UNITED STATES ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 t»CT ' OFFICE OF KEHORAKDUK SUBJECT: Transmittal of OSWER Directive /923fe.2-25: "Guidance for Evaluating the Technical ^praejtic^bility pf Ground- Water Restoration" FROM: Richard J. Guimond Assistant Surgeon GenirVl", Acting Assistant Administrator TO: Director, Waste Management Division Regions I, IV, V, VII Director, Emergency and Remedial Response Division Region II Director, Air and Waste Management Division Region II Director, Hazardous Waste Management Division Regions III, VI, VIII, IX Director, Hazardous Waste Division Region X Director, Environmental Services Division Regions I, VI, VII PURPOSE The purpose of this memorandum is to transmit the OSWER Directive on evaluating the technical impracticability of ground- water restoration. This guidance will apply to both the RCRA Corrective Action and Superfund programs. BACKGROUND Restoration of contaminated ground water is one of the primary objectives of both the Superfund and RCRA Corrective Action programs, as ground water contamination is present at over 85% of Superfunds National Priorities List (NPL) sites and a large number of RCRA facilities. While both programs have had a great deal of success reducing the immediate threats posed by contaminated ground water, experience over the past decade has shown that achieving the required final cleanup standards may not be practicable at some sites due to the limitations of remediation technology. The Technical Impracticability (TI) Workgroup was formed to address the issue of how to determine whether ground-water cleanup goals are technically achievable at TUT 007 1921 IMM SOS i*e«e<*« "»*' *65013* 65013 a particular site, and how to establish an alternative, protective cleanup strategy where restoration is determarad to be technically impracticable. The technical challenges to remediating contaminated ground water include many complex factors related to site hydrogeology and chemistry. One of the most difficult of these challenges is the problem presented by DNAPL (dense, nonaqueous phase liquid) contamination. DNAPLs include such diverse organic compounds as chlorinated solvents, PCBs, creosote, and certain pesticides. These compounds, which a recent EPA study indicates may be present as DNAPLs at up -to 60% of NPL sites, are often very difficult to locate and remove from the subsurface environment and may continue to contaminate ground water for many hundreds of years despite best efforts to remediate them. The prevalence and intractability of DNAPL contamination are among the principal reasons this guidance was developed by EPA. OBJECTIVE Use of this guidance will provide the basis for EPA to determine whether ground-water restoration is technically impracticable, as well as for establishing alternative remedial strategies at such sites. As this guidance applies to both RCRA and Super fund, it will also promote consistency between the two programs. Specifically, the guidance provides: • Clarification of the regulatory basis for TI determinations; • A recommended approach for the management and remediation of sites contaminated by DNAPLs; • Recommendations for the data and analyses necessary for TI evaluations; • Expectations and recommendations for establishing alternative remedial strategies where ground-water restoration is technically impracticable; and • A discussion of the TI evaluation and decision, process, and of other administrative issues. You will notice several important messages as you review the guidance that arexworth highlighting here: This guidance does not signify a scaling back of BPA's •fforts to restore contaminated ground water. Rather it- promotes a careful, technically sound approach to determine if critical limitations to ground-water restoration exist at a particular facility. Where ground-water restoration is technically impracticable EPA will select an alternative remedial strategy TUT 007 1922 that will ensure protection of human health and the environment. EPA may make Tl decisions either after a full-scale remedy has been implemented and operated for a period of time, or, in what is anticipated to be a smaller number of cases, before a final remedy decision document has been signed. Where the decision is made during the operation of a remedy, EPA expects the existing remedy to have been rigorously monitored, and modified or enhanced where appropriate to demonstrate that best efforts have been made to achieve the required cleanup levels. Where EPA is evaluating TI prior to remedy implementation, site characterization efforts must be especially thorough and must clearly and convincingly demonstrate that the attainment of cleanup levels is not practicable. The guidance promotes the use of a phased approach to site remediation, particularly where a moderate to high level of uncertainty exists regarding the potential outcome of restoration efforts. Early actions to control plume migration and remove contaminant sources are encouraged as part of this strategy. Such actions, where properly designed and monitored, can not only reduce risks posed by contaminated ground water, but also provide information useful in evaluating the restoration potential of the site. Use of language in final remedy decision documents which addresses the uncertainty in achieving reguired cleanup levels is appropriate in certain cases. However, language that identifies a TI decision (e.g., an ARAR waiver) as a future contingency of the remedy is discouraged. Such language (e.g., a contingency for an ARAR waiver) is not necessary, as a TI evaluation may be performed (and a decision made) by EPA at any site regardless of whether such a contingency is provided in the decision document. It is important to note that where such contingency language has been used in an existing decision document, the demonstration reguired to justify a TI decision should be consistent with that recommended in this guidance. As sites with extensive DNAPL contamination are aore likely to involve an evaluation of TI, the guidance also provides a brief discussion of DNAPL contamination, as veil as strategies for characterising and remediating such sites. The approach recommended includes locating and removing subsurface DNAPL sources where practicable and in general, where significant reduction of current* or future risk will result. DNAPL sources that cannot practicably be removed generally should be contained to limit further contamination of ground water. In such cases, remediation goals in the aqueous contaminant plume (as opposed to the DNAPL source areas) will depend on the effectiveness of source area containment and the particular circumstances of the site. Where practicable, EPA expects to restore those portions TUT 007 of the aqueous plume which lie