Hazard Ranking System Issue Analysis: Containment Rating Factor MITRE ------- Hazard Ranking System Issue Analysis: Containment Rating Factor Stuart A. Haus March 1987 MTR-86W199 SPONSOR: U.S. Environmental Protection Agency CONTRACT NO.: EPA-68-01-7054 The MITRE Corporation Civil Systems Division 7525 Colshire Drive McLean, Virginia 22102-3481 ------- irmw Department Approva MITRE Protect Approval: /< LsCS <>J ------- ABSTRACT This report addresses issues related to the modification of the current HRS containment factor. Advantages and disadvantages of basing the containment factor on the RCRA Subtitle C land disposal regulations are discussed. A number of alternative approaches for modifying the containment factor so as to base it on evaluation criteria other than the RCRA Subtitle C regulations are identified and briefly examined. Two of these alternatives are recommended for possible further development. These two alternatives are: • The integration of evidentiary and predictive criteria in the containment factor. • The integration of a physical state factor into the containment factor. Suggested Keywords: Superfund, Hazardous waste, Hazard ranking, Containment. iii ------- ACKNOWLEDGEMENT The author wishes to acknowledge Greg Vogel for his valuable contribution in the development and analysis of the options presented in this report. iv ------- TABLE OF CONTENTS LIST OF TABLES 1.0 INTRODUCTION 1.1 Background 1 1.2 Issue Description 3 1.3 Scope and Approach 5 1.4 Organization of the Report 6 2.0 REVIEW OF CURRENT HRS CONTAINMENT EVALUATION METHODOLOGY 9 2.1 Overview of HRS Containment Factor 9 2.2 Analysis of Current HRS Containment Factor 14 2.3 Issues Relevant to the Revision of the HRS Containment 19 Factor 2.3.1 Screening Out Sites 20 2.3.2 Differentiation of Sites 22 2.4 Summary of Containment Factors in Other Ranking Systems 23 3.0 CONTAINMENT EVALUATION ALTERNATIVES 27 3.1 Overview of Alternatives 27 3.2 Updating of Present RCRA-Based Criteria 30 3.3 Integration of Containment Factor With a Factor Based 32 on Waste Quantity 3.4 Development of Evidentiary Criteria With a Zero Value 33 3.5 Development of Evidentiary Criteria Without a Zero Value 36 3.6 Integration of Evidentiary and Predictive Criteria 37 3.7 Development of Time-Dependent Criteria 38 3.8 Use of Criteria Based on Waste Disposal Location 41 3.9 Development of Criteria Based on Site Drainage 44 3.10 Integration of Containment and Physical State Factors 46 3.10.1 Use of Evidentiary Criteria 48 3.10.2 Use of Predictive Criteria 49 4.0 SUMMARY AND RECOMMENDATIONS 51 APPENDIX A - REVIEW OF CONTAINMENT FACTORS IN OTHER SITE 59 RANKING SYSTEMS APPENDIX B - BIBLIOGRAPHY 87 ------- LIST OF TABLES Table Number Page 2-1 HRS Containment Factor for Ground Water Route 12 2-2 HRS Containment Factor for Surface Water Route 13 2-3 Distribution of HRS Containment Factor Values 16 for Facilities in the Automated NPL Technical Data Base 3-1 Illustrative Containment Factor Values Based 43 on Waste Disposal Location vii ------- 1.0 INTRODUCTION 1.1 Background The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA) (PL 96-510) requires the President to identify national priorities for remedial action among releases or threatened releases of hazardous substances. These releases are to be identified based on criteria promulgated in the National Contingency Plan (NCP). On July 16, 1982, the Environmental Protection Agency (EPA) promulgated the Hazard Ranking System (HRS) as Appendix A to the NCP (40 CFR 300; 47 FR 31180). The HRS comprises the criteria required under CERCLA and is used by EPA to estimate the relative potential hazard posed by releases or threatened releases of hazardous substances. The HRS is a means for applying uniform technical judgment regarding the potential hazards presented by a release relative to other releases. The HRS is used in identifying releases as national priorities for further investigation and possible remedial action by assigning numerical values (according to prescribed guidelines) to factors that characterize the potential of any given release to cause harm. The values are manipulated mathematically to yield a single score that is designed to indicate the potential hazard posed by each release relative to other releases. This score is one of the criteria used by EPA in determining whether the release should be placed on the National Priorities List (NPL). ------- During the original NCP rulemaking process and the subsequent application of the HRS to specific releases, a number of technical issues have been raised regarding the HRS. These issues concern the desire for modifications to the HRS to further improve its capability to estimate the relative potential hazard of releases. The issues include: • Review of other existing ranking systems suitable for ranking hazardous waste sites for the NPL. • Feasibility of considering ground water flow direction and distance, as well as defining "aquifer of concern," in determining potentially affected targets. • Development of a human food chain exposure evaluation methodology. • Development of a potential for air release factor category in the HRS air pathway. • Review of the adequacy of the target distance specified in the air pathway. • Feasibility of considering the accumulation of hazardous substances in indoor environments. • Feasibility of developing factors to account for environmental attenuation of hazardous substances in ground and surface water. • Feasibility of developing a more discriminating toxicity factor. • Refinement of the definition of "significance" as it relates to observed releases. • Suitability of the current HRS default value for an unknown waste quantity. • Feasibility of determining and using hazardous substance concentration data. ------- • Feasibility of evaluating waste quantity on a hazardous constituent basis. • Review of the adequacy of the target distance specified in the surface water pathway. • Development of a sensitive environment evaluation methodology. • Feasibility of revising the containment factors to increase discrimination among facilities. • Review of the potential for future changes in laboratory detection limits to affect the types of sites considered for the NPL. Each technical issue is the subject of one or more separate but related reports. These reports, although providing background, analysis, conclusions, and recommendations regarding the technical issue, may not directly affect the HRS. Rather, these reports will be used by an EPA working group that will assess and integrate the results and prepare recommendations to EPA management regarding future changes to the HRS. Any changes will then be proposed in Federal Register notice and comment rulemaking as formal changes to the NCP. The following section describes the specific issue that is the subject of this report. 1.2 Issue Description The HRS containment factor is currently a measure of the methods (either engineered or natural) that have been employed to minimize or prevent the migration of hazardous substances to ground water or surface water. Examples of engineered containment methods include landfill liners, leachate collection systems, diversion ------- structures, and diking. Examples of natural containment methods include locating waste management sites in areas where the terrain precludes the overland migration* of hazardous substances to surface water or in areas where there is no ground water or surface water in the vicinity of the site. The criteria in the current HRS containment factor are based upon those land disposal regulations developed, as of early 1982, under Subtitle C of the Resource Conservation and Recovery Act (RCRA). Since that time, there have been significant changes in the RCRA Subtitle C land disposal regulations, and further modifications are currently being developed in response to requirements mandated by the Hazardous and Solid Waste Amendments of 1984 (HSWA). Consequently, concerns have been raised within EPA that the HRS containment factor should be modified to be more consistent with the current RCRA Subtitle C land disposal regulations. Public commenters have not raised technical issues about the HRS containment factor in either NCP or NPL rulemakings. Those comments that have been received from the public have been related to site-specific applications of the containment factor (e.g., whether a specific site should have been assigned a containment factor value of 2 or 3). *For reasons beyond the scope of this discussion, the HRS does not assess the potential for hazardous substances to be released to surface water via the ground water. Consequently, containment methods relating to ground water discharges to surface water are not included in this discussion. ------- 1-3 Scope and Approach In the absence of public comments on the MRS containment factor, the original objective of this effort was the development of options for modifying the MRS containment factor to make it more consistent with the current RCRA Subtitle C land disposal regulations. This objective did not include the examination and development of additional alternative technical approaches for evaluating containment in the HRS. However, the analysis presented in Chapter 2 indicates that few, if any, sites currently eligible for the NPL are likely to have employed containment measures that even come remotely close to meeting the requirements of the 1982 RCRA Subtitle C regulations, let alone those currently in effect. For this reason, as well as for other reasons discussed in Chapter 2, modification of the HRS containment factor to ensure more consistency with the current RCRA regulations is considered to be primarily a policy issue rather than a technical issue. Consequently, the primary focus of this paper has been redirected toward a discussion of the issues related to any modification of the containment factor, rather than the development of specific options for modifying the containment factor. Advantages and disadvantages of basing the containment factor on the RCRA land disposal regulations are discussed. A number of alternative approaches for modifying the containment factor so as to base it on evaluation criteria other than the RCRA Subtitle C regulations are ------- identified. The advantages and disadvantages of each approach are discussed. One option for the development of a containment factor under each of the alternative approaches is briefly outlined to illustrate the types of options possible under each of these alternative approaches. The full development of options under any of the alternative approaches will, however, require the identification, by EPA, of the preferred options and the resolution of specific issues (both policy and technical) discussed under each alternative. For example, decisions on whether data will be collected for determining environmental concentrations of contaminants greatly influence the types of MRS containment evaluation criteria that can be developed (see discussion in Chapter 3). The discussion in this paper is limited to the development of containment options for the current HRS surface water and ground water pathways because the current HRS has no air pathway containment factor. Options for containment factors for the air pathway are discussed in a companion paper that addresses the development of a potential for air release factor category (Wolfinger, 1986). 1.4 Organization of the Report The current HRS containment evaluation criteria are presented and analyzed in Chapter 2. Issues associated with any modifications to the current HRS containment evaluation criteria (including ------- modifications to ensure more consistency with the current RCRA Subtitle C land disposal regulations) are discussed. Several possible objectives to be considered in the development of alternatives to the containment factor are identified. Important similarities and differences between the current HRS containment factor and the containment factors in ten other site ranking systems and three EPA hazardous waste policy analysis models are also summarized in Chapter 2. (Appendix A presents a detailed review of the use of containment in these other hazardous waste site ranking systems and the EPA policy analysis models.) Chapter 3 outlines nine options for modifying the containment factor to meet the various objectives discussed in Chapter 2. Advantages and disadvantages of each option are described. Chapter 4 presents a summary and recommendations. Appendix B contains the bibliography. ------- 2.0 REVIEW OF CURRENT HRS CONTAINMENT EVALUATION METHODOLOGY This chapter presents a summary of the current HRS containment evaluation methodology. This is followed by an analysis of the current containment factor. The analysis identifies several issues associated with the containment factor that need to be considered in any modification of the containment factor. Other issues generic to modification of the containment factor are then identified and discussed. This chapter concludes with a summary of the important similarities and differences between the current HRS containment factor and the containment factors incorporated in ten other systems used to rank hazardous waste sites and in three EPA hazardous waste policy analysis models. A more detailed review of these systems is provided in Appendix A. 2.1 Overview of HRS Containment Factor The HRS migration score currently reflects the potential threat to humans or the environment from the migration of a hazardous substance away from a site by three possible routes—ground water, surface water, and air. A migration score for each applicable route is calculated by rating the site with respect to a number of factors that characterize a) the potential for hazardous substances to migrate from the site by that route, b) the hazardous substances present at the site, and c) the presence and proximity of targets (e.g., human populations). The potential for migration along a route may be evaluated in one of two ways—either through direct evidence of a release (i.e., ------- an observed release) or through the use of factors that evaluate the potential for a release in terms of route characteristics and containment. If there is a documented observed release for any route, a rating value of 45 is assigned for that route and the route characteristics and containment are not evaluated. If there is no evidence of an observed release, the observed release is assigned a value of zero. In this case, route characteristics and containment are evaluated for the ground water or surface water routes, but not for the air route. The route characteristics score and the containment score are multiplied to assign a rating to the potential for a release; their product has a maximum value of 45, the same as the value for an observed release. The migration score for a route is then determined by multiplying either the observed release score or the potential for a release score (for ground water or surface water) by scores for waste characteristics and targets and normalizing the product so that it ranges from 0 to 100. The individual route scores are combined to obtain the overall migration score for the site. The components of the