OSWER Directive 9360.4-16
                                                     EPA xxx/x-xx/xxx
                                                         PBxx-xxxxxx
                                                       December 1995
           SUPERFUND PROGRAM

REPRESENTATIVE SAMPLING GUIDANCE


    VOLUME 5: WATER AND SEDIMENT

    PART I -- Surface Water and Sediment

                    Interim Final
               Environmental Response Team

          Office of Emergency and Remedial Response
         Office of Solid Waste and Emergency Response

            U.S. Environmental Protection Agency
                 Washington, DC 20460

-------
                                               Notice
This document has been reviewed in accordance with U.S. Environmental Protection Agency policy and approved
for publication.

The  policies and  procedures established in this document are intended solely for the guidance of government
personnel for use in the Superfund Program. They are not intended, and cannot be relied upon, to create any rights,
substantive or procedural, enforceable by any party in litigation with the United States.  The Agency reserves the right
to act at variance with these policies and procedures and to change them at any time without public notice.

For more information on water sampling procedures, refer to the U. S. EPA Compendium ofERT Surface Water and
Sediment Sampling Procedures, OSWER Directive 9360.4-03. Topics covered in the compendium include sampling
equipment decontamination, surface water and sediment sampling procedures,  sampling equipment, and quality
assurance/quality control (QA/QC) methods.

Please note that the procedures in this document should be used only by individuals properly trained and certified
under a 40-hour hazardous waste site training course that meets the requirements set forth in 29 CFR 1910.120(e)(3).
This document should not be used to replace or supersede any information obtained in a 40-hour hazardous waste site
training course.

Questions,  comments, and recommendations are welcomed  regarding the Superfund Program Representative
Sampling Guidance, Volume 5 —  Water and Sediment, Part I - Surface Water and Sediment. Send remarks to:


                                       Mr. William  A. Coakley
                             Chairman, Representative Sampling Committee
                                           U.S. EPA-ERT
                                  Rantan Depot - Building 18, MS-101
                                       2890 Woodbridge Avenue
                                        Edison, NJ 08837-3679

-------
                                          Disclaimer

This document has been reviewed under U.S. Environmental Protection Agency policy and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

The following trade name is mentioned in this document:

Teflonฎ is a registered trademark of E.I. DuPont de Nemours and Company of Wilmington, Delaware.
                                                in

-------
                                            Contents

Notice	 ii

Disclaimer 	iii

List of Figures	  vii

List of Tables  	  vii

1.0 INTRODUCTION	 1

        1.1      Objective and Scope  	 1
        1.2      Characteristics of Surface Water and Sediment	 1
                1.2.1    Surface Water 	 1
                1.2.2    Sediment 	 3
        1.3      Representative Sampling	 4
        1.4      Conceptual Site Model	 4
        1.5      Representative Sampling Objectives  	 6
                1.5.1    Determine Hazard and Identify Contaminant	 6
                1.5.2    Establish Imminent or Substantial Threat	 6
                1.5.3    Determine Long-Term Threat  	 6
                1.5.4    Develop Containment and Control Strategies	 6
                1.5.5    Identify Available Treatment/Disposal Options	 6
                1.5.6    Verify  Treatment Goals or Clean-up Levels	 7
        1.6      Example Site	 7

2.0 SURFACE WATER AND SEDIMENT SAMPLING DESIGN  	 8

        2.1      Introduction	 8
        2.2      Sampling Plan	 8
                2.2.1    Historical Data Review 	 9
                2.2.2    Site Reconnaissance	 9
                2.2.3    Physiographic and Other Factors	 9
        2.3      Migration Pathways and Receptors  	  10
        2.4      Surface Water and Sediment Sample Types	  11
                2.4.1    Grab Sample 	  11
                2.4.2    Composite Sample	  11
        2.5      Surface Water and Sediment Characteristics	  12
        2.6      Sampling Considerations	  12
        2.7      Quality Assurance Considerations	  12
        2.8      Data Quality Objectives	  13
        2.9      Analytical Screening	  13
        2.10     Analytical Parameters 	  13
        2.11     Representative Sampling Approaches  	  14
                2.11.1   Judgmental Sampling	  14
                2.11.2   Random Sampling  	  14
                2.11.3   Systematic Grid Sampling	  15
                2.11.4   Systematic Random Sampling	  17
                2.11.5   Transect  Sampling	  17
                2.11.6   Stratified Sampling  	  18
                2.11.7   Three Dimensional (3D)	  18
        2.12     Sampling Locations and Numbers	  18
                                                 IV

-------
        2.13    Example Site	 19
               2.13.1   Background  	 19
               2.13.2   Historical Data Review and Site Reconnaissance	 19
               2.13.3   Identification of Migration Pathways, Transport Mechanisms, and Receptors	 21
               2.13.4   Sampling Objectives  	 21
               2.13.5   Selection of Sampling Approaches	 21
               2.13.6   Analytical Screening, Geophysical Techniques, and Sampling Locations  	 22
               2.13.7   Parameters for Analysis	 23

3.0 FIELD ANALYTICAL SCREENING AND SAMPLING EQUIPMENT	 24

        3.1     Introduction	 24
        3.2     Field Analytical Screening Equipment	 24
        3.3     Surface Water and Sediment Sampling Equipment and Selection  	 24
        3.4     Example Site	 25
               3.4.1    Selection of Analytical Screening Equipment  	 25
               3.4.2   Selection of Geophysical Equipment	 26
               3.4.3    Selection of Sampling Equipment  	 26

4.0 FIELD SAMPLE COLLECTION AND PREPARATION	 33

        4.1     Introduction	 33
        4.2     Sample Volume and Number	 33
        4.3     Surface Water Sample Collection 	 33
               4.3.1    Rivers, Streams, and Creeks	 35
               4.3.2   Lakes, Ponds, and Impoundments  	 35
               4.3.3    Estuaries  	 36
               4.3.4   Wetlands  	 36
        4.4     Sediment Sample Collection  	 37
        4.5     Sample Preparation  	 38
               4.5.1    Removing Extraneous Materials	 38
               4.5.2   Homogenizing	 38
               4.5.3    Splitting	 38
               4.5.4   Compositing	 38
               4.5.5    Final Preparation 	 39
        4.6     Example Site	 39
               4.6.1    Sampling  	 39
               4.6.2   Sample Preparation  	 40

5.0 QUALITY ASSURANCE/QUALITY CONTROL 	 41

        5.1     Introduction	 41
        5.2     Data Categories	 41
        5.3     Sources of Error	 41
               5.3.1    Sampling Design 	 41
               5.3.2   Sampling Methodology  	 42
               5.3.3    Sample Heterogeneity  	 42
               5.3.4   Analytical Procedures  	 42
        5.4     QA/QC Samples 	 42
               5.4.1    Field Replicate Samples	 42
               5.4.2   Collocated Samples	 43
               5.4.3    Background Samples	 43
               5.4.4   Rinsate Blank Samples	 43
               5.4.5    Field Blank Samples  	 43
               5.4.6   Trip Blank Samples	 43
               5.4.7    Performance Evaluation/Laboratory Control Samples  	 43

-------
                5.4.8   Matrix Spike/Matrix Spike Duplicate Samples	  44
                5.4.9   Laboratory Duplicate Samples	  44
        5.5      Evaluation of Analytical Error	  44
        5.6      Correlation Between Field Screening Results and Definitive Laboratory Results  	  44
        5.7      Example Site	  45
                5.7.1   Data Categories  	  45
                5.7.2   Sources of Error	  45
                5.7.3   Field QA/QC Samples	  45
                5.7.4   Laboratory QA/QC  	  46

6.0 DATA PRESENTATION AND ANALYSIS  	  47

        6.1      Introduction	  47
        6.2      Data Posting 	  47
        6.3      Cross-Section/Fence Diagrams 	  47
        6.4      Contour Mapping	  47
        6.5      Statistical Graphics 	  47
        6.6      Recommended Data Interpretation Methods  	  48
        6.7      Example Site	  48

Appendix A      EXAMPLE  OF FLOW DIAGRAM FOR CONCEPTUAL SITE MODEL	49

References 	  52
                                                 VI

-------
                                       List of Figures






1      Conceptual Site Model for Water Sampling  	  5




2      Random Sampling	  16




3      Systematic Grid Sampling	  16




4      Systematic Random Sampling	  17




5      Transect Sampling	  18




6      ABC Plating Site	  20




A-l    Migration Routes of a Gas Contaminant 	  49




A-2    Migration Routes of a Liquid Contaminant	  50




A-3    Migration Routes of a Solid Cotaminant	  51







                                        List of Tables






1      Surface Water and Sediment Field Analytical Screening Equipment	  27




2      Surface Water Sampling Equipment  	  29




3      Sediment Sampling Equipment	  31




4      Surface Water and Sediment Sample Method Location	  34
                                               vn

-------
                                    1.0  INTRODUCTION
1.1     OBJECTIVE AND SCOPE

This is Part I  of the fifth volume in a series of
guidance documents that assist Superfund Program
Site  Managers, On-Scene Coordinators  (OSCs),
Remedial Project Managers (RPMs), and other field
staff in obtaining representative samples at Superfund
sites.   In the Superfund Program, surface water or
sediment  sampling  can  be  conducted  during:
emergency responses, site assessments, and removal
or early action activities.  The representative sampling
principles discussed in this document are applicable
throughout the  Superfund Program.  This guidance
document presents basic and general principles for
sampling  approaches,  methods,  and  equipment.
Surface water or sediment sampling specifically for
remedial investigations and at remediation sites is not
discussed directly in this guidance. However, general
sampling decisions discussed in this document could
be applicable  to more  detailed surface water or
sediment sampling instances such as those performed
for remedial investigations. More samples  may be
collected or more specific analytical parameters may
be established  for remedial investigations,  but the
sampling objectives and methods remain similar to
those in this guidance.

The objective of representative sampling is to ensure
that  a sample  or  a group of samples  accurately
characterizes site conditions.  The  selected sample
must possess the same qualities or properties as the
location and source under investigation.  In order to
conduct representative  sampling,  proper sampling
techniques  and sample handling must be used to
maintain the integrity of the sample (preserving the
original form  and chemical  composition).  The
following chapters will help field personnel to assess
available information, select an appropriate sampling
approach, select and utilize field analytical screening
methods and sampling equipment, incorporate suitable
types and numbers of quality assurance/quality control
(QA/QC)  samples, and interpret  and present  site
analytical data.

As  the Superfund Program  has  developed,  the
emphasis has shifted beyond addressing  emergency
response and short-term  cleanups.  Each planned
response action must consider a variety of sampling
objectives, including identifying threat, determining
the need for long-term action, delineating sources of
contamination,  and confirming the achievement of
clean-up standards. Because many important  and
potentially costly decisions are based on the sampling
data, Site Managers and other field personnel must
characterize site conditions accurately.  Inappropriate
sample collection procedures can seriously bias the
representativeness of a sample as well as its analytical
results.  This document emphasizes the use of cost-
effective  field  analytical  screening techniques in
characterizing sites and aiding  in the  selection of
sampling locations.
1.2    CHARACTERISTICS OF
        SURFACE WATER AND
        SEDIMENT

1.2.1  Surface Water

Surface waters are water bodies that rest or flow over
land,  with a surface that is open to the atmosphere.
Surface water sampling consists of the collection of
representative samples from streams, lakes, rivers,
ponds,  creeks,  lagoons,  estuaries,  and   surface
impoundments. It includes samples collected from the
depth of the water as well as the surface.  Water
sampling typically involves sampling low to medium-
hazard wastes rather than the more concentrated high-
hazard wastes found in drums or  storage facilities.
(For high-hazard waste sampling, see  U.S.  EPA
Superfund   Program  Representative  Sampling
Guidance,  Volume  4  — Waste, OSWER Directive
9360.4-14,  1995.)  Surface water sampling requires
recognition of special properties and precautions.  The
following   aspects  of surface water   should  be
considered  in developing a representative sampling
design:

•       Stratification - Stratification in a water body
        can be thermally or  chemically induced.
        The  temperature  profile  is   often  the
        controlling force in the circulation of a water
        body.  The warm, less dense surface water
        (epilimnion) and the deeper cold water mass
        (hypolimnion) become stratified and create
        a thermocline region where the temperature
        changes rapidly with depth. The position of
        the thermocline varies in surface water
        bodies, but is typically less than 30 meters
        below  the surface.   Chemically-induced
        stratification  generally results  when  two
        levels  of a water body are separated by a
        steep salinity gradient. Still water bodies,

-------
such as lakes or reservoirs, have a much
greater tendency to  stratify than rivers or
streams.

The   epilimnion   is   exposed   to   the
atmosphere, whereas the hypolimnion is a
"confined"  stratum which  is vented  only
during seasonal overturn. These two zones
may thus have very different concentrations
of contaminants if:  1) the point of discharge
is to one zone only; 2) the contaminants are
volatile (thus  vented in the epilimnion but
possibly not in the hypolimnion); or 3) the
surface stratum is influenced by short-term
flushing due to inflow or outflow of shallow
streams.

Current - A  current is a large  portion of
water  moving  in  a  certain  direction.
Currents  can disturb  mixing  zones  and
reduce  the   chances  of   obtaining   a
representative sample.   For  example,  a
strong current may carry  and distribute
contamination over  a larger area or move
contaminated sediments further downstream,
complicating source identification.

Storm  events - Storms may turn over strata
in  a   water   body   and  reduce   the
representativeness of the sample. Increased
precipitation   or runoff may  increase  or
decrease  representative concentrations of
contaminants. For example, a  large storm
will dilute the concentration of contaminants
present in  a water  body, possibly below
detection  levels.   A  water body which
receives surface  runoff may show a higher
concentration of contaminants from the
ensuing runoff than are representative of the
water body under "normal" conditions.

Precipitation may affect a field screening
instrument's operation and accuracy through
water or humidity interference during field
use. This interference may affect screening
for  sample  locations or put samplers at risk
for health and safety concerns.

Time   of year -  Temperate water bodies
(usually lakes) experience  two periods of
overturn annually. As air temperature cools
in the  fall,  the epilimnion becomes cooler
and eventually isothermal conditions exist in
the  lake.   Overturning and total mixing
occurs. Similar overturning occurs again in
the  spring.   The chemical  composition of
lakes and  ponds  can  vary  considerably
depending  on the  season.  Variations can
occur during  periods  of increased  water
movement  due to temperature variations,
vegetation  decay, freezing and thawing, as
well as turnovers and inversions.

The time of year also influences rainy and
dry periods.  For most areas of the United
States, precipitation is greater in the late fall
through  spring with   an  accompanying
increase in volume and flow in surface water
bodies.  In  the spring, flowing water bodies
may swell from upland headwaters receiving
melting snow.  By summer, water bodies
may reduce in volume and velocity due to
drying or drought  conditions.  Some water
bodies, such as in intermittent streams, may
actually be dry during  certain times of the
year.

Circulation - Lakes shallower than 5 meters
are subject to mixing by  wind action.  Large-
scale water motion in  lakes may be either
wind  driven  or  the   result  of  density
gradients.  Sediment distribution  may  be
dominated  by  either or both types of water
motion.  If a water  body lacks stratification,
the entire lake may be  circulated or mixed
by wind-generated motion.

Velocity -  The speed  at which a surface
water body flows can affect the selection of
sampling locations, times, equipment, and
techniques.  Varying flow rates across or
within the  cross-section of the water body
can lead to non-homogeneous mixing of
contaminants, producing different phases,
increasing  the difficulty  of collecting  a
representative sample.

Turbidity   - Surface  water may  contain
suspended  particles of fine sediments or
solid contaminants.  These particles may
have a higher concentration of contaminants
adhering to  their surface  area  than  is
dissolved  in the   aqueous  portion of the
sample.  Turbidity will vary due to mixing
and settling in the water body.

Salinity - The natural salt concentration, or
salinity, of a water body may vary with its
proximity  to   the ocean   and  seasonal
gradients/stratification.    An   estuary  is
generally categorized as one of three types,
depending  upon fresh  water  inflow  and

-------
        mixing properties:   mixed  estuary, salt
        wedge estuary,  or oceanic estuary.   Tidal
        phases  of the water  body  must also be
        considered when sampling in saline waters.
        Salt concentration in the surface water may
        alter concentrations of contaminants due to
        chemical   reaction/transformation.    See
        Section 4.3.3 for additional details regarding
        estuaries.

1.2.2  Sediment

The  characteristics of sediment  are  dependent on
biological,  chemical,  and physical  phenomena.
Sediments consist  of particles derived from rocks or
biological materials  that  are either transported by
flowing water bodies (e.g., rivers, streams) or situated
beneath a static aqueous layer (e.g.,  lakes, ponds,
impoundments).  They include solids and sludges,
suspended or settled in the water.  Sediment types are
classified  by  particle   size,  mineralogy,  source
materials, and other potential variables.  Analysis of
sediment  can determine  whether concentrations of
specific contaminants exceed  established threshold
action levels or pose a risk to  public health or  the
environment. Media-specific variables that can affect
sediment sampling include:

•       Particle size (grain size) - Particle size can
        affect sampling  results  because  many
        pollutants adhere to particle  surfaces and
        therefore  occur in highest concentrations in
        small-grained material, where total surface
        area  is   greater, than  in  large-grained
        material.

•       Terrigenous  sediments  - Sediments may
        consist of  material eroded  from a land
        surface,  transported and deposited  in  the
        water body.  The origin of the sediment may
        influence the selection of analytical methods
        to determine soil physical characteristics and
        the presence  of chemical  contaminants.
        Terrigenous  sediments  may  exhibit  a
        historical  release not  associated with  the
        water body.     For  example,  chemical
        reactions from sediments which originated in
        mining areas may result in changes in iron,
        sulfate, and pH concentrations in the surface
        water.

•       Chemical   constituents   -    Chemical
        constituents associated with sediments may
        reflect  an  integration  of  chemical  and
        biological processes.  Sediments may reflect
the historical  input with  respect to time,
application  of chemicals, and  land use.
Bottom sediments, especially  fine-grained
particles, may act as a reservoir for adsorbed
heavy    metals   and   trace    organic
contaminants. Organic materials and metals
are more concentrated  and readily found in
sediment than in water and can be detected
in  sediment  analysis  if  they  have  not
degraded.  Ion exchange properties of certain
clays may affect concentrations of soluble
inorganic  ions by  removing them from
solution.  The clay-based sediments may
remain suspended in water and thereby not
provide a representative sediment sample.
The clay or other suspended sediments may
serve to transport  contaminants that have
adhered  to  the solid  particles,  to other
locations in the water body.

Depositional/erosional  areas  -  Sediment
accumulation  depends  on depth of water,
water flow rate, and bottom configuration as
well as temperature, rainfall,  and latitude.
Surface   water   velocity   and   flow
characteristics  can  directly   affect  the
distribution  of substrate  particle  size  and
organic content.  Contaminants are more
likely  to  be  concentrated in sediments
typified  by fine particle size  and  high
organic content. This  type of sediment is
most likely to be collected from depositional
zones.   In contrast, coarse sediments with
low  organic content,  found  in erosional
zones,   do   not   typically   concentrate
pollutants.    Identify  depositional   and
erosional  zones and  plan the sampling
design accordingly.

Anaerobic/aerobic   conditions   -   Deep
sediments subject  to  no disturbance  or
mixing may exhibit anaerobic conditions, or
lack      of     oxygen.           The
transformation/degradation  of  historical
deposits of contaminants will be affected by
either  anaerobic  or  aerobic  processes
depending on the  substrate  conditions.
Knowledge of whether  anaerobic or aerobic
conditions exist in the substrate at a specific
sampling  location  will  help to  identify
transformation   products  of  suspected
contaminants.      Detection   of   these
transformation  products   can  be used to
delineate the spread of  contamination.

-------
1.3     REPRESENTATIVE SAMPLING
1.4     CONCEPTUAL SITE MODEL
Representative surface water and sediment sampling
ensures that a sample or group of samples accurately
reflects the concentration of the contaminant(s)  of
concern  at  a  given time  and location.  Analytical
results  from  representative  samples  reflect the
variation  in pollutant presence and concentration
throughout a site.

In addition to the variables introduced due to the
characteristics of the sample media (as is discussed in
Section 1.2), this document concentrates on those that
are introduced in the  field.  These  latter variables
relate to the site-specific conditions, the sampling
design approach, and the techniques for collection and
preparation of samples.  The following variables
affect the representativeness of samples  and their
method of collection:

•       Media   variability  -  The   physical  and
        chemical characteristics of surface water and
        sediments, such as stratification, flow rate,
        particle  size,  and deposition.  (Section  1.2
        provides additional  specifics  of  media
        variability.)

•       Contaminant  concentration  variability   -
        Variations in the contaminant concentrations
        throughout  the site  and/or the  variables
        affecting the  release of site contaminants
        into surface water bodies away from the site.

•       Collection and preparation variability - Bias
        introduced   during   sample   collection,
        preparation, and transportation (for  analysis)
        can cause deviations in analytical results.

