j
i

 o
 T
.0*
   Performance Monitoring of
   MNA Remedies for VOCs in
   Ground Water
                                   -."=
                                  \
   Source
    Area
             Low
          Concentration
                      Non-Hazardous
                      Degradation
                     Products & Other
   High    Plume Fringe Geochemrca| |ndkators
Concentration
 Plume Core       Ground-Water Flow

-------
 Office of Research and
 Development
 Washington, DC 20460
Office of Superfund
Remediation and
Technology Innovation
Washington, DC 20460
OSWER 9355.4-25
PB 2003 103270
EPA/540/R-03/004
September 2003
Performance Monitoring of
     MNA Remedies for
   VOCs in Ground Water

-------

-------
                                  CONTENTS
ACRONYMS AND ABBREVIATIONS	vi
FOREWORD	vii
NOTICE AND DISCLAIMER	viii
ACKNOWLEDGMENTS	ix
ABSTRACT	x

CHAPTER 1     INTRODUCTION	1

        1.1   Purpose	1
        1.2   Scope	1

CHAPTER 2     PERFORMANCE MONITORING SYSTEM DESIGN	3

        2.1   Introduction	3
        2.2   Objectives of Performance Monitoring	3
        2.3   Developing Site-Specific Monitoring Objectives and Performance
             Criteria for MNA	5
        2.4   The MNA Conceptual Site Model	 8
             2.4.1   Hydrogeology	9
             2.4.2   Contaminant Distribution, Migration, and Fate	13
             2.4.3   Geochemistry	15
             2.4.4   Receptor Locations	17
        2.5   Monitoring Network Design	17
             2.5.1   Introduction	17
             2.5.2   Monitoring Locations	19
                    2.5.2.1   Typical Target Zones	20
                    2.5.2.2   Screen Lengths	23
             2.5.3   Monitoring Parameters	23
             2.5.4   Monitoring Frequency	25
        2.6   Demonstrating MNA Effectiveness with Respect to
             Remedial Objectives	30
             2.6.1   #1 - Demonstrate that Natural Attenuation is
                    Occurring According to  Expectations	30
                    2.6.1.1   Temporal Trends in Individual Wells	31
                    2.6.1.2   Estimation of Contaminant Mass Reduction	32
                    2.6.1.3   Comparisons of Observed Contaminant Distributions
                            with Predictions and Required Milestones	34
                    2.6.1.4   Comparison of Field-Scale Attenuation Rates	35
             2.6.2   #2 - Detect Changes in Environmental Conditions that
                    May Reduce the Efficacy of Any of the Natural Attenuation
                    Processes	35

-------
                                                                             Page

                    2.6.2.1   Geochemical Parameters	36
                    2.6.2.2   Hydrogeologic Parameters	37
              2.6.3  #3 - Identity Any Potentially Toxic and/or Mobile
                    Transformation Products	37
              2.6.4  #4 - Verify that the Plume is Not Expanding Downgradient,
                    Laterally or Vertically	38
              2.6.5  #5 - Verify No Unacceptable Impacts to Downgradient
                    Receptors	39
              2.6.6  #6 - Detect New Releases of Contaminants	40
              2.6.7  #7 - Demonstrate the Efficacy of Institutional Controls	41
              2.6.8  #8 - Verify Attainment of Remediation Objectives	42
        2.7    Monitoring Plan Contents	43
              2.7.1  Introduction	43
              2.7.2  Background and Site Description	43
              2.7.3  Conceptual Site Model for Natural Attenuation	43
              2.7.4  Objectives and Decision Points	43
              2.7.5  Monitoring Network and Schedule	45
              2.7.6  Monitoring of Institutional  Controls	45
              2.7.7  Evaluations of Remedy Effectiveness	45
              2.7.8  Plan for Verifying Attainment of RAOs	45
              2.7.9  Sampling and Analysis Plan	46
              2.7.10 Quality Assurance Project Plan	46

CHAPTER 3     ANALYSIS OF PERFORMANCE MONITORING DATA	47

        3.1    Introduction	47
        3.2    The DQA Process	47
        3.3    Interpreting the Data	49
              3.3.1  Introduction	49
              3.3.2  Preliminary Presentation and Evaluation of the Data	50
              3.3.3  Data Comparisons	50
                    3.3.3.1   Comparisons  of Concentrations Within and
                             Outside the Plume	51
                    3.3.3.2   Trend Analyses	51
                    3.3.3.3   Comparisons with Existing Literature and
                             Laboratory Studies	51
                    3.3.3.4   Comparisons with Threshold Values	52
              3.3.4  Statistics	52
        3.4    Elements of a Performance Monitoring Report	52
              3.4.1  Introduction	52
              3.4.2  Summary	53
              3.4.3  Background and Site Description	53
              3.4.4  Monitoring Network and Schedule	53
              3.4.5  Evaluation of New Data	55
              3.4.6  Evaluation of Institutional Controls	55

-------
                                                                             Pac
             3.4.7  Conceptual Site Model Evaluation	55
             3.4.8  Recommendations	56

CHAPTER 4 APPLICATION OF MONITORING DATA TO
             REMEDIAL DECISIONS	57

        4.1   Introduction	57
        4.2   Decision 1 - Continue Monitoring Program Without Change 	57
        4.3   Decision 2 - Modify the Monitoring Program	57
        4.4   Decision 3 - Modify Institutional Controls	59
        4.5   Decision 4 - Implement a Contingency or Alternative Remedy	59
             4.5.1  Decision Criterion 1: Contaminant Concentrations in
                    Soil or Ground Water at Specified Locations Exhibit an
                    Increasing Trend Not Originally Predicted During Remedy
                    Selection	60
             4.5.2  Decision Criterion 2: Near-Source Wells Exhibit
                    Large Concentration Increases Indicative of a New or
                    Renewed Release	61
             4.5.3  Decision Criterion 3: Detection of a Contaminant in
                    Monitoring Wells Located Outside of the Original Plume
                    Boundary or Other Compliance Monitoring Boundaries	61
             4.5.4  Decision Criterion 4: Contaminant Concentrations Are
                    Not Decreasing at a Sufficiently Rapid Rate to Meet the
                    Remediation Objectives	62
             4.5.5  Decision Criterion 5: Changes in Land and/or Ground-
                    Water Use that Have the Potential to Reduce the
                    Protectiveness of the MNA Remedy	62
             4.5.6  Decision Criterion 6: Contaminants Are Identified in
                    Locations Posing or  Having the Potential to Pose
                    Unacceptable Risk to Receptors	62
        4.6   Decision 4 - Terminate Performance Monitoring	63

REFERENCES	65

GLOSSARY	73

APPENDIX A Variability in Measured Parameters and the Effects on
              Performance Monitoring	A-l

        A.I   Introduction	A-2
        A.2   Spatial and Temporal Variability	A-2
        A.3   Measurement Variability	A-3
        A.4   Variability in Data Interpretation	A-4
                                        ui

-------
                               LIST OF FIGURES
Figure                                                                        Page

  1.    Steps in the establishment of data quality objectives
       (modified from U.S. EPA, 2000a) 	7

  2.    Elements of a conceptual site model for monitored natural attenuation	10

  3.    Geologic block diagram and cross section depicting a stream environment
       in -which sediments have accumulated as valley fill.  In the figure, there are
       numerous coarse-grained deposits as well as finer-grained materials with
       lower hydraulic conductivity	12

  4.    Example of a network design for performance monitoring, including
       target zones for monitoring effectiveness with respect to specific
       remedial objectives	19

  5.    Cross section A to A  through monitoring network in general direction
       of ground-water flow	20

  6.    Cross section B to B' through monitoring network perpendicular to
       ground-water flow	21

  7.    Examples of possible  changes in monitoring frequency over the
       monitoring life cycle	27

  8.    Monitoring frequency effects on sampling data collection and
       interpretation	29

  9.    Potential effects of changes in ground-water flow direction on
       temporal trends in contaminant concentrations	33

 10.    Conceptual monitoring network for verifying lack of impact to
       surface water from ground-water discharge	42
                                        IV

-------
                               LIST OF TABLES
Table

  1.    Objectives for Performance Monitoring of MNA (U.S. EPA, 1999a)	4

  2.    Examples of MNA-Relevant Decisions to be Addressed Using the
       DQO Process	8

  3.    Source Characterization Information for  Conceptual Site Model
       Development	14

  4.    Elements of a Performance Monitoring Plan	44

  5.    Elements of a Performance Monitoring Report	54

-------
         LIST OF ACRONYMS AND ABBREVIATIONS
Acronym

BTEX
CERCLA

COCs
DCE
DNAPL
DQA
DQO
EPA
LNAPL
MCL
MTBE
MNA
NAPL
OSWER

PCE
PRGs
QA/QC
RAOs
RCRA
ROD
TCE
TEA
TICs
VC
             Definition

Benzene, Toluene, Ethylbenzene, Xylenes
Comprehensive Environmental Response
      Compensation and Liability Act
Contaminants of Concern
Dichloroethene
Dense Nonaqueous Phase Liquid
Data Quality Assessment
Data Quality Objectives
U.S. Environmental Protection Agency
Light Nonaqueous Phase Liquid
Maximum Contaminant Level
Methyl-t-Butyl Ether
Monitored Natural Attenuation
Nonaqueous Phase Liquid
Office of Solid Waste and
      Emergency Response
Perchloroethene (tetrachloroethene)
Preliminary Remediation Goals
Quality Assurance/Quality Control
Remedial Action Objectives
Resource Conservation and Recovery Act
Record of Decision
Trichloroethene
Terminal Electron Acceptor
Tentatively Identified Compounds
Vinyl Chloride
                               VI

-------
                                  FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with protecting the
Nation's land, air, and water resources. Under a mandate of national environmental laws,
the Agency strives to formulate and implement actions leading to a compatible balance
between human activities and the ability of natural systems to support and nurture life.
To meet this mandate, EPA's research program is providing data and technical support for
solving environmental problems today and building a science knowledge base necessary
to manage our ecological resources wisely, understand how pollutants affect our health,
and prevent or reduce environmental risks in the future.

The National Risk Management Research Laboratory (NRMRL) is the Agency's center
for  investigation of technological and management approaches for preventing and
reducing risks from pollution that threatens human health and the environment. The
focus of the Laboratory's research program is on methods and their cost-effectiveness for
prevention and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems; remediation of contaminated sites,
sediments and ground water; prevention and control of indoor air pollution; and
restoration of ecosystems. NRMRL collaborates with both public and private sector
partners to foster technologies that reduce the cost of compliance and to anticipate
emerging problems. NRMRL's research provides solutions to environmental problems by:
developing and promoting technologies that protect and improve the environment;
advancing scientific and engineering information to support regulatory and policy
decisions; and providing the technical support and information transfer to ensure
implementation of environmental regulations and strategies at the national, state, and
community levels.

Effective performance monitoring for remedies that rely on the natural attenuation of
contaminants is a crucial  element of remedial design and implementation.  Effective
monitoring system designs are formulated from an enhanced understanding of the
migration and ultimate fate of the contaminants in the site-specific environment.  This
document provides technical guidance regarding the types of monitoring parameters and
analyses useful for evaluating the effectiveness of the natural attenuation component of
ground-water remedial actions. The information will be  helpful during the design  of the
performance monitoring plan as well as during its implementation.
                                  Stephen G. Schmelling, Direij
                                  Ground Water and Ecosystems'Restoration Division
                                  National Risk Management Research Laboratory
                                        Vll

-------
                                    NOTICE
The U.S. Environmental Protection Agency through its Office of Research and
Development funded and managed the research described here under EPA Contract Nos.
68-C-99-256 and 68-C-02-092 to Dynamac Corporation, Ada, Oklahoma. It has been
subjected to the Agency's peer and administrative review and has been approved for
publication as an EPA document.  Mention of trade names or commercial products does
not constitute endorsement or recommendation for use.

All research projects making conclusions or recommendations based on environmental
data and funded by the U.S. Environmental Protection Agency are required to participate
in the Agency Quality Assurance Program. This project did not involve the collection or
use of environmental data and, as such, did not require a Quality Assurance Plan.
                                 DISCLAIMER
This document provides guidance to EPA Regions concerning how the Agency intends to
exercise its discretion in implementing one aspect of EPA cleanup programs, including
CERCLA remedial actions and RCRA corrective measures. The guidance is designed to
implement national policy on these issues.

This document cannot impose legally-binding requirements on  EPA, States, or the
regulated community, and may not apply to a particular situation based upon the
circumstances. Any decisions regarding a particular remedial action or corrective measure
-will be made based on the statute and regulations, and EPA decision makers retain the
discretion to adopt approaches on a case-by-case basis that differ from this guidance
where appropriate.

Interested parties are free to raise  questions and objections about the substance of this
guidance and the appropriateness  of the application of this guidance to a particular
situation, and the Agency welcomes public input on this document at any time. EPA may
change this guidance in the future.
                                       vui

-------
                           ACKNOWLEDGMENTS
This document was a joint effort of several offices and divisions -within the U.S. EPA.
The principal authors were:

                                 Daniel F. Pope 1
                                Steven D. Acree 2
                                Herbert Levine 3
                               Stephen Mangion 4
                                 Jeffrey van Ee 5
                                  Kelly Hurt 1
                                'Barbara Wilson 1

The authors express their appreciation for helpful comments received from numerous
organizations and individuals including the U.S. EPA Ground Water Forum, the U.S.
EPA Federal Facilities Forum, Ken Lovelace, Guy Tomassoni, Hal White, Terry
Evanson, Dr. Aristeo M. Pelayo, Patricia Ellis, Todd Wiedemeier, Dr. Frank Chapelle,
Dr. Ryan Dupont, Dr. John Wilson, Dr. Jim Weaver, and Dr. Robert Ford. The authors
would also like to acknowledge Carol House for her support in creating the graphics
and layout for this report.
                   The document was prepared under contract to:
                              Dynamac Corporation
                EPA Contract Numbers: 68-C-99-256 and 68-C-02-092
                                 Project Officer
                                David S. Burden 2
       1 Dynamac Corporation, Ada, OK
       2 U.S. EPA, Office of Research and Development, National Risk
        Management Research Laboratory, Ground Water and Ecosystems
        Restoration Division, Ada, OK
       3 U.S. EPA, Region 9, Superfund Division, San Francisco, CA
       4 U.S. EPA, Office of Research and Development, New England
        Region, Boston, MA
       5 U.S. EPA, Office of Research and Development, National Exposure
        Research Laboratory, Environmental Sciences Division, Las Vegas, NV

-------
                                  ABSTRACT
Environmental monitoring is the major component of any remedy that relies on natural
attenuation processes. The objective of this document is to identify data needs and
evaluation methods useful for designing monitoring networks and determining remedy
effectiveness. Effective monitoring of natural attenuation processes involves a three-
dimensional approach to network design and clearly defined performance criteria based
on site-specific remedial action objectives.  Objectives for the monitoring program will be
met through routine evaluations of institutional controls and measurements of
contaminant, geochemical, and hydrologic parameters. These data are used to evaluate
changes in three-dimensional plume boundaries, contaminant mass and concentration,
and hydrological and geochemical changes that may indicate changes in remedy
performance.

Data interpretation focuses on detection of spatial and temporal changes, and assessment
of their impacts on the achievement of site-specific goals. Particular changes of interest
include:

         Progress toward contaminant removal objectives and indications of additional
          contaminant releases,

         Contaminant detections at the horizontal and vertical plume boundaries that
          may indicate plume expansion,

         Geochemical changes {e.g., oxidation-reduction (redox) conditions) indicative
          of possible changes in contaminant transformation rates,

         Changes in ground-water flow rates or directions such that contaminants may
          move into previously unimpacted areas, and

         Changes in land and resource uses that threaten the effectiveness of
          institutional controls.

Decisions regarding remedy effectiveness and the adequacy of the monitoring program
will generally result in either continuation of the program, program modification, or
implementation of a contingency or alternative remedy. Such decisions are appropriately
based on specific, quantifiable performance criteria defined in the monitoring plan.
Continuation of the program -without modification would be supported by contaminant
concentrations behaving according to remedial expectations -while ground-water flow and
geochemical parameters remain within acceptable ranges. Modification of the program,
including increases or decreases in monitoring parameters, frequency, or locations, may
be -warranted to reflect changing conditions or improved understanding of natural

-------
attenuation processes at the site.  Situations that may trigger implementation of a
contingency or alternative remedy include:

         Increasing contaminant concentrations or trends not predicted during remedy
          selection or indicative of new releases,

         Contaminant migration beyond established plume or compliance boundaries,

         Contaminants not decreasing at a rate sufficient to meet remediation
          objectives,

         Changes in land or ground-water use that have the potential to reduce the
          protectiveness of the remedy, and

         Contaminants observed at locations posing or having the potential to pose
          unacceptable risks to receptors.
                                        xt

-------

-------
                                      Chapter 1
                                   INTRODUCTION
1.1    Purpose

The term "monitored natural attenuation," as used in this document and in the Office of Solid
Waste and Emergency Response (OSWER) Directive 9200.4-17P (U.S. EPA, 1999a), refers to "the
reliance on natural attenuation processes (within the context of a carefully controlled and
monitored site cleanup approach) to achieve site-specific remediation objectives within a time
frame that is reasonable compared to that offered by other more active methods." Performance
monitoring for monitored natural attenuation (MNA) remedies in ground water will present
unique, remedy-specific, economic and technical challenges to be evaluated -when comparing
MNA to other remedies. To reduce uncertainty in assessing remedy performance, more intensive
and unique monitoring may be needed for remedies that rely on natural attenuation processes
than for remedies that rely on engineered systems. This document is designed to be used during
preparation and review of long-term monitoring plans for sites where MNA has been selected as
part of the site remedy.  Nothing in this document changes Agency policy regarding remedial
selection criteria, remedial expectations, or the selection and implementation of MNA. This
document does not supercede any previous guidance and is intended for use in conjunction -with
several other documents, including OSWER Directive 9200.4-17P, "Use of Monitored Natural
Attenuation at Superfund, RCRA Corrective Action, and Underground Storage Tank Sites" (U.S.
EPA, 1999a),  and "Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents
in Ground Water" (U.S. EPA,  1998a).

As stated in the Directive (U.S. EPA, 1999a), "EPA expects that MNA will be an appropriate
remediation method only -where its use will be protective of human health and the environment
and it will be  capable of achieving site-specific cleanup remediation objectives within a time frame
that is reasonable compared to other alternatives." Environmental monitoring -will be an
essential component of the remedy to ensure site-specific objectives are achieved. This document
was designed  to provide a technical framework for monitoring  program development.
Discussions include details of technical issues concerning monitoring parameters, locations (i.e.,
three-dimensional monitoring locations relative to the plume),  and monitoring frequencies. This
document does not provide details of particular methodologies  for sampling,  analysis, modeling,
or other characterization tools.

1.2    Scope

This document focuses on chlorinated solvent compounds and  common fuel-related aromatic
compounds (i.e., benzene, toluene, ethylbenzene, and xylenes (BTEX)) dissolved in ground water
within porous media. These compounds comprise a significant portion of current ground-water
pollution problems, and there is a considerable body of information available concerning their
behavior in the subsurface. It is necessary to have a detailed understanding of the behavior of a
contaminant in the subsurface in order to properly evaluate natural attenuation processes and
develop an adequate monitoring plan. The limited data that may be available on subsurface
behavior of other contaminants increase the difficulty of proper evaluation of MNA.

-------
Contaminants such as inorganic compounds, radionuclides, fuel oxygenates (e.g., methyl-t-butyl
ether (MTBE)), explosives, and wood treating chemicals are not specifically addressed in this
guidance.  Many of these contaminants behave differently from those on -which this document is
focused. Such compounds may not be attenuated at rates sufficient to be protective of human
health and the environment -while other compounds associated with a given release may be
rapidly attenuated. For example, MTBE is a common fuel component that is described in detail in
U.S. EPA (2003a). At many sites, the potential for significant migration of MTBE maybe greater
than that of the BTEX contaminants.  The determination of conditions under which MTBE may
be readily attenuated in the subsurface is an area of ongoing research. Although the details of an
appropriate performance monitoring program are specific to the contaminant under
consideration, the type of monitoring and evaluation methods that are discussed -will often be
applicable to contaminants such as MTBE that are not explicitly considered in this document.
Soils and/or sediment-only remedies are discussed only in reference to contaminant sources for
ground water.

This document -will focus on contaminants in the aqueous-phase plume. As stated in U.S. EPA
(1999a), it is expected that source control will be a fundamental component of any MNA remedy.
The degree to -which contaminant sources are removed or controlled directly affects the
effectiveness and remedial time frame for MNA remedies. Even small amounts of source
materials remaining at a site (e.g., residual nonaqueous phase liquids  (NAPL)) may greatly extend
the time frame for achieving remedial goals. At such sites, the extensive source characterization
and monitoring typically needed to accurately estimate source lifetime and evaluate source effect
on plume behavior may make MNA an inappropriate remedial choice. In general, the discussions
provided in this document will be most applicable  to sites with controlled sources or no materials
that provide continuing sources for ground-water contamination (e.g., NAPL). Accordingly, the
contaminant release scenario used in figures throughout the document is one in -which only a
small volume  of NAPL -was released to the subsurface and effective source removal actions -were
subsequently  implemented.

This document is limited to evaluations performed in porous-media settings. Although similar
concepts are sometimes applicable to other geologic settings such as karst and fractured rock
formations, ground -water and contaminants move preferentially through discrete path-ways (e.g.,
solution channels, fractures, and joints) in these settings. Existing techniques are often incapable
of fully delineating the path-ways along -which contaminated ground -water migrates. This greatly
increases the uncertainty and costs of assessments of contaminant migration and fate and is
another area of continuing research.  Detailed guidance on performance monitoring system design
in fractured rock, karst, and other such highly heterogeneous settings is beyond the scope of this
document. As noted in OSWER Directive 9200.4-17P (U.S. EPA, 1999a), "MNA will not
generally be appropriate -where site complexities preclude adequate monitoring."  The directive
provides additional discussion regarding the types  of sites -where the use of MNA may be
appropriate.

This document focuses on monitoring the saturated zone, but site characterization and monitoring
for MNA or any other remedy typically -would include monitoring of all significant path-ways by
-which contaminants may move from source areas and ground-water plumes to impact receptors
(e.g., surface water and indoor air).

-------
                                       Chapter 2
                 PERFORMANCE MONITORING SYSTEM DESIGN
2.1    Introduction

Designing a monitoring system to assess natural attenuation processes and the effectiveness of
those processes with respect to achieving remedial objectives involves making site-specific
decisions regarding:

       Monitoring parameters,

       Number and location of monitoring points,

       Monitoring frequency, and

       Methods to analyze and interpret the data obtained from monitoring points.

Sections 2.2 and 2.3 of this chapter discuss objectives of performance monitoring and ways to
develop site-specific objectives and performance criteria such that an MNA remedy can be
evaluated based on measurable criteria.

Sections 2.4 and 2.5 discuss general principles of development of the MNA conceptual site model
and the monitoring network design, based on data collected during the site characterization
efforts. These general principles provide a framework for understanding the controlling factors
that guide the monitoring decisions listed above.  The data that generally should be available from
the site characterization effort are discussed in detail.

Section 2.6 of this chapter addresses the specific application of these general principles to
demonstrating effectiveness of MNA for attaining the performance monitoring objectives in
Table 1.

The final section of this chapter (Section 2.7) provides suggested content and format for
performance monitoring plans.

2.2    Objectives of Performance Monitoring

The OSWER Directive 9200.4-17P (U.S.  EPA, 1999a) provides eight specific objectives to be met
by the performance monitoring program of an MNA remedy (Table 1). This document will
discuss the technical aspects of monitoring systems typically necessary to meet these and similar
objectives.  The objectives usually will be met by implementing a performance monitoring
program that routinely evaluates the effectiveness of institutional controls and measures
contaminant concentrations, geochemical parameters (e.g., oxidation-reduction (redox)
parameters, dissolved organic carbon,  pH), and hydrologic parameters. These data will be used to
evaluate the dynamic behavior of the plume over time, including:

-------
Table 1.        Objectives for Performance Monitoring of MNA (U.S. EPA, 1999a)


    1)  Demonstrate that natural attenuation is occurring according to expectations,

    2)  Detect changes in environmental conditions (e.g., hydrogeologic, geochemical, microbiological, or other
       changes) that may reduce the efficacy of any of the natural attenuation processes,

    3)  Identify any potentially toxic and/or mobile transformation products,

    4)  Verify that the plume(s) is not expanding downgradient, laterally or vertically,

    5)  Verify no unacceptable impact to downgradient receptors,

    6)  Detect new releases of contaminants to the environment that could impact the effectiveness of the
       natural attenuation remedy,

    7)  Demonstrate the efficacy of institutional controls that were put in place to protect potential receptors,
       and

    8)  Verify attainment of remediation objectives.
       Changes in three-dimensional plume boundaries,

       Changes in the geochemical setting (i.e., as indicated by the geochemical parameters,
       especially the redox parameters such as redox potential, dissolved oxygen, nitrate/nitrite,
       manganese (II), iron (II), sulfate, and methane) that may be indicative of changes in biotic
       or abiotic processes affecting the rate and extent of natural attenuation, and

       Contaminant mass and/or concentration reductions indicative of progress toward
       contaminant reduction objectives.