route characteristics rating category for the ground water route are the depth to the aquifer of concern, the net precipitation, the permeability of the unsaturated zone, and the physical state of the waste. The components of the route characteristics rating category for the surface water route are the facility slope and intervening terrain, 24-hour rainfall, the 10 ------- distance to the nearest surface water, and the physical state of the waste. For both routes, the values assigned to the component factors of the route characteristics category are weighted and then added prior to being multiplied by the containment factor value. The HRS containment factor is currently a measure of the methods (either natural or engineered) that have been used to minimize or prevent the migration of hazardous substances to ground water or surface water. The containment factor is assigned a value for either route by applying criteria that are specific both to the route and to the type of waste management method being evaluated for that route. The specific criteria for rating containment are presented in Table 2-1 for the ground water route and in Table 2-2 for the surface water route. The containment factor value assigned to a route is the highest containment value (i.e., value for the least effective containment measure) assigned to any of the containment measures applicable to that route. The evaluation criteria in Tables 2-1 and 2-2 are based on the technical requirements of the RCRA Subtitle C land disposal regulations (40 CFR 264 and 265) promulgated or proposed prior to mid-1982. Wastes whose containment meets the specified RCRA technical requirements are considered to pose a lesser threat to human health and the environment than wastes whose containment does not meet the specified requirements and are currently assigned the lowest relative containment factor value of zero. A containment factor value of zero for a route, in the 11 ------- TABLE 2-1 HRS CONTAINMENT FACTOR FOR GROUND WATER ROUTE Assign containment a value of 0 if: (1) all the hazardous substances at the facility are underlain by an essentially non-permeable surface (natural or artificial) and adequate leachate collection systems and diversion systems are present; or (2) there la no ground water in the vicinity. The value "0" does not Indicate no risk. Rather, it indicates a significantly lower relative risk when compared with more serious sites on a national level. Otherwise, evaluate the containment for each of the different means of storage or disposal at the facility using the following guidance. Assigned Value •• Surface Impoundment Sound run-on diversion structure, essentially non-permeable liner (natural or artificial) compatible with the 0 waste, and adequate leachate collection system Essentially non-permeable compatible liner with no leachate collection system; or inadequate freeboard 1 Potentially unsound run-on diversion structure; or moderately permeable compatible liner 2 Unsound run-on diversion structure; no liner; or incompatible liner 3 B. Containers Containers sealed and In sound condition, adequate liner, and adequate leachate collection system " Containers sealed and In sound condition, no liner or moderately permeable liner 1 Containers leaking, moderately permeable liner 2 Containers leaking and no liner or Incompatible liner 3 C. Piles Piles uncovered and waste stabilized; or piles covered, waste uns tabilized , and essentially non-permeable liner 0 Pilea uncovered, waste unstabllced, moderately permeable liner, and leachate collection system 1 Piles uncovered, waste uns tabilized , moderately permeable liner, and no leachate collection system 2 Piles uncovered, waste unstabillzed, and no liner 3 D. Landfill Essentially non-permeable liner, liner compatible with waste, and adequate leachate collection system 0 Essentially non-permeable compatible liner, no leachate collection system, and landfill surface precludes ponding 1 Moderately permeable, compatible liner, and landfill surface precludes ponding 2 No liner or incompatible liner; moderately permeable compatible liner; landfill surface encourages ponding; no 3 run-on control Source: 47 PR 31180, July 1982. ------- TABLE 2-2 HRS CONTAINMENT FACTOR FOR SURFACE WATER ROUTE A»lgn containment a value of 0 If: (1) all the waste at the site Is surrounded by diversion structures that are In sound condition and adequate to contain all runoff, spills, or leaks fron the waste; or (2) Intervening terrain precludes runoff fron entering surface water. Otherwise, evaluate the containment for each of the different Beans of storage or disposal at the site and assign a value as follows: Assigned Value A. Surface Impoundment Sound diking or diversion structure, adequate freeboard, and no erosion evident 0 Sound diking or diversion structure. Inadequate freeboard 1 Diking not leaking, but potentially unsound 2 Diking unsound, leaking, or In danger of collapse 3 B. Containers Containers sealed, In sound condition, and surrounded by sound diversion or containment system 0 !_. Containers sealed and In sound condition, but not surrounded by sound diversion or containment system 1 U> Containers leaking and diversion or containment structures potentially unsound 2 Containers leaking, and no diversion or containment structures or diversion structures leaking or In danger of 3 collapse C. Haste Piles Files are covered and surrounded by sound diversion or containment system 0 Piles covered, waste unconsolldated, diversion or containment system not adequate 1 Piles not covered, waste unconsolldated, and diversion or containment system potentially unsound 2 Piles not covered, waste unconsolldated, and no diversion or containment structures or diversion system 3 leaking or in danger or collapse D. Landfill Landfill slope precludes runoff, landfill surrounded by sound diversion system, or landfill has adequate cover 0 material Landfill not adequately covered and diversion system sound 1 Landfill not covered and diversion system potentially unsound 2 Landfill not covered and no diversion system present, or diversion system unsound 3 Sourcei 47 PR 31180, July 16, 1982. ------- absence of an observed release, results in the route receiving a migration score of zero because the other rating factor values are multiplied by the containment factor to obtain the migration score. Because of this multiplicative property, the containment factor is one of the two most influential rating factors in the HRS (along with observed release) for the ground water and surface water route scores. Accordingly, very specific criteria have been established for assigning the containment factor value. For example, a surface impoundment having a sound run-on diversion structure, an adequate leachate collection system, and an essentially non-permeable liner that is compatible with the waste in the impoundment is assigned a zero value for the containment factor for the ground water route. A zero value can also be assigned for the ground water route if there is no ground water in the vicinity. If no waste containment is attempted, the containment factor is assigned a value of three. Containment factor values of 1 or 2 are assigned for various intermediate levels of containment. 2.2 Analysis of Current HRS Containment Factor This section presents an analysis of the current HRS containment factor. The analysis identifies several issues that need to be considered in any subsequent modification of the containment factor. Although data are not available to determine the actual distribution of containment factor values that have been assigned to all wastes sites ranked using the HRS, it is apparent from data for 14 ------- sites in the automated NPL technical data base that there is little variation in the containment values assigned among these sites. Table 2-3 presents the distribution of the containment factor values (for ground water and surface water migration routes) that have been assigned to the approximately 600 sites in the automated NPL technical data base that do not have an observed release and for which a containment value has been reported. Approximately 99 percent of the relevant NPL sites, 93 percent of the relevant "other sites," and 96 percent of all such sites in the automated NPL technical data base have been assigned the maximum containment factor value for the ground water pathway. For the surface water pathway, these numbers are 87, 80, and 84 percent, respectively. Not one site has been assigned a containment factor value of zero for either pathway. This is not especially surprising since sites with a containment factor value of zero would have a route score of zero; only sites with migration scores of 25 or higher are required to be forwarded to EPA Headquarters for review. The data in Table 2-3 illustrate that the current HRS containment factor provides little discrimination, based on containment practices, among those sites submitted to EPA Headquarters for quality assurance review. For these sites, the primary function of the containment factor appears to be the screening out of those few sites that have relatively high levels of containment. For example, about 65 to 85 percent of sites in the 15 ------- TABLE 2-3 DISTRIBUTION OF HRS CONTAINMENT FACTOR VALUES FOR FACILITIES IN THE AUTOMATED NPL TECHNICAL DATA BASE* Number of NPL Facilities** Containment Value 0 1 2 3 Total Ground Water No. % Surface Water No. % 0 0 1 2 2 1 3 216 Total 219 0 1 1 99 100 0 19 16 252 287 0 7 6 87 100 Number of Other Facilities*** Ground Water No. % 0 0 10 3 12 4 300 93 322 100 Surface Water No. % 0 0 35 11 29 9 256 80 320 100 Total Number of Facilities**** Ground Water No. % 0 00 1 12 2 2 13 2 3 516 96 Total 541 100 Surface Water No. % 0 0 54 9 45 7 508 84 607 100 *This distribution applies to facilities in the automated NPL technical base that do not have an observed release and for which a containment value has been reported. **Based on the 888 facilities listed on or proposed for NPL as of August 1986. ***Based on facilities in the automated NPL technical data base that have not been proposed for or listed on the NPL as of August 1986; containment values for these facilities have not all been subjected to a quality assurance review. ****Summation of the NPL and other facilities identified above. 16 ------- automated NPL technical data base assigned a containment value of 1 for any route are not on the NPL.* No conclusions can be drawn from the data in Table 2-3 about how containment factor values affect the overall ability of the current HRS to discriminate between NPL and non-NPL sites. Low scoring sites (i.e., those with migration scores less than 25) are not generally submitted to EPA Headquarters for quality assurance review. It is possible that low containment factor values are much more prevalent among these sites and do, in fact, assist in discriminating such non-NPL sites from NPL sites. Data to prove or disprove this are not available. One observation possible from the data in Table 2-3 is that wastes at many CERCLA sites are not well contained** and that the use of technical criteria based on the RCRA Subtitle C land disposal regulations may not provide much differentiation among containment practices at these sites. For most sites eligible for the NPL, this is even more likely to be true if the HRS containment factor is modified to reflect the current RCRA regulations which are considerably more stringent than those in effect during early 1982. Another consideration with regard to the present containment evaluation criteria is that the criteria may lead to the assignment *Some of these sites are still under quality assurance review for placement on the NPL and may ultimately be placed on the NPL. **A containment factor value of 3, which most sites appear to receive, indicates that there is essentially no containment provided at the sites for at least some of the wastes present at the site. 17 ------- of containment ratings values that are based, to a large extent, on subjective judgments. The evaluation criteria currently require a determination of whether the relevant containment technologies (e.g., liners) exist at a site and, if so, an assessment of the effectiveness of each containment technology. For example, the HRS user needs to differentiate between sound, potentially unsound, and unsound diversion structures. Similarly, assessments are required as to whether liners are essentially non-permeable or moderately permeable and whether they are compatible or incompatible with wastes. Hie adequacy of covers and leachate collection systems must also be assessed. In light of the type of containment likely to be present at CERCLA sites, it may be more meaningful for HRS screening purposes just to distinguish whether a containment technology is present at a site and not to try to make fine distinctions about how effective it is. For example, with regard to liners, the containment could be ranked on whether a liner is present, not on whether it is permeable or compatible with wastes. Data to make such distinctions are not generally available from current site inspections. The above analysis suggests a need for examining whether alternatives exist for modifying the current containment factor that: • Meaningfully provide greater differentiation among sites by including evaluation criteria other than RCRA-based technical requirements and/or • Reduce the subjectivity of the containment evaluation criteria 18 ------- 2.3 Issues Relevant to the Revision of the HRS Containment Factor One issue that needs to be considered is whether the containment factor should continue to be included in the HRS. The containment factor currently appears to be effective in screening out sites that have relatively high levels of containment. However, the number of such sites appears to be relatively small, at least based on data for the sites in the automated NPL technical data base. Of the sites in the automated NPL technical data base that have containment values assigned, only 4 percent do not receive the maximum factor value for the ground water route and about 16 percent do not receive the maximum factor value for the surface water route. Chapter 3 examines several alternative approaches for evaluating containment in the HRS. A further alternative is to replace or integrate the current containment factor with another factor that addresses the retardation (e.g., geochemical removal) of containments in the subsurface environment and in surface water. The development of options for considering retardation in the HRS is currently being examined in two companion studies (Sayala, 1986; Wang, 1986). Once the development of these options is completed, it will be possible to assess how they could most effectively be used in the HRS. Assuming that containment is to remain a component of the HRS and that the HRS is to remain a screening tool, there are several other important issues that are generic to any modification of the current HRS containment factor. These issues are reviewed in this section. 