outside of DNAPL containment areas to required cleanup levels. IMPLEMENTATION This guidance should be carefully considered by all staff involved in the management of Superfund sites or RCRA facilities with ground-water contamination; its content is relevant to all phases of ground-water cleanup. Further, copies of the guidance should be forwarded by Regional management to the appropriate personnel in State^ agencies. EPA should involve the appropriate State agencies as early in the TI evaluation and decision process as possible, since state ground-water resource management considerations (e.g., aquifer classification and wellhead protection areas) may heighten concern about TI decisions and the long-term management strategy chosen at sites where restoration is technically impracticable. TI evaluations will generally require specialized expert analysis of technical data. Thus, Regional decision makers are encouraged to involve technical personnel as part of a site review team as early as possible in the TI decision process. Technical assistance should be sought first from Regional staff hydrogeologists and engineers. However, Regional resources may be augmented through the Technical Support Project, by which scientists from the Office of Research and Development laboratories provide assistance for site-specific technical evaluations. Requests for support may be routed through a Regional ORD scientist (where available), or through the Technical Support Project manager in the appropriate laboratory. Additional support for site-specific reviews may also be obtained from Headquarters staff. Regional personnel with such requests or questions on the attached guidance should contact Peter Feldman at (703) 603-8768 (Superfund), or Guy Tomassoni at (703) 308-8622 (RCRA). Attachment cc: OERR Division Directors David Ziegele, OUST Walter Kovalick, TIO Bruce Diamond, OWPE James Makris, CEPPO Jeffery Denit, OSW Tim Fields, SRO Lisa Friedman, OGC William White, OE TUT OO7 Courtney Riordan, ORD Regional Super fund Branch and Sect:'on Chiefs Regional RCRA Branch and Section Chiefs TUT 007 1925 Directive 9234.2-25 September 1993 Guidance for Evaluating the Technical Impracticability of Ground-Water Restoration Interim Final Office of Solid Waste and Emergency Response U.S. Environmental Protection Agency Washington, DC 20460 • Notice The policies set out in this document are intended solely as guidance to U.S. Environmental Protection Agency (EPA) personnel; they are not final EPA actions and do not constitute rulemaldng. These policies are not intended, nor can they be relied upon, to create any rights enforceable by any party in litigation with the United States. EPA officials may decide to follow the guidance provided in this document, or to act at variance with the guidance, based on an analysis of specific site circumstances. EPA also reserves the right to change this guidance at any time with- out public notice. TUT O07 1927 -Contents 1.0 IntrOdUCtlOn i . I M . . I . . I . . . . . I I . I . . I I I . . . . . I .......... . . . . . l l I . . I J I I . . I . . L I I I I . ! . I M . . - L . . . I - I I I Ml. in -------- ——————————— - ——— --- 1 1.1 Background ........ ,,......,,..,..,,...•.. ,,...,....- ...i,....,,,,.,,.,..^^.,., _ ^^....^^.....^......^.......^^ ______ — -- _ — -• 1 12 Purpose of the Guidance ~~. — „ —— ............. — ..... ————— ................................................. ———— ....2 2.0 Ground Water Remedy Decision Framework ... — ,. — M — - —— „ — ; —————— . ——— .. —— ~~~........ — 2 2.1 Use of the Phased Approach „ — ».. —— .. — «.„.. _ ................................. — .....«»».....^._.-_..~«. — ..~2 23. Documenting Ground-Water Remedy Decisions Under CERCLA ——————— -~~« — ~ — . —— ........4 22.1 Removal Actions — « ——————— _____ ,. __ ....................... ——————— ... —— .....4 222 Interim RODs ....................... ..........M.............~......................~.^.....~~~..".""~™ -~.........~~...~..«.~....5 22.3 Final RODs ———————————— „„..„„».„., —— ..... — . ————— — ———————— - —— .5 22.4 ROD Contingency Remedies and Contingency Language ————— —— M ————— ~— ., — —.5 23 Documenting Ground-Water Remedy Decisions under RCRA~~.~.~~~~..~.~.~.«~~~ — » — .—. ———— ..—6 2.3.1 Permits/Orders Addressing Stabilization ___ ——————————————— ——— . —— .~~..6 232 Permits/Orders Addressing Final Remedies ———— „. ———— .. ——— . —— ••••«-• ——— . ——— 6 3.0 Remedial Strategy for DNAPL Sites——————;—„.,—...—...—...——————————..-.———————..6 4.0 TI Decisions and Supporting Information————..—————....————.....—————————.........—.~....9 4.1 Regulatory Framework for TT Decisions..___,~__.«.»».«_.—————————»——.——...-9 4.1.2 KCRA...........r^.....^..................1...............1..........1..^.t....................._....~_..........,........,,.,.,....,....,.,,, 10 42 Timing of TI Evaluations ~^.~—...——__.—..__...»——————*...-.——.—~...,—».——10 4.4 Supporting Information for TI Evaluations ...........—«...—__».^...~^™~.»—~~..~«.«~.^..~...~.~~«.» 11 4.4.1 Specific ARARs or Media Cleanup Standards.......__.._„.......,..............————...—————..... 12 4.42 Spatial Extent of TI Decisions....————....—»....-.~~^.v—~———.»—~~——...W..M.».... 12 4.43 Development and Purpose of the Site Conceptual Model....—...———————————————. 13 4.4.4 Evaluation of Restoration Potential -__,—.___,~~,___.———————————.»«——.13 4.4.4.1 Source Control Measures........—.......——............————.—————................................ 13 4.4.42 Remedial Action Performance Appraisal....—————————.~.~~~——...———»——16 4.4.43 Restoration Timeframe Analysis »——,—..........—————...———,~.———....—16 4.4.4.4 Other Applicable Technologies «.....««^».....«M........~^.^.n..».«.....~..~.^~.M»M.~.»..^.»»...».... 18 4.4.4.5 Additional Considerations ——«»-.-».«...-..«.-————„———————————————.-18 5.0 Alternative Remedial Strategies .. ——— „ — ,~~~ — .. ————— . — ~». ——— « ——— .. ————— . —— . 19 5.1 Options and Objectives for Alternative Strategies ....-.., —————————————— _ .„ ___ .......... 19 5.1.1 Exposure Control ~— ».~».»~.^._Mn_n»».n..»~-»«.««^......M».M«...~^.«».^.«.MWM^».M«.-.»......« 19 f « A P^.|..M>jt ^on^fnl * 10 5.13 Aqueous Plume Remediation ———— ... ————— .......... ————— ..« ______ Mm —— .20 52 Alternative Remedy SelectionMM«.n«M...M.».MM.M.»M.«MU.«»».M».n»«.MWMnM.MM.WMMM.n.MMM.n>.n.MM..M.«»21 52,1 Superfund-..-.~~~.M ———————————— .. —— «... —————— « — . ———— ~_ ———— .-21 523 Additional Remedy Selection Considerations ~_~~~~~, ————————— , __ ..M.MM..UMM. —— 22 52.4 Relation to Alternate Concentration Limits — ^^^.— .— ~«, ——— __ . _______ . _ ,...22 6.1 TI Review and Decision Process —————————— - — ~ —————— ————— _ — »„„ —— .23 O« 1 •» SUpCrftinCl mim •>•••••»• »»i*i«»»««««iim»«m»n»tt«»»«»««»«»»»*»»«**«»»««*ii*tii*»«««»«»i Ml »>•»*•» •>»•»»••!