Plume behavior can then be evaluated to judge the effectiveness of the MNA remedy, the
adequacy of the monitoring program, and the adequacy of the conceptual site model for MNA.
On the basis of these judgements, decisions may be made for subsequent phases of site operations,
including:

       Continue the performance monitoring program without change,

       Modify the performance monitoring program,

       Modify the institutional controls,

       Implement a contingency or alternative remedy, or

       Terminate performance monitoring.

As is the case for other remedies, performance monitoring for MNA remedies continues until all
remedial action objectives have been met  (e.g., contaminant concentrations throughout the site
meet remedial requirements).  The last phase of performance monitoring (verification
monitoring) may involve changes to the performance monitoring program as appropriate to verify
remedial goals have been met (e.g., a period of more frequent monitoring and/or more dense

-------
monitoring, especially if monitoring locations or frequency had been reduced over an extended
performance monitoring period).

2.3    Developing Site-Specific Monitoring Objectives and Performance Criteria for MNA

Site-specific performance monitoring objectives for MNA are derived from site-specific remedial
action objectives (RAOs) and preliminary remediation goals (PRGs). RAOs provide a general
description of what the cleanup will accomplish (e.g., restoration of ground water). PRGs are the
more specific statements of the desired endpoint concentrations or risk levels, for each exposure
route, that are believed to provide  adequate protection of human health and the environment
based on preliminary site information. For additional guidance concerning remedial objectives
refer to current, program-specific guidance (e.g., U.S. EPA, 1997a; U.S. EPA, 2002c).

The site-specific performance monitoring objectives are general statements of what is to be
required  of the monitoring network (e.g., detect plume expansion). Performance criteria are
detailed statements that set forth standards or requirements based on specific measurements.  For
instance,  it might be specified that the monitoring system be able to detect a contaminant
concentration of 1 |ig/L in ground water in the area between the known contaminated aquifer
and a receptor.  Clearly stated performance monitoring objectives, accompanied by specific,
quantifiable performance criteria,  are useful for designing and evaluating the performance
monitoring system, and assessing  MNA remedy effectiveness. RAOs, PRGs, performance
monitoring objectives, and performance criteria may be specified in remedy decision documents
(e.g., Record of Decision, Corrective Action Plan)  to provide the basis for development of the
performance monitoring plan.

Common remedial objectives include lack of plume expansion and reduction of contaminant
concentrations to established limits. Examples of performance criteria for monitoring such
objectives include:

      The ability of the monitoring network to detect a specified contaminant at a specified
       concentration at specified sampling locations (e.g., detect the occurrence of vinyl chloride
       at a transect of wells located between the plume boundaries and potential receptors at  a
       concentration of 2 |j,g/L), and

      The ability of the monitoring network to detect a specified decrease in contaminant
       concentrations throughout the site within a specified time frame (e.g., detect a 50%
       decrease in concentrations throughout the plume by the end of ten years).

In order to develop such performance  criteria and realistic means of evaluating performance with
respect to these criteria,  a systematic process should be followed.  This process should take into
account the data needs and the methods available to obtain the data.  For example, if statistical
tests are to be used to assess MNA performance, a systematic process can be used to select  the
appropriate tests,  and to choose the type of data and data  collection techniques necessary to
obtain data required for the test. One such systematic development process that is highly effective
and is endorsed in current guidance (U.S. EPA, 2000a) is  the Data Quality Objectives (DQO)
Process.

The DQO process is a systematic planning approach for data collection that is based on the
scientific method.  The DQO process identifies goals of the data collection and decision-making
process, and assesses consequences of incorrect decisions. Although the process is typically

-------
described in linear terms (Figure 1), it is really a flexible process that relies on iteration and
modification as the planning team works through each step, allowing earlier steps to be revised
considering new information.  The basic steps in the process begin with identification of the
problem to be solved and the resources available to support the solution.  Based on a statement of
the problem, a specific decision that requires the acquisition of new data is formulated. Once the
specific question to be answered has been stated, the data needed to make the decision can then
be specified. Decisions regarding the boundaries of the study required to answer the question of
concern are then made.  Such boundaries may include spatial boundaries (e.g., the boundaries of
the contaminated aquifer) and time boundaries (e.g., the time frame for data collection).  The
decision to be made is then simplified through  development of a logical if-then statement
describing the conditions under which different alternatives would be chosen. At this point,
acceptable limits on errors in the decision and methods for limiting uncertainty in the data are
established. Finally,  cost-effective sampling designs to provide the needed data are produced.
Although these steps are largely intuitive, steps may be overlooked if a specified framework is not
used.

The DQO process provides a framework for addressing  the issues of subsurface heterogeneity and
data variability (Appendix A) that  cause uncertainty about site characteristics, and often present
obstacles to development of monitoring plans and data interpretation. Subsurface geology,
hydrology, geochemistry, biology, and contaminant distribution may be highly variable and
interact in complex ways.  Therefore, it is important to have a defined process for dealing with the
variability in order to design an appropriate monitoring system, and interpret the data derived
from the monitoring system within an acceptable range of uncertainty.  Uncertainty may be
expressed mathematically using statistical techniques, if feasible, and it may be expressed
qualitatively as "professional judgement" concerning the reliability of a  certain interpretation of
the data. The DQO process involves identification of data gaps that may cause an erroneous
decision to be made, and assessment of the cost-benefit ratio of filling those gaps to reduce
uncertainty.

When the DQO process identifies additional data needed to reduce uncertainty and facilitate
development of the performance monitoring plan, further investigation of a site is warranted.
Use of the Triad approach (U.S. EPA, 2001a) to planning  and conducting the investigation can
provide the information rapidly and cost-effectively, allowing the performance monitoring plan to
be developed and deployed expeditiously. The  Triad approach combines systematic planning to
ensure that the characterization goals are clearly defined with dynamic work plans (U.S. EPA,
2003b)  and quick-turnaround analytical techniques, including field analytical techniques (U.S.
EPA, 2003c), to provide data meeting site-specific requirements in a condensed time frame.  The
approach promotes the use  of new  science and  technology tools to identify and manage
information gaps (i.e., uncertainties) that could lead to unacceptable decision errors. The Triad
approach is particularly effective where site heterogeneity increases uncertainty about the
representativeness of isolated data points. The Triad approach is intended to produce an accurate
conceptual site  model.  Using the dynamic work plan element of the Triad allows efficient
refinement of the model in  real time by specifying sampling locations and analyses based on data
from previous samples.

Examples of typical problems and concerns that may be addressed using the DQO process during
development of a performance monitoring plan are  listed in Table 2. Important outcomes from
the DQO process include the spatial and temporal scales for the collection of data, sample
collection methods, acceptable decision error rates, and number of samples needed to support
decision-making.

-------
                           1. STATE THE PROBLEM
             Summarize the problem that will require newdata,and
               identify the resources available; develop site-specific
              conceptual model for monitored natural attenuation.
                        2. IDENTIFYTHE DECISION
                  Identify the decision that requires new data
                        (e.g., Is the plume expanding?).
                   3. IDENTIFY INPUTS TO THE DECISION
             Identify the information needed to support the decision
              and specify which inputs require new measurements.
                    4. DEFINE THE STUDY BOUNDARIES
            Specify the spatial and temporal aspects of the media that
                the data must represent to support the decision.
                       5. DEVELOP A DECISION RULE
             Develop a logical "if...then..."statement that defines the
               conditions that would cause the decision maker to
                      choose among alternative actions.
                  6. SPECIFY LIMITS ON DECISION ERRORS
            Specify the decision maker's acceptable limits on decision
              errors, which are used to establish performance goals
                      for limiting uncertainty in the data.
               7. OPTIMIZE THE DESIGN FOR OBTAINING DATA
         Identify the most resource-effective sampling and analysis design
            for generating data that are expected to satisfy the DQOs.
Figure 1. Steps in the establishment of data quality objectives (modified from U.S.EPA,2000a).

-------
Table 2.   Examples of MNA-Relevant Decisions to be Addressed Using the DQO Process


    Is natural attenuation occurring according to expectations?

    Q   Determine that plume is behaving as specified in decision documents.
    Q   Determine that contaminant mass reduction is proceeding as specified in decision documents.
    Q   Determine that contaminant concentrations in source areas and the downgradient plume are declining at
        sufficient rates to meet RAO's.

    Are there any changes in conditions that may affect the efficacy of natural attenuations?

    Q   Determine that there are no changes in ground-water flow rates and directions that would impact plume stability.
    Q   Determine that there is no change in the geochemical environment (e.g., redox conditions) that would impact
        plume stability or contaminant reduction.

    Are there potentially toxic and/or mobile transformation products?

    Q   Determine if naturally-occurring arsenic, manganese or other potentially problematic species are being
        transformed and mobilized due to changes in the geochemical environment.
    Q   Determine if previously unrecognized toxic and/or mobile transformation products are present.

    Is the plume expanding?

    Q   Determine that the plume is not migrating beyond current horizontal or vertical boundaries.
    Q   Determine that no contaminants are found  in the wells placed between the downgradient edge of the plume and
        receptors at concentrations above a specified limit.

    Are there any unacceptable impacts to receptors?

    Q   Determine that there is no unacceptable impact to surface-water bodies, wetlands, or other ecological receptors.
    Q   Determine that there is no impact to indoor air in adjacent buildings.
    Q   Determine that there is no impact to water-supply wells.

    Are there any new contaminant releases that may impact remedy effectiveness?

    Q   Determine that there are no new releases of contaminants from the source area.

    Are the specified institutional controls effective?

    Q   Determine that institutional controls are currently effective in eliminating exposure to contaminants.

    Have remedial objectives been met?

    Q   Verify that contaminant concentrations are  below required levels.
2.4     The MNA Conceptual Site Model

The monitoring network design is based on the data derived from site characterization and any
other site studies. These data are used to develop a conceptual model of the site before remedy
selection; the conceptual model can then be used to facilitate the performance monitoring design
process. The conceptual site  model for natural attenuation is the site-specific qualitative and
quantitative description of the migration and fate of contaminants with respect to possible
receptors and the geologic, hydrologic, biologic, geochemical, and anthropogenic factors that
control contaminant distribution. Essentially, the conceptual site model expresses an
understanding  of the site structure, processes, and factors that affect plume development and
behavior.  It is  built on assumptions and hypotheses that have been evaluated using site-specific
data, and is continually reevaluated as new data are developed throughout the site lifetime.  The
conceptual model typically should be developed and evaluated using a team approach (i.e., using a

-------
team of subject-matter experts including hydrogeologists, microbiologists, statisticians/modelers,
chemists and other experts as appropriate for the specific site).

A three-dimensional conceptual site model that incorporates temporal changes is often needed to
provide a framework for interpreting the site data, judging the significance of changes in site
conditions, and predicting future behavior of the source and plume.  Understanding plume
formation and behavior is the basis for predicting future plume behavior, and therefore, predicting
whether the MNA remedy will be able to achieve site remedial goals within specified time frames.
Conceptual models are expressed tangibly in text, site maps (e.g., contaminant isoconcentration
maps and potentiometric surface maps), cross sections (e.g., hydrogeologic and chemical
distributions), and other graphical presentations, and in terms of mathematical calculations
describing the plume and site.

The development of quantitative models (i.e., mathematical representations of site conditions)
based on the conceptual site model generally is an essential part of site characterization, remedy
selection, and performance monitoring for MNA. Quantitative models may be as simple as linear
regressions for estimation of contaminant attenuation rates, or as complex as numerical models of
ground-water flow and contaminant fate and migration.  These mathematical representations are
used to help understand site processes, locate monitoring points, estimate attenuation rates, and
evaluate possible effects of different conditions on plume behavior.  Quantitative assessments
require particular types of data. The data collection effort should be designed with the chosen
evaluation methods, calculations, and model (s) in mind.

The data and analyses necessary for formulation of an adequate three-dimensional conceptual
site model for MNA depend on site-specific conditions.  However, the types of information and
analyses that are generally needed for model development are illustrated in Figure 2 and
described below.

2.4.1  Hydrogeology

Hydrogeology is the foundation of the conceptual site model for natural attenuation (National
Research Council, 2000).  Detailed knowledge of site hydrogeology is crucial to understanding
how ground water flows and contaminants may be transported in the subsurface. A general
discussion of the effects of geology and hydrology on ground-water flow follows.

Much of the spatial variability in observed contaminant concentrations is the result of geologic
heterogeneity. In a sedimentary geologic setting, spatial changes in geology are present at scales
that can vary from fractions of an inch to miles.  Sedimentary fades  (i.e., sedimentary bodies that
are internally similar in characteristics) determine  the three-dimensional geometries,
connectivity, and heterogeneity (i.e., variability) of transmissive zones and barriers to flow
(Galloway and Sharp, 1998). These characteristics and the resulting variability in hydraulic
properties (e.g., hydraulic conductivity and porosity) are generally the result of the original
geologic depositional processes. In similar fashion, anthropogenic features such as buried  utility
corridors and heterogeneous fill materials may also result in the formation of transmissive zones
and flow barriers.  Transmissive zones are subsurface units where ground water flows in paths
constrained or bounded by lower hydraulic conductivity materials (i.e., geologic impediments to
flow) or hydrologic  barriers  (e.g., hydraulic head boundaries). An example of a transmissive zone
is a deposit of discontinuous, interbedded sands and silt bounded above by the water table and
below by a thick, locally continuous clay of low hydraulic conductivity.  These bounds act  to keep
ground water flowing within the transmissive zone. However, even within a given transmissive


                                              9

-------
            Contaminant Source and Source
                 Control Information

Q Location, nature, and history of contaminant releases
   or sources
Q Locations and characterizations of sources for ground-
   water contamination [e.g., nonaqueous phase liquid
   (NAPL)]
Q Locations and descriptions of source control and other
   ongoing and proposed remedial actions
         Geologic and Hvdroloaic Information
Q Regional and site geologic and hydrologic settings,
   including controls on ground-water flow
Q Analyses of depositional environments and geologic
   features that may serve as zones of preferential flow or
   barriers to flow, including geometry and physical
   properties of geologic facies (e.g., texture, porosity, bulk
   density) and their variability
Q Stratigraphy, including thickness and lateral continuity
   of geologic units,and bedding features
Q Anthropogenic features (e.g., buried corridors and
   heterogeneous fill materials) that control ground-water
   flow, and may serve as migration pathways or barriers
Q Depth to ground water and temporal variation
Q Characteristics of surface water bodies (e.g., locations,
   depths, and flow rates), their interactions with ground
   water,and temporal variations
Q Ground-water recharge and discharge locations, rates
   and temporal variability
Q Hydraulic gradients, including horizontal and vertical
   components,and their variations in response to
   fluctuations in site hydrology (e.g., seasonal or longer
   term precipitation patterns and changes in patterns of
   ground-water withdrawal or irrigation)
   Hydraulic properties (e.g., hydraulic conductivities,
   storage properties, and effective porosities) and their
   variability and anisotropy within geologic units
   Quantitative description of the ground-water flow field
Q Chemical properties of the subsurface matrix including
   mineralogy and organic matter
Q
Q
                                                                             Receptor Information

                                                              Q Aquifer classification, current usage information, and
                                                                reasonably anticipated future usage
                                                              Q Locations and production data for water-supply wells
                                                              Q Locations and information on human and ecological
                                                                receptors under current and reasonably anticipated
                                                                future conditions
                                                              Q Areas susceptible to impact by vapor-phase
                                                                contaminants (e.g., indoor air)
                                                              Q Information on local historical and cultural uses of land,
                                                                water,and other resources used to identify receptor
                                                                populations
                                                              Q Descriptions of institutional controls currently in place
                                                                   Contaminant Distribution, Transport ana Fate
Q Distribution of each contaminant phase (/.e.,gaseous,
   aqueous, sorbed, NAPL) and estimates of mass
Q Mobility of contaminants in each phase
Q Temporal trends in contaminant mass and
   concentrations
Q Sorption information,including retardation factors,
   sorption mechanisms,and controls
Q Contaminant attenuation processes and rate estimates
Q Assessment of facilitated transport mechanisms (e.g.,
   complexation or colloidal transport)
Q Geochemical characteristics that affect or are indicative
   of contaminant transport and fate, and mineralogy, if
   needed
Q Potential for mobilization of secondary contaminants
   (e.g., arsenic)
Q Effects of other proposed or ongoing remedial
   activities on contaminant transport, fate and  natural
   attenuation processes
     Figure 2.  Elements of a conceptual site model for monitored natural attenuation.
                                                       10

-------
zone, ground water may move   in sinuous paths due to small-scale diffences in hydraulic
conductivity resulting from heterogeneous geologic materials or to temporally variable, three-
dimensional hydraulic gradients.

Transmissive zones may be separate and distinct pathways for contaminant movement. For
example, the degree of hydrologic connection between different sedimentary fades depicted in
Figure 3 is small.  Monitoring points in different sedimentary fades may appear to be similar and
may be hydraulically downgradient of one another without significant ground-water flow and
contaminant migration from one unit to the next.  A reduction in contaminant concentrations
between monitoring points may be an artifact of sample location and not representative of actual
conditions. Inferences about natural attenuation based on apparent decreases in contaminant
concentration in the downgradient direction are likely to be  incorrect in these situations unless
ground-water flow paths are determined and monitored.  Determining flow paths is often difficult
to accomplish, especially when using small numbers of monitoring points.

Differentiation of the ground-water flow field by  means of a detailed chacterization of site
geology is crucial for effective performance monitoring of a natural attenuation remedy. In this
respect, evaluation of sedimentary depositional environments is an especially useful framework
for the understanding of site stratigraphy  and  the distribution of lithologic controls on ground-
water flow.   TlaE-dimensional characterization is used to evaluate and predict the effects of
natural attenuation processes (e.g., advection, dispersion, sorption, and transformation processes)
on contaminants. Requirements for monitoring transmissive zones of interest typically should be
considered in the development of the site-specific DQOs for  a performance monitoring plan.

Temporal variations in the ground-water flow field due to natural or anthropogenic changes in
ground-water recharge or withdrawal may also be important influences on contaminant
migration. These fluctuations may lead to changes in the geometry of the plume that will affect
monitoring system design and operation.  For  example, there can be seasonal changes in water
elevations that depend upon the temporal patterns of rainfall, and, in the north, snowfall and
snowmelt. Because there can be fixed controls on water levels (i.e., the fixed elevation of the
oceans or large water bodies), hydraulic gradients can also change seasonally.  In addition,  there
can be longer term fluctuations in water elevations associated with sequences of unusually wet or
dry years. With increasing development,  land use  changes can alter patterns of recharge,
discharge, or withdrawal that in turn may impact contaminant migration.

Available ground-water elevation data are used to determine, if possible, the expected range of
variation of velocities over the life of the plume that is being treated by  MNAFor example,
seasonal variations in precipitation may change hydraulic gradients and the depth to ground
water resulting in fluctuations in horizontal and vertical plume boundaries, changes in
contaminant concentrations in individual wells, changes in direction of plume migration, and
plume discharge to intermittent streams and wetlands.  Anthropogenic influences on site
hydrology such as changes in ground-water withdrawal or irrigation rates and patterns may have
similar effects on plume behavior but may occur on frequencies and timing other than those
corresponding to precipitation patterns. Temporal variations in plume behavior affect the choice
of performance monitoring well locations and the analyses of data obtained from the well
network.

Geologic and hydrologic data and interpretations that are used in the development of the
performance monitoring plan are shown on Figure 2. The level of detail needed will be site
specific.   Much of the geologic information is obtained fim  geologic cores and supplemented with

-------
                     Area of Former Solvent Tank
                     and Ongoing Source Controls
                                 A1
               Target Monitoring Zones

             1. Source area I      _|
             2. Contaminated zones of highest
               concentrations and mobility
             3. Plume fringes exhibiting low
               contaminant concentrations
             4. Plume boundaries
             5. Recalcitrant zone determined
               from historical trends
             6. Upgradient locations
  A
             Legend

      Gravel,gravel-sand mixtures

      Medium to coarse-grained sand

      Fine-grained silty sand

      Bedrock
!  Dissolved Plume
Figure 3. Geologic block diagram and cross section depicting a stream environment in which sediments have
         accumulated as valley fill. In the figure, there are numerous coarse-grained deposits as well as finer-
         graiimed materials with lower hydraulic conductivity.
                                                 12

-------
information from surface and, particularly, borehole geophysical methods.  Innovative
characterization technologies, such as the cone penetrometer and geologic sampling using direct-
push methods, offer the potential for cost-effectively evaluating the geologic controls on ground-
water flow and their variability in greater detail than previously possible.  Hydrologic information
will generally be obtained through hydraulic testing in the field (e.g., pumping tests, slug tests, and
tracer tests) and through the measurement of hydraulic heads in wells and piezometers that are
appropriately screened in individual hydrostratigraphic units. Additional information regarding
geologic and hydrologic site characterization concepts and techniques may be obtained from a
variety of sources (e.g., Butler,  1998; Kruseman and de Ridder, 1989; U.S. EPA, 1991a, 1993a,
1993b, 1997b, 2002a).

At a minimum, the hydrogeologic data base generally should be sufficient to:

       Define geologic and hydrologic controls on the ground-water flow field (e.g., transmissive
       units, barriers to flow, and the horizontal and vertical components of hydraulic gradients),

       Quantify ground-water flow rates and directions and their spatial and temporal variations
       within transmissive units, and

       Support identification of possible receptors  and characterization of natural attenuation
       rates and the relative effects of dominant natural attenuation processes.

It is desirable that the hydrogeologic data base be developed and kept in electronic form, for ease
of adding, sorting, analyzing and transferring data,  developing and publishing reports, and
interfacing with geographic information systems. There are no widely-recognized standard
formats for such data bases, but the interested parties may agree to adhere to a particular format
for a given site.

The scale and intensity of the characterization are determined by the variability in site geology
and hydrology and the acceptable level of uncertainty in the outcome of the evaluations.  Spatial
and temporal variability in these parameters and their effects on the performance monitoring
network and sampling frequencies typically should be explicitly considered. This process
generally should consider the observed variability, the sources of that variability, and the degree to
which the variability affects decisions regarding the monitoring network design and monitoring
strategies.

2.4.2    Contaminant Distribution, Migration, and Fate

Contaminant distribution and behavior should be well characterized within both former source
areas and the downgradient plume  (i.e., the plume downgradient of all known source materials).
Source characterization data (Table 3) are  used to help identify appropriate performance
monitoring constituents and monitoring locations and depths as well as interpret historical data
and predict future behavior of the dissolved plume.  Some  of the data, such as the history of source
release, may be unavailable.

Some of the more common sources for continued ground-water contamination by organic
compounds are NAPL and sorbed organic  contaminants within the vadose and, especially,
saturated zones (Palmer and Fish, 1992; U.S. EPA,  1993c). NAPL infiltration into the subsurface
is complex and is influenced by physical properties  of the NAPL and by macro- and micro-scale
geologic features (Mercer and Cohen,  1990; Cohen  and Mercer, 1993). Careful evaluation of the

                                              13

-------
Table 3.  Source Characterization Information for Conceptual Site Model Development
                   Data Type
                 Utility
  Three-dimensional distribution of physical contaminant
  phases (e.g., nonaqueous phase liquid (NAPL) and
  sorbed materials) that are continuing sources for
  ground-water contamination.
  Contaminant release and source removal/control
  histories (i.e., timing and descriptions).
  Constituents that were released, those that are currently
  present, and toxic transformation products.
Conceptualize, in conjunction with hydrogeologic data,
three-dimensional migration pathways of both NAPL
and aqueous-phase contaminants.

Project possible behavior of dissolved plume.

Identify need for source controls and estimate range of
possible time frames for restoration of the plume.

Identify areas where contaminant attenuation rate may
not be sufficient to meet contaminant reduction
objectives (i.e.,"recalcitrant zone").

Aid in identifying appropriate monitoring locations and
depths.

Constrain interpretations of contaminant migration
and historical trends in observed contaminant
concentrations.

Identify potentially dominant transport and fate
processes based on chemical properties.

Identify monitoring constituents.
extent of NAPL infiltration and migration is needed because NAPL may have migrated both
vertically and horizontally far from the original site of the release. Source materials may greatly
affect MNA remedy performance with respect to attainment of contaminant reduction goals, the
geometry of the associated dissolved plume, and design of the performance monitoring program. It
is expected that source control will be a fundamental component of any MNA remedy (U.S. EPA,
1999a).

Detailed definition of contaminant distribution throughout the three-dimensional boundaries of
the contaminated aquifer provides data that can be correlated with the hydrogeologic
characterization data to determine the effects of the hydrogeologic controls on contaminant
migration.  The data generally should be sufficient to define the zones of greatest contamination,
rapid contaminant migration and greatest risk to possible receptors in addition to defining the
plume boundaries in order to target these zones during the performance monitoring program.