•       Analytical variability - The manner in which
        the sample   was  stored,  prepared,  and
        analyzed by the on-site or off-site laboratory
        can affect the analytical results.  Analytical
        variability can falsely lead to the conclusion
        that error is due to sample collection and
        handling procedures, although it cannot be
        corrected through representative sampling.
A conceptual site model is a useful tool for selecting
sampling  locations.   It helps ensure that sources,
migration pathways, and receptors throughout the site
are considered before sampling locations are chosen.
The  conceptual model assists the Site Manager in
evaluating the interaction of different site features.
Risk assessors use conceptual models to help plan for
risk assessment activities. A conceptual model may
be created as a site map (see Figure 1) or it may be
developed as a flow diagram which describes potential
migration of  contaminants to  site receptors  (see
Appendix A).

A conceptual site model follows contaminants from
their sources through migration  pathways (e.g., air,
surface water) to the assessment endpoints. Consider
the following when creating  a conceptual site model:

•       The state(s) of each contaminant  and its
        potential mobility

•       Site topographic features

•       Meteorological   conditions   (e.g.,   wind
        direction/speed,    average   precipitation,
        temperature, humidity)

•       Human/wildlife activities on or near the site

The conceptual site model in Figure 1 is an example
created for  this document.   The  model  assists  in
identifying the following site characteristics:

        Potential Sources:  Site (waste pile, lagoon);
        drum   dump;   sewage   plant   outfall;
        agricultural activities.

        Potential  Migration   Pathway (Surface
        Water): Runoff from the waste pile, lagoon,
        drum dump, or agricultural activities; outfall
        from the lagoon or  sewage plant.

        Potential Migration Routes:  Ingestion or
        direct contact with water in the river, lake,
        or aquifer (e.g., ingestion of drinking water,
        direct contact with  water at the  public
        beach).

-------

-------
        Potential Receptors of Concern'.

               Human Population
               (Residents/Workers/Trespassers):
               Ingestion  or direct contact  with
               contaminated water in the river,
               lake,  or aquifer (e.g., swimming,
               drinking).

               Biota:      Endangered/threatened
               species or  human  food  chain
               organisms suspected of ingesting or
               being  in  direct  contact  with
               contaminated water.

Preliminary  site  information  may  provide  the
identification of the contaminant(s) of concern and the
level(s) of the contamination.  Develop a sampling
plan based upon the receptors of concern and the
suspected  sources and pathways. The model may
assist in the selection of on-site and off-site sampling
locations.
1.5    REPRESENTATIVE SAMPLING
        OBJECTIVES

Representative sampling applies to all phases of a
Superfund  response  action.    The  following are
representative sampling objectives for surface water
and sediment:

•        Determine if the contaminant is hazardous
        by   identifying   its   composition   and
        characteristics.

•        Determine  if  there  is  an  imminent  or
        substantial threat to public health or welfare
        or to the environment.

•        Determine the need for long-term action.

•        Develop containment and control strategies.

•        Evaluate  appropriate  disposal/treatment
        options.

•        Verify treatment goals or clean-up levels.

1.5.1   Determine Hazard and Identify
        Contaminant

One of the first objectives during a response action at
a site is to determine  the presence, identity, and
potential threat of any  hazardous materials.   Field
screening  techniques  can  be used  (discussed in
Chapter 3) for rapid detection of contaminants. Upon
confirming  the presence  of hazardous materials,
sample and/or continue screening to identify their
compositions and determine their concentrations.

1.5.2  Establish Imminent  or
        Substantial Threat

Establishing threat to the public or the environment is
a primary objective during a response action.  The
data obtained from characterizing the contaminants
will help the Site Manager to determine whether an
imminent or substantial threat exists  and whether a
response action is necessary.  The type and degree of
threat determines the rate at which a response action
is taken.

1.5.3  Determine Long-Term Threat

Site conditions may establish  a long-term threat that
is not imminent or substantial. Characterization of the
contaminants can assist the Site Manager to determine
the need for long-term remediation  and  response.
Samples should be collected in a manner that enables
their use to support evaluating the site under the
Hazard Ranking System.

1.5.4  Develop Containment and
        Control Strategies

Once the chemical constituents and threat have been
determined,  many strategies  for  surface water and
sediment containment and  control  are  available.
Analytical data indicating the presence of chemical
hazards are not in themselves sufficient to select a
containment or control strategy.  Site reconnaissance
and historical site research provide information on site
conditions and the physical state of the contaminant
sources; containment and control strategies are largely
determined by this information. For example, harbor
booms, sorbent booms, sorbent pad strings, and filter
fences can prevent  spread of contamination in a
surface water body.

1.5.5  Identify Available Treatment/
        Disposal Options

The contaminants should be identified, quantified, and
compared to selected action levels. Where regulatory
action levels do not exist, site-specific clean-up levels
are determined by the Region (often in consultation
with the Agency for Toxic Substances and Disease
Registry  [ATSDR]) or by  State identification of

-------
Applicable or Relevant and Appropriate Requirements
(ARARs).  If action levels are exceeded, a series of
chemical  and physical  tests may be  required to
evaluate possible treatment and/or disposal options.

1.5.6  Verify Treatment Goals or
        Clean-up Levels

After treatment or disposal, representative sampling
results should either confirm that the response action
has met the site-specific treatment goals or clean-up
levels, or  indicate whether  further treatment or
response is necessary.

Sampling   to   verify   cleanup   requires  careful
coordination with demobilization  activities.  After
treatment of a water body, verification sampling can
begin by using field screening and on-site analysis.
Lab confirmation of the screening performed can help
ensure accuracy of subsequent screening to meet data
quality  objectives, as is  discussed in Section  5.2.
Sediment sampling can be conducted in phases before,
during,   and  after  cleanup.    While  verification
sampling on a  previously treated  area is being
conducted, treatment on other areas can begin.

1.6     EXAMPLE SITE

An example site, presented at the end of each chapter,
illustrates the development of a representative surface
water  and  sediment sampling plan  that  meets
Superfund Program objectives for early actions or
emergency responses.

-------
        2.0  SURFACE  WATER AND SEDIMENT SAMPLING  DESIGN
2.1     INTRODUCTION

There  is no universal  sampling method to fully
characterize surface water and sediment contaminants
because site characteristics and sampling situations
vary widely.  The sampling methods and equipment
must be suited to the specific sampling situation.  A
properly developed surface water/sediment sampling
design defines the sampling purpose, protects site
worker  health  and  safety,  effectively  utilizes
resources, and minimizes errors.  The sampling design
will vary according to the type and characteristics of
the water body (e.g., river, estuary) being sampled, as
well as  the characteristics of the  site.   When
developing a sampling design, consider:

•       Prior actions at the site (e.g., prior sampling
        practices, compliance inspections)
•       Properties   and   characteristics  of   the
        suspected contaminants
•       Site waste  sources (e.g., impoundments,
        waste piles, buried drums)
•       Topographic,  geologic,  hydrologic,  and
        meteorologic conditions of the site
•       Flora, fauna, and human populations in the
        area

Surface water and sediment samples can vary greatly
in composition, therefore making it difficult to obtain
truly representative samples. Variation is due to both
the location within the body of water being sampled
and the time of collection. The change in composition
of flowing waters such as streams or rivers is subject
to the variance in flow and depth. Real-time field
analytical   screening  techniques  can  be  helpful
throughout the response action.  The  results can be
used to modify the site sampling plan as the extent of
contamination becomes known.  Emergency response
sampling may require the use of a generic but media-
specific sampling plan.
2.2    SAMPLING PLAN

The purpose of sampling is to obtain  a  small but
representative portion  of the  medium  of interest.
Planning  to  ensure proper  sample collection  is
essential. Many site-specific factors are important in
the development of a good sampling plan, including:
data use and quality assurance objectives, sampling
objectives,  sampling  equipment   and  sampling
methodology,  sampling design, standard  operating
procedures  (SOPs),  field  analytical  screening,
analytical method selection, decontamination, sample
handling and shipment, and data validation. Each of
these components  should be  addressed in  one
document, a site-specific sampling plan, to be used
throughout the investigation. A sampling plan should
be referred to throughout the field activities, along
with the site-specific quality assurance/quality control
plan, and the health and safety plan.

The U.S. EPA Quality Assurance Sampling Plan for
Environmental Response software (QASPER), is a
database  that was  designed  to  assist  with  the
development of sampling plans for response actions.
QASPER is menu driven software that prompts the
user to  input background information and to select
prescribed parameters in  order to develop a  site-
specific sampling plan. It also gives the user access
to any previously  developed site-specific sampling
plans.

The  following procedures are recommended  for
developing  a  thorough  surface  water/sediment
sampling  plan.    Many  steps  can  be  performed
simultaneously, and the sequence is flexible.

•       Review the history of the  site and adjoining
        areas, including regulatory and reported spill
        history; note current and former locations of
        buildings, tanks, and process, storage, and
        disposal areas.

•       Perform a site reconnaissance; categorize
        physical/chemical properties and hazardous
        characteristics of materials involved.

•       Identify    topographic,   geologic,   and
        hydrologic  characteristics  of  the   site,
        including surface water, ground-water, and
        soil characteristics,  as well as potential
        migration pathways and receptors.

•       Determine  geographic  and  demographic
        information, including  population size and
        its proximity to the site (e.g., public health
        threats,  source of drinking water); identify
        threatened environments  (e.g.,  potentially
        contaminated wetlands or other sensitive
        ecosystems).

-------
•        Select sampling strategies considering field
         analytical    screening    and    statistical
         applications when appropriate.

•        Determine data quality and quality  assurance
         objectives  for field  analytical  screening,
         sampling,  and analysis;  as the  extent of
         contamination  becomes   quantified,  the
         sampling plan can be  modified to better
         achieve sampling objectives throughout the
         response action.

It is recognized that many  of these steps (described in
detail  below)  may  not be  applicable during an
emergency response because of the lack of advance
notice.  Emergency response sampling nevertheless
requires good documentation of sampling events.

2.2.1   Historical Data Review

The first step  in developing a  sampling  plan  is a
review  of historical site data, examining past and
present site operations  and disposal practices to
provide clues  on possible  contaminants and waste
sources.  Available sources of information include:
federal, state and local agencies and officials; federal,
state, and local agency  files  (e.g.,  site inspection
reports  and legal actions);  deed or title records;
current  and former facility employees; potentially
responsible parties  (PRPs); local  residents;  and
facility  records or  files.   Where  possible,  data
regarding  adjoining   properties  should   also  be
reviewed.

A  review of previous  sampling  information should
include sampling locations,  matrices, methods of
collection  and analysis,  and relevant contaminant
concentrations.  The reliability  and  usefulness of
existing analytical data  should be assessed,  including
data which are not substantiated by documentation or
QA/QC controls, but which may still illustrate general
site trends.

Information  that   describes   specific   chemical
processes, raw materials  used, products and wastes,
and waste storage and disposal practices should also
be collected.  Information on materials handled at a
site  may  provide  guidance  in the  selection of
analytical  parameters.   Review any  available site
maps,   facility  blueprints, and  historical  aerial
photographs  detailing past  and  present  storage,
process, and waste disposal locations.  Areas on a site
where particular processes occurred are good choices
as  sampling  locations.   U.S.   Geological  Survey
(USGS) topographic maps should be reviewed to
identify  possible contamination  overland  flow or
migration routes to surface water bodies.  County
property and tax records are also useful sources of
information about the site and its surroundings.

2.2.2   Site  Reconnaissance

A site reconnaissance can be conducted  at an earlier
date or on the same day immediately prior to sampling
activities.  It allows field  personnel  to  assess site
conditions, evaluate areas of potential contamination,
evaluate potential hazards associated with sampling,
and finalize a sampling plan.  Site reconnaissance
activities include:  observing and photographing the
site; noting site access routes and potential evacuation
routes; noting potential safety hazards; recording label
information from drums, tanks, or other containers;
mapping  effluent  pipes   or  other  point  source
discharges; mapping potential contaminant migration
routes such as streams and  irrigation ditches; noting
the condition of animals and/or vegetation; and noting
topographic and structural features (e.g., bridges or
piers).   Field  personnel  should use  appropriate
personal protective equipment when engaged in any
site activities.  A site reconnaissance for a surface
water body should focus  on  collecting as  much
information as possible on the physical and chemical
parameters of the water body.   National  Oceanic
Atmospheric Administration (NOAA) tide tables and
USGS  freshwater surface  water flow  records are
useful in determining the water body type. Common
measurement tools  and  means for a  surface water
body  reconnaissance  include:    boat, recording
fathometer,   salinometer,   and   conductivity   and
dissolved oxygen meters.

2.2.3   Physiographic and Other
         Factors

Other procedures, such as  determining data quality
and  QA/QC objectives, utilizing  field analytical
screening  techniques,   identifying   topographic,
geologic,   and  hydrologic   characteristics,   and
determining geographic and demographic information
are important steps in an overall sampling plan.  The
remainder of this chapter includes a brief discussion
of  many  of these procedures.   Field analytical
screening techniques and equipment are discussed in
greater  detail  in Chapter  3;   QA  objectives are
discussed in Chapter 5. For additional guidelines on
preparing a sampling plan, please refer to the U.S.
EPA Superfund Program Representative Sampling
Guidance,  Volume 1  — Soil, OSWER Directive
9360.4-10.

-------
2.3     MIGRATION PATHWAYS AND
         RECEPTORS

The historical data review and site reconnaissance are
the initial  steps  in  defining  the  source areas  of
contamination which could pose a threat to human
health and the environment.   Source areas could
include  waste  impoundments,  landfills,   spills,
contaminated soil, drums, tanks and other containers,
and other waste  management  areas.   Often these
source areas are not directly  located in  or  even
adjacent to the surface water body.  The contaminants
are transported or migrate to  the surface water or
sediments. This section addresses how to delineate
the spread of  contamination away from the source
areas.  Included are pollutant migration pathways and
the routes by which persons or the environment may
be exposed to the on-site chemical wastes.

The fate  of a contaminant is dictated by the source,
the characteristics of the  contaminant, and by the
physical  environment into which it is released.  By
defining   the   contaminants   and   the  physical
environment,  the  fate  of  contaminants  can  be
predicted and the migration pathway can be identified.
Knowing the migration pathway ensures that samples
are collected in the most appropriate location(s).

Migration pathways are routes by which contaminants
have  moved  or may  be  moved  away  from  a
contamination  source. Pollutant migration pathways
may   include  man-made   pathways,    surface
drainage/topography, vadose zone transport, and wind
dispersion. Human activity (such as foot or vehicular
traffic)   and   animal   activity   also   transport
contaminants away from a source area.  These five
transport mechanisms are described below.

•        Man-made pathways - A site located in an
         urban/suburban  setting has the following
         man-made  pathways  which   can  aid
         contaminant migration to  surface  water
         bodies:  storm and sanitary sewers, drainage
         culverts, sumps and sedimentation basins,
         French  drain systems,  and underground
         utility lines.  A facility might utilize effluent
         pipes or point source discharges.

•        Surface drainage/topography -  Contaminants
         can be adsorbed onto sediments, suspended
         independently  in the water column,  or
         dissolved in surface water  runoff.  The
         runoff, following natural topography, can be
         rapidly   carried  into  drainage  ditches,
         streams, rivers, ponds, lakes,  and wetlands.
        Historical   aerial  photographs  can   be
        invaluable  for  delineation of past surface
        drainage patterns.  A search of historical
        aerial photographs can be requested through
        the U.S. EPA Regional  Remote  Sensing
        Coordinator.  The U.S. Soil Conservation
        Service and local county planning offices are
        also excellent  sources of historical  aerial
        photographs.

•       Vadose  zone  transport  -  Vadose  zone
        transport  is the vertical  or horizontal
        movement  of  water and  of soluble and
        insoluble    contaminants    within    the
        unsaturated  zone  of  the  soil  profile.
        Contaminants from  a surface source or a
        leaking  underground storage  tank can
        percolate through the vadose zone and  be
        adsorbed  onto  subsurface  soil or  reach
        ground water.  Contaminants might migrate
        to surface water through a ground-water
        discharge area.

•       Wind dispersion - Contaminants deposited
        over or adsorbed onto soil may migrate from
        a  waste  site   as  airborne  particulates.
        Depending on the particle-size distribution
        and    associated  settling   rates,   these
        particulates may be deposited downwind or
        remain     suspended,     resulting     in
        contamination  of surface  soils,  surface
        waters,   and/or  exposure  to  nearby
        populations.

•       Human and animal  activity - Foot and
        vehicular traffic of facility workers, response
        personnel,   and  trespassers   can   move
        contaminants away from a source.  Animal
        burrowing, grazing, and migration can also
        contribute to contaminant migration.

Once the migration pathways have been determined,
identify all possible receptors (i.e., potentially affected
human and environmental receptors) along  these
pathways.  Human  receptors include on-site and
nearby residents, workers, and school children. Note
the attractiveness and accessibility of site wastes to
children and other nearby residents.  Environmental
receptors include edible aquatic species, federal- or
state-designated endangered or threatened  species,
habitats for these species,  wetlands, and other federal -
or state-designated wilderness, critical, and natural
areas.
                                                   10

-------
2.4    SURFACE WATER AND
        SEDIMENT SAMPLE  TYPES

Sampling  procedures should  be designed  to  be
consistent with sampling objectives.  The type of
sample   collected  may  depend   on   suspected
contaminant types and characteristics; projected extent
of water contamination; type of water body to be
sampled (e.g., stream, impoundment); target analytes;
and health and safety requirements.  The  following
section describes and gives examples of the  two types
of surface water and sediment samples.

2.4.1  Grab Sample

A grab sample is a discrete aliquot from one specific
sampling location at a specific point in time, and may
be   considered   representative  of  homogenous
conditions over a period of time and/or geographical
area. When obtaining grab samples from a water body
having stratified layers, sample each phase or stratum
separately; the separate aliquots are representative of
their respective stratum.  When sampling stratified
sources, determine as many properties as possible for
the contaminants  through historical data and  site
reconnaissance prior to sampling. Grab samples can
be collected for both surface water and sediments, and
are generally the preferred  method for  screening
investigations.  However,  because  the release of a
contaminant in  a  surface water body  is subject to
variance over time and distance, a grab sample may
not be a representative sample.

For  many  sampling  situations  grab  sampling
techniques are preferred over composite  sampling.
Grab  sampling  minimizes the amount of time and
expense required  for multiple  samples; minimizes
sampling personnel's exposure to potential hazardous
substances; reduces risks associated with compositing
unknowns;  and eliminates  physical and  chemical
changes that might occur due to compositing. Grab
sampling also documents contamination at a specific
point or location which can be  easily identified and
also  re-located  in later  investigations for possible
remedial or enforcement purposes.

2.4.2  Composite Sample

A composite  sample  is  a  non-discrete  sample
composed of two or more aliquots (of equal volume)
collected at various sampling points or times.  It can
represent portions collected at  various  locations,
various times, or a combination of both location and
time variables.   Composite  samples are  made by
combining grab samples collected at defined intervals.
There are four types of composite samples:  areal,
vertical, flow proportional, and time.   The areal
composite  is  composed   of  individual  aliquots
collected  over a defined  area.  It is made  up of
aliquots  (of  equal volume)  from   grab  samples
collected  in an  identical  manner (e.g., sediment
aliquots collected along a streambed).  A variation of
this  approach is the equal-width-increment  (EWI)
technique, in which equally-spaced vertical samples
are collected across a stream with the sampling device
passing through the water column at the same velocity
at each location.  This technique ensures that water
and suspended particles are collected  equally  across
the water body.   Another variation is the  equal-
discharge-increment (EDI) technique, which positions
the sampling locations across  the stream based on
incremental discharges  rather  than width  (i.e.,
locations in deeper or higher velocity areas  of the
stream's cross-section are spaced more  closely). This
technique measures total discharge of contaminants in
poorly mixed water bodies, but it requires knowledge
of the cross-sectional stream flow distribution.  Both
techniques, however, are very  time-consuming and
expensive to employ.  (Both techniques, as well as
other depth integration approaches, are discussed in
detail in ASTM standards, such as Standard D4411, in
the 1989 Annual Book of ASTM Standards - Volumes
11.01  and  11.02,   Water   and   Environmental
Technology.)

A vertical, also referred to  as a zonal, composite is
composed of individual aliquots collected at different
depths but along the same vertical  line. Like an areal
composite, it is made up of aliquots collected in an
identical manner.  A flow proportional composite is
a sample collected proportional to the flow rate during
the  compositing   period   by  either  a  time-
varying/constant  volume  or time-constant/varying
volume method. A time composite, or  chronological
sampling,  is composed of a  varying  number of
discrete aliquots collected  at  equal  time  intervals
during  the  compositing   period.     Both   flow
proportional and time composite samples are most
appropriate for sampling flowing water bodies.

By design, composite samples reflect an "average"
concentration within the composite  area,  flow, or
interval.     Compositing   is   appropriate  when
determining the   general   characteristics  or  the
representativeness of certain sources for treatment or
disposal.  Samples collected along the length of the
watercourse or at different times may reflect  differing
inputs  or  dilutions.   It   should  be  noted  that
compositing can  mask problems by diluting isolated
concentrations of  some  contaminants to  below
                                                  11

-------
detection levels.  When compositing samples from a
water body, note that  resulting concentrations are
representative   of  the  water  body's   average
concentration,  but not  of discrete areas within the
water body. Compositing is not recommended where
volatile compounds are  a concern.