* »»»*•»» **Mi»»»»mn»»i«>««i •••»*•••»* 23 6.1 3 Technical Review and Support ————— .« _ ....-_.»,. — .. —————— »._ __ „ ___ —— 25 62 Duration of TI Decisions— — . ————————————————— .. —— .. ———— „«_ __ ..„.. __ ^....»25 References «.~.~.~~.~~—«~~.~—~»~~.—~~.«...M^«...«.—~~.—~~...™™.~—."~~.~........~~~W.~»...M.MM«.^M^.M— ...... .~..2o w TUT 007 1920 1.0 Introduction 1.1 Background Restoration1 of contaminated ground waters is one of the primary objectives of both die Superfund and RCRA Corrective Action programs. Ground-water contamination problems are pervasive in both pro- grams; over 85 percent of Superfund National Priori- ties List (NPL) sites and a substantial portion of RCRA facilities have some degree of ground-water contamination. The Superfund and RCRA Corrective Action programs share the common purposes of pro- tecting human health and the environment from con- taminated ground waters and restoring those waters to a quality consistent with their current, or reason- ably expected future, uses. The National Contingency Plan (NCP), which pro- vides the regulatory framework for the Superfund program, states that "EPA expects to return usable ground waters to their beneficial uses wherever practicable, within a timeframe that is reasonable given the particular circumstances of the site" (NCP §300.430(a)(l)(ui)(F)). Generally, restoration cleanup levels in the Superfund program are established by applicable or relevant and appropriate requirements (ARARs), such as the use of Federal or State standards for drinking water quality. Cleanup levels protective of human health and the en- vironment are identified by EPA where no ARARs for particular contaminants exist (see Section 4.1.1). The RCRA Corrective Action program for releases from solid waste management facilities (see 40 CFR. 264.101)2 requires a facility owner/operator to: "...institute corrective action as necessary to pro- tect human health and the environment for all releases of hazardous waste or constituents from any solid waste management unit-.." The goal of protectiveness is further clarified in the Preamble to the Proposed Subpart S to 40 CFR 264: "Potentially drinkable ground water would be cleaned up to levels safe for drinking throughout the contaminated plume, regardless of whemer the water was in fact being consumed... Alternative levels protective of the environment and safe for other uses could be established for ground water that is not an actual or reasonably expected source of drinking water.**^ While both programs have had a great deal of success reducing the immediate threats posed by contami- nated ground waters, experience over the past decade has shown that restoration to drinking water quality (or more stringent levels where required) may not al- ways be achievable due to the limitations of available remediation technologies (EPA 1989b, 1992d). EPA, therefore, must evaluate whether ground-water resto- ration at Superfund and RCRA ground-water cleanup sites is attainable from an engineering perspective. This document outlines EPA's approach to evalu- ating the technical impracticability of attaining re- quired ground-water cleanup levels and establish- ing alternative, protective remedial strategies where restoration is determined to be technically impracticable. Many factors can inhibit ground-water restoration. These factors may be grouped under three general categories: • Hydrogeologic factors; • Contaminant-related factors; and • Remediation system design inadequacies. Hydrogeologic limitations to aquifer remediation in- clude conditions such as complex sedimentary depos- its; aquifers of very tow permeability; certain types of For this guidance, "restoration" refers to me reduction of contaminant concentrations to levels required under me Superfund or RCRA Collective Action programs. For ground water currently or potentially used for drinking water purposes, these lev- els may be Maximum Contaminant Levels (MCLs) or non-zero Maximum Contaminant Levels Goals (MCLOs) established under the Safe Drinking Water Act; State MCLs or other cleanup requirements; or risk-based levels for compounds not cov- ered by specific State or Federal MCLs or MCLOs. Other cleanup levels may be appropriate for ground waters used for non- drinking water purposes. At this time, this guidance is not applicable to corrective actions for releases from Subpart F regulated units mat are subject to corrective actions under 40 CFR 264.91-264.100. "Corrective Action for Solid Waste Management Units (SWMUs) at Hazardous Waste Management Facilities," 55 FJR. 30798- 30884, July 27,1990, Proposed Rules, is currently used as guidance in the RCRA Corrective Action program. When final regulations under Subpart S are promulgated, certain aspects of this guidance pertaining to the RCRA program may need to be revised to reflect new regulatory requirements. TUT OO7 1929 fractured bedrock; and other conditions mat presently make extraction or in situ treatment of contaminated ground water extremely difficult (Figure 1). Contaminant-related factors, while not independent of hydrogeologic constraints, are more directly re- lated to contaminant properties that may limit the success of an extraction or in situ treatment process. These properties include a contaminant's potential to become either sorted onto, or lodged within, the soil or rock comprising the aquifer. Nonaqueous phase liquids (NAPLs) are examples of contaminants that may pose such technical limitations to aquifer resto- ration efforts. NAPLs that are denser than water (DN APLs) often are particularly difficult to locate and remove from the subsurface; their ability to sink through the water table and penetrate deeper portions of aquifers is one of the properties that makes them very difficult to remediate figure 1). The widespread use of DNAPLs in manufacturing and many other sectors of the economy prior to the advent of safe waste-management practices has led to their similarly widespread occurrence at ground-wa- ter contamination sites. Most of the sites where EPA already has determined that ground-water restoration is technically impracticable have DNAPLs present The potential impact of DN APL contamination on at- tainment of remediation goals is so significant that EPA is developing specific recommendations for DNAPL site management; the key elements of this strategy are presented in Section 3.0 below. The third factor that may limit ground-water restoration is inadequate remediation system design and imple- mentation. Examples of design inadequacies in a ground-water extraction system include an insufficient number of extraction points (e.g., ground water or va- por enaction wells) or wells whose locations, screened intervals, or pumping rates lead to an inability to capture the plume. Design inadequacies may result from incomplete site characterization, such as inaccu- rate measurement of hydraulic conductivity of the af- fected aquifer or not considering the presence of NAPL contamination. Poor remediation system operation, such as excessive downtime or failure to modify or enhance the system to improve performance, also may limit the effectiveness of restoration efforts. Failure to achieve desired cleanup standards re* suiting from inadequate system design or opera- tion is not considered by EPA to be a sufficient justification for a determination of technical im- practicability of ground-water cleanup. 