19 ------- The first issue that needs to be resolved is the specification of the primary function that a containment factor is to serve in the HRS. There are two different functions that a containment factor can provide. It can be used primarily to screen out (assign low rating values to) those relatively few sites that have adequate containment* or it can be used primarily to provide differentiation among all sites without an observed release, through modifications designed to provide a more uniform distribution of containment rating values. Based on the data presented above, it is unlikely that any single containment factor (especially one based on technical requirements for containment structure design)** can be developed to do both effectively. For those sites in the automated NPL technical data base, the current HRS containment factor appears to be serving the former function rather than the latter function. Data are not available to determine which function it actually serves for all sites ranked with the HRS; however, for reasons discussed above, it is likely to serve the former function. 2.3.1 Screening Out Sites Because containment is currently one of only two multiplicative factors in the HRS (the other is observed release), it can be extremely effective as a factor for screening out sites that exhibit *Issues related to the definition of adequate containment are discussed below. **Technical requirements for containment structure design include such requirements as covers, liners, and leachate collection systems rather than the specification of performance objectives for the containment structure. 20 ------- specified levels of containment. If it is decided to use the containment factor primarily to screen out those sites that have adequate containment, the basic structure of the containment factor would likely be: • No containment—assign maximum value • Limited containment*—assign intermediate value • Adequate containment—assign minimum value The main issue that would need to be resolved is how the concept of adequate containment is to be defined and evaluated for HRS purposes. Ideally, it would be defined such that, at sites meeting the definition, there would be little or no potential for hazardous substances to migrate from the site via the route being evaluated. This would tend to argue for basing the definition of adequate containment on RCRA technical requirements since they provide, for screening purposes, the clearest definition of adequate containment. However, to be effective, the definition must also be consistent with the site characteristics that would actually be encountered in the universe of sites being evaluated for the NPL. If no sites meet the definition of adequate containment, then there is likely to be little or no benefit, for ranking purposes, to such a definition. Chapter 3 identifies and examines several possible alternatives that *There could be one or more levels of limited containment defined, depending on how fine a distinction can realistically be developed from available data. 21 ------- could be further developed for screening out sites, if this is the desired approach. A further issue that needs to be considered is whether or not the minimum value to be assigned for adequate containment should be zero. With a multiplicative containment factor, a zero value would result in a zero score for that migration route. A non-zero (but relatively low value) would result in a low score for the migration route and would reflect the fact that almost no containment is likely to be 100 percent effective over a sufficiently long period of time. This issue is primarily a policy issue, not a technical issue. 2.3.2 Differentiation of Sites If, on the other hand, a primary objective is to provide greater differentiation within the types of containment practices likely to be present at most sites eligible for the NPL, the issues are more technical in nature. (One reason for such an objective might be to provide some greater differentiation among those sites whose scores are high enough for placement on the NPL.) The primary issue in this case is the definition of containment characteristics that meaningfully delineate actual, but likely small, differences in containment effectiveness and that at the same time are representative of the containment technologies likely to be present at most sites eligible for NPL listing. Based on Tables 2-1 through 2-3, most sites in the automated NPL technical data base have no containment for at least a portion of the wastes present at the site. Thus, it is unlikely that 22 ------- any useful definition could be based solely on technical requirements for land disposal. Chapter 3 identifies and examines several possible alternatives for non-RCRA based criteria that could be further developed for differentiating among sites, if this is the desired approach. 2.4 Summary of Containment Factors in Other Ranking Systems In addition to the above analysis, ten ranking systems that consider containment in rating the threat posed by hazardous waste sites have been identified along with three EPA hazardous waste policy analysis models that also account for waste containment. The containment factors in these systems have been reviewed. The review focuses on how containment is used in the various systems and how containment effectiveness is defined and evaluated. Important similarities and differences between these factors and the HRS containment factor have been identified. The findings of this review are presented in Appendix A and are summarized below. The ten ranking systems reviewed are: • HARM • S.P.A.C.E. for Health • HARM II • PERCO • GSR • Illinois Rating Scheme • ADL • Rating Methodology Model • SAS • Dames and Moore Methodology The three EPA hazardous waste policy analysis models reviewed are: • Liner Location Risk and Cost Analysis Model • Hazardous Waste Tank Failure Model • RCRA Risk-Cost Analysis Model 23 ------- Of the 10 ranking systems reviewed, one (S.P.A.C.E. for Health) has a containment factor that is identical to the HRS containment factor and four (HARM, HARM II, CSR, and ADL) have multiplicative containment factors that are very similar to the HRS containment factor in concept and application. There are two major differences between these latter four systems and the HRS. Three of these systems (HARM, HARM II, and CSR) have a non-zero lower limit for the containment score to indicate that no containment is likely to be 100 percent effective. Three of the systems (HARM, CSR, ADL) also employ fewer intermediate containment levels (either one or none) than the HRS to avoid making fine distinctions about the degree of containment present at a site. Two other ranking systems (SAS and PERCO) do not explicitly use containment in rating a site. SAS uses containment only as a means of evaluating waste quantity. PERCO uses containment only for identifying sites similar to the site being evaluated. One other ranking system (Illinois Rating System) considers operational history, rather than containment effectiveness, in the ranking of sites. This approach does not appear practical to apply to CERCLA sites where information about the operational history of the site is extremely limited and often unavailable. The remaining two ranking systems (Rating Methodology Model and Dames and Moore Methodology) have containment factors that are additive rather than multiplicative and that are rated on the same scale (0 to 3) as the HRS containment factor. These factors were considered 24 ------- in the original development of the HRS and were judged not to be comprehensive, nor well defined. Consequently, we have concluded that these other ranking systems do not provide any further basis for developing modifications to the HRS containment factor. The three EPA hazardous waste policy analysis models address containment in a very different manner than the HRS. Two of the models (the Liner Location Risk and Cost Analysis Model and the Hazardous Waste Tank Failure Model) use fault tree analysis and Monte Carlo simulation (see Appendix A) to estimate both the probability and timing of containment failure events that lead to releases of hazardous substances and the release volumes associated with the failure events. The failure and release components of the two models are currently developed for application to a limited number of facility designs (e.g., several specified containment configurations) and operating conditions, most of which are not likely to be representative of CERCLA sites. The third model (RCRA Risk-Cost Analysis Model) is a more deterministic model than the other two with regard to the consideration of containment failure. In this model, all synthetic liners are assumed to fail within 35 years and to have a linear failure rate over this 35-year period; leachate collection systems are assumed to release a fixed fraction of the leachate they handle; and surface releases from storm-water management units and surface impoundments are 25 ------- assumed to occur solely from overtopping during storm events, and overtopping is assumed to have a constant probability of occurrence in any year. As currently structured, these three models are intended primarily for use in analyzing the risks and costs associated with alternative regulatory strategies (e.g., alternative standards for waste containment). They are not meant for, nor are they applicable to, site-specific comparisons of containment effectiveness among different facilities, especially among abandoned, uncontrolled waste disposal sites. Consequently, we have concluded that the models themselves do not provide any further basis for developing modifications to the HRS containment factor. However, some results from regulatory analyses conducted with these models may be useful in developing portions of various containment rating scales to be incorporated in the HRS containment rating factor. This will depend, in part, on the nature of the containment rating factor alternatives that are developed. (Chapter 3 identifies various alternative approaches for evaluating containment in the HRS.) Limitations in the release pathways and facility types and designs accounted for in the three models will, however, limit any such application. 26 ------- 3.0 CONTAINMENT EVALUATION ALTERNATIVES Nine alternative approaches for evaluating containment in the HRS are identified and examined in this chapter. These alternative approaches have been developed to collectively illustrate ways that the current HRS containment factor could be modified to: • Increase its consistency with RCRA Subtitle C land disposal regulation • Include criteria other than RCRA-based technical requirements • Provide greater differentiation among sites, or • Reduce the subjectivity of the current containment evaluation criteria No one alternative approach does all four. However, depending on data availability, several of the alternative approaches could be combined to encompass various objectives. One option for the development of a containment factor under each alternative approach is briefly outlined to illustrate the types of containment options possible under each alternative. The advantages and disadvantages of each alternative are discussed, and issues that need to be resolved before more complete options can be developed under the alternative are identified. 3.1 Overview of Alternatives The nine alternatives are as follows: 1. Modify the present criteria to incorporate changes in the RCRA technical requirements since 1982 (Section 3.2). 2. Integrate the containment rating factor with a rating factor based on waste quantity (Section 3.3). 27 ------- 3. Revise the evaluation criteria to be evidentiary rather than predictive, and retain the zero factor value (Section 3.4). 4. Revise the evaluation criteria to be evidentiary rather than predictive, and eliminate the zero factor value (Section 3.5). 5. Integrate evidentiary and predictive criteria (Section 3.6). 6. Develop time-dependent evaluation criteria to replace the present criteria (Section 3.7). 7. Develop evaluation criteria based on waste disposal location rather than on waste management technology (Section 3.8). 8. Develop evaluation criteria based on site drainage to replace the present criteria (Section 3.9). 9. Integrate the containment rating factor with the physical state rating factor (Section 3.10). The first alternative provides an illustration of the criteria that would be developed if the containment factor was modified to be more consistent with current RCRA Subtitle C land disposal regulations. Two other alternatives examine methods for providing a more uniform distribution of containment values. One of these alternatives (Alternative 9) considers the integration of the current HRS physical state factor into the containment factor. As discussed in Section 3.10, most sites in the automated NPL technical data base have been assigned values of 3 for both of these factors. This approach seeks to provide a more uniform distribution of assigned values by integrating these factors into a revised 28 ------- containment factor. The other alternative (Alternative 2) considers the integration of the current containment factor with a factor based on the quantity of waste present in the various waste management areas at a site. The current HRS containment value assigned to a route is the highest rating value assigned to any waste management area for that route. This alternative attempts to define a more representative containment value by weighting the containment value for each waste management area by the portion of the waste at the site that is present in that waste management unit. Two additional alternatives (Alternatives 3 and 4) examine the use of evidentiary criteria, instead of RCRA-based criteria, in evaluating containment. The criteria in the present HRS containment factor, which are based on RCRA requirements, are predictive, equating the application of good containment technology to the existence of good containment. These two alternatives consider criteria based on evidence of containment efficiency at a site rather than on the predictive technical requirements. Such evidentiary criteria would equate the visual or analytical evidence of hazardous substances in contact with the environment with the existence of poor containment. Good containment would also need to be proven through analytical evidence. A third alternative (Alternative 5) then considers the integration of evidentiary and predictive criteria in evaluating containment. 29 ------- Another alternative (Alternative 6) addresses the use of time-dependent criteria, rather than RCRA-based criteria, in evaluating containment. This alternative examines whether containment values could be assigned based on the time elapsed since waste deposition, recognizing that no containment is likely to be 100 percent effective over time. Two other alternatives examine ways to reduce the subjectivity in the current containment evaluation criteria and to make the criteria more applicable to the differences in containment likely to be present at CERCLA sites. One alternative (Alternative 7) examines the use of containment factors based on waste location (i.e., above ground or below ground) rather than on the type of waste management unit in which the wastes are placed. The second alternative (Alternative 8) equates good containment with good site drainage and bases the evaluation criteria on site drainage. 