When compositing either surface water or sediment
samples, specify in the  sampling plan the method of
selecting the aliquots that are composited and the
compositing factor.   The compositing factor is the
number of aliquots to be composited into one sample
(e.g., 3  to 1,  10 to  1).  Determine this factor by
evaluating detection limits for parameters of interest
and comparing them with the selected action level for
that parameter.

Compositing requires that each discrete aliquot be the
same in terms  of volume or weight  and that they be
thoroughly homogenized.   Because  compositing
dilutes high concentration aliquots, the  applicable
detection limits should be reduced accordingly. If the
composite value is to be compared to a selected action
level, then the action level must be divided by the
number of aliquots  that make up the composite in
order to determine the  appropriate detection limit.
The  detection level need not  be  reduced if the
composite  area  is  assumed  to be  homogenous in
concentration. Generally the number  of samples to be
taken for a composite depends upon the width, depth,
discharge,  and suspended sediments of the water
body. The greater number of individual aliquots, the
more  likely  the  composite   sample   is  truly
representative  of the overall  characteristics of the
water body.
2.5    SURFACE WATER
        AND SEDIMENT
        CHARACTERISTICS

The  physical and  chemical characteristics of the
surface water and sediments, including stratification,
current/flow   rate,    salinity,   particle    size,
depositional/erosional  areas,   and   degradation
conditions, among other factors, influence the number
and types of samples collected. These characteristics
may also assist in determining sampling approaches
and analytical parameters. Many of the characteristics
of surface water  and sediments  are  defined in
Section 1.2.
2.6    SAMPLING CONSIDERATIONS

Factors to consider when designing a sampling plan
include: hydrology, topography, water quality data,
and  water  quality  measurements  such  as  pH,
conductivity, temperature,  dissolved oxygen,  and
salinity.    Hydrology  and  morphometrics  (e.g.,
measurements of volume, depth) of the surface water
should be  determined  prior to sampling.   Before
sampling, identify the presence of phases or layers in
impoundments and lakes, flow patterns in streams,
and/or appropriate sample locations and depths.

Water  quality  data  should  be   collected   in
impoundments and non-flowing (static) water bodies
to determine if stratification is present.  Measurements
of dissolved oxygen,  pH, temperature, conductivity,
and oxidation-reduction potential can indicate if strata
exist  which  would  affect   analytical  results.
Measurements should  be collected  at  one-meter
intervals from the substrate to  the surface using an
appropriate instrument (e.g., Hydrolab or equivalent).
Knowing these variables assists in selecting locations
and depths and interpreting analytical data.
2.7    QUALITY ASSURANCE
        CONSIDERATIONS

Quality assurance components are defined as follows:

•       Precision - Measurement of variability in the
        data collection process

•       Accuracy (bias) - Measurement of bias in
        the  analytical  process;  the  term  "bias"
        throughout this document refers  to  the
        QA/QC accuracy measurement

•       Completeness  - Percentage  of  sampling
        measurements which are judged to be valid

•       Representativeness  -   Degree  to  which
        sample   data   accurately  and   precisely
        represent    the    characteristics    and
        concentrations of the site contaminants

•       Comparability - Evaluation of the similarity
        of conditions (e.g., sample depth, sample
        homogeneity) under which separate sets of
        data are produced
                                                  12

-------
To   ensure  that  the  analytical   samples  are
representative of site conditions, quality assurance
measures must be associated with each sampling and
analysis event. The sampling plan must specify these
QA measures, which include, but are not limited to,
sample collection, laboratory   standard operating
procedures (SOPs),  sample container preparation,
equipment decontamination, field  blanks, replicate
samples, performance  evaluation  samples,  sample
preservation  and  handling,  and  chain-of-custody
requirements    (see    Chapter     5,     Quality
Assurance/Quality Control).
2.8     DATA QUALITY OBJECTIVES

Data  quality objectives (DQOs)  state the level of
uncertainty  that is acceptable for data collection
activities and define  the data quality  necessary to
make certain decisions.  When establishing DQOs for
a particular project, consider:

•        Decision(s) to be made or question(s) to be
         answered by  the data
•        Why analytical data are needed and how the
         results will be used
•        Time  and resource constraints  on data
         collection
•        Descriptions of the analytical data to be
         collected
•        Applicable  model  or data  interpretation
         method used to arrive at a conclusion
•        Detection limits for analytes of concern
•        Sampling and analytical error

In  addition  to  these considerations, the  quality
assurance components of precision, accuracy (bias),
completeness, representativeness,  and comparability
should also be considered.   These  components are
defined in Section 2.7 and  are discussed  in further
detail  in Chapter  5,  Quality  Assurance/Quality
Control.
2.9    ANALYTICAL SCREENING

There are two primary types of analytical data that can
be generated during a response action: field analytical
screening data and laboratory analytical data.  Field
analytical   screening  instruments and  techniques
provide real-time or direct (or colorimetric) reading
capabilities.  They include: flame ionization detectors
(FIDs), photoionization detectors (PIDs), colorimetric
tubes,  portable  X-ray  fluorescence  (XRF)  units,
portable   gas   chromatography   (GC)   units,
immunoassay tests, and hazard categorization (hazcat)
kits.  These screening methods can  assist with the
selection of sample locations and depths or samples to
be  sent for laboratory  analysis by  narrowing the
possible groups or  classes of  chemicals.  They are
effective and economical for gathering large amounts
of site data.  Once an area has been characterized
using field screening techniques, a subset of samples
can be sent for laboratory analysis to substantiate the
screening results.

Under a limited sampling budget, analytical screening
(with laboratory  confirmation) will generally result in
more analytical data from a site than will sampling for
rigorous laboratory analysis alone.  To minimize the
potential  for   false   negatives  (not   detecting
contamination),  use  only  those  field  analytical
screening methods  which  provide detection limits
below applicable action levels.  If these methods are
not available, field  analytical  screening can still be
useful for detecting grossly contaminated areas, as
well as for health and safety  determination.  Field
analytical screening techniques  to support  surface
water and sediment sampling are discussed in greater
detail in Chapter 3.

Geophysical techniques (e.g., ground penetrating radar
[GPR], magnetometry, electromagnetic conductivity
[EM]) may be  utilized during a response action to
locate potential buried or disturbed waste source
areas.   These  techniques  are generally  not used
directly   with  representative  surface  water  and
sediment  sampling.   Please  refer  to U.S.  EPA
Superfund  Program   Representative  Sampling
Guidance,  Volume 1  —  Soil, OSWER Directive
9360.4-10, for a discussion of geophysical techniques.
2.10   ANALYTICAL PARAMETERS

Designing a representative surface water and sediment
sampling plan includes selecting analytical parameters
and methods.  Use data collected during the historical
data review (e.g., past site operations and processes,
materials stored on site, effluent discharges) to select
appropriate analytical parameters and methods. If the
historical  data reveal little information about the
possible types of contaminants on site, use applicable
field analytical  screening  methods to  narrow the
parameters for analysis by ruling out the presence of
high concentrations of certain contaminants.  If the
screening results are inconclusive, send a subset of
samples from the areas of concern for a full chemical
characterization by an  off-site laboratory.   These
analyses can identify all contaminants of concern
                                                   13

-------
(even at low detection levels).  Methods often used
for characterization include gas chromatography/mass
spectrometry (GC/MS) for tentatively  identified
compounds  (TICs) in the volatile and semivolatile
organic fractions,  infrared  spectroscopy (IR)  for
organic compounds, and inductively coupled plasma
(ICP) for inorganic substances.

After characterization, future sampling and analysis
efforts can focus on substances identified above the
action level.  This will result in significant  cost
savings over a full chemical characterization of each
sample. Utilize U.S. EPA-approved methodologies
and  sample preparation,  where  possible,  for all
requested off-site laboratory analyses. Knowledge of
the  analytical methodology  and requirements is
helpful when selecting sampling devices.  Refer to the
American  Public  Health   Association  Standard
Methods  for the  Examination of  Water  and
Wastewater, Seventeenth Edition, 1989, for detailed
descriptions of analytical procedures/methodologies.
2.11    REPRESENTATIVE SAMPLING
         APPROACHES

Representative    sampling    approaches    include
judgmental, random,  systematic  grid,  systematic
random,  transect, stratified,  and three-dimensional
(3D) sampling.  The random and systematic random
approaches  are not very practicable for sampling
water systems. When these two approaches are used,
however,  they  are  more appropriate to sediment
samples  than  to  surface  water.   The remaining
approaches may be applied to both surface water and
sediment   sampling  plans.      Selection   of  a
representative sampling approach must also consider
the practicability of reaching sediments and obtaining
a sample from a specific location, particularly difficult
in surface waters. A representative sampling plan
may use one or a combination of the approaches, each
of which is described below.

2.11.1  Judgmental Sampling

Judgmental sampling  is the  biased  selection of
sampling locations based on historical information,
visual  inspection,  and  professional  judgment.
Judgmental  sample  collection is  most appropriate
when knowledge  of the contaminant or its origin is
available or when sampling non-static systems, such
as flowing bodies of  water.   Judgmental sampling
includes no randomization in the sampling strategy,
precluding statistical interpretation of the sampling
results.  Criteria for selecting the  sampling location
depend  on  the  sampling  objectives  and  best
professional judgment. Judgmental sampling does not
necessitate sampling from the  middle of the water
body, but  may  consider factors  such  as  source
locations, tributaries, or depositional areas for more
representative samples.  Judgmental sampling also
enables the investigator to select sampling locations
with the fewest physical barriers impeding  sample
collection  (e.g.,  docks,  piers,  stumps, dry  stream
beds).  For surface water and sediment sampling for
site assessments,  emergency responses, and some
early actions, judgmental sampling is often utilized.

Judgmental sampling allows no  statistical analysis of
error or bias. It is not always representative of site
conditions,  and  tends to document  "worst-case"
scenarios.  Judgmental sampling meets the objective
to qualify  hazardous substances on site, but not to
quantify them. The judgmental approach is best used
as a screening investigation to be  followed with a
statistical  approach  when  determining extent  of
contamination or action  alternatives.   Judgmental
approaches  should  be  incorporated into sampling
designs  for  remedial investigations and  large-scale
early and long-term response actions.

2.11.2  Random Sampling

Random sampling, also referred to as simple random
sampling, is the arbitrary collection of samples having
like contaminants within defined boundaries of the
area of  concern (see  Figure 2).    Obtaining  a
representative sample depends on random  chance
probabilities.  Random sampling is useful when there
are many sampling locations available and no criteria
for selecting one location over  another.  Choose
random sampling locations using a random selection
procedure (e.g., a random number table).  (Refer to
Ford and  Turina, July  1984, for an  example of a
random  number table.)  The arbitrary selection of
sampling points ensures that each sampling point is
selected independently from all other points, so that
all locations within the area of concern have an equal
chance of being sampled. Randomization is necessary
in order to make probability or confidence statements
about the sampling results.  The key to interpreting
these statements is  the assumption that  the site or
water body is homogeneous  with respect  to the
parameters being sampled. The higher the degree of
heterogeneity, the less the random sampling approach
will adequately characterize true  conditions. Random
sampling is useful for sites with little background
information available or for sites where  obvious
contaminated areas do not exist or are not evident.
Random sampling is not recommended  in flowing
                                                  14

-------
water bodies and is only practicable for sediment bed
sampling in non-flowing (static) water bodies.

2.11.3  Systematic Grid Sampling

Systematic grid sampling involves subdividing the
area of concern by using a square or  triangular grid
and collecting samples from the nodes (intersections
of the grid lines) (see Figure 3).  Select the origin and
direction for placement of the grid using an initial
random point.  From that point, construct a coordinate
axis and grid over the area of concern.  Generally, the
more  samples collected (and the smaller the  grid
spacing), the more reproducible and representative the
results. Shorter distances between sampling locations
improve representativeness. Systematic grid sampling
can be used to characterize non-flowing (static) water
bodies and their sediment beds.
                                                  15

-------
               Figure 2:  Random Sampling
    200-
    150--
  k;
  ffi 1004
     50-
                                            r STATIC WATER
                                            i BODY BOUNDARY
                  -\	h
-\	h
      0     50   100   150    200   250   300   350   400

                             FEET
         KEY

X SELECTED SAMPLE LOCATION
           Figure 3: Systematic Grid Sampling
200-
150-
ti
K 100-
50-
0

c
* *
N* '
^^
f
*
f

•^^
50 100
KEY
X SELECTED SAMPLE LOCATION


)
\

f *
f ^
t 5
— ^ _
f ซ
/ ซ
f *


f \
f *

• A
/STATIC WATER
I BODY BOUNDARY
* ^^^
• •">
'S

II III
150 200 250 300 350 400
FEET
                            16

-------
2.11.4  Systematic Random Sampling

Systematic random sampling is a flexible design for
estimating the average pollutant concentration within
grid cells (see Figure  4).  Subdivide the area  of
concern  using a  square  or  triangular  grid  (as
mentioned above) then  collect samples from within
each grid cell using random selection procedures.
Systematic random sampling allows for the isolation
of cells that may require  additional sampling and
analysis. Like systematic grid sampling, systematic
random sampling can be used to characterize sediment
in an impoundment  or non-flowing (static)  water
body; it is not recommended or practicable for surface
water in any system.
                           Figure 4:  Systematic Random Sampling
200-
150-
ti
ffi 100-
50-

/X


X
V X

„ — ' —
X
X
X
^d
X
X
X
X

X
X
X
X


X
X
X
X


/*v^
/ ^
/ STATIC WATER
I BODY BOUNDARY
X \
X ^y

                               50    100    150   200    250    300   350    400
                                                   FEET
                           KEY
                 X SELECTED SAMPLE LOCATION
2.11.5  Transect Sampling

Transect sampling involves establishing one or more
transect lines across a surface (see Figure 5). Collect
samples at regular intervals along the transect lines at
the surface and/or at one or more given depths.  The
length of the transect line and the number of samples
to be  collected  determine  the spacing between
sampling points along the transect. Transect sampling
can best be accomplished when surface water bodies
are small in size and the sampling locations within the
transect grid boundaries are easily accessible.  This is
not the most desirable method in large lakes and
ponds, or inaccessible areas  where surface  water
samples can be  obtained only by boat.  Multiple
transect lines may be parallel or non-parallel to one
another, or may intersect.  If the lines are parallel, the
sampling objective  is similar to systematic grid
sampling. The primary benefit of transect sampling is
the ease  of establishing and  relocating individual
transect lines.   Transect sampling is applicable to
characterizing    water   flow   and   contaminant
characteristics   and   contaminant   depositional
characteristics  in sediments, such as distinguishing
erosional versus depositional zones.
                                                  17

-------
                                  Figure 5:  Transect Sampling
                      20-
                                                                       35
                       40
                           KEY
                X SELECTED SAMPLE LOCATION
2.11.6  Stratified Sampling

Stratified sampling involves dividing the area to be
sampled into mutually exclusive strata or areas where
different sampling strategies may be employed in each
stratum.   Strata are  chosen either based on areas
where separate clean-up decisions need to be made or
where variable  strata contamination constituents or
levels are expected.  Where access is not a problem,
stratified sampling is more appropriate for collecting
representative sediment samples than surface water
samples.  Prior knowledge of stratification is required
in order for this method to be most effective.

2.11.7  Three Dimensional (3D)

Three-dimensional  (3D)  sampling is similar to
systematic sampling. First, the water body is divided
along three axes (x,  y, z),  as opposed to the  two
horizontal axes in grid sampling. Then, a systematic
approach (random or grid) is used to select sampling
locations across the surface and at depth.   Three-
dimensional sampling is useful in static water bodies
which exhibit distinct strata with depth but for which
few  data  are  available  on  contaminants  and/or
contaminant locations.
2.12   SAMPLING LOCATIONS AND
        NUMBERS

Selection of a surface water or sediment sampling
location is based on many factors, including sampling
objectives, surface water use, point source discharges,
nonpoint source discharges, mixing zones, tributaries,
changes in  stream characteristics,  stream  depth,
turbulence, presence of structures (e.g., dams, weirs),
and accessibility  to the sampling location.  Tidal
movement must  also be considered when selecting
sampling locations in tidal zones.  Seasonal salinity
ranges should be considered in estuaries.

The   sampling   objective  can  determine  which
characteristics of the surface water body warrant more
attention.  For example,  when investigating a water
body that serves as a source of water supply, factors
such as accessibility, flow, and velocity  are not as
critical  as  they  would   be  when  determining
contaminant impact on wetlands or sediments. This
is  because  water supply intakes draw water from
across the water body, also drawing in contaminants,
while contaminants settle  into  wetlands by natural
flow or mixing.  When multiple sampling locations
need to be investigated to determine pollution patterns
or to obtain data for mathematical modeling purposes,
several related factors may need to  be considered.
(See A Practical Guide to Water Quality Studies of
                                                  18

-------
Streams, F.W. Kittrells, for additional guidelines on
extensive or complicated sampling designs.)

The sampling objective will also influence the number
of  samples  collected.    When  determining the
presence/absence of a contaminant, few samples are
required.  More samples are needed if the objective is
to identify  the  characteristic concentrations of a
contaminant  or  the  extent  of  contamination.
Judgmental and statistical sampling techniques can be
used together to thoroughly address an area.  Some
samples may be obtained from locations considered
potentially affected areas by a judgmental approach
(e.g., sediments downstream  of a discharge outfall
pipe). For areas less likely to be affected or with little
available  historic information,  a random  or grid
approach may be used to adequately assess the entire
water body or site.

To determine whether a water body has been affected
by site contaminants, two sample sets are generally
required:  one surface water and sediment sample each
from the point (or slightly downstream) where on-site
contaminants are suspected to have entered the water
body  (also referred to as the probable point of entry
[PPE]), and another  surface water  and sediment
sample set from an upstream, unaffected background
location.  If multiple sources or contaminants from
other sites upstream of the PPE are suspected in the
water  body, additional sample  locations will be
needed  downstream  of those  alternate  sources,
upstream of the PPE.

Where the  sampling objective  is to delineate the
extent of sediment contamination for response action
alternatives,  a   greater number  of samples and
sampling locations will be required.  In this situation,
a systematic approach will be needed (e.g., transect or
systematic  grid)    to   accurately    "map"  the
contamination.  The exact number of samples required
will be determined by the analytical parameters and
the size of the line or grid and their intersects.
2.13    EXAMPLE  SITE

2.13.1  Background

The  ABC  Plating Site  is  located in northeastern
Pennsylvania approximately 1.5  miles north of the
town of Jonesville. Figure 6 provides a layout sketch
of the site and surrounding area.  The  site  covers
approximately 4 acres and was operated as a multi-
purpose specialty electroplating facility from 1947 to
1982.  During its years of operation, the company
plated automobile and airplane parts with chromium,
nickel, and copper.  Cyanide solutions were used in
the  plating  process.    ABC   Plating  deposited
electroplating wastes into two unlined shallow surface
settling lagoons in the northwest portion of the site.
Surface drainage from this area then entered a nearby
stream.

Pennsylvania Department of Environmental Resources
(PADER) personnel cited the owner/operator for the
operation of an  unpermitted treatment system and
ordered the owner to submit a remediation plan for
state approval. Before PADER could follow up on the
order, the lagoons were partially backfilled with the
wastes  in place.   The process  building was  later
destroyed by  a fire of suspicious origin.  The owner
abandoned the facility  and could not be located by
enforcement  and legal  authorities.   Several vats,
drums,  and  containers were  left unsecured and
exposed to the elements.  The state contacted EPA for
an  assessment of the site for a possible federally
funded   response  action;  an  EPA   On-Scene
Coordinator (OSC) was assigned to the task.

2.13.2 Historical Data  Review and
         Site Reconnaissance

The EPA OSC reviewed the PADER site file.  In 1974
the owner was cited for violating the Clean Streams
Act and for storing and treating industrial waste
without a permit.  The owner was  ordered to file a site
closure  plan and  to remediate the settling lagoons.
The owner, however, continued operations and was
then ordered to begin remediation in 90 days or be
issued a cease and desist order.  Soon after, a follow-
up  inspection revealed that the lagoons had been
backfilled without removing the  waste.

