Directive 9283.1-12EPA 540/R-96/023

PB96-963508October 1996

PRESUMPTIVE RESPONSE STRATEGY AND EX-SITU TREATMENTTECHNOLOGIES FOR CONTAMINATED GROUND WATER

AT CERCLA SITES

FINAL GUIDANCE

Office of Solid Waste and Emergency ResponseU.S. Environmental Protection Agency

Washington, DC 20460

901290

NOTICE

This document provides guidance to EPA staff. It also provides guidance to the public and to theregulated community on how EPA intends to exercise its discretion in implementing the NationalContingency Plan. The guidance is designed to implement national policy on these issues. Thedocument does not, however, substitute for EPA's statutes or regulations, nor Is it a regulationitself. Thus, it cannot impose legally-binding requirements on EPA, States, or the regulatedcommunity, and may not apply to a particular situation based upon the circumstances. EPA maychange this guidance in the future, as appropriate.

CONTENTS

F I G U R E S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

HIGHLIGHTS .......................................................................... iii

ACRONYMS USED IN THIS GUIDANCE .................................................. iv

PREFACE ............................................................................. v

1.0 INTRODUCTION ................................................................... 11.1 Purpose of Guidance ........................................................... 11.2 Expectations and Objectives for Ground-Water Cleanup .............................. 2

1.2.1 Program Expectations .................................................. 21.2.2 Objectives for Site Response Actions ...................................... 2

1.3 Lessons Learned .............................................................. 31.3.1 Sources and Types of Contaminants ....................................... 31.3.2 Factors Limiting Restoration Potential ..................................... 31.3.3 Assessing Restoration Potential .......................................... 5

2.0 PRESUMPTIVE RESPONSE STRATEGY ............................................... 52.1 Definition and Basis for Strategy ................................................. 5

2.1.1 Benefits of Phased Approach ............................................ 62.1.2 Early Actions ......................................................... 62.1.3 Monitoring........................................................... 8

2.2 Phased Response Actions ....................................................... 82.2.1 Two Separate Actions .................................................. 82.2 2 Phasing of a Single Action ................................'.............. 8

2.3 Post-Construction Refinements ................................................. 112.3.1 Types of Refinements ................................................. 112.3.2 Documenting Refinements ............................................. 11

2.4 Integrating Response Actions ................................................... 122.4.1 Integrating Source Control and Ground-Water Actions ....................... 122.4.2 Combining Ground-Water Restoration Methods ............................ 12

. 2.5 Strategy for DNAPL Sites ..................................................... 132.5.1 Site Characterization .................................................. 142.5.2 Early Actions ........................................................ 142.5.3 Long-Term Remedy .................................................. 14

2.6 Areas of Flexibility in Cleanup Approach .......................................... 152.6.1 Beneficial Uses and ARARs ............................................ 152.6.2 Remediation Timeframe ............................................... 162.6.3 Technical Impracticability .............................................. 172.6.4 Point of Compliance .................................................. 172.6.5 Natural Attenuation ................................................... 182.6.6 Alternate Concentration Limits .......................................... 18

3.0 PRESUMPTIVE TECHNOLOGIES .................................................... 193.1 Presumptive Technologies for Ex-Situ Treatment ................................... 19

3.1.1 Design Styles within Presumptive Technologies ............................ 203.1.2 Benefits of Presumptive Technologies .................................... 203.1.3 Consideration of Innovative Technologies ................................. 20

3.2 Basis for Presumptive Technologies ............................................. 213.2.1 Sources of Information ................................................ 213.2.2 Rationale for mdentifying Presumptive Technologies ........................ 21

3.3 Remedy Selection Using Presumptive Technologies ................................. 223.3.1 Use of Technologies in Treatment Systems ................................ 223.3.2 This Guidance Constitutes the FS Screening Step ........................... 233.3.3 Deferral of Final Technology Selection to RD .............................. 23

3.4 Information Needed for Selecting Technologies .................................... 243.4.1 When Should this Information be Collected? .............................. 243.4.2 Extraction Flow Rate ................................................. 253.4.3 Discharge Options and ARARs ......................................... 263.4.4 WaterQuality of Treatment Influent ..................................... 263.4.5 Treatability Studies ................................................... 26

3.5 Treatment Technologies for Aquifer Tests ......................................... 273.5.1 Treatment Needs during Aquifer Tests .................................... 273.5.2 Treatment Technologies for Aquifer Tests ................................. 27

4.0. REFERENCES .................................................................... 28

APPENDICESA. Additional Background Information

Al Background on DNAPL Contamination ...................................... A-2A2 Contaminants Most Frequently Reported in Ground Water at CERCLA NPL Sites .... A-4A3 Examples ofhi-Situ Treatment Technologies .................................. A-6A4 Definition and Discussion of Pulsed Pumping .................................. A-8

B. ROD Language Examples For Selected RemedyBl Phased Implementation of Ground-Water Remedy .............................. B-lB2 Phased Implementation of Extraction Component of Remedy at a DNAPL Site ....... B-3B3 Deferring Selection of Treatment Components to Remedial Design ................. B-5B4 Suggested ROD Language &om 1990 OSWER Directive ........................ B-7

C. Ex-Situ Treatment Technologies for Ground WaterCl Ex-Situ Technologies Considered in Sample of 25 Sites ......................... C-lC2 Other Components Needed for Treatment Trains ............................... C-3C3 Information Needed for Selection of Technologies and Design of Treatment Train ..... C-4C4 Advantages and Limitations of Presumptive Treatment Technologies ............... C-9

D. Descriptions of Presumptive Treatment TechnologiesDl Air Stripping ............................................................ D-lD2 Granular Activated Carbon ................................................ D-3D3 Chemical/UV Oxidation ................................................... D-4D4 Aerobic Biological Reactors ................................................ D-7D5 Chemical Precipitation .................................................... D-9D6 Ion Exchange/Adsorption ................................................. D-l 1D7 Electrochemical Methods ................................................. D-13D8 Aeration of Background Metals ............................................ D-15

FIGURES

Figure Page

1 Examples of Factors Affecting Ground-Water Restoration Potential ...................... 4

2 Phased Ground-Water Actions: Early Action Followed by Long-Term Remedy ............. 9

3 Phased Ground-Water Actions: Long-Term Remedy Implemented in Phases .............. 10

Al-1 Components ofDNAPL Sites ................................................... A-2

Al-2 Types of Contamination and Contaminant Zones ofDNAPL Sites (Cross-Section) ........ A-2

HIGHLIGHTS

Highlight Page

1 Presumptive Response Strategy ................................................... 6

2 Early Actions mat Should be Considered ............................................ 7

3 Remedy Refinements for Extraction/Treatment Remedies ............................. 12

4 Presumptive Technologies for Treatment of Extracted Ground Water .................... 20

5 Summary of Site Information Needed for Treatment Train Design. ...................... 25

in

ACRONYMS USED IN THIS GUIDANCE

ACL Alternate Concentration Limit NPL

ARAR Applicable or Relevant and OERRAppropriate Requirement

CERCLA Comprehensive Environmental ORDResponse, Compensation, andLiability Act of 1980, as amended by OSWERSARA

CERI Center for Environmental Research PCBInformation

CFR Code of Federal Regulations POTW

CSGWPP Comprehensive State Ground Water RARAProtection Program

DNAPL Dense Nonaqueous Phase Liquids RD

EPA Environmental Protection Agency RD/RA

ESD Explanation of Significant RIDifferences

RI/FSFS Feasibility Study

GAC Granular Activated Carbon ROD

LNAPL Light Nonaqueous Phase Liquids SACM

MCL Maximum Contaminant LevelSARA

MCLG Maximum Contaminant Level Goal

NAPL Nonaqueous Phase Liquid UV

NCP National Oil and Hazardous VOCSubstances Pollution ContingencyPlan

National Priorities List

Office of Emergency and RemedialResponse

Office of Research and Development

Office of Solid Waste and EmergencyResponse

Polychlorinated BiphenylCompounds

Publicly Owned Treatment Works

Resource Conservation and RecoveryAct

Remedial Design

Remedial Design/Remedial Action

Remedial Investigation

Remedial Investigation/FeasibilityStudy

Record of Decision

Superiund Accelerated CleanupModel

Superfund Amendments andReauthorizanon Act of 1986

Ultra Violet (light)

Volatile Organic Compound

IV

PREFACE

Presumptive Remedies Initiative. The objective of the presumptive remedies initiative is to use meSuperfund program's past experience to streamline site investigations and speed up selection of cleanupactions. Presumptive remedies are expected to increase consistency in remedy selection and implementation,and reduce the cost and time required to clean up similar types of sites. The presumptive remedies approachis one tool within the Superfund Accelerated Cleanup Model (SACM) (EPA, 1992d).

Presumptive remedies are preferred technologies for common categories of sites, based on historicalpatterns of remedy selection and EPA's scientific and engineering evaluation of performance data ontechnology implementation. Refer to EPA Directive, Presumptive Remedies: Policy and Procedures (EPA,1993d) for general information on the presumptive remedy process and issues common to all presumptiveremedies. This directive should be reviewed before utilizing a presumptive remedy and for furtherinformation on EPA expectations concerning the use of presumptive remedies. "Presumptive remediesare expected to be used at all appropriate sites," except under unusual site-specific circumstances (EPA,1993d).

Other Presumptive Remedy Guidance. Previous fact sheets from EPA's Office of Solid Waste andEmergency Response (OSWER) have established presumptive remedies for municipal landfill sites (EPA,1993f), for sites with volatile organic compounds in soils (EPA, 1993e) and for wood treater sites (EPA,1995g). A presumptive response selection strategy for manufactured gas plant sites is under development.Additional fact sheets are in progress for sites contaminated with polychlorinated biphenyl compounds(PCBs), metals in soils and for grain storage sites.

Relation of this Guidance to Other Presumptive Remedies. The fact sheets mentioned above providepresumptive remedies (or a strategy for selecting remedies) for "source control" at specific types of sites.With respect to ground-water response, source control refers to containment or treatment of materials matmay leach contaminants to ground water, or a combination of these approaches. In general, treatment isexpected for materials comprising (he principal threats posed by a site, while containment is preferred for lowlevel threats (EPA, 199 Ic). Where contaminants have reached ground water and pose an unacceptable risk tohuman health or the environment, a ground-water remedy will generally be required in addition to the sourcecontrol remedy and this guidance should be consulted.

Instead of establishing one or more presumptive remedies, this guidance defines a presumptive responsestrategy. EPA expects that some elements of this strategy will be appropriate for all sites with contaminatedground water and all elements of the strategy will be appropriate for many of these sites. In addition, thisguidance identifies presumptive technologies for me ex-situ treatment component of a ground-waterremedy, mat are expected to be used for sites where extraction and treatment is part of the remedy. (The termpresumptive technology is used in this guidance to denote only the ex-situ treatment component of a ground-water remedy.) Other remedy components could include methods for extracting ground water, enhancingcontaminant recovery or degradation of contaminants in me subsurface, discharging treated water, preventingcontaminant migration, and institutional or engineering controls to prevent exposure to contaminants.

Applicability to RCRA Corrective Action Program. EPA continues to seek consistency between cleanupprograms, especially in me process of selecting response actions for sites regulated under the ComprehensiveEnvironmental Response, Compensation and Liability Act (CERCLA or Superfund program) and correctivemeasures for facilities regulated under me Resource Conservation and Recovery Act (RCRA). In general,

even though the Agency's presumptive remedy guidances were developed for CERCLA sites, they shouldalso be used at RCRA Corrective Action sites to focus RCRA Facility Investigations, simplify evaluation ofremedial alternatives in the Corrective Measures Study, and influence remedy selection in the Statement ofBasis. For more information refer to the RCRA Corrective Action Plan (EPA, 1994c), the proposed SubpartS regulations (Federal Register, 1990b), and the May 1,1996 RCRA Corrective Action Advance Notice ofProposed Rulemaking (Federal Register, 1996).