3.2 Updating of Present RCRA-Based Criteria HSWA imposes additional technical requirements for RCRA Subtitle C hazardous waste management facilities, particularly land disposal units. The containment evaluation methodology could be revised to incorporate these additional requirements, as well as additional requirements promulgated under RCRA in the period after 1981. For example, HSWA requires double liners, leachate collection systems, and ground water monitoring at surface Impoundments, disposal piles, and landfills, although waivers are allowed. The 30 ------- revised evaluation criteria applicable to surface impoundments for the ground water route might be the following: Rating* Criteria 0 = Sound run-on diversion structure, non-permeable and compatible double liners, functioning leachate collection and ground water monitoring systems, sound diking with cover, adequate freeboard, weekly inspection. 1 = Sound run-on diversion structure, non-permeable and compatible liner, no leachate collection or ground water monitoring systems, sound diking, adequate freeboard. 2 = Potentially unsound run-on diversion structure or diking, liner installed but integrity unknown, or inadequate freeboard. 3 = Unsound or no run-on diversion structure, no liner, or unsound diking. Similar criteria could be developed for landfills, waste piles, tanks and containers under both the ground water and surface water routes. These revisions would have similar characteristics to the present methodology. Specifically, they may not result in differentiation among sites considered for placement on the NPL. Hie major advantage of using these criteria is that for screening purposes they do provide a clear definition of adequate containment, and sites that are well engineered to contain the wastes would likely be assigned very low scores. *Note the rating values indicated here and throughout the remainder of Chapter 3 are meant only to illustrate relative rankings of containment. They are not specifically being recommended for use in the HRS. 31 ------- 3.3 Integration of Containment Factor With a Factor Based on Waste Quantity In the current HRS each containment area is assigned a containment factor value for each of the two water routes. The containment value for each route is the highest rating value assigned to any waste management area for that route. A more representative measure of containment at the site may be obtainable by weighting the containment factor rating value for each waste management area by the portion of the hazardous waste at the site that is present in that waste management area. This weighted value could then be used to assign the overall containment rating value to the route. In addition to providing a more representative containment value, this approach would also provide increased differentiation among sites that have some degree of containment for at least some of the wastes at the site. However, this alternative does not appear to be practical to implement unless current data acquisition problems can be overcome. Presently, hazardous waste quantity estimates cannot be derived for about 20 percent of the sites in the automated NPL technical data base (Kushner, 1986). For many of the remaining sites, waste quantity estimates can be derived only for a fraction of the hazardous wastes deposited at the site. Unless more complete estimates of hazardous waste quantities can be obtained for CERCLA sites, this is not a viable alternative. To derive more complete estimates of hazardous waste quantity, the current site inspection 32 ------- program would have to be expanded considerably. (A companion report [Wolfinger, 1986] discusses issues involved in the collection of waste concentration data which could be used to estimate waste quantity at CERCLA sites.) Because of the data limitations, this alternative is not developed further in this report. 3.4 Development of Evidentiary Criteria With a Zero Value The present containment evaluation criteria are based on the presumption that if a waste is contained using technology specified by the RCRA regulations, it poses a lesser threat to human health and the environment than waste not contained using such technology; this level of containment is currently assigned a value of zero. Reductions in the containment technologies present at a site or in their operating efficiencies result in higher containment factor values being assigned. Several subjective evaluations are presently required to assign a containment factor value, such as liner permeability, diversion structure soundness, and container condition. The information required to make these evaluations may be difficult to obtain. For example, the presence of a liner may be established through design drawings or site inspections, but the condition or permeability may be difficult to determine without laboratory testing or on-site monitoring that disturbs the site. The approach under this alternative is to base the containment factor value primarily on criteria that can be evaluated through observation and limited site sampling and analysis, such as the following: 33 ------- Rating Criteria 0 = Demonstrated compliance with current RCRA waste containment requirements, such as impermeable and compatible double liners, covers, functioning diversion and leachate collection systems, and non-leaching waste stabilization. 1 = Evidence of waste containment measures, efficiency unknown, no evidence of uncontained waste. 2 = Evidence of waste containment measures and presumptive evidence of uncontained waste. 3 = Visual or analytical evidence of uncontained waste, or no evidence of waste containment present. The above criteria for assigning a zero value are similar to the present criteria for assigning a zero containment value, except that they reflect the current RCRA regulations. If compliance with RCRA requirements cannot be documented, but waste containment measures of unknown efficiency exist, a higher value (e.g., 1) could be assigned under this alternative. This category would include sites at which a liner or cover is installed but at which the liner's compatibility and integrity were not evaluated, or at which a diversion structure or leachate collection system is present, but at which its operating efficiency was not determined. A still higher value (e.g., 2) could be assigned when there is analytical evidence of uncontained waste, even if some waste containment measures exist. The analytical evidence could include contaminated soil, ponded water, or leachate where the contaminants can be attributed to the wastes disposed at a site. While this 34 ------- analytical evidence would be insufficient to score an observed release, it does indicate that containment is poor and that waste constituents have a higher potential for entering the environment. The criteria for assigning the highest value (e.g., 3) are similar to the present containment factor criteria. Sites at which no waste containment measures are evident, or at which waste contact with the environment can be visually observed, would be assigned this highest value. This alternative illustrates the use of evidentiary criteria, rather than RCRA-based technical criteria, to rate most levels of waste containment. RCRA-based criteria, not evidentiary criteria are, however, used to rate the highest level of containment because the lack of analytical evidence of uncontained waste does not necessarily indicate that a release has not or will not occur. It may just mean that the evidence has not been found. For example, where monitoring has been done, the lack of such evidence may be the result of temporal or spacial variations in environmental concentrations or faulty monitoring procedures. It may also mean that substances being released were not those tested for in the sample analysis. This is discussed further in Wolfinger (1986). Consequently, it is possible for sites with identical wastes and identical containment measures to be assigned different containment ratings under this alternative because evidence has been found at one site but not at the other. 35 ------- It should be noted that this evidentiary approach may be more applicable to the surface water route than to the ground water route, Certain types of evidence that would be of extreme importance in the ground water route (e.g., leachate or soil samples from below landfills, surface impoundments, etc.) are not easily obtainable or have risks inherent in their collection (e.g., drilling through a landfill could open conduits for waste migration). The collection of these additional data would likely require an expansion of the current site inspection program. Without further analysis, it is not possible to determine whether the data required to use the above criteria or the data required to use the current ground water containment factor are more costly to obtain. This approach is not likely to result in any significant shift in the distribution of containment factor scores. The highest and lowest evaluation criteria under this alternative are essentially the same as those of the current containment factor. As a result, there would likely be little change in the distribution of these two values. There would be some limited shift in the distribution of the two intermediate values, but probably almost no shift in the absolute number of sites being assigned an intermediate value. 3.5 Development of Evidentiary Criteria Without a Zero Value This alternative is the same as that in Section 3.4, except that the zero rating value is eliminated. The following evaluation criteria illustrate this approach: 36 ------- Rating Criteria 1 = Waste containment measures evident, no evidence of uncontained waste. 2 = Waste containment measures evident with presumptive evidence of uncontained waste. 3 = No waste containment present, or visual or analytical evidence of uncontained waste. The zero rating value has been eliminated both to remove dependence of the criteria on RCRA land disposal requirements and to indicate that complete waste containment over time cannot be absolutely assured even with measures that comply with RCRA requirements. (This latter consideration could also be incorporated in the alternative in Section 3.4 by assigning a value of 1, not zero, to containment that meets RCRA requirements and by adjusting all other values accordingly.) The advantages and disadvantages of this alternative are essentially the same as those discussed in Section 3.4. However, no differentiation is made for different levels of containment when there is no evidence of uncontained wastes. This is addressed in Section 3.6. 3.6 Integration of Evidentiary and Predictive Criteria This alternative illustrates the integration of evidentiary and predictive criteria in the rating of containment. Under this approach, sites at which there is evidence that hazardous substances have been released from a containment area would be assigned a higher value than sites at which there is no evidence of release. Where there is no evidence, the rating would be based on predictive 37 ------- technical requirements. It is anticipated that the type of evidence that would be used under this alternative would be evidence that shows migration from a waste containment area, but which is not adequate for demonstrating an observed release to surface or ground water (e.g., contaminated soil, leachate, erosion trails). This alternative would have essentially the same advantages and disadvantages as the alternative presented in Section 3.4. However, it would have the added advantage of being able to differentiate among sites at which there was no evidence of uncontained waste. One possible option under this alternative is illustrated below (for the ground water route for a landfill): Rating Criteria 0 = Demonstrated compliance with RCRA-based land disposal requirements. 1 = No evidence of uncontained waste; single liner and functioning leachate collection system. 2 = No evidence of uncontained waste; single liner, no functioning leachate collection system. 3 = Presumptive evidence of uncontained waste. 4 = Visual or analytical evidence of uncontained waste, or no evidence of waste containment present. 3.7 Development of Time-Dependent Criteria Since the physical, chemical, and biological processes that result in contaminants being released from waste sites are continuous, the duration of time these processes have been in operation affects the integrity of containment structures at the 38 ------- site, the quantity of any constituents released from the site, and the nature of the wastes that remain at the site. The longer wastes are at a site, the greater the overall potential for there to be a release from the site, especially if the site is abandoned or not maintained. The approach under this alternative is to use criteria that are based on the elapsed time since waste deposition to evaluate the containment at a site. Such criteria would likely result in a more uniform distribution of containment values. The following evaluation criteria are an example of criteria that could be developed (they are presented solely for illustration purposes; they are not based on an analysis of containment failure data): Rating Criteria 0 = Less than two years elapsed time since the initial disposal of hazardous substances. 1 ™ At least two years, but less than ten years, since initial disposal. 2 = At least ten years, but less than twenty years, since initial disposal. 3 = Twenty years or more since initial disposal or no evidence of waste containment. The use of time-dependent criteria would, by its very nature, tend to differentiate sites on the basis of the timing of a potential release (i.e., how soon is it likely to occur). For example, under this approach, older sites would be assigned higher rating values than those newer sites where releases are equally likely to occur within the same period of time from waste deposition. Consequently, 39 ------- newer sites would, in general, tend to have lower HRS migration scores than older equivalent sites. The net effect would be that sites that are close to (or are already) releasing hazardous substances would be more likely to be placed on the NPL than sites that are equally likely to, but further away in time from, releasing hazardous substances. The desirability of such an impact is a policy issue. Furthermore, there would be several drawbacks to the application of time-dependent criteria. First, it would be necessary to define a viable measure of time. Documentation of the wastes managed at a site has proven difficult (see Section 3.3); documentation of the dates of deposition would be even more difficult. Disposal dates are often unavailable because of poor or non-existent recordkeeping. Further, wastes were often deposited at different times and in different amounts and mixtures at a site. Surrogate measures, such as years since site opening or site closing, suffer from similar information collection problems and may differ significantly from the actual years since waste deposition. Second, the interactions between site age, containment structure integrity, and contaminant mobility are very complex. It is not clear that they could be adequately defined or be simplified enough to be incorporated in the HRS. Site-specific and waste-specific characteristics would be very important and would have to be considered. For example, buried drums in a desert environment 40 ------- generally corrode more slowly than those buried in locations with high ground water tables where the drums are continually in contact with water and quickly corrode. Furthermore, waste characteristics, such as corrosivity and reactivity, also strongly affect the drum life. For these reasons, time-dependent criteria do not appear to be a viable alternative for use in the HRS containment factor. 