The OSC and a sampling contractor (Team) arrived on
site to  interview local and  county  officials, fire
department officers, neighboring residents (including
a   former   facility   employee),   and  PADER
representatives regarding site operating practices and
other site details.  The former  employee  sketched
facility process features on a map copied from state
files. The features included two  settling lagoons and
a feeder trench which transported plating wastes from
the process building to  the  lagoons.   The OSC
obtained copies of aerial photographs of the site area
from  the local  district  office  of  the U.S.  Soil
Conservation Service.  The state provided the OSC
with copies of all historical site and violation reports.
These  sources indicated the possible presence and
locations of chromium, copper, and  nickel plating
process areas.
                                                  19

-------
                   Figure 6: ABC Plating Site
A
-\
r
A A >
o o ^
T"^ f
                                   TREELINE
                            SUSPECTED
                             LAGOONS
                                  SUSPECTED
                                   TRENCH
                                HOUSE
                               TRAILER
      SCALE IN FEET
100
     50
                    100
                               LEGEND
                       DAMAGED
                       BUILDING
                        AREA
—	   SURFACE FLOW

	SITE BOUNDARY
                             20

-------
The Team mobilized to the site with all the equipment
needed to perform multi-media sampling.  The OSC
and Team  made a site entry,  utilizing appropriate
personal protective equipment and instrumentation, to
survey the general site conditions. They observed 12
vats, likely containing  plating solutions, on a concrete
pad where the original facility process building once
stood.   Measurements of pH ranged from 1  to  11.
Fifty drums and numerous smaller containers (some
on the concrete  pad, others sitting  directly on the
ground) were leaking and bulging, because of the fire.
Some rooms of  the  process building could not be
entered due to unsafe  structural  conditions caused by
the fire.  The Team noted many areas of stained soil,
which  indicated  container  leakage,  poor  waste
handling practices, and possible illegal dumping of
wastes.

2.13.3 Identification of Migration
         Pathways, Transport
         Mechanisms, and Receptors

During the site entry, the  OSC  and Team noted that
several areas were devoid of vegetation, threatening
wind erosion which could transport heavy metal- and
cyanide-contaminated  soil particulates off site.  These
particulates could be deposited on residential property
downwind or be inhaled by nearby residents.

The site entry  team  observed that the site was not
secure and there were signs of trespass (confirming a
neighbor's  claim that children  play  at the  facility).
These activities  could  lead to  direct  contact with
cyanide and heavy metal contaminants, in addition to
the potential for chemical burns from direct contact
with strong acids and bases as might be found in
leaking or unsecured  drums or containers.

After interviewing residents, it was established that
the homes located to the south and nearest to the site
rely upon private wells for their  primary  drinking
water supply. Ground  water is also utilized by several
small community systems which have wells located
within 2 miles of the site. The on-site settling lagoons
were unlined and therefore posed a threat to ground
water,   as  did  precipitation   percolating  through
contaminated  soils.   Contamination  might  have
entered  shallow  or deeper aquifers  and  potentially
migrated to off-site drinking water wells.

Erosion gullies located on site indicated soil erosion
and water transport due to storms.  Surface drainage
sloped toward the west and northwest, including a
distinct drainage path topographically downgradient of
the former lagoon area.  The Team observed stressed
and discolored  vegetation along the  surface water
drainage path.  Surface runoff of heavy metals and
cyanide was a direct contact hazard to local residents.
Surface water systems were also potentially affected.
Further downgradient, site runoff and the drainage
path entered an intermittent tributary of Little Creek.
The naturally eroded tributary flows west/southwest
into  a heavily  wooded  area  off-site prior to  its
convergence with Little Creek.  Little Creek in turn
feeds Barker Reservoir, located southwest of the site.
This reservoir is the primary water supply for the City
of Jonesville  and neighboring communities, which are
located 2.5 miles downgradient of the site. Shallow
ground-water discharges into the creek and reservoir
at several  locations,  serving  as another possible
contaminant migration route.

2.13.4  Sampling Objectives

The  OSC initiated a removal assessment  with a
specific sampling objective, as follows:

•       Phase  1  -  Determine whether a threat to
        public health, welfare, and the environment
        exists.  Identify sources of contamination to
        support  an immediate CERCLA-funded
        activation for containment of contaminants
        and security  fencing  (site  stabilization
        strategies) to reduce direct contact concerns
        on  site.

Once CERCLA funding was obtained and the site was
stabilized:

•       Phase 2 - Define the extent of contamination
        at the site and adjacent areas.  Estimate the
        costs for early action options and review any
        potential long-term remediation objectives.

•       Phase 3 - After early actions are completed,
        document the attainment of goals.  Assess
        that the response action was completed to
        the selected level and is suitable for  long-
        term goals.

2.13.5  Selection of Sampling
         Approaches

The OSC, Team, and PADER reviewed all available
information to formulate a sampling plan.  The OSC
selected a judgmental sampling approach for Phase 1.
Judgmental  sampling supports the immediate action
process by best  defining on-site contaminants in the
worst-case scenario  in order to evaluate the threat to
human health, welfare, and the environment. Threat
                                                  21

-------
is  typically  established using a  relatively  small
number of samples (fewer than 20) collected from
source areas or suspected contaminated areas based on
the historical data review and site reconnaissance.  For
this  site,  containerized wastes  were  screened to
categorize the contents and to establish a worst-case
waste volume, while bias-selected soil, ground-water,
surface water, and sediment samples were collected to
demonstrate whether a release had already occurred.

For Phase 2, a stratified systematic grid design was
selected  to define the  extent  of contamination in
soils.   The grid could  accommodate  analytical
screening and geophysical surveys.   A block grid
with 50-foot grid spacing was selected.  This grid size
ensured a 10 percent or less probability of missing a
"hot spot."   The grid was extended to  adjacent
residential  properties when contaminated  soil was
identified at grid points near the boundary of the site.
Based  on the results of soil sampling, a judgmental
approach was used to locate sample locations along
the drainage path.  A judgmental approach was also
used for the intermittent tributary and Little Creek.
Based on the results of soil  sampling and geophysical
surveys, a judgmental  approach was used to select
locations for installation of monitoring wells; at "hot
spots"; along the perimeter  of the  suspected plume
established from  analytical results and geophysical
survey  plots; and at background ("clean") locations.
Subsurface  soil   and ground-water samples were
collected  from  each  of  the  15 monitoring  well
locations for laboratory  confirmatory analysis to
establish the presence and, if applicable, the degree of
contamination at depth.

A  judgmental approach was selected  for Phase  2
sampling in the surface water migration route. During
Phase  1, samples were collected of soils along the
drainage path and of surface water and sediments in
the intermittent tributary. For purposes of EPA target
and  listing criteria,  surface water at  this  site was
considered to begin  at Little Creek, the perennially
flowing stream.  Phase 1 samples exhibited limited
site-related contamination along the drainage path.
Because of Little Creek's distance from the site  and
the tributary traversing through the wooded area,
detection of contamination in the surface water body
had  to be determined first.  For this reason, during
Phase 2 biased locations were selected for sampling in
Little Creek, the  intermittent tributary,  and along the
drainage path topographically  downgradient of the
former lagoons, to establish contaminant migration.
A  surface  water and  sediment sample  set was
collected along Little Creek upstream of the tributary
PPE to determine background conditions.
2.13.6 Analytical Screening,
         Geophysical  Techniques,
         and Sampling Locations

During Phase 1, containerized wastes were screened
using hazard categorization techniques to identify the
presence of acids, bases, oxidizers, and flammable
substances.  Following this procedure, photoionization
detector (PID) and flame  ionization detector (FID)
instruments, a radiation meter, and a cyanide monitor
were used to  detect the presence of volatile organic
compounds,  radioactive  substances, and cyanide,
respectively,  in the containerized wastes.   Phase  1
screening indicated the presence of strong acids and
bases and the absence of volatile organic compounds.
The Team collected a total of 12 surface soil samples
(0-3  inches),  3 ground-water samples, one surface
water sample, and one sediment sample during this
phase and sent them to  a laboratory for analysis.  The
soil sampling locations included stained soil areas,
erosion  channels, and  soil adjacent to  leaking
containers.  Background samples were not collected
during Phase 1 because they were unnecessary for
activating  immediate action  response  funding.
Ground-water samples were collected from  three
nearby  residential wells.   The surface water and
sediment samples were collected from the observed
PPE at  the confluence of  the unnamed intermittent
tributary and the on-site  surface  water  drainage
pathway.   Based on Phase  1 analytical results,
chromium was selected as the  target compound for
determination of  extent  of contamination in all
media/pathways.

During Phase 2 sampling activities,  the OSC used a
transportable X-ray fluorescence (XRF) unit installed
in an on-site trailer  to screen soil and  sediment
samples for  total chromium in order  to  limit the
number  of samples to be sent for laboratory analysis.
Soil sampling was performed at all grid nodes at the
surface  (0-4 inches) and subsurface (36-40 inches).
The  36-40 inch  depth   was   selected based on
information obtained  from  state reports and local
interviews, which indicated that lagoon wastes were
approximately 3 feet below ground surface. Twenty-
four surface  and  subsurface samples were  sent for
laboratory  confirmation  analysis  following  XRF
screening.  The analytical results from these samples
allowed for site-specific calibration of the XRF  unit.
Once grid nodes with a contamination level greater
than a  selected  target action   level were  located,
composite  samples   were  collected  from   each
adjoining grid cell. Based on  the XRF data,  each
adjoining cell was either identified as "clean" (below
                                                   22

-------
action level) or designated for response consideration
(at or above action level).

Also  during  Phase  2,  the   OSC  oversaw  the
performance of ground penetrating radar (GPR) and
electromagnetic   conductivity  (EM)  geophysical
surveys to help delineate the buried trench and lagoon
areas, any conductive ground-water plume, and any
other waste burial areas.   The  GPR survey was
conducted over the original  grid and run along the
north-south grid axis across the suspected locations of
the trench and lagoons.  For the comprehensive EM
survey,   the  original  50-foot  grid  spacing was
decreased to 25 feet along the north-south grid axis.
The EM survey was run along the north-south axes
and readings were obtained at  the established grid
nodes.  The EM survey was utilized throughout the
site to detect the presence of buried metal objects
(e.g., buried pipe leading to the lagoons) and potential
subsurface contaminant plumes.

Using the data obtained during soil sampling and the
geophysical surveys,  a ground-water investigation
plan for Phase 2 was prepared. Monitoring wells were
located  in  areas shown to be heavily  contaminated
during soil sampling; along the  outer perimeter of a
contaminant plume based on soil XRF results and the
geophysical  surveys;  and  apparent  upgradient
locations  for  background  conditions  comparison.
Fifteen wells were located at grid nodes established
using the  above  data.   Upon  monitoring  well
installation and sampling, a hydraulic (pump) test was
completed of the bedrock monitoring wells to gather
information about aquifer characteristics, which help
assess the ability of contaminants to migrate through
ground water.

Three soil grid samples collected along the bank of the
surface   water   drainage   path,   topographically
downgradient  of the former lagoon area, exhibited
chromium contamination ranging from 772  to 2,060
mg/kg.   The  samples were  from  random  locations
according to the layout  of the sampling grid.  This
chromium contamination suggests that a contaminant
plume   may    have   traveled    topographically
downgradient from the  lagoons along the drainage
path.  (Contamination was not detected at  depth in
these samples.) Based on these results, it was decided
that the  surface water  migration  route should  be
further evaluated.
The tributary PPE sample set collected during Phase
1 did not exhibit any contamination at the time of
sampling.  However, the Team observed that the
drainage path and tributary  became very level and
shallow  prior to, and in, the heavily wooded area.
Contaminants may settle out in this area due to its
level  terrain  and  many flow  obstructions.   Any
contaminants here would be transported downstream
only during heavy  flow or storm events.   It was
decided  to  collect  additional  surface  water  and
sediment sample sets along the drainage path and
tributary using a judgmental approach during Phase 2
activities.   If the  site were  to continue  under
Superfund remedial site  evaluation for consideration
of the surface water migration route, contamination
must  have  been  detected or  suspected  in  the
perennially flowing stream, Little Creek.  A  surface
water and sediment  sample set at the PPE  for the
tributary to  Little Creek was collected to establish
whether the contamination had migrated to the  surface
water body.  The sediment  sample would establish
historical contamination,  while the surface water
aliquot   would   indicate   current   contamination
migration.    (Phase  2   sampling  activities  were
scheduled to occur while the intermittent tributary was
flowing.)  A background sample set was collected in
Little Creek by obtaining  surface water and sediments
upstream of the tributary confluence (PPE).

Phase 3 activities are discussed in Section 6.7.

2.13.7  Parameters for Analysis

During  Phase  1  sampling  activities, full  priority
pollutant metals and total  cyanide analyses  were
conducted on all soil, ground-water, surface water,
and sediment samples sent to the laboratory.  These
parameters were initially selected based on research of
plating  chemistry  (plating facilities generally use
either an acid bath or basic cyanide bath to achieve the
desired coating on their metal products).  Since Phase
1 samples were collected from  the areas of  highest
suspected contaminant concentration (i.e., sources and
drainage pathways), Phase 2 samples (all media types)
were  analyzed  for total   chromium,  hexavalent
chromium  (in water only),  and cyanide,  the only
analytes detected consistently during the Phase 1
analyses.  During Phase 3, the samples sent to the
laboratory for definitive analysis were analyzed  for
total chromium and cyanide.
                                                   23

-------
   3.0  FIELD ANALYTICAL SCREENING AND SAMPLING  EQUIPMENT
3.1     INTRODUCTION

Sample collection requires an understanding of the
capabilities of the sampling equipment, since using
inappropriate  equipment may result in biased or
nonrepresentative samples.   The limitations, uses,
construction,  and ease of use of the equipment or
techniques must be understood prior to designing a
sampling plan.

Section 3.2  provides an  overview of the  most
commonly  utilized  field  analytical  screening
equipment  and  techniques that  are applicable to
surface water and sediment sampling.  Section 3.3
provides  information   for   selecting   sampling
equipment.  The example site synopsis continues at
the end of the chapter.
3.2    FIELD ANALYTICAL
        SCREENING EQUIPMENT

Field analytical screening techniques and equipment
may provide valuable information for developing
sampling strategies. Field analytical screening can
determine chemical classes of contaminants  and in
some  cases  can identify  particular  substances of
concern.   Real-time  or direct-reading capabilities
narrow the possible groups or classes of substances,
which aids in selecting the appropriate laboratory
analytical  method.  These screening techniques are
useful and economical when gathering large amounts
of site data.   The screening techniques can also be
utilized to select sample locations, as well as samples
to  be  sent   for  off-site  laboratory  analysis  or
confirmation.   The analytical  screening  methods
provide  on-site measurements  of  contaminants of
concern, limiting the number of samples which need
to be sent  for  off-site  analysis.    All  screening
equipment and methods described in this section are
portable (the equipment is hand-held and generally no
external power source is  necessary).  Screening
techniques for surface water and sediment sample
analysis are  discussed in Table 1;  the methods are
presented in a general order of those most utilized and
applied  shown first.   Field  analytical  screening
methods  are  most often used to identify waste or
contaminant  source areas  and may not be required
during  all surface water  and  sediment sampling
events.
Field screening generally provides analytical data of
suitable quality for site characterization, monitoring
response activities, and health and safety decisions.
Its  application with surface  water  and sediment
sampling may be more limited than with other sample
media.  For investigations of water bodies,  these
methods may assist  with sample  selection  for
laboratory analysis or for a preliminary determination
of the extent of contamination in sediments or of a
contaminant plume in a static water body. Screening
methods can provide rapid, cost-effective, real-time
data;  however,  results are often not compound-
specific and not quantitative.

When selecting one screening  method over another,
consider relative cost, sample analysis time, potential
interferences or instrument limitations, applicability to
the  sample  medium,  detection limit,  QA/QC
requirements, level of training required for operation,
equipment availability  and durability, and data bias.
Also consider which elements, compounds, or classes
of compounds the screening instrument is designed to
analyze.  As discussed in Section 2.9, the screening
method  selected  should  be  sensitive enough to
minimize the potential for false negatives.   When
collecting samples  for  screening  analysis  (e.g.,
portable gas chromatograph), evaluate the detection
limits and bias of the screening method by sending a
minimum of 10 percent of the samples for laboratory
confirmation. For additional information on specific
field screening analytical techniques and equipment,
please refer to the U.S. EPA Compendium of ERT
Waste  Sampling  Procedures, OSWER Directive
9360.4-07 or Superfund Program Representative
Sampling Guidance, Volume  4  -  Waste, OSWER
Directive 9360.4-14.
3.3    SURFACE WATER AND
        SEDIMENT SAMPLING
        EQUIPMENT AND
        SELECTION

Sample collection requires an understanding of the
capabilities of the sampling equipment, since the use
of   inappropriate   equipment  may   result  in
nonrepresentative samples.  Select approved sampling
equipment based on the sample type and medium,
matrix,  physical  location  of the sample  point,
sampling  objectives,   and   other   site-specific
conditions. Site-specific conditions may dictate that
                                                 24

-------
only one method or type  of equipment will work.
Also consider the equipment design.  For example, a
device which aerates a sample during collection might
release volatile organic compounds and thus not yield
a sample representative of actual conditions.

Also consider the compatibility of the contaminants
being sampled with the composition of the sampling
device.   All sampling devices  should be of good
quality.  They should be made of material that will not
affect the outcome of analytical results; they must not
contaminate the sample being collected and must be
able to be cleaned easily in order to reduce the risk for
cross-contamination.  Use of a device constructed of
undesirable material may compromise sample quality
by having components of its material leach into the
sample  or adsorb  constituents of the sample.  If a
sampling device cannot be easily decontaminated,
consider  the   cost-effectiveness  of  expendable
equipment. Standard construction materials typically
include  Teflonฎ, polyvinyl chloride (PVC), glass,
stainless steel,  and steel.   Selection is  commonly
determined  by   considering the  substance  to be
sampled and the cost of sampling.

Select,  when possible,  equipment that  is easy to
operate,  in order to decrease training requirements and
when wearing   cumbersome  personal   protective
equipment. Complicated sampling procedures usually
require  increased training and introduce a greater
likelihood of procedural errors; SOPs help to avoid
such errors.  Follow SOPs for the proper use and
decontamination of all sampling equipment. The  U.S.
EPA  Compendium of ERT Surface  Water  and
Sediment Sampling Procedures, OSWER Directive
9360.4-03, provides SOPs for some standard surface
water and sediment sampling equipment and methods.

This section provides appropriate uses, advantages,
and disadvantages of select examples of surface water
and sediment sampling equipment.  Representative
sampling  requires  that   appropriate   sampling
equipment be chosen for each sampling objective and
location. The  surface water sample collected  may
represent all phases or a specific stratum present in the
water,   as  required  by  the sampling  objective.
Construction  material,  design  and   operation,
decontamination procedures, and the procedures for
proper use are  factors to  consider when selecting
equipment.  The following characteristics of surface
water can affect the representativeness of a sample:
density,  analyte solubility, temperature, and currents.
A sampling device should have a capacity of at  least
500 milliliters, if possible, to reduce the number of
times the  liquid must  be  disturbed and to reduce
sediment agitation.
When  selecting  sediment  sampling  equipment,
consider the  width,  depth, flow,  and  the  bed
characteristics of the area to be sampled.  Sediment
may be sampled in both flowing and standing water.
Samples may be recovered using a variety of methods
and equipment, depending on the depth of the aqueous
layer, the portion of the sediment profile required
(surface vs. subsurface), the type of sample required
(disturbed vs.  undisturbed) and the sediment type.
Sediment is collected from beneath an aqueous layer
either directly using a hand-held device, or indirectly
using a remotely-activated device.  Selection of a
sampling device is most often contingent upon the
depth of water at the sampling location as well as the
physical characteristics of the medium to be sampled.
Take  care to  minimize  disturbance  and  sample
washing  as the sample  is  retrieved through the
aqueous layer.  It is important to get a representative
sample  of all  horizons  present  in  the  sediments.
Maintain sample integrity by preserving the sample's
physical form and thus its chemical composition.

Tables 2 and 3 provide examples of commonly used
surface  water and sediment sampling  equipment,
respectively, but the list  is not  exhaustive.   The
advantages and disadvantages listed represent only
highlights of the equipment use. Additional details on
surface water and sediment sampling equipment and
procedures  are  provided   in   the  U.S.   EPA
Compendium of ERT Surface Water and Sediment
Sampling Procedures, OSWER Directive 9360.4-03.
3.4    EXAMPLE SITE

3.4.1  Selection of Analytical
        Screening Equipment

Phase 1 sampling identified the sources and types of
on-site contaminants in order to establish a threat.
Hazard  categorization  techniques,  organic vapor
detecting  instruments,  and radiation  and  cyanide
monitors  were  utilized  to   tentatively   identify
containerized liquid wastestreams in order to select
initial judgmental sampling locations. During  Phase 2
sampling, a portable XRF unit was used to determine
the extent of soil  contamination  and to  identify
additional "hot spots."  Soil samples to be  sent for
laboratory analysis were placed into sampling jars.
An  organic   vapor  detecting   instrument  (PID)
continued to be utilized throughout all field activities
for health and safety monitoring during Phases 1
through 3.

-------
The portable XRF was used during soil screening,
monitoring well installation, and sediment sampling.
Ground-water and  surface  water  samples  were
screened in  the  field  for  pH, conductivity, and
temperature   using  a   three-in-one   monitoring
instrument.  The instrument probe was placed into a
clean  glass jar  containing an aliquot of the  water
sample.  The instrument was decontaminated prior to
and after each sample screening.

3.4.2  Selection of Geophysical
        Equipment

The GPR instrument delineated buried trench and
lagoon  boundaries.    The  EM   meter  detected
subsurface conductivity changes due to buried metal
containers  and contaminants.    The  EM-31D,  a
shallower-survey ing instrument than the EM-34, was
selected because expected contaminant depth was less
than  10 feet  and because  of  the  instrument's
maneuverability and ease of use.