Use of this Guidance. The presumptive response strategy, described in Section 2.1, integrates sitecharacterization, early actions, remedy selection, performance monitoring, remedial design and remedyimplementation activities into a comprehensive, overall response strategy for sites with contaminated groundwater. By integrating these response activities, the presumptive strategy illustrates how the SuperfundAccelerated Cleanup Model (SACM) can be applied to ground-water cleanup. Although mis responsestrategy will not necessarily streamline the remedial investigation/feasibility study (RI/FS) phase, EPAexpects that use of the presumptive strategy will result in significant time and cost savings for the overallresponse to contaminated ground water. By providing a mechanism for selecting achievable remediationobjectives, me presumptive strategy will minimize the need for changing these objectives during remedyimplementation. By optimizing the remedy for actual site conditions during implementation, me effectivenessof the selected remedy can be greatly increased, which will reduce the time and cost required to achieveremediation objectives.

The presumptive technologies for treating extracted ground water, identified in Section 3.1, are thetechnologies that should generally be retained for further consideration in me Detailed Analysis portion of mefeasibility study (or in the remedial design as explained in Section 3.3.3). This guidance and its associatedAdministrative Record will generally constitute the Development and Screening of Alternatives portion of mefeasibility study (FS) for me ex-situ treatment component of a ground-water remedy (see Section 3.3.2). Inthis respect, the presumptive technologies will streamline the FS for this component of a ground-waterremedy in the same way (hat other "presumptive remedies" streamline the FS for me overall remedy for theirrespective site types (see EPA, 1993d).

VI

1.0 INTRODUCTION

In implementing me Superiund and otherremediation programs, cleanup of contaminatedground water has proven to be more difficult mananticipated. For many sites, me programexpectation of returning ground waters to men-beneficial uses (see Section 1.2.1) often requiresvery long time periods and may not be practicablefor all or portions of me site. Thus, the ultimatecleanup goal for ground water may need to bedifferent over different areas of the site (seeSection 1.3.1). For sites where achieving theultimate goal will require a long time period,interim remediation objectives will generally beappropriate, such as preventing further plumemigration. Therefore, a critical first step in theremedy selection process is to determine thefun range of remedial objectives that areappropriate for a particular site.

This guidance is intended to emphasize meimportance of using site-specific remedialobjectives as me focus of the remedy selectionprocess for contaminated ground water. Thoseremedy components that influence attainment ofremedial objectives should receive me greatestattention. For example if restoring me aquifer tobeneficial use is me ultimate objective, remedycomponents that influence attainment of cleanuplevels in the aquifer include: methods forextracting ground water, enhancing contaminantrecovery, controlling subsurface contaminantsources (e.g., nonaqueous phase liquids orNAPLs, discussed in Appendix Al) or in-situtreatment of contaminants. Some or all of theseremedy components should be included inremedial alternatives that are developed andevaluated in detail in the feasibility study (FS)when aquifer restoration is a remedialobjective.

Although me technologies employed for treatingextracted ground water and the types of dischargefor the treated effluent are important aspects of aremedy, they have little influence on reducingcontaminant levels or minimizing contaminantmigration in the aquifer. In developing this

guidance, historical patterns of remedy selectionand available technical information were reviewedin order to identify presumptive technologies forex-situ treatment of ground water. By providingpresumptive technologies, this guidanceattempts to streamline selection of thesetechnologies and shift the tune and resourcesemployed in remedy selection to other, morefundamental aspects of the ground-waterremedy.

Although extraction and treatment has been andwill continue to be used as part of the remedy formany sites with contaminated ground water, itmay not be the most appropriate remediationmethod for all sites or for all portions of a givencontaminant plume. Also, remedial alternativesmat combine extraction and treatment with othermethods, such as natural attenuation (defined inSection 2.6.5) or in-situ treatment, may haveseveral advantages over alternatives that utilizeextraction and treatment alone (see Section 2.4.2).(Remedial alternatives are evaluated againstremedy selection criteria defined in me NationalContingency Plan at §300.430(eX9)(iii) (FederalRegister, 1990a).) In general, the remedyselection process should consider whetherextraction and treatment can achieve remedialobjectives appropriate for the site and how misapproach can be most effectively utilized toachieve these objectives. This guidance alsodescribes a presumptive response strategywhich facilitates selection of both short andlong-term remediation objectives duringremedy selection, and allows the effectivenessof the remedy to be improved duringimplementation.

1.1 Purpose of Guidance

In summary, this guidance is intended to:

• Describe a presumptive responsestrategy, at least some elements of whichare expected to be appropriate for an siteswith contaminated ground water;

• Identify presumptive technologies fortreatment of extracted ground water (ex-situ treatment) that are expected to beused (see EPA,1993d) for sites whereextraction and treatment is part of theremedy;

• Simplify the selection of technologies forme ex-situ treatment component of aground-water remedy, and improve thetechnical basis for these selections; and

• Shift the time and resources employedin remedy selection from ex-situtreatment to other, more fundamentalaspects of the ground-water remedy, asdiscussed above.

13, Expectations and Objectives for Ground-Water Cleanup

Careful consideration should be given to nationalprogram expectations as well as site-specificconditions when determining cleanup objectivesthat are appropriate for a given site.

1.2.1 Program Expectations. Expectations forcontaminated ground water are stated in theNational Oil and Hazardous Substances PollutionContingency Plan (NCP), as follows:

"EPA expects to return usable groundwaters to their beneficial uses whereverpracticable, within a timeframe mat isreasonable given the particularcircumstances of the site. Whenrestoration of ground water to beneficialuses is not practicable, EPA expects toprevent farther migration of the plume,prevent exposure to the contaminatedground water, and evaluate farther riskreduction." (Federal Register, 1990a;§300.430 (aXlXi")(F). emphasis added.)

The Preamble to the NCP explains that (heprogram expectations are not "bindingrequirements." "Rather, the expectations areintended to share collected experience to guide

those developing cleanup options" (FederalRegister, 1990a; at 8702).

12.1 Objectives for Site Response Actions.The program expectations can be used to definethe following overall objectives for site responseactions, which are generally applicable for all siteswith contaminated ground water.

• Prevent exposure to contaminated groundwater, above acceptable risk levels;

• Prevent or minimize farther migration ofthe contaminant plume (plumecontainment);

• Prevent or minimize farther migration ofcontaminants from source materials toground water (source control); and

• Return ground waters to their expectedbeneficial uses wherever practicable(aquifer restoration).

In this guidance die term "response action" is usedto indicate an action initiated under eitherCERCLA removal or remedial authority."Response objective" is the general description ofwhat a response action is intended to accomplish.Source control is included as ah objective becausethe NCP expectation of aquifer restoration willnot be possible unless farther leaching ofcontaminants to ground water is controlled, fromboth surface and subsurface sources. Theobjectives, given above, are listed in thesequence in which they should generally beaddressed at sites.

Monitoring of ground-water contamination is nota separate response objective, but is necessary toverify that one or more of the above objectives has

, been attained, or will likely be attained (seeSection 2.1.3). Other response objectives mayalso be appropriate for some sites, depending onme type of action being considered and siteconditions (e.g., maximizing the reuse of extractedground water may be an appropriate objective forsome sites). Response objectives may be

different over different portions of thecontaminant plume, as discussed in Section1.3.1.

U Lessons Learned

The most important lesson learned duringimplementation ofSupernmd and otherremediation programs is that complex siteconditions are more common man previouslyanticipated, including those related to the sourceand type of contaminants as well as sitehydrogeology. As a result of these sitecomplexities, restoring all or portions of thecontaminant plume to drinking water or similarstandards may not be possible at many sites usingcurrently available technologies.

U.I Sources and Types of Contaminants.Approximately 85 percent of sites on theCERCLA National Priorities List (NPL sites)have some degree of ground-water contamination.Contaminants have been released to ground waterat a wide variety of site types and can include avariety of contaminants and contaminantmixtures. Sources of contaminants to groundwater not only include facilities from which meoriginal release occurred (e.g., landfills, disposalwells or lagoons, storage tanks and others) butalso include contaminated soils or othersubsurface zones where contaminants have cometo be located and can continue to leach into groundwater (e.g., NAPLs, see Appendix Al). Thus, meplume of contaminated ground water mayencompass NAPLs in me subsurface (sources ofcontamination) as well as dissolved contaminants.m mis case, different response objectives may beappropriate for different portions of the plume.For example, source control (e.g., containment)may be the most appropriate response objectivefor portions of the plume where NAPLs arepresent and can not practicably be removed, whileaquifer restoration may be appropriate only for theremaining portions of the plume (see Section2.5.3).

Although originating from a variety of sources,contaminants which reach ground water tend to be

those that are relatively mobile and chemicallystable in the subsurface environment (e.g., lesslikely to sorb to soil particles or degrade above thewater table). Organic and inorganic contaminantsmost frequently found in ground water atCERCLA sites are listed in Appendix A2.Sixteen of me 20 most common organiccontaminants are volatile organic compounds(VOCs). Of the 16 VOCs, 12 are chlorinatedsolvents and four are chemicals found inpetroleum fuels. Petroleum fuels are lightnonaqueous phase liquids (LNAPLs, with adensity lighter man water); while most chlorinatedsolvents are dense nonaqueous phase liquids(DNAPLs) in pure form (see Appendix Al).

133, Factors Limiting Restoration Potential.At many sites, restoration of ground water tocleanup levels denned by applicable or relevantand appropriate requirements (ARARs) or risk-based levels may not be possible over all orportions of the plume using currently availabletechnologies. Two types of site conditions inhibitthe ability to restore ground water:

• Hydrogeologic factors, and

• Contaminant-related factors.

Recent studies by EPA and others have concludedthat complex site conditions related to thesefactors are more common at hazardous waste sitesthan originally expected (EPA, 1989a, 1992b,1992g, and 1993b; and the National ResearchCouncil, 1994). Examples ofhydrogeologic orcontaminant-related factors affecting the difficultyof restoring ground water are given in Figure 1.These types of site conditions should beconsidered in the site conceptual model, which isan interpretive summary of me site informationobtained to date (not a computer model). Refer toEPA, 1993b and 1988a for additional informationconcerning me site conceptual model. For everysite, data should be reviewed or new datashould be collected to identify factors thatcould increase (or decrease) the difficulty ofrestoring ground water.

Certsgene(Figu

S£w

s

5

S'e

Rgurel. Examplestin site characteristics may limralized. The particular factor (re 1 is taken from EPA. 1993b

Site/ContaminantCharacteristics

Nature of Release

Biobc/Aboitic DecayPotential

Volatility

ContaminantRetardation (SofpUon)Potential

Contaminant Phase

Volume o(Contaminated Media

Contaminant Depth

HydrogeologicCharacteristics

Stratigraphy

Texture ofUnconsoGdated Deposits

Degree of Heterogeneity

Hydraulic Conductivityof Aquifer

Temporal Variationof Flow Regime

Vertical Flow

of Factors Affecting Ground-Water Restoration Potentialit the effectiveness of subsurface remediation. The examples listed below are highlyor combination of factors that may critically limit restoration potential will be site specific.with minor modifications.)

