EPA is in the process of developing a Hazard Ranking System Guidance Document that will address this need for individuals or groups who are scoring sites.

In general, the HRS is structured so that the collection of more data leads to a higher score. Lack of data or uncertainty in inputs tends to skew the results towards lower values (Haness and Warwick, 1991). This "reward" structure for greater data-collection effort may make it possible for interested parties to manipulate the HRS score to meet their underlying objectives. States or communities that wish to keep a site off the NPL can limit the sampling and data-collection effort at that site. Conversely, states or communities that are motivated by economic or political factors to have a site placed on the NPL might continue the sampling effort until enough data is uncovered to push the score above the 28.5 threshold. Indeed, affluent communities might be more able than poor communities to invest in the necessary sampling to accomplish this, yielding a potential for socioeconomic inequity in the site selection process. This phenomenon can limit the ability of the HRS to provide a truly representative ranking based on objective environmental criteria.

The presence of the 28.5 threshold can lead to a pattern of behavior among scorers that further limits the utility of the HRS score for subsequent comparison of sites. Many site evaluators will collect the data necessary to push the HRS score above 28.5, then stop. Once the score passes this threshold, the site will be on the NPL, and there is no need for a further scoring effort. This limits the utility of the HRS score as an indicator of relative risk between sites once they are on the NPL; some sites with scores of 29 or 30 might have received higher scores had not the scorers focused their effort solely on the 28.5 threshold.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

A number of validation studies have been performed for the HRS model, in an attempt to relate the HRS score to the results of more formal risk analyses or in-depth expert panel studies of a group of sites (Applied Decisions Analysis, Inc., 1987; OTA, 1989; Dory and Travis, 1990). In these studies, little correlation was found between the HRS score and the more rigorous risk estimates. An underlying assumption in these comparisons is that the full risk assessments or panel studies represent "truth". Given the high degree of uncertainty in risk analysis procedures and the wide range of expert opinion that might pertain to the various complexities of a hazardous-waste site, these measures of truth might themselves be highly uncertain and suspect. Furthermore, in the analysis performed by Doty and Travis (1990), the HRS scores were compared to risk estimates based solely on human health impacts. Still, the referenced studies provide the only current comparisons of HRS scores to more detailed site hazard estimates, and based on these more objective and rigorous measures of risk and threat, it is likely that the HRS does yield a significant number of false positives (sites included on the NPL that should not be included) and false negatives (sites left off the NPL that should be included.

Given the potential for errors in NPL decisions based on the HRS score, an ability to consider the accuracy, precision, and sensitivity of the score should be available. Sensitivity analyses of the HRS model have been performed, although the multiplicative nature of the model precludes the determination of an absolute sensitivity since the response to a particular factor depends on the values assigned to the other factors in the model (Haness and Warwick, 1991). In a sensitivity analysis performed by EPA to assist in the development of the PA method (the simplified version of the HRS used for screening at the preliminary assessment stage intended to provide a conservative first estimate of the subsequent HRS score), the model results for 110 test sites were found to be particularly sensitive to the combined contaminant charac-

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

teristic score, which incorporates toxicity, mobility, persistence, bioaccumulation potential, and ecotoxicity factors (EPA, 1991c). The setting of these contaminant characteristic factors to their maximum value, which can occur at actual Superfund sites, was the only simplification to the full HRS considered by EPA that re-suited in PA method scores being significantly different (higher) at many sites from the HRS scores ultimately determined with PA and SI information.

Although selective sensitivity analyses have been performed, no formal mechanism for sensitivity or uncertainty analysis is provided as part of the regular HRS scoring procedure. The process does provide for a review of scores by EPA headquarters and a quality assurance check by contractor personnel. This review plus the QA efforts help to eliminate major errors and ensure a degree of consistency in the scoring process, and are conducted for all sites with initial scores above 25.0. Although the review and QA efforts are helpful, the small range of underscoring allowed for further review (between 25.0 and 28.5) and the exactness of the final cutoff value (28.5, clearly well beyond the precision of environmental assessment models of any type) dictate that a more formal consideration and allowance for uncertainty should be incorporated as part of the HRS process. Such a recommendation, based on detailed review of sites with a wider range of HRS scores, has been put forward by the U.S. Congress Office of Technology Assessment (OTA) and is considered later in this chapter.