3.4.3  Selection of Sampling
        Equipment

Disposable plastic scoops were used for Phase 1 soil
and sediment sampling.  Phase 1  ground-water and
surface  water samples were collected directly into
sample containers.  For Phase 2, soils were collected
from  the near  surface  (0-4 inches) and  at depth.
Stainless steel trowels were used to collect shallow
samples.   Subsurface samples  were collected by
advancing boreholes using  a hand-operated power
auger to just above the sampling zone and then using
a stainless steel split spoon to retrieve the soil. The
split spoon was advanced with a manual hammer
attachment.

Monitoring wells were installed using a dual-tube, air
percussion drill rig.  Borehole  soil samples were
retrieved using 2-foot  stainless steel split  spoon
samplers.  Soil from the split spoons was transferred
to sample containers using disposable plastic scoops.
Ground water was  sampled in  Phase 2 from  the
monitoring wells installed on site. The ground-water
samples were obtained using  dedicated bottom-fill
Teflonฎ bailers.  The bailer was attached to  nylon
rope, which was selected because  less material would
be adsorbed onto the nylon and brought out of the
well.    Residential  ground-water  samples  were
collected directly into the sample  containers from the
kitchen sink tap. Water level and depth measurements
were   obtained  from   monitoring   wells   using
decontaminated electronic measuring equipment.

As  in  Phase  1, Phase 2  sediment  samples were
collected using dedicated disposable  plastic scoops.
Surface water samples were collected directly into the
sample containers.   The shallow depth and narrow
breadth of the intermittent tributary and Little  Creek
did not require any specialized equipment or remote
sampling devices.
                                                 26

-------
                   TABLE 1:  SURFACE WATER AND SEDIMENT FIELD ANALYTICAL SCREENING EQUIPMENT
   Instrument
                 Use(s)
            Advantage(s)
          Disadvantage(s)
Direct -
Reading/
Real - Time
Instruments
Portable monitoring instruments used to
measure or identify specific parameters
under field conditions including: pH,
specific conductivity, temperature, salinity,
and dissolved oxygen	
1 Portable and easy to operate and
 maintain in the field
1 Qualitative identification
1 May be used with probes placed
 directly into the sample medium
May return a reading with a high degree
of error
Field Test Kits
and
Colorimetric
Indicator Tubes
Used for detecting specific compounds,
elements, or compound classes in surface
water and sediment
• Rapid results
• Easy to use
• Kits may be customized to user needs
Limited number of kit types available
Interference by other analytes is
common
Subjective interpretation is needed
Can be prone to error
May have limited shelf life
Colorimetric tubes may be used for
ambient air only
Photoionization
Detector (PID)
Detects and measures total concentration of
volatile organic compounds (VOCs) and
some non-volatile organic and inorganic
contaminants in ambient air or container
headspace; used to evaluate existing
conditions, identify potential sample
locations, or identify extent of
contamination
1 Immediate results
1 Easy to operate and maintain
1 Detects to parts per million (ppm) level
 for headspace analysis
Limited use to quantify specific
substances
Does not detect methane
Readings can be affected by high winds,
humidity, condensation, dust, power
lines, and portable radios
Probe should not be placed directly into
sample medium	
Flame
lonization
Detector (FID)
Detects and measures the level of total
organic compounds (including methane) in
ambient air or container headspace; used to
evaluate existing conditions, identify
potential sample locations, or identify
extent of contamination
1 Immediate results
1 Detects to ppm level for headspace
  analysis
1 Rugged
1 Available with a GC mode to detect
  specific VOCs
Does not respond to inorganic
substances
Does not recognize and may be
damaged by acids
Requires training and experience
Requires a hydrogen fuel source
Probe should not be placed directly into
samnle medium	
                                                                    27

-------
               TABLE 1: SURFACE WATER AND SEDIMENT FIELD ANALYTICAL SCREENING EQUIPMENT (Cont'd)
   Instrument
                 Use(s)
            Advantage(s)
           Disadvantage(s)
Hazard
Categorization
(hazcat)
Performed as an initial screen for hazardous
substances to provide identification of the
classes/types of substances in the individual
surface water or sediment sample
• Rapid categorization of unknown
 liquids
• Good for screening and determining
 contaminant compatibility
 Not analyte-specific, yields only basic
 information (e.g., base vs. acid,
 chlorinated vs. non-chlorinated
 substance)
 Requires numerous chemical reagents
 Requires interpretation of results	
Portable Gas
Chromatograph
(GC)
Used to measure occurrence and
concentration of VOCs and some semi-
VOCs
1 Can screen "hot spots"
1 Determines potential interferences
1 Conducts headspace analysis
1 Semi-quantitation of VOCs and semi-
 VOCs
• Highly temperature sensitive
• Requires set-up time, many standards,
 and extensive training
Radiation
Detector
Detects the presence of selected forms of
radionucliides in sediments
1 Easy to use
1 Probes for one or combination of
 alpha, beta, or gamma emitters
 Units and detection limits vary greatly
 Time intensive for detailed surveys
 Experienced personnel required to
 interpret results	
Portable X-ray
Fluorescence
(XRF)
Used to detect heavy metals in sediments
1 Rapid sample analysis
1 Detects to ppm level (detection limit
 should be calculated on a site-specific
 basis)
 Requires trained operator
 Sediment must be dried
 Potential matrix interferences
 Detection limit may exceed action level
 Radioactive source
 Cannot be used  for surface water
 samnles	
                                                                    28

-------
                                                 TABLE 2:  SURFACE WATER SAMPLING EQUIPMENT
  Sampler
                               Uses
                                                               Advantages
                                                                                 Disadvantages
Laboratory-
cleaned
Sample
Container
(Direct
Method)
Used to collect samples from surface and
shallow depths of surface water bodies
' Quick and easy to use
' No decontamination required
' Disposable
' Reduces risk of cross-contamination from sampling equipment
' Reduces the loss of volatile fraction during transfer to a sample
 container
' Preferred if there is an oily layer on the sample surface; the
 layer will not stick to a sampling device and thus miss being
 transferred to the sample container	
• Cannot be used for other water bodies, such as waste
 impoundments, where contact with concentrated
 contaminants is a concern
• Labelling can be difficult
• May not be possible when containers are pre-preserved
Scoop,
Ladle,
Beaker
(Transfer
Devices)
Stainless steel, Teflonฎ, or other inert
composition material devices to transfer
the sample directly into a sample
container at a near shore location
' Easy to use and decontaminate
' Allows collection without a loss of preservative in the sample
 container
• Difficult to maneuver sample especially if placing into VOA
 vials
• Avoid equipment with painted or chrome-plated surfaces
• May aerate sample releasing VOCs, or some contaminants
 may adhere to the surface of the transfer device
Weighted
Bottle
Sampler
Used to collect samples in a water body
or impoundment at predetermined depth
' Easy to decontaminate
' Simple to operate
' Sampler remains unopened until at desired sampling depth
• Cannot be used to collect liquids that are incompatible with
 the weight sinker, line or actual collection bottle
• Sample container may not fit into sampler, thus requiring
 additional equipment
• Sample container exposed to matrix	
Pond           Used for near shore sampling where
Sampler        cross-sectional sampling is not
               appropriate and for sampling from outfall
               pipe or along a disposal pond, lagoon, or
               pit bank where direct access  is limited
                                        ' Easy to fabricate using a telescoping tube; not usually
                                         commercially available
                                        ' Can sample at depths or distances up to 3.5 meters (can sample
                                         areas difficult to reach with extension)
                                                            • Difficult to obtain representative samples in stratified water
                                                             bodies
                                                            • Sample container may not fit into sampler, thus requiring
                                                             additional equipment
Peristaltic      Used to extend the reach of sampling
Pump          effort by allowing the operator to reach
               into the water body, sample at depth, or
               sweep the width of narrow streams
               through the use of Teflonฎ or other
               tubing
• Very versatile
• Easy to carry and operate; fast
• With medical-grade silicone, it is suitable to sample almost
 any parameter including most organic contaminants
• Sample large bodies of water
• Capable of lifting water from depths in excess of 6 meters
                                                                                                    • Depth limited to 7.5 meters/25 feet
                                                                                                    • Cannot be used if volatile compounds are to be analyzed
                                                                                                    • Lift ability decreases with higher density fluids, increased
                                                                                                     wear on silicone pump tubing, and increases with altitude
                                                                                                    • Oil and grease contaminants may adhere to tubing and thus
                                                                                                     decrease concentration in sample
                                                                                                    • Must often change tubing between locations to decrease
                                                                                                     cross-contamination; must always have extra tubing on hand
                                                                                                    • At high flow, must weight tubing in stream
                                                                                    29

-------
                                           TABLE 2: SURFACE WATER SAMPLING EQUIPMENT (Cont'd)
Sampler
Bailer
Kemmerer
Bottle/Van
Dorn
Sampler
Bacon
Bomb
Sampler
Wheaton
Dip Sampler
Depth-
Integrating
Samplers
PACS Grab
Sampler
Uses
Used for collecting samples in deep
bodies of water where cross-sectional
sampling is not appropriate
Used when access is from a boat or
structure such as a bridge or pier, and
where discrete samples at specific depths
are required
Used to collect samples from discrete
depths within a water body; generally
used when access is from a boat or
structure
Useful for sampling liquids in shallow
areas or from areas where direct access is
limited; also useful when sampling from
an outfall pipe
Used to collect water and suspended
sediment samples; used with the EWI
and EDI composite sampling techniques
Used to collect water samples from
impoundments, or ponds with restricted
work areas
Advantages
• Easy to use
• No power source needed
• Bailers can be dedicated to sample locations
• Disposable equipment available
• Can be constructed of a varietv of materials
• Can take discrete samples at specific depths
• Can sample at great depths
• Kemmerer Bottle lowers vertically; Van Dorn Sampler lowers
horizontally, which is more appropriate for estuary sampling
• Remains unopened until the sampling depth
• Can collect a discrete sample at desired depth/stratum
• Widely used and available
• Long handle allows access from a discrete location
• Sample container is not opened until specified sampling depth
• Sampler can be closed after sample is collected ensuring
integrity
• Easy to operate
• Allows for collection of representative samples of suspended
materials
• Samples proportionate to the velocity of the water body
• Allows discrete samples to be collected at depth
Disadvantages
• Transfer of sample may cause aeration, thus not appropriate
for VOCs
• Inappropriate for strong currents or where a discrete sample
at a specific depth is required
• Sampling tube is exposed to material while traveling down to
sampling depth
• Transfer of sample into sample container may be difficult
• May need extra weight
• Often constructed of materials incompatible with sample
• Difficult to decontaminate
• Difficult to transfer sample to sample container
• Tends to aerate sample thereby losing volatile organic
constituents
• Depth of sampling is limited by length of extension poles
• Exterior of sample container may come in contact with
sample
• Sample container may not fit into sampler
• Requires experienced operator
• Depth of sampling is limited by length of extension pole
• Difficult to decontaminate
Note:  Standard operating procedures and example figures of some of the equipment is available in the U.S. EP Compendium ofERT Surface Water and Sediment Sampling Procerfwre^OSWER
Directive 9360.4-03.

Abbreviations

EWI = equal-width-increment
EDI = equal-discharge-increment
                                                                               30

-------
TABLE 3: SEDIMENT SAMPLING EQUIPMENT
Sampler
Scoops,
Trowels,
Dippers,
Shovels
(Direct Method)
Vertical-pipe,
Core Sampler
Ponar/Ekman/
Peterson





Thin-Wall Tube
Auger
Veihmeyer
Sampler
Uses
Used for surface sediments where
water depth is shallow (limited to
near surface)
Used to collect samples of most
sediments to depths of 75 cm (30 in.)
Ponar dredge is used to sample most
types of sediments
Ekman dredge is used where bottom
material is unusually soft, such as
organic s u ges
Peterson dredge is used when
bottom is rocky, in deep water or in a


Used to collect consolidated
sediments at surface and at depth
Used for sampling most types of soil
and sediments, except very wet or
stony sediments
Advantages
• Quick and easy to use
• Easy to decontaminate
• Available in a variety of materials
• Appropriate for consolidated sediments
• Disposability reduces the risk for cross-
contamination
• Laboratory scoop is less subject to
corrosion or chemical reactions than
commercially available garden or
household tools (less risk for sample
contamination)
• Easy to use
• Can collect undisturbed sample
(minimum loss of fine fraction) that
can profile any stratification as a result
of changes in deposition
• Provides historical record of deposition
• Ponar is easily operated by one person;
light weight
be operated without a winch or crane
• Appropriate for most sediment types
rom si s o granu ar ma ena s
• Ekman can obtain samples of bottom
fauna
• Peterson can be used in rocky
substrates and high velocity water
• Easily operated by one person
• Easy to use
• Preserves core sample
• Can achieve substantial depths with
appropriate length of tubing
• Various driveheads available for
different sediment types
Disadvantages
• Disturbs the water/sediment interface and may alter sample integrity; fine fraction is lost
• Not efficient in mud or other soft substrates
• Difficult to release secured undisturbed samples to readily permit subsurface sampling
• Difficult to maneuver sample especially if placing into VOA vials
• Limited by depth of aqueous layer
• Avoid equipment with painted or chrome-plated surfaces (common with garden trowels)
• When used in impoundments, penetration depths could exceed that of substrate and
damage the liner material
• A relatively small surface area and sample size result in the need for repetitive sampling
to obtain an adequate amount for analysis
• Dredges are normally used from a boat, bridge or pier due to the weight of the
equipment which may require a boom for lowering or raising
• Not capable of collecting undisturbed sample and may cause agitation currents that may
temporarily resuspend some settled solids
• Ekman is not suitable for sand rock and hard bottoms ve elation covered bottoms
and streams with high velocities
• Should not be used from a bridge more than a few feet high because spring mechanism
• Not capable of collecting an undisturbed sample and may cause agitation currents that
may temporarily resuspend some settled solids
• Peterson can displace and miss light materials if allowed to drop freely
• Limited by the depth of the aqueous layer
• May be difficult to remove core sample from auger
• Possible washout during retrieval
• Very difficult to clean
• Parts needed for sampler are not appropriate for certain analyses
• Not appropriate in rocky substrate
                   31

-------
                                               TABLE 3: SEDIMENT SAMPLING EQUIPMENT (Cont'd)
Sampler
PACS Grab
Sampler/Sludge
Getter
Sampling Trier
Soil Coring
Device/ Silver
Bullet Sampler
Sludge Judge
Hand Corer
Gravity Corer
Bucket and
Pesthole
Augers
Uses
Used for collecting grab samples
from ponds and impoundments at
depth
Used to collect sediments up to 40
inches depth from water surface
Used when a core sample is required
Used to collect a core of sediments
or water and sediments
Used for sediments in water that is
very shallow (a few inches)
Collects core samples from most
sediments; can be used in water
deeper than 5 feet
Used for direct method samples
Advantages
• Allows discrete samples to be collected
at depth
• Can be used in heavy sediments or
sludges, or moderately viscous
materials
• Preferred for moist or sticky samples
• Contains a collection tube which holds
core relatively intact
• Bit of silver bullet sampler is
replaceable
• Easy to use
• Core allows delineation of settled state
of sediments or physical state of water
body
• Easy to use
• Preserves sequential layer of deposit
(useful for historical information)
• Appropriate for trace organic
compounds or metals analyses
• May have a check valve on top to
prevent wash-out during retrieval
• Collects undisturbed samples
• Can collect to a depth of 75 cm (30 in.)
within the sediment substrate
• Preserves sequential layer of deposit
(useful for historical information)
• Has a check valve to prevent washout
during retrieval
• Direct sample recovery
• Fast and easy to use
• Provides a large volume sample
Disadvantages
• Not useful in very viscous materials
• Depth of sampling is limited by length of extension pole
• Heavy, possibly requiring more than one person to operate
• Difficult to use in stony or sandy substrates
• May be difficult to remove sample from sampling device
• Difficult to use in rocky or tightly packed substrates
• Depth restrictions
• Use is limited due to possible reactivity of construction material
• Difficult to decontaminate
• Not useful in thick sediments
• Can be disruptive to water/sediment interface
• May cause disruption to sample integrity
• Delivers small sample size requiring repetitive sampling
• May damage liners in impoundments if penetration is too deep
• Not suitable for obtaining coarse-grained samples
• Disturbs sediment horizons
• May cause disruption to sample integrity
• Pesthole augers that are designed to cut through fibrous, rooted swampy areas have
limited sample collection utility
Note:  Standard operating procedures and figures of many of these equipment types are available in the U.S. EP'Compendium ofERT Surface Water and Sediment Sampling Procerfwre^OSWER
Directive 9360.4-03.
                                                                              32

-------
            4.0 FIELD SAMPLE  COLLECTION AND PREPARATION
4.1     INTRODUCTION

When sampling a water body, the following critical
factors must be considered to ensure that the sample
is representative:  points of sampling, frequency of
sampling,  and maintenance of integrity of sample
prior to analysis.  During a response action, proper
field sample collection and preparation methods are as
important  as proper sampling equipment selection.
Sample collection refers to the physical removal of
water or sediments from a water body for the purposes
of either screening or laboratory analysis, and includes
sample quantity and sample volume.  Field sample
preparation refers to all aspects of sample handling
from collection to the time the sample is received by
the laboratory.  This chapter provides information on
sample collection and preparation for various sample
types and sources.

The collection of samples from water bodies presents
unique challenges.   Some samples involve merely
collection by a direct  method in shallow waters.
Often however, site-specific conditions may dictate
the  use of special  equipment to access  the sample
location, increased health and safety concerns,  and
proper timing to consider tidal fluctuations and/or
flow rates.
4.2    SAM RLE VOLUME AND
        NUMBER

How   a  sample  is   collected  can  affect  its
representativeness.   The  greater the number of
samples collected from a site  and the larger the
volume of each sample, the more representative the
analytical  results should  be.   However, sampling
activities are often limited by sampling budgets and
project schedules.

Sampling  objectives and analytical  methods  are
considerations in determining  appropriate sample
volume and number. The volume of a sample should
be  sufficient  to  perform all  required laboratory
analyses with an additional amount remaining to
provide for analysis of QA/QC samples (including
duplicate analyses).  The volume of water samples can
vary depending on the requirements of the laboratory
and the analytical method(s).  The minimum volume
collected should  be three to four times the amount
required for the analysis.  Typically, no more than 8
liters are required for each water sample.  The amount
of sediment required for analysis can also vary but
will not usually exceed 16 ounces.  Always consult
the analytical laboratory during sampling design to
determine the  adequate volume required  for  each
matrix and location.  Sometimes site conditions may
limit the available sample volume; creek waters may
be shallow during a dry season or the sediments may
consist  of  a rocky  substrate.   Review  the  site
conditions when selecting laboratory analyses.  Where
sample volume may be limited, it may be necessary to
reduce the number of analyses to those most critical to
the investigation  and its objectives.

The number of sample locations  will depend upon
site-specific requirements  and  must satisfy  the
investigation objectives.   A few  selected  locations
may  be enough  to  identify  the  existence  of
contamination,  or multiple location,  systematic
sampling may be required to delineate the full extent
of contamination.  Both strategies may be used during
different phases of a site investigation. The physical
characteristics  of the water body might also dictate
sample numbers.  A  complicated,  well-developed
system of tributaries, changes in flow, and  sediment
deposition will necessitate additional sample  locations
to ensure that samples  are  representative  of  site
contaminant migration conditions.  The number of
samples may vary according to the particular  sampling
approach used at the site.   Chapter 2  provides
additional information on sampling  approaches and
sample locations  and numbers.
4.3    SURFACE WATER SAMPLE
        COLLECTION

Sampling  situations vary widely and therefore no
universal sampling procedure can be recommended.
Sampling considerations and guidelines, however, do
apply to every case.   Prior to sample collection,
review the characteristics of the water body. When
sampling surface waters and sediments, always collect
the water samples before sediment samples to avoid
disturbing sediments into the water and biasing the
water sample.  Avoid surface scum.  Sampling should
proceed from downstream to upstream locations to
minimize disturbance. Determine tidal influences and
flow rates, which can affect sample collection.
                                                 33

-------
Surface water samples are generally collected as grab
samples  because of the natural  mixing effect  of
flowing waters.  However, compositing samples may
assist in the attempt to document intermittent  or
sporadic contaminant discharges. This is particularly
of concern with effluent releases which  are highest
during certain  times  of the  day.   Representative
sampling  would   seek to  obtain  an  average
concentration from release and no release conditions.
Section  2.4.2  describes composite  samples and
compositing approaches. Surface water compositing
is  generally  completed using  the surface  water
collection  equipment described in  Chapter  3.   A
programmable composite sampler is  available for time
compositing. This electronic pumping tool collects  an
aliquot of the sample water from a stationary location
over designated time intervals (e.g.,  30 or 60 minutes)
for a certain study  period  (e.g.,  24 hours).   This
equipment  allows the collection of an "averaged,"
uniform,  representative sample,  but   will not
distinguish a particular interval when contaminant
levels are high or low. The criteria for selection of the
                                       "automatic sampler"  are  the same as  for  other
                                       sampling equipment, including compatibility, sample
                                       integrity, etc.   (Automatic sampling equipment is
                                       generally not used at EPA CERCLA sites prior to
                                       remedial investigations and is therefore not discussed
                                       in greater detail in this document; please refer to U.S.
                                       EPA, 1986 and Krajca, 1989 for further discussion of
                                       these devices.)