The effects of the dominant attenuation processes may be evaluated and field-scale attenuation
rates estimated (U.S. EPA, 1998a; Wiedemeier etal, 1999; U.S. EPA, 2002b) in order to identify
and monitor the controls on plume stability and project progress toward remedial action
objectives of reduction in  contaminant concentrations to target levels. Several processes may
control the fate of the dissolved plume (e.g., the processes that are the components of the
attenuation rate: dispersion, dilution, sorption processes, volatilization, and chemical and
biological degradation).  The need for determining the contribution of each component of the
attenuation rate will vary  depending on remedial goals.  For instance, only chemical and
biological degradation actually destroy contaminant mass; the other processes that  only lower
contaminant concentration, retard contaminant migration, or move contaminants to other media
may not produce acceptable remedial results at all sites.
                                                14-

-------
The controls on each process and the potential for continuation of the processes at current rates
throughout the remediation time frame typically should be considered because the effectiveness of
some of the processes may vary over time, invalidating predictions of future effectiveness based
on historical rates. For example,  continued biodegradation of chlorinated solvent contamination
may be contingent on the continued supply of readily degradable organic carbon compounds to
serve as electron donors in the biotransformation processes (Wiedemeier etal., 1999; Leahy and
Shreve, 2000).

Data on the shape and dynamic behavior of the dissolved contaminant plume collected over a
period of several years are helpful in order to evaluate natural attenuation processes and develop
the monitoring plan. Three to five years of periodic monitoring (during the site characterization
and remediation effort) may, in many cases, be sufficient to form a conceptual model of plume
behavior adequate for developing the performance monitoring plan. However, more complex or
highly variable sites may require  longer-term characterization to adequately evaluate the range of
plume behavior due to variations in factors such as ground-water flow and biological activity
(Barcelona etal., 1989).  For the purposes of performance monitoring plan development, these
data are used to determine whether significant temporal (e.g., seasonal) fluctuations  in plume
boundaries are occurring.  This information is needed to site wells to monitor long-term plume
stability and to trigger implementation of contingency remedies based on observed plume
migration. The data may also indicate portions of the plume where the progress toward
contaminant reduction goals may be slow and enhanced monitoring may be warranted to
determine the cause and any necessary remedy modifications.  It is important to note that a few
years of site characterization monitoring data are not  reasonably sufficient to accurately predict
plume behavior (and performance monitoring needs)  for decades. However, such data can
appropriately be used to refine a conceptual model sufficient for initial design of a performance
monitoring program. The monitoring program in operation  can then provide data for continual
refinement of the conceptual model and, subsequently, of the monitoring program itself.

2.4.3   Geochemistry

Characterization of subsurface geochemical environments and their variability provides
important insights into the types of biotic and abiotic  processes that may be affecting plume
behavior.  Many of the processes  driving plume behavior cannot be measured directly (e.g.,
biological transformation of contaminants). However, the processes may cause changes in
geochemical parameters, leaving an observable geochemical "footprint" that can be related
qualitatively and quantitatively to the biotic and abiotic processes (National Research Council,
2000).

In general, fuels serve as electron donors during microbial degradation of the fuels (i.e., the fuels
are oxidized during microbial metabolism). More oxidizing redox conditions and greater
availability of electron acceptors at a site lead to  more efficient biodegradation processes for fuel
degradation  (Ludwig etal., 2000). In contrast, chlorinated solvents may serve as electron
acceptors in microbial metabolism (i.e., the solvents can be reduced during microbial metabolism),
and more reducing redox conditions lead to more efficient biodegradation (Chapelle, 1996). Note
that some chlorinated solvents such as trichloroethene (TCE), dichloroethene (DCE), and vinyl
chloride (VC) can also act as electron donors, and degrade under more oxidizing conditions.

Since  the degradation of fuels and solvents is influenced by redox conditions, assessment of
ambient redox conditions is an important component  of any monitoring program for monitored
natural attenuation of fuels or solvents (U.S.  EPA, 2000b). The nature of this assessment may  be

                                             15

-------
as simple as delineating the distribution of oxic/anoxic ground water, or it may be more in depth
requiring the delineation of oxic, iron(III)-reducing, sulfate-reducing, and methanogenic zones at
the site (Chapelle etal, 2000).  The role of Eh measurements in redox assessments is subject to
numerous uncertainties  (Barker, 2000), but Eh measurements are qualitative indicators of redox
conditions (Westall, 2000).  The list of redox parameters given by Wilson (2000) can be used to
develop a site-specific monitoring program for redox parameters.  The appropriate level of
monitoring can only be determined on a site-by-site basis (Wilson, 2000).

Detailed discussions of biogeochemical reactions pertinent to fuel and chlorinated solvent
contamination in ground water and the geochemical patterns associated with these processes may
be found in Azadpour-Keeley etal. (1999), Azadpour-Keeley etal. (2001), National Research
Council (2000), U.S. EPA (1998a), and Wiedemeier etal. (1999).

Geochemistry can provide the following kinds of information:

      Whether ambient redox conditions favor the natural attenuation of the contaminants of
       concern,

      Whether stoichiometric relationships between electron acceptor (oxygen, nitrate, sulfate
       etc.) utilization and contaminant degradation are observable,

      Whether redox conditions or other geochemical conditions could enhance the mobility of
       certain contaminants of concern (e.g., manganese or arsenic), or

      Whether certain geochemical changes (e.g., depletion of dissolved oxygen or other electron
       acceptors) can be used to trace the movement of impacted ground water.

With respect to development of a performance monitoring plan, geochemical data serve to:

      Identify dominant degradation processes and long-term monitoring parameters  indicative
       of  the continuing effectiveness of those processes,

      Identify geochemical environments that may result in mobilization of naturally-occurring
       inorganic compounds such as arsenic or manganese species, and

      Identify zones  beyond the current  plume boundaries where soluble electron acceptors or
       donors are depleted or nonhazardous reaction products are enriched with respect to
       ambient ground water but contaminants  are not detected. The water in these zones has
       been called "treated water" (i.e., water that once was contaminated but has been
       remediated by natural attenuation). Since the plume would travel into these zones if it
       expands, such zones can serve as target zones for monitoring plume stability.

Geochemical parameters and trends that are often useful indicators of biotransformation
processes at sites with fuel and chlorinated solvent contamination are discussed in U.S. EPA
(1998), Wiedemeier etal. (1999), and Wiedemeier and Haas (2002).  The individual parameters
diagnostic of dominant processes and most useful in a performance monitoring program depend
on site-specific conditions.  Parameters to  be measured are chosen with regard to their potential to
affect site-related decisions  (i.e., it would not be useful to measure a given parameter if the
information would not be used to change site-related decisions).

                                             16

-------
2.4.4  Receptor Locations

For the purpose of specifying performance monitoring locations, identification of human and
ecological receptors that may be affected by the contaminant plume under current and reasonably
anticipated future conditions is a critical element. In addition to possible receptors within the
areas that are currently contaminated, locations outside of the plume that may be subject to
impact during the remedy performance period should be identified.  Such receptor locations may
include, but are not limited to, water supply wells, buildings, aquifers, wetlands, sediments, and
surface-water bodies.  In many instances, this list may include aquifers that are below zones of
shallow contamination.  The vapor intrusion to indoor air pathway is also of considerable
importance (U.S. EPA, 2002d).  This  document focuses on monitoring the saturated zone, but site
characterization and monitoring for any remedy should include appropriate monitoring of all
significant pathways by which contaminants may migrate and impact receptors. Sources of
additional information concerning receptor identification include U.S. EPA (1997c) and U.S. EPA
(1998b).

Information on receptors and possible pathways for impact is used to identify appropriate
locations for monitoring points between the plume and the possible receptor as well as locations
where impact may occur. Monitoring at these locations provides confidence that no unacceptable
impacts occur.

2.5    Monitoring Network Design

2.5.1  Introduction

The media of primary concern and, therefore, the focus of the monitoring network design, will be
ground water at many sites. However, monitoring of other media (e.g., indoor air, soil gas, soils,
surface water, and sediments) may be necessary to determine possible impacts to receptors and
other measures of remedy effectiveness. Guidance regarding the assessment of the indoor air
exposure pathway is provided in U.S.  EPA (2002d).  Recent discussion of the assessment of the
interactions of ground water and surface water useful in formulating effective monitoring
networks may be found in U.S. EPA (2000e). At many sites, the assessment of such cross-media
transfers of contaminants is essential to adequate performance monitoring for MNA remedies.
The focus of the following discussion is primarily on the monitoring of ground water as this is
often an exposure pathway of significant concern. Performance monitoring is extended to other
media as warranted by site conditions.

A monitoring system designed for evaluating the performance of an  MNA remedy with respect to
specific remedial action objectives may be very different from the network established during
earlier phases of site characterization, feasibility studies, or interim actions.  Existing wells were
often incrementally installed for different purposes, such as defining the extent of contamination
during the remedial investigation or evaluating an engineered remedy during a  feasibility study.
The location and number of these wells may not be well suited for MNA performance
monitoring.  Specification of a monitoring network design should be based on consideration of all
available information concerning the  processes and factors expected to control contaminant
distribution. The network is designed to provide data to demonstrate attainment of all the
remedial action objectives for an MNA remedy.

A plume is a dynamic, three-dimensional distribution of contaminants in ground water, that
generally necessitates three-dimensional monitoring. Plume shape is influenced by many factors,

                                             17

-------
including original source distribution, geology, hydrology, and biologic processes. The resulting
spatial and temporal variability significantly impact choice of monitoring locations and
frequencies and necessitate continual re-evaluation of the performance monitoring network.
Appendix A provides a brief discussion of some issues regarding variability.

The density of sampling points in a monitoring network will depend on the geology and
hydrology, the spatial scales at which contaminant distribution varies horizontally, vertically and
temporally, and the desired level of confidence in the evaluation. Plumes often vary significantly
in concentration in transverse and vertical cross sections (e.g., Cherry,  1996) making evaluation of
contaminant distribution and remedy performance difficult. In these cases, a dense network of
monitoring points will often be needed to support many of the performance monitoring
evaluations described below.

One approach to this problem is to define and monitor the plume using clustered monitoring
points positioned in transects across and through the plume, perpendicular to the direction of
ground-water flow (Figures 4, 5 and 6). The horizontal and vertical spacings of the monitoring
points in each transect are determined by the hydrogeologic conditions that control contaminant
migration and the dimensions and spatial heterogeneity of the resulting contaminant  distribution.
The horizontal distance between transects  is generally based on changes in contaminant
concentration along the  plume, and the  location of the source and distal portions of the plume.
For example, transects may be placed:

       Immediately upgradient of the source area to monitor contaminant and electron acceptor
       flux into the plume,

       Immediately downgradient of the source area to monitor contaminant flux to ground
       water,

       Near the downgradient and sidegradient plume boundaries to monitor contaminant
       concentration increases indicating possible plume expansion,

       Immediately downgradient of the plume or other compliance boundary to monitor for
       plume expansion, and

       Along the plume to provide information on plume configuration and contaminant
       attenuation. Horizontal spacing of the transects for determination of contaminant
       attenuation may be based on the number of locations needed for attenuation calculations,
       and changes in contaminant  concentration along the plume (e.g., spacing conforming to
       order-of-magnitude decreases in contaminant concentration).

The elevations of sampling intervals are generally based on stratigraphy (i.e., sampling the
different stratigraphic intervals), the vertical component of hydraulic gradients, and contaminant
distribution (i.e., sampling  the top, bottom, and "core" of the plume, and, possibly, above and
below the plume to vertically bound it).

The use of a transect-based approach to monitoring may greatly reduce the uncertainty in
performance monitoring evaluations at  many sites by improving the definition of contaminant
distribution and its variability.  The transect approach helps to locate ground-water flow lines and
contaminant migration paths. Transects also provide a better definition of contaminant
distribution under conditions of changing hydraulic gradients.  A detailed example of the use of a

                                             18

-------
                                     Site (Map View)
           Upgradient
            Transect
Lateral (Side Gradient)

              B'n
             Source
              Area
                    High Concentration
                       Plume Core
                          (2)
    Low Concentration /
         Plume
        Non-Hazardous
          Degradation
4)      Products & Other
     Geochemical Indicators
                                    Ground-Water Flow
                   Target Monitoring Zones

                 1. Source area
                 2. Contaminated zones of highest
                   concentrations and mobility
                 3. Plume fringes
                 4. Plume boundaries
                 5. Recalcitrant zone determined
                   from historical trends
                 6. Upgradient and sidegradient
                   locations
                 x  Monitoring well cluster
                 A Piezometer
                -*- Transect of well clusters
Figure 4. Example of a network design for performance monitoring, including target zones for
         monitoring effectiveness with respect to specific remedial objectives. In this example,
         monitoring network design is based on transects of wells oriented perpendicular to the
         ground-water flow direction. Sampling locations for target monitoring zones were chosen
         based on site characterization. Piezometers provide additional data for evaluation of changes
         in potential ground-water flow direction.
transect-based approach in the evaluation of natural attenuation processes at a site where
petroleum products were released may be found in Kao and Wang (2001).

2.5.2  Monitoring Locations

Generally, each distinct zone of contaminant migration and geochemical regime is monitored to
assess its impact on remediation. For instance, if part of a plume of tetrachloroethene is
anaerobic with high levels of electron donors available and another part of the plume is aerobic
with few electron donors available, degradation of the tetrachloroethene may be very active in the
anaerobic zone but nonexistent in the aerobic zone. For each zone with distinctly different
conditions or controls on contaminant migration and fate, the following locations would be
monitored: areas hydraulically Upgradient and sidegradient to the plume, source area, main body
of the plume, and distal portions and boundaries of the plume.
                                                19

-------
             Area of Former Solvent
               Tankand Ongoing
                Source Controls
Monitoring Well
    Cluster
       o.
                                                                                  A1
     150
            Target Monitoring Zones

          1. Source area  I	_l
          2. Contaminated zones of highest
            concentrations and mobility
          3. Plume fringes exhibiting low
            contaminant concentrations
          4. Plume boundaries
          5. Recalcitrant zone determined
            from historical trends
          6. Background locations
                     Legend

              Gravel,gravel-sand mixtures

              Medium to coarse-grained sand

              Fine-grained silty sands

              Dissolved Plume
Figure 5. Cross section A-A'through monitoring network in general direction of ground-water flow.
         Placement of monitoring points within target zones is based on geologic controls and
         contaminant distribution characterized prior to remedy selection and is periodically
         modified, as warranted, based on evaluation of performance monitoring data. In this
         scenario, detailed site characterization data would be used to define the limits of the
         source area, the distribution of any NAPL and aqueous-phase contaminants, and the
         effectiveness of source removal and control actions. Source control activities  and
         monitoring associated with the release from the former solvent tank are not pictured.
2.5.2.1 Typical Target Zones

Typical target zones for monitoring a contaminant plume (Figures 4, 5 and 6) include:

       Source areas - within and immediately downgradient of remediated source areas

       The monitoring objectives are to determine and demonstrate whether any further
       contaminant releases to ground water occur and to estimate contaminant reduction over
       time.  In situations where the source is contained, increased contamination or new
                                            20

-------
                                        Monitoring Well
                                           Cluster
            150
                                            Legend

                                    Gravel, gravel-sand mixtures

                                    Medium to coarse-grained sand

                                    Fine-grained silty sand

                                    Dissolved Plume
Figure 6. Cross section B-B'through monitoring network perpendicular to ground-water flow.
         Monitoring points are placed to define the plume horizontally and vertically, as well as
         monitor contaminant concentration zones and geochemical zones.
       contaminants could be indicative of such conditions as cap failure, buried drums that
       rupture, a rise in the water table transferring additional contaminants from the vadose
       zone, or slurry -wall failure. These new contaminant releases could be greater than the
       capacity of the subsurface to attenuate concentrations without significant plume
       expansion or could include contaminants not effectively remediated by MNA.

       Transmissive zones with highest contaminant concentrations or hydraulic conductivity

       A change in conditions in these zones, such as an increase in contaminant mass
       from source areas or increased ground-water velocity, may lead to a relatively rapid
       impact to a downgradient receptor.

       Distal or fringe portions of the plume

       These are areas where reduction of contaminant concentrations to levels required by
       remedial action objectives may be attained most rapidly or where plume expansion may be
       observed.
                                             21

-------
Plume boundaries and other compliance boundaries

Multilevel monitoring points typically would be placed at the sidegradient,
downgradient, and vertical plume boundaries, and between these boundaries and
possible receptors.  Multilevel monitoring generally should also be performed at any
other compliance boundaries specified in remedy decision documents. Results from
these monitoring locations may directly demonstrate unacceptable plume expansion
and changes in ground-water flow directions.

Zones in which contaminant reduction rates appear to be lower than required to meet
remediation goals (i.e., recalcitrant zone)

These are the areas where attaining cleanup standards within accepted time frames
may be impeded due to site conditions (e.g., presence of previously undetected
source materials or low flux of electron acceptors).  Such areas, if present, will be
delineated through evaluation of data obtained throughout the performance
monitoring period. These areas may require additional characterization to
determine if additional remedial actions are necessary to reduce contaminant
concentrations to desired levels.

Areas representative of contaminated and uncontaminated geochemical settings

Sampling locations for monitoring the geochemical setting include monitoring
points that are hydraulically upgradient and sidegradient with respect to the plume.
Since assumptions  concerning the geochemical setting and naturally occurring
changes in geochemical parameters affect interpretation of data from the plume,
such assumptions should be tested and evaluated like other parts of the conceptual
site model.  Therefore, multiple monitoring points generally should be used to
determine the  variability of geochemical conditions outside the plume. Data
concerning the movement of electron acceptors, donors, and any contaminants into
the plume aid in understanding and interpreting data from the plume. These
geochemical data are used to determine whether the observed differences in
geochemical parameter concentrations within the plume are due to contaminant
transformation processes rather than natural variations  in the background
geochemical conditions. The locations sidegradient to the plume help to evaluate
changes in plume geochemistry with time as ground water migrates  through
uncontaminated aquifer materials.  Changes in geochemistry within the plume may
not be directly related to attenuation of the contaminants, so geochemical changes
outside the plume generally should be assessed and compared to geochemical
changes taking place within the plume. If upgradient and lateral monitoring points
show geochemical changes similar to changes in the plume, such changes may not
be attributed solely to contaminant-related processes (i.e., degradation), and,
therefore, may not serve as supporting evidence for degradation processes.

Areas supporting the monitoring of site hydrology

At some sites,  monitoring of ground-water elevations at locations additional to
those used for the monitoring of chemical parameters may be needed to
determine hydraulic gradients. At such sites, appropriate locations for placing
piezometers will often include positions that are upgradient and sidegradient to

                                      22

-------
       the contaminant plume, as well as in zones above and below the plume.  Piezometers
       are usually spaced across the site so that ground-water elevation measurement errors
       are relatively small compared to the difference in ground-water elevations between
       piezometers.

2.5.2.2 Screen Lengths

The screen length for a given well constitutes an important part of the three-dimensional
monitoring location.  The screen is sized to sample the interval of interest.  The interval may be
defined by stratigraphy, contaminant loadings, or geochemistry, based on the site characterization
data.  Factors to consider in determining screen length include the following:

       Well screens can be matched to stratigraphic intervals if the intervals are relatively
       small, and the geochemical and contaminant values are similar throughout the vertical
       extent of the interval intersected by the well screen.  For instance, a five-foot thick
       sand layer within which contaminated water migrates  could be sampled with a five-
       foot screen to intersect the entire interval.  Any interval of significantly different
       conductivity could be targeted  by a specific screen length. Well clusters can be used to
       provide  coverage of the entire contaminated unit, as needed.

       Monitoring wells typically should screen comparable intervals.  For example, well
       screen lengths can be matched to contaminant loadings.  Suppose that dissolved
       petroleum contaminants (e.g., benzene, toluene,  ethylbenzene, and xylenes) were
       transported primarily within a 5 ft  to 10 ft thick interval of a 30-foot thick sandy unit.
       In this situation, a monitoring  well screen could be sized to sample the most
       contaminated part of the unit to help determine attenuation of the contaminants in
       that interval.  The use of longer screens may result in artificially lower measured
       containment concentrations, or even lack of detections, due to the mixing of water
       with different chemical compositions. Calculated attenuation rates or estimated plume
       boundaries may reflect variations in screen length and placement rather than actual
       attenuation if monitoring well  screen intervals are not matched to contaminant
       distribution in the contaminated stratigraphic interval.

       Well screen lengths can be matched to geochemistry, to sample a zone where a particular
       geochemistry prevails.  Since attenuation of some contaminants is highly sensitive to the
       geochemical environment, it is often desirable to be able to accurately identify and
       discretely sample locations in the plume where  a particular geochemistry prevails.

2.5.3   Monitoring Parameters

The primary classes of parameters to be monitored during these evaluations generally will be the
contaminants of concern (COCs), geochemical indicators of transformation processes, and
hydrogeologic parameters. Contaminants of concern are those chemicals identified during site
investigations that are required to be addressed by the response action proposed in the remedy
decision documents.

Contaminants of concern may be identified from:

       Ground-water and soil monitoring  data,

       Contaminant source histories,  and

                                             23

-------
       Evaluation of contaminants that potentially may be formed or mobilized as a result of
       biotransformation processes or changes in the geochemical environment. For example,
       toxic products such as dichloroethene and vinyl chloride would be measured at sites
       contaminated with chlorinated solvents where they might be expected to be present.

At some sites, it may be necessary to identify and include chemicals that have only been
tentatively identified in previous sampling (e.g., Tentatively Identified Compounds  (TICs)). Since
the effectiveness of natural attenuation processes varies markedly for different contaminants, it is
often inappropriate to assume that other components of the plume will be adequately remediated
along with the primary contaminants of concern during MNA.  Potentially hazardous plume
components should be positively identified, evaluated to determine if they will be sufficiently
mitigated by MNA, and monitored, as appropriate, if they are a threat to human or environmental
health. For instance, MTBE (commonly found in gasoline-sourced plumes), was overlooked at
most sites for many years because  sampling and analysis at most such sites emphasized BTEX.
However, at many such sites, BTEX components may be attenuated to meet standards before
MTBE, so monitoring only BTEX  may result in a mistaken interpretation of the effectiveness of
MNA.  Another example of an overlooked contaminant or TIC is 1,4-dioxane, which may be
associated with certain plumes of volatile organic contaminants. Recent advances in analytical
methods allow detections in ground-water samples at much lower concentrations (e.g., 1.5 |ig/L)
than previously possible.  1,4-Dioxane has been used as a solvent stabilizer in 1,1,1-
trichloroethane at concentrations of 2% to 8% by volume (Mohr, 2001) and is a class II-B
probable carcinogen.  This contaminant is highly soluble in and mobile in ground water. At sites
where this contaminant is present, it may be expected to migrate further than the associated
plume of chlorinated solvent compounds (Fetter, 1994). At such sites, a monitoring program that
does not include 1,4-dioxane would likely result in a mistaken interpretation of the effectiveness
of MNA.

At a well-characterized site, the DQO process can be used to  choose parameters and monitoring
frequencies for each monitoring location based on the value of the data to monitoring MNA, as
an alternative to the measurement of all parameters at all locations. For example, if several years
of monitoring indicate that the geochemistry in the central portion of a BTEX plume is stable, it
may not be necessary or useful to continue to analyze samples from these locations for all of the
geochemical parameters at each sampling event.  However, all contaminants normally would be
measured in all samples on a periodic basis to verify the conceptual site  model. The most
important monitoring parameters are derived from a thorough understanding of site history,
ambient geochemistry, contaminant sources, and geochemical changes induced by contaminant
degradation.

The purposes for long-term monitoring of geochemical parameters in addition to ground-water
contaminants include:

       Confirmation of dominant biotransformation processes,

       Evaluation of the potential for continued transformation, and

       Identification of the zones  of contaminant migration  (e.g., where geochemistry indicates
       contaminant degradation has taken place).

These purposes are discussed in Azadpour-Keeley etal. (1999), Azadpour-Keeley etal. (2001),
U.S. EPA (1998a), Wiedemeier etal. (1999), and Wiedemeier and Haas  (2002).

                                             24

-------
The specific parameters useful at a given site depend on site contaminants, natural geochemical
settings, and dominant biotransformation processes. In most cases, a select group of parameters
that indicate the geochemical environment (e.g., oxidation-reduction potential, dissolved oxygen,
pH), identify geochemical regimes affecting contaminant degradation (e.g., nitrate, iron (II),
sulfate, methane), or are products of contaminant degradation (e.g., ethane, ethene) will be
measured in most samples.  In addition to the measurement of geochemical parameters in ground-
water samples, periodic monitoring of solid-phase electron acceptors, such as bioavailable  iron, in
aquifer materials may be useful at some sites to evaluate the supply of such materials relative to
the mass of contaminants to be degraded (Ruling etal, 2002). The geochemical parameters
measured generally should be chosen based on the utility of the data for affecting site-related
decisions (i.e., if no decisions would be changed based on the data,  then the data need not  be
collected).