1.2 Purpose of the Guidance This guidance clarifies how EPA will determine whether ground-water restoration is technically im- practicable and what alternative measures or actions must be undertaken to ensure that the final remedy is protective of human health and the environment Topics covered include the types of technical data and analyses needed to support EPA's evaluation of a particular site and the criteria used to make a determi- nation. As technical impracticability (TI) decisions are pan of the process of site investigation, remedy selec- tion, remedial action, and evaluation of remedy perfor- mance, the guidance also briefly discusses the overall framework for decision making during these phases of site cleanup. This guidance does not signal a scaling back of EPA's efforts to restore contaminated ground wa- ters at Superfund sites and RCRA facilities. Rather, EPA is promoting the careful and realistic as- sessment of the technical capabilities at hand to man- age risks posed by ground-water contamination. This guidance provides consistent guidelines for evaluat- ing technical impracticability and for maintaining protectiveness at sites where ground water cannot be restored within a reasonable timeframe. EPA will continue to conduct fund, and encourage research and development in the fields of subsurface assess- ment remediation, and pollution prevention so that an ever decreasing number of sites will require the analysis described in this document 2.0 Ground-Water Remedy Decision Framework 2.1 Use of the Phased Approach At sites with very complex ground-water contamina- tion problems, it may be difficult to determine whether required cleanup levels are achievable at the time a remedy selection decision must be made. This is especially true when such decisions must be based on site data collected prior to implementation and monitoring of pilot or full-scale remediation systems. EPA recognizes this limitation and has recommended several approaches to reduce uncertainty during the site characterization, remedy selection, and remedy implementation processes (EPA 1989a, 1992a). Determining the restoration potential of a site may be aided by employing a phased approach to site char- acterization and remediation. Each phase of site TUT O07 Ce big wi I 1 •! | a Figure 1. Ext man site characterisdcs may 1 hly generalized. TheparticuU 1 be site specific. Contaminant Characteristics Nature of Release tmples of Factors Affecting Ground-Water Restoration mit the effectiveness of subsurface remediation. The examples listed bebw are IT factor or combination of factors that may critically limit restoration potential Generalized Remediation Difficulty Scale Increasing difficulty . ^ ~ Small Volume Large Volume Short Duration —————————— ————— > Long Duration Slug Release Continual Release Biotic/Abiotic Decay Potential Volatility Contaminant Retardation (Sorption) Potential High ——————————— • ————————— >• Low High ———————————————————— *• Low Low ————————————————————— *» High Contaminant Phase Volume of Contaminated Media Contaminant Depth Aqueous, Gaseous — *• Sorted -> LNAPLs • — * DNAPLs Shallow —— - ———————————————— *- Deep * * Hydrogeologlc Characteristics Stratigraphy Texture of Unconsolidated Deposits Degree of Heterogeneity Simple Geology, ————— * Complex Geology, e.g., Planar Bedding e.g., Interbedded and Discontinuous Strata Sand —————————— * Clay Homogeneous —————— * Heterogeneous (e.g., interbedded sand and (e.g., well-sorted sand) silts, clays, fractured media, karst) Hydraulic conductivity Temporal Variation Vertical Flow High (>10* cm/sec) ——— * Low (< 104 cm/sec) Little/None ——————— > High Little —————————— * Large Downward Flow Component TUT 007 1931 1 characterization should be designed to provide infor- mation necessary for the next phase of characteriza- tion. Likewise, site remediation activities can be con- ducted in phases to achieve interim goals at the out- set, while developing a more accurate understanding of the restoration potential of the contaminated aqui- fer. An example of how mis approach might be ap- plied at a site is provided below in Section 4.4.3. The timing of phased cleanup actions (early, interim, final) should reflect the relative urgency of the action and die degree to which the site has been character- ized. Early actions should focus on reducing the risk posed by site contamination (e.g., removal of con- tamination sources) and may be carried out before de- tailed site characterization studies have been com- pleted. Interim remedial actions may abate the spread of contamination or limit exposure but do not fully address the final cleanup levels for the site. In- terim actions generally will require a greater degree of site characterization man early actions. However, implementation of interim actions still may be appro- priate prior to completion of site characterization studies, such as the Remedial Investigation/Feasibil- ity Study (RI/FS) or RCRA Facility Investigation (RFI) and Corrective Measures Study (CMS). Final remedial actions must address the cleanup levels and other remediation requirements for the site and, there- fore, must be based on completed characterization re- ports. Information from early and interim actions also should be factored into these reports and final remedy decisions. Phasing of activities generally should not delay or prolong site characterization or remediation. In fact, such an approach may accelerate die implementation of interim risk reduction actions and lead more quickly to the development of achievable final reme- diation levels and strategies. A phased approach should be considered when there is uncertainty re- garding die ultimate restoration potential of the site but also a need to quickly control risk of exposure to, or limit further migration of, die contamination. It is critical tiiat die performance of phased remedial actions (e.g., control of plume migration) be monitored carefully as pan of die ongoing effort to characterize die site and assess its restoration potential Data collec- tion activities during such actions not only should be designed to evaluate performance wim respect to die action's specific objectives but also contribute to die overall understanding of die site. Indus manner, actions implemented early in the site remediation process can achieve significant risk reduction and lead to development of technically sound, final rem- edy decisions. 