                                       Fresh water environments are commonly separated
                                       into  three groups: flowing waters,  such as rivers,
                                       streams, and creeks; static water bodies, such as lakes,
                                       ponds, and impoundments;  and estuaries.   These
                                       waterways differ in characteristics, therefore sample
                                       collection must be adapted to each.  A discussion of
                                       special considerations for sampling in wetlands is also
                                       included in this section. This section provides general
                                       information on  sampling several  types of water
                                       bodies.     Table  4  compares  advantages  and
                                       disadvantages of  sample  method  locations.   For
                                       specific sampling information, refer to the U.S. EPA
                                       Compendium of ERT Surface Water and Sediment
                                       Sampling Procedures, OSWER Directive 9360.4-03.
          TABLE 4:  SURFACE WATER AND SEDIMENT SAMPLE METHOD LOCATION
  Location
  Water Body Type
    Advantages
   Disadvantages
  Bridge, Pier
Rivers, streams, large ponds or
impoundments
• Provide ready access;
 allow sampling at any
 point across water body
• Little disturbance
• Structure can alter water
 flow and influence
 sediment deposition and
 scouring
• Not always in ideal
 location
  Wading,
  Shore
Lakes, ponds, slow-moving
rivers and streams
• Ease of collecting
 sediment samples
• Disturbs bottom
 deposits; introduces
 particulate and sediments
 into water
• Samplers must carry
 large amounts of
 equipment	
  Boat
Slow-moving, deep water, and
estuaries
• Appropriate for
 locations where no other
 means are available
• Safety concerns
• Difficult to
 decontaminate
• Requires a means of
 launching  and
 transporting boat
• May affect flow of
 water
• Depending on depth,
 may disturb sediments
                                                  34

-------
4.3.1   Rivers,  Streams, and Creeks

This group  of water bodies includes  outfalls  and
drainage features (e.g., ditches and culverts), as well
as rivers, streams, and creeks.  Methods for sampling
flowing water bodies vary from the simplest  direct
methods to more sophisticated multipoint sampling
techniques.  The size of the stream or river and its
amount of turbulence can affect the number and type
of sampling locations. In small streams (less than 20
feet wide) it is possible to select a location with well-
mixed water for grab  sampling.  A grab sample
collected at mid-depth in moving water at the main
flow line would represent the entire cross-section.
(The main flow line is not necessarily the center of the
stream; observe flow patterns across the surface to
identify this area.)   Slightly larger streams or rivers
would require multiple samples at locations across the
channel  width.   At the  minimum,  one vertical
composite  (consisting of grab locations from  just
below surface, mid-depth, and just above the bottom)
collected at the main flow line would be necessary.
Identifying sampling locations that  are  well mixed
vertically or ones that are horizontally  stratified is
useful  prior to sampling.   When sampling rivers,
streams, or creeks,  locate the  area that exhibits the
greatest degree of cross-sectional homogeneity.  Since
mixing is primarily attributed to turbulence and water
velocity, selecting a site  immediately downstream of
a mixing zone will  ensure good vertical mixing. In
the absence of mixing zones, the selection of a site
without  any  immediate point  sources, such  as
tributaries and industrial and municipal effluent, is
preferred for the collection of representative  water
samples.

For fast flowing rivers and streams, it may be difficult
to collect a mid-channel sample at a specific location;
health  and safety concerns  must dictate  where to
collect the sample.  For  low flowing streams, health
and safety concerns are reduced,  but obtaining a
specific representative location may be difficult.  For
low flow or intermittent streams, either locate an area
where a pool has been created or, in the most extreme
situations, use a cleaned trowel to create a pool in the
sediments for water to accumulate.

When  sampling a point source, two samples from
channel mid-depth are typically drawn:  one upstream
and one adjacent to, or  slightly downstream of, the
site PPE or the  point  of discharge.   Additional
samples may be required if multiple discharges or
additional tributaries are present. Structural features
such as dams, weirs, and bridges can cause changes in
the physical characteristics of a stream or river by
creating shallow pools. When water travel times are
long through these areas, sampling locations should be
established in them.   Some stream structures allow
overflow that significantly re-aerates oxygen-deficient
water.   This requires locations to  be close (both
upstream and downstream) to the structures in order to
measure the rapid and artificial increase in dissolved
oxygen (DO), which may cause the sample to be non-
representative.  Also collect a sample at a  location
well away from the aeration effect of the obstacle.

4.3.2  Lakes, Ponds, and
        Impoundments

The number of samples collected in these three types
of water bodies will  vary according to the size and
shape  of  the  water  body.    Stratification  from
temperature differences  is often present in these
bodies and is more prevalent than in rivers or streams.
Different layers can be detected visually as well as by
compiling a temperature profile.  In ponds and small
impoundments, a  single  vertical composite at the
deepest point would  be adequate to characterize the
water body. (The deepest point of a naturally formed
pond is generally near the center (although this may
need to be determined), and near  the  dam  in  an
impoundment.) Measure DO, pH, and temperature in
each aliquot of the vertical composite. Fewer mixing
zones require more samples to be collected. One way
to obtain representative samples is to divide the area
into a grid and then  perform systematic grid sampling
at each node. If stratified, collect a sample from each
stratum at each node location (three-dimensional or
stratified sampling).   Transect  sampling may also
apply.

Lakes  and larger impoundments require  several
vertical aliquots to be collected which can then  be
composited. Sampling locations may be determined
by  a  transect  or  grid.   Separate  composites  of
epilimnetic and hypolimnetic zones may be collected
if desired;  however,  a composite should consist of
several vertical aliquots collected at various depths.
Irregularly  shaped lakes may  require  additional
separate composite samples to be collected. Lakes
where discharges, tributaries, land use characteristics,
and other  such  factors may affect mixing,  water
quality and/or the  accuracy of representative water
body sampling may also require additional composite
samples.   Compositing is  discussed  further  in
Section 2.4.2.

Surface impoundments (such as wastewater lagoons)
which contain concentrated wastes are addressed in
U.S.  EPA  Superfund  Program  Representative

-------
Sampling Guidance,  Volume 4 - Waste, OSWER
Directive 9360.4-14. Precautions and concerns exist
when dealing with waste impoundments which are not
addressed in  general surface water  and sediment
sampling.

4.3.3   Estuaries

Estuaries are  areas where inland fresh water (both
surface water and ground water) mixes with oceanic
saline water.  Estuaries  are generally categorized  as
mixed, salt wedge, or oceanic, dependent upon inflow
and mixing properties. Determining estuary category
is critical to establishing sample locations. Estuaries
may be  classified as critical  areas,  wetlands,  or
fisheries,  and therefore also present special target
considerations.

Mixed estuaries are characterized by  homogenous
salinity in the water column and a gradual increase in
salinity toward  the  sea.  This type of estuary  is
typically shallow and well mixed.  Locating specific
sampling points, particularly in the vertical water
column, is not critical due to this mixing. Location
with respect to  the open  sea is more important  in
mixed estuaries.

Salt wedge estuaries are characterized by a significant
vertical increase in salinity and stratified fresh-water
flow along the surface. Density differential between
fresh and saline waters overrides any vertical mixing;
a salt wedge tapering inland moves horizontally with
the tide. Contamination  entering from upstream may
be missed if sampling into the salt wedge.

Oceanic estuaries exhibit salinity levels near to full-
strength ocean waters. Seasonally, fresh-water inflow
is  low compared to the fresh-saline water  mixing
occurring near, or at, the shoreline.

Sampling in estuary zones is typically performed on
successive slack  tides.   Estuary studies  can be
complex and are usually performed  in two  phases,
during both  wet  and dry periods. Estuary dynamics
can be affected  by fresh-water  inflow sources and
therefore  cannot  be  studied in  a  single  season.
Samples are generally collected at mid-depth in areas
where the depth is less than 10 feet, unless the  salinity
profile indicates the presence  of salinity  stratification.
In those cases,  samples  are collected  from each
stratum.   Measurements of dissolved oxygen  and
temperature should  accompany the  sampling.    In
estuaries where the depth is greater than 10 feet, water
samples may be collected at the one-foot depth, mid-
depth, and one foot from the bottom.
True salt-water bodies (e.g., oceans, salt lakes) are
rarely sampled at Superfund sites.  Salt-water bodies
would be sampled according to the fresh water and
estuary guidance above.    Review  stratification,
flow/turbulence,  and  other site  factors  prior  to
developing the sampling plan.  As with fresh water
bodies, sampling in estuaries can demonstrate current
and historical contamination through surface water
and sediment samples, respectively.  Be certain to
evaluate the  effect of the salt concentration on the
contaminants  of concern and their analytical methods
in order to accurately document a contaminant plume
or establish  connection  to  a  source or site.  Also
consider the  salt concentration and its compatibility
with sampling equipment.  For estuarine sampling, the
Van Dorn horizontal sampler is often utilized.

4.3.4  Wetlands

Wetlands are  considered a sensitive environment and
generally include swamps, marshes, bogs and similar
areas.   Wetlands can be  natural  or man-made.
Wetlands  can include fresh and estuarine water
systems and are commonly contiguous to open waters
(e.g., rivers, lakes, bays).  As defined in 40 CFR Part
230.3, as part  of Superfund's Hazard Ranking System
(HRS), wetlands are those  "areas that are inundated or
saturated by  surface  or ground water at a frequency
and duration sufficient to support, and  that under
normal conditions   do  support,  a  prevalence  of
vegetation typically adapted for life in saturated soil
conditions."  Wetlands are also identified using other
definitions, including a classification system of the
U.S. Fish and Wildlife Service (USFWS) and the
1989 Federal Manual for Identifying and Delineating
Jurisdictional Wetlands, as is used by the U.S. Army
Corps of Engineers.

National Wetlands Inventory maps use the USFWS
classifications.  These maps serve as  an excellent
starting point for identifying wetlands  at a site, but
should not be  used as the sole source of identification.
(A detailed comparison of the relationship between
the HRS and the USFWS definitions of wetlands is
addressed in  the U.S. EPA Hazard Ranking System
Guidance Manual,  OSWER Directive 9345.1-07,
Section A.2.) Where possible, an attempt should be
made to field verify and document (e.g., logbook,
photograph) the wetlands location and area.

In some instances, historical data may document the
presence of wetlands which no longer exist during the
site reconnaissance.  Attempt to determine whether
the wetlands were eliminated or filled, particularly if
the alteration was due to site activity.  Dredged or
                                                   36

-------
filled former wetlands may affect sampling design,
methods, and results due to the potential effects from
non-native soils, confined or void subsurface spaces,
or buried organic layers and on-site contaminants.

Special care should be taken when sampling wetlands
to collect surface water and sediment samples free of
vegetation and other organic matter or detritus.  As
with  other  surface water and sediment  samples,
consider curves and bends, slow versus fast flow,  and
depositional areas when selecting locations. Due to
the  slow movement of water through the vegetated
wetlands, contaminants may tend  to collect in
wetlands sediments. Wetlands may also serve as a
valuable source to  document historical contaminant
releases.     For   some   purposes  (e.g.,   HRS
documentation), an aqueous sample is preferred or
required to document contamination within wetlands,
therefore surface water samples should be  collected
where possible for all response action considerations.
As with other water bodies, wetlands can demonstrate
historical contamination through sediment samples,
current contamination through surface water samples,
and concern for future contamination if the wetlands
can be  documented to be the receiving body for a
contaminant drainage pathway or surface water route,
although not  currently exhibiting any  site-related
contamination.   The probable point of entry for a
tributary or drainage path into a surface water body
may be located within adjoining wetlands.   As a
sensitive environment, wetlands present special threat
and target considerations beyond those of other water
body systems.

Depending on the type of wetlands and the season,
wetlands  may  contain fresh or  salt  water,  and
saturated or dry sediments.  Follow the protocols  and
procedures  discussed  throughout  this  guidance
document for sampling each medium, respectively,
depending on the site-specific characteristics  of the
wetlands. Wetlands,  if periodically dry, should be
sampled during a wet period, if possible, to establish
the  wetlands sample as a sediment versus  a surface
soil. For complex sites with extensive surface water,
sediment and wetlands concerns, a wetlands expert
should be consulted for identification, delineation  and
sampling.
4.4    SEDIMENT SAM RLE
        COLLECTION
and  are  subject  to  variations  in  texture,  bulk
composition, water content, and pollutant content.
Therefore, large numbers of samples may be required
to characterize a small area.  Many sediment samples
along the cross-section of a river or stream need to be
collected  in order to  accurately  characterize  the
deposits. Generally, samples are collected at quarter
points  along the cross-section of the  water  body.
Aliquots can usually be combined  into  a  single
composite   sample except  for  those of  unlike
composition.  For small streams, one single sediment
sample can be collected at the main flow line of the
water body.  In most cases,  a sediment sample is
collected at the same location(s) as a surface water
sample.

Sediments in low flowing waters are largely the
products of erosion and may contain  a variety  of
organic matter. Sediment samples from ponds, lakes,
and reservoirs should be collected approximately in
the deepest point of the water body.  This is especially
applicable to reservoirs formed by impoundments of
rivers or streams.  Coarser grain sediments are found
near the  headwaters  of the  reservoir, while bed
sediments are composed of fine-grained materials
which  may  have an  increased  concentration  of
contaminants. Sediment sampling locations  can be
influenced  by  the  shape,  flow  pattern,   depth
distribution, and circulation of the water body.

Sediment samples from ponds  and lakes  can be
collected from each node  of the grid or transect set up
for  sampling surface water.  For streams or rivers,
collect a sediment sample in  at least two locations:
one  upstream  and one  adjacent to, or  slightly
downstream of, the site PPE or at point of discharge.
Consider depositional versus erosional areas against
the  objectives for sampling;  contaminants  tend  to
concentrate  in   the  fine-grained sediments  in
depositional zones.

Take care  to  minimize  disturbance   and  sample
washing as the sediment is retrieved through the water
column.  Fine fractions lost during sample collection
can result in a non-representative sample. Any liquid
collected   when   sampling  can  be  considered
representative of sediment conditions. Wet sediments
which  are to be analyzed while still wet should be
collected in rigid containers, not collected or stored in
bags.
As with water sampling, determine tidal influence and
its possible effect on  sediment sample collection.
Sediments are typically heterogenous in composition
                                                   37

-------
4.5    SAMPLE PREPARATION
                        4.5.2   Homogenizing
Sample  preparation  depends  on  the   sampling
objectives and analyses to  be performed.   Proper
sample preparation  and handling  maintain  sample
integrity.  Improper handling can render samples
unsuitable for analysis. For example, homogenizing
and compositing samples result in  a loss of volatile
constituents and are thus inappropriate methods when
volatile contaminants are of concern.  The effective
use of SOPs can ensure that the same methods are
used for all samples and by  all samplers.  Where
possible, the same person should sample all of one
matrix per water body to ensure similar methodology
in collection.   Sample  preparation for water and
sediments may include, but is not limited to:

•       Removing extraneous materials
•       Homogenizing
•       Splitting
•       Compositing
•       Final preparation
4.5.1  Removing
        Materials
Extraneous
During  sample  collection,  identify and  discard
materials from the sample which are not relevant or
vital for characterizing the site.  Avoid the collection
of floating or suspended debris (e.g., leaves, paper
trash, etc.) in the surface water flow or column. For
sediments, avoid collecting decaying or other organic
material, such as twigs, leaves, roots, and insects.
Avoid trash and other unrelated materials. Remove
the materials with the cleaned sampling tool, not with
your hand or other instrument which might cross-
contaminate the sample. The presence of extraneous
materials may introduce an error into the sampling or
analytical procedures.

Not all  external materials are extraneous, however.
For example, some  contaminants may be adsorbed
onto inert materials, such as fly ash or other industrial
by-products or waste,  which  settle onto the bottom
sediments. Collect samples of any material thought to
be a potential source of contamination. Discuss any
special  analytical   requirements  for  extraneous
materials  with  the  project  team   (e.g.,  project
management, geologist, chemist),  and notify  the
laboratory   of   any   special   sample   handling
requirements or method changes.
Homogenizing is the mixing or blending of a grab or
composite   sample  to  distribute   contaminants
uniformly  within  the  sample.   Ideally,  proper
homogenizing ensures that all portions of the sample
are equal  or identical  in  composition  and  are
representative  of  the  total  sample  collected.
Incomplete homogenizing thus introduces  sampling
error.  All samples to be composited or split should be
homogenized after all aliquots have been combined.

Homogenizing  generally  does not apply  to  water
samples; unless stratified, surface water is assumed to
be homogenous due to natural mixing.  If phases
occur, treat each stratum  as a unique homogenous
medium and sample each  separately.  The mixing of
sediments may  release  some contaminants into the
water phase of the sediment sample. If homogenizing
is required, manually mix  the sediment sample using
a spoon or scoop and a  tray or bucket constructed of
inert  or compatible materials (stainless  steel is
preferred).  Samples can also be homogenized using
a mechanically operated  stirring device as depicted in
ASTM  Standard  D422-63.   Do  not  homogenize
samples for volatile compound analysis.

4.5.3  Splitting

After collection, samples  are split into two or more
equivalent parts when two or more portions of the
same sample need to be analyzed separately.  Split
samples  are  most often  collected  in  enforcement
actions to compare sample results obtained by  EPA
with  those obtained by the potentially responsible
party. Split samples also provide measures of sample
variability  and analytical error (field replicates).
Homogenize the samples  before  splitting,  when
collecting only non-VOC sediment samples.  For each
parameter,  split water samples by alternately filling
sample collection jars for the sample and its split from
the same sampling device.  For sediment, alternate
spoonfuls of homogenized  sample between collection
jars.  Surface water and sediment samples for VOC
analysis should not be homogenized; instead, collect
two uniform samples concurrently from the same
location (collocated samples).

4.5.4  Compositing

Compositing is the process of physically combining
and homogenizing (if applicable) several individual
aliquots  of the  sample.   The field preparation
technique of compositing of  samples requires that
each discrete aliquot be equal, and that the aliquots be
                                                 38

-------
thoroughly homogenized.   Compositing samples
provides an "average" concentration of contaminants
over a  certain  number  of  sampling points, which
reduces both  the  number  of required laboratory
analyses and the sample variability. Compositing can
be a useful technique but must always be implemented
with caution. Compositing is not recommended where
volatile  organic  compounds   are  of  concern.
Compositing may dilute an isolated contaminant to
below  detection limits,  thus  masking  a possible
problem. Additional information  on compositing for
surface water and sediment sampling is provided in
Sections 2.4.2,  Composite Sample, and  4.3, Surface
Water Sample Collection.

4.5.5  Final Preparation

Obtain sample containers from a vendor that certifies
their decontamination/cleanliness.   Consider their
compatibility  with the  material being  sampled,
resistance to breakage,  volume, container color,
storage and transport, and decontamination procedures
(see U.S. EPA Compendium of ERT Surface Water
and Sediment  Sampling   Procedures   OSWER
Directive  9360.4-03).   Additional  information on
containers and  cleaning procedures is  available in
U.S.  EPA's  Specifications   and  Guidance  for
Obtaining Contaminant-Free  Sample  Containers,
OSWER Directive 9240.0-05. Volume and containers
will vary according to the parameter(s) to be analyzed.
Glass is appropriate for most sampling because it is
chemically inert to most substances, although some
metals may adhere to the sides of glass containers.
Glass is not recommended for samples containing
strong   alkali   solutions and hydrofluoric  acid.
Polyethylene plastic bottles are suitable for metals,
cyanide,  and   sulfide  in  water,   but  are   not
recommended for organic analyses since plasticizers
may leach into the sample. Amber glass bottles  help
preserve sample integrity  for extractable organic
constituents in water which may degrade in light, such
as hydrocarbons, pesticides, and petroleum residues.
Sample containers must be tightly capped in order to
prevent oxidation  from the air  and/or the loss  of
volatile components. Most sample aliquots for VOC
analysis are stored in 40-milliliter glass Teflonฎ
septum vials, which allow for easy  syringe removal of
the sample for analysis, without the loss of headspace
gases.  VOC sample containers must be completely
filled to the  top with no  air pockets.  Improper
decontamination of sampling equipment may result in
cross-contamination of samples.

Keep low and medium concentration surface water
and sediment samples  to be analyzed  for organic
constituents at not more than 4E C by using ice or
"blue ice" when shipping. This cooling is to retard the
transformation     of    contaminants    through
biodegradation or reaction while awaiting laboratory
analysis. If required, add any preservatives to specific
samples before  shipping.  The analytical laboratory
will  recommend   or  provide   any    chemical
preservatives  prior  to  sampling.    Follow  the
laboratory's instructions for quantity and timing of
preservative addition; many  laboratories will provide
the sample containers already chemically preserved.
Refer to the laboratory, as well as 40 CFR 136, and
the U.S. EPA Compendium of ERT Surface Water
and  Sediment  Sampling   Procedures,   OSWER
Directive  9360.4-03, for  actual  sample volumes,
appropriate containers,  and holding times.  Label all
sample containers in accordance  with the analytical
laboratory or Regional procedures and place them into
reclosable  plastic  bags  prior  to  packaging  for
shipment.  Package all samples in compliance with
current U.S. Department of Transportation (DOT) or
International  Air  Transport Association (IATA)
requirements. Be certain the sample container meets
these  requirements, and check the shipping/packing
instructions about preservatives.