Generalized Remediation Difficulty ScaleIncreasing difficulty

Small Volume Large VolumeShort Duration ———————————————————^> Long DurationSlug Release Continual Release

High ————————————————————————^- Low

High ————————————————————————^- Low

1^, ^ High

Aqueous, Gaseous —^- Sorbed —^" LNAPLs——^- DNAPLs

e.g.. Planar Bedding e.g., Interbedded and DiscontinuousStrata

Sand Clay

Homogeneous Heterogeneous e.g., Interbedded sand ande.g., well-sorted sand silts, days, fractured media, karst

High (>104 cm/sec)———————^- Low(<1o1 cnVsec)

Little/None———————————^. High

Little——————————————^. Large Downward FlowComponent

133 Assessing Restoration Potential.Characterizing all site conditions that couldincrease the difficulty of restoring ground water isoften not possible. As a result, the likelihood thatARAR or risk-based cleanup levels can beachieved (restoration potential) is somewhat tohighly uncertain for many sites, even after arelatively complete remedial investigation. Thisuncertainty can be reduced by using remedyperformance in combination with sitecharacterization data to assess the restorationpotential. By implementing a ground-waterremedy in more than one step or phase (as twoseparate actions or phasing of a single action asdescribed in Section 22), performance data froman initial phase can be used to assess dierestoration potential and may indicate matadditional site characterization is needed. Inaddition to providing valuable data, the initialremedy phase can be used to attain short-termresponse objectives, such as preventing furtherplume migration. Phased implementation ofresponse actions also allows realistic long-termremedial objectives to be determined prior toinstallation of me comprehensive or "final"remedy.

A detailed discussion of factors to consider forassessing restoration potential is provided inGuidance for Evaluating the TechnicalImpracticability of Ground-Water Restoration(EPA, 1993b; Section 4.4.4). An especiallyimportant tool for this evaluation is the siteconceptual model, which should integrate datafrom site history, characterization and responseactions. This assessment could providejustification for waiving ARARs due to technicalimpracticability from an engineering perspectiveover all or portions of a site (EPA, 1993b). It isrecommended that technical assistance be enlistedfrom regional technical support staff or theTechnical Support Project (EPA, 1994d) whenevaluating technical impracticability.

Data from remedy performance are not alwaysnecessary to Justify an ARAR waiver due totechnical impracticability (see Section 2.6.3).At the completion of the remedial investigation

(RI), site conditions may have been characterizedto me extent needed for EPA (or the lead agency)to determine that ground-water restoration istechnically impracticable from an engineeringperspective (EPA, 1993b; EPA 1995b). For thiscase, an ARAR waiver request can be submittedto EPA (or the lead agency), and if approved,included in the Record of Decision (ROD). It willoften be appropriate to include an ARAR waiverin the ROD for portions of a site where DNAPLshave been confirmed in the aquifer (see Section2.5.3).

2.0 PRESUMPTIVE RESPONSESTRATEGY

2.1 Definition and Basis for Strategy

Key elements of the presumptive strategy aresummarized in Highlight 1. In the presumptiveresponse strategy, site characterization andresponse actions are implemented in a severalsteps, or in a phased approach. In a phasedresponse approach, site response activities areimplemented in a sequence of steps, or phases,such that information gained from earlier phases isused to refine subsequent investigations,objectives or actions (EPA, 1989a, 1992b,1993b).

In general for sites with contaminated groundwater, site characterization should becoordinated with response actions and bothshould be implemented in a step-by-step orphased approach.

Performance data from an initial response actionare also used to assess (he likelihood mat ARARor risk-based cleanup levels can be attained bylater, more comprehensive actions. Although it isrecognized that phased implementation may notbe appropriate for all ground-water remedies, EPAexpects that some elements of this strategy will beappropriate for all sites with contaminated groundwater and that all elements will be appropriate formany of these sites. For this reason, theresponse approach given in

Highlight 1. Presumptive ResponseStrategy

• For sites with contaminated groundwater, site characterizationshould be coordinated withresponse actions and both shouldbe Implemented In a phasedapproach (Sections 1.3.3 and 2.1).

• Early or Interim actionsshould beused to reduce site risks (bypreventing exposure to and furthermigration of contaminants) and toprovide additional site data (Section2.1.2).

• Site characterization andperformance data from early orinterim ground-water actions shouldbe used to assess the likelihoodof restoring ground watertoARAR or risk-based cleanup levels(restoration potential). (Sections1.3.3 and 2.1.2.)

• The restoration potential should beassessed prior to establishingobjectives for the long-termremedy (Sections 1.3.3 and 2.1.2).

• All ground-water actions shouldinclude provisions for monitoringand evaluating their performance(Section 2.1.3).

• Ground-water response actions,especially those using extractionand treatment, should generally beImplemented In more than onephase - either as two separateactions or phasing of a single action(Sections 2.2.1 and 2.2.2).

• In addition to phasing, post-construction refinements willgenerally be needed for long-termremedies, especially those usingextraction and treatment (Section2.3.1).

Highlight 1 is a presumptive strategy forcontaminated ground water.

Also, this response strategy is consideredpresumptive because the basic elements wereincluded in all previous policy directivesconcerning ground-water remediation from EPA'sOffice of Solid Waste and Emergencyrecommended use of a phased approach for sitecharacterization and response actions, and morefrequent use of early actions to reduce site risks.Better integration of site activities and morefrequent use of early actions are also essentialcomponents of (he Superfund AcceleratedCleanup Model (SACM), defined in EPA, 1992d.

2.1.1 Benefits of Phased Approach.Implementing investigations and actions in phasesprovides me following major benefits:

• Data from earlier response actions areused to further characterize the site andassess restoration potential;

• Attainable objectives can be set for eachresponse phase;

• Flexibility is provided to adjust theremedy in response to unexpected siteconditions;

• Remedy performance is increased,decreasing remediation timeframe andcost; and

• Likely remedy refinements are built intothe selected remedy, better defining thepotential scope and minimizing the needfor additional decision documents.

2.1.2 Early Actions. "Early" refers to the timingof uie start of an action with respect to otherresponse actions at a given site. For Superfundsites, early actions could include removal actions,interim remedial actions, or early final remedialactions (EPA, 1992bandEPA, 1991b). Althoughinitiated prior to other actions, some early ground-water actions may need to operate over a long time

period (e.g., hydraulic containment actions). Inthis guidance the later, more comprehensiveground-water action is called the "long-termremedy," consistent with SACM terminology(EPA, 1992e). Early actions that should beconsidered in response to contaminated groundwater are listed in Highlight 2, categorized byresponse objective. Early or interim actionsshould be used to reduce site risks (bypreventing exposure to contaminated groundwater and further migration of contaminants)and to provide additional site data.

Factors for determining which responsecomponents are suitable for early or interimactions include: me timeframe needed to attainspecific objectives, the relative urgency posed bypotential or actual exposure to contaminatedground water (e.g., likelihood that contaminantswill reach drinking water wells), me degree towhich an action will reduce site risks, usefulnessof information to be gained from the action, sitedata needed to design me action, and compatibilitywith likely long-term actions (EPA, 1992e).Whether to implement early response actions andwhether to use removal or remedial authority forsuch actions should be determined by me"Regional Decision Team" defined under SACM(EPA, 1992f) or similar decision-making body forthe site.

Early or interim actions should be integrated asmuch as possible with site characterization andwith subsequent actions in a phased approach.Once implemented, early actions will oftenprovide additional site characterizationinformation, which should be used to update mesite conceptual model. Also, treatability studies(see Section 3.4.5) needed for selection or designof me long-term remedy should be combined withearly actions whenever practical. Sitecharacterization and performance data from earlyor interim ground-water actions should be used toassess the likelihood of restoring ground water toARAR or risk-based cleanup levels (restorationpotential). The restoration

Highlight 2. Early Actions That ShouldBe Considered

Prevent exposure to contaminated groundwater

• Plume containment

• Alternate water supply

• Well head treatment

• Use restrictions

Prevent further migration of contaminantplume:

• Plume containment

• Contain (and/or treat) plume "hotspots"

Prevent further migration of contaminantsfrom sources

• Source removal and/or treatment

Excavate wastes or soilsand remove from site

Excavate soils and treat ex-situ

Treat soils in-situ

Extract free-phase NAPLs(see Appendix A1)

• Source containment

Contain wastes or soils

Contain subsurface NAPLs

Provide additional site data:

• Assess restoration potential

• Combine actions with treatabilitystudies

potential should be assessed prior toestablishing objectives for the long-termremedy (see Section 1.3.3).2.13 Monitoring.Monitoring is needed to evaluate whether theground-water action is achieving, or will achieve,the intended response objectives for die site (seeSection U.I) and other performance objectivesfor the action (e.g., discharge requirements). ASground-water actions should includeprovisions for monitoring and evaluating theirperformance. A monitoring plan should bedeveloped for both early and long-term actions. Ingeneral, the monitoring plan should include:

• Response objectives and performancerequirements for me ground-water action;

• Specific monitoring data to be collected;

• Data quality objectives;

• Methods for collecting, evaluating andreporting the performance monitoringdata; and

• Criteria for demonstrating that responseobjectives and performance requirementshave been attained.

Flexibility for adjusting certain aspects ofmonitoring during the life of the remedy should beincluded in the monitoring plan, such as changesin the monitoring frequency as the remedyprogresses or other changes in response to remedyrefinements (see Section 2.3.1). A detaileddiscussion of the data quality objectives process isprovided in EPA, 1993J. Methods for monitoringdie performance of extraction and treatmentactions are discussed in EPA, 1994e.

13. Phased Response Actions

In general, ground-water response actions,especially those using extraction andtreatment, should be implemented in morethan one phase. There are two options forphasing response actions - implementation of twoseparate actions, or implementation of a single.

action in more than one phase. It is recognizedthat phased implementation may not beappropriate for all ground-water remedies. Insome cases, it may be more appropriate to installdie entire remedy and men remove from servicethose components mat later prove to be unneeded.

2.2.1 Two Separate Actions. In mis approach anearly or interim ground-water action is followedby a later, more comprehensive action (me long-term remedy). A flow chart of mis approach isgiven in Figure 2. Earlier ground-water actionsare used to mitigate more immediate threats, suchas preventing further plume migration. Responseobjectives for die long-term remedy are notestablished until after performance of me earlieraction is evaluated and used to assess melikelihood diat ground-water restoration (or omerappropriate objectives) can be attained. Twoseparate decision documents are used, in whichresponse objectives are specified mat areappropriate for each action. The earlier decisiondocument could be an Action Memorandum or anInterim Record of Decision (Interim ROD), sinceme early action could be initiated under eitherCERCLA removal or remedial audiority. Thisapproach should be used when sitecharacterization data are not sufficient todetermine the likelihood of attaining long-termobjectives (e.g., restoring ground water) overall or portions of the plume, which will be thecase for many sites. In order to providesufficient data for assessing me restorationpotential, me early or interim action may need tooperate for several years.

2.2.2 Phasing of a Single Action. In thisapproach die long-term remedy for ground wateris implemented in more man one design andconstruction phase. A flow chart of mis approachis given in Figure 3. Response objectives for melong-term remedy are specified in a single Recordof Decision (ROD) prior to implementing meremedy. Provisions for assessing die attainabilityofdiese objectives using performance data froman initial remedy phase are also included in meROD. Thus, phased remedy implementation andassessment of remedy performance are specified

in one ROD. A second decision document couldstill be required if evaluation of the first phase

Figure 2. Phased Ground-Water Actions: Eariy Action Followed by Long-Term Remedy

f This approach should be used when site characterization data flreJial̂ uffideA to detennlne the tkelihood of attaining long-term 1I objectfr/es (e.g., restoring ground-water) over aB or portions of the plume. }

DecisionDocuments

Remedy

10

Figure 3. Phased Ground-Water Actions: Long-Term Remedy Implemented in Phases——————————————————————————————————————————This approach should be used when site characterization dataare sufficient to determine that the likelihoodof attaining long-term objectives Is relatively high.

DecisionDocuments

RemedyPhase Remedy Selection/ Implementation Steps

11

indicates (hat long-term objectives or otheraspects of the remedy require modification, andthe modified remedy differs significantly from meselected remedy in terms of scope, performance orcost (EPA, 1991a). This approach should beused when site characterization data indicatethat the likelihood of attaining long-termobjectives is relatively high.