Although recognizing that the HRS cannot provide an accurate absolute measure of environmental threat at the level of a derailed risk assessment, proponents of the HRS have generally found the simplicity of the model to be advantageous for consistent application and transparent evaluation of model results, even by non-experts. However, early reports indicate that this user-friendliness has been compromised to some extent with the promulgation of the revised HRS, which is considered next.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

THE EFFECT OF THE 1990 HRS REVISIONS

The revised HRS evolved through an extensive procedure of model development, field testing, peer review, and public comment (EPA, 1988; Caldwell and Ortiz, 1989). The model was modified by eliminating the direct contact pathway and the fire and explosion pathway, and by adding a soil exposure pathway. The eliminated pathways were designed to determine the need for immediate removal (emergency) action, and it was thought that they could be better addressed outside the long-term scope of the NPL process. Additions to the model included

  • a new exposure pathway for contact with contaminated soils;

  • consideration of chronic noncarcinogenic toxicity;

  • expansion of the ecological components of the model, allowing for consideration of a wider range of sensitive environments;

  • consideration of the potential for air emissions;

  • a groundwater-to-surface water migration pathway;

  • use of concentration data in determining waste quantity; and

  • higher weights for actual exposure and for potential exposure closer to the site.

Some these changes were intended to allow the revised HRS to correspond more closely to a risk-assessment procedure, as described in Johnson and Zaragoza (1991).

The changes to the HRS have resulted in a tool that is superior to the previous version in both the range of issues considered and the types of input data used. However, this improvement has occurred at the cost of a significant increase in model complexity and the amount of effort required to collect the input data and perform the model evaluation. This increase in complexity and in the amount of resources required has taken the model from the realm of an easy-to-use, accessible tool to one that is significantly

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

more difficult to use. The model can no longer be taught as a lab project for engineering students, and it is no longer possible to teach the use of the model in a workshop for groups of lay citizens interested in scoring a site or checking the scoring of a site performed by their EPA or state officials (L. Greet, National Resources Defense Council, Washington, D.C., pets. comm., 1991). Perhaps because of this, a number of states have recently adopted simpler ranking procedures for their non-NPL sites (see Chapter 7). Although dissatisfaction with the added complexity of the revised HRS might diminish as regular EPA scorers gain experience with the model, it is apparent that the model is now less accessible to other users than was previously the case.

A related concern with the revised HRS is that the improvements in the coverage and rigor with which environmental inputs are evaluated may be lost within the empirical, ad hoc algorithm that underlies the HRS calculation. As such, the additional effort at input data collection (which now approaches that of a full risk assessment) might be wasted when the data are processed through the HRS algorithm. From this perspective, structured-value models are only useful when they are kept simple and transparent. Once they are complicated to the extent that the required level of effort and understanding approaches that for a more rigorous scientific procedure, then arguably the structured-value model should be abandoned in favor of the more rigorous approach.

An important issue that arose in revising the HRS model was whether to modify the 28.5 cutoff value used for inclusion of sites on the NPL. To address this issue, EPA considered a different cutoff score that would be "functionally equivalent" to the 28.5 score in the original HRS, where functional equivalency could be defined based on the following: statistical correlation between the original and revised scores, an equivalent number of sites above the cutoff threshold, or an equivalent risk level for inclusion of sites on the NPL. The criterion of equivalent risk was considered appropriate by most commentators to EPA, including states and

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

industry (Zaragoza, 1990). To support the evaluation, EPA compared original and revised HRS scores at 110 test sites (Zaragoza, 1990). The revised HRS scores tended to be somewhat lower, with fewer of the 110 sites scoring above the 28.5 threshold with the revised model (see Figure 4-4). However, based on a qualitative assessment of risks at selected sites (and given the administrative and legal difficulties that could result from a change in the threshold value), EPA did not feel there was sufficient cause to lower the threshold value to maintain a risk equivalency. The 28.5 threshold for inclusion on the NPL was thus maintained.

FIGURE 4-4 Scatter plot of site scores for the original and revised HRS. Results were obtained from an EPA study to evaluate the 28.5 cutoff. Source: Wells et al., 1990). Reprinted with permission; copyright 1990, Hazardous Materials Control Research Institute, Rockville, Md.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

PRIORITY SETTING AT LATER STAGES OF SUPERFUND

A backlog of unremediated sites developed as a result of the unexpectedly large number of sites listed on the NPL as well as the significant time and effort required to complete detailed site studies and reach agreement on appropriate plans for remedial action. Such a backlog dictated that EPA develop procedures for priority-setting activities in the later stages of the Superfund process. These procedures include the regional remedial investigation (RI) and feasibility study (FS) priority-setting process and the remedial action (RA) priority-setting process. These processes are not mandated by law and are considerably simpler and less formal than the HRS. However, they do affect the priority assigned to different Superfund sites for remediation and thus need to be considered in this review of alternative priority-setting methods.