The primary hydrologic parameter  of interest is the elevation of ground water in monitoring wells
and piezometers. Estimates of hydraulic gradients and changes in the elevation of the water table
with respect to locations of residual source materials (i.e., any source materials that remain after
all appropriate source control actions have been completed in accordance with OSWER Directive
9200.4-17P (U.S. EPA, 1999a)) are fundamental to all evaluations  conducted during performance
monitoring.  The wells and piezometers used in these evaluations should be discretely  screened in
individual hydrogeologic units within and adjacent to the contaminant plume. Additional  data
that are often essential in assessing ground-water flow include surface-water elevations, local
rates and schedules of irrigation, local precipitation data, and pumping rates and schedules for
nearby wells. The evaluations should be three-dimensional  in nature, including both horizontal
and vertical  components of hydraulic gradients that control three-dimensional contaminant
migration. In many cases, the vertical component of the hydraulic gradient may be larger  and
may display  more variability than the horizontal component. Depending on the anisotropy in the
hydraulic conductivity of aquifer materials and the magnitude, direction, and duration of the
hydraulic gradients, vertical movement of ground water and contaminants may be significant and
fluctuations  in both the horizontal  and vertical components of ground-water flow may exist.

2.5.4   Monitoring Frequency

    Monitoring frequency affects the ability of the performance monitoring program to:

       Provide timely warning of impact to receptors,

       Detect contaminant releases to ground water that warn of possible plume expansion,

       Detect changes in plume size/concentration,

       Determine temporal variability of data,

       Detect changes in geochemistry that warn of changes in attenuation, and

       Yield data necessary to reliably evaluate progress toward contaminant reduction
       objectives.

The most appropriate frequency for ground-water sampling depends on the rate with which
contaminant concentrations change due to ground-water flow and natural attenuation processes,
the degree to which the causes of this variability are known, the types of evaluations to be

                                             25

-------
performed, the location (s) of possible receptors, and the RAOs for the site.  Based, in part, on
previous studies (e.g., Barcelona et al, 1989), quarterly monitoring will often be an appropriate
frequency to establish baseline conditions over a period of time sufficient to observe seasonal
trends, responses to recharge, and to estimate attenuation rates for key contaminants. Quarterly
monitoring for several years provides baseline data to determine trends at new monitoring points
and test key hypotheses of the conceptual site model.  In situations where hydrologic, geochemical
and contaminant trends are stable and the conceptual site model is verified by measured site data,
reductions in sampling frequency may be warranted.  In situations where variability is high,
increases in monitoring frequency and additional investigations to determine the source of the
variability may be warranted.  More frequent monitoring may be appropriate under
circumstances where ground-water flow is rapid and/or contaminant travel time to receptors is
short.

More frequent monitoring of ground-water elevations may be warranted, particularly during the
establishment of baseline conditions, to improve the characterization of ground-water flow
patterns. In addition, more frequent monitoring may be  needed to observe changes in ground-
water flow patterns in response to other site  activities, such as the start or cessation of ground-
water extraction, source control activities,  and other significant changes in the hydrologic system.
In some cases, monitoring of this parameter on a very frequent basis using automated recording
equipment may be needed to determine the effects of variability in recharge and discharge rates or
locations on ground-water flow patterns. This will aid in specification of appropriate long-term
monitoring frequencies and locations for evaluating fluctuations in hydraulic gradients and trends
in chemical parameters. Based on the results of such assessments, initial monitoring frequencies
may be adjusted to adequately capture the  fundamental features of the trends. In situations
where temporal trends in hydraulic gradients are absent or well characterized, monitoring of
ground-water elevations at  a frequency no less  than that  of the chemical parameters is generally
appropriate.

Several years of monitoring data are typically necessary for estimation of the site variability and
expected rates of change in ground-water flow, contaminant concentrations, and geochemistry
(Barcelona etal., 1989). Once site characterization and initial performance monitoring activities
have provided these data, re-evaluation of the monitoring frequency may be warranted if trends
are established and the remedy is progressing as expected. Increases and decreases in monitoring
frequency may occur several times over the life of the remedy in response to changes in site
conditions and monitoring  needs (Figure 7).   For example, note in Figure 7 that monitoring
frequency was decreased in response to stable contaminant concentration declines, increased in
response to a sudden increase in contaminant concentration, decreased again based on the
reestablishment of stable contaminant concentration declines, and increased yet again with the
implementation of the verification monitoring  program.

Specifically, monitoring frequency generally should be related to detecting changes in site
parameters that indicate ability of the MNA  remedy to achieve site-related remedial action
objectives, and to provide early warning of possible impact to receptors. The monitoring
frequency may be adjusted  to capture information regarding trends of interest while eliminating
unnecessary redundancy. However, monitoring frequency determinations should take into
account that apparently stable site parameters may sometimes change rapidly and nonlinearly due
to natural or anthropogenic causes.  For  example, contaminant releases or remediation activities
on adjacent properties may cause chemical or hydrologic changes in plume properties at the site.
Increased rainfall can change  recharge rates,  ground-water extraction can cause changes in
ground-water flow, and fertilization of fields, golf courses or lawns can cause an influx of nitrate.

                                             26

-------
(LU)
jaje/v\-punoJ3
                                                          c
                                                          o
                                                          o
                                                          o

                                                          o
                                                         o
                                                         o
                                                        o
                                                         o
                                                         o
                                                         o
                                                      o

                                                      o
                                                    o
                                                 o
                                            o
                                          o
                                        o
                       01
-Q

 Oi


 "o.

 E
 re
"1/1
                                            O
                                       "0

                                        O
                                     O

                                     O
                                        O

                                       O

                                       o
                                    O

                                 O

                                     o
                                I	I-
                                                      Dl
                                                      3.

                                                      o
                                                                                 01
                                                                                 4-
                                                                                 

                                                                                                             o
                                                                                                             ^
                                                                                                             u
                                                                                                             c
                                                                                                             o>
                                                                                                             3
                                                                                                             cr
                                                                                                             01
                                                                                                            vt
                                                                                                             01
                                                                                                             o
                                                                                                             E
                                                                                                             c

                                                                                                             i/i
                                                                                                             Ol
                                                                                                             01
                                                                                             u

                                                                                            _OJ
                                                                                                             O
                                                                                                             Q.
_OJ

 Q.


 I
 X
LJJ





 2
 3
 O)
                                 27

-------
Water-table variations and changes in velocity are common in flood plains, particularly near large
rivers that have major changes in the stage of the river. In some climates, most of the recharge to
aquifers due to precipitation occurs in only a few months of the year. In some settings, water
extracted by trees from shallow water table aquifers can influence the direction of ground-water
flow.  Irrigation or municipal water supply wells that pump intermittently can change ground-
water flow in an irregular manner.

Examples of factors influencing specification of the monitoring frequency are provided below.
Each example  considers only the point under discussion.  The  actual monitoring frequency
selected would be based on consideration of all the factors bearing on the given parameter and
monitoring location.

       Monitoring Frequency Determination Based on Possible Contaminant Travel Time to
       Receptors

       Consider a site with the following conditions: the travel time for ground water from the
       downgradient plume boundary to a receptor was expected to be  two years, based on
       analysis of the ground-water flow field. The time required to design,  construct and bring
       an alternative remedy online to intercept the contaminant and protect the receptor would
       be added to the ground-water travel time. In this hypothetical situation, monitoring
       contaminant concentrations at wells placed downgradient of the current plume boundary
       to detect plume expansion  would be conducted at least  annually, and, in most cases, more
       frequently (e.g., quarterly)  to allow time for confirmation  of detections and other
       contingencies. Monitoring  may be more frequent for wells near  and downgradient of the
       plume  boundary or other compliance boundaries than for interior wells. In this example,
       the travel time for a conservative constituent was used  in the calculations to allow a more
       conservative estimate of the minimum contaminant migration time than that obtained
       through use  of an estimated transport velocity for a contaminant that may undergo
       sorption or biotransformation reactions.

       Monitoring Frequency Determination Based on Evaluation of a Cyclic Change

       Some sites, such as sites where the climate produces a pronounced wet season/dry season
       with associated changes in recharge rates and ground-water flow characteristics, may
       display pronounced cyclical trends in contaminant concentrations and plume boundaries.
       For example, seasonal  changes may involve seasonal inputs of contaminant from the
       vadose zone, and may  require more frequent monitoring or changes in monitoring timing
       relative to cyclical changes (i.e., the particular dates when samples are taken) (Figure 8).
       By reviewing the historical variability in water levels for the site and recorded climatic
       variability (i.e., drought frequencies, periods of above-average rainfall), a plan for
       specifying the monitoring frequency may be established.  The objective of such a plan is
       to increase monitoring frequency sufficiently to prevent unmonitored expansion of the
       plume  while avoiding collection of unnecessary data. Typically,  data gathered over several
       years are necessary to  evaluate seasonal trends and determine what frequency and time of
       monitoring are appropriate to capture changes related to seasonal trends (Barcelona et al,
       1989).  Examination of prevailing annual weather patterns (available from National
       Weather Service historical  data) may be helpful for determining appropriate sampling
       frequencies.
                                             28

-------
                                                            Effects of Monitoring Frequency
c
o
COC Concentrat
35.0
30.0
25.0
20.0
15.0
10.0
5.0

A. /\ A A
/ v \ / \ /





               Quarterly 2, Starting in February
                         Date
The following example is used to illustrate the effect of different
monitoring frequencies and timing on monitoring data and data
interpretation.

Site Characteristics

The site had a leaking underground storage tankthat contained gasoline.
The tank has been removed and source control actions implemented, but
some undetected petroleum product remains at residual saturation in the
vadose zone and saturated zone extending downgradient of the source
area. The water table varies about four feet in elevation throughout the
year.

This example is focused on one monitoring well from the multi-well
performance monitoring network. The well is constructed  with a ten-foot
screen and located 100 feet downgradient from the LNAPL release area.
Travel time for contamination to move from the release area to the
monitoring well varies, but averages about five months. The top of the
well screen is located at 20 feet below ground surface (bgs)(/.e., about
one foot above the highest observed  water table elevation). The bottom
of the well screen is at 30 feet bgsand is terminated at a regionally
extensive, uniform clay layer.

The ground water is contaminated from the water table to a depth of
about 28 feet bgs. There is an uncontaminated zone about two feet thick
over the clay confining layer, so that some clean water is intercepted by
the lower part of the well screen.

During  the rainy season in the fall (November-December) and spring
(March-May),the water table reaches its"maximum"elevation at about 22
feet bgs, which is about two feet below the top of the well screen. The
higher water table intercepts more of the previously undetected residual
LNAPL and sorbed contaminants in the vadose zone in the release area,
causing a pulse of contaminants to enter the ground water.

From December to March and May to June,the water table declines to its
"average"elevation,at about 24 feet bgs. During July,a nearby irrigation
well comes online,and continues pumping until sometime in September,
causing the water table to decline to elevations as low as 26 feet bgs, its
"minimum"elevation.Also,ground-waterflowis shifted so that the
monitoring well is not directly downgradient of the source zone.

During  periods of high water table elevations, flushing and interception of
contaminated vadose zone by infiltration and the elevated water table
cause pulses of contaminant to move toward the monitoring well,arriving
several  months after the rain.  During  periods of low water table
elevations, less residual NAPLis intercepted by the ground water, so
water with lower contaminant concentrations moves toward the
monitoring well. Also,the shift in ground-water flow direction caused by
the irrigation well causes the monitoring well to intercept the fringe of
the plume, rather than the more central, highly contaminated volume of
the plume.

Results

The sampling data from the monitoring well are shown in the adjacent
panels. The trend line in each graph shows the estimated linear trend of
contaminant concentration based on the data in the chart. It can be seen
that at sites where conditions cause significant changes in the
contaminant concentration throughout the year, the frequency of sam-
pling and the timing of the sampling cycle relative to the contaminant
concentration cycle can strongly affect the data obtained and the
interpretation of the data. These data highlight the importance of site
characterization in choosing monitoring frequency and, in this example,
identifying the need for additional source controls for effective
remediation.
Figure 8. Monitoring frequency effects on sampling data collection and interpretation. The relationship
           of monitoring frequency and timing to cyclical changes in contaminants may significantly
           affect data collection and interpretation.
                                                           29

-------
       Monitoring Frequency Determina tion Based on Relevance of Parameters

       If a parameter is not expected to significantly influence evaluations of remedy
       performance at  a site, then monitoring frequency for that parameter could be greatly
       reduced or eliminated.  If information gained from frequent sampling for a given
       parameter would not be reasonably expected to change site-related decisions, sampling for
       that parameter could be reduced in frequency. However, absent specific reasons for
       excluding a parameter, it generally would be expected that the entire suite of geochemical
       parameters relevant to the site, as well as contaminants and hydrogeological parameters,
       would be measured at all monitoring points on a set schedule. This allows an evaluation
       of continued stability in the geochemical setting and the potential for changes in
       biotransformation processes and attenuation rates to be performed.

       Monitoring Frequency Determination Based on Information Redundancy

       If, over a period of several years, data trends are stable, a reduction in monitoring
       frequency may be warranted (Figure 7). Also,  if two or more wells sampling the same
       zone are located close together, and consistently produce similar data, changes in the
       monitoring frequency of one or more of the wells may be considered. For example, if the
       geochemical indicators at a given monitoring location are stable over a long period (e.g.,
       within a range indicating suitable conditions for degradation of the contaminants), then
       monitoring for these geochemical indicators could be reduced in frequency, to the point of
       being measured annually, biennially, or at even greater intervals. However, the possibility
       of a rapid change in a previously stable pattern may affect determination of the
       monitoring frequency.  For instance, nitrate concentrations, that can 1) change quickly
       due to anthropogenic inputs and 2) strongly inhibit reductive dechlorination of the
       chlorinated solvent compounds, generally should continue to be monitored frequently  at
       sites where reductive dechlorination is an important component of the MNA remedy.

       Contrastingly, for example, if a sudden change in contaminant concentrations were noted
       (Figure  7), an increase in monitoring frequency may be warranted to provide information
       to facilitate understanding of the change and provide earlier warning of further change.

2.6    Demonstrating MNA Effectiveness with Respect to Remedial Objectives

Techniques for  demonstrating the effectiveness of MNA, with respect to the remedial objectives
provided in Table 1, are discussed below. The discussions are intended to provide general
guidance on the types of assessments, monitoring network designs, parameters, and evaluation
frequencies. However,  specific study designs depend on site conditions and the site-specific limits
on decision errors. For additional information concerning remedial objectives, refer to current
program-specific guidance.

2.6.1   #1 - Demonstrate that Natural Attenuation is Occurring According to Expectations

Although remedial expectations and, consequently, appropriate performance monitoring analyses
are site specific  in nature, reduction in contaminant concentrations to specified levels is a  general
expectation for  most selected remedies. Other common goals, such as the prevention of
additional contaminant migration, are discussed in the following sections. Data analyses useful
in evaluating progress toward contaminant reduction  objectives include evaluation of temporal
trends in contaminant concentrations or mass, comparisons of observed contaminant
distributions with predictions or required milestones, and, in some cases, comparison of
calculated attenuation rates with the range of rates required to meet remedial objectives within

                                             30

-------
the required time frame. Evaluations of adequate progress toward restoration objectives are
difficult due, in large measure, to subsurface variability and to a lesser extent, measurement
variability (Appendix A).  This will often necessitate use of multiple lines of evidence (e.g.,
temporal trends and estimates of contaminant mass loss, as discussed in the following sections)
and relatively dense monitoring networks to reduce uncertainty to acceptable levels.

2.6.1.1 Temporal Trends in Individual Wells

Temporal trends in the concentrations of all contaminants of concern  measured in ground-water
samples are essential indicators of plume stability and progress toward contaminant reduction
objectives.  Temporal trends in an individual well can sometimes be used to estimate the potential
lifetime of the plume at that location  (U.S. EPA, 2002b).  However, contaminant trends at
monitoring points located throughout the plume will be needed to adequately interpret progress
toward most contaminant reduction goals.

Wells used for analysis of temporal trends generally should be screened in portions of the plume
representative of areas where different processes may be dominant (e.g., different
biotransformation processes or processes that mobilize naturally occurring toxic metals) and
representative of the range of contaminant concentrations observed at the site.  Such areas
(Figures 4, 5, and 6) include:

       In and immediately downgradient of suspected or former source areas,

       In the most contaminated zones downgradient of these areas,

       In areas (recalcitrant zones) where contaminants may be attenuating at rates too low to
       meet remediation goals due to differences in geochemistry or other factors,

       In fringe areas of the plume,

       In locations surrounding the plume or other compliance boundaries,

       In contaminated zones with the highest ground-water flow or contaminant migration
       rates that may serve as pathways for rapid migration.

Statistical techniques, such as those described in Gibbons (1994), Gilbert (1987), Helsel (1995),
and U.S. EPA (2000c), may be used to objectively determine whether contaminant
concentrations are  increasing or declining with time, and to compare trends between wells. A
trend may be assessed to determine if the trend would be sufficient to  meet remedial goals in the
desired time frame.

Temporal trends observed in wells may result from processes such as biotransformation that
reduce contaminant mass or concentrations.  However, temporal trends may be the result  of
contaminant release history and effects of source control actions.  Trends may also result from
processes such as changes in ground-water elevations (Figure 8) or ground-water flow directions
(Figure 9) that merely change the shape of the plume or retard its movement, rather than
degrading contaminants.  For instance, changes in ground-water flow patterns may cause a
change in plume shape, perhaps causing some wells to have lower contaminant concentrations,
and other wells to have higher concentrations.  Ground-water flow patterns vary in response to
temporal differences in the rates and  locations of recharge (e.g., precipitation or irrigation) and

                                             31

-------
discharge (e.g., pumping of potable or agricultural wells or operation of drainage systems) at many
sites. Variations in ground-water flow directions and rates may occur on scales of hours to years
and may influence contaminant concentration trends in individual wells.

Variations in recharge may cause pulses of contaminants to be released from poorly controlled
sources, or may provide pulses of relatively uncontaminated ground water.  For instance, rain may
infiltrate through a vadose zone source, or higher water tables may intercept vadose zone
contamination, causing increased contaminant concentrations in ground water. As pulses of
contaminants or uncontaminated ground water move downgradient,  contaminant concentrations
in monitoring wells may increase or decrease temporarily, perhaps leading to false inferences
about contaminant attenuation.

Trends in contaminant concentrations may also result from measurement variability. For
example, changes in sample collection and analysis procedures and personnel can introduce
variability into contaminant data. Consideration of these sources of variability may be helpful for
interpretation of patterns in the data, especially when abrupt changes in established patterns are
noted.

Due to the  variety of possible causes, it may be difficult to isolate or identify specific causes of
contaminant concentration trends either in individual wells or groups of wells.  The range of
viable hypotheses that may explain observed trends generally should  be incorporated into the
conceptual site model and consistently used for evaluating site data in order to obtain accurate
interpretations of the data.  Contaminant concentration changes in wells ordinarily should be
evaluated in the context of the conceptual site model and sampling history before they are
attributed to natural attenuation processes or other possible causes.

A site-wide trend cannot be satisfactorily elucidated from monitoring one sampling location, or
one target monitoring zone within a plume. Significant reductions in contaminant
concentrations from all or most sampling locations may indicate decreases in contaminant mass
and imply progress toward restoration.  Mixed trends in different portions of the aquifer (Figure
9) may imply temporal changes in plume  migration and an inadequate conceptual site model
rather than contaminant mass loss.

2.6.1.2 Estimation of Contaminant Mass Reduction

Estimation of the reduction in total contaminant mass is another tool that may be used to
evaluate progress toward achieving contamination reduction objectives.  The production of
daughter products from parent contaminants is generally primary evidence of biotransformation
processes, particularly at chlorinated solvent sites. However, certain  compounds that can be
daughter products (e.g., trichloroethene) may be or have been present as  a substantial portion of
the source materials at many sites.  Thus, source characterization information is needed to ensure
their appearance is not incorrectly attributed to natural attenuation processes.

Qualitatively, simple comparisons of the stoichiometric ratios of parent contaminants to daughter
products in different parts of the plume provide important evidence of the effects of degradation
processes with migration distance and should be considered during evaluations of contaminant
mass reduction.  However, data for quantitative assessments of contaminant mass reduction
would consist of the concentrations of all contaminants of concern in samples obtained
throughout the contaminated aquifer. As with all methods used to evaluate subsurface
remediation performance, the most appropriate spacing of monitoring points for evaluation of
mass reduction depends on the variability in contaminant distribution. Uncertainties in these


                                             32

-------
Contaminant   Monitoring Well
Concentration      clusters
   in |jg/l
     Ground-Water Flow
          Time 7
         Active Irrigation Well   o
         Ground-Water Flow
              Time 2
                Active Irrigation Well   

                    New Monitoring Wells
                             A     A
                 Ground-Water Flow
                     Time 3
      Well Cluster
         W3
          Well Cluster
             W3
(Offset)*
Irrigation
  Well
Well Cluster
   W3
                                                                          ABC
(Offset)*
Irrigation
  Well
                                                                 Source
                                                                  Area  y
                                Offset from plane of cross section
                        o
                        c
                        QJ
                        U
                        c
                       3
10-
                                                         W2-B
                                                         W1-B
  W3-B


  W4-B
  W5-B
  Figure 9.  Potential effects of changes in ground-water flow direction on temporal trends
             in contaminant concentrations.
                                                33

-------
methods normally result from use of sparse data sets relative to in-situ variability.  Application of
transect-based approaches with monitoring at multiple depths may be needed to provide adequate
data for evaluation of mass reduction.  Given a sufficient density of monitoring points, evaluation
of contaminant mass reduction may provide a more consistent indication of MNA effectiveness
than analysis of temporal contaminant concentrations trends in individual monitoring wells.

Cohen etal (1994) describe methods for estimating total contaminant mass in the aquifer if no
NAPL or other source material is present. Although the authors describe this methodology for
evaluation of pump-and-treat performance, the techniques would be identical for estimating total
contaminant mass within the plume during MNA performance evaluations. Depending on site
conditions and contaminant properties, periodic sampling of both ground water and aquifer
materials may be needed to better determine total contaminant (dissolved and sorbed phase) mass
trends.

Data from transects of clustered sampling points, placed perpendicular to the longitudinal axis of
the plume, may also be used to estimate contaminant flux within the plume, and, therefore, mass
loss along the plume axis  (Weaver etal., 1996).  Such transects provide three-dimensional cross
sections of contaminant distribution. If used with hydrogeologic data, contaminant mass flux
through each cross section can be  calculated.  Contaminant mass loss and, consequently,
attenuation rates may be estimated using the differences in the fluxes between transects.
Sampling density and frequency for transects generally should be based on the factors identified
in Sections 2.5.2 Monitoring Locations and 2.5.4 Monitoring Frequency.

As noted in other sections, evaluations of cross-media transfers of contaminants are important
elements of performance monitoring for MNA.  Subsurface contaminant mass within one
medium may be reduced due to contaminant  migration into other media, so the potential for such
loss should be evaluated in order to provide a comprehensive assessment of contaminant mass
reduction (e.g., evaluate the potential for cross-media transfer into soil gas).  Contaminant mass in
all physical phases  (e.g., NAPL, aqueous phase, vapor phase, sorbed)  should be assessed within
each medium to estimate the actual contaminant mass reduction.

2.6.1.3 Comparisons of Observed Contaminant Distributions with Predictions and Required
       Milestones

Contaminant distribution data obtained during performance monitoring may periodically be
compared with previous predictions or specified  milestones (e.g., at least 50% contaminant
reduction at all monitoring points within a specified number of years) to ensure that suitable
progress toward contaminant reduction goals is achieved.  Quantitative models may be used to
help set or evaluate the milestones. For example, models may be used to estimate the range of
time frames for achieving remedial objectives (e.g., contaminant concentrations below Maximum
Contaminant Levels (MCLs) in all wells within 20 years) and set intermediate milestones (e.g.,
concentrations in all wells at the current plume periphery below MCLs in five years). Although
these comparisons  may be as simple as visual comparisons of contoured maps or comparisons of
predicted and observed total contaminant mass estimates, the use of statistical methods provides a
more objective basis for evaluation. Types of data comparisons for such evaluations are discussed
in Chapter 3.

The goal of these comparisons is to identify any  significant deviations from predictions or other
progress requirements that may signify flaws in the conceptualization of site conditions.  For
instance, a gradual  long-term trend of increasing water levels may cause increased  dissolution of


                                             34

-------
inadequately removed or contained vadose zone contaminants, causing increased ground-water
contaminant concentrations in the source area and, potentially, resulting in plume expansion.
Comparison of the observed contaminant concentrations with predicted concentrations or
milestones would highlight the change and trigger re-evaluation of the conceptual site model. If
the potential impact on the remedy is negligible, revision of the conceptual site model may be the
only outcome of such evaluations. However, if the potential impact is significant, implementation
of alternative remedies may be triggered or, at a minimum, additional characterization to better
define the factors limiting  contaminant reduction and determine  methods for mitigating the
magnitude of their impact  on the remedy would be warranted.

Monitoring parameters for these periodic  comparisons may include all contaminants of concern
but, at some sites, may be limited to the most mobile, toxic, or recalcitrant compounds as
indicators of the behavior of other site contaminants.  The most appropriate frequency for
performing this type  of evaluation depends on the expected rate of change in contaminant
concentrations and the ability of the monitoring network to reliably measure significant changes
(Section 2.5.4.).  It is likely that performance of such evaluations at a frequency coincident with
major, comprehensive site  reviews (e.g., every few years) may be appropriate for most sites.