When phased remedy implementation is specifiedin a ROD, the Agency should ensure mat theproposed plan contains sufficient informationregarding me nature, scope timing and basis offuture decision points and alternatives that mepublic is able to evaluate and comment on theproposed remedy. Example language illustratinghow such an approach can be specified in theselected remedy portion of me ROD is included inAppendices Bl and B2 for hypothetical sites.These examples follow the suggested RODlanguage given in EPA, 1990b, although mewording has been updated to reflect this and otherrecent guidance (EPA, 1993b). For comparison,suggested ROD language from the EPA, 1990b isincluded as Appendix B4.

Phased implementation of a remedy can often bebeneficial even for relatively simple ground-wateractions. For example, one extraction well couldbe installed as the initial phase and theperformance of this well would be used todetermine whether any additional wells are neededand whether long-term objectives need to be re-evaluated.

Phased implementation of an extraction andtreatment remedy will require mat the treatmentsystem be designed to accommodate phasedinstallation of the extraction system. Presumptivetechnologies for the treatment system and otherdesign considerations are discussed in Section 3.Use of modular treatment components, which canbe easily added or removed from me treatmentsystem, may facilitate phased implementation orother changes in flow or contaminantconcentration that may occur during me life of aremedy. Another approach is to design metreatment system for me higher flows expected

from all phases of the extraction system. Somecomponents of the remedy, such as buriedportions of the piping distribution system, aredifficult to install in phases and should bedesigned to cany the highest expected flows.

23 Post-Construction Refinements

Even after phased implementation of a ground-water remedy, post-construction refinements willgenerally be needed because of me long timeperiod over which the remedy will operate,especially for extraction and treatment remedies.The refinement portion of the long-term remedy,after phased design and construction, is shown inboth Figures 2 and 3.

2 J.I Types of Refinements. Post-constructionrefinements that should be considered forextraction and treatment remedies are given inHighlight 3. These refinements are intended to berelatively minor changes to the remedy (i.e., forwhich an Explanation of Significant Differences(ESD) or ROD Amendment would generally notbe required). For example, adding a newextraction or reinjection well, or a few additionalmonitoring wells should be considered a minormodification to a remedy that includes a relativelylarge number of such wells, because the overallscope, performance and cost of the remedy are notsignificantly changed (EPA, 1991a). One or moresuch refinements should generally be implementedwhen the results of a remedy evaluation indicatemat they are needed to increase the performanceof the remedy or to decrease the remediationtimeframe.

2.3.2 Documenting Refinements. Potential post-construction refinements should be included in meROD as part of me selected remedy. Listingspecific remedy refinements in the ROD serves tocommunicate the anticipated mil scope of theremedy to all concerned parties at an early date,and also minimizes the likelihood that asubsequent ESD or ROD Amendment will beneeded. When remedy refinements are specifiedin a ROD, the Agency should ensure mat the

12

Highlight 3. Remedy Refinements forExtraction/Treatment Remedies

• Change the extraction rate in someor all wells.

• Cease extraction from some wells.

• Initiate "pulsed pumping" (seeAppendix A4).

• Add or remove extraction orreinjection wells, or drains.

• Add or remove monitoring wells.

• Refine source control componentsof remedy.

• Refine enhanced recovery or in-situdegradation components of remedy(see Note).

• Refine ex-situ treatmentcomponents

NOTE: A ground-water remedy couldinclude both extraction and treatment and In-situ treatment methods.

proposed plan contains sufficient informationregarding the nature, scope timing and basis offuture decision points and alternatives that thepublic is able to evaluate and comment on theproposed remedy. Example ROD languagespecifying likely post-construction refinements forme extraction portion of the selected remedy isgiven in Appendices Bl and B2. Even if an ESDis not required, a letter or memorandum should beincluded in the post-ROD portion of theAdministrative Record explaining the minorremedy modifications and the reasons for mem.Additional information concerning documentationof remedy modifications can be found in the EPAfact sheet entitled Guide to Addressing Pre-RODand Post-ROD Changes (EPA, 1991a).

2.4 Integrating Response Actions

In general, actions in response to contaminatedground water should be planned and implementedas part of an overall strategy. Earlier actions (seeHighlight 2 for examples) should be compatiblewith and not preclude implementation of lateractions. For example, permanent facilities shouldnot be constructed which could interfere withpossible later actions (e.g., structures that wouldinterfere with later construction of extraction wellsor of a cap).

2.4.1 Integrating Source Control and Ground-Water Actions. Restoration of contaminatedground water generally will not be possible unlesscontaminant sources have been controlled in somemanner. Source control is a critical component foractive restoration remedies (e.g., extraction andtreatment and in-situ methods) as well as fornatural attenuation (defined in Section 2.6.5).Selection of appropriate source control actionsshould consider whether other contaminantsources (i.e., NAPLs) are likely to be present inaddition to contaminated soils. If NAPLs arepresent, the vast majority of contaminant masswill likely reside in the subsurface NAPLs ratherthan in the surficial soils. Therefore, for mis casesource control actions that are intended tominimize further contamination of ground watershould focus on controlling migration ofcontaminants from the subsurface NAPLs. Also,capping or treatment of surficial soils may beneeded to prevent exposure to contaminants fromdirect soil contact or inhalation, but these actionsalone would be ineffective in preventing furthercontamination of ground water at sites whereNAPLs are present.

1.4.2 Combining Ground-Water RestorationMethods. A remedy could include more than onemethod for restoring ground water to its beneficialuses, such as combining extraction and treatmentwith natural attenuation or in-situ-treatment withextraction and treatment. Extraction andtreatment is especially useful for providinghydraulic containment of those portions of the

13

plume where contaminant sources are present(e.g., subsurface NAPLs or contaminated soils), orfor containing or restoring those plume areas withrelatively high concentrations of dissolvedcontamination ("hot spots"). However, extractionand treatment may not be the best method forrestoring large areas of the plume with lowcontaminant levels.

Once source areas are controlled, naturalattenuation may be able to restore largeportions of the plume to desired cleanup levelsin a timeframe that is reasonable (see Section2.6.2) when compared with the timeframe andcost of other restoration methods. Thus,natural attenuation of some plume areas combinedwith extraction and treatment to contain sourceareas and/or plume "hot spots" may be the mostappropriate restoration approach for many siteswith relatively large, dilute plumes. Whether ornot natural attenuation is used alone or combinedwith other remediation methods, me Agencyshould have sufficient information to demonstratethat natural processes are capable of achieving theremediation objectives for the site. EPAiscurrently preparing a directive that will providemore detailed discussion ofEPA policy regardingthe use of natural attenuation for remediation ofcontaminated ground water (EPA, 1996c).

By combining in-situ treatment and extraction andtreatment methods it may be possible tosignificantly increase the effectiveness with whichcontaminants are removed from me aquifer. Inthis guidance, in-situ treatment methods forground water are divided into two types:

• Methods that can be used to enhancecontaminant recovery during extractionand treatment (e.g., water, steam orchemical flooding; hydraulic or pneumaticfracturing); and

• Methods for in-situ degradation ofcontaminants generally involve addingagents to the subsurface (i.e., via wells ortreatment walls) which facilitate chemicalor biological destruction, and have the

potential to be used as an alternative toextraction and treatment for long-termrestoration of ground water.

Examples of both types of in-situ treatmentmethods are given in Appendix A3. Reinjectionof treated ground water can be used as a methodfor enhancing contaminant recovery as well as adischarge method, if the reinjection is designed formis purpose as part of an extraction and treatmentremedy. When considering enhanced recoverymethods for sites with subsurface NAPLs,potential risks of increasing the mobility ofNAPLs should be evaluated. Methods of in-situdegradation of contaminants most frequently usedat Superfund sites include air sparging, varioustypes of in-situ biological treatment andpermeable treatment walls or gates (EPA, 1995e).Additional information concerning air spargingand permeable treatment walls is available inEPA, 1995fand EPA, 1995d, respectively. EPAencourages me consideration, testing and use ofin-situ technologies for ground-water remediationwhen appropriate for (he site.

IS Strategy for DNAPL Sites

Dense nonaqueous phase liquids (DNAPLs) posespecial cleanup difficulties because they can sinkto great depths in the subsurface, continue torelease dissolved contaminants to me surroundingground water for very long time periods, and canbe difficult to locate. Due to the complex natureof DNAPL contamination, a phased approach tocharacterization and response actions is especiallyimportant for sites where DNAPLs are confirmedor suspected. A recent EPA study concluded matsubsurface DNAPLs may be present at up to 60percent ofCERCLA National Priorities List sites(EPA, 1993c). Refer to Appendix Al foradditional background information on DNAPLs.

Two types of subsurface contamination can bedefined at DNAPL sites, me:

• DNAPL zone, and me

• Aqueous contaminant plume.

14

The DNAPL zone is that portion of (hesubsurface where immiscible liquids (free-phaseor residual DNAPL) are present either above orbelow the water table. Also in the DNAPL zone,vapor phase DNAPL contaminants are presentabove me water table and dissolved phase belowthe water table. The aqueous contaminantplume is that portion of the contaminated groundwater surrounding the DNAPL zone whereaqueous contaminants derived from DNAPLs aredissolved in ground water (or soibed to aquifersolids) and immiscible liquids are not present.

2.5.1 Site Characterization. If DNAPLs areconfirmed or suspected, the remedial investigation(RI) should be designed to delineate the:

• Extent of aqueous contaminant plumes,and the

• Potential extent of DNAPL zones.

Methods and strategies for characterizing DNAPLsites as well as suggested precautions arediscussed in other guidance (EPA, 1992a and1994b) and by Cohen and Mercer, 1993. Thereason for delineating these areas of the site is thatresponse objectives and actions should generallybe different for the DNAPL zone man for meaqueous contaminant plume. It is recognized thatfor some sites complete delineation of theDNAPL-zone may not be possible.

2.5.2 Early Actions. The early actions listed inHighlight 2.should be considered. Also, mefollowing early actions are specificallyrecommended for DNAPL sites (EPA 1992b,1993b):

• Prevent further spread of me aqueousplume (plume containment);

• Prevent further spread of hot spots in theaqueous plume (hot spot containment);

• Control further migration of contaminantsfrom subsurface DNAPLs to thesurrounding ground water (sourcecontrol); and

• Reduce me quantity of source material(free-phase DNAPL) present in theDNAPL zone, to the extent practicable(source removal and/or treatment).

At DNAPL sites, hot spots in me aqueous plumeoften are associated with subsurface DNAPLs.Therefore, me second and third actions listedabove are essentially me same.

2.5.3 Long-Term Remedy. The long-termremedy should attain those objectives listed abovefor the DNAPL zone, by continuing early actionsor by initiating additional actions. Althoughcontaminated ground waters generally are notconsidered principal threat wastes, DNAPLsmay be viewed as a principal threat because theyare sources of toxic contaminants to ground water(EPA,1991c). For this reason EPA expects toremove or treat DNAPLs to the extent practicablein accordance with me NCP expectation to '"usetreatment to address the principal threats posed bya site, wherever practicable" (Federal Register,1990a; §300.430 (a)(l)(ui)(A)). However,program experience has shown that removal ofDNAPLs from the subsurface is often notpracticable, and no treatment technologies arecurrently available which can attain ARAR orrisk-based cleanup levels where subsurfaceDNAPLs are present. Therefore, EPA generallyexpects that the long-term remedy will controlfarther migration of contaminants fromsubsurface DNAPLs to the surroundingground water and reduce the quantity ofDNAPL to the extent practicable.

For me aqueous plume, me long-term remedyshould:

• Prevent further spread of me aqueousplume (plume containment);

15

• Restore the maximum areal extent of theaquifer to those cleanup levelsappropriate for its beneficial uses (aquiferrestoration).