The regional RI/FS priority-setting process is not a formal model, but rather a systematic procedure that individual EPA regions must establish to determine priorities for RI/FS projects. The process is applicable to sites and individual operating units and is based on a "worst-first" principle that allocates resources so as to have the "greatest impact on human health and the environment." (EPA, 1990c). The method is applicable only to sites where the costs of the RI/FS could be covered by the Superfund budget; sites where no Superfund dollars are spent, such as federal facility activities or state-initiated enforcements, are exempt from the policy. In addition, other management considerations can be evoked by the regional office to override the evaluation of priority level, including enforcement considerations such as the presence of a willing and financially viable potentially responsible party, or the desire to push forward in-house projects for training, operating unit projects needed for site completion, or projects for sites with multiple, interdependent operating units. State involvement in the priority-setting process is also encouraged.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

To implement the worst-first principle, regions are directed to use the HRS package as a starting point and to consider additional information using "standard environmental criteria." These criteria, also used to determine remedial action priorities, include: the risk of contaminants (nature of principal threats), the stability of contaminants, whether human populations are exposed, and threats to significant environments. The result of the priority-setting process is to classify projects which are candidates for RI/FS into three tiers: highest, next-highest, and relatively low priority.

The RA priority-setting process is somewhat more formal, but stir quite simple compared with the HRS. Regions determine scores for the standard environmental criteria and for program management considerations, based on questionnaires and a panel review. The scores are combined in a structured-value model, using the weights shown in Table 4-2.

Based on this result, remedial action starts are classified into three categories:

  • Priority 1: Immediate or imminent threat.

  • Priority 2: Threat from current situation.

  • Priority 3: Threat from future situation.

The RA priority-setting process incorporates aggregate, subjective evaluations, but it is simple and quite transparent, so that the reasons for a particular site receiving a given score are clear. This type of model is appropriate for an internal, administrative function, but lacks the formality and replicability of a priority-setting process required by law, such as the NPL selection process wherein the HRS is used.

PROPOSALS FOR IMPROVING SUPERFUND SITE SELECTION AND PRIORITY SETTING

Several suggestions for improving the Superfund site-selection

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

TABLE 4-2 Weights for Remedial Action Priority Setting

Criteria

Score

Weight

Maximum Score

Risk of contaminants

1-5

× 5

25

Stability

1-5

× 5

25

Population exposed

1-5

× 4

20

Significant environment

1-5

× 3

15

Program management

1-5

× 3

15

 

 

 

100

 

Source: D. Evans, EPA, unpublished data presented to the committee, April 10, 1991.

process have been put forth in recent proposals by OTA and other sources. The OTA report, "Coming Clean, Superfund Problems Can Be Solved," (OTA, 1989), notes that many of the features of the site-selection process have been motivated by institutional management constraints rather than environmental or cost-benefit considerations for society as a whole. As mentioned previously, the initial selection of the HRS cutoff value of 28.5 was not based on any inherent environmental-risk threshold or cost-benefit trade-off, but rather the desire at the start of the program to allow an administratively manageable number of sites onto the NPL. The number of sites on the NPL has since grown considerably; however, the HRS cutoff of 28.5 has remained, reflecting the reality that regulatory criteria, once in place, are difficult to change.

Motivated largely by the criticism it received through the early years of Superfund, EPA has modified the Superfund process to encourage quicker progress along the path towards final remediation and closure. The need to demonstrate better administrative progress and control, mandated in part by SARA, has also provided a motivation for EPA to limit the number of sites in the selection pipeline. Fewer sites entering the NPL allow for a better record of progress on those sites that do enter. OTA believes that

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

this motivation has discouraged EPA from undertaking an active-site discovery program, which might lead to hundreds of thousands of sites being placed on CERCLIS, but would avoid future problems that otherwise would occur as these sites are discovered in a delayed and random manner, often by (unpleasant) surprise. Similarly, the motivation to demonstrate progress has encouraged EPA to implement the prescreening evaluations shown in Figure 4-2 (such as the PA method) to further trim the number of sites in the pipeline. Although these decisions are a logical result of the administrative pressures placed on EPA, they do not necessarily encourage environmentally sound decisions. In particular, the strong push to avoid false positives can potentially lead to an increase in the rate of false negatives.