3.8 Use of Criteria Based on Waste Disposal Location This alternative illustrates the use of criteria that are based on waste location rather than on the type of waste management unit in which wastes are deposited. For each water route, the criteria are differentiated by whether the wastes are deposited below the ground surface or deposited at or above the ground surface. The assumption is that waste containment requirements are similar for wastes deposited in similar locations (i.e., above or below ground), regardless of the type of unit in which they are placed. For example, for the surface water route diking is important for wastes deposited above ground, but is not important for wastes deposited below ground. For each pathway under this alternative, containment is evaluated for both locations using predictive criteria similar to those in the current containment factor. The criteria are modified, however, to make them more applicable to the differences in containment likely to be present at CERCLA sites and to remove some 41 ------- of the subjectivity that results from use of current site inspection data. For example, under this alternative distinctions are made as to whether or not liners are present, not on how permeable the liners are or whether they are compatible with the wastes at the site. Such distinction may not be appropriate for CERCLA sites since it appears that the vast majority of these sites do not have liners. Further, based on current site inspection data, asssessments of liner integrity, permeability, and compatibility are often subjective. Illustrative criteria for this alternative are presented in Table 3-1. It should be noted that the values listed in Table 3-1 are based on engineering judgments, not an analysis of modeling results or site data. Such analysis may be necessary if this approach is to be considered further. In using the criteria in Table 3-1, it is recommended that evidence of properly functioning waste containment measures be required to assign values corresponding to excellent or good containment. A lack of evidence would result in a rating of poor containment if containment measures were present. When both types of disposal locations are present at a site, rating values would be assigned to each and the highest rating value would be used to compute the route score. One advantage of this approach is that it can be applied to any type of waste containment unit at a site; it is not limited to the four types in Tables 2-1 and 2-2. It also eliminates some of the subjectivity present in the current containment factor. Further, it 42 ------- TABLE 3-1 ILLUSTRATIVE CONTAINMENT FACTOR VALUES BASED ON WASTE DISPOSAL LOCATION Rating Value Criteria Ground Water Surface Water Below-Grade Disposal Excellent containment (4 of 5 technologies functioning) Good containment (2 of 5 technologies functioning) Poor containment (1 of 5 technologies functioning or efficiency unknown) No containment Above-Grade or Surface Disposal Excellent containment (4 of 5 technologies functioning) Good containment (2 of 5 technologies functioning) Poor containment (1 of 5 technologies functioning or efficiency unknown) No containment 2 3 2 3 0 0 1 2 2 3 Containment Technologies • Engineered Cover • Leachate collection • Double liner • Run-on diversion Diking with adequate freeboard 43 ------- may provide a somewhat more uniform distribution of containment values since only one containment technology needs to be present to achieve a value of 1. One major disadvantage of this approach is that it assumes that each containment technology in Table 3-1 is of equal importance. This may not be the case. For example, leachate collection may be more important than run-on diversion. A further disadvantage of the approach is that the five listed containment technologies may not be applicable, nor even meaningful, for every type of waste management unit. For example, surface impoundments are not likely to have covers, while almost all landfills would have some limited cover. Additional refinements would thus be necessary if this approach was deemed to warrant further investigation. 3.9 Development of Criteria Based on Site Drainage This alternative represents a further simplification of the approach in Section 3.8. The simplification is directed at providing a more uniform distribution of containment values and further i reducing subjectivity in the evaluation of containment. To do this, the evaluation criteria are based on those containment technologies more likely to be present at CERCLA sites (i.e., some type of cover material or cap and drainage or diversion structures). Technologies (e.g., liners, leachate collection systems) not likely to be present or whose evaluations are likely to be subjective are not included. 44 ------- Under this alternative, the level of waste containment could be equated with good site drainage using criteria similar to the following: Rating Criteria 0 = Cap or cover, functioning drainage or diversion structure with collection system. 1 = Cap or cover, functioning drainage or diversion structure. 2 = No or non-functioning drainage or diversion structure, no ponding observed. 3 = No or non-functioning drainage or diversion structure, ponding observed. These criteria are similar to the current surface water route containment evaluation methodology and, except for liner and leachate collection system evaluations, they are also similar to the ground water criteria. This approach eliminates the evaluations of liner permeability and leachate collection system operation from the present HRS criteria because accurate information on their effectiveness is difficult to obtain without extensive monitoring programs. For example, the presence of a clay or synthetic liner is not by itself evidence that the waste is contained. The synthetic liner may be torn or the clay may be cracked or channeled such that the liner is quite permeable. There is no way of making this determination without more extensive monitoring data than is presently available from site inspections. 45 ------- The primary advantage of this approach is that it would provide a more uniform distribution of site scores, if this is desired. In addition, the proposed criteria should be simpler to implement than the existing criteria because they are not specific to each waste management technology. The necessary evaluations could be based on visual evidence; evaluations of liner permeability or compatibility are not required. The criteria could also be integrated with the evidentiary criteria discussed in Sections 3.4 to 3.6 to obtain a more uniform distribution among containment values than presently exists and to incorporate other factors affecting waste containment. The major disadvantage with this alternative is that it ignores elements (e.g., liners) that are important to good containment. It is structured to provide a more uniform distribution of values, not to meaningfully evaluate the overall effectiveness of containment at a site. 3.10 Integration of Containment and Physical State Factors The physical state of a waste is included as a rating factor under the current HRS route characteristics category. It is evaluated when there is no evidence of an observed release for the ground water or surface water routes. The following criteria are presently used to assign a value to physical state: Rating Criteria 0 = Consolidated or stabilized solid. 1 = Unconsolidated or unstabilized solid. 46 ------- Rating Criteria 2 = Powder or fine material. 3 = Liquid, sludge or gas. The physical state of a waste affects the potential for the waste (or its hazardous constituents) to migrate from a site or alternatively for it to be contained at a site. Physical state can thus be used as a measure of either waste containment or waste migration potential. It is currently used as a measure of waste migration potential in the ground water and surface water route characteristics factor category of the HRS. Physical state could, however, be integrated with the current containment factor (or a revised containment factor) and deleted from the route characteristics factor category. The advantages and disadvantages of this are discussed below. Most of the sites in the automated NPL technical data base for which there are physical state and containment values have a value of 3 assigned to both factors, indicating that there are at least some uncontained liquids, sludges, or gases present at most of these sites.* Of the 698 sites in the automated NPL technical data base for which a value was assigned to the physical state factor, 617 of the sites (88 percent) received a value of 3. Eighteen sites *The physical state factor value assigned to a route is the highest physical state factor value assigned to any of the substances at the site that can migrate along that route. Thus, there could be solids, in addition to liquids, sludges, or gases, present at a site that receives a factor value of 3. 47 ------- (3 percent) were assigned a value of zero, 32 sites (5 percent) a 1, and 31 sites (4 percent) a value of 2. Based on the above data, it appears that the primary function that physical state currently serves in the HRS is to assist in screening out (i.e., lowering scores) of sites that do not contain at least some liquids, sludges, or gases. If this is the objective, it could most likely be carried out more effectively by integrating physical state into the containment factor, since this would place greater weight on physical state in the overall route score. Under this approach for each waste management unit, both the type of containment and the physical state of the wastes in that unit would be considered in assigning a rating value to that waste management unit. This integration could at the same time provide some further differentiation among the vast majority of CERCLA sites at which uncontained liquids, sludges, and gases are present in some, but not all, waste management units at the site. This would, however, have to be verified by further analysis if this alternative was deemed to warrant further investigation. There are numerous ways in which physical state could be integrated into the containment factor. Two examples are illustrated below, using both evidentiary and predictive approaches. 3.10.1 Use of Evidentiary Criteria The physical state and containment factors could be integrated for each waste management unit using evidentiary criteria similar to 48 ------- those presented in Section 3.4. The following criteria illustrate this approach for rating a waste management unit: Rating Criteria 0 = Evidence of only stabilized or contained solid in the waste management unit 1 = Unstabilized, contained solid in the waste management unit with presumptive evidence of containment failure, or containment efficiency unknown -or- Evidence of contained liquid or sludge in the waste management unit 2 = Uncontained solid in the waste management unit -or- Contained liquid or sludge in the waste management unit with presumptive evidence of containment failure, or containment efficiency unknown 3 = Uncontained liquid, sludge, or gas in the waste management unit The effective use of these criteria focuses on the determination of containment. For example, a solid must be stabilized or contained for the waste management unit to be assigned a zero value. Containment or stabilization must be supported by evidence similar to that discussed in Section 3.4. The advantages and disadvantages of this approach would be similar to those advantages and disadvantages discussed for the evidentiary approach in Section 3.4. 3.10.2 Use of Predictive Criteria The physical state and containment rating factors could also be integrated for each waste management unit using predictive criteria similar to those of the current containment factor or to those illustrated in the alternatives. The following criteria illustrate 49 ------- how some of the waste location criteria in Section 3.8 could be integrated with the physical state factor: Rating Criteria 0 = Stabilized solid, or solid disposed with four of the five functioning containment technologies.* 1 = Unstabilized solid disposed with two of the five functioning containment technologies -or- Liquid or sludge disposal with four of the five functioning containment technologies 2 = Unstabilized solid disposal with one of the five functioning containment technologies -or- Liquid or sludge disposal with two of the five functioning containment technologies -or- Analytical evidence of leaking containment 3 = No waste containment for solids, liquids, sludges, or gases Evidence of functioning containment technologies, similar to that discussed in Section 3.8, would be required to ensure consistent evaluation. The advantages and disadvantages of the above criteria would be similar to those discussed in Section 3.8. *From Section 3.8, the five containment technologies are: dike with adequate freeboard, run-on diversion structure, double liner, engineered cover, and leachate collection system. 