In general, restoration of the aquifer to ARARor risk-based cleanup levels in a reasonabletimefirame will not be attainable in the DNAPLzone unless the DNAPLs are removed. For thisreason, it is expected mat ARAR waivers due totechnical impracticability will be appropriate formany DNAPL sites, over portions of sites wherenon-recoverable DNAPLs are present (EPA,1995c). Also, EPA generally prefers to utilizeARAR waivers rather than ARAR complianceboundaries for such portions of DNAPL sites (seeSection 2.6.4). A waiver determination can bemade after construction and operation of meremedy or at me time of remedy selection (Le., inthe ROD), whenever a sufficient technicaljustification can be demonstrated (EPA, 1993b;EPA 1995b). For further information refer toSection 2.6.3 of mis guidance and EPA'sGuidance for Evaluating the TechnicalImpracticability of Ground-Water Restoration(EPA, 1993b). Restoration of the aqueous plumemay also be difficult due to hydrogeologic factors,such as sorption of dissolved contaminants tosolids in finer grained strata. For some sites,ARAR waivers may also be appropriate for all orportions of the aqueous plume when supported byadequate justification.

2.6 Areas of Flexibility in Cleanup Approach

The current response approach to contaminatedground water, as defined in the NCP and otherguidance, includes several areas of flexibility inwhich response objectives and the timeframe inwhich to meet them can be adjusted to meet sitespecific conditions. These are briefly discussedbelow.

2.6.1 Beneficial Uses and ARARs. Since EPAgenerally expects to return contaminated groundwaters to their beneficial uses whereverpracticable, the required cleanup levels for a givensite should be determined from applicable or

relevant and appropriate requirements (ARARs)based on the current and expected futurebeneficial uses of the ground water at that site.Depending on state requirements and waterquantity or quality characteristics, some groundwaters are not expected to provide a future sourceof drinking water (e.g., EPA Class in groundwaters (EPA, 1986) or similar state designations).In general, drinking water standards are relevantand appropriate cleanup levels for ground watersthat are a current or future source of drinkingwater, but are not relevant and appropriate forground waters that are not expected to be a futuresource of drinking water (Federal Register, 1990a;Preamble at 8732). (Drinking water standardsinclude federal maximum contaminant levels(MCLs) and/or non-zero maximum contaminantlevel goals (MCLGs) established under me SafeDrinking Water Act, or more stringent statedrinking water standards.) Ground waters mayhave other beneficial uses, such as providing baseflow to surface waters or recharging otheraquifers. For contaminated ground waters (hatdischarge to surface water, water quality criteriaestablished under me Clean Water Act, or morestringent state surface water requirements, mayalso be cleanup level ARARs (Federal Register,1990a; Preamble at 8754). Thus, the beneficialuses of contaminated ground water at a particularsite will generally provide the basis fordetermining which federal or state environmentalrequirements are applicable or relevant andappropriate cleanup levels. For additionalinformation on the determination of cleanuplevels, refer to EPA, 1988b, Chapter 4.

Determination of current and expected futurebeneficial uses should consider state ground-waterclassifications or similar designations. Severalstates have developed ground-water use or prioritydesignations as part of a Comprehensive StateGround Water Protection Program (CSGWPP),defined in EPA, 1992h. EPA is currentlydeveloping a directive (EPA, 1996a) which willrecommend (hat EPA remediation programsshould generally defer to state determinations offuture ground-water use — even when thisdetermination differs from the use that would

16

otherwise have been determined by EPA - whensuch determinations are:

• Developed as part of an CSGWPP that isendorsed by EPA, and

• Based on CSGWPP provisions that canbe applied at specific sites (EPA, 1996a).

This provision of the directive, when final, isintended to supersede previous guidance containedin the Preamble to the NCP (Federal Register,1990a; at 8733). Refer to EPA, 1996a foradditional information concerning me role ofCSGWPPs in the selection of ground-waterremedies. When information concerningbeneficial uses is not available ftom a CSGWPP,ground-water classifications defined in EPA, 1986(i.e., EPA Classes I, n or ffl) or "more stringent"state ground-water classifications (or similar statedesignations) should generally be used todetermine the potential future use, in accordancewith the NCP Preamble (Federal Register, 1990a;at 8732-8733). Regardless of the ground-wateruse determination, remedies selected underCERCLA authority must protect humanhealth and the environment and meet ARARs(or invoke an ARAR waiver).

Many states have antidegradation or similarregulations or requirements that may be potentialARARs. Such requirements typically focus on 1)prohibiting certain discharges, 2) maintainingground-water quality consistent with its beneficialuses, or 3) maintaining naturally occurring(background) ground-water quality. Regulationsof the third type do not involve determination offuture ground-water use, and often result incleanup levels that are more stringent than thedrinking water standard for a particular chemical.Such requirements are potential ARARs if theyare directive in nature and intent and establishedthrough a promulgated statute or regulation that islegally enforceable (see Federal Register, 1990a;Preamble at 8746). For further informationconcerning issues related to state ground-waterantidegradation requirements, refer to EPA,1990a.

1.6.2 Remediation Timeframe. "Remediationtimeframes will be developed based on thespecific site conditions" (Federal Register, 1990a;Preamble at 8732). Even though restoration tobeneficial uses generally is me ultimate objective,a relatively long time period to attain thisobjective may be appropriate for some sites. Forexample, an extended remediation timeframegenerally is appropriate where contaminatedground waters are not expected to be used in thenear term, and where alternative sources areavailable. In contrast, a more aggressive remedywith a correspondingly shorter remediationtimeframe should generally be used forcontaminated ground waters that are currentlyused as sources of drinking water or are expectedto be utilized for this purpose in the near future(Federal Register, 1990a; at 8732). A state'sCSGWPP may include information helpful indetermining whether an extended remediationtimeframe is appropriate for a given site, such asthe expected timeframe of use, or me relativepriority or value of ground-water resources indifferent geographic areas.

A reasonable timeframe for restoring groundwaters to beneficial uses depends on the particularcircumstances of the site and the restorationmethod employed. The most appropriatetimeframe must be determined through an analysisof alternatives (Federal Register, 1990a; Preambleat 8732). The NCP also specifies that:

"For ground-water response actions, thelead agency shall develop a limitednumber of remedial alternatives thatattain site-specific remediation levelswithin different restoration time periodsutilizing one or more differenttechnologies." (Federal Register, 1990a;§300.430(e)(4).)

Thus, a comparison of restoration alternativesfrom most aggressive to passive (i.e., naturalattenuation) will provide information concerningme approximate range of time periods needed toattain ground-water cleanup levels. Anexcessively long restoration timeframe, even with

17

&e most aggressive restoration methods, mayindicate that ground-water restoration istechnically impracticable from an engineeringperspective (see Section 2.6.3). Where restorationis feasible using both aggressive and passivemethods, the longer restoration timeframerequired by a passive alternative may bereasonable in comparison with the timeframeneeded for more aggressive restorationalternatives. The most appropriate remedialoption should be determined based on me nineremedy selection factors denned in the NCP(Federal Register. 1990a; §300.430 (e)(9Xiii)).Although restoration timeframe is an importantconsideration in evaluating whether restoration ofground water is technically impracticable, nosingle time period can be specified which wouldbe considered excessively long for all siteconditions (EPA, 1993b). For example, arestoration timeframe of 100 years may bereasonable for some sites and excessively long forothers.

2.6 J Technical Impracticability. Whererestoration of ground water to its beneficial uses isnot practicable from an engineering perspective,one or more ARARs may be waived by EPA (orthe lead agency) under me provisions defined inCERCLA §121(d)(4XC)). The types of data usedto make such a determination are discussed inGuidance for Evaluating the TechnicalImpracticability of Ground- Water Restoration(EPA, 1993b). Alternative remedial strategies, tobe considered when restoration ARARs arewaived, are also discussed in EPA, 1993b. Afinding of technical impracticability may be madein the Record of Decision (ROD) prior to remedyimplementation, or in a subsequent decisiondocument after implementation and monitoring ofremedy performance.

2.6.4 Point of Compliance. The area over whichARAR or risk-based cleanup levels are to beattained is defined in me NCP as follows:

"For ground water, remediation levelsshould generally be attained throughoutthe contaminated plume, or at and beyond

the edge of me waste management areawhen waste is left in place" (FederalRegister, 1990a; Preamble at 8713).

Thus, the edge of the waste management area canbe considered as me point of compliance, becauseARAR or risk-based cleanup levels are notexpected to be attained in ground water within thewaste management area. In general, the term"waste left in place" is used in the NCP to refer tolandfill wastes that, at the completion of meremedy, will be contained or otherwise controlledwithin a waste management area.

For the purposes of ARAR compliance, EPAgenerally does not consider DNAPLs as "wasteleft in place." DNAPLs are typically not locatedin a waste management area, as envisioned in theNCP. This is because me full extent ofDNAPLcontamination is often not known, DNAPLs cancontinue to migrate in me subsurface, andmeasures for controlling their migration are eitherunavailable or have uncertain long-term reliability.Also, as discussed in Section 2.5.3, restoration ofme aquifer to ARAR or risk-based cleanup levelsgenerally will not be attainable in a reasonabletimeframe unless me DNAPLs are removed. Forthese reasons, EPA generally prefers to utilizeARAR waivers rather than an alternate pointof compliance over portions of sites where non-recoverable DNAPLs are present in thesubsurface (EPA, 1995c).

The NCP Preamble also acknowledges mat "analternative point of compliance may also beprotective of public health and (he environmentunder site-specific circumstances" (FederalRegister, 1990a; at 8753). For example, wherethe contamination plume is "caused by releasesfrom several distinct sources that are in closegeographical proximity.-.the most feasible andeffective cleanup strategy may be to address meproblem as a whole, rather than source by source,and to draw the point of compliance to encompassthe sources of release" (Federal Register, 1990a;at 8753). The NCP Preamble goes on to say that"...where there would be little likelihood ofexposure due to the remoteness of the site,

18

alternate points of compliance may be considered,provided contamination in me aquifer is controlledfrom further migration" (Federal Register, 1990a;at 8734). The Agency has not developedadditional guidance on the use of alternate pointsof compliance at Superfund sites.

2.6.5 Natural Attenuation. Natural attenuationis defined in the NCP as "biodegradation,dispersion, dilution, and adsorption" ofcontaminants in ground water (Federal Register,1990a; Preamble at 8734). The NCP goes on toexplain (hat natural attenuation may be a usefulremedial approach if site-specific data indicatethat these processes "will effectively reducecontaminants in the ground water toconcentrations protective of human health [and theenvironment] in a timeframe comparable to thatwhich could be achieved through activerestoration." This approach differs from the "noaction" alternative because natural attenuation isexpected to attain cleanup levels in a reasonabletimeframe (discussed in Section 2.6.2). The NCPrecommends use of natural attenuation where it is"expected to reduce the concentration ofcontammants in the ground water to theremediation goals [ARAR or risk-based cleanuplevels] in a reasonable timeframe."

Natural attenuation may be an appropriateremedial approach for portions of the contaminantplume when combined with other remedialmeasures needed to control sources and/orremediate "hot spots" (also see Section 2.4.2).Whether or not natural attenuation is used alone orcombined with other remediation methods, meAgency should have sufficient information todemonstrate that natural processes are capable ofachieving the remediation objectives for the site.One caution is mat natural attenuation may not beappropriate for sites where contaminantsbiodegrade to intermediate compounds mat aremore toxic and degrade more slowly.

Additional EPA policy considerations regardingthe use of natural attenuation for remediation ofcontaminated ground water are provided in EPA,1996c. Although currently in draft, this EPA

directive recommends mat remedies utilizingnatural attenuation should generally include: 1)detailed site characterization to show mat thisapproach will be effective; 2) source controlmeasures to prevent further release ofcontaminants to ground water; 3) performancemonitoring to assure that natural attenuation isoccurring as expected; and 4) Institutionalcontrols and other methods to ensure matcontaminated ground waters are not used beforeprotective concentrations are reached. Also,contingency measures may be needed in theevent that natural attenuation does not progress asexpected.