2.2 Documenting Ground-Water Remedy Decisions Under CERCLA The phased approach to site characterization and remediation can be employed using die existing deci- sion document options widtin die Supermnd program. 22 J Removal Actions Removal authority can be used for early actions as pan of a phased approach to ground-water cleanup and decision making and should be considered where early response to ground-water contamination is advantageous or necessary. Within the context of ground-water actions, removals are appropriate where contamination poses an actual or potential tiveat to drinking water supplies or threatens sensi- tive ecosystems. Examples of actions tiiat might qualify for use of removal authority include removal of surface sources (e.g., drums or highly contami- nated soils), removal of subsurface sources (e.g., NAPL accumulations, highly contaminated soils, or other buried waste), and containment of migrating ground-water contamination "hot spots" (zones of high contaminant concentration) or plumes to protect current or potential drinking water supplies. Removals of subsurface sources most likely will be non-time-critical actions, although time-critical ac- tions may be appropriate for removal of NAPL ac- cumulations or other sources, depending on die ur- gency of die threat. Documentation requirements for removal actions include a Removal Action Memorandum and, for non-time critical actions, an Engineering Evaluation/Cost Analysis report.4 Removal actions must attain ARARs to die extent practicable, considering die exigencies of die situation. The urgency of die situation and die scope of die removal action may be considered when determining die practicability of attaining ARARs (NCP §300.41500). Standards or regulations typically used to establish ground-water cleanup levels for final actions (e.g., MCLs/MCLGs) may not be ARARs, depending on me scope of die removal. Further 4 See "Guidance on Conducting Non-Time Critical Removal Actions under CERCLA," OSWER Publication 9360.0-32, August 1993 (EPA 1993b). TUT 007 1932 information on removal actions may be found in other EPA guidances (EPA 1990b, 199ld). 2.2.2 Interim RODs Interim RODs may be appropriate where there is a moderate to high degree of uncertainty regarding at- tainment of ARARs or other protective cleanup lev- els. As mentioned before, an interim action may be used to minimize further contaminant migration and reduce the risk of exposure to contaminated ground water. Interim actions include containment of the leading edge of a plume to prevent further contami- nation of unaffected portions of an aquifer, removal of source material, remediation of ground-water hot spots, and in some cases, installation of physical barriers or caps to contain releases from source ma- terials. Interim actions should be monitored care- fully to collect detailed information regarding aqui- fer response to remediation, which should be used to augment and update previous site characterization efforts. This information then can be used at a later date to develop final remediation goals and cleanup levels that more accurately reflect the particular con- ditions of the site. It is important to note that for interim actions. ARARs must be attained only if they are within the scope of that action. For example, where an interim action will manage or contain migration of an aque- ous contaminant plume, MCLs and MCLGs would not be ARARs, since the objective of the action is containment, not cleanup (although requirements such as those related to discharge of the treated water still would be ARARs, since they address the disposi- tion of treated waste). Furthermore, a requirement that is an ARAR for an interim action may be waived under certain circum- stances. An "interim action" ARAR waiver may be invoked where an interim action that does not attain an ARAR is part of, or win be followed by, a final action that does (NCP §300.430(f)(l)(ii)(C)). For ex- ample, where an interim action seeks to reduce con- tamination levels in a ground-water hot spot, MCLs/ MCLGs may be ARARs since the action is cleaning up a portion of the contaminated ground water. If, however, this interim action is expected to be fol- lowed by a final, ARAR-comptiant action that ad- dresses the entire contaminated ground-water zone, an interim action ARAR waiver may be invoked. Final RODs Where site characterization is very thorough and there is a moderate to high degree of certainty that cleanup levels can be achieved, a final decision docu- ment should be developed mat adopts those levels. Conversely, in cases where Acre is a high degree of certainty that cleanup levels cannot be achieved, a final ROD that invokes a TI ARAR waiver and establishes an alternative remedial strategy may be the most appro- priate option.5 Note mat for ROD-stage waivers, site characterization generally should be sufficiently de- tailed to address the data and analysis requirements for TI determinations set forth in this guidance. 22.4 ROD Contingency Remedies and Contingency Language Where a moderate degree of uncertainty exists re- garding the ability to achieve cleanup levels, a final ARAR-compliant ROD generally still is appropriate. However, the ROD may include contingency lan- guage that addresses actions to be taken in the event the selected remedy is unable to achieve the required cleanup levels (EPA 1990a, 1991a). The contingency language may include requirements to enhance or augment the planned remediation system as well as an alternative remedial technology to be employed if modifications to the planned system foil to signifi- cantly improve its performance. Use of language in final remedy decision documents that addresses the uncertainty in achieving required cleanup levels also is appropriate in certain cases. However, language that identifies a TI decision (e.gn an ARAR . waiver) as a future contingency of the remedy should be avoided. Such language is not necessary, as a TI evaluation may be performed (and a decision made) by EPA at any site regardless of whether such a contingency is provided in the decision document Note that in cases of existing RODs that already include a contingency for invoking a TI ARAR waiver, the conditions under which the ARAR may be waived should be consistent with, and as stringent as, those presented in this guidance or a future update. Furthermore, the fact that such contingency lan- guage has been included in an existing ROD does not alter the need to enhance or augment a rem- edy to improve its ability to attain ARARs before concluding that a waiver can be granted. It also 5 At sites where a TI ARAR waiver is invoked in the ROD, preparation of the pre-referral negotiation package ("mini-lit" pack- age) must include analysis of die model Consent Degree language to ensure mat appropriate consideration of the waiver's im- pact is incorporated. TUT OO7 should be noted that remediation must be conducted for a sufficient period of time before its ability to re- store contaminated ground water can be evaluated. This minimum time period will be determined by EPA on a site-specific basis. 