Packaging should be performed by  someone trained in
current DOT shipping procedures.  Be certain all
containers  are  packaged  to prevent  breakage  or
leakage.   For all  samples, be certain to  maintain
secure chain-of-custody from collection to shipment
to the analytical laboratory.
4.6     EXAMPLE  SITE

4.6.1   Sampling

During Phase 1, soil samples were collected as grab
samples from shallow surface locations.  The sample
locations were cleared of surface  debris,  then  the
samples  were retrieved with disposable scoops and
placed directly into sample containers. During Phase
2, soil samples were  collected using trowels and split
spoon samplers.   The shallow soil  samples  were
collected in the  same manner as the Phase 1 soil
samples.  The subsurface soil samples were retrieved
from the split spoon sampler using a disposable plastic
scoop which transferred the soil into a stainless steel
bowl.  Several scoopfuls were collected along  the
length of the split spoon sampler and composited in
the bowl.  The composite sample was then transferred
directly into the sample container using a disposable
plastic scoop.
                                                  39

-------
Phase  1 and Phase 2 residential well ground-water
samples were collected  directly  from the taps of
homes, which used private wells near the site. Fifteen
monitoring wells were installed at the site with 4-inch
Schedule 40 PVC casing and 0.010 slot screen in
lengths appropriate to each well. Shallow wells were
drilled to approximately 40 feet below ground surface,
and  bedrock  contact  wells  were   drilled   to
approximately 55 to 60 feet.  Continuous split spoon
sampling was completed at each well location from 4
feet to  well completion depth.  Upon completion, all
monitoring   wells  were  developed  using   a
decontaminated submersible pump and flexible PVC
hose.

After development, the 15 on-site monitoring wells
were sampled for analysis of ground water.  Each
monitoring well was purged to obtain a representative
sample. Wells with sufficient yield were purged three
well volumes. Low-yielding wells were purged once
to dryness.

Each monitoring well was sampled after purging and
recovery.  Ground-water samples were collected using
dedicated disposable Teflonฎ bailers.  Each bailer
was attached to a clean nylon rope and introduced into
the well. After well sampling, a  hydraulic (pump) test
was performed to determine aquifer characteristics for
mathematical  modeling  of  potential  contaminant
plume  migration.  To generate  accurate gradient and
well location maps, the fifteen  newly  installed
monitoring wells were surveyed for vertical location
using feet above mean sea level (MSL) units.

Surface water  and  sediment  samples were also
collected as grab samples during Phase 1  and Phase 2.
Sampling activities occurred when  the intermittent
tributary  was flowing  in order to  obtain  water
samples.  Because of the shallow depth and narrow
breadth of the tributary  and Little  Creek,  samples
could be obtained by reaching into the near center in
the main flow line of the water  body  from the stream
bank.  The sampler stood downstream of the desired
sampling location and created as little disturbance of
the  streambank and water body as possible.  This
caution reduced the potential for cross-contamination
of the sample locations.

Sampling  proceeded  from the  most  downstream
location in Little Creek, to upstream, and the surface
water  aliquot  was  sampled  prior  to sediment
collection  at each location  to reduce entraining
suspended material into the water samples.  Cleaned
and labeled surface water sample containers were
placed directly into the flow of the water body for
sample collection.   The sediment samples were
collected (using dedicated disposable plastic scoops)
from the substrate directly beneath the location where
the water sample was retrieved. The sample material
was then transferred immediately into a clean, labeled
sample container.

All non-disposable equipment, including drill rig and
equipment, stainless steel bailers, submersible pumps,
water level indicators, and depth  sounders, were
decontaminated between sampling at each location
and prior to the first sampling event each day.

4.6.2  Sample Preparation

All sample containers were supplied by the contracted
analytical laboratory. Chemical preservation was also
provided by  the  laboratory through pre-preserved
bottleware.   Sample  containers  for surface  water
samples consisted of:

•       1-liter  polyethylene  bottles  for   total
        chromium, pre-preserved with reagent-grade
        nitric acid to result in, after sample addition,
        a pH of less than 2
•       1-liter polyethylene bottles for hexavalent
        chromium
•       1-liter polyethylene bottles for cyanide, pre-
        preserved with sodium hydroxide

Sample containers for sediments consisted of 8-ounce
glass jars with Teflonฎ caps for all parameters.

All samples were preserved to 4E C by placing them
in coolers packed with "blue ice"  immediately after
collection and during shipment. (The laboratory was
responsible for cooling and refrigeration of samples
upon arrival.)

The samples were packaged in compliance with IATA
requirements  for environmental samples.  Chain-of-
custody paperwork was prepared for the  samples.
Laboratory paperwork was completed as appropriate
and the samples were shipped to  the predesignated
laboratories for  analysis. Holding times  for total
chromium and cyanide are less than six months, but
hexavalent chromium has a holding time of less than
24 hours.  This was coordinated in advance with the
analytical laboratory   and  required daily ground
delivery of samples to the laboratory.
                                                   40

-------
                 5.0  QUALITY ASSURANCE/QUALITY CONTROL
5.1      INTRODUCTION

The  goal of representative  sampling is to obtain
analytical results that accurately depict site conditions
during a defined time interval.  The goal of quality
assurance/quality control (QA/QC) is to implement
correct methodologies which limit the introduction of
error into the sampling and analytical procedures, and
ultimately into the analytical data.

QA/QC samples evaluate three types of information:
1) the magnitude of site variation; 2) whether samples
were cross-contaminated during sampling and sample
handling procedures; and 3) whether a discrepancy in
sample results is a result of laboratory handling and
analysis procedures.  The QA/QC sample results are
used  to  assess the  quality  of analytical results  of
environmental samples collected from a site.
5.2     DATA CATEGORIES

EPA has established data quality objectives (DQOs)
which   ensure  that  the   precision,   accuracy,
representativeness, and quality of environmental data
are  appropriate  for  their   intended  application.
Superfund  DQO   guidance  defines  two  broad
categories  of  analytical data:   screening  and
definitive.

Screening data are  generated by rapid, less precise
methods of  analysis with  less  rigorous sample
preparation than definitive data.  Sample preparation
steps may be restricted to simple procedures such as
dilution  with  a solvent, rather than   elaborate
extraction/digestion and cleanup.  At least 10 percent
of the screening data are  confirmed using analytical
methods  and  QA/QC  procedures  and  criteria
associated  with definitive data.   Screening  data
without  associated  confirmation  data   are  not
considered to be data of known quality.   To  be
acceptable, screening data must include the following:
chain-of-custody, initial and continuing calibration,
analyte  identification,  and analyte  quantification.
Streamlined  QC requirements  are  the  defining
characteristic of screening data.

Definitive data are generated using rigorous analytical
methods (e.g.,  approved EPA reference methods).
These data are analyte-specific, with confirmation of
analyte identity and concentration.  Methods produce
tangible raw  data  (e.g.,  chromatograms, spectra,
digital values) in the form  of  paper printouts or
computer-generated electronic files.   Data may be
generated at the site or at an off-site location, as long
as the QA/QC requirements are satisfied. For the data
to be definitive, either analytical or total measurement
error must be determined. QC measures for definitive
data contain  all  of the elements associated  with
screening data, but also may include trip, method, and
rinsate blanks;  matrix spikes; performance evaluation
samples;   and   replicate   analyses   for  error
determination.

For further information on these  QA/QC objectives,
please refer to U.S. EPA's Data  Quality Objectives
Process for Superfund, 1993, pp. 42-44.
5.3     SOURCES OF ERROR

Identifying and quantifying the error or variation in
sampling and laboratory analysis can be difficult.
However, it is important to limit their effect(s) on the
data.  The four most common potential sources of data
error in surface water and sediment sampling are:

•        Sampling design
•        Sampling methodology
•        Sample heterogeneity
•        Analytical procedures

Refer to U.S. EPA's Data Quality Objective Process
for Superfund, for further discussion on error.

5.3.1   Sampling Design

Site variation includes the variation both in the types
and in the concentration  levels  of  contaminants
throughout a water body.  Representative sampling
should accurately identify and define this variation.
However, error can be introduced by the selection of
a sampling design which "misses" this variation. For
example,  a  sampling grid  with  relatively large
distances between  sampling points  or  a biased
sampling approach  (i.e., judgmental sampling) may
allow   significant   contaminant   trends   to   go
unidentified.   Surface water might have  multiple
strata;  failure  to  account  for  differences  in
composition  of  multiple  phases  can introduce
sampling error. The  sampling design must account for
all phases and strata which might contain hazardous
substances.
                                                  41

-------
The  sampling design should utilize approved  SOPs
and previously approved sampling designs to ensure
uniformity and comparability between samples. The
actual sample collection process should be determined
prior to sampling.  All samples should be collected
using a uniform surface area and/or depth to ensure
data comparability.  Sampling  equipment must be
standardized for similar sampling situations.

The sampling design should fulfill sampling and data
quality objectives.  Data quality objectives should be
built into the sampling design, including all necessary
QA/QC samples.

5.3.2  Sampling  Methodology

Sampling   methodology   and   sample   handling
procedures have possible sources of error, including:
cross-contamination from inappropriate use of sample
collection equipment,  unclean  sample containers,
improper sampling equipment decontamination, and
improper   shipment  procedures.     Standardized
procedures for collecting, handling, and  shipping
samples identify potential source(s) of error and help
minimize them.  Use SOPs to ensure that all  given
sampling  techniques  are  performed  in  the  same
manner, regardless of  the sampling team, date, or
location of sampling  activity.   Use  field blanks,
replicate samples, trip blanks,  and rinsate blanks
(discussed in Section 5.4) to identify errors due to
improper sampling methodology and sample handling
procedures.

Site   screening  methods might  employ  hazard
categorization kits or "cookbook" procedures requiring
interpretations based on  chemical reactions which
produce a color change.  The degree of subjectivity
inherent in interpretation, and the complexity of some
of the procedures, introduce a  significant source of
potential error.

5.3.3  Sample Heterogeneity

Sample heterogeneity is a potential source of error in
sediment sampling.  Unlike water,  sediment is rarely
a homogeneous medium. Sediments exhibit variations
with lateral distance and  depth.  This heterogeneity
may  also be present in the sample container unless the
sample was  homogenized in  the field or in the
laboratory. The laboratory uses  only a small aliquot
of the sample for analysis; poor reproducibility from
heterogenous samples  is a common error.  If the
sample is not properly homogenized, the analysis may
not be truly representative of the sample and of the
corresponding site.  Thorough homogenization of
samples  limits the error associated with  sample
heterogeneity.   (Note: Do not homogenize when
analyzing for VOCs.)

5.3.4  Analytical Procedures

Analytical  procedures may  introduce  errors from
laboratory cross-contamination, inefficient extraction,
and  inappropriate  methodology.    Matrix  spike,
laboratory  duplicate,  performance evaluation,  and
laboratory  control  samples  help  to   distinguish
analytical error from sampling error.
5.4    QA/QC SAMPLES

QA/QC samples are collected at the site or prepared
for or by the laboratory.  Analysis of QA/QC samples
provides information on the variability and usability of
sampling data,  indicates possible field sampling or
laboratory error,  and provides a basis  for future
validation and usability of the analytical data.  The
most common  field QA/QC  samples  are  field
replicate, collocated, background, and rinsate, field,
and trip blank samples. The most common laboratory
QA/QC samples are performance evaluation  (PE),
matrix spike (MS), and matrix spike duplicate (MSD)
samples.  QA/QC results may suggest the need for
modifying sample collection, preparation, handling, or
analytical procedures if the resultant data do not meet
site-specific  quality assurance objectives.

The following sections briefly describe the types of
QA/QC samples appropriate for surface water and
sediment sampling.

5.4.1   Field Replicate Samples

Field replicates, also referred to as field  duplicates and
split samples, are field samples obtained from one
sampling point, homogenized (where appropriate),
divided into separate  containers,  and  treated as
separate samples throughout  the remaining sample
handling and analytical processes.  (Splitting samples
for  surface  water  and sediments  is  discussed in
Section 4.5.3.)  Use replicate samples  to assess error
associated   with   sample  heterogeneity,  sample
methodology,  and  analytical procedures.    Field
replicates can also be used when determining total
error for  critical   samples   with   contamination
concentrations near the action level. In such a case, a
minimum of eight replicate samples is recommended
for valid statistical analysis. Field replicates may be
sent to two or more laboratories  or to  the  same
laboratory  as  unique  samples.    For total  error
                                                  42

-------
determination, samples should be analyzed by the
same laboratory. Generally, one field replicate per 20
samples per day is recommended.

5.4.2  Collocated Samples

Collocated samples are collected adjacent to the
routine field sample to determine local variability of
the sample location and contamination at the site.
Typically,  collocated  samples for  sediments are
collected side by side, but no more than 3  feet away
from  the  selected  sample  location.   Collocated
samples for surface water are collected from the same
location and depth.  Collocated samples are collected
and  analyzed  as discrete  samples;  they are not
composited.   Analytical  results from  collocated
samples can be used to assess site variation, but only
in the immediate sampling area. Because of the non-
homogeneous nature of sediment at sites, collocated
samples should not be used to assess variability across
a site and  are not recommended for assessing error.
Collecting  many samples can demonstrate variation in
sediments  in  a water  body.    Determine  the
applicability of collocated  samples on a site-by-site
basis.

5.4.3  Background Samples

Defining  background  conditions may  be difficult
because of natural variability  and the  physical
characteristics of the site, but it is important in order
to quantify true changes in contaminant concentrations
due  to  a  source or  site.   Defining  background
conditions is critical for avoiding false positives and
for enforcement purposes  in  naming  responsible
parties.  Background samples are collected upstream
of the area(s) of contamination  (either on or off site)
where there is little or no chance of migration of the
contaminants of concern.   Background samples
determine the natural composition of the surface water
and sediments and are considered "clean" samples.
They provide a basis for comparison of contaminant
concentration levels with samples collected on site.
Collect at least one background surface water and one
background sediment sample. Additional samples are
often warranted by site-specific factors such as natural
variability of local sediments, multiple sources, and
discharges from off-site facilities. Tidal influences
must  be considered when  selecting  a  background
location. Background samples may also be collected
to evaluate potential error associated with sampling
design,  sampling   methodology,  and   analytical
procedures.
5.4.4  Rinsate Blank Samples

A rinsate blank,  also referred to as an equipment
blank,  is used to assess  cross-contamination from
improper  equipment  decontamination  procedures.
Rinsate blanks are samples  obtained  by running
analyte-free water over  decontaminated sampling
equipment.  Any residual contamination should appear
in the rinsate sample data. Analyze the rinsate blank
for the same analytical parameters as the field samples
collected that day. Handle and ship the rinsate like a
routine  field sample.   Where dedicated sampling
equipment is not utilized, collect one rinsate blank per
type of sampling device per day.

5.4.5  Field Blank Samples

Field blanks are samples prepared in the field using
certified clean water (HPLC-grade water [carbon-free]
for organic analyses and deionized or distilled water
for inorganic analyses) or sand, which are submitted
to the laboratory for analysis. A field blank is used to
evaluate contamination or error associated  with
sampling      methodology,      preservation,
handling/shipping,  and   laboratory   procedures.
Handle, ship, and analyze a field blank like a routine
field sample.  Submit one field blank per day.

5.4.6  Trip Blank Samples

Trip blanks are samples prepared prior to going into
the field.   They  consist of  certified  clean  water
(HPLC-grade) or sand and are not opened until they
reach the laboratory.   Utilize trip blanks for volatile
organic analyses only.  Handle, transport,  and analyze
trip blanks in the  same manner as the other volatile
organic samples collected that day.  A trip blank
should be included with each shipment or two-day
sampling event. Trip blanks are used to evaluate error
associated with shipping and handling, and analytical
procedures.

5.4.7  Performance Evaluation/
        Laboratory  Control Samples

A performance evaluation (PE) sample evaluates the
overall error contributed by the analytical laboratory
and detects any bias in the analytical method being
used. PE samples contain known quantities of target
analytes manufactured under  strict quality control.
They are usually prepared by a third party under an
EPA certification program.  The samples are usually
submitted  "blind" to  analytical  laboratories  (the
sampling team knows the contents of the samples, but
                                                  43

-------
the laboratory does not).  Laboratory analytical error
may  be evaluated by the percent recoveries and
correct identification of  the components in the PE
sample. Note: Even though they are not available for
all analytes, analysis ofPE samples is recommended
in order to obtain definitive data.

A blind PE sample may be included in a set of split
samples provided to the  PRP.  The PE sample will
indicate PRP laboratory accuracy, which may be
critical during enforcement litigation.

A laboratory control  sample (LCS)  also contains
known quantities of target analytes in certified clean
water. In this case, the laboratory knows the contents
of the sample (the LCS  is usually prepared by the
laboratory). PE and LCS  samples are not affected by
waste matrix interference, and thus can provide a clear
measure of laboratory error.

5.4.8  Matrix Spike/Matrix Spike
        Duplicate Samples

Matrix spike and matrix  spike duplicate samples
(MS/MSDs) are field samples that are spiked in the
laboratory with a known concentration of  a  target
analyte(s) in order to determine percent recoveries in
sample  extraction.   The  percent recovery  from
MS/MSDs indicates the degree to which  matrix
interferences  will affect  the  identification  of  a
substance.  MS/MSDs can also be used to monitor
laboratory performance.  When four or more pairs of
MS/MSDs are analyzed, the  data  obtained may be
used to evaluate  error due to laboratory bias and
precision.  Analyze one MS/MSD pair to assess bias
for every  20  samples of each matrix, and use the
average percent recovery for the  pair.   To assess
precision, analyze at least eight matrix spike replicates
from the  same sample, and determine the standard
deviation and the coefficient of variation. MS/MSDs
are recommended for screening data and are required
as one of several methods for determining analytical
error for definitive data.  Since the MS/MSDs are
spiked  field samples, provide sufficient volume for
three separate analyses (i.e., triple volume).

5.4.9  Laboratory  Duplicate Samples

A laboratory duplicate is a sample that undergoes
preparation and analysis twice.  The laboratory takes
two  aliquots of one sample  and analyzes them  as
separate samples.  Comparison of data from the two
analyses   provides  a  measure  of  analytical
reproducibility within a sample set.  Discrepancies in
duplicate analyses may indicate poor homogenization
in the field or other sample preparation error, either in
the field or in the laboratory.
5.5    EVALUATION OF
        ANALYTICAL ERROR

The acceptable level of error  in sampling data is
determined by the intended use of the data and the
sampling objectives, including: the degree of threat to
public health, welfare, or the environment; selected
action levels;  litigation concerns;  and  budgetary
constraints.

Error may be determined with replicate samples. To
evaluate the total error of samples with contaminant
concentrations near the selected action level, prepare
and analyze a minimum of eight replicates of the same
sample.  Analytical data from replicate samples also
serve as a quick  check on errors  associated  with
sample  heterogeneity,  sample  methodology,   and
analytical procedures.  Different analytical  results
from two or more replicate samples could indicate
improper  sample  preparation  (e.g.,  incomplete
homogenization),  or improper sample  handling,
shipment, or analysis.

Although a quantified  confidence  level may be
desirable, it may not always be possible.  A  95%
confidence level (5% acceptable error)  should be
adequate for most Superfund activities. Note that the
use of confidence levels is based on the assumption
that a sample is homogeneous.
5.6    CORRELATION  BETWEEN
        FIELD SCREENING RESULTS
        AND DEFINITIVE
        LABORATORY RESULTS

One cost-effective approach for delineating the extent
of site contamination is to correlate inexpensive field
screening  data and other field measurements  with
definitive laboratory results.  The relationship between
the two methods can then be described by a regression
analysis.  The resulting  equation  can be used to
predict laboratory results based on field  screening
measurements.  In this manner, cost-effective  field
screening results may be used in conjunction with off-
site laboratory analysis.

Statistical regression involves developing an equation
that relates two or more variables at an acceptable
level of correlation. In this case, the two variables are
                                                 44

-------
field  screening results  and  definitive  laboratory
results.    The  regression equation  can be used to
predict a laboratory value based on the results of the
screening device.  The model can  also be used to
place   confidence  limits   around   predictions.
Additional discussion of correlation and regression
can be found in most introductory statistics textbooks.
A simple linear regression equation can be developed
on many calculators or computer databases. Consult
a statistician to check the accuracy of more complex
models.