2.6.6 Alternate Concentration Limits.Alternate concentration limits (ACLs) areintended to provide flexibility in establishingground-water cleanup levels under certaincircumstances, In the Superfund program, EPAmay establish ACLs as cleanup levels in lieu ofdrinking water standards (e.g., MCLs) in certaincases where contaminated ground waterdischarges to surface water. The circumstancesunder which ACLs may be established atSuperfund sites are specified in CERCLA§121(d)(2)(B)(ii), and can be summarized asfollows:

• The contaminated ground water musthave "known or projected" points of entryto a surface water body;

• There must be no "statistically significantincreases" of contaminant concentrationsin me surface water body at those pointsof entry, or at points downstream; and

• It must be possible to reliably preventhuman exposure to the contaminatedground water through the use ofinstitutional controls.

Each of these criteria must be met and must besupported by site-specific information. Suchinformation also must be incorporated into theappropriate portions of me Administrative Record(e.g., the RI/FS and ROD).

19

The NCP Preamble also advises that ACLs not beused in every situation in which the aboveconditions are met, but only where activerestoration of the ground water is "deemed not tobe practicable" (Federal Register, 1990a; at8754). This caveat in the Preamble signals thatEPA is committed to me program goal ofrestoring contaminated ground water to itsbeneficial uses, except in limited cases. In thecontext of determining whether ACLs could orshould be used for a given site, the term"practicability" refers to an overall finding of theappropriateness of ground-water restoration,based on an analysis of remedial alternatives usingthe Superfund remedy selection criteria, especiallythe "balancing" and "modifying" criteria (EPA,1993b). (These criteria are defined in part§300.430(e)(9Xui) of the NCP (Federal Register,1990a.) This is distinct from a finding of"technical impracticability from an engineeringperspective," which refers specifically to anARAR waiver and is based on the narrowergrounds of engineering feasibility and reliabilitywith cost generally not a major factor, unlessARAR compliance would be inordinately costly(see Section 2.6.3 and EPA, 1993b). Where anACL is established, such an ARAR waiver is notnecessary. Conversely, where an ARAR is waiveddue to technical impracticability, mere is no needto establish CERCLA ACLs, as defined above.When establishing an ACL, a detailed site-specificjustification should be provided in theAdministrative Record which documents that theabove three conditions for use of ACLs are met,and that restoration to ARAR or risk-based levelsis "not practicable" as discussed above.

Although alternate concentration limits are alsodefined in the RCRA program, users of misguidance should be aware of several importantdifferences in the use of ACLs by the RCRAand Superfund programs. For "regulated units"(defined in 40 CFR 264.90) ACLs are one of methree possible approaches for establishingconcentrations limits of hazardous constituents inground water. Those options are described in 40CFR 294.94(a). Factors considered whendetermining whether an ACL is appropriate for a

particular facility are provided in 40 CFR264.94(b). The use of RCRA ACLs is not strictlylimited to cases where contaminated ground waterdischarges to surface water, or to cases whereground-water restoration is considered "notpracticable" (as is me case in Superfund).However, the factors considered in the RCRAACL decision are meant to ensure thatestablishment of ACLs will be protective ofhuman health and the environment.

A specific reference to ACLs is not made in theexisting framework for implementing RCRACorrective Action at "non-regulated units"(Federal Register, 1990b and 1996). However,the Corrective Action framework recommendsflexibility for me development and use of risk-based cleanup standards, based on considerationssimilar to those used for establishing ACLs under40 CFR 264.94.

3.0 PRESUMPTIVE TECHNOLOGIES

3.1 Presumptive Technologies for Ex-SituTreatment

Presumptive technologies for the treatmentportion of an extraction and treatment remedy (ex-situ treatment) are identified in Highlight 4.Descriptions of each of the presumptivetechnologies are presented in Appendices Dlthrough D8. These technologies are presumptivefor treatment of contaminants dissolved inground water that has been extracted from thesubsurface, and are expected to be used for thispurpose at "all appropriate sites." (Refer to thePreface of (his guidance and EPA, 1993d forfurther information concerning the Agency'sexpectations concerning me use of presumptivetreatment technologies.)

20

Highlight 4. Presumptive TechnologiesFor Treatment Of Extracted GroundWater

For treatment of dissolved organiccontaminants, volatiles, semivolatiles andothers (see Note):

• Air stripping

• Granular activated carbon (GAC)

• Chemteal/UV oxidation (for cyanidesalso)

• Aerobic biological reactors

For treatment of dissolved metals:

• Chemical precipitation

• Ion exchange/adsorption

• Electrochemical methods (whenonly metals are present)

• Aeration of background metals

For treatment of both organic andInorganic constituents:

• A combination of the technologieslisted above

NOTE: A given treatment train could includea combination of one or more of thepresumptive technologies for treatment ofdissolved contaminants as well as othertechnologies for other purposes (e.g.,separation of solids) as indicated inAppendix C2.

3.1.1 Design Styles within PresumptiveTechnologies. The presumptive technologiesidentified in Highlight 4 refer to technology typesrafter than specific designs (design styles). Eachpresumptive technology represents a singleprocess falls within one of these technology types(e.g., innovative air stripper designs, orinnovative media for ion exchange/adsorption of

metals). A listing of design styles of thepresumptive technologies typically consideredduring Superfund remedy selection are listed inAppendix Cl.

3.12 Benefits of Presumptive Technologies.Use of the presumptive technologies identified inthis guidance will simplify and streamline theremedy selection process for me ex-situ treatmentportion of a ground-water remedy by:

• Simplifying the overall selection process,since the large number and diverseassortment of these technologies havebeen reduced to relatively few technologytypes;

• Eliminating the need to perform thetechnology screening portion of thefeasibility study (FS), beyond the analysiscontained in mis guidance and itsassociated Administrative Record. (SeeSection 3.32);

• Allowing, in some cases, furtherconsideration and selection among mepresumptive technologies to be deferredfrom the FS and ROD to the remedialdesign (RD), which prevents duplicationof effort and allows selection to be basedon additional data collected during the RD(see Section 3.3.3);

• Shifting the time and resources employedin remedy selection from ex-situtreatment to other, more fundamentalaspects of the ground-water remedy (seeSection 1.0); and

• Facilitating the use of extraction andtreatment for early actions, whereappropriate, since selection of thetreatment component is simplified.

3.13 Consideration of InnovativeTechnologies. Use of presumptive technologiesfor treatment of extracted ground water isintended to simplify the remedy selection process,

21

but does not preclude the consideration ofinnovative technologies for this purpose in the FSor RD. Refer to the EPA fact sheet. PresumptiveRemedies: Policy and Procedures (EPA, 1993d),for additional information. Many innovative oremerging technologies for ex-situ treatment areactually design variations of one of thepresumptive technology types, as discussed above,and others may be considered on a site-specificbasis. In addition, EPA encourages considerationofin-situ treatment technologies for ground-waterremedies, either when combined with extractionand treatment or as an alternative to such methods(see Section 2.42).

3.1 Basis for Presumptive Technologies

3 J.I Sources of Information. Three sources ofinformation were used to determine whichtechnologies should be identified as presumptivefor ex-situ treatment of ground water:

• Review of me technologies selected in allRODs signed from fiscal years 1982through 1992;

• Review of capabilities and limitations ofex-situ treatment technologies fromengineering and other technical literature;and

• Detailed evaluation of the technologiesconsidered in the FS and selected in theROD or RD for a sample of 25 sites forwhich at least one ex-situ treatmenttechnology was selected.

The above information is summarized in aseparate report entitled Analysis of RemedySelection Results for Ground-Water TreatmentTechnologies at CERCLA Sites (EPA, 1996b). Atotal of 427 RODs selected at least one ex-situtechnology for treatment of ground water, as ofSeptember 30,1992. From these RODs, a sampleof 25 sites were selected for detailed evaluation ofme rationale used to select these technologies aspart of the ground-water remedy.

3.2.2 Rationale for Indentifying PresumptiveTechnologies. At least one of me eightpresumptive technologies, identified in Highlight4, was selected as part of the ground-water remedyin 425 of 427 RODs; or 99.5 percent of the time.m only five RODs were technologies other thanthe presumptive technologies selected as part ofthe treatment train. Therefore, presumptivetechnologies were the only technologies selectedfor ex-situ treatment of dissolved ground-watercontaminants in 420 of the 427 RODs.

More importantly, all the presumptivetechnologies are well understood methods thathave been used for many years m thetreatment of drinking water and/or municipalor industrial wastewater. Engineering Bulletinsor Technical Data Sheets have been developed byEPA and me Naval Energy and EnvironmentalSupport Activity, respectively, for five of me eightpresumptive technologies. These publicationsgenerally include site specific performanceexamples, and are included as references, alongwith other publications, with the description ofeach technology in Appendix D.

In the 25 site sample, the presumptivetechnologies, identified in Highlight 4, were meonly technologies selected in the ROD for all sitesand the only technologies implemented in the RDfor 24 sites. Other technologies were consistentlyeliminated from further consideration, usually inthe technology screening step, based on technicallimitations which were verified by me engineeringliterature. As part of this evaluation the largenumber and diverse assortment of technologiesconsidered for ex-situ treatment of ground waterwere categorized according to the underlyingtreatment process. A complete listing of thetechnologies considered in me FS, ROD or RD forthe 25 sites is given in Appendix Cl, categorizedby process type and with the presumptivetechnologies identified.

Some technologies are identified as presumptiveeven though they were selected in relatively fewRODs. Aeration of background metals wasidentified as presumptive because this technology

22

is often used for removal of iron and manganese,and was considered and selected for this purposeat two of me 25 sample sites. Electrochemicalmethods for metals removal were also identifiedas presumptive because these methods wereconsidered at all three sample sites where metalswere the only contaminants of concern, and wereselected at two of these sites. Chemical/UVoxidation and aerobic biological reactors wereidentified as presumptive technologies for treatingorganic contaminants for the following technicalreasons:

• A range of chemical, physical andbiological treatment methods should beincluded in the presumptive technologies,because air stripping and granularactivated carbon, alone or combined, maynot provide cost effective treatment (seeSection 3.4.5) for all organiccontaminants.

• These methods destroy organiccontaminants as part of (he treatmentprocess instead of transferring mem toother media, which reduces me quantityof hazardous treatment residuals (e.g.,spent carbon) that will require furthertreatment.

• Ongoing research and developmentefforts, by EPA and others, are expectedto increase the cost effectiveness of thesetreatment methods.

33 Remedy Selection Using PresumptiveTechnologies

Selection of technologies for long-term treatmentof extracted ground water requires anunderstanding of me types of technologies matwill be needed, how they will be used in thetreatment system and site-specific information fordetermining the most appropriate and cost-effective technologies. The presumptivetechnologies for treating dissolvedcontaminants in extracted ground water,

Identified in Highlight 4, are the technologiesthat should be retained for fartherconsideration In the Detailed Analysis portionof the feasibility study (FS). This guidance andits associated Administrative Record willgenerally constitute me Development andScreening of Alternatives portion of me FS for theex-situ treatment component of a ground-waterremedy, as discussed in Section 3.3.2.

Site information needed to select cost-effectivetreatment technologies (see Section 3..4) is oftennot collected until the remedial design (RD) phase.In such cases, it will generally be appropriateto specify performance requirements for thetreatment system in the ROD, but deferselection of specific technologies until the RD,as discussed in Section 3.3.3.

3 J.I Use of Technologies in TreatmentSystems. Complete treatment of extracted groundwater generally requires that units of more thanone technology, or multiple units of a singletechnology (unit processes), be linked together ina treatment train. A given treatment train couldinclude some combination of treatmenttechnologies for me following purposes:

1. Separation of mineral solids and/orimmiscible liquids from the extractedground water during initial treatment(pretreatment);

2. Treatment of dissolved contaminants;

3. Treatment of vapor phase contaminantsfrom the extracted ground water or thosegenerated during treatment;

4. Separation of solids generated duringtreatment;

5. Final treatment of dissolvedcontaminants prior to discharge(polishing); and

23

6. Treatment of solids generated duringtreatment.

Presumptive technologies for treatment ofdissolved contaminants in extracted groundwater (No. 2 and 5, above) are identified inHighlight 4. Examples of the types oftechnologies used for other purposes are given inAppendix C2, along with a listing of the generalsequence of unit processes used in a treatmenttrain. Solid residuals (such as sludges fromchemical or biological processes, or spent carbonmedia) will generally require additional treatmentor disposal, either as part of me treatment train orat a separate facility. Presumptive technologiesfor purposes other than for treatment of dissolvedcontaminants have not been identified in thisguidance.