2.3 Documenting Ground-Water Remedy Decisions under RCRA The instruments used for implementing the RCRA Corrective Action program (permits and orders) also are amenable to a phased approach to remedy selec- tion and facility remediation. The RCRA program can use permits or orders to compel both interim measures and final remedies. 23.1. Permits/Orders Addressing Stabilization RCRA permits or orders can require the stabilization of releases from solid waste management units (SWMUs) at the facility. The Stabilization Initiative focuses on taking interim actions to prevent the fur- ther spread of existing contamination and reduce risks. Examples of measures used for stabilization include capping, excavation, and plume containment. Since the long-term or final cleanup of the facility is not the objective of stabilization (although stabiliza- tion should be consistent with the final remedy), TI decisions are not applicable at this early stage. Infor- mation gained during stabilization should be used to help determine the restoration potential of the facility and the objectives of the final remedy. 232. Permits/Orders Addressing Final Remedies Where achieving ground-water cleanup standards is determined by EPA to be technically impracticable, the permit or order addressing final remedies should include practicable and protective alternative reme- dial measures. EPA's decision to make a TI determi- nation will be based on clear and convincing infor- mation provided by the owner/operator. EPA gener- ally will seek public comment on TI determinations prior to implementation. EPA's preliminary TI deter- minations and justification for these determinations should be documented in a Statement of Basis. As discussed above, uncertainty in the ability to restore an aquifer should be reduced through phased charac- terization and the use of interim remedial measures, where appropriate. ">« Permits and orders that address "final" remedies should specify the remediation cleanup levels selected by the implementing Agency. Such permits and orders, how- ever, generally should not incorporate contingency TI language. The permit or order will need to be modified to document the TI determination and to specify, as appropriate, alternative cleanup levels and alternative remedial measures mat have been determined to be technically practicable and protective of human health and the environment. 3.0 Remedial Strategy for DNAPL Sites Many of the subsurface contaminants present at Su- perfund sites and RCRA facilities are organic com- pounds mat are either lighter-than-water NAPLs (LNAPLs) or DNAPLs. As mentioned in Section 1.1, the presence of N APL contamination, and in particu- lar DNAPL contamination, may have a significant impact on site investigations and the ability to restore contaminated portions of the subsurface to required cleanup levels. Furthermore, DNAPL contamination may be a relatively widespread problem. A recent EPA study (EPA 1993a) concluded that up to 60 per- cent of National Priorities List (NPL) sites may have DNAPL contamination in the subsurface; a signifi- cant percentage of RCRA Corrective Action facilities also are thought to be affected by DNAPLs. As proven technologies for the removal of certain types of DNAPL contamination do not exist yet, DNAPL sites are more likely to require TI evaluations than sites with other types of contamination. Although this guidance pertains to TI evaluations at all site types, EPA believes the significance of the DNAPL contamination problem warrants the following brief discussion of DNAPL contamination and recom- mended site management strategies. DNAPLs comprise a broad class of compounds, in- cluding creosote and coal tars, polychlorinated biphe- nyls (PCBs), certain pesticides, and chlorinated or- ganic solvents such as trichloroethylene (TCE) and tetrachloroethylene (PCE). The term "DNAPL" re- fers only to liquids immiscible in, and denser man, water and not to chemicals that are dissolved in water that originally may have been derived from a DNAPL source. DNAPLs may occur as "free-phase" or "re- sidual" contamination. Free-phase DNAPL is an im- miscible liquid in'the subsurface that is under positive pressure; mat is, the DNAPL is capable of flowing into a well or migrating laterally or vertically through an aquifer. Where vertically migrating free-phase DNAPL encounters a rock or soil layer of relatively low permeability (e.g., clay or other fine-grained layer), a DNAPL accumulation or "pool" may form. Residual DNAPL is immiscible liquid held by capillary forces TUT OO7 1934 within the pores or fractures in soil or rock layers; residual DNAPL, therefore, generally is not capable of migrating or being displaced by normal ground- water flow. Both free-phase and residual DNAPL, however, can slowly dissolve in ground water and produce "plumes" of aqueous-phase contamination. DNAPLs also can produce subsurface vapors capable of migrating through the unsaturated zone and con- taminating ground water (EPA 1992c). Figure 2 de- picts the various types of contamination that may be encountered at a DNAPL site. The three areas mat should be delineated at a DNAPL site are the DNAPL entry location, the DNAPL zone, and the aqueous contaminant plume. The entry locations are those areas where DNAPL was released and likely is present in the subsurface. Entry locations include waste disposal lagoons, drum burial sites, or any other area where DNAPL was al- lowed to infiltrate into the subsurface. The DNAPL zone is defined by that portion of the subsurface con- taining free-phase or residual DNAPL. Thus, the DNAPL zone includes all portions of the subsurface where the immiscible-phase contamination has come to be located. The DNAPL zone may occur within both the saturated zone (below the water table) and the unsaturated zone (above the water table). The DNAPL zone also may contain vapor and aqueous- phase contamination derived from the DNAPL. The DNAPL zone may include areas at relatively great depths and lateral distances from the entry locations, depending on the subsurface geology and the volume of DNAPL released. The aqueous contaminant plume contains organic chemicals in the dissolved phase. The plume originates from the DNAPL zone and may extend hundreds or thousands of feet