50 ------- 4.0 SUMMARY AND RECOMMENDATIONS The current HRS containment factor uses evaluation criteria that are based on the requirements of the RCRA Subtitle C land disposal regulations as of early 1982. The analysis in Chapter 2 indicates that many wastes at CERCLA sites are not well contained and that the use of evaluation criteria based on the RCRA Subtitle C regulations may not provide much differentiation between sites. This is even more likely to be true if the HRS containment factor is modified to reflect the current land disposal regulations. Furthermore, the current HRS containment evaluation criteria may be more complex and subjective than is warranted. In light of the type of containment likely to be present at CERCLA sites, it may be more meaningful for HRS screening purposes just to distinguish whether a containment technology is present at a site and not to try to make fine distinctions about how effective it is. Based upon the analysis in Chapter 2, several issues that need to be considered in any modification of the HRS containment factor have been identified. One issue is whether the containment factor should continue to be included in the HRS. While the containment factor appears to be effective in screening out sites that have relatively high levels of containment, the number of sites receiving less than the maximum containment factor value appears to be small. Assuming that containment is to remain a factor in the HRS, the primary issue that needs to be addressed is whether the containment 51 ------- factor is to be used primarily to screen out those relatively few sites that have adequate containment or to provide increased differentiation among all sites without an observed release. For reasons discussed in Chapter 2, it is unlikely that any single containment factor can be developed to do both effectively. If it is decided to use the containment factor to screen out sites with adequate containment, there are two basic issues that need to be addressed: • How adequate containment is to be defined and evaluated for HRS purposes. • Whether the rating assigned for adequate containment is to be a zero or a non-zero value. If it is decided to use the containment factor to differentiate among all sites without observed releases, the primary issue is more technical. The primary issue is defining containment characteristics that meaningfully delineate actual, but likely small, differences in containment effectiveness and that at the same time are representative of the containment technologies likely to be present at most sites eligible for NPL listing. For reasons discussed in Chapter 2, it is unlikely that any useful definition could be based solely on technical requirements for land disposal. Nine alternative approaches for evaluating containment in the HRS are identified and examined. These alternatives have been developed to collectively illustrate ways that the current HRS containment factor could be modified to: 52 ------- • Increase its consistency with RCRA Subtitle C land disposal regulations • Include criteria other than RCRA-based technical requirements • Provide greater differentiation among sites, and/or • Reduce the subjectivity of the current containment evaluation criteria No one alternative approach does all four. One option for the development of containment factors under each alternative approach is briefly outlined to illustrate the types of containment options possible under each alternative. The advantages and disadvantages of each alternative are discussed, and issues that need to be resolved before more complete options can be developed under the alternative are identified. As part of the development of the alternatives, ten existing ranking systems that consider containment in rating the threat posed by hazardous waste sites were reviewed along with three EPA hazardous waste policy analysis models that also account for containment. Five of the ranking systems were found to have containment factors that are very similar in concept and application to that of the HRS. Two ranking systems were found not to explicitly evaluate containment in ranking a site. The other three ranking systems were found to have containment factors that are not comprehensive nor well defined. The three EPA hazardous waste policy analysis models were found not to be applicable for use in making site-specific comparisons of containment effectiveness. 53 ------- Of the nine alternatives developed, one focuses on updating the current RCRA-based containment evaluation criteria. One examines the use of time-dependent, rather than RCRA-based, criteria. Two others examine the use of evidentiary criteria rather than RCRA-based criteria. Three others examine the integration of waste quantity, physical state, or evidentiary criteria into the containment factor. The remaining two examine ways to reduce the subjectivity in the current evaluation criteria and to make the current criteria more applicable to the differences in containment likely to be present at CERCLA sites. It is recommended that two of these alternatives be considered for possible further development and that six others be eliminated from further consideration. Further development of the ninth alternative, Updating of the Current Containment Factor, is entirely a policy issue. The two alternatives that are recommended for possible further development are the following: • Integration of Evidentiary and Predictive Criteria in the Containment Factor. • Integration of the Physical State Factor Into the Containment Factor. The integration of evidentiary and predictive criteria offers several possible advantages. First, it could be used to assign higher rating values to sites at which there is evidence that hazardous substances are leaving waste management areas than to sites 54 ------- at which there is no detected evidence that hazardous substances are leaving the waste management area. Such evidence is currently not considered in the HRS unless it can be used to document an observed release. Second, where there is no detected evidence of a release, predictive criteria could also be used to assign containment factor values based on differences in containment technologies. For reasons previously discussed, it is recommended that such criteria only distinguish whether a technology is present and not whether there are differences in the effectiveness of the technology. Third, this approach would likely result in a slightly more uniform distribution of containment values; however, any such change is not expected to be large. The major disadvantage of this approach is that sites that have releases that are not detected, due for example to environmental variations in concentrations, would receive a lower rating than they should. However, the same problem can also occur with the current observed release factor. From the analysis presented in Section 3.10, it appears that there is not much differentiation of sites based on the current physical state factor. Instead, the physical state factor appears to assist in screening out sites that do not contain at least some liquids, sludges, or gases. Regardless of which containment alternative, if any, is selected for further development, it is recommended that the integration of the physical state and containment factors be evaluated further. Such an integration may 55 ------- increase the effectiveness of the HRS in screening out those sites that have an adequate level of containment for the types of wastes present at the site. It may also increase the differentiation among the vast majority of CERCLA sites that have at least some uncontained liquids, sludges, or gases. The other six alternatives are not recommended for further development for a variety of reasons. The integration of the waste quantity and containment factor is not recommended for further development because of significant problems with data availability. If these could be resolved, this alternative would be the preferred approach because it could be used to both screen out sites with adequate containment and to provide a considerably more uniform distribution of containment values. The use of time-dependent criteria is not recommended because it is unlikely that the complex relationships between site age, containment integrity, and contaminant mobility could be adequately defined or simplified for effective use in the HRS. This approach also suffers from data acquisition problems. The two evidentiary approaches that are not integrated with predictive criteria are not recommended because the integrated approach offers several advantages over these two. The other two approaches are not recommended because many of their simplifications either ignore or equate differences in containment that may be significant for certain types of containment 56 ------- structures. Further, some of the more meaningful simplifications can still be incorporated into whatever containment factor is ultimately used in the HRS. However, it should be noted that, if it is decided to modify the containment factor to provide a more uniform distribution of containment values, these latter two approaches are the only ones identified that are likely to be able to achieve this. 57 ------- APPENDIX A REVIEW OF CONTAINMENT FACTORS IN OTHER SITE RANKING SYSTEMS This appendix reviews the containment factors that have been incorporated in other systems. Ten ranking systems that consider containment in rating the threat posed by hazardous wastes sites have been identified along with three EPA hazardous waste policy analysis models that also account for containment. The review of these systems focuses on how containment is used in each system and how containment effectiveness is defined and evaluated. Important similarities and differences between these factors and the HRS containment factor are identified. The ten ranking systems reviewed are: • HARM • SAS • HARM II • PERCO • GSR • Illinois Rating Scheme • ADL • Rating Methodology Model • S.P.A.C.E. for Health • Dames and Moore Methodology The three EPA hazardous waste policy analysis models reviewed are: • Liner Location Risk and Cost Analysis Model • Hazardous Waste Tank Failure Model • RCRA Risk-Cost Analysis Model Before reviewing these systems, it should be noted that the HRS containment factor is a multiplicative factor that is rated on a scale of 0, 1, 2 or 3. Since it is a multiplicative factor, if it were to be divided by three and the divisors used to normalize the corresponding HRS pathway scores were also to be divided by three, 59 ------- there would be no change in the HRS pathway scores. In this case, the HRS containment factor would be rated on a scale of 0 to 1 and would have values of 0, 0.33, 0.67- and 1. A.I HARM Hie Hazard Assessment Rating Methodology (HARM) is used by the U.S. Air Force to rank hazardous substance sites for priority attention for follow-on site investigations and confirmation activities under Phase II of the Air Force's Installation Restoration Program (IRP). HARM is designed to use data developed during the Record Search (Phase I) portion of the IRP (Engineering-Science, 1983). Record Searches are essentially equivalent to EPA Preliminary Assessments. The HARM score is developed from four subscores: Receptors, Pathways, Waste Characteristics, and Waste Management Practices. A total risk score is determined by averaging and normalizing the first three subscores. The total risk score is then multiplied by the Waste Management Practices subscore to produce the HARM score. The Waste Management Practices subscore is a measure of the containment at the site and ranges from 0.1 to 1 as follows: • No containment 1.0 • Limited containment 0.95 • Fully contained and in full compliance 0.10 Table A-l shows the guidelines for a determination of fully contained. These guidelines are quite similar to those used to assign a zero rating to the current HRS containment factor. 60 ------- TABLE A-l FULLY CONTAINED SITES—HARM GUIDELINES Landfills; • Clay cap or other impermeable cover • Leachate collection system • Liners in good condition • Adequate monitoring walls Surface Impoundments; • Liners in good condition • Sound dikes and adequate freeboard • Adequate monitoring wells Spills; • Quick spill cleanup action taken • Contaminated soil removed • Soil and/or water samples confirm total cleanup of the spill Fire Protection Training Areas; • Concrete surface and berms • Oil/water separator for pretreatment of runoff • Effluent from oil/water separator to treatment plant Source: Engineering-Science, Comparison of U.S. Air Force Hazard Assessment Rating Methodology (HARM) with U.S. Environmental Protection Agency Hazard Ranking System (HRS) at Four Air Force Bases Evaluated under the Phase I Installation Restoration Program, Engineering-Science, Atlanta, GA, April 1983. 61 ------- The HARM Waste Management Practices subscore is thus very similar in concept and application to the HRS Containment Factor. The main differences are that its lower limit is 0.1, rather than zero, and that it has one, rather than two, intermediate rating categories. The primary disadvantage of the HARM factor is that it does not address route-specific differences in containment. A.2 HARM II The Hazard Assessment Rating Methodology II (HARM II) is a modification and extension of the HARM system that is intended to permit the use of site-specific monitoring data in setting priorities. HARM II is used by the U.S. Air Force in Phase II of the IRP program to set priorities for detailed site investigations and possible remedial action (Barnthouse et al., 1986). HARM II has a Waste Containment Effectiveness factor that is based on the HRS containment factor (Barnthouse et al., 1986) and that is consequently very similar in concept and application to the HRS containment factor. The Waste Containment Effectiveness factor is a multiplicative factor used to determine surface water or ground water pathway subscores. Tables A-2 and A-3 present the Waste Containment Effectiveness factor for the surface water and ground water pathways, respectively. Again, the main difference compared to the HRS, is that the lower limit is 0.1, not zero. 62 ------- TABLE A-2 HARM II WASTE CONTAINMENT EFFECTIVENESS FACTORS FOR SURFACE WATER PATHWAY Containment Structure Score Landfills Clay or other cover in sound condition without evidence 0.1 of severe erosion; if in floodplain, diked effectively to prevent floodwater encroachment Some problems with cover or dike/diversion systems; 0.5 no wastes exposed Problems with both cover and dike/diversion system; 0.8 or serious problem with one of these; no exposure of wastes yet Cover and/or diking system (if needed) absent; wastes 1.0 exposed Surface Impoundments Sound dikes, adequate freeboard, and no erosion evident 0.1 Inadequate freeboard 0.5 Potentially unsound dike 0.8 Dikes unsound, leaking, or in danger of collapse, or 1.0 evidence of past spillover Spills Contaminants material apparently removed completely; 0.1 but possible occurrence of a spill Contaminated area covered with impervious material 0.5 Contaminants completely covered with soil, area 0.8 revegetated No cleanup action or covering done 1.0 63 ------- TABLE A-2 (Concluded) Containment Structure Score Tanks Tanks in sound condition and inspected regularly; tank 0.1 area effectively bermed to contain any spills (and subsequent rainfall) Tanks in sound condition and bermed, but berm system 0.5 or inspection system possibly insufficient Tanks in sound condition, but not bermed 0.8 Tanks unsound 1.0 Fire Protection Areas Berms; oil-water separator for treatment of runoff; 0.1 oil-water separator effluent to treatment plant Some deficiency in the above 0.8 None of the above 1.0 Source: Barnthouse, L.W. et al., Development and Demonstration of a Hazard Assessment Rating Methodology for Phase II of the Installation Restoration Program, (ORNL/TM-9857), Oak Ridge National Laboratory, TN, 1986. 