2.6.1.4  Comparison of Field-Scale Attenuation Rates

Information obtained during periodic performance monitoring may also be used to estimate  field-
scale attenuation rates and compare these estimates with the range of rates determined to be
necessary to meet remediation objectives within specified time frames.  Methods that have been
used to estimate attenuation rates and the biodegradation component of the attenuation rate are
described in U.S. EPA (1998a), Wiedemeier etal  (1999), and U.S. EPA (2002b).  However,
commonly used calculation methods (e.g.,  Buscheck and Alcantar, 1995) often suffer from
significant uncertainty regarding the applicability of assumptions required in the calculations as
well as the limited data available from many sites  (McNab and Dooher, 1998). Careful evaluation
of site conditions and the assumptions inherent in the method of analysis is required to produce
useful results.  Due to the  increased uncertainties inherent in the estimation of attenuation rates,
it is anticipated that periodic evaluation of such rates during performance monitoring often may
have few advantages  over the direct comparisons of contaminant distributions described above.

2.6.2   #2 -Detect Changes in Environmental Conditions that May Reduce the Efficacy of
       Any of the Natural Attenuation Processes

Changes in hydrogeological and geochemical conditions may affect microbiological populations,
the transformation processes that result in contaminant destruction, and rates or directions of
ground-water flow and contaminant migration. Reductions in the effective rates of contaminant
transformation, sustained increases in hydraulic gradients, or, in some cases, exceedence  of
sorptive capacity may lead to plume expansion. Sustained changes in ground-water flow
directions may result in risk to receptors that were not originally identified at the time of remedy
selection.  The monitoring network should be designed to detect changes in such parameters to
allow determination of their effects on remedy performance.

It is particularly important to monitor and evaluate the effects of other remedial actions that may
be taking place on or near  the site.  Many remedial actions (e.g., capping of the site, pump and
treat,  air sparging,  chemical oxidation) would be expected to change hydrogeologic and/or
geochemical conditions,  potentially adversely affecting the MNA remedy.  For instance, a site cap
could reduce the flux of oxygen to a shallow aquifer, potentially slowing degradation of petroleum

                                            35

-------
hydrocarbons, or a pump and treat remedy could alter the flow of ground water at the site, and
potentially change the geochemical environment.  In addition, the potential effect of planned
remedial actions should be taken into account when predicting the efficacy of MNA during the
remedy selection phase, and designing and implementing the performance monitoring system.

Routine monitoring parameters for these determinations are primarily geochemical indicators of
the contaminant transformation reactions that may occur within the aquifer, and ground-water
elevations in wells and piezometers that are used to estimate hydraulic gradients in each
hydrogeologic zone.

2.6.2.1 Geochemical Parameters

The values and patterns of geochemical parameters other than ground-water contaminants
provide evidence of the potential for continued biotransformation. The usefulness  of
geochemistry for performance monitoring comes from these factors:

       Degradation of some contaminants (e.g., some highly chlorinated ethenes, highly
       chlorinated methanes, and highly chlorinated chlorobenzenes)  appears to occur only
       under fairly specific geochemical conditions, and

       Degradation of contaminants causes specific geochemical changes that may be, at least
       qualitatively, correlated with microbial activity and contaminant degradation.

Though rates may differ, some contaminants  (e.g., BTEX) can be degraded under a variety of
geochemical conditions, but others (e.g., tetrachloroethene (PCE))  may require a more narrow
range of conditions (Wiedemeier et al, 1999). The controls  on biotransformation processes at
chlorinated solvent sites are very complex and not as well understood as at sites contaminated
solely by more readily degradable hydrocarbons such as the  aromatic components of fuels (i.e.,
BTEX). Appreciable amounts of a biodegradable carbon compound can produce appropriate
geochemical conditions for reductive dechlorination (Wiedemeier  et al., 1999; Leahy and Shreve,
2000). Therefore, depletion of the carbon source may limit  the extent  of biotransformation of
chlorinated ethenes if reductive dechlorination is a significant biodegradation process.
Biotransformation of some contaminants, such as MTBE and other ethers that may be present in
fuel contamination, is not well understood and is the subject of continuing research (National
Research Council,  2000).

Most of the geochemical signatures seen at sites of chlorinated solvent  contamination are
reflections of biological processes utilizing various  carbon sources and  often are not directly
caused by biotransformation  of the chlorinated solvent contaminants. However, these data have
value  as continuing indicators of appropriate  geochemical environments for biotransformations of
contaminants in many situations.  The analyses provide evidence regarding the continued
effectiveness of MNA and the adequacy of the conceptual site model.  If biotransformation is a
major natural attenuation mechanism, then the geochemistry generally should be monitored to
determine that biotransformation continues.  If a strong correlation exists between changes in
geochemical indicators, decreasing contaminant levels, and the products of known  biological
transformation processes in a particular portion of the plume, this may be considered supporting
evidence for continued microbial destruction  of contaminants. However, in areas of decreasing
contaminant concentration that do not bear a strong geochemical signature, other processes
typically  should be considered. The relationship between geochemical patterns and degradation
of organic carbon other than the contaminants of concern should also be considered.


                                            36

-------
Geochemical parameters generally should be measured throughout the plume in order to establish
a correlation between reductions in contaminant concentrations and microbial activity (U.S.
EPA,  1998a; Wiedemeier et al, 1999). In addition, geochemical samples generally should be
obtained upgradient and sidegradient of the plume (Figure 4) to determine the range and
variation in ambient conditions, in order to differentiate changes in geochemistry related to
microbial metabolism of contaminants from unrelated processes. The specific parameters useful
at a given site include:

       Parameters displaying correlations with contaminant distribution that are indicative of
       dominant biotransformation processes,

       Parameters showing conditions that are necessary for contaminant degradation, and

       Parameters whose appearance may indicate changes detrimental to continued
       biotransformation.

Discussion of these parameters and the interpretation of geochemical data is found in U.S.  EPA
(1998a) and Wiedemeier et al. (1999).

At sites where the total available supply of electron acceptors utilized in contaminant
biotransformation is predominantly composed of solid phase material (e.g., bioavailable iron
species), periodic monitoring of bioavailable electron acceptors in soil samples may provide
information concerning their rates of depletion (Ruling et al., 2002) and the potential for
renewed plume migration. This type of monitoring may be useful where a large mass of
biodegradable contaminant relative to the total electron acceptor supply is present. Where
warranted, it is anticipated that this type of assessment would be performed on a relatively
infrequent basis determined by the possible time frames over which significant changes are
expected to occur.

2.6.2.2  Hydrogeologic Parameters

Long-term changes in ground-water flow directions due to sustained changes in recharge or
discharge locations or rates may result in risks to receptors that were not originally identified at
the time of remedy selection. In addition, increases in ground-water flow rates may result in
expansion of the plume. Such changes may result from many factors including  land use changes
(e.g., irrigation of adjacent areas), installation of new pumping wells or  increase in pumping rates
of existing wells, and installation of drainage systems, as well as climatic patterns.

Monitoring of ground-water and surface-water elevations should be performed at a frequency
sufficient to determine significant variations in hydraulic gradients. The frequency would usually
be no less than the monitoring frequency for the chemical parameters.  Other hydrologic data
such as rates of local ground-water pumping, irrigation, and precipitation may also be needed at
some  sites.  As previously noted, the evaluations generally should be three-dimensional in nature,
monitoring horizontal and vertical components of hydraulic gradients.  The monitoring network
should be dense enough to accurately estimate spatial variations in gradients. The number of
monitoring points needed at any given site depends in large measure on the sources of aquifer
recharge and discharge in the vicinity of the site and the degree of geologic heterogeneity.

2.6.3   #3- Identify Any Potentially Toxic and/or Mobile Transformation Products

Contaminant transformation processes may result in the formation of new chemicals that may be
more  toxic or more mobile than their "parent" compounds. Transformation products may appear

                                             37

-------
sequentially along the direction of ground-water flow and all possible products may not be present
throughout the plume.  Potentially toxic daughter products should be evaluated based on
information regarding contaminant transformation pathways and monitoring data.  All such
products typically should be included as contaminants of concern in the monitoring program and
monitored at the same frequency as other contaminants of concern.  Mobile but nontoxic
transformation products may be monitored as an aid to determining flowpaths and demonstrate
continued biotransformation processes, as discussed in Sections 2.6.1, Demonstrate that Natural
Attenuation is Occurring According to Expectations, and 2.6.2, Detect Changes in Environmental
Conditions that May Reduce the Efficacy of Any of the Natural Attenuation Processes.

Minerals in the aquifer matrix have specific chemical compositions and crystalline structure and
are in dynamic equilibrium with surrounding pore water. Large changes in oxidation-reduction
potential caused by biological processes within the contaminant plume may result in
transformation of some naturally occurring mineral species into more soluble and mobile forms
that pose risks to human health or the environment (Brady etal.,1999, U.S. EPA,  1998a,
Wiedemeier etal., 1999). Mobilization of arsenic and manganese species has been observed at
some sites. Where such mobilization is possible, these species generally should be monitored as
potential contaminants,  and also for interpretation of their possible use as electron acceptors.
Metals that are possible contaminants of concern due to human health or surface-water quality
criteria should be included in the ground-water monitoring program at the  same frequency as
other contaminants.  Removal of these constituents from the monitoring program may be
warranted if data demonstrate lack of appropriate mineral species or that such mobilization does
not occur under site conditions.

2.6.4   #4 - Verify that the Plume is Not Expanding Downgradient, Laterally, or Vertically

Plume  expansion may result in a plume extending to areas that were previously uncontaminated,
migrating beyond established compliance boundaries, degrading additional  ground-water
resources, and potentially increasing risk to receptors. Previous site characterization and
performance monitoring data obtained using  a three-dimensional monitoring network will likely
be needed to assess hydrogeological changes, temporal variability in the lateral and vertical extent
of the plume, and to determine the effective boundary for demonstrating compliance with the
nonexpansion  objective. Assessment of plume expansion is complicated because ground-water
flow systems, like all natural systems, are temporally dynamic.  For example, precipitation may
vary seasonally and/or over longer time frames. This may result in changes in water table
elevations and possible changes in hydraulic gradients and contaminant inputs to a plume.
Anthropogenic effects (e.g., variations in ground-water extraction rates or loss of water from
distribution systems) also can be a source of changes in subsurface conditions. Such changes may
result in three-dimensional changes in plume geometry, causing the plume  boundary to be in
continual flux.  This constant variation in plume geometry means that, in some cases, it may
require several years of monitoring to definitively determine if a plume is not expanding. In such
cases, particularly where receptors are immediately susceptible to impact, implementation of
alternative remedies may be more appropriate than waiting to verify a lack of plume expansion.

Plume  expansion may be caused by many geochemical and biologic processes as well as variability
in site  hydrology.  Examples of these processes include:

      Increases in contaminant inputs to the plume that exceed the capacity of transformation
       and hydrologic processes to attenuate  the concentrations within the current plume
       boundaries, and

                                             38

-------
       Changes in oxidation-reduction potential due to factors such as the depletion of electron
       donors or acceptors that reduce the effective rates of transformation processes.  For
       example,  depletion of a degradable carbon source prior to complete biotransformation of
       tetrachloroethene to lesser chlorinated products may result in additional tetrachloroethene
       migration.

A three-dimensional performance monitoring network (Figures 4, 5 and 6) will often be
necessary to meet this objective.  Monitoring of points throughout the plume, including locations
in or near existing or suspected source areas and in the zones of highest contaminant
concentrations, generally will be needed to evaluate changes that may lead to plume expansion.
Monitoring points of the most immediate concern will often be those points located near the
horizontal and vertical plume boundaries and any other compliance boundaries specified in the
remedy decision.  Trends of increasing contaminant concentrations in these wells will often be
direct evidence of plume expansion.  In addition, evaluations of temporal trends in individual
monitoring points and comparisons with trends observed in different areas of the site may aid in
confirming the roles of the dominant processes (e.g., advection, dispersion, and  transformation)
controlling  contaminant migration. The appropriate number and locations of monitoring points
will be dependent on such factors as the size of the plume, ground-water velocity, proximity to
receptors, and presence of preferential pathways for contaminant migration.

In practice, wells located at the plume and other compliance boundaries, and between the
boundaries and receptors, may often be used to detect contaminant increases and trigger the
implementation of contingency or alternative remedies to prevent impact to receptors.  Such wells
would be located sufficiently upgradient of receptors to allow adequate time for implementation
of the contingency remedy and demonstration of its effectiveness given  the ground-water velocity
within the most transmissive contaminated zones.  Monitoring at the receptor point generally
should also be included in all settings. For example, monitoring of public and private water
supply wells that would be at risk if the plume expanded should be  incorporated into the
performance monitoring program, as warranted. Wells chosen for sampling should be based on
evaluations of the capture zone for each well.  In similar fashion, monitoring at locations of
possible impact to ecological or other receptors should also be included.

At some sites, the geochemical fingerprint of ground water can be established and used to trace
water downgradient to distinguish ground water that has never been contaminated from ground
water that was previously contaminated (National  Research Council 2000, Weidemeier and Haas
2002). Such information may be used to site wells near the current plume boundary in zones
where contaminant migration would be expected if plume expansion occurred.  Depletion  of
electron acceptors, and presence of metabolic by-products and nonhazardous daughter products
may be used as indicators of appropriate monitoring locations.  The most useful parameters for
sites with hydrocarbon contamination may include nitrate, sulfate, iron,  methane, and dissolved
oxygen. The most useful tracers in plumes of chlorinated solvent compounds are often their
reduced transformation products, particularly ethane or ethene, but also include the same
parameters as for petroleum hydrocarbon plumes.  The most appropriate parameters for
determining locations for monitoring wells downgradient of a contaminant plume depend  on site-
specific correlations of contaminants and geochemical indicators.

2.6.5   #5 - Verify No Unacceptable Impacts to Downgradient Receptors

Impacts to receptors may result from plume migration to wells used for drinking and other
domestic purposes, irrigation or industrial purposes; contaminant migration to  indoor air;  or

                                            39

-------
discharge of ground water to wetlands or surface-water bodies. Prevention of unacceptable
impacts to receptors includes continuing verification of plume stability and the reliability of
institutional controls.  Adequate demonstration of the lack of unacceptable impact to receptors
will often require monitoring of wells located between the plume boundaries and the receptors, in
the transmissive zones where contaminants may migrate if plume expansion occurs.  In the case
of potential impacts to receptors associated with water production wells, it would generally
include periodic monitoring of the production wells. The potential for impacts to both human and
ecological receptors should be evaluated, and the performance monitoring system designed to
verify that no unacceptable impacts occur.

Performance monitoring with respect to this objective also includes monitoring of the cross-media
transfer of contaminants. Cross-media transfers of concern include transfers between ground
water and soil gas, air, sediments, and surface water.  For example, in some situations, particularly
where volatile contaminants are found in the vicinity of existing buildings or other structures, the
potential may exist for contaminant migration through the vadose zone to indoor air. In such
situations, periodic monitoring of soil gas and indoor air may be needed to determine that no
impacts occur. Additional guidance regarding techniques for the monitoring of soil gas may be
found in U.S. EPA (1988, 1993b).  Fate and migration of volatile contaminants from the
subsurface to indoor air is a rapidly evolving area of research and guidance/policy development.
The reader is referred to U.S. EPA (2002d) for further information.

Another major cross-media transfer process of concern is movement of contaminants between
ground-water and surface water/sediments. The ground-water/surface-water transition zone,
where ground water and surface water mix in the saturated sediment beneath and beside surface
water, may exert a major influence on input of contaminants to surface water because of the
enhanced chemical and biological activity often found in these zones. The hydrologic and
geochemical conditions in areas where ground water interacts with surface water often differ
markedly from those in the main body of the plume and may require more intensive monitoring
to determine the effect on remedial goals (Winter, 2000).

The locations and characteristics of contaminated ground-water discharges generally should be
determined. Areas of discharge and interaction may vary rapidly both temporally and spatially.
Plume discharge to the surface-water body  or wetlands may not be at the immediate shoreline or
channel edge, and the discharge may vary spatially through time.  In addition, contaminant
plumes may also migrate beneath streams and drainage features in some settings. Monitoring
generally should occur in the interface between ground water and surface water as well as in the
surface-water column (U.S. EPA, 1991b).

Tools to characterize the hydraulic relationships between ground water and surface water (Figure
10) include piezometers, pore water sampling devices, devices for the in situ measurement of
ground-water velocity, and certain geophysical techniques. Such tools may be used to define
ground-water flow from the plume into the surface-water bodies and to aid in siting monitoring
points for determining the impact of the discharging water on the sediments and surface-water
quality.  It should be noted that mobilized metals may be a primary concern during these
investigations due to the relatively low thresholds for unacceptable  ecological impact.  Multilevel
monitoring is generally needed to identify contaminant discharge locations.  Once the hydraulic
relationships are characterized, water and sediment sampling locations may be specified.

2.6.6   #6 - Detect New Releases of Contaminants

Increases in contaminant concentrations or detection of new contaminants at monitoring points
located within and immediately downgradient of source areas may be indications of new releases.


                                             40

-------
Releases may result from such causes as the failure of source control measures, increased water
infiltration or water table rise into contaminated portions of the vadose zone, or new releases to
the environment. In many cases, it may be difficult to determine whether increased contaminant
concentrations are due to new sources or other causes.  Increases in contaminant mass may be
sufficient to result in expansion of the plume and/or extension of time frames required to attain
contaminant reduction objectives, warranting implementation of additional source control/
removal measures, other remedy modifications, or alternative remedies.

2.6.7  #7 - Demonstrate the Efficacy of Institutional Controls

Institutional controls may be put in place to prevent access to or use of contaminated ground
water. These controls are an integral and necessary  part of an MNA remedy because they are
relied upon to ensure that the remedy is protective to human health and the environment. For
this reason they should be monitored with the same degree of thoroughness as other components
of the remedy.  Institutional controls (U.S. EPA, 2000d) include:

       Governmental controls (e.g., zoning restrictions, ordinances, statutes, building
       permits),

       Proprietary  controls (e.g., easements, covenants),

       Enforcement and permit tools (e.g., administrative orders or consent decrees with
       institutional control), and

       Informational devices (e.g., state registries, deed notices, hazard advisories).

At a minimum, the  following questions should be answered during development of the
monitoring plan to allow an initial assessment of the potential efficacy of institutional controls:

    1)  Are institutional  controls  in place?

    2)  What are the criteria for determining that institutional controls are effective and
       operating as intended?

    3)  Who is responsible for monitoring and reporting on the integrity and effectiveness of
       institutional controls for the site?

    4)  Are institutional  controls  being monitored regularly and at  an appropriate frequency?

    5)  Are institutional  control monitoring results  included in either a performance monitoring
       report or a separate report?  Are these reports completed at  an appropriate frequency and
       forwarded to the appropriate regulator?

    6)  What procedures are in place to report breaches or failures of the institutional controls to
       the appropriate U.S. EPA and/or state regulator, local or tribal government, and the
       designated party or entity responsible for reporting?

    7)  Is the property likely to be transferred in the near future, and if so, who will be
       responsible for the future  monitoring and reporting and enforcement of the institutional
       controls?
                                              41

-------
                                             Lakebed
                              Sediment      Piezometer
                              Sampling
                              Locations
Monitoring
  Wells
               Pore Water
                Sampler
                                                                         Contaminant
                                                                            Plume
                                                Ground-Water Flow
        Figure 10.    Conceptual monitoring network for verifying lack of impact to ground-
                     water discharge.
    8)  Are there provisions in place to notify regulators of any impending property transfer such
       as in a consent decree or order?

The MNA performance monitoring plan should include a description of activities to initiate and
periodically monitor institutional controls, or refer to a separate plan for monitoring institutional
controls. The assumptions and uncertainties associated with the site-specific institutional controls
should be carefully considered in designing the monitoring plan and determining effectiveness.
Compliance monitoring may include field inspection of affected areas, particularly if affected
properties have been sold or land use patterns have changed, as well as determinations that the
specified controls have been enacted. Each control should be periodically investigated to
determine whether the control continues to be implemented as specified.

2.6.8   #8 - Verify Attainment of Remediation Objectives

One of the fundamental cleanup objectives for most sites is the reduction of contaminant
concentrations in subsurface media (i.e., soil and/or ground water) to specified levels. The
remedial action objective of attaining permitted standards throughout the plume should be
demonstrated before monitoring is  terminated to ensure that the required standards are actually
achieved. The techniques for determining compliance with RAOs under a natural attenuation
remedy are similar to those used for determining compliance following application of other
remediation methods. The demonstration of the attainment of cleanup objectives should include
sufficient verification monitoring (e.g., three to  five years) once the standards are met to evaluate
the effects of natural variations in site  conditions, based on objective statistical analyses of the
data.  Statistical methods useful in these evaluations include analyses of temporal trends in
contaminant  concentrations and  comparisons with the specified concentration standard.
Guidance regarding verification of compliance with cleanup objectives is provided in Cohen et al.
(1994) and U.S. EPA (1992a).
                                            42

-------
2.7    Monitoring Plan Contents

2.7.1   Introduction

The following material has been prepared to provide guidance regarding appropriate formats
and components of performance monitoring plans. Although there is no standard format that
is universally accepted, most monitoring plans make extensive use of maps, cross sections, and
figures to convey monitoring requirements and approaches. Ideally, monitoring data would be
acquired, transmitted, evaluated and presented in an electronic form for ease of transmission
and analysis. The format and content of a monitoring plan will vary according to site-specific
needs, but most plans would contain or reference the elements listed in Table 4 and briefly
discussed in the  following sections.

2.7.2   Background and Site Description

Background information on the site, including a discussion of the site setting, history,
characteristics, remedial goals, past and present remedial actions, and any institutional
controls, typically would be provided in abbreviated fashion or incorporated by reference to
existing documents.  Essential aspects (e.g., hydrogeologic setting, contaminant distribution,
remedial goals and actions) generally should be addressed at sufficient length that the
technical reviewer can judge the adequacy of the monitoring efforts. The geologic and
hydrogeologic controls on ground-water flow at the regional scale and site scale should be
discussed in detail with references to the supporting data or provided by reference to specific
portions of existing documents.

2.7.3   Conceptual Site Model for Natural Attenuation

Essential elements of the conceptual site model for natural attenuation include:

       Descriptions and locations of possible  receptors,

       Geologic  and hydrologic controls on ground-water flow,

       Contaminant distribution and behavior, and

       Relationship of geochemistry and anthropogenic factors to attenuation processes

The monitoring  plan typically would provide a detailed discussion of how the conceptual site
model and the supporting data and analyses were used to  design the network.

2.7.4   Objectives and Decision Points

The monitoring  plan typically should specify the RAOs, monitoring objectives, and
performance criteria used to guide the performance monitoring plan development and provide
decision points for continuing, modifying, or terminating performance monitoring.  The
decisions to be made should be thoroughly discussed in relation to site objectives, specific
decision points, methods of determining attainment of goals, and how site activities would
proceed based on decisions made.  The criteria upon which each decision is to  be based should
be explicitly stated.

                                             43

-------
Table 4.     Elements of a Performance Monitoring Plan
  Background and Site Description

  The following information typically would be discussed or incorporated by reference:

  Q Site setting, history, and characteristics
  Q Remedial goals
  Q Past and present remedial actions and institutional controls

  Conceptual Site Model for Natural Attenuation

  The following information typically would be discussed or incorporated by reference:

  Q Descriptions and locations of potential receptors (text and maps)
  Q Geologic and hydrologic controls on ground-water flow (text, maps, and cross sections)
  Q Contaminant sources, distribution, migration, and fate (text, maps, and cross sections)
  Q Relationship of geochemistry to attenuation processes

  Objectives and Decision Points

  Q Remedial Action Objectives (RAOs)
  Q Monitoring Objectives
  Q Performance  Criteria
  Q Decisions to be made based on monitoring data (e.g., monitoring network changes, implementation of
     contingency/alternative remedy, implementation of verification monitoring, terminate performance
     monitoring) and the criteria for making each decision

  Monitoring Network and Schedule

  Q Detailed discussion of relationship of conceptual model, supporting data, and analyses to design of the monitoring
     network
  Q List and map  of monitoring locations
  Q Description and construction details for each monitoring  point
  Q Monitoring schedule specifying monitoring parameters, analytical methods, sampling frequency
  Q Maps, cross sections, and other visual aids to show where, when, and by what methodologies
     samples are to be taken and analyzed
  Q Statistical methods and test designs to be used for data interpretation

  Monitoring of Institutional Controls (ICs)

  Q Descriptions of ICs and the procedures for their implementation
  Q Procedures for verifying establishment of ICs
  Q Frequency for monitoring ICs
  Q Procedures for routine monitoring of the effectiveness of  ICs and parties responsible for the monitoring

  Evaluations of Remedy Effectiveness

  Q Description of evaluations to be performed to demonstrate effectiveness of natural attenuation
     with respect to site-specific remedial action objectives
  Q Determination of temporal and spatial trends in contaminant concentrations or mass
  Q Comparisons of contaminant concentrations with previous predictions or milestones
  Q Comparisons of contaminant concentrations in areas outside the previous plume boundaries
     or other compliance boundaries with specified action  levels
  Q Frequency (e.g., quarterly, annually) on which each evaluation is to be conducted and the data that are to be used
  Q Data presentations to be used (e.g., tables, maps, cross sections, and other figures)

  J Methodology or plan for verifying attainment of remedial objectives and termination of performance
     monitoring

  J Sampling and Analysis Plan

  J Quality Assurance Project Plan
                                                      44

-------
2.7.5  Monitoring Network and Schedule

The plan should include information on the monitoring network including a list and map of
monitoring locations; a description of the construction of each monitoring point; and a
monitoring schedule specifying monitoring parameters, analytical methods, sampling frequency,
and statistics to be used for data interpretation.  Information regarding Data Quality Objectives
(DQO's) should be provided or incorporated by reference. It is particularly important that maps,
cross-sections, and other visual aids be extensively used so that the reviewer can determine
exactly where, when, and by what methodologies samples are to  be taken and analyzed.  In
situations where different hydrogeologic units are monitored, separate maps with wells/sampling
locations for each major lithologic unit may provide additional clarity in visualization of the
network design.