downgradient (in the direction of ground-water flow). Figure 3 illustrates the various components of a DNAPL site. Since each DNAPL site component may require a different remediation strategy,it is important to char- acterize these components to the extent practicable. Thus, the properties and behavior of DNAPL con- tamination require consideration when planning and conducting both site investigation and remediation. The potential for DNAPL occurrence at the site should be evaluated as early as possible in the site in- vestigation. Recent publications such as "Estimating Potential for DNAPL Occurrence at Superfund Sites" (EPA 1992c) and "DNAPL Site Evaluation" (Cohen and Mercer, 1993) provide detailed guidance on these topics. At sites where DNAPL disposal is known or suspected to have occurred, likely DNAPL entry locations should be identified from available historical waste-management information and sub- surface chemistry data. This information can assist in the delineation of the DNAPL zone. Characterization and delineation of the DNAPL zone is critical for remedy design and evaluation of the restoration potential of the site. At many sites, a sub- surface investigation strategy that begins outside of the suspected DNAPL zone may be appropriate ("outside-in" strategy), in pan to minimize the possi- bility of inadvertent mobilization of DNAPLs to Figure 2. Types of Contamination and Contaminant Zones at DNAPL Sites (Cross-sectional view) AqiMOua Contaminant Plum* TUT O07 1935 Figure 3. Components of DNAPL SHes ONAPLZem contain* frw-phaM DNAPL In pools or Itnm and/or nMldual DNAPL V ONAPL Entry Location such as a tomw watt* pond Greund-WattfFtow lower aquifers. Delineation of the extent of the DNAPL zone may be difficult at certain sites due to complex geology or waste disposal practices. In such cases, the extent of the DNAPL zone may need to be inferred from geologic information (e.g., thickness, extent, structure, and permeability of soil or rock units) or from interpretation of the aqueous concen- tration of contaminants derived from DNAPL sources. At some sites, however, geologic complex- ity and inadequate information on waste disposal may make the delineation of the DNAPL zone difficult. A phased approach, as discussed in Section 2.1, is recommended for DNAPL sites; such an approach may facilitate identification of appropriate short- and long-term site remediation objectives. Note also that technical approaches appropriate for the DNAPL zone (e.g., tree-phase DNAPL removal, vapor extrac- tion, excavation, and slurry walls aided by limited pump-and-treat) may differ significantly from those appropriate for the aqueous contaminant plume (typi- cally pump-and-treat). Short-term remediation objectives generally should include prevention of exposure to contaminated ground water and containment of the aqueous con- taminant plume. Where sufficient information is available, early removal of DNAPL sources also is recommended. Information gathered during these actions should be used to help characterize the site and identify practicable options for further remediation. The long-term remediation objectives for a DNAPL zone should be to remove the free-phase, residual, and vapor phase DNAPL to the extent practicable and contain DNAPL sources that cannot be removed. EPA recognizes that it may be difficult to locate and remove all of die subsurface DNAPL within a DNAPL zone. Removal of DNAPL mass should be pursued wherever practicable and, in general, where significant reduction of current or future risk will re- sult.6 Where it is technically impracticable to remove subsurface DNAPLs, EPA expects to contain the DNAPL zone to minimize further release of contami- nants to the surrounding ground water, wherever practicable.7 Where it is technically practicable to contain the long-term sources of contamination, such as the DNAPL zone, EPA expects to restore the aqueous contaminant plume outside the DNAPL zone to re- quired cleanup levels. Effective containment of the DNAPL zone generally will be required to achieve this long-term objective because ground-water ex- traction remedies and analyses that support any assertion that attainment of ARARs or media cleanup standards is technically impracticable from an engineering perspective (See Section 4.4.4). At a minimum, this generally should jichtfc* a. A demonstralkm uto combination sources have been identified and have been, or will be, removed and contained to the extent practicable; b. An analysis of the performance of any ongo- ing or completed remedial actions; c. Predictive analyses of the timeframes to attain required cleanup levels using available tech- nologies; and d. A demonstration that no other remedial tech- nologies (conventional or innovative) could reliably, logically, or feasibly attain the cleanup levels at the site within a reasonable tiraeframe. 5. Estimates of the cost of tiie existing or pro- posed remedy options, including construction, operation, and maintenance costs (See Section 4.4.5). 6. Any additional information or analyses that EPA deems necessary for the TI evaluation. The data and analyses needed to address each of these components of a TI evaluation should be de- termined on a site-specific basis. Where outside parties are preparing the TI evaluation, its contents generally should be identified and discussed prior to submittal of the evaluation to EPA. Early agreement between EPA and PRPs or owner/operators on the type and quantity of data and analyses required for TI deci- sions will promote efficient review of TI evaluations. References to other documents in the administrative record, such as theRI/FS and RFI, likely will be nec- essary to produce a concise evaluation; however, these references should be as explicit as possible (e.g., cite specific page or table numbers). Technical discussions and conclusions should be supported by data compilations, statistical analyses, or other types of data reduction included in the evaluation. 4.4 Supporting Information for TI Evaluations Most, if not all, of the information needed to evaluate TI could be obtained during a thorough site investiga- tion and, where appropriate, remedy performance monitoring efforts. At some sites, however, addi- tional analysis of existing data or new information may be required before EPA can determine accu- rately me technical practicability of me restoration goals. Not all of the data or analyses outlined in this guidance will be required at all sites; specific infor- mation needs will depend on site conditions and any ongoing remediation efforts. 12 For this guidance a TI evaluation" comprises the data and analyses necessary to make a TI determination. The TI evaluation may be performed by PRPs at enforcement-lead Superfund sites, or by State or other Federal agencies, where appropriate. Similarly, owner/operators at RCRA facilities may perform TI evaluations. However, me actual TI "determination," or "deci- sion.'' will be made by EPA (or other lead agency, as appropriate). 