64 ------- TABLE A-3 HARM II WASTE CONTAINMENT EFFECTIVENESS FACTORS FOR GROUND WATER PATHWAY Containment Structure Score Landfills Liner essentially impermeable, intact, and compatible 0.1 with waste; cover of low permeability, and intact; leachate collection system; adequate monitoring wells Physical containment system suitable; no monitoring 0.5 Deficiencies in physical containment system (e.g., 0.8 moderately permeable liner, no leachate collection, or defective cover) Cover and/or diking system (if needed) absent; wastes 1.0 exposed Surface Impoundments Liner essentially impermeable and compatible with 0.1 waste; double liner or leakage detection system System essentially sound, but no double liner or 0.5 leakage detection system Liner moderately permeable or in deteriorating condition 0.8 No liner or incompatible liner; soil contaminated 1.0 by leakage from impoundment Spills Contaminated material apparently removed completely; 0.1 but possible occurrence of a spill Contaminated area covered with impervious material 0.5 No cleanup action or covering done 1.0 65 ------- TABLE A-3 (Concluded) Containment Structure Score Tanks Tanks in sound condition and inspected regularly; tank 0.1 area lined adequately Tanks in sound condition; no liner 0.5 Tanks leaking 1.0 Fire Protection Areas Concrete surface 0.1 No concrete surface 1.0 Source: Barnthouse, L.W. et al., Development and Demonstration of a Hazard Assessment Rating Methodology for Phase II of the Installation Restoration Program, (ORNL/TM-9857), Oak Ridge National Laboratory, TN, 1986. 66 ------- A.3 GSR The Confirmation Study Rating (GSR) model is used by the U.S. Navy in the Navy Assessment and Control of Installation Pollutants (NACIP) Program to assign priorities for further study to hazardous substance sites. The GSR model is based on the HARM system (Luecker, 1982). In the GSR model, Receptor, Pathway, and Waste Characteristics rating factors are multiplied by a Waste Management factor to obtain the site score. The Waste Management factor is a measure of the containment at the site and is identical to the Waste Management Practices subscore of HARM, with one exception. The one difference is that limited containment is given a value of 0.8 in the CSR model rather than 0.95 as in HARM. A.4 ADL The Arthur D. Little, Inc. (ADL) system is an adaption of the HRS that was developed for the Chemical Manufacturers Association (Arthur D. Little, Inc., 1981). Like the HRS, the ADL system contains a multiplicative containment factor that is used only when there is no observed release for the ground water or surface water pathway. The containment factor is scored as either a 0 or a 1. For both pathways, it is scored 0 if the site is certified to have an impermeable liner (natural or artificial), the high water table is below the liner, and an impermeable cover has been installed. Otherwise, it is scored 1. 67 ------- The major difference between the HRS containment factor and the ADL containment factor is that the ADL containment factor does not attempt to evaluate intermediate levels of containment. It is based on the assumption that the rater should not try to make fine distinctions in the degree of containment present at a site. The site is scored as either being contained or not contained. A.5 S.P.A.C.E. for Health The System for Prevention, Assessment, and Control of Exposures and Health Effect from hazardous sites (S.P.A.C.E. for Health) was developed by the Centers for Disease Control (CDC) for use in public health assessments of hazardous sites (French et al., 1984; Kay and Tate, 1984). The system is used to assign priorities to sites, based on the potential of the site to endanger human health. Site Characteristics is one of four factors used in S.P.A.C.E. for Health for determining the site priority. There are seven elements used to rate Site Characteristics. One element, Site Management and Containment, provides a measure of the containment at the site. Site Management and Containment is evaluated through the use of criteria that are identical to those of the HRS containment factor (see Tables 2-1 and 2-2). A.6 SAS The Site Assessment System (Michigan, 1983) is used to assess and prioritize release sites for further investigation and possible remedial action. 68 ------- SAS considers containment through a Release Potential Category. All containment structures (e.g., landfill) are first rated for each pathway as to whether their containment effectiveness is unquestionably adequate, inadequate, or unknown for that pathway. A rating of unquestionably adequate containment for a pathway must be based on information that indicates the structure was designed, constructed, and is operating so that emissions of hazardous substances are effectively prevented from entering the environment through that pathway. Limited guidelines are provided for this determination. The Release Potential Category is scored for each containment structure that is not rated unquestionably adequate. This is done by first determining the fraction of the total waste quantity at the site that is present in each containment structure (this results in a value between 0 and 1 for each structure). The physical state of the waste in each containment structure without unquestionably adequate containment is then scored as follows: solid—1.0, semi-solid—1.5, liquid or gas—2.0. The quantity fraction and the physical state score are then multiplied to give a release potential score for each structure. The release potential scores for each structure without adequate containment for a specific pathway are then summed to obtain the release potential score for the pathway. The aggregated pathway release potential scores may not exceed 2. 69 ------- The aggregated pathway release potential score is then used as a multiplicative factor in determining the overall pathway score. SAS thus differs from the HRS in that it uses the concept of containment strictly as a means of measuring the quantity of waste that can migrate along a pathway and not as a measure of the relative effectiveness of the containment structure at a site. In the HRS containment is used for both purposes. Any waste in a containment structure receiving an HRS containment score of zero for a pathway is not considered in the rating of the HRS waste quantity factor for that pathway. A.7 PERCO The Prioritization of Environmental Risks and Control Options (PERCO) model (Arthur D. Little, Inc., 1983) was developed for the Massachusetts Department of Environmental Quality Engineering for use in ranking contaminated sites in terms of immediate and long-term environmental and human health hazards. The ranking is used to provide a rationale for allocation of state remedial action funds. Containment is not explicitly considered in ranking a site with PERCO. Rather containment is used in the following manner. PERCO contains a multiplicative factor called contaminant severity. It is calculated based on the measured concentrations of contaminants in the environment of a site and acceptable ambient levels of those contaminants, using relationships defined in PERCO. Contaminant 70 ------- severity is assigned a value ranging from zero to 100. If monitoring data are not available for a site, then the contaminant severity is approximated for the site, using information from other similar sites for which monitoring data are available. To do this, at least ten similar sites, whose contaminant severity scores fall within an acceptable range, are identified; the contaminant severity scores for those sites are then averaged. This average contaminant severity score is used as the contaminant severity score for the site. As one option, similar sites may be identified through professional judgment. For two pathways (air and surface water), there is also an option for identifying similar sites through the calculation of a similarity score. One of the factors used to calculate the similarity score is containment integrity. Containment integrity is a measure of the potential for leakage from three types of containment structures: drums; tanks; and pools, pits, or lagoons. Drum integrity is scored as follows: many drums leaking—8; a few widely separated drums leaking—4; all drums intact—2; no drums on surface—0. Tank integrity is scored as follows: at least one tank leaking—10; all tanks intact—6; no tanks on site—0. Pools, pits and lagoons are scored as follows: large pools or lagoons evident, but not associated with leaking drums or tanks—10; no such pools evident—0. 71 ------- For the air pathway, there is one further option for calculating similarity scores to identify similar sites. This involves calculating a release potential score for sites that may be similar. The release potential score is calculated for each site, based on the condition of drums and tanks at the site; the volume of liquids in pools, pits, and lagoons at the site; and the areal extent over which liquid wastes are buried at the site. The condition of drums is scored on a scale of 0 to 40, primarily based on the percent of drums present that are corroded, damaged, have loose covers, or are otherwise capable of leaking contents to the environment. The condition of tanks is scored for each tank on a scale of 0 to 7, based on whether the tank is leaking, has air vents to the outside, and/or has an open top. The liquid volume, the surface area, and the two scores for drum and tank condition are manipulated mathematically and then added to obtain the release potential score. PERCO thus differs from the HRS in that it does not consider waste containment in the rating of a site. Each site is rated solely on monitoring data obtained either from that site or from at least ten similar sites. Containment is considered only in the identification of the sites that are similar to the one being rated. The containment evaluation factors used in PERCO are appropriate only for this purpose; they are not intended for, nor are they appropriate for, rating the relative effectiveness of containment structures. 72 ------- A.8 Illinois Rating Scheme The Illinois Rating Scheme is used by the State of Illinois as a screening tool, for regional planning, to identify and prioritize sites or areas potentially affecting ground water for more detailed study and evaluation (Gibb et al., 1983). The Illinois Rating Scheme consists of four additive factor categories, each of which is made up of several additive elements. One of these factor categories, Health Risk of Waste and Handling Mode, contains an element that is a measure of site containment. This element, Recorded Management of Waste, is evaluated differently for active industrial sites, active landfills, and abandoned sites as indicated in Table A-4. In using the criteria in the table, no guidance is provided as to what constitutes controlled or uncontrolled operation, a violation, or a site being well operated. The Illinois Rating Scheme differs from the HRS (and the nine other ranking systems reviewed) in that it considers limited features of a site's operational history (e.g., source of waste, types of violations) rather than its waste management characteristics in the ranking of the site. While containment effectiveness is related to operational history to a limited extent, this kind of approach does not appear practical to apply to CERCLA sites where information about the operational history of the site is extremely limited and often unavailable. 73 ------- TABLE A-4 ILLINOIS RATING SCHEME RATING RECORDED MANAGEMENT OF WASTE Criteria Rating Active Industrial Sites Principal On-site Storage or Disposal Method Incineration 0 Secure Containers 2 Treatment/Discharge 4 Land Application 6 Landfill 8 Surface Impoundments 10 Active Landfills Operational History No violations whatsoever; operated up to best 0 expectations No violations; generally well operated 2 No violations of a serious nature; generally well 4 operated Some violations of a serious nature; history doubtful 6 Selected violations of a serious nature; past history 8 unknown History of serious violations; essentially 10 uncontrolled for periods of time Abandoned Waste Sites Operational History Controlled operation; solely municipal wastes 0 involved Controlled operation; predominantly municipal wastes 2 involved Controlled operation; municipal and industrial wastes 4 involved Uncontrolled operation; municipal and industrial 5 wastes involved 74 ------- TABLE A-4 (Concluded) Criteria Rating Operational History (Concluded) Uncontrolled operation; predominantly industrial 8 wastes involved Uncontrolled operation; wastes of all types probably 10 present Source: Gibb, J. et al., Hazardous Waste in Ogle and Winnebago Counties; Potential Risk via Groundwater Due to Past and Present Activities, Illinois Department of Energy and Natural Resources, Document No. 83/26, Springfield, Illinois, 1983. 75 ------- A. 9 Rating Methodology Model The Methodology for Rating the Hazard Potential of Waste Disposal Sites (Rating Methodology Model) was developed for use by the EPA Office of Enforcement and the Oil and Special Materials Division in setting priorities for sites based upon preliminary assessment data (JRB Associates, Inc., 1980). The Rating Methodology model consists of four rating areas whose scores are added to obtain an overall site score. One of the four rating areas is Waste Management Practices which consists of eight rating factors, four of which are related to waste containment. These four rating factors are: • Use of liners • Use of leachate collection systems • Use of gas collection systems • Use and conditions of containers The criteria and rating scales used to evaluate each of these four rating factors are shown in Table A-5. The eight Waste Management Practices rating factors are weighted and summed to determine the score for the Waste Management Practices rating area. The rating factors in the Rating Methodology model were considered in the original development of the HRS. The rating factors for containment were found not to be suitable for inclusion in the HRS, in part because they were not comprehensive enough, nor well enough defined, for use in evaluating site containment. 76 ------- TABLE A-5 CONTAINMENT RATING FACTORS IN THE RATING METHODOLOGY MODEL Rating Factors Criteria Rating Scale Levels Use of Liners Use of Leachate Collection Systems Use of Gas Collection Systems Use and Condition of Containers Clay or other liner resistant to organic 0 compounds Synthetic or concrete liner 1 Asphalt-base liner 2 No liner used 3 Adequate collection and treatment 0 Inadequate collection or treatment 1 Inadequate collection and treatment 2 No collection or treatment 3 Adequate collection and treatment 0 Collection and controlled flaring 1 Venting or inadequate treatment 2 No collection or treatment 3 Containers are used and appear to be 0 in good condition Containers are used but a few are leaking 1 Containers are used but many are leaking 2 No containers are used 3 Source: JRB Associates, Inc., Methodology for Rating the Hazard Potential of Waste Disposal Sites (Draft Final Report), McLean, VA, 1980. 77 ------- Furthermore, the use of an additive containment factor was judged unsuitable because with an additive factor, a site could have total containment and still receive a high score. This is not the case with a multiplicative containment factor. A. 10 Dames and Moore Methodology The Dames and Moore Methodology was developed to evaluate waste disposal sites with respect to their potential for ground water and surface water contamination (Dames and Moore, undated). The Dames and Moore Methodology was adapted from the Rating Methodology model and consists of four rating areas. One of these rating areas is Spill Potential. Spill Potential applies only to landfills and consists of ten rating factors, four of which are related to waste containment. These four rating factors are: • Cover condition • Leachate management • Gas management • Personnel training The criteria and rating scales used to evaluate each of these four rating factors are shown in Table A-6. The ten Spill Potential rating factors are weighted and summed to determine the score for the Spill Potential rating area. These four rating factors have the same limitations as the rating factors in the Rating Methodology model from which they were derived. 