2.7.6   Monitoring of Institutional Controls

The monitoring plan should include detailed descriptions of the  institutional controls and their
implementation procedures. Procedures for verifying that institutional controls are put in place,
for monitoring the effectiveness of the controls, and for reporting monitoring results to
appropriate government entities should be described.  The frequency for monitoring and
reporting effectiveness should be specified. The parties responsible for implementing, verifying,
and monitoring effectiveness of the institutional controls should be listed, along with complete
contact  information.

2.7.7  Evaluations of Remedy Effectiveness

The monitoring plan should specify the evaluations that will be performed to demonstrate the
effectiveness of natural attenuation with respect to the site-specific remedial action objectives.
The frequency (e.g., quarterly, annually)  on which each evaluation is to be conducted and the data
that are to be used would also be specified. The evaluations should be designed to objectively
determine performance in relation to specific performance criteria established in remedy decision
documents.  Pertinent evaluations include:

      Determination of temporal and spatial trends in contaminant concentrations or mass,

      Comparisons of observed contaminant concentrations with previous predictions or
       established milestones, and

      Comparisons of contaminant concentrations in areas outside of previous plume
       boundaries with specified action levels (e.g., drinking-water standards)

In addition,  the plan should specify the data presentation methods (e.g., tables, maps, cross
sections, and other figures). Submission of data in  electronic formats should be considered for
ease of manipulation and independent analysis.

2.7.8  Plan for Verifying Attainment ofRAOs

The performance monitoring plan should include discussion of the methodology or plan for
verifying attainment of all remedial action objectives and initiating termination of performance
monitoring.  In general, the plan developed at the initiation  of performance monitoring should be
flexible  and  allow for future modifications based on observations of actual remedy performance.
The methodology should include:

                                             45

-------
       Proposed time period for verifying attainment of objectives, considering any observed
       seasonal or other temporal trends,

       Description of monitoring locations, parameters, and frequency, including discussion of
       any possible changes that may be warranted, and

       Specification of statistical methods of data analysis to be used in demonstrating
       attainment.

It is recommended that the plan for the verification of RAOs be re-evaluated prior to its
implementation to insure that the verification methodology initially proposed is still valid
considering the additional data obtained and any changes in the conceptual site model that
occurred during the performance monitoring period.

2.7.9   Sampling and Analysis Plan

The sampling and analysis plan typically should include a description of:

       Sample collection methods,

       Sample preservation and handling,

       Chain-of-custody procedures,

       Analytical procedures, and

       Field and laboratory quality assurance/quality control.

Additional guidance regarding the development and specific elements of sampling and analysis
plans is found in a variety of sources  including U.S. EPA (1986a, 1986b, 1992c, 1993a, 2002a).

2.7.10  Quality Assurance Project Plan

The quality assurance project plan (QAPP) documents the manner in which quality assurance
and quality control activities will be implemented throughout the performance monitoring
program.  The QAPP is composed of the following elements:

       Description of project tasks, data quality objectives, and management,

       Description of data acquisition and management,

       Description of assessments, responses, and oversight, and

       Description of data validation, verification, and usability.

Detailed guidance regarding preparation of quality assurance project plans is available in U.S.
EPA (1998c).
                                             46

-------
                                       Chapter 3
                ANALYSIS OF PERFORMANCE MONITORING DATA
3.1    Introduction

Data interpretation involves these basic steps of the data quality assessment (DQA) process :

      Placing the data in context of time, location, sampling and analytical methods,

      Preliminary assessment of the data with basic statistical measures (e.g., means and ranges),
       and graphs, charts, maps, time-series plots, and cross-sections,

      Applying appropriate statistical tests to detect changes, trends, and assess attainment of
       goals, and

      Making decisions based on the data.

The conceptual site model for MNA and the monitoring program  are continually refined as new
data are gathered and interpreted. Each new round of data should help develop a better
understanding of the site and site processes, including plume shape, location, stability, and
dynamics; attenuation rates; geochemical regimes; and site hydrogeology.

The criteria used for site-related decision-making are usually based on either presence of specific
contaminant concentrations (e.g., MCLs at a specified group of wells), or spatial and temporal
trends in concentrations (e.g., a decreasing trend in contaminant concentrations that indicates
progress toward contaminant reduction objectives). Other site-related data (e.g.,  geochemical and
ground-water flow data) are important for corroborating contaminant migration, fate, and
attenuation processes.  These data may be used to provide evidence for contaminant degradation
and the continuation of appropriate conditions for attenuation at  acceptable rates.

3.2    The DQA Process

The Data Quality Assessment (DQA) process is an iterative procedure used to evaluate, analyze,
and interpret data. Guidance for the DQA process is found in U.S. EPA (2000c). The assessment
methods used at each site  should be specified in the performance monitoring plan.

Briefly, the DQA process for performance monitoring of an MNA site is as follows:

      Review the Data Quality Objectives (DQOs) and the sampling design to ensure that the
       design is suited to  achieving the DQOs. The data derived  from the monitoring should
       have the proper characteristics for their intended use; that is, data should be of the correct
       type, quality, and quantity in order to be useful for making site-related decisions.

      Conduct a preliminary review of the data, such as calculating means and other basic
       statistics, graphing the data,  and mapping the data to identify obvious patterns,


                                            47

-------
       relationships, and anomalies, such as possible "outliers". Simple scatter plots, for example,
       may show relationships between two variables that are not evident using purely numerical
       methods. Mapping the data in plan view and cross-section is helpful for understanding
       patterns of contaminant distribution and migration, and for understanding relationships
       between geochemical zones and contaminant attenuation.

       Verify the selection of the appropriate statistical test(s) (if needed) for analyzing the data,
       considering the types of decisions that are to be made {e.g., attainment of contaminant
       reduction objectives) and the sampling design. Identify the assumptions that underlie the
       chosen statistical test (e.g., normally distributed data). The appropriate statistical tests are
       identified through the application of the DQO process during monitoring system design,
       because the design is chosen considering the requirements of the particular tests to be
       used.

       Verify that  the data meet the assumptions of the statistical test, and whether any
       departures  from the assumptions are acceptable. Given the nature of many sites, it is
       possible that data will not always meet the assumptions of either parametric or non-
       parametric statistical tests.  Ground-water monitoring data typically exhibit extremely
       high variability and skewed distributions. Although certain assumptions of particular tests
       may be in question, statistical tests are often still a worthwhile guidance tool for decision-
       making.  Statistical analyses should be considered as a way to bring a measure of
       objectivity  to the decision-making framework.

       As the design of the sampling and analysis plan can minimize or eliminate violations of
       statistical assumptions, it is recommended that statisticians be included in the
       performance monitoring planning phase. The data analyst, however, bears the ultimate
       responsibility of determining if the assumptions of a particular statistical test are violated
       and how these violations may affect site decisions.

       Interpret and draw conclusions from the data analysis by performing the appropriate
       statistical tests. The sampling design should be evaluated to determine if any changes are
       needed to enhance the usability of the data.

       Interpret and draw conclusions from the data in the context of the conceptual site model
       for MNA and site remediation goals.

In addition to the steps presented here for assuring the quality of the data, it is recommended that
a process be implemented to assure the quality of the conceptual site model for MNA. With each
new round of sampling, the conceptual site model for MNA should be re-evaluated and, if
necessary, modified.  In particular, the conceptual site model for MNA should be re-evaluated
from the standpoint of its power to explain observed plume shape, location, stability, and
dynamics; ground-water flow patterns; and geochemical regimes.

At many sites, source control and interim remedial activities will occur between the original site
characterization  process and implementation of performance monitoring. Because these activities
are expected to change the site's characteristics and behavior, it is possible that performance
monitoring will require development of a conceptual site model for MNA that is substantially
different from the MNA conceptual site model that was based on the initial site characterization
data.  Development and implementation of the new conceptual site model for MNA may require
collection of additional data to better depict existing site characteristics and parameter variability.


                                             48

-------
These data are used to determine trends, determine plume characteristics, and assess the
likelihood of meeting site remediation goals.

3.3    Interpreting the Data

3.3.1   Introduction

Data analysis and interpretation focus on two fundamental aspects: detection of changes or trends
in the data, and assessment of the changes or trends in the context of their impact on the potential
for MNA to achieve site-related goals. Data can be interpreted with regard to variability so that
attenuation-related changes and trends can be distinguished from other sources of variation in the
data. In addition, once changes or trends are identified, the  importance of the trend should be
considered (i.e., Does the trend indicate that MNA will or will not meet site-specific goals?).

Particular changes that are of interest include:

      Changes in ground-water flow rates or directions that indicate contaminants may move
       farther downgradient, laterally or vertically into previously unimpacted areas,

      Changes in contaminant concentrations -within the plume, that may indicate new releases,
       significant changes in the rate of release from any source materials that remain following
       implementation of all appropriate source control measures, or changes in the attenuation
       rate,

      Detections of contaminants outside the known plume or other compliance boundaries,
       indicating unacceptable plume expansion or additional source  areas, and

      Changes in geochemistry that may indicate changes in attenuation rates, such as changes
       in availability of electron donors and electron acceptors, changes in oxidation-reduction
       potential, or other primary geochemical indicators; or changes in sorption characteristics
       (e.g., adsorption, precipitation).

Data interpretation is complicated by data variability (Appendix A). This variability is inherent
in the measurement process, in -which case it may be evaluated in a carefully designed and
implemented sampling  and analysis-related QA/QC program, and it is also associated -with natural
causes. Due to the dynamic nature of natural processes, data typically exhibit fluctuations  of
various magnitudes. For example, there is an expected degree of natural variability -within  a
plume that may include small-scale expansion and shrinkage in response to changes in ground-
water flow rates and biological degradation rates throughout the year.  Contaminant
concentrations in individual -wells may fluctuate with changes in plume configuration caused by
oscillations in ground-water flow.  Also, changes in ground-water elevations can cause changes in
contaminant concentrations measured in monitoring -wells if there are vertical differences in
-water quality or other factors.

Examination of historical and spatial patterns may provide information useful in assessing the
source and significance of variability in the measured parameters.  For instance, if there are
multiple -wells in the same hydrogeologic setting that show a consistent trend for a number of
monitoring variables across two monitoring events, then one may suspect a natural variation,
such as recharge differences, and investigate a correlation of recharge or other external hydrologic
factors -with concentration as indicative of the likely cause of the change.  If, on the other hand, a


                                             49

-------
specific well shows an unusually large change in concentration between monitoring events,
relative to nearby monitoring -wells, then one may suspect a measurement factor is involved in
that localized change and look at the possible influences of sampling technique or other such
factors on the concentration change.

Because plume variability should be taken into account -when interpreting monitoring data, it is
important to have a good estimate of natural variability derived from the site investigation. In
addition, it is recommended that performance monitoring data be continually evaluated with
respect to variability, so that there is an ever-expanding data base for the site that can be used for
comparison -with the changes in parameters noted between monitoring events as well as for
comparison to predicted,  acceptable trends.

Statistical procedures are available that allow assessment of site characteristics and trends. If a
change in site characteristics or a departure from a predicted trend is found, it can be evaluated to
determine what impact the change may have on MNA's ability to achieve site-related goals.
Conceptual models for MNA can be used to conduct sensitivity analyses to help determine what
changes may be significant in the context of the site-related goals. For instance, sensitivity
analyses could be used to determine if a two-fold  or three-fold change in attenuation rate will
significantly affect attainment of remedial goals.  Departures from expected trends may not mean
remedy failure, because there may be a range of attenuation rates that will achieve the desired
remedial goals in the allotted time.

3.3.2  Preliminary Presentation and Evaluation of the Data

The data analyst conducts an initial evaluation and review of the data by calculating means,
modes, medians, and ranges, graphing the data, and mapping the data in plan and cross sectional
views to identify obvious patterns, relationships,  and anomalies.  Simple graphical representations
include site maps and cross sections with posted and contoured data, plots of data as a function of
time, plots of the relationship between two or more variables, and measures of central tendency
(e.g., means, modes, medians) and data dispersion (e.g., standard  deviations). It is recommended
that the data analyst consider as many graphical techniques as possible to maximize the amount of
information gained in this step.

As part of this initial evaluation process, it is essential that each data point be interpreted in the
context of its derivation in time and space.  For instance, monitoring -wells produce data that
should be interpreted in terms of the three-dimensional location  of the -well screen -with respect to
contaminant sources, site stratigraphy, the ground-water flow field, the plume, possible receptors,
season of the year, and, potentially, numerous other factors. The sampling methodology should
also be considered, due to possible differences in contaminant concentrations derived from
different sampling techniques or devices. The analytical methodology may be evaluated in terms
of accuracy, precision, and detection limits.  This list of data considerations is not exhaustive, but
merely emphasizes that a number  (i.e., a data point) should be placed in context to be interpreted.
The better a data analyst understands the derivation and context of the data, the more likely it is
that the data analyst can make the appropriate decision based on the data.

3.3.3  Data Comparisons

The following types of data comparisons are commonly useful during performance monitoring
evaluations.
                                             50

-------
3.3.3.1 Comparisons of Concentrations Within and Outside the Plume

Data obtained within the plume are compared to data from monitoring points upgradient of the
source area and sidegradient to the plume. Data acquired upgradient of the source area may be
used to monitor for contaminants coming from other sources and for electron donors and
acceptors migrating into the contaminated area  (U.S. EPA, 1998a; Wiedemeier etal., 1999). It
often is important to obtain geochemical data from locations sidegradient to the plume because
there may be geochemical changes in the ground water as it flows through the aquifer that are
unrelated to processes in the plume.  The range  of values and spatial patterns for the particular
parameter should be considered during data interpretation. If the range of values in the
geochemical setting surrounding the plume is similar to (or greater than) the range of values
for the parameter in the plume, then it may be problematic to attach any particular
interpretation to the changes of the parameter in the plume. For instance, suppose sulfate
values in uncontaminated ground water adjacent to a plume vary from less than 1 mg/L to 30
mg/L; then a drop in sulfate concentration from 20 mg/L to 5 mg/L within the plume may not
necessarily be conclusive evidence that sulfate reduction is important in plume geochemistry.

3.3.3.2 Trend Analyses

Data are compared to determine if temporal and spatial trends exist -within the plume and in
surrounding areas. Trends of interest include trends in contaminant and daughter product
concentrations, electron acceptors and donors, oxidation-reduction potential, and other general
geochemical indicators. These comparisons include trend-to-trend comparisons {e.g., a
decreasing trend in tetrachloroethene compared to an increasing trend in trichloroethene in the
same transmissive zones may indicate degradation of tetrachloroethene to trichloroethene).

Trends at individual sampling points or groups of sampling points may be compared to other
sampling points, or to trends in other groups of  sampling points.  For instance, contaminant
concentrations at individual sampling points may show different trends. However, evaluating
trends in data from all sampling locations in the plume will determine if the plume exhibits
stability or reduction in contaminant concentrations. Similarly, data from a group of sampling
points at the downgradient limits of a plume may be compared to data from previous sampling
rounds to determine if the  plume seems to be stable, shrinking, or expanding.

In some cases, particularly with petroleum hydrocarbons, known stoichiometric relationships
between usage of electron acceptors and degradation of contaminants may be used to relate
trends in geochemistry qualitatively or semi-quantitatively to degradation of contaminants.

3.3.3.3 Comparisons with Existing Literature and Laboratory Studies

Parameter values from other sites and the literature may be used to determine how ranges of
values at a particular site compare to the ranges  that have been found through research or
experience at other sites or laboratory studies. Typically, these comparisons are made during
site characterization, but they may also be useful in the performance monitoring planning
phase. For instance, a calculated degradation rate for a contaminant may be compared to values
found at other times and places at the same site  and to values found at other sites or in
laboratory studies.  If literature reports indicate  that a compound does not degrade under
conditions similar to those found at the site under investigation, but data from the site seem to
indicate the compound is disappearing at a substantial rate, it would be advisable to carefully
re-evaluate the monitoring operations before concluding that the compound is indeed


                                             51

-------
degrading.  Note that literature values cannot substitute for values determined at the site for
evaluating and monitoring MNA.

3.3.3.4  Comparisons with Threshold Values

Contaminant concentrations at the site may be compared to values set by regulation or other
factors.  For instance, MNA may be required to reduce the concentration of a contaminant to an
MCL of 5 mg/L at a particular sampling point.  Of particular interest for this comparison is that
the threshold value is typically assumed to be a true value, not an estimate of a population
parameter; that is, it has no variability associated with it. This affects the statistical procedures
used for comparing the values, as discussed in the references given below.

3.3.4  Statistics

The previously discussed data comparisons can be conducted by simply comparing measured
values, calculated (or graphed) trends, or set values for contaminants or geochemical parameters.
However, statistical procedures and models provide a formal, quantitative method for assessing
the relationship of sample measurements to characteristics of the sampled system, for using
sample data to make decisions, and for predicting future states of the sampled system.

Statistical procedures are often used to evaluate the variability associated with data, and to use
estimates of variability to guide decision-making processes. For example, if multiple analyses are
performed, statistical procedures can be used to express a measure of the  analytical variability
associated with a reported contaminant concentration {e.g., 4.9 mg/L +/- 3.2 mg/L, representing
the 95 % confidence limits on the mean value).

Statistical methods are also available to facilitate analysis and comparison of trends by considering
data variability through time.  For instance, changes in contaminant concentrations over space or
time can be used to calculate attenuation rates, and the variability associated with those rates can
be quantified with confidence intervals about the rates. These confidence intervals can be used to
determine the likelihood of attaining site-related remedial goals.  If all values of the attenuation
rate falling -within the confidence intervals lead to predictions that site remedial goals will be
attained in the desired time frame, then confidence that MNA can attain  remedial goals is
increased.

Implementing formal methods that compare data by taking into account data variability is
especially important for decision-making purposes.  Gibbons  (1994), Gilbert (1987) and Helsel
(1995) contain extensive discussions of the issues concerning use of statistics in environmental
and ground-water monitoring. For a detailed discussion of these points, as well as step-by-step
guidance on calculations for the various types of comparisons mentioned above, see also U.S. EPA
(2000c) and U.S. EPA (1992a).  Software resources include the GRITS package (U.S. EPA,
1992b) that is available from the U.S. EPA Office of Solid Waste.  DataQUEST (U.S. EPA,  1997d)
is another useful software package that is available from the EPA's Office of Research and
Development.

3.4    Elements of a Performance Monitoring Report

3.4.1   Introduction

Compilation and presentation of monitoring data in an easily usable form that facilitates
interpretation requires significant effort.  As was discussed for the monitoring plan, there is no

                                             52

-------
standard format that is universally acceptable, but, ideally, monitoring data would be acquired,
transmitted, evaluated and presented in an electronic form, and the report would make extensive
use of maps, cross sections, and figures to convey the results of monitoring efforts.

The following material has been prepared to provide some guidance regarding appropriate
components of performance monitoring reports. The format and content of these documents will
vary according to site-specific needs but most reports would contain or reference the elements
listed in Table 5 and briefly discussed in the following sections.

Elements of a monitoring report (Table 5) include:

       Summary of data interpretations  and recommendations,

       Background and site description,

       Monitoring network and schedule description,

       Evaluation of new data and comparisons with previous data and established performance
       criteria,

       Interpretation of new data with respect to the conceptual site model for natural
       attenuation, and

       Recommendations for action.

3.4.2  Summary

The summary typically would contain a brief description of the site, remedial goals, a narrative
summary of new data and their interpretation, and any recommended actions.  This portion of
the report should be written to convey essential findings to both technical and nontechnical
readers.

3.4.3  Background and Site Description

Background information on the site,  such as a discussion of the site setting, history,
characteristics, remedial goals, past and present remedial activities, and any institutional controls,
typically would be provided in abbreviated fashion or incorporated by reference to existing
documents.  The most salient issues  (e.g., hydrogeological setting, contaminant distribution,
remedial goals and actions) generally should be addressed at sufficient length that the technical
reviewer can judge the adequacy of the monitoring efforts.  The geologic and hydrogeologic setting
of the site, including controls on ground-water flow at the regional scale and site scale, would be
illustrated in map and cross-sectional views.

3.4.4  Monitoring Network and Schedule

The report typically would include information on the monitoring network including a list and
map of monitoring locations; a description of the construction of each monitoring point; and a
monitoring schedule specifying monitoring parameters, analytical methods, and sampling
frequency. Information regarding Data Quality Objectives (DQOs) may be provided or
incorporated by reference. It is particularly important that maps, cross-sections, and other visual


                                             53

-------
Table 5.   Elements of a Performance Monitoring Report
Summary of Data Interpretations and Recommendations
    Q   Brief description of site
    Q   Remedial goals
    Q   Narrative summary of new data and their interpretation
    Q   Recommended actions

Background and Site Description
    The following information typically would be discussed or incorporated by reference:
    Q   Site setting,history,characteristics
    Q   Remedial goals
    Q   Past and present remedial activities,and any institutional controls
    Q   Map and cross-sectional views illustrating geologic and hydrogeologic setting of the site, including controls on ground-
         water flow at the regional scale and site scale

Monitoring Network and  Schedule
    Q   List and maps of monitoring locations for each sampled medium and each major hydrogeologic unit
    Q   Description and construction details for each monitoring point
    Q   Monitoring schedule specifying monitoring para meters, analytical met hods, and sampling frequency for each
         monitoring location
    U   Data Quality Objectives (DQOs) to be met

Evaluation of New Data
    Q   Detailed discussion of new results and evaluations
    Q   Data in tables and  electronic files
    Q   Potentiometric surface maps for each hydrostratigraphic unit
    Q   Hydrographs of ground-water elevations for key wells in each hydrostratigraphic unit and surface-water monitoring
         points
    Q   Contaminant data posted and contoured on maps for each media and major hydrogeologic unit
    Q   Hydrochemical cross sections along and perpendicular to ground-water flow directions depicting contaminants,
         monitoring points,and hydrogeology
    Q   Geochemicaldata posted and contoured on maps for each major hydrogeologic unit
    Q   Cross sections depicting geochemical data, monitoring points,and hydrogeology
    Q   Comparison of the new data with previous data and established performance criteria
    Q   Results of statistical comparisons
    Q   Discussion of trend sand the relation of any trends to remedial goals
    Q   Assessment of measurement variability from analysis of QA/QC data
    Q   Observed changes in land use

Evaluation of Institutional Controls  (ICs)
    Q   Description of ICs that are in place with appropriate verification
    Q   Evaluation of the effectiveness of ICs
    U   Discussion of any pending changes in property ownership
    Q   Observed changes in land or resource uses

MNA Conceptual Site Model Evaluation
    Q   Evaluation of the conceptual site model incorporating any new data and data trends
    Q   Discussion of consistency of previous conceptual site  model with new data
    Q   Suspected sources for continued ground-watercontamination (e.g., number,location(s),characteristicsof sources)
    Q   Trends in contaminant and geochemistry values
    Q   Discussion of any observed  changes in site hydrology (e.g., water elevations,g round-water velocities)
    Q   Discussion of refinements/modifications to conceptual site model
    Q   Consistency of current data  with previous predictions
    Q   Discussion of changes in land use and potential effects on the conceptual model

Recommendations (as warranted)
    Q   Recommended changes in monitoring locations
    Q   Recommended changes in monitoring frequencies
    Q   Recommended changes in sampling methods
    U   Recommended changes in analyses
    U   Discussion of new data in relation to performance criteria previously established to trigger implementation of a
      contingency remedy
    Q   Discussion of changes in land use and potential effects on site remedies, and remedy protect! veness for human and
         ecological receptors
    U   Recommended remedy modifications (e.g.,additional source removal actions)
    Q   Recommendations for starting verification monitoring,or terminating performance monitoring
    U   Rationale for recommended changes
                                                        54

-------
aids be extensively used so that the reviewer can determine exactly where, when, and by what
methodologies each sample was obtained and analyzed.  In situations where different
hydrogeologic units are monitored, separate maps with wells/sampling locations for each major
lithologic unit may provide additional clarity in visualization of the network design.