11 1939 The data and analyses identified and discussed below address the TI evaluation components provided in Section 43. 4.4.1. Specific ARARs or Media Cleanup Standards The TI evaluation should identify die specific ARARs or media cleanup standards 0.e., the specific contaminants) for which the determination is sought. Such contaminants generally should include only those for which attainment of the required cleanup levels is technically impracticable. Factors EPA will consider when evaluating contaminants that may be included in the TI decision include: 1) the technical feasibility of restoring some of the con- taminants present in the ground water; and 2) the potential advantages of attaining cleanup levels for some of the contaminants. For example, consider a Superfund site with a DNAPL contamination problem (e.g., TOE), including a wide- spread subsurface DNAPL source area for which con- tainment or restoration are technically impracticable. The aqueous plume also contains inorganic contamina- tion (e-g., chromium) from on-site sources. Although it would be feasible to reduce chromium concentrations to the required cleanup level within a reasonable time- frame, TCE concentrations would remain above cleanup levels much longer due to the continued pres- ence of the DNAPL or slow desorption of TCE from aquifer materials. However, in such cases, EPA may choose to limit the TI ARAR waiver to TCE alone, while requiring cleanup of the chromium.13 Two situations would favor use of this approach. The first would be where attaining chromium cleanup levels in the ground water will make future ex silu treatment of the (TCE-contaminated) ground water less complex and less expensive. This may be advan- tageous where a community wishes to extract the TCE-contaminated water, perform ex situ treatment, and put the treated water to beneficial use. A related consideration is whether removal of the chromium will facilitate future subsurface remediation using a newly developed technology. The second situation favoring this approach is where one of the contami- nants (e.g., TCE) is being naturally biodegraded and the other (e.g., chromium) is not Therefore, cleanup of the chromium may result in more rapid attainment of the long-term cleanup goals at the site. -Where the balance of conditions at such a site do not indicate that it is practicable to attain the cleanup levels for only some of the contaminants present, EPA may conclude that cleanup levels for the re- maining contaminants need not be attained, depend- ing on the circumstances of the site. As discussed further in Section 5.0, however, this decision does not preclude EPA from selecting (or continuing op- eration of) a remedy that includes active measures (e.g., pump-and-treat) along with measures to pre- vent exposure (e.g., institutional controls) needed to address site risks. 4,42 Spatial Extent of TI Decisions The TI evaluation should specify the horizontal and vertical extent of the area for which the TI determina- tion is sought Where EPA determines mat ground- water restoration is technically impracticable, the area over which the decision applies (the "TI zone") generally will include all portions of the contami- nated ground water that do not meet the required cleanup levels (contaminated ground-water zone), un- less the TI zone is otherwise defined by EPA. In certain cases, EPA may restrict the extent of the TI zone to a portion or subarea within the contami- nated ground-water zone. For example, consider a DNAPL site where it is technically impracticable to remove the residual DNAPLs from the subsurface but it is feasible and practicable to: 1) limit further migration of contaminated ground-water using a containment system; and 2) restore that portion of the aqueous plume outside of the containment area. The TI zone in mis case should be restricted to that portion of the site that lies within the containment area. Outside of the TI zone, ARARs or media cleanup standards still would apply. The potential to spatially restrict the TI zone, therefore, will de- pend on the ability to delineate and contain non-re- movable subsurface contamination sources and re- store those portions of the aqueous plume outside of the containment area. The spatial extent of the TI zone should be limited to as small an area as pos- sible, given the circumstances of the site. A TI zone should.be delineated spatially, both in area and depth. Depth of a TI zone may be defined in ab- solute terms (e.g., feet above mean sea level) or in relative terms (e.g., with respect to various aquifers within multi-aquifer systems), as appropriate. Where 13 The extracted ground water would likely need to be treated for both TCE and chromium to satisfy treatment and waste dis- posal ARARs. 12 TIT, the TI zone will be restricted to a portion of the con- taminated ground-water zone, the limits of the TI zone should be delineated clearly on site maps and geologic cross-sections. Delineation of the TI zone based on the location of a particular mapped contami- nant concentration contour interval (e.g., the 200 pan per billion isoconcentration line) generally should be avoided. This is because the location of such mapped contours often is highly interpretive, and their posi- tion may change with time. While concentration data may be appropriate to consider when determining the size of a containment area or the extent of a TI zone, the limits of that TI zone should be fixed in space, both horizontally and vertically. 4.43 Development and Purpose of the Site Conceptual Model Decisions regarding the technical practicability of ground-water restoration must be based on a thor- ough characterization of the physical and chemical aspects of the site. Characterization data should de- scribe site geology and hydrology; contamination sources, properties, and distribution; release mecha- nisms and rates; fate and transport processes; current or potential receptors; and other elements that define the contamination problem and facilitate analysis of site restoration potential. While the elements of such a model may vary from site to site, some generaliza- tions can be made about what such a model would contain. Examples of these elements are provided in Figure 4. The site conceptual model synthesizes data acquired from historical research, site characteriza- tion, and remediation system operation. The site conceptual model typically is presented as a summary or specific component of a site investigation report. The model is based on, and should be sup- ported by, interpretive graphics, reduced and analyzed