78 ------- TABLE A-6 CONTAINMENT RATING FACTORS IN THE DAMES AND MOORE METHODOLOGY MODEL Containment Rating Factors Rating Scale Multiplier Criteria Levels Cover Condition Leachate Management 10 Well contoured, no cracking or vigorous vegetative cover, no signs of subsidence Signs of cover failure visible at 25% of site area Signs of cover failure visible at 50% of site area Large areas of subsidence, poor contouring, no vegetative cover in most of the site, exposed waste 8 Arid region; or leachate monitors installed but none detected No leachate monitoring, but leachate generation unlikely (arid weather, good cover, etc.) Some evidence of leachate generation at only certain portions of the site No leachate monitoring, or positive proof of leachate generation 0 79 ------- TABLE A-6 (Concluded) Containment Rating Factors Multiplier Criteria Rating Scale Levels Gas Management Personnel Training No gases are generated or generated gases are successfully managed. Inadequate information. Gas generation anticipated. Some venting measures Installed. Inadequate information. Gas generation anticipated. No venting measures installed. Gas generation is evident or anticipated. Signs of cap distress visible. All waste disposal related personnel are formally trained in safety and environmental control. Only waste disposal supervisory personnel are formally trained in safety and environmental control. Some informal training of waste disposal supervisory personnel. None 0 Source: Dames and Moore, Overview of Methodology for Rating Potential for and Significance of Ground and Surface Water Contamination from Waste Disposal Sites, Bethesda, Maryland, Undated. 80 ------- A.11 Liner Location Risk and Cost Analysis Model The Liner Location Risk and Cost Analysis Model is designed to be used by EPA to investigate cost/risk and cost/effectiveness implications of the land disposal of hazardous wastes under different technology, location, and waste stream scenarios. The model estimates the relative chronic risk to human health from land disposal facilities with different design technology, location, and waste stream characteristics. The model also estimates the cost of facilities with differing technologies and sizes. The model uses a series of submodels to predict contaminant releases, subsurface and atmospheric transport, human exposure, and health risk based upon dose-response factors (U.S. Environmental Protection Agency, 1985). Containment is taken into account within the Failure and Release Submodel of the Liner Location Model. For land disposal facilities (i.e., landfills, surface impoundments, waste piles, and land treatment), this submodel estimates both the probability of various types of failures (i.e., release of leachate) at selected times for different facility designs in different climates and the quantity of leachate released by the failures. Various combinations of liners and cover types and materials are used to define the different facility types. The model does not consider failure events that result in overland surface run-off releases. The Failure Analysis Submodel utilizes a "fault tree" approach to evaluate the frequency of occurrence of a failure event at a 81 ------- facility due to various elementary failure events (e.g., liner aging, infiltration of liquid, drum failure, cover consolidation, liner breach due to rise of ground water table). The fault tree approach traces a failure event backwards to identify all relevant elementary failure events which could combine to cause the failure. Each elementary failure event is given a probability distribution based upon theoretical and empirical data. The volume of liquid entering or exiting the facility if the elementary failure event occurs is also given a probability distribution. (Each elementary failure event is assumed to occur at random and is given either a triangular, binomial or time-dependent binomial distribution. The volume of liquids associated with an event is given a uniform distribution.) This submodel uses the fault tree structure to compute the probabilities of failure events which occur if some sequence of prerequisite elementary events have occurred. Monte Carlo simulation* is then used to estimate the cumulative probability of occurrence of each failure event for each facility type and climate and to estimate the volume of leachate released by the failure event. The volume of leachate released is estimated by incorporating a series of hydrologic inputs (e.g., infiltration from *Monte Carlo simulation consists of sampling from the probability distribution of the elementary events. The sampling is done through the use of a random number generator. After all the elementary failure events in a fault tree are evaluated, the model determines whether a failure event occurs and the volume released if it occurs. This process is repeated a large number of times using a computer to obtain an estimate of the relative frequency of occurrence of each failure event. 82 ------- precipitation) in the elementary failure events and tracking them with a mass balance analysis. The Liner Location Risk and Cost Analysis Model thus addresses containment in a very different manner than the HRS. However, as currently structured, it is not meant for, nor is it applicable to, site-specific comparisons of containment effectiveness. Rather it is intended primarily for use in analyzing the risks and costs associated with alternative regulatory strategies (e.g., alternative standards for waste containment at land disposal facilities). A.12 Hazardous Waste Tank Failure Model The Hazardous Waste Tank Failure (HWTF) Model is one of the models being used by the EPA Office of Solid Waste to support the development of regulations for hazardous waste tanks (IGF Incorporated and Pope-Reid Associates, Inc., 1986; Pope-Reid Associates, Inc., 1986). The HWTF model uses fault trees and Monte Carlo simulation (see Section A.11) to predict the timing of failure events (e.g., leaks, ruptures) for hazardous waste tank systems and to estimate release volumes associated with these failure events over a 20-year operating life. The HWTF model is currently designed to analyze four types of tanks: RCRA-permitted treatment tanks, RCRA-permitted storage tanks, accumulation tanks (storage for less then 90 days), and small-quantity generator tanks. For each type of tank, the HWTF estimates the timing of failure events and the magnitude of associated releases for five 83 ------- regulatory (containment) scenarios: a baseline scenario that assumes all tanks are managed and operated in compliance with existing RCRA Subtitle C regulations and four alternatives that are more stringent than the baseline scenario. As structured, the HWTF model is not meant for, nor is it applicable to, site-specific comparison of containment effectiveness. Rather it is intended for use in analyzing alternative regulatory strategies. A. 13 RCRA Risk-Cost Analysis Model The RCRA Risk-Cost Analysis Model is used by the EPA Office of Soild Waste to support the development of regulations under RCRA. The model produces relative risk and cost estimates for different management configurations of waste streams; waste transportation, treatment and disposal technologies; and environments. The model estimates human health, ecosystem and sensory risks from steady state releases of RCRA contaminants (and selected other contaminants) to ground water, surface water, and air. The model also calculates the costs of each technology in a management configuration as an annual revenue requirement. The model treats each management configuration in a generic fashion employing standard risk assessment methods (e.g., emissions estimates coupled with transport and fate models aligned with dose response models) and numerous simplifying assumptions (ICF Incorporated, 1984; Males, 1984). 84 ------- Containment is taken into account in the RCRA Risk-Cost Analysis Model in the Waste Disposal Technologies component of the model. This component contains six kinds of disposal technologies: landfills, surface impoundments, waste piles, land treatment, incineration, and deep well injection. To represent a wide range of designs, the model includes a number of different configurations for each kind of disposal technology. For landfills and surface impoundments, the configurations differ primarily in the type of liner system used, ranging from unlined to double liners. Except for the unlined configuration, all configurations include leachate collection, monitoring wells, and collection of surface run-off. A number of the configurations that employ liners comply with the RCRA Subtitle C requirements for landfills and surface impoundments. With regard to the other four disposal technologies, the various configurations in the model all comply with the RCRA Subtitle C regulations appropriate to the technology. In general, the model accounts for leachate migration to ground water by first estimating leachate generation rates over time and combining these with estimates of liner failure probabilities and saturated flow limitations imposed by underlying clays and soils. Synthetic liners are assumed to have a life of 35 years and to have a uniform failure rate over this period. When a synthetic liner fails, it is assumed to disintegrate (i.e., leak everywhere). Clay 85 ------- liners are assumed to retain their integrity and low permeability over the operating life of the waste disposal technology. Leachate collection and treatment systems are assumed to release a fixed portion of the leachate they handle. All leachate released by the system is assumed to be lost to the environment. The leachate is distributed among air, surface water, and ground water by first estimating the amount of constituents that evaporate (based on volatility), and then assuming that 70 percent of the remainder moves toward surface water and 30 percent moves toward ground water. Overland run-off to surface water is also assumed to occur from storm events exceeding the holding capacities of surface impoundments and/or storm water run-off management systems. Overtopping is assumed to have a constant probability of occurrence in any year. The model does not account for other overland run-off release mechanisms (e.g., leaks or dike failures). The model also does not account for the transport of contaminants by ground water to surface water. As structured, the RCRA Risk-Cost Analysis Model is not meant for, nor is it applicable to, site-specific comparisons of containment effectiveness. Rather it is intended for use in analyzing alternative regulatory strategies. 86 ------- APPENDIX B BIBLIOGRAPHY Arthur D. Little, Inc., Proposed Revisions to MITRE Model, Arthur D. Little, Inc., Cambridge, MA, September 23, 1981. Arthur D. Little, Inc., PERCO; A Model for Prioritization of Environmental Risks and Control Options at Hazardous Waste Sites, Arthur D. Little, Inc., Cambridge, MA, September 12, 1983. Barnthouse, L.W. et al., Development and Demonstration of a Hazard Assessment Rating Methodology for Phase II of the Installation Restoration Program, (ORNL/TM-9857), Oak Ridge National Laboratory, TN, 1986. Dames and Moore, Overview of Methodology for Rating the Potential for and Significance of Ground and Surface Water Containment from Waste Disposal Sites, Bethesda, MD, Undated. Engineering-Science, Comparison of U.S. Air Force hazard Assessment Rating Methodology (HARM) with U.S. Environmental Protection Agency Hazard Ranking System (HRS) at Four Air Force Bases Evaluated Under the Phase I Installation Restoration Program, Engineering-Science, Atlanta, GA, April 1983. French, Jean G. et al., A System for Prevention, Assessment, and Control of Exposures and Health Effects from Hazardous Sites (S.P.A.C.E. for Health), Centers for Disease Control, Atlanta, GA, January 1984. Gibb, J., Barcelona, M., Schock, S., and Hampton, M., Hazardous Waste in Ogle and Winnebago Counties; Potential Risk via Groundwater Due to Past and Present Activities, Illinois Department of Energy and Natural Resources, Document No. 83/26, Springfield, IL, 1983. ICF Incorporated, The RCRA Risk-Cost Analysis Model Phase III Report and Report Appendices, Washington, DC, 1984. ICF Incorporated and Pope-Reid Associates, Inc., Hazardous Waste Tanks Risk Analysis (Draft Report), prepared for the U.S. Environmental Protection Agency: Office of Soild Waste, Washington, DC, 1986. JRB Associates, Inc., Methodology for Rating the Hazard Potential of Waste Disposal Sites (Draft Final Report), McLean, VA, 1980. 87 ------- Kay, Robert L., Jr. and Chester L. Tate, Jr., "Public Health Significance of Hazardous Waste Sites," Proceedings of the Fifth National Conference on Management of Uncontrolled Hazardous Waste Sites, held on November 7-9, 1984 in Washington, DC, Hazardous Materials Control Research Institute, Silver Spring, MD, 1984, pp. 232-238. Kushner, L., Hazard Ranking System Issue Analysis; Superfund Sites With Unknown Waste Quantity; MTR-86W83, The MITRE Corporation, McLean, VA, 1986. Luecker, Elizabeth B., "Navy Assessment and Control of Installation Pollutant (NACIP) Confirmation Study Ranking Model," Proceedings of the Twelfth Annual Environmental Systems Symposium, held on May 20-21, 1982, at Langley Air Force Base, Langley, VA, American Defense Preparedness Association, Arlington, VA, 1982. Males, E., RCRA Risk-Cost Analysis Model, presented at the AICHE Conference held on August 21, 1984, U.S Environmental Protection Agency, Washington, DC, 1984. Michigan Department of Natural Resources, Site Assessment System (SAS) for the Michigan Priority Ranking System Under the Michigan Environmental Response Act (Act 307, P.A., 1982), Lansing, MI, 1983. Pope-Reid Associates, Inc., Hazardous Waste Tank Failure Model; Description of Methodology and Appendices A-E, prepared for the U.S. Environmental Protection Agency, Office of Soild Waste, Washington, DC, 1986. Sayala, D., Hazard Ranking System Issue Analysis; Subsurface Geochemical Processes (Draft Report), MIR-86W171, The MITRE Corporation, McLean, VA, 1986. U.S. Environmental Protection Agency, Liner Location Risk and Cost Analysis Model (Draft Report and Draft Appendices), Office of Solid Waste, Washington, DC, 1985. Wang, M., Hazard Ranking System Issue Analysis; Alternative Methods for Ranking the Persistence of Hazardous Substances in Surface Water (Draft Report), MTR-86W172, The MITRE Corporation, McLean, VA. 1986. Wolfinger, T., Hazard Ranking System Issue Analysis; Options for Revising the AirTathway (Draft Report), MTR-86W53, The MITRE Corporation, McLean, VA, 1986. Wolfinger, T., Hazard Ranking System Issue Analysis; The Use of Concentration Data (Draft Report), MTR-36W40, The MITRE Corporation, McLean, VA, 1986. 88 ------- |