3.4.5  Evaluation of New Data

The evaluation of new data and comparisons with previous data typically should provide a:

       Detailed discussion of new results and evaluations with presentation of data in tables,
       maps, and figures,

       Comparison of the new data with previous data and established performance criteria,

       Discussion of uncertainty with statistical measures of variability, including discussion of
       measurement variability assessed through evaluation of QA/QC data, and

       Discussion of trends and the relation of any data trends to the remedial goals.

Tables ordinarily would include hydraulic head data, contaminant data, and geochemical data in
both in-well and between-well comparisons. In other -words, all data for each sampling point
would be tabulated to  facilitate assessment of the geochemistry at that location, and a tabulation of
data by monitoring parameter -would be provided to facilitate comparison of each parameter
across all sampling locations.  Tables and figures depicting contaminant concentrations and
geochemical data for individual -wells over time -would be included to aid in the evaluation  of
trends. Maps and cross-sections of the plume -would be prepared and compared to previous
conditions. Contour maps of the contaminant concentrations and geochemical parameters  are
helpful for visualizing broad trends. These comparisons aid in the evaluation of temporal
changes.  They may also serve to identify areas where additional data are needed. Data
submission in electronic formats should also be considered.

3.4.6   E/aluation of Institutional Controls

The performance  monitoring  report typically -would include a full description of all institutional
controls implemented or planned at the site, along -with verification  of their implementation.  The
results of the monitoring activities, including an evaluation of the effectiveness of the individual
institutional controls, should be discussed at length so that the reviewer  can judge -whether the
controls are likely to continue to meet performance objectives. Any observed or pending changes
in land or resource uses or ownership (e.g., property ownership change,  housing developments,
-well installations) should be discussed in view of their current and possible future impact on the
effectiveness of the controls and the performance monitoring operations.

3.4.7  Conceptual Site Model Evaluation

The report typically -would contain an evaluation of the conceptual site model incorporating any
new data and data trends.  Included in this section -would be a discussion of the following
questions:

       (1)    Do the new data fit the previous conceptual site model?

-------
              Consider:

                Suspected sources for continued ground-water contamination (e.g., number,
                 location(s), characteristics of sources),

                Trends in contaminant and geochemistry values, and

                Changes in hydrologic factors {e.g., ground-water elevations, hydraulic
                 gradients, velocities)

       (2)     Should the conceptual site model be refined/modified?

       (3)     Do current data support previous predictions?

Consider uncertainty in previous predictions and implications for the effectiveness of natural
attenuation mechanisms and the current monitoring network.

3.4.8  Recommendations

Recommendations for action, based on interpretation and evaluation of the data in reference to
RAOs, monitoring objectives and performance criteria should be provided and discussed.
Recommendations for action may include changes in monitoring locations, frequencies, methods,
and analyses with a rationale for  changes. Recommended actions may also include additional
source removal or other remedy modifications, implementation of contingency or alternative
remedies, advancement to verification monitoring, or termination of performance  monitoring
based on achievement of site remedial goals.

-------
                                      Chapter 4
        APPLICATION OF MONITORING DATA TO REMEDIAL DECISIONS
4.1    Introduction

Following data evaluations, decisions are routinely made regarding the effectiveness of the MNA
remedy, the effectiveness of institutional controls, the adequacy of the monitoring program, and
the adequacy of the conceptual site model for MNA. Important decisions that may be made
include:
       Continue monitoring program without change,

       Modify the monitoring program,

       Modify the institutional controls,

       Implement the contingency or alternative remedy, or
       Verify remedial goals have been met and terminate performance monitoring.

Site-specific criteria should be developed to define conditions that indicate the appropriateness of
increased or decreased monitoring, additional characterization, re-evaluation of the conceptual
site model,  modification of institutional controls, implementation of a contingency or alternative
remedy, or termination of performance monitoring. The following discussion briefly considers
each decision and points relative to why it may be chosen.

4.2    Decision 1 - Continue Monitoring Program Without Change

Evidence leading to this decision would include contaminant concentrations, including toxic
transformation products and any mobilized by-products or secondary contaminants (e.g., arsenic,
manganese) remaining within the bounds of acceptable trends.  Ground-water flow parameters
would not have changed outside previously identified acceptable ranges.  The geochemistry would
not have changed in such a way as to indicate that the contaminant degradation or other natural
attenuation processes would be significantly affected. Significant changes in geochemistry
include a marked reduction of electron acceptors or donors essential to the important natural
attenuation processes identified at the site, an influx of substances such as oxygen or nitrate that
could interfere with desired processes, or a change in oxidation-reduction potential bringing the
potential into a range not suitable for the identified important natural attenuation processes
(Ferrey et al.,  2001). Performance monitoring continues until all remedial action objectives have
been met.

4.3    Decision 2 - Modify the Monitoring Program

Modification of the monitoring program may be warranted to better reflect changing conditions
or increased understanding of natural attenuation processes at the site. Changing conditions, the


                                            57

-------
need to test key assumptions, or providing final confirmation of remedial goal achievement may
require new sampling points or increases in monitoring parameters or frequency. Examples
include:

       A change in ground-water flow rate and direction caused by installation of an irrigation
       well may necessitate re-evaluation of the monitoring frequency and locations (Figure 9),

       Input of nitrate from fertilizer applications may sufficiently change the geochemistry to
       inhibit reductive dechlorination. Increased monitoring frequency and locations may be
       necessary to delineate the effect of nitrate influx,

       Periodic pulses of contaminants from the former source area may be caused by flushing of
       contaminant from undetected residual material in the vadose zone. Monitoring in the
       vadose zone as well as the saturated zone may be needed to determine the nature and
       characteristics of the  contaminant pulses and provide data for assessing the appropriate
       response, and

       Contradictory indications of the potential for plume expansion in some areas or the data
       may be so highly variable that interpretation is  difficult and unacceptably uncertain. High
       data variability, making interpretation difficult, may be caused by variability inherent in
       sampling and, in some cases, analysis. In addition, contaminant concentration changes
       caused by seasonal changes  in ground-water flow, degradation rates, contaminant releases
       from poorly controlled source (s),  and influx of electron acceptors can make interpretation
       difficult unless monitoring is sufficiently frequent to allow delineation of the changes due
       to seasonal cycling. A determination may be made to collect more data sufficient to
       interpret trends. However, if the protectiveness of the remedy is in question,
       implementation of the specified contingency or an alternative remedy will generally be
       warranted.

In other instances, decreases in monitoring parameters, frequency, or locations may be
appropriate.  For example, decreases in monitoring frequency for certain parameters may be
warranted if the remedy is proceeding according to expectations and trends are stable after
evaluation of data from a sufficient number of monitoring periods (e.g., many years). To support
such a decision, the available data generally should cover a time period sufficient to allow
evaluation of seasonal trends and other long-term cycles and trends. Evidence supporting
reduced monitoring includes trends in  contaminant concentrations continuing as expected and
geochemistry and ground-water flow conditions remaining stable (i.e., in the ranges suitable for
continued natural attenuation).  Also, the time required for implementation of the contingency
remedy after established decision criteria are met would be  considered in any decision to reduce
monitoring frequency.  In any case, performance monitoring should continue until all remedial
action objectives have been met (e.g., contaminant concentrations  are below levels of concern).

Once performance  monitoring data indicate that site remedial goals have been met, a period of
verification monitoring may be initiated.   Verification monitoring  may include different sampling
frequencies or additional locations from those used during routine performance monitoring,
especially if monitoring locations or frequency had been reduced during the performance
monitoring period.  For instance, the number of monitoring locations and monitoring frequency
may have been reduced if data had indicated that the MNA remedy was proceeding as expected.
In such cases, some of the sampling locations or times may have been eliminated from the
sampling program as unnecessary.  However, it often may be appropriate to include these
sampling locations and times during verification sampling in order to increase certainty that goals


                                             58

-------
are met at all locations and times.  Once verification monitoring has shown that all MNA remedy
goals have been met, then termination of performance monitoring generally would be warranted.

4.4    Decision 3 - Modify Institutional Controls

Some changes in land or ground-water use have the potential to decrease the effectiveness or the
protectiveness of the MNA remedy. Such changes could include installation of new pumping
wells in the vicinity of the site. New pumping -wells could change ground-water flow patterns and
increase the potential for plume migration, thereby increasing the risk to receptors.  Some
changes in land use  could affect subsurface geochemical conditions. For example, excessive
application of nitrogen-containing fertilizers could cause an influx of nitrate to ground water and
possibly inhibit reductive dechlorination. Changes in land or ground-water use should be
monitored and evaluated to  determine potential impacts to the MNA remedy.  If existing
institutional controls or procedures for monitoring the controls are not sufficient to maintain the
effectiveness or protectiveness of the MNA remedy, the site manager should consider modifying
the institutional controls or monitoring procedures. If the changes in land or ground-water use
resulted from a breach of the institutional controls established for the site, the procedures used to
monitor and enforce the existing institutional controls should be evaluated and modified to
prevent future breaches.  If the changes in land or ground-water use were allowed by the current
institutional controls, the site manager should consider modifying the institutional controls to
minimize future adverse impacts to the MNA remedy.

4.5    Decision 4 - Implement a Contingency or Alternative Remedy

Remedies relying upon MNA may have an associated contingency remedy in case the MNA
component fails to perform  at the desired effectiveness (U.S. EPA, 1999a).  In any case,
alternative remedies are  usually considered during remedy selection and these or other remedies
can be considered for use if it is determined that the current remedy has failed. Criteria for
determining specific failures and implementing modifications, a contingency, or alternative
remedy may be explicitly stated in the remedy decision documents. If not provided in previous
documents, it is recommended that objective and quantitative decision criteria for determining
remedy failure be developed for use in designing the performance monitoring system and
evaluating data from the system.

Development and evaluation of specific criteria to trigger implementation  of a contingency or
alternative remedy are generally based on and related to the purpose behind site RAOs. That
relationship not only provides a rationale for a specific trigger, but also provides context for
evaluating the trigger once it occurs, and deciding on the appropriate response. For example, a
site RAO may require that the MNA remedy control the plume.  Suppose that contaminants were
detected in wells outside the previously-known plume boundary in 1)  an existing well with a long
history of no contamination, or 2) a newly installed well. If the aim of the RAO was that there
should be no contaminant beyond a specific point (the previously-known plume boundary), then
the response to situation 1)  or 2) may be the same.  However, if the aim  of the RAO was  that the
plume should not expand, then the response to situations 1) and 2) may differ, because
contaminants in the existing well indicate plume expansion, -whereas contaminants in the new
-well may or may not indicate plume expansion (i.e., the contaminants may have been at the new
location for many years,  but there -were no samples previously collected  in the  new location).

Situations that may  -warrant implementation of a contingency or other remedy modifications
depend on site conditions and include, but are not limited to, the following examples:


                                            59

-------
       Contaminant concentrations in soil or ground -water at specified locations exhibit an
       increasing trend not originally predicted during remedy selection,

       Near-source -wells exhibit large concentration increases indicative of a new or renewed
       release,

       Contaminants are identified in monitoring -wells located outside the original plume
       boundary or other specified compliance boundary,

       Contaminant concentrations are not decreasing at the predetermined rate necessary to
       meet the remediation objectives,

       Changes in land and/or ground-water use will adversely affect the protectiveness of the
       MNA remedy, and

       Contaminants are identified in locations posing unacceptable risk to human or ecological
       receptors.

4.5.1   Decision Criterion 1:    Contaminant Concentrations in Soil or Ground Water at Specified
                             Locations Exhibit an Increasing Trend Not Originally Predicted
                             During Remedy Selection

This criterion involves  the identification and assessment of trends in contaminant concentration
data, including toxic transformation products or  mobilized inorganic constituents, that indicate
plume expansion, lack of progress toward cleanup objectives, or additional contaminant releases.
Statistical tests such as  those described in the references may be used to objectively identify
trends in these data.  An example of a criterion based on a trend -would be detection of a
consistent, increasing trend in contaminant concentrations in -wells near the downgradient edge
of the plume. Such an event may be evidence of  plume expansion, and,  if so, may trigger the
contingency or alternative remedy based on the previously-established, site-specific guidelines for
assessing trends and remedy effectiveness.

For some locations in the plume,  a detection of a significant change may not trigger immediate
implementation of a contingency or alternative remedy.  For  instance, a low-level or otherwise
limited contaminant concentration increase near the source area does not necessarily mean that
remediation time frames will be expanded,  or that by the time the pulse moves downgradient to
the boundary of the plume there -will be a significant plume expansion.  Criteria for determining
acceptable limits for increases in  contaminant concentrations in locations near the source area
should be based on site conditions and specified in decision documents. Because the contaminant
travel time from the source area to the downgradient edge of the plume may be many years at
some  sites, there may be time  for further monitoring and assessment before a contingency or
alternative remedy -would have to be implemented. In some cases, if contaminants will not reach
receptors for several years, it may be possible to allow verification procedures to continue until
the next sampling event (i.e., the next quarterly or semiannual sampling event). In contrast, if a
significant trend of increasing contaminant concentration in -wells near the downgradient edge of
the plume -was noted, there may be much less time available to assess the trend if receptors -were
only a short travel time downgradient. In this case, a contingency or alternative remedy may be
triggered immediately.

In some cases, the  site characterization may indicate that concentrations in some portions of the
plume are expected to increase temporarily, yet MNA -would  still meet remedial objectives.  If so,

-------
the temporary increase would not trigger the contingency or alternative remedy as long as the
increase conformed to the predictions and MNA remained protective. Specific criteria for
determining that a significant change in trend has occurred may be developed to ensure that
sampling variability or acceptable seasonal fluctuations do not unnecessarily trigger a
contingency or alternative remedy.

4.5.2  Decision Criterion 2:    Near-Source Wells Exhibit Large Concentration Increases
                             Indicative of a New or Renewed Release

A release may result from such conditions as drum rupture or increased flushing of the source
area caused by changes in hydrologic conditions. The expectation is that the new or renewed
release may increase the time needed to meet remedial objectives or cause the plume to expand.
Contaminant concentration increases beyond those previously predicted and stated in decision
documents could trigger a contingency or alternative remedy, or implementation of sampling
efforts to determine the causes of the increases and the impact on the remedy. In some cases,
detailed characterization and modeling of contaminant fate and migration may indicate that the
increased concentrations would be expected to attenuate without causing plume expansion,
allowing observation for several years to test model predictions, if time is available. In
situations where evaluation shows the observed contaminant concentration increases are
expected to unacceptably increase the time for plume restoration, or cause unacceptable plume
expansion, triggering the  contingency or alternative remedy generally would be warranted.

4.5.3  Decision Criterion 3:    Detection of a Contaminant in Monitoring Wells Located
                             Outside of the Original Plume Boundary or Other Compliance
                             Monitoring Boundaries

Detections of contaminants outside of the predetermined horizontal or vertical plume
boundaries or other compliance boundaries  may indicate unacceptable plume expansion.
Procedures for verifying contaminant detections typically should be included in the monitoring
plan. The choice of procedures depends on many factors including: the locations of the new
detections, distance between these monitoring points and receptors, possible contaminant
migration rates to receptors,  time frames required for implementation of the contingency  or
alternative remedy, and monitoring procedures.  Appropriate verification procedures will often
include:

      Verification of the detection, by verifying the analytical procedures showing the
       detection and identity of the contaminant. This verification would be conducted by
       the analytical laboratory.  It may also be possible for the laboratory to analyze
       another aliquot of the same sample, if available, and

      Verification by immediately resampling the well. Criteria defining the necessary
       level of agreement between results may be developed to facilitate evaluations. For
       example, a conservative decision criterion -would be implementation of the
       contingency remedy if the contaminant concentration measured in the second
       sample exceeded the action level.

If these are verified, this may indicate remedy failure, depending on the RAO for the site.
Verified detections of contaminants in these wells may be used to trigger implementation  of an
alternative or contingency remedy.  Similar criteria may be developed to support decisions
regarding other types of compliance boundaries.


                                             61

-------
4.5.4  Decision Criterion 4:     Contaminant Concentrations Are Not Decreasing at a Sufficiently
                              Rapid Rate to Meet the Remediation Objectives

This assessment will involve such evaluations as the comparison of current data with predictions
used to support the remedy decision and evaluation or projection of temporal trends in
contaminant concentrations. There will generally be significant uncertainty in these evaluations
due to uncertainty in the projection of future conditions and sampling and measurement
variability. Specific, objective criteria and milestones specified in the remedy decision should be
reviewed. If specific criteria and milestones are not stated, they may be developed in the
performance monitoring plan. Specific milestones  (e.g., 50% contaminant concentration decrease
in all wells within a specified number of years) may be developed using such techniques as
projection of calculated attenuation rates. Failure to meet specified milestones or criteria could
result in triggering of the contingency or alternative remedy.

4.5.5  Decision Criterion 5:     Changes in Land and/or Ground-Water Use that Have the Potential
                              to Reduce the Protectiveness of the MNA Remedy

This involves re-evaluating the conceptual site model for MNA and determining whether there
are receptors at higher risk than originally conceptualized or whether the change in land or
ground-water use has resulted in hydrologic changes that affect plume stability, or geochemical
changes that affect biotransformation processes. Such changes could increase the potential for
contaminant migration and increased risk to the receptors as noted in Section 4.4. In such
situations where modification of the institution controls is not possible, appropriate, or sufficient
to restore the protectiveness of the remedy, triggering of a contingency or alternative remedy may
be appropriate.

4.5.6  Decision Criterion 6:     Contaminants Are Identified in Locations Posing or Having the
                              Potential to Pose Unacceptable Risk to Receptors

In addition to  monitoring contaminants in ground water at locations that indicate unacceptable
risk to human receptors, cross-media transfer of contaminants from ground water to surface
water, to indoor air, or other human or ecological receptors should be monitored based on a site-
specific evaluation of the risk posed by the contaminants.  Monitoring may include, for example:

     Sampling in the aquifer near the point of discharge to the surface-water body,

     Pore water or sediment samples where ground water moves into the surface water,

     Samples from the surface-water body, and

     Soil gas and indoor air in enclosed areas near plume or source materials.

Decision criteria may be based on contaminant concentrations exceeding action levels in these
media. Assessment of this decision criterion would use statistical tests and verification
procedures similar to decision criteria 1 and 3.
                                             62

-------
4.6    Decision 4 - Terminate Performance Monitoring

Specific methods and criteria for demonstrating the attainment of all remedial action objectives
should be developed as noted in Sections 2.6.8 and 2.7.7.  In general, once a thorough analysis of
performance/verification monitoring data shows that all MNA-related site remedial goals have
been achieved, termination of performance monitoring -would be warranted.

-------
64

-------
                                     REFERENCES
Azadpour-Keeley, A., H.H. Russell, and G.W. Sewell,  1999. Microbial Processes Affecting
   Monitored Natural Attenuation of Contaminants in the Subsurface, EPA/540/S-99/001. U.S.
   Environmental Protection Agency,  Office of Research and Development, National Risk
   Management Research Laboratory,  Cincinnati, OH.
   http://www.epa. gov/ada/download/issue/microbial .pdf

Azadpour-Keeley, A., J.W. Keeley, H.H. Russell, and G.W. Sewell, 2001. Monitored natural
   attenuation of contaminants in the subsurface: Processes. Ground Water Monitoring and
   Remediation, 21(2): 97-107.

Barcelona, M.J., H.A. Wehrmann, M.R. Schock, M.E. Sievers, and J.R. Karny, 1989. Sampling
   Frequency for Ground-Water Quality Monitoring, EPA/600/4-89/032. U.S. Environmental
   Protection Agency, Environmental  Monitoring Systems Laboratory, Las Vegas, NV.

Barker, J., 2000. Redox processes in petroleum hydrocarbon remediation or Why I never want to
   measure Eh again. In: Workshop on Monitoring Oxidation-Reduction Processes for Ground-Water
   Restoration, Workshop Summary, Dallas, Texas, April 25-27, 2000, EPA/600/R-02/002. U.S.
   Environmental Protection Agency,  Office of Research and Development, National Risk
   Management Research Laboratory,  Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa 600 r02  002.pdf

Brady, P.V., B.P. Spalding, K.M. Krupka, R.D. Waters, P. Zhang, D.J. Borns, and WD. Brady,
   1999.  Site Screening and Technical Guidance for Monitored Natural Attenuation at DOE Sites.
   SAND99-0464. US Department of Energy,  Sandia National Laboratory, Albuquerque, NM.
   http://www.sandia.gov/eesector/gs/gc/na/mnatoolbox.pdf

Buscheck, T.E., and C.M. Alcantar, 1995. Regression  techniques and analytical solutions to
   demonstrate intrinsic bioremediation. Bioremediation, 3(1):109-116.

Butler, J.J., 1998. Design, Performance, and Analysis of Slug Tests. Lewis Publishers, Boca Raton,  FL.

Chapelle, F.H., 1996. Identifying redox conditions that favor the natural attenuation of
   chlorinated ethenes in contaminated ground-water systems. In: Symposium on Natural
   Attenuation of Chlorinated Organics in Ground Water, Hyatt Regency Dallas, Dallas, TX,
   September 11-13,  1996, EPA/540/R-96/509. U.S.  Environmental Protection Agency, Office of
   Research and Development, Washington, DC.
   http://www.epa. gov/ordntrnt/ORD/WebPubs/natural/natural .pdf

Chapelle, F.H., 2000.  Identifying the distribution of terminal electron-accepting processes
   (TEAPS) in ground-water systems. In: Workshop on Monitoring Oxidation-Reduction Processes
   for Ground-Water Restoration, Workshop Summary, Dallas, Texas, April 25-27, 2000,
                                            65

-------
   EPA/600/R-02/002. U.S. Environmental Protection Agency, Office of Research and
   Development, National Risk Management Research Laboratory, Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa  600 r02 002.pdf

Cherry, J.A., 1996. Conceptual models for chlorinated solvent plumes and their relevance to
   intrinsic remediation. In: Symposium on Natural Attenuation of Chlorinated Organics in
   Ground Water, Hyatt Regency Dallas, Dallas, TX, September 11-13, 1996, EPA/540/R-96/509.
   U.S. Environmental Protection Agency, Office of Research and Development, Washington, DC.
   http://www.epa. gov/ordntrnt/ORD/WebPubs/natural/natural .pdf

Cohen, R.M., and J.W. Mercer, 1993. DNAPL Site Evaluation. CRC Press, Boca Raton, FL.

Cohen, R.M., A.H. Vincent, J.W. Mercer, C.R. Faust, and C.P. Spaulding, 1994. Methods for
   MonitoringPump-and-Treat Performance, EPA/600/R-94/123. U.S. Environmental Protection
   Agency, Office of Research and Development,  Ada, OK. 74820.
   http://www.epa.gov/rlOearth/offices/oea/gwf/issue20.pdf

Ferrey, M.L., J.B. Lundy, and P. Estuesta, 2001. The effect of groundwater aeration on PCE
   natural attenuation patterns. Bioremediation Journal, 5(3):211-224.

Fetter, C.W, 1994. AppliedHydrogeology. Macmillan College Publishing Company, New York.

Galloway, WE., and J.M. Sharp, Jr., 1998. Characterizing Aquifer Heterogeneity within
   Terrigenous Clastic Depositional Systems.  In: SEPM Concepts in Hydrogeology and
   Environmental Geology No. 1:  Hydrogeologic Models of Sedimentary Aquifers. G.S.  Fraser and
   J.M. Davis, eds., Society for Sedimentary Geology, Tulsa, OK.

Gibbons, R.D., 1994.  Statistical Methods for Groundwater Monitoring.  John Wiley & Sons, New
   York.

Gilbert, R.O., 1987. Statistical Methods for Environmental Pollution Monitoring. VanNostrand
   Reinhold, New York.

Helsel, D.R., and R.M. Hirsch, 1995. Statistical Methods in Water Resources. Elsevier, Amsterdam,
   The Netherlands.

Huling, S. G., B. E. Pivetz, and R. Stransky, 2002.  Terminal electron acceptor mass balance:
   NAPLS and natural attenuation.  Journal of Environmental Engineering, 128(3): 246-252.

ITRC (Interstate Technology and Regulatory Cooperation Work Group), 2000. Dense Non-
   Aqueous Phase Liquids (DNAPLs): Review of Emerging Characterization and Remediation
   Technologies.
   http://www.itrcweb.org/DNAPL-1 .pdf

Kao, C.M., and YS. Wang, 2001. Field investigation of the natural attenuation and intrinsic
   biodegradation rates at an underground storage tank site. Environmental Geology, 40 (4-
   5):622-630.

Kruseman, G.P., and N.A.de Ridder,  1989. Analysis and Evaluation of Pumping Test Data,
   International Institute for Land Reclamation and Improvement. Wageningen, The Netherlands.


                                            66

-------
Leahy, J.G., and G.S. Shreve, 2000.  The effect of organic carbon on the sequential reductive
   dehalogenation of tetrachloroethylene in landfill leachates. Water Research, 34(8):2390-2396.