Use of modular treatment components, which canbe easily added or removed from me treatmentsystem, may facilitate phased implementation orother changes that may occur during the life of aremedy. Phased implementation of the extractionportion of a remedy may require that somecomponents of the treatment system also beinstalled in stages. Also, modification of thetreatment system over time may be needed inresponse to changes in me inflow rate orcontaminant loadings, or to increase theeffectiveness or efficiency of me treatment system.

3.3.2 This Guidance Constitutes the FSScreening Step. This guidance and its associatedAdministrative Record will generally constitutethe "development and screening of alternatives"portion of the feasibility study (FS), for me ex-situtreatment component of a ground-water remedy.When using presumptive technologies, the FSshould contain a brief description of this approach(see fact sheet entitled Presumptive Remedies:Policy and Procedures (EPA, 1993d)), and referto this guidance and its associated AdministrativeRecord. Such a brief description should fulfill meneed for the development and screening oftechnologies portion of me FS for the ex-situtreatment component of me remedy.

3JJ Deferral of Final Technology Selection toRD. Although EPA prefers to collect the siteinformation needed for technology selection priorto the ROD, it is sometimes impracticable tocollect some of the necessary information until theremedial design (RD) phase. (See Section 3.4 fora summary of site information generally neededfor selection of these technologies.) m reviewingremedy selection experience for a sample of sites,EPA found that at seven of 25 sites (28 percent)the type of technology selected in me ROD fortreatment of extracted ground water was laterchanged in me RD because of additional siteinformation obtained during the design phase(EPA, 1996b). Where EPA lacks importantinformation at me ROD stage, it may beappropriate to defer final selection among mepresumptive ex-situ treatment technologies (aswell as selection of specific design styles) to theRD phase.

In mis approach, EPA would identify and evaluatethe technologies and provide an analysis ofalternative technologies in the FS (this guidanceand its associated administrative record willgenerally constitute that discussion). Theproposed plan would identify me technologies thatmay be finally selected and specify the timing ofand criteria for the future technology selection insufficient detail that the public can evaluate andcomment on the proposal. The ROD would alsoidentify all ARARs and other performancespecifications and information associated withdischarge and treatment of the extracted groundwater, including me types of discharge, effluentrequirements, and specifications developed inresponse to community preferences. Specifyingthe performance criteria and other requirements inthe ROD (using a type of "performance basedapproach") ensures that the remedy will beprotective and meet ARARs. Overall, the RODshould be drafted so mat the final selection oftechnologies at the RD phase follows directlyfrom the application of criteria and judgmentsincluded in the ROD to facts collected during theRD phase. If the ROD is drafted in mis fashion,documenting me final technology selection cangenerally be accomplished by including a

24

document in fee post-ROD portion of feeAdministrative Record, which explains fee basisof technology selection (e.g.. Basis of DesignReport, or memorandum to fee RD file).

Advantages of deferring selection ofex-situtreatment technologies to fee RD include:

• The remedy selection process is furtherstreamlined, since final selection and feeaccompanying detailed analysis for thesetechnologies is performed only in fee RDnot in both fee FS and fee RD,minimizing duplication of effort;

• Site information collected during fee RDcan be used to make final technologyselections as well as to design feetreatment train, which facilitates selectionof fee most cost effective technologies(see Section 3.4.5);

• The likelihood feat changes in feetreatment train will be made during feeRD is explicitly recognized in fee ROD;and

• The time and resources employed in feeFS can focus on other components of feeground-water remedy feat have moredirect influence on attainment ofremedial objectives for contaminatedground water (see Section 1.0).

Cost estimates for remedial alternatives,including fee ex-situ treatment component, willneed to be included in fee FS regardless ofwhether or not technology selection is deferred totheRD. For cost estimating purposes whendeferring technology selection to fee RD,reasonable assumptions should be madeconcerning fee treatment system, includingassumptions concerning fee presumptivetechnologies and likely design styles to be used.To assist in making such assumptions, advantagesand limitations for fee presumptive technologiesare summarized in Appendix C4. Also, briefdescriptions of fee presumptive technologies and

references for additional information are providedin Appendix D. Assumptions used for estimatingtreatment costs should be consistent across allremedial alternatives. All assumptions should beclearly stated as such in fee FS and ROD.

Example ROD language for deferring technologyselection to fee RD is given in Appendix B3 for ahypothetical site. This language is only for fee ex-situ treatment portion of an extraction andtreatment remedy and should appear in feeselected remedy portion of fee ROD whenfollowing this approach.

3.4 Information Needed for SelectingTechnologies

The site information listed in Highlight 5 isgenerally needed to determine fee treatmentcomponents of a complete treatment train forextracted ground water and to select fee mostappropriate technology type and design style foreach component. Further detail regarding site dataneeded and fee purpose of this information isprovided in Appendix C3. Much of thisinformation is also needed for design of feeextraction component of an extraction andtreatment remedy.

3.4.1 When Should this Information beCollected? The information listed in Highlight 5is needed for design of fee treatment train.Therefore, it must be collected prior to or duringfee design phase, for either an early action or long-term remedy. Much of this information shouldalso be available for selecting among feepresumptive technologies, since it is generallyneeded to determine fee technologies mostappropriate for site conditions. The timing ofinformation needed during remedy selection isdifferent when deferring technology selection tofee RD than when selecting technologies in feeROD, as discussed in Section 3.3.3. However,much of this information can be collected alongwife similar data gathered during fee remedialinvestigation (RI). m general, it is recommendedfeat as much of this information as possible beobtained prior to fee RD in order to minimize fee

25

Highlight 5. Summary of Site InformationNeeded For Treatment Train Design

• Total extraction flow rate

• Discharge options and requirements

• Target effluent concentrations

Contaminants

Degradation products

Treatment additives

Natural constituents

• Ottier requirements

Regulatory

Operational

• Community concerns orpreferences

• Water quality of treatment influent

• Contaminant types andconcentrations

• Naturally occurring constituents

• Other water quality parameters

• Treatability information

NOTE: Further detail Is provided in AppendixC3.

need for additional site investigations during theRD and to accelerate the RD phase.much of this information can be collected alongwith similar data gathered during the remedialinvestigation (RI). In general, it is recommendedthat as much of this information as possible beobtamed prior to the RD in order to minimize theneed for additional site investigations during meRD and to accelerate the RD phase.

3.4.2 Extraction Flow Rate. Inflow to thetreatment system is the total flow from allextraction wells or drains. Estimates of totalextraction flow rate often have a high degree ofuncertainty (i.e., one or more orders ofmagnitude), depending on type of data andestimation method used. Expected flow ratesfrom extraction wells are typically estimated fromhydraulic properties of the aquifer. Aquiferhydraulic properties may have considerablenatural variation over the site and accuratemeasurement of these properties is often difficult.In order to reduce uncertainty during design of thetreatment system, aquifer properties used inestimating the inflow should generally beobtained from pumptng-type aquifer tests andnot from "slug tests," laboratory measurements onborehole samples or values estimated from theliterature.

Pumping-type aquifer tests provide a much betterestimate of average aquifer properties than othermethods, because a much larger volume of aquiferis tested. For the same reason, ground waterextracted during pumping tests is morerepresentative of that which will enter thetreatment system, and should generally be used fortreatability studies ofex-situ treatmenttechnologies instead of samples obtained frommonitoring wells. Suggested procedures forconducting pumping-type aquifer tests are givenin EPA, 1993L Methods for treatment ofcontaminated ground water extracted duringpumping-type aquifer tests are discussed inSection 3.5.

The likely variability in the total extraction rateduring the life of the remedy should also beestimated. Variability in the extraction rate couldresult from addition or removal of extractionwells, short-term operational changes in thesystem (e.g., changing me pumping rates) orseasonal fluctuations in me water table. Thenumber of extraction wells could change as aresult of implementing the remedy in phases orfrom post-construction refinement of me remedy(see Section 2.3.1).

26

3A3 Discharge Options and ARARs. Alloptions for discharge of ground water afterextraction and treatment should be identified andconsidered in the FS, especially options matinclude re-use or recycling of the extracted groundwater. Water quality requirements for the treatedeffluent (i.e., effluent ARARs) may be differentfor each discharge option. Examples of regulatoryrequirements include those promulgated under thefederal Safe Drinking Water Act and Clean WaterAct, which would apply to discharges to adrinking water system or to surface waters,respectively; and state requirements for thesetypes of discharge. Effluent requirements couldalso include those for chemicals added duringtreatment, contaminant degradation products, andnaturally occurring constituents (e.g., arsenic), inaddition to those for contaminants of concern. Ingeneral, one or more types of discharge forextraction and treatment remedies should beselected in the ROD, not deferred to the RD.ARARs for me treated effluent will determine theoverall level of treatment needed, which in turndetermines the type of components needed in thetreatment train (see Section 3.3.1) and is a criticalfactor in selecting appropriate treatmenttechnologies.

In some cases it may be appropriate to select morethan one type of discharge for the selected remedy.One type of discharge may be preferred, but maynot be capable of accepting the entire flow oftreated effluent. For example, it may be possibleto re-use or recycle a portion but not all of (hedischarge. It may also be desirable to reinject aportion of the treated effluent for enhancedrecovery of contaminants (aquifer flushing) butprohibitively costly to reinject the entire discharge.

In addition to the types of discharge, ARARsand other specifications related to technologyselection or operating performance of thetreatment system should be specified in theROD. Regulatory requirements for all wastestreams from the treatment system should bespecified, including those for me treated effluent;releases to the air; and those for handling,treatment and disposal of solid and liquid

treatment residuals. Other specifications couldinclude those preferred by the affected community,such as requirements to capture and treatcontaminant vapors (even though not required byARARs) or limits on operating noise. Otherspecifications may also be needed to maintaincontinued operation of the system, such as waterquality conditions necessary to minimize chemicaland/or biological clogging of injection wells ordrains.

3.4.4 Water Quality of Treatment Influent. Inorder to design me treatment system, contaminanttypes and concentrations and other water qualityparameters must be estimated for the total flowentering the system. Since some technologies aremore effective than others in removing certaincontaminant types, this is an important technologyselection factor. Concentrations of naturallyoccurring constituents as well as background andsite-related contaminants in the extracted groundwater should also be measured, as discussed inAppendix C3.

3.4.5 Treatability Studies. Treatability studiesinvolve testing one or more technologies in melaboratory or field to assess their performance onthe actual contaminated media to be treated from aspecific site. These studies may be needed duringthe RI/FS to provide qualitative and/orquantitative information to aid in selection of theremedy, or during the RD to aid in design orimplementation of me selected remedy. Threetiers of testing may be undertaken: 1) laboratoryscreening, 2) bench-scale testing, or 3) pilot-scaletesting. Treatability studies may begin with anytier and may skip tiers that are not needed (EPA,1989c).

For treatment of extracted ground water,treatability studies are generally needed toaccurately predict the effectiveness and total costof a technology for a given site, includingconstruction and operating costs; and me costs ofother components mat may be needed in thetreatment train (see Section 3.3.1). Optimizingthe cost effectiveness of the treatment train isespecially important for systems designed to

27

operate over a long time period. (In this guidance,optimizing the cost effectiveness of the treatmentsystem is denned as meeting all treatment andother performance requirements while minimizingtotal costs per unit volume of water treated.)Treatability studies may also indicate that sometechnologies provide cost effective treatment whenall of the above factors are considered, eventhough these technologies were infrequentlyselected in past RODs (e.g.» chemical/UVoxidation or aerobic biological reactors). Forthese reasons treatability studies will be helpful inselecting among the presumptive technologies.Similarly, a presumptive treatment technologyshould not be eliminated from furtherconsideration in the FS or RD simply because atreatability study is required to determine itsapplicability for a given site. In general, sometype of treatability study should be performedprior to or during the design of any systemexpected to provide long-term treatment ofextracted ground water, including systems usingpresumptive technologies.