Two of the recommendations put forth by OTA to improve the administration of Superfund directly concern the use of the HRS in site selection. The first is to eliminate the 28.5 score as an exact threshold for inclusion on the NPL. Instead, two HRS scores, one higher and one lower, would be used to classify the candidate sites into three groups. Those above the high score would be selected for immediate inclusion on the NPL. Those below the low score would be deleted as NFRAP cases. Those with scores in the range between the low and high score would then be subject to further review by an expert panel with authority to make the final recommendation for NPL nomination. This procedure would help to eliminate both false positives and false negatives, by allowing more careful evaluation of those sites where selection errors are most likely to occur.

The second OTA recommendation is to combine the PA and SI and RI/FS phase of the Superfund process into a single site-evaluation step, which would then be followed by the HRS scoring. This recommendation, intended to streamline and expedite the overall process, would provide a higher level of site information for use in the HRS scoring. Indeed, as discussed above, many think that the increased data needs of the revised HRS now require information beyond the typical PA and SI effort.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

Additional proposals to reform the Superfund process have appeared recently in studies by the MIT Center for Technology, Policy and Industrial Development (MIT, 1992) and by Putnam, Hayes, & Bartlett, Inc. (Butler and Jones, 1992) for the Coalition on Superfund. The MIT study recommends the development of special assessment and remediation procedures for sites with common characteristics, such as landfills, and earlier categorization of sites into immediate action versus no-action but monitor pathways. The Putnam, Hayes, & Bartlett study goes further, and specifically recommends setting priorities for individual actions across sites, rather than setting priorities for the sites themselves, and the identification of early actions that might be taken at sites to significantly reduce risk, even before the NPL decision is made (Butler and Jones, 1992). Both studies recommend that the HRS should be rescored as remediation actions are implemented and that, as warranted, the rescored sites should be deleted from the NPL. The Putnam, Hayes, & Bartlett study specifically recommends that predicted reductions in HRS scores associated with alternative remedial actions be used to assign priority to these actions (Butler and Jones, 1992). This use of HRS score differences to reflect risk reduction benefits runs counter to the committee's scientific assessment that the HRS scores can only be used to reflect ordinal differences in sites, and not cardinal or continuous differences in absolute risks.

EPA has recently responded to the recommendations discussed above as well as others by considering new approaches, such as the Superfund Accelerated Cleanup Model (Inside E.P.A., 1992). That plan would combine the site-screening and risk-assessment studies for the preremedial, removal, and remedial phases into a single study. This change would allow elimination of the distinction between EPA's early removal and long-term remedial programs and encourage expedited progress through the Superfund time frame. Whatever the outcome of this particular proposal, it is likely that further efforts will be made to compress the Superfund timeline. Those efforts, consistent with recommendations of

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

the OTA and others recommendations, suggest that more data could be available when the HRS scoring is performed. Again, the issue arises of whether a structured-value model such as HRS should be used, in contrast to a formal risk assessment, when these additional data are available. However this issue is resolved, current statutory requirements make it probable that the HRS will remain a key part of the EPA priority-setting process for NPL selection.

SUMMARY EVALUATION OF EPA PRIORITY SETTING FOR HAZARDOUS-WASTE SITES

To provide a summary evaluation of the EPA priority-setting process, the evaluation criteria identified in Chapter 2 are examined with the primary focus upon the current (revised) version of the HRS model.

General Issues in HRS Model Development and Application

Clearly Defined Purpose: The HRS model has a well-defined purpose within the Superfund process—site selection for the NPL—and a specified user population made up of those responsible for site scoring. The priority-setting processes for the later stages of Superfund are similarly well defined.

Credibilities and Acceptability: Although certain technical limitations to the HRS model have been identified by the committee and others, including particular aspects of the likelihood of release or exposure category for certain pathways, and questions on the handling of the toxicity component of the waste characteristic category, the committee finds that, in general, the HRS

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

model is (within the context of a structured-value model) generally consistent with accepted scientific understanding and knowledge of the environment. Extensive scientific peer review, public participation, and public comment have been included as part of the model development process.

Appropriate Logic and Implementation of Mathematics: The HRS model includes a combination of additive and multiplicative calculations to obtain pathway and total site scores from the individual factor scores. The calculation of pathway scores as the product of the contaminant release (or exposure), chemical characteristics, and target receptor category scores is patterned according to the multiplicative model for risk. However, the often ad hoc procedures for determining the factor and category scores, and the chemical and site factors that combine source, transport, and exposure-toxicity into single measures, make it difficult to interpret the resulting HRS score in any absolute sense. This problem, endemic to structured-value models, precludes the use of the FIRS for evaluating risk-reduction benefits obtained from a proposed or completed remedial activity.

Model Documentation: The documentation for the HRS model is generally adequate, though little guidance has been provided to ensure consistent sampling and collection of input data. This need is being addressed in the HRS guidance document being developed by EPA.

