»3 environmtrtiai rrotocuon Aatncy
                 Washington. DC 20460
OSWER Directive Initiation Request
                   1. Directive Number

                                 2. Originator Information
     Name of Contact Person
        John Z. Cannon
                  Mail Cod*
Telephone Code
      3. Title
         Considerations in Ground Water Remediation at Superfund Sites
      4. Summary of Directive (include brief statement of purpose)
        The purpose of this memorandum is  to transmit our findings from a recently
        completed study of several sites where ground water extration is being conducted
        to contain or reduce levels of contaminants in the ground water.  In addition,
        this memorandum presents several recommendations for modifying the Superfund
        approach to ground water remediation.	
      S. Keywords
      6a. Does THIS Directive supenseae previous uirecuve^sjy
      b. Does It Supplement Previous Ofrectrve(s)?
                       |   | No   [   I Yea    What directive (number. We)

                                  Yes    What elective (number. We)
      7. Draft Level
          A-SIgnedbyAA/DAA    [  [ B - Signed by Office Director      C - For Review 4 Comment   |   | D - Hi Development
Document to
be distributed
*-t .
by Headquarters?
This Request Meets OSWER Directives System Format Standards.
9. Signature of Lead Office Directives Coordinator
Betti VanEpps
10. Name and Title of Approving Official
John Z. Cannon ^.
     EPA Form 1315-17 (Rev. 5-17) Previous editions are obsolete.
   OSWER          OSWER               OSWER              O


                      WASHINGTON DC  20460
                           OCT I  8 1989
                                 Directive No.  9355.4-03
Considerations in Ground Water Remediation at Superfur.d
Sites   /
Jonathan*2. cannon
Acting Assistant Administrator

Waste Management Division Directors
   Regions I, IV, v, VI, vii, viii
Emergency and Remedial Response Division Director
   Region II
Hazardous Waste Management Division Directors
   Region III, IX
Hazardous Waste Division Director
   Region X

     The purpose of this memorandum is to transmit our findings
from a recently completed study o"f several sites where ground
water extraction is being conducted to contain or reduce levels c:
contaminants in the ground water.  In addition, this memorandum
presents several recommendations for modifying the Superfund
approach to ground water remediation.


     The most common method for restoring contaminated ground
water is extraction and treatment of contaminated ground water.
Recent research has suggested that in many cases, it may be more
difficult than is often estimated to achieve cleanup concentration
goals in ground water. In response to these findings, the Office  zi
Emergency and Remedial Response  (OERR) initiated a project to
assess the effectiveness of ground water extraction systems in
achieving specified goals.  Nineteen case studies were developed
from among Superfund and State-lead sites, RCRA and Federal
facilities.  These sites were selected primarily on the basis that
the ground water extraction systems had been operating for a period
of time sufficiently long to allow for an evaluation of the syster..


     The objective of this memorandum is to describe the findings
 cf  this study and to recommend the consideration of certain
 factors and approaches in developing and implementing ground vater
 response actions at Superfund sites.

 Findings of Study

     Several trends were identified from the case studies:

     o   The extraction systems are generally effective in
         containing contaminant plumes,  thus preventing further
         migration of contaminants.

     o   Significant mass removal of contaminants (up to 130,CCC
         pounds over three years) is being achieved.

     o   Concentrations of contaminants have generally decreased
         significantly after initiation of extraction but have
         tended to level off after a period of time.  At the sites
         examined, this leveling off usually began to occur at
         concentrations above the cleanup goal concentrations
         expected to have been attained at that particular point
         in time.

     o   Data collection was usually not sufficient to fully
         assess contaminant movement and system response to

     Several factors appear to be limiting the effectiveness c:
the extraction systems examined, including:

     o   Hydrogeological factors, such as the heterogeneity of the
         subsurface, the presence of low permeability layers, and
         the presence of fractures;

     o   Contaminant-related factors, such as sorption to the
         soil, and presence of non-aqueous phase liquids
         (dissolution from a separate non-aqueous phase or
         partitioning of contaminants from the residual non-
         aqueous phase);

     o   Continued leaching from source areas;

     o   System design parameters,  such as pumping  rate, screened
         interval, and location of  extraction wells.

     The report summarizing the study and  findings, entitled
Evaluation of Ground Water Extraction Remedies is attached.
Additional copies of the report are available through the  Public
Information Center ((202) 382-2080) or the Center for

 Environmental Research  Information  (FTS 634-7391 or (513) 569-
 7391) .