Ludwig, R., M. Barcelona, and K. Piontek, 2000. Redox processes in petroleum hydrocarbon site
   characterization and remediation. In: Workshop on Monitoring Oxidation-Reduction Processes
   for Ground-Water Restoration, Workshop Summary, Dallas, Texas, April 25-27, 2000, EPA/
   600/R-02/002. U.S. Environmental Protection Agency, Office of Research and Development,
   National Risk Management Research Laboratory, Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa 600 r02 002.pdf

McNab, W.W. and B.P. Dooher, 1998.  Critique of a steady-state analytical method for estimating
   contaminant degradation rates. Ground Water, 36(6):983-987.

Mercer, J.W., and R.M. Cohen, 1990. A review of immiscible fluids in the subsurface: Properties,
   models, characterization and remediation. Journal of Contaminant Hydrology, 6:107-163.

Mohr, T.K.G., 2001. Solvent Stabilizers, white paper, Santa Clara Valley Water District, Santa
   Clara, CA.

National Research Council, 2000. Natural Attenuation for Groundwater Remediation. National
   Academy Press, Washington, DC.
   http://books.nap.edu/books/0309069327/html/index.html

Palmer,  C.D., and W Fish, 1992. Chemical Enhancements to Pump-and-Treat Remediation, EPA/
   540/S-92/001. U.S. Environmental Protection Agency, Office of Research and Development,
   National Risk Management Research Laboratory, Cincinnati, OH.
   http://www.epa. gov/ada/download/issue/chemen .pdf

U.S. EPA,  1986a. RCRA Ground-Water Monitoring Technical Enforcement Guidance Document,
   OSWER-9950.1. U.S. Environmental Protection Agency, Office of Solid Waste and
   Emergency Response, Washington, DC.

U.S. EPA,  1986b. Test Methods for Solid Waste, SW-846.  U.S. Environmental Protection Agency,
   Office of Solid Waste and Emergency Response, Washington, DC.
   http ://www. epa. gov/epaoswer/hazwaste/test/sw8 4 6. htm

U.S. EPA,  1988. Soil Gas Sensing for Detection and Mapping of Volatile Organics, EPA/600/8-87/
   036. U.S. Environmental Protection Agency, Environmental Monitoring and Support
   Laboratory, Las Vegas, NV

U.S. EPA,  1991a. Seminar Publication: Site Characterization for Subsurface Remediation, EPA/
   625/4-91/026. U.S. Environmental Protection Agency, Office of Research and Development,
   Cincinnati, OH.
   http://www.epa.gov/swerustl/cat/sitchasu.pdf

U.S. EPA,  1991b. A Review of Methods for Assessing Nonpoint Source Contaminated Ground-Water
   Discharge to Surface Water, EPA/570/9-91/010. U.S. Environmental Protection Agency,
   Washington, DC.
                                            67

-------
U.S. EPA, 1992a. Methods for Evaluating the Attainment of Cleanup Standards, Volume 2: Ground
   Water, EPA/230/R-92/014. U.S. Environmental Protection Agency, Office of Policy, Planning,
   and Evaluation, Environmental Statistics and Information Division, Washington, DC.
   http://www.epa.gov/superfund/resources/gwdocs/per eva.htm

U.S. EPA, 1992b. GRITS/STAT v4.2, A GRound Water Information Tracking System with
   STATistical Analysis Capability, EPA/625/11-91/002. U.S. Environmental Protection
   Agency, Office of Research and Development, Washington, DC.
   http://www.epa.gov/correctiveaction/resource/guidance/sitechar/gwstats/gritsstat/
   gritsstat.htm

U.S. EPA, 1992c. Statistical Analysis of Ground-water Monitoring at RCRA Facilities, Addendum to
   Interim Final Guidance.  U.S. Environmental Protection Agency, Office of Solid Waste,
   Washington, DC.
   http://www.epa.gov/epaoswer/hazwaste/ca/resource/guidance/sitechar/gwstats/gritsstat/
   download/addendum.pdf

U.S. EPA, 1993a. Subsurface Characterization and Monitoring Techniques: A Desk Reference Guide,
   Volume 1: Solids and Ground Water, EPA/625/R-93/003a. U.S. Environmental Protection
   Agency, Office of Research and Development, Washington, DC.

U.S. EPA, 1993b. Subsurface Characterization and Monitoring Techniques: A Desk Reference Guide,
   Volume 2: The VadoseZone, Field Screening and Analytical Methods, EPA/625/R-93/003b. U.S.
   Environmental Protection Agency,  Office  of Research and Development, Washington, DC.

U.S. EPA, 1993c. Seminar on Regional Considerations for Dense Nonaqueous Phase Liquids at
   Hazardous Waste Sites: Implementation and Enforcement Issues, EPA/600/K-93/004. U.S.
   Environmental Protection Agency,  Office  of Emergency and Remedial Response,
   Washington, DC.

U.S. EPA, 1994.  DNAPL Site Characterization, EPA/540/F-94/049. U.S. Environmental
   Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC.

U.S. EPA, 1997a. Rules of Thumb for Superfund Remedy Selection, EPA/540/R/97/013, OSWER
   9355.0-69. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency
   Response, Washington, DC.

U.S. EPA, 1997b. Expedited Site Assessment Tools for Underground Storage Tank Sites: A Guide for
   Regulators, EPA/510/B-97/001. U.S. Environmental Protection Agency,  Office of
   Underground Storage Tanks, Washington, DC.

U.S. EPA, 1997c. Ecological Risk Assessment Guidance for Superfund: Process for Designing and
   Conducting Ecological Risk Assessments, Interim Final, EPA/540/R-97/006. U.S.
   Environmental Protection Agency,  Office  of Solid Waste and Emergency Response.
   Washington, DC.
   http://www.mng-ltd.eom/ert/p ecorisk.htm

U.S. EPA, 1997d. DataQUEST Data Quality Evaluation Statistical Toolbox, EPA/600/R-96/085.
   U.S. Environmental Protection Agency, Office of Research and Development, Washington, DC.
   http://www.epa.gov/quality/qa docs.html


                                            68

-------
U.S. EPA, 1998a. Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in
   Ground Water, EPA/600/R-98/128. U.S. Environmental Protection Agency, Office of Research
   and Development, National Risk Management Research Laboratory, Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/protocol.pdf

U.S. EPA, 1998b. Risk Assessment Guidance for Superfund (RAGS): Volume I Human Health
   Evaluation Manual, PartD: Standardized Planning, Reporting, and Review of Superfund Risk
   Assessments, EPA/540/R-97/033, OSWER 9285.7-01 D-l, PB 97-963305. U.S. Environmental
   Protection Agency, Office of Solid Waste and Emergency Response, Washington, DC.
   http://www.epa.gov/superfund/programs/risk/ragsa/

U.S. EPA, 1998c. EPA Guidance for Quality Assurance Project Plans, EPA QA/G-5, EPA/600/R-
   98/018. U.S. Environmental Protection Agency, Office of Research and Development,
   Washington, DC.
   http://www.epa.gov/swerustl/cat/epaqag5.pdf

U.S. EPA, 1999a.  Use of Monitored Natural Attenuation at Superfund, RCRA Corrective Action,
   and Underground Storage Tank Sites, Directive 9200.4-17P. U.S. Environmental Protection
   Agency, Office of Solid Waste and Emergency Response, Washington, DC.
   http://www.epa.gov/swerustl/directiv/d9200417.pdf

U.S. EPA, 1999b. A Guide To Preparing Superfund Proposed Plans, Records Of Decision, And Other
   Remedy Selection Decision Documents. EPA/540/R-98/031, OSWER 9200.1-23P, PB98-963241.
   U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response,
   Washington, DC.
   http://www.epa.gov/superfund/resources/remedy/rods/

U.S. EPA, 2000a. Data Quality Objectives Process for Hazardous Waste Sites, EPA OQ/G-4HW,
   EPA/600/R-00/007. U.S. Environmental Protection Agency, Office  of Environmental
   Information, Washington, DC.
   http://www.epa.gov/quality/qa  docs.html

U.S. EPA, 2000b. Workshop on Monitoring Oxidation-Reduction Processes for Ground-Water
   Restoration, Workshop Summary, Dallas, Texas, April 25-27, 2000, EPA/600/R-02/002. U.S.
   Environmental Protection Agency, Office of Research and Development, National Risk
   Management Research Laboratory, Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa 600  r02  002.pdf

U.S. EPA, 2000c. Guidance for Data Quality Assessment: Practical Methods for Data Analysis
   (G-9), EPA/600/R-96/084. U.S. Environmental Protection Agency, Washington, DC.
   http://www.epa.gov/quality 1 /qa docs.html

U.S. EPA, 2000d. Institutional Controls: A Site Manager's Guide  to Identifying, Evaluating and
   Selecting Institutional Controls at Superfund and RCRA Corrective Action Cleanups, OSWER
   9355.0-74FS-P.  U.S. Environmental Protection Agency, Office of Solid Waste and Emergency
   Response, Washington, DC.
   http://www.epa. gov/superfund/action/ic/guide. htm

U.S. EPA, 2000e. Proceedings of the Ground-Water and Surface-Water Interactions Workshop,
   EPA/542/R-00/007. U.S. Environmental Protection Agency, Washington, DC.
   http://www.epa.gov/tio/pubitech.htmtfgwc

                                            69

-------
U.S. EPA, 2001a.  Using the Triad Approach to Improve the Cost-Effectiveness of Hazardous
    Waste Site Cleanups. Current Perspectives in Site Remediation and Monitoring.  EPA 542-R-01-
    016, October 2001.  U.S. Environmental Protection Agency, Office of Solid Waste and
    Emergency Response (5102G), Washington, DC.
    http://www.epa.gov/tio/triad/

U.S. EPA, 2001b. Current Perspectives in Site Remediation and Monitoring: Clarifying DQO
    Terminology Usage To Support Modernization of Site Cleanup Practice, EPA/542/R-01/014.
    U.S. Environmental Protection Agency, Washington, DC.
    http://www.clu-in.org/tiopersp/tfcurr pers

U.S. EPA, 2002a. Site Characterization Library, Volume 1  (Release Z5),EPA/600/C-02/002. U.S.
    Environmental Protection Agency, Office of Research and Development, National Exposure
    Research Laboratory, Las Vegas, NV.

U.S. EPA, 2002b. Calculation and Use of First-Order Rate Constants for Monitored Natural
    Attenuation, EPA/540/S-02/500. U.S. Environmental Protection Agency, Office of Research
    and Development, National Risk Management Research Laboratory, Cincinnati, OH.
    http: //www. epa. gov/ada/pubs/issue .html

U.S. EPA, 2002c. Handbook of Groundwater Protection and Cleanup Policies for RCRA, EPA/5 30/R-
    01/015.  U.S. Environmental Protection Agency, Office of Solid Waste and Emergency
    Response, Washington, DC.
    http ://www. epa. gov/epaoswer/hazwaste/ca/resource/guidance/gw/gwhandbk/gwhndbk. htm

U.S. EPA, 2002d. Draft Guidance for Evaluating the Vapor Intrusion to Indoor Air Pathway from
    Groundwater and Soils (Subsurface Vapor Intrusion Guidance), RCRA-2002-033. U.S.
    Environmental Protection Agency,  Office of Solid Waste and Emergency Response,
    Washington, DC.
    http ://www. epa. gov/correctiveaction/eis/vapor. htm

U.S. EPA, 2003a. MTBE (methyl tertiary-butyl ether) and Underground Storage Tanks. U.S.
    Environmental Protection Agency.  OSWER/OUST. Washington, DC .
    [Online] Available:  http://www.epa.gov/swerustl/mtbe/index.htm (September  30, 2003).

U.S. EPA, 2003b. Using Dynamic Field Activities for On-Site Decision Making: A Guide for
    Project Managers. OSWER No.  9200.1-40,  EPA/540/R-03/002, May 2003. U.S. Environ-
    mental Protection Agency, Office of Solid Waste  and Emergency Response, Washington,  DC.
    http ://www. epa. gov/superfund/programs/df a/guidoc. htm

U.S. EPA,  2003c.  Field Analytic  Technologies Encyclopedia (FATE).  U.S. EPA  Technology
    Innovation Office, Washington,  DC.
    [Online] Available:  http://fate.clu-in.org/  (September 30,  2003).

Weaver, J.W, J.T Wilson, and D.H.  Kampbell,  1996. Extraction of degradation rate constants
    from  the St. Joseph, Michigan, trichloroethene site. In: Symposium on Natural Attenuation of
    Chlorinated Organics in Ground  Water, Hyatt Regency Dallas, Dallas, TX, September 11-13,
    1996, EPA/540/R-96/509. U.S.  Environmental Protection Agency, Office of Research and
    Development, Washington, DC.
    http ://www. epa. gov/ordntrnt/ORD/WebPubs/natural/natural .pdf


                                            70

-------
Westall, J.C. 2000. Geochemical equilibrium and the interpretation of Eh. In: Workshop on
   Monitoring Oxidation-Reduction Processes for Ground-Water Restoration, Workshop Summary,
   Dallas, Texas, April 25-27, 2000, EPA/600/R-02/002. U.S. Environmental Protection Agency,
   Office of Research and Development, National Risk Management Research Laboratory,
   Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa 600 r02  002.pdf

Wiedemeier, T.H., H.S. Rifai,  C.J. Newell, and J.T. Wilson, 1999. Natural Attenuation of Fuels
   and Chlorinated Solvents in the Subsurface. John Wiley & Sons, New York.

Wiedemeier, T.H., P.E. Haas, 2002. Designing monitoring programs to effectively evaluate the
   performance of natural attenuation. Ground Water Monitoring & Remediation, 22(3):124-135.

Wilson, J. 2000.  Current state of practice for evaluation of oxidation reduction processes
   important to the biological and chemical destruction of chlorinated organic compounds in
   ground water. In: Workshop on Monitoring Oxidation-Reduction Processes for Ground-Water
   Restoration, Workshop Summary, Dallas, Texas, April 25-27, 2000, EPA/600/R-02/002. U.S.
   Environmental Protection Agency, Office of Research and Development, National Risk
   Management Research Laboratory, Cincinnati, OH.
   http://www.epa.gov/ada/download/reports/epa 600 r02  002.pdf

Winter, T.C., 2000. Interaction of ground water and surface water.  In: Proceedings of the Ground-
   Water and Surface-Water Interactions Workshop, EPA/542/R-00/007.  U.S. Environmental
   Protection Agency, Washington, DC.
   http: 7/www. epa. gov/tio/tsp/issue. htm
                                            71

-------
72

-------
                                      GLOSSARY
abiotic not relating to living things, not alive.

adsorption process by which molecules collect on and adhere to the surface of an adsorbent solid
       because of chemical and/or physical forces.

aerobic living, active, or occurring only in the presence of oxygen.

alkalinity the capacity to accept protons (acid) while maintaining the pH above a predetermined
       value. Ground-water alkalinity may be increased as carbon dioxide emitted during
       biodegradation causes bicarbonate production.

alluvial relating to material deposited by moving water.

anaerobic living, active, or occurring only in the absence of oxygen.

anthropogenic man-made.

attenuation a lessening in concentration or mass.

biodegradable refers to a material or compound that can be broken down by natural processes of
       living things such as metabolization by microorganisms.

biodegradation act of breaking down material by natural processes of living things such as
       metabolization by microorganisms.

contaminants of concern those chemicals identified during site investigations that are required
       to be addressed by the response action proposed in the remedy decision documents.

daughter product degradation product of a compound. Vinyl chloride is a daughter product of
       the reductive  dechlorination of dichloroethene.

diffusion process by which ionic and molecular species move from a region of higher
       concentration to a region of lower concentration.

dispersion phenomenon by which a solute in flowing ground water mixes with uncontaminated
       water, becoming reduced in concentration. Dispersion is due both to differences in -water
       velocity at the pore level and differences in the rate at -which -water moves through
       different  strata. Also refers to statistical measures of how -widely a set of data vary.

dispersivity property that quantifies dispersion in a medium.
                                            73

-------
electron acceptor a compound capable of accepting electrons during oxidation-reduction (redox)
       reactions.  Microorganisms obtain energy by transferring electrons from electron donors
       such as organic compounds (or sometimes reduced inorganic compounds such as sulfide)
       to an electron acceptor. Electron acceptors are compounds that are relatively oxidized and
       include oxygen, nitrate, iron (III), manganese (IV), sulfate, carbon dioxide, or in some
       cases the chlorinated aliphatic hydrocarbons such as tetrachloroethene, trichloroethene,
       dichloroethene, and vinyl chloride.

electron donor a compound capable of supplying (giving up) electrons during oxidation-
       reduction reactions. Microorganisms obtain energy by transferring electrons from electron
       donors such as organic compounds (or sometimes reduced inorganic compounds such as
       sulfide) to an electron acceptor. Electron donors are compounds that are relatively
       reduced and include fuel hydrocarbons and native organic carbon.

hydraulic conductivity relative ability of soil, sediment, or rock to transmit water; a coefficient
       of proportionality describing the rate at -which -water can move through a permeable
       medium.

hydraulic gradient the  change in total hydraulic head -with a change in distance in a given
       direction.

hydraulic head sum of the elevation head, the pressure head, and the velocity head at a given
       point in an aquifer; also referred to as the total head.

hydrostratigraphic unit in -which the geologic materials have similar hydrologic properties.

in situ refers to a technology or treatment process that can be carried out in place at the site of
       contamination.

metabolic byproduct a product of the reaction bet-ween an electron donor and an electron
       acceptor. Metabolic byproducts can include volatile fatty acids, daughter products of
       chlorinated aliphatic hydrocarbons, methane, chloride, carbon dioxide, and -water.

oxidation chemical process that results in a net loss of electrons in an element or compound.

porosity the ratio of void volume to total volume of a rock or sediment.

pump and treat treatment method in -which contaminated -water is pumped out of the
       contaminated aquifer, then treated.

reduction chemical process that results in a net gain of electrons to the reduced element or
       compound.

sorb to remove a substance from the aqueous phase to the solid phase.

sorption movement of a substance from the aqueous phase to the solid phase, -whether by
       adsorption, absorption, fixation or precipitation.  Sorption may be reversible or
       irreversible.

substrate substance(s) that provides growth and energy requirements for cells.


                                            74

-------
tentatively identified compounds (TICs) compounds appearing in a chemical analysis of
       environmental media from a site that have not been definitely identified.

transmissive zones subsurface units where ground-water flow is constrained or bounded by
       lower hydraulic conductivity materials (i.e., geologic impediments to flow)  or hydrologic
       barriers (e.g., hydraulic head boundaries). Transmissive zones may be bounded by
       components as obvious as the water table or units with low hydraulic conductivity, or by
       conditions as subtle as small differences in grain size, sorting, and packing of seemingly
       uniform sands.

uncertainty reduction of confidence in a conclusion -when more than one estimate is available
       for a variable.

vadose zone zone between the ground surface and the water table.
                                            75

-------
76

-------
               Appendix A

    VARIABILITY IN MEASURED PARAMETERS
AND THE EFFECTS ON PERFORMANCE MONITORING
                   A-l

-------
                                     Appendix A
                    VARIABILITY IN MEASURED PARAMETERS
              AND THE EFFECTS ON PERFORMANCE MONITORING

A.I    Introduction

Monitoring data often exhibit significant variability. The variability in the measured values
reflects both the inherent spatial and temporal variability in the subsurface as well as variability
introduced into the data by the measurement techniques. The variability in the data introduce
uncertainty into the decision-making process, increasing the probability of making incorrect
decisions. Therefore, it is important to be able to assess the nature of data variability to
understand its importance toward achieving remedial goals.  Identification and quantitative
assessment of data variability may often be more important in remedies that rely solely on natural
attenuation processes than for engineered remedies as neither contaminant migration nor
attenuation processes are actively controlled.

A.2    Spatial and Temporal Variability

A major source of variability in measured and interpreted contaminant distributions will
generally be  related to the placement and screening of monitoring points -within the spatially and
temporally heterogeneous subsurface. Current methods for defining subsurface contaminant
distributions rely on techniques that generally sample only a small volume  of material
immediately surrounding a well or borehole. Therefore,  the locations of sample collection often
greatly affect the interpretation of natural attenuation processes. For example,  a vertical series of
ground-water samples collected using short-screened (e.g., 0.5 ft to 1 ft) wells, such as those
installed by direct push technologies, may display differences in contaminant concentrations  of
several orders of magnitude over short vertical distances. In comparison, a nearby conventional
monitoring -well screened over the same  total interval as the series of short-screened wells may
yield samples with contaminant concentrations that are essentially a flow-weighted average of the
samples taken from the short-screened wells.

The issue of obtaining "representative" samples from heterogeneous media is complex and has
been termed sample "support" (U.S. EPA, 2000a). Proper sample support involves ensuring that
the sample is representative of the original matrix under investigation (U.S. EPA, 2001b).
Sample support and the scale at which remedy performance will be evaluated should be
considered during design of the performance monitoring system because it strongly affects
monitoring network density and the specification of appropriate data evaluation methods. For
instance,  as in the example given above, it may be possible to find a narrow zone in the subsurface
-where contaminant concentrations exceed allowable limits, but ground -water at the receptor
point (e.g., an irrigation -well) may not exceed limits because the receptor derives ground -water
from a much thicker zone in  the subsurface. In such a case, the sample support chosen can
determine the remedial decision, because a sample representative of the narrow highly
contaminated zone may lead  to a decision different than that derived from  consideration of a
sample representative of the thicker zone with a lower mean contaminant concentration.
                                           A-2

-------
Three-dimensional spatial variability in dissolved contaminant concentrations occurs due to such
complex factors as the:

       Non-uniform distribution of the original source materials for ground-water
       contamination,

       Heterogeneity in site geology, and

       Differences in the geochemical environment.

Source materials in various forms such as sorbed materials and NAPL may be non-uniformly
distributed throughout the vadose and saturated zone. This variability in source distribution
means that source contact with recharge water and ground water (and therefore dissolution of
contaminants into the ground water) is significantly different at varying locations in the aquifer.
Also, the geochemical environment varies throughout the aquifer due to the interactions of
microbial communities with the supply of terminal electron acceptors (TEA) and electron donors
carried by ground water from upgradient locations or released locally from the aquifer matrix.
Therefore, types and rates of contaminant degradation can vary throughout the plume.

After installation of the performance monitoring network, temporal variations in contaminant
distribution may become major influences on observed contaminant trends and data
interpretations.  Sources of this variability include:

       Changes in contaminant input to ground water,

       Changes in degradation processes,

       Short-term changes  due to seasonal factors {e.g., recharge or other physical factors such as
       temperature), and

       Long-term changes due to substrate or TEA  depletion, or changes in site hydrology.

Changes in contaminant input to ground water may result from source control actions, additional
releases to the environment, dissolution of vadose zone contamination during infiltration of
precipitation, and dissolution during increases in water table elevation.  Changes in site
hydrology, such as drought, extremely wet periods, and changes in ground-water extraction or
recharge due to different land uses may change ground-water flow rates, directions, and,
correspondingly, contaminant concentrations at established monitoring points.

A.3    Measurement Variability

Other sources of variability are related to the measurement processes for monitoring parameters
and variability in the interpolation and data interpretations. Numerous factors, such as:

       Differences in sample collection methods {e.g., changes in equipment, pumping volumes or
       rates, or pump intake location within a well  screen),

       Differences in sample preparation methods {e.g., sample filtration or lack thereof, sample
       preservation, and adherence  to holding times), and
                                            A-3

-------
       Analytical variability (e.g., incorrect instrument calibration, improper operating
       parameters).

may result in variability in measured contaminant concentrations. The variability from these
sources is often much less than the variability due to subsurface heterogeneity and temporal
variability. However, measurement variability can be significant, and caution should be exercised
in the use of samples taken or analyzed by different methods to assess plume changes (e.g,
calculating attenuation rates). Measurement variability is generally more readily definable and
controllable than spatial and temporal variability, using proper quality assurance/quality control
procedures such as those discussed in U.S. EPA (1998b). Additional discussion of the sources of
variability in subsurface investigations of contaminant migration and fate is provided by
Barcelona et al. (1989).

A.4    Variability in Data Interpretation

In addition to spatial and temporal variability, and variability in measured contaminant
concentrations, variability is also introduced during the data interpretation process due to the
non-uniqueness of possible explanations of measured site conditions.  That is, there may be a
variety of configurations of the  conceptual site model that could explain the available data {e.g.,
with regard to source factors, ground-water flow, and components of the attenuation rate). It may
be possible to reduce the number of alternative explanations by designing a focused sampling
program to fill the data gaps. The alternative explanations may be grouped by their probable
impact on attainment of remedial goals, and their likelihood of accurately describing site
conditions, in order to assess the desirability of performing  additional work to verify or eliminate
important alternative explanations.
                                             A-4

-------
4>EPA
United States
Environmental Protection Agency
(5102G)
Washington, DC 20460

Official Business
Penalty for Private Use $300
Office of Research and       Office of Superfund          OSWER 9355.4-25
Development            Remediation and            PB 2003 103270
Washington, DC 20460       Technology Innovation        ERA/540/R-03/004
                     Washington, DC 20460        September2003
EPA/540/R-03/004

-------