Model Validation: The HRS model has been compared in a number of studies with more detailed site assessments based on risk analysis or expert panels. The degree of correlation with these other estimates has generally been low to modest.

Model Sensitivity and Uncertainty Analysis: A number of studies have been conducted by EPA and others to evaluate the sensitivity of the model to various factors and factor categories. In practice, the scoring outcome is quite sensitive to the overall effort exerted in data collection at the site, with a potential for manipulation of this effort by interested parties.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

Specific HRS Technical Features

Applicability to All Waste Sites: The HRS model is broadly applicable to the range of hazardous-waste sites encountered.

Allowance for Dynamic Tracking: Proposals have been made to use the HRS model at various stages of the Superfund process to set priorities and track alternative remedial actions. While this could provide a useful administrative tool, such use is not consistent with the ordinal (non-absolute) nature of the HRS normalized score.

Discrimination between Immediate- and Long-Term Risk: The model is intended for long-term risk (greater than 20 years), because immediate threats have been addressed by EPA prior to the NPL-listing decision. The recent EPA plan to remove the sharp division between immediate response and long-term remediation, if implemented, would dictate the need to reintroduce immediate threats to the HRS model.

Inclusion of Cost Estimates of Remediation: The HRS model does not consider costs or timing issues associated with remediation. These issues are considered in the priority-setting procedures for later stages of Superfund.

Transparency: The original HRS model was relatively transparent, but the recent revisions have made the model significantly more difficult to understand. The effective weights for human health versus ecological impacts are difficult to assess.

User-Friendliness: The model is presented in a straightforward manner that should be relatively easy to follow for regular site scorers. However, the revised HRS is too complex for routine use by lay citizens. The committee is unaware of any interactive computer implementation of the HRS.

Appropriate Security: The hard-copy format for HRS Scoring and the quality-assurance checks provided by EPA and a contractor limit the potential for security problems.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×

The HRS model has been described as providing a consistent, expedient format for site-scoring and site-setting priorities consistent with legislative mandate (Wu and Hilger, 1984). However, some of the particular features of the model are subject to challenge, and the overall appropriateness of a structured-value approach is questioned, given the availability of risk- assessment procedures. This question might continue to be raised as more detailed site study and evaluation procedures are performed earlier in the Superfund process. To the extent that EPA remains committed to an early decision for inclusion on the NPL, the HRS model may remain the best alternative available. Modifications to allow for more detailed review of sites with intermediate scores could, however, help to reduce the number of false positive and false negative decisions.

Subsequent EPA Priority-Setting Process

The HRS represents a critical, first step for remediation priority setting: deciding which sites to place on the National Priority list. Sites that are placed on this list then are subject to subsequent priority setting to determine which ones to investigate first through the Remedial Action/Feasibility Study (RI/FS) process, following which sites are selected for remedial action (RA). The steps in these priority-setting processes are well defined and open to public comment and scrutiny, though the selections themselves are generally not open to outside review. The RA process is somewhat formal and involves a structured-value model with weighted consideration of the risk of contaminants, stability, population exposed, significant environments, and program management. The process is well defined, relatively simple, and transparent. It is consistent with an administrative program which is not mandated by law, but still needed to ensure effective management of the Superfund program.

Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
This page in the original is blank.
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 85
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 86
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 87
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 88
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 89
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 90
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 91
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 92
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 93
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 94
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 95
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 96
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 97
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 98
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 99
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 100
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 101
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 102
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 103
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 104
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 105
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 106
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 107
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 108
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 109
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 110
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 111
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 112
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 113
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 114
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 115
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 116
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 117
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 118
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 119
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 120
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 121
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 122
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 123
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 124
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 125
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 126
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 127
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 128
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 129
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 130
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 131
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 132
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 133
Suggested Citation:"4 EPA'S PRIORITY SETTING." National Research Council. 1994. Ranking Hazardous-Waste Sites for Remedial Action. Washington, DC: The National Academies Press. doi: 10.17226/4781.
×
Page 134
Next: 5 DOD'S PRIORITY SETTING »
Ranking Hazardous-Waste Sites for Remedial Action Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The United States may not be able to make all hazardous-waste sites as clean as possible. Therefore, priorities must be set for the timing of waste site remediations. This book assesses several of the government's methods of ranking sites for remediation and compares the performance of three such models using input data developed from the same set of waste sites. Because inconsistent methods may be neither effective nor prudent, the book recommends that the government consider developing a unified national approach to setting priorities to replace the current multiple approaches.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!