3.5 Treatment Technologies for Aquifer Tests

Although pumping-type aquifer tests are (hepreferred method of determining average aquiferproperties (see Section 3.4.2) and this informationis useful for remedy selection, such testing is oftendeferred to the RD phase because of the need todetermine how to treat and/or dispose of theextracted ground water. To facilitate use of suchtests earlier in the site response, ex-situ treatmenttechnologies most suitable for (his application arediscussed below.

35.1 Treatment Needs during Aquifer Tests.In comparison to an extraction and treatmentremedy, pumping-type aquifer tests (see Section3.4.2) generate relatively small flows ofcontaminated ground water over a short period oftime. At the time of such tests, me estimatedpumping rates and contaminant loadings generallyhave a high degree of uncertainty. Often the totalvolume of ground water extracted during testing isheld in storage tanks or lined ponds to prevent thedischarge from affecting water levels in

observation wells and interfering with the testStorage of (he extracted ground water also allowssubsequent flow to a treatment system to becontrolled and optimized. For example, if storagevessels are used for both the untreated and treatedwater, me extracted water can be routed throughthe treatment system as many times as necessaryto meet discharge and/or disposal requirements.Therefore, the cost effectiveness of treatmenttechnologies (see Section 3.4.5) is less importantfor aquifer testing than for me long-term remedy,because of the much smaller volume of groundwater to be treated and the much shorter period ofoperation.

3.5.1 Treatment Technologies for AquiferTests. Technologies for treating ground waterextracted during aquifer tests should be able totreat a wide range of contaminant types, beavailable in off-the-shelf versions (short lead timefor procurement), have a short on-site startuptime, be relatively simple to operate, and beavailable in easily transportable units. Of thepresumptive technologies identified above, thethree most suitable for this application are:

• Granular activated carbon,

• Air stripping, and

• Ion exchange/adsorption.

Granular activated carbon can effectively removemost dissolved organic contaminants and lowconcentrations of some inorganic compounds. Ionexchange/adsorption can remove most metals. Airstripping may be applicable for volatile organiccontaminants (VOCs) and generally is more costeffective than granular activated carbon fortreating VOCs when flow rates are greater thanabout three gallons per minute (Long, 1993).Granular activated carbon may still be needed inconjunction with air stripping, for treatingdissolved semivolatile organic contaminants, orfor reaching stringent effluent requirements forVOCs. Granular activated carbon may also beneeded for treatment of vapor phase contaminantsseparated by an air stripper. Also, treatability

28

studies generally are not required for the abovethree technologies, especially for short-termapplications. Additional information regardingthe availability and field installation of skid ortrailer mounted treatment units (package plants) isavailable in EPA, 1995a.

Other presumptive ex-situ treatment technologies(chemical/UV oxidation, aerobic biologicalreactors, chemical precipitation, andelectrochemical methods) generally are lesssuitable for aquifer testing purposes. In general,these other technologies require longer lead timesfor procurement and longer time on-site forstartup; and have more complex operatingrequirements and higher capital costs.

4.0. REFERENCES

Cohen, R.M., and J.W. Mercer, 1993. DNAPLSite Evaluation. CJC Smoley, Boca Raton, FLand ORD Publication EPA/600/R-93/022.

EPA, 1986. "Guidelines for Ground-WaterClassification Under me EPA Ground-WaterProtection Strategy, Final Draft," November,1986.

EPA, 1988a. "Guidance for Conducting RemedialInvestigations and Feasibility Studies UnderCERCLA, Interim Final," OSWER Directive9355.3-01, EPA/540/G-89/004, October 1988.

EPA, 1988b. "Guidance on Remedial Actions forContaminated Ground Water at Superfund Sites,"OSWER Directive 9283.1-2, EPA/540/G-88/003,December 1988.

EPA,1989a. "Considerations in Ground WaterRemediation at Superfund Sites," OSWERDirective 9355.4-03, October 18,1989.

EPA, 1989b. "Interim Final Guidance onPreparing Superfund Decision Documents,"OSWER Directive 9335.3-02, October 1989.

EPA, 1989c. "Guide for Conducting TreatabilityStudies Under CERCLA, Interim Final,"OERR/ORD Publication EPA/540/2-89/058,December 1989.

EPA, 1990a. "ARARs Q's & A's: State Ground-Water Antidegradation Issues," OSWERPublication 9234.2-11/FS. July 1990.

EPA, 1990b. "Suggested ROD Language forVarious Ground Water Remediation Options,"OSWER Directive 9283.1-03, October 10,1990.

EPA, 1991a. "Guide to Addressing Pre-ROD andPost-ROD Changes," OSWER Publication9355.3-02FS-4, April 1991.

EPA, 1991b. "Guide to Developing SuperfundNo Action, Interim Action, and ContingencyRemedy RODs," OSWER Publication 9355.3-02FS-3, April 1991.

EPA, 1991c. "Guide to Principal Threat and LowLevel Threat Wastes," OSWER Publication9380.3-06FS, November 1991.

EPA, 1992a. "Estimating Potential forOccurrence of DNAPL at Superfund Sites,"OSWER Publication, 9355.4-07FS, January1992.

EPA, 1992b. "Considerations in Ground-WaterRemediation at Superfund Sites and RCRAFacilities - Update," OSWER Directive 9283.1-06, May 27,1992.

EPA, 1992c. "Guidance on Implementation of theSuperfund Accelerated Cleanup Model (SACM)under CERCLA and the NCP," OSWER Directive9203.1-03, July 7,1992.

EPA, 1992A "The Superfund AcceleratedCleanup Model (SACM)," OSWER Publication9203.1-021, November 1992.

EPA,1992e. "Early Action and Long-TermAction Under SACM - Interim Guidance,"OSWER Publication 9203.1-051, December 1992.

29

EPA, 1992f. "SACM Regional Decision Teams -Interim Guidance," OSWER Publication 9203.1-051, December 1992.

EPA,1992g. "Evaluation of Ground-WaterExtraction Remedies: Phase n. Volume 1Summary Report," OSWER Publication 9355.4-05, February 1992.

EPA, 1992h. "Final Comprehensive State GroundWater Protection Program Guidance," PublicationEPA 100-R-93-001, December 1992.

EPA, 1993a. "Guidance on Conducting Non-Time-Critical Removal Actions Under CERCLA,"OSWER Publication 9360.0-32, EPA/540-R-93-057, August 1993.

EPA,1993b. "Guidance for Evaluating TechnicalImpracticability of Ground-Water Restoration,"OSWER Directive 9234.2-25, EPA/540-R-93-080, September 1993.

EPA, 1993c. "Evaluation of me Likelihood ofDNAPL Presence at NPL Sites, National Results,"OSWER Publication 9355.4-13, EPA/540-R-93-073, September 1993.

EPA, 1993d. "Presumptive Remedies: Policy andProcedures," OSWER Directive 9355.0-47FS,EPA/540-F-93-047, September 1993.

EPA,1993e. "Presumptive Remedies: SiteCharacterization and Technology Selection ForCERCLA Sites With Volatile OrganicCompounds In Soils," OSWER Directive 9355.0-48FS, EPA/540-F-93-048, September 1993.

EPA, 1993f. "Presumptive Remedy for CERCLAMunicipal Landfill Sites," OSWER Directive9355.0-49FS, EPA/540-F-93-035, September1993.

EPA, 1993g. "Innovative TreatmentTechnologies: Annual Status Report (FifthEdition)", Publication EPA 542-R-93-003,September 1993.

EPA,1993h. "m-situ Treatment of Contaminants:An Inventory of Research and FieldDemonstrations and Strategies for ImprovingGround Water Remediation," OSWERPublication EPA/500/K-93/001, January 1993.

EPA, 1993L "Ground Water Issue, SuggestedOperating Procedures for Aquifer PumpingTests," OSWER Publication EPA/500/S-93/503,February 1993.

EPA,1993j. "Data Quality Objectives Processfor Superfund, Interim Final Guidance" OSWERPublication 9355.9-01, EPA/540/R-93/071,September 1993.

EPA, 1994a. "Alternative Methods for FluidDelivery and Recovery," ORD/CERI PublicationEPA/625/R-94/003, September 1994.

EPA, 1994b. "DNAPL Site Characterization,"OSWER Publication 9355.4-16FS, EPA/540/F-94/049, September 1994.

EPA,1994c. "RCRA Corrective Action Plan,"OSWER Directive 9902.3-2A, EPA/520/R-94/004, May 1994.

EPA, 1994A "Technical Support Project, DirectTechnical Assistance for Site Remediation,"OSWER Publication EPA/542-F-94/004, October1994.

EPA,1994e. "Methods for Monitoring Pump-and-Treat Performance," ORD PublicationEPA/600/R-94-94/123, June 1994.

EPA,1995a. "Manual: Ground Water andLeachate Treatment Systems," ORD/CERIPublication EPA/625 R-94/005, January 1995.

EPA,1995b. "Consistent Implementation of theFY 1993 Guidance on Technical Impracticabilityof Ground-Water Restoration at Superfund Sites,"OSWER Directive 9200.4-14, January 19,1995.

30

EPA,1995c. "SuperfundGroundwaterRODs:Implementing Change This Fiscal Year," OSWERMemorandum from EUiott P. Laws to RegionalAdministrators and others, July 31,1995 (nopublication number).

EPA,1995d. "In-Situ Remediation TechnologyStatus Report: Treatment Walls," OSWERPublication EPA/542 K-94-004, April 1995.

EPA, 1995e. "Innovative TreatmentTechnologies: Annual Status Report (SeventhEdition);' OSWER Publication EPA-542-R-95-008 Number 7, Revised, September 1995.

EPA,1995f. "Soil Vapor Extraction (SVE)Enhancement Technology Resource Guide,"OSWER Publication EPA/542-B-95-003,October 1995.

EPA, 1995g. "Presumptive Remedies for Soils,Sediments and Sludges at Wood Treater Sites,",OSWER Directive 9200.5-162, EPA/540-R-95/128, December 1995.

EPA,1996a. "Consideration of'ComprehensiveState Ground Water Protection Programs' by EPARemediation Programs," Draft OSWER Directive9283.1-09 dated June 1996. Final Directiveexpected by November 1996.

EPA, 1996b. "Analysis of Remedy SelectionExperience for Ground Water TreatmentTechnologies at CERCLA Sites," Draft FinalReport dated July 1996. Final Report expected byNovember 1996.

EPA,1996c. "Use of Natural Attenuation atSuperfund, RCRA Corrective Action, andUnderground Storage Tank Sites," Draft OSWERDirective__dated September 1996. FinalDirective expected by February 1997.

Federal Register, 1990a. Volume 55, No. 46,March 8,1990; 40 CFR Part 300, "National Oiland Hazardous Substances Pollution ContingencyPlan; Final Rule" (NCP).

Federal Register, 1990b. Volume 55, No. 145,July 27,1990; 40 CFR Parts 264,265,270 and271, "Corrective Action for Solid WasteManagement Units at Hazardous Waste Facilities;Proposed" (proposed Subpart S regulations).

Federal Register, 1996. Volume 61, No. 85, May1,1996; "Corrective Action for Releases fromSolid Waste Management Units at HazardousWaste Management Facilities, Advance Notice ofProposed Rulemaking."

Long, G. M., 1993. "Clean up HydrocarbonContamination Effectively," ChemicalEngineering Progress, Vol. 89, No. 5.

National Research Council, 1994. Alternativesfor Ground Water Cleanup. National AcademyPress, Washington, DC.

31