 P.ecc-r.endat ions

      The  findings of the study substantiate previous research and
 confirm .that ground water remediation is a very new, complex
 field.  Based on this study, I am recommending consideration cf
 certain factors and approaches in developing and implementing
 ground water response actions.  The major recommendation is to
 orient our thinking so  that we initiate early action on a small
 scale, while gathering  more detailed data prior to committing to
 full-scale restoration.  These recommendations are consistent
 with  the  Guidance on Remedial Actions for Contaminated Ground
 Water at  Superfund Sites and do not alter Superfund's primary
 goal  of returning ground water to its beneficial uses in a ti.-?.e
 frame that is reasonable given the particular site circumstances.
 The recommendations do, however, encourage the collection of data
 to allow  for the design of an efficient cleanup approach that
 more  accurately estimates the time frames required for
 remediation and the practicability of achieving cleanup goals.

     While standard procedures for the more refined data
 collection techniques suggested below are being developed, it
 will be beneficial at most sites to implement the ground water
 remedy in stages.  This might consist of operating an extractirr.
 system on a snail scale that can be supplemented incrementally as
 information on aquifer  response is obtained.

     These recommendations are described further below.  The
 attached  flow chart illustrates how the recommendations  fit ir.tc
 the Superfund ground water response process.

 Recommendation 1:  Initiate Response Action Early.

     The bias for action should be considered early in the site
 management process. Response measures may be implemented to
 prevent further migration of contaminants if they will prevent
 the situation from getting worse, initiate risk reduction, and/or
 the operation of such a system would provide information useful
 to the design of the final remedy.  Because the data needed to
 design * ground water containment system are often more  limited
 than that needed to implement full  remediation, it will  in a
 number of cases be possible and valuable to prevent the
contaminant plume from  spreading while the investigation to
 select the remediation  system progresses.  The determination  of
whether to implement a  containment  system should be based  on
existing  information, data defining the approximate plume
boundaries, hydrologic  data, contaminants present,  and
 approximate concentrations, and best professional  judgment.
 Examples of situations  where this type of action will probably  be
 warranted include sites where ground water plumes  are migrating

rapidly (e.g.,  highly permeable  aquifers,  mobile contaminants,
potential migration thorough  fractures)  and  sites near drir.kir.g
water wells that are potentially affected  by the plume.
                 Decision (ROD)  for  an  interim  remedy may be
                limited evaluation of alternatives  that comoars
      A  Record  of
 prepared  with  a lx.oi'_eu «va.j.u
 the  remedy nay reveal  that  it  is technically impracticable to
 achieve health-based concentrations throughout the area of
 attainment,  and that another remedy or a contingent remedy may be

      '•v'here sufficient  information is available to specify an
 alternative  or contingent remedy at the time of remedy selection,
 the  ROD should discuss  the  contingency in equal detail to the
 primary remedial  option, and should provide substantive criteria
 by which the Agency will decide whether or not to implement the
 contingency.   See Interim Final Guidance on Preparing Superfund
 Decision Documents, OSWER Directive 9355.3-02 (May 1989), at page
 9-17.1   The  ROD may also discuss the possibility that an ARARs
 waiver  will  be invoked  when MCLs or other Federal or State
 standards  cannot  practicably be attained in the ground water; a
 written  waiver finding  should  be issued at the time the
 contingency  is invoked, or  in  limited circumstances, in the ROD

     The public should  be informed of the decision to invoke the
 contingency  (and,  perhaps,  the waiver) through issuance of an
 Explanation  of Significant  Differences (ESD) which involves a
 public notice.  A formal public comment period is not retired
 when a decision is made to  invoke a contingency specified in the
 ROD; however,  the Region may decide to hold additional public
 comment periods pursuant to NCP section 300.825(b) (proposed)
 (Dec. 21,  1988, 53FR at 51516).  In any event, the public may
 submit comments after ROD signature on any significant new
 information  which "substantially support[s] the need to
 significantly  alter the response action" NCP Section 300.825(c)

     There may also be  situations where the Region finds that it
 is impracticable  to achieve the levels set out in the ROD, but no
contingency  had been previously specified in the ROD.  In such
cases, a ROD amendment  would be necessary to document fundamental
changes that are  made in the remedy based on the information
gained during  implementation;  an ESO would be necessary to
     1  idr instance, the ROD may provide that a contingent
remedy will be implemented if there  is a levelling-off of
contaminant concentrations despite continued ground water
extraction over a stated period of time.

     2  It may be possible to invoke a waiver at the  time  of  ROD
signature (a "contingent waiver") where, for example, the  ROD is
detailed and establishes an objective level or situation at which
the waiver would be triggered.  However, the use of contingent
waivers should only be considered on a case-by-case basis  after
discussion with OERR\OWPE.

 document  significant but non-fundamental changes in the remedy
 based  on  the additional information.

     For  sites where there is substantial uncertainty regarding
 the  ability of the remedy to return the ground water to its
 beneficial uses  (e.g., dense non-aqueous phase liquids in
 fractured bedrock) it  is appropriate to indicate that the initial
 action  is interim with an ultimate remedy to be determined at
 some specified future date.  The action should be designed to
 achieve the basic goal and carefully monitored over time to
 determine the feasibility of achieving this goal.  In many of
 these cases, this can only be determined after several years of
 operation.  The  five year review may be the most appropriate tine
 to make this evaluation.  When sufficient data have been
 collected to specify the ultimate goal achievable at the site
 (e.g.,  first or  second five year review), a final ROD for ground
 water would be prepared specifying the ultimate goal, including
 anticipated time frame, of the remedial action.

     Although overall system parameters must be specified in the
 ROD, it is usually appropriate to design and implement the ground
 water response action as a phased process.  An iterative process
 of system operation, evaluation, and modification during the
 construction phase can result in the optimum system design.
 Extraction wells might be installed incrementally and observed
 for one to three months to determine their effectiveness.  This
 will help to identify appropriate locations for additional wells
 and can assure proper sizing of the treatment systems as the
 range of contaminant concentrations in extracted ground water is

     If it is determined that some portion of the ground water
within the area of attainment cannot be returned to its
beneficial uses,  an evaluation of an alternate goal for the
ground water should be made.  Experience to date on this phase of
ground water remediation is extremely limited and more definitive
guidance on when to terminate ground water extraction will be
provided later.  When the point at which contaminant
 concentrations in'ground water level off, however, this should be
viewed a* a signal that some re-evaluation of the remedy is
warranted.  In many cases, operation of the extraction system on
 an intermittent basis will provide the most efficient mass
 removal.  This allows contaminants to desorb from the soil in the
saturated zone before ground water is extracted providing  for
maximum removal of contaminant mass per volume of ground water

     Ground water monitoring should continue for two to three
years after active remediation measures have been completed  to
 ensure that contaminant levels do not recover. For cases where
 contaminants remain above health-based levels, reviews to  ensure

 that  protection  is  being  maintained at the site will take place
 at  least  every five years.

 Recommendation 3:   Collect Data to Better Assess Contaminant
      Movement and Likely  Response of Ground Water to Extraction.

      In addition to the traditional plume characterization data
 normally  collected,  the following data is of particular
 importance to .the design  and evaluation of ground water remedies
 and should be considered  in scoping ground water RI/FSs.
 Assessments of contaminant movement and extraction effectiveness
 can be greatly enhanced by collecting more detailed information
 on vertical variations in stratigraphy and correlating this to
 contaminant concentrations in the soil during the remedial
 investigation.  More frequent coring during construction of
 monitoring wells and the  use of field techniques to assess
 relative  contaminant concentrations in the cores are methods that
 may be used to gain this  information.  More detailed analysis of
 contaminant sorption to soil in the saturated zone can also
 provide the basis for estimating the time frame for reducing
 contaminant concentrations to established levels and identifying
 the presence of non-aqueous phase liquids. Cores taken from
 depths where relatively high concentrations of contaminants were
 identified might be analyzed to assess contaminant partitioning
 between the solid and aqueous phases.  This might involve
measuring the organic carbon content and/or the concentration of
the contaminants themselves.

     The  long-term  goal is to collect this information during the
RI so that more definitive decisions can be made at the ROD
stage.  Standardized sampling and analytical methods to support
these analyses are  currently being evaluated.

     For  further information, please consult the appropriate
Regional Ground Water Forum member, Jennifer Haley at
FTS 475-6705 or Caroline  Roe at FTS 475-9754 in OERR's Hazardous
Site Control Division, or Dick Scalf at the Robert S. Kerr
Environmental Research Laboratory (FTS 743-2308)

Attachment:  Flow Chart
       vqj     Summary Report

cc: Superfund Branch Chiefs, Regions I - X
    Superfund Section Chiefs, Regions I - X wo/summary  report

     Scoping ol
1 *

Site 1
Ml ,.«•» 1 1
Very Hiqtw

                                                                                                                           I iisl 01 S««>nd
                                                                                                                            !> YI.-.II Huview
                                                                                            •;te»|  I nidi Action
                                                               Associated with
                                                               Ralurning Ground
                                                                 Water to its
                                                               Beneficial Uses
   •  ktontty d*to ootocton
   •  ktenttypi
Inttoll gradwnl control wells
in phased process
Monrtor aqurfer response
•  Design and Implement ground
  water extr acton system in
  phased process
•  Monitor aquifer response
  (. v.iluate data bom syslitm
  Determine practicable goals
  Idunbfy any areas whare
  long letm institutional
  contiols wiH bo nucossaiy
  Admlntotntlv* Contktontlons:
                          ROD (Early Acton)
                                                          A) ROD (Inlehm Remedy)
                                                          B). ROD (Contingency)
                                                          C) ROD (Final)
A) ROD (Final)
  Enforcement Considerations:
   Negottele Rl^FS Scope:
   • Data collection
   • Earty acton
                                                          Nogohalu Consent Decree
A). Negohale Consent Owciue
B) Possible stipulation 01
   amondmonl to Consent