Evaluation of the accuracy of a model relies in part on
statistical correlation, which involves computing an
index  called  the  correlation coefficient  (r)  that
indicates the degree and nature  of the relationship
between two or more sets of values.  The correlation
coefficient ranges  from  ! 1.0 (a  perfect inverse, or
negative, relationship), through 0 (no relationship), to
+1.0 (a perfect direct, or positive, relationship).  The
square  of the correlation coefficient, called the
coefficient of  determination,  or  simply R2, is an
estimate of the proportion of variance in the dependent
variable.  The value of an acceptable coefficient of
variation depends  on the sampling objectives and
intended data uses.  As a rule of thumb,  statistical
relationships should have an R2 value of at least 0.6 to
determine  a reliable model.   However,  for health
assessment purposes, the acceptable R2 value may be
more stringent (e.g., 0.8).  Analytical calibration
regressions have an R2 value of 0.98 or greater.

Field screening data can be used to predict laboratory
results if there  is an acceptable correlation between
them. The predicted values can be located on a base
map and contoured. These maps can be examined to
evaluate the estimated extent of contamination and the
adequacy of the sampling program.
5.7     EXAMPLE SITE

5.7.1   Data Categories

Screening  data  which  generate   non-definitive,
unconfirmed  results were used to select analytical
parameters  and  samples to  be  sent for off-site
laboratory confirmation analysis. Samples were sent
to the analytical  laboratory under protocols which
provided  definitive data.   The rigorous  laboratory
analyses  provided  definitive identification   and
quantitation of contaminants.
5.7.2   Sources of Error

All direct reading instruments were maintained and
calibrated  in  accordance with their  instruction
manuals. Many of these instruments are class-specific
(e.g., volatile organic vapors) with relative response
rates  that  are  dependent  on  the  calibration  gas
selected.   Instrument  response to  ambient  vapor
concentrations may differ by  an order of magnitude
from response to calibration standards.  If compounds
of interest are known, site-specific standards may be
prepared.

The number and location of initial field samples were
based on observation and professional judgment (as
outlined in Section 2.13.5). Field standard operating
procedures, documented in the  site sampling plan,
established   consistent  screening  and sampling
procedures among all sampling personnel, reducing
the chances for variability and error during sampling.
Site briefings were conducted prior to all sampling
and screening  events to review the use of proper
screening and sampling techniques.

Other  steps  taken  to  limit  error included proper
sample  preparation, adherence  to sample  holding
times,  and  the  use  of proper IATA shipment
procedures.  All off-site laboratory sample analyses
were  performed using EPA  standard methods  and
protocols.

5.7.3   Field QA/QC Samples

Field QA/QC samples were collected during surface
water and sediment sampling at the ABC Plating  site.
One each of field duplicates were collected for surface
water and sediment, respectively, plus duplicates for
other  media.  Rinsate  blanks were collected from
ground-water  and  soil sampling  equipment after
decontamination by  pouring deionized water through
the respective piece of equipment and  then  into a
sample  container.   The field duplicates and blanks
were  preserved  and  prepared  as  "regular"  field
samples.   A trip blank for VOC  analysis  and a
performance evaluation (PE) sample for metals were
sent to the laboratory. (The PE sample is not affected
by matrix interferences.) The trip blank was provided
by the laboratory (pre-filled and preserved) and  sent
with the sample containers prior to sample collection.
One trip blank  per   day  was submitted to  the
laboratory.  Additional volume was  collected  and
provided to the laboratory for matrix spike/matrix
spike  duplicate analyses for one per  ten sample
locations for each medium.
                                                   45

-------
5.7.4  Laboratory QA/QC

Instructions  on  matrices,  target compounds,  and
QA/QC criteria of particular interest were provided to
the laboratory to help ensure that analytical results
met the required objectives. The laboratory analyzed
for metals using the methods of inductively coupled
plasma  (ICP) spectrometry  and atomic absorption
(AA).  Two methods were conducted for hexavalent
chromium: Method 7196, a colorimetric method, and
Method  2185,  a chelation  method.   These  two
methods were utilized in an attempt to better quantify
hexavalent results.  The presence of cyanide  was
confirmed in the laboratory using total and amenable
cyanide analyses (colorimetric manual Method 9010).
                                                 46

-------
                    6.0  DATA PRESENTATION AND ANALYSIS
6.1     INTRODUCTION
6.4    CONTOUR MAPPING
Data  presentation  and  analysis  techniques  are
performed with analytical or field screening results.
The  techniques  discussed  below can be  used to
compare analytical  values, to evaluate  numerical
distribution of data, and to reveal the location of "hot
spots" and the extent of contamination at a site. The
appropriate methods to present and analyze sample
data depend on the sampling objectives, the number of
samples collected, the sampling approaches used, and
other considerations.
6.2    DATA POSTING

Data posting involves the placement of sample values
on a site base map or cross-section. Data posting is
useful for displaying the distribution of sample values,
visually depicting the location of contaminants with
associated assessment data.  Data posting requires
each sample to have a specific location (e.g., x, y, and
sometimes  z  coordinates).   Ideally,  the sample
coordinates are surveyed values or marked sampling
locations facilitating placement on a scaled map.  Data
posting is useful for depicting concentration values for
both surface water and sediments.
Contour maps can depict contaminant concentration
values in surface waters and sediments throughout the
water body. This method may be useful for sediment,
but is not typically used for surface water.  Contour
mapping requires an accurate, to-scale base map of the
site.  After data  posting sample values  on  the base
map, insert contour lines (or isopleths) at a specified
contour interval, interpolating values between sample
points. Contour lines can be drawn manually or can
be generated by computer using contouring software.
Although the  software makes the contouring process
easier, computer programs have a limitation:  as they
interpolate between  data points,  they  attempt  to
"smooth" the values by fitting contour intervals to the
full range of data values. This can result in a contour
map that does not accurately represent  general site
contaminant trends.  Typical Superfund sites have low
concentration/non-detect areas and "hot spots."  If
there is a big difference in concentration between the
"hot spot" and the surrounding  area, the computer
contouring program, using  a contour interval that
attempts to smooth the "hot spots," may eliminate
most of the  subtle site features and general trends.
Contour mapping is generally best used with non-
flowing, static water bodies, or over large areas.
6.3    CROSS-SECTION/FENCE
        DIAGRAMS

Cross-section diagrams (two-dimensional) and fence
diagrams (three-dimensional) depict layers or phases
of contaminants in the surface waters or sediments of
rivers, lakes, and impoundments. Two-dimensional
cross-sections  may be  used to illustrate  vertical
profiles  of contaminants  in  surface  water  and
sediment.   Three-dimensional fence  diagrams are
often  used to interpolate data between sampling
locations,  particularly where contaminants  do not
form horizontal layers. Both cross-sections and fence
diagrams can provide useful visual interpretations of
contaminant concentrations and migration.
6.5    STATISTICAL GRAPHICS

The distribution or spread of the data set is important
in determining which statistical techniques to use.
Common statistical analyses, such as the t-test, rely on
normally  distributed  data.    The histogram  is  a
statistical bar graph which displays the distribution of
a data set. A normally distributed data set takes the
shape of a bell curve, with the mean and median close
together about halfway between the maximum  and
minimum  values.    A  probability  plot  depicts
cumulative percent against the concentration of the
contaminant of concern.  A normally distributed data
set, when  plotted as a probability plot, appears as a
straight line.  A histogram or probability plot can be
used to see trends and anomalies in the data from a
contaminant source prior to conducting more rigorous
forms  of  statistical  analysis.  As  with  contour
mapping, statistical data interpretation applications are
typically used for sediment analysis.
                                                  47

-------
6.6    RECOMMENDED DATA
        INTERPRETATION METHODS

The data interpretation methods chosen depend on
project-specific considerations, such as the number of
sampling  locations  and their  associated  range in
values.  Data which are dissimilar in composition
should not be compared using statistical interpretation
methods.  Data posting, screening, and sampling data
sheets, and cross-section/fence diagrams  may  be
appropriate.  A site showing  extremely  low data
values (non-detects), with significantly higher values
(e.g., 5,000 ppm) from neighboring hot spots and little
or no concentration gradient in between, does not lend
itself to contour mapping.  Data posting would be
useful at  such a site to illustrate hot spots and clean
areas.
6.7    EXAMPLE SITE

A water table contour map was generated with the
water  level  data  for  the  shallow  overburden
monitoring wells.  This indicated a westward flow
direction, which generally coincides with the surface
topography.

All  surface  water and  sediment  samples  were
analyzed for total chromium and cyanide.  Cyanide
and chromium were not found above the  50 ppm
detection limit in any of the surface water or sediment
samples. Chromium was detected in soil and ground-
water samples at the site.
The  rate  of  chromium contaminant migration in
ground water and the potential long-term impact to
nearby residential wells  was  estimated  using  a
mathematical model which  included  worst  case
assumptions   and   evaluated   attenuation   of
contaminants through  soil and ground water.   The
OSC concluded that the potential for residential well
contamination was minimal  and  therefore,  the
potential for contamination of surface water through
the discharge of ground water was also considered
minimal.    Removal   of soil,  the  source  of
contamination, was recommended. This decision met
the Phase 2 objective  of  establishing early action
options and consideration of long-term remediation
requirements for ground water.

All containers of wastes were removed from the site.
Soil  treatment/disposal  was  completed using the
existing  grid  design.   Cells were  sampled  and
designated as clean or excavated. Excavated material
was stockpiled while treatment/disposal options were
evaluated.  Excavated cells were filled with stone and
clean soil. Composite sampling in each cell verified
cleanup, using an action level of 100 mg/kg chromium
in the soil  composite.   (The clean-up level was
established based on the earlier mathematical model
and soil attenuation calculations.) The soil response
served as an early action to meet the Phase 3 objective
originally established for the site.
                                                  48

-------
      APPENDIX A -- Example of Flow Diagram For Conceptual Site Model
                           Figure A-1
   Migration Routes  of a Gas Contaminant
              from Origin to Receptor
Gas
Pathway
from
origin
> Air
condc
solldl
Change of
contaminant
state In
pathway
insatlon
> Liquid
* *
t. ^^orป
> OaS
^> Solid
Flcatlon




Final
pathway
to receptor
> SO
> sw
> SO
> AT

^ /A -I
> sw
p-> SO

I > sw

Receptor
Human
G,D
G,D
I,D
I,D
G,D
G,D
G,D
Ecological Threat
Terrestrial
G,D
G,D
I,D
I,D
I,D
G,D
G,D
Aquatic
N/A
G,D
N/A
N/A
G,D
N/A
G,D
        *  May be a transformation product
        ** Includes vapors
Receptor Key

D  - Dermal Contact
]  - Inhalation
G  - Ingestlon
N/A = Not Applicable
Pathway Key

Al -Air
SO - Soil
SW - Surface Water
(Including sediments)
GW - Ground Water
                             49

-------
               Figure A-2
Migration Routes of a Liquid Contaminant
         from Origin to Receptor
Change of
Original state Pathway contaminant
of contaminant from state In
of concern* origin pathway
Liquid
* May be a t
** Includes v
> Liquid
-t SW > ftfl<5**
solidification SฐllCl
k Of"l k 1 i n 1 1 n H —
' OU * L-LUU-LU
leachate,
Infiltration
—> AT t fine**
A\J. P Vjdo —
ransformation product
apors

Final
pathway
to receptor
> SW
h AT
* Al
k CIA/
" oW
^ SW
> so
>> SW
> GW
^ SO
^ AI
* SW
Receptor
Human
G,D
I,D
G,D
G3D
Ecological Threat
Terrestrial
G,D
I,D
G,D
G,D
Aquatic
G,D
N/A
G,A
G,D

G,D
G,D
G,D
G,D
G,D
N/A
N/A
G,D
N/A

G,D
I,D
G,D
G,D
I,D
G,D
N/A
N/A
G,D

Receptor Key
D - Dermal Contact
I - Inhalation
G - Ingestlon
N/A - Not Applicable
Pathway Key
AI - Air
SO - Soil
SW - Surface Water
(Including sediments)
GW - Ground Water
                 50

-------
                         Figure A-3
      Migration  Routes of a Solid Contaminant
                from Origin to Receptor
 Original state
 of contaminant
 of concern*
Solid
            AI
               partlculates/
                 dust
            Solid
sw
^  Solid

->  Liquid
            SO
                         Gas
                            **
            Solid
                        Liquid
 * May be a transformation product
 ** Includes vapors
AI
SW
so

sw

sw
Receptor Key
D - Dermal Contact
I - Inhalation
G - Ingestlon
N/A - Not Applicable

Pathway Key
AI . Air
SO - Soil
SW - Surface Water
(Including sediments)
GW - Ground Water
Receptor
Human
I,D
G,D
G,D
Ecological Threat
Terrestrial
I,D
G,D
G,D
Aquatic
N/A
G,D
N/A
G,D
G,D
G,D
G,D
G,D
G,D
so
AI
SW
SO
SO
sw
G3D
I,D
G,D
G,D
G,D
G,D
G,D
G,D
I,D
G,D
G,D
G,D
N/A
G,D
N/A
N/A
G,D
N/A
N/A
N/A
G,D
                            51

-------
                                           References
American Public Health Association.  1989.  Standard Methods for the Examination of Water and Wastewater.
        Seventeenth Edition. Washington, DC.

American Society for Testing and Materials (ASTM).  1989. 1989 Annual Book ofASTM Standards - Volumes
        11.01 and 11.02, Water and Environmental Technology. American Society for Testing and Materials,
        Philadelphia, Pennsylvania.

Ford, Patrick I, and Paul J. Turina. July 1984. Characterization of Hazardous Waste Sites — A Methods Manual.
        Volume  I  -- Site Investigations.   EPA/600/S4-84/075.   U.S. Environmental Protection Agency.
        Environmental Monitoring Systems Laboratory. Las Vegas, Nevada.

Kittrells, F.W. 1969. A Practical Guide to Water Quality Studies of Streams. U.S. Federal Water Pollution Control
        Administration.  Washington, DC.

Krajca, Jaromil M.  1989.  Water Sampling.  Ellis Horwood Ltd., Chichester, England

National Research Council.  1990. Managing Troubled Waters - The Role of Marine Environmental Monitoring.
        Committee on a Systems Assessment of Marine Environmental Monitoring. National Academy Press,
        Washington, DC.

New Jersey Department of Environmental Protection, Hazardous Waste Programs.  February 1988. Field Sampling
        Procedures Manual.

Pavoni, Joseph L.  1977.  Handbook of'Water Quality Management Planning. Van Rostrand Reinhold Co., New
        York, New York.

Tchobanoglous, George, and Edward D. Schroeder.  1985. Water Quality - Characteristics, Modeling, Modification.
        Addison-Wesley Publishing Co., Reading, Massachusetts.

U.S. Environmental Protection Agency.  January 1995.  Quality Assurance Sampling Plan for Environmental
        Response  (QASPER),  User's Guide.  (Based on Office of Solid Waste  and Emergency Response
        Directive 9360.4-01.)

U.S. Environmental Protection Agency. 1995a. Superfund Program Representative Sampling Guidance, Volume
        1 — Soil.  Office of Solid Waste and Emergency Response Directive 9360.4-10.

U.S. Environmental Protection Agency. 1995b. Superfund Program Representative Sampling Guidance, Volume
        4 — Waste.  Office of Solid Waste and Emergency Response Directive 9360.4-14.

U.S. Environmental Protection Agency.  September 1993.  Data Quality Objectives Process for Superfund.  Office
        of Emergency and Remedial Response Directive 9355.9-01.

U.S. Environmental Protection Agency.  November 1992. Hazard Ranking System Guidance Manual. Office of
        Solid Waste and Emergency Response Directive 9345.1-07

U.S. Environmental Protection Agency.  September  1992.  Guidance for Performing Site Inspections  Under
        CERCLA. Office of Solid Waste and Emergency Response Directive 9345.1-05.

U.S. Environmental Protection Agency.  1992.  Specifications and Guidance for Obtaining Contaminant-Free
        Sample Containers. Office of Solid Waste and Emergency Response Directive  9240.0-05.
                                                 52

-------
U.S. Environmental Protection Agency.  January 1991. Compendium of ERT Surface Water and Sediment Sampling
        Procedures.  Office of Solid Waste and Emergency Response Directive 9360.4-03.

U.S. Environmental Protection Agency.  1990. Samplers Guide to the Contract Laboratory Program. Office of
        Solid Waste and Emergency Response Directive 9240.0-06.

U.S. Environmental Protection Agency.  1986. Manual-Sampling for Hazardous Materials, Parti.

U.S.  Environmental Protection Agency, Region 3 Remedial Engineering Management.  1986. REM III Program
        Guidelines. Ebasco Services, Incorporated, Langhorne, Pennsylvania.

U.S. Environmental Protection Agency.  1983. Characterization of Hazardous Waste Sites - A Methods Manual.
        Volume II — Available Sampling Methods.  EPA/600/4-83/040. Environmental Monitoring Systems
        Laboratory, Las Vegas, Nevada.
                                                 53

-------
                                              OSWER Directive 9360.4-16
                                                      EPA xxx/x-xx/xxx
                                                          PBxx-xxxxxx
                                                        December 1995
           SUPERFUND PROGRAM

REPRESENTATIVE SAMPLING GUIDANCE


    VOLUME 5: WATER AND SEDIMENT

           PART II--  Ground Water

                     Interim Final
               Environmental Response Team

          Office of Emergency and Remedial Response
         Office of Solid Waste and Emergency Response

            U.S. Environmental Protection Agency
                  Washington, DC 20460

-------
                                               Notice
This document has been reviewed in accordance with U.S. Environmental Protection Agency policy and approved
for publication.

The  policies and  procedures established in this document are intended solely for the guidance of government
personnel for use in the Superfund Program. They are not intended, and cannot be relied upon, to create any rights,
substantive or procedural, enforceable by any party in litigation with the United States.  The Agency reserves the right
to act at variance with these policies and procedures and to change them at any time without public notice.

For more information on Ground-Water Sampling procedures, refer to the U.S. EPA Compendium ofERT Ground-
Water Sampling Procedures, OSWER Directive 9360.4-06.  Topics covered in this compendium include: sampling
equipment  decontamination; ground-water monitoring well  installation, development, and sampling; soil gas
sampling; water level measurement; controlled pump testing; slug testing.

Please note that the procedures in this document should be used only by individuals properly trained and certified
under a 40-hour hazardous waste site training course that meets the requirements set forth in 29 CFR 1910.120(e)(3).
This document should not be used to replace or supersede any information obtained in a 40-hour hazardous waste site
training course.

Questions,  comments, and recommendations are welcomed  regarding the Superfund Program Representative
Sampling Guidance, Volume 5 —  Water and Sediment, Part II —  Ground Water. Send remarks to:


                                       Mr. William A. Coakley
                             Chairman, Representative Sampling Committee
                                           U.S. EPA-ERT
                                  Rantan Depot - Building 18, MS-101
                                       2890 Woodbridge Avenue
                                        Edison, NJ 08837-3679

-------
                                           Disclaimer


This document has been reviewed under U.S. Environmental Protection Agency policy and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or recommendation for use.

The following trade names are mentioned in this  document:

Teflonฎ is a registered trademark of E.I. DuPont de Nemours and Company of Wilmington, Delaware

Geoprobeฎ is a registered trademark of Geoprobe, Inc. of Salina, Kansas

Gilianฎ is a registered trademark of Gilian Instrument Corporation  of Wayne, New Jersey
                                                in

-------
                                          Contents


Notice	  ii

Disclaimer  	iii

List of Figures  	  vii

List of Tables  	  vii

1.0     INTRODUCTION	  1

       1.1      OBJECTIVE AND SCOPE	  1
       1.2      UNIQUE CHARACTERISTICS OF GROUND WATER	  1
       1.3      REPRESENTATIVE SAMPLING 	  2
       1.4      REPRESENTATIVE SAMPLING OBJECTIVES 	  2
               1.4.1    Identify Contamination and Determine Hazard  	  2
               1.4.2    Establish Imminent or Substantial Threat	  2
               1.4.3    Determine Long-Term Threat  	  3
               1.4.4    Develop Containment and Control Strategies	  3
               1.4.5    Evaluate Treatment Options	  3
       1.5      CONCEPTUAL SITE MODEL	  3
       1.6      OVERVIEW OF GROUND-WATER MONITORING WELL INSTALLATION AND
               GROUND-WATER MODELING	  5
               1.6.1    Ground-Water Monitoring Well Installation	  5
               1.6.2    Ground-Water Modeling  	  6
       1.7      EXAMPLE SITE	  6

2.0     GROUND-WATER SAMPLING DESIGN 	  7

       2.1      INTRODUCTION	  7
               2.1.1    Pre-Sampling Plan Investigation 	  7
               2.1.2    Types of Information Provided by Ground-Water Sampling Assessment	  8
               2.1.3    Site Reconnaissance	  9
       2.2      PARAMETERS OF CONCERN, DATA QUALITY OBJECTIVES, AND QUALITY
               ASSURANCE MEASURES	  10
               2.2.1    Parameters of Concern	  10
               2.2.2    Data Quality Objectives	  10
               2.2.3    Quality Assurance Measures  	  11
       2.3      REPRESENTATIVE GROUND-WATER SAMPLING APPROACHES
               AND SAMPLE TYPES	  11
               2.3.1    Judgmental Sampling	  11
               2.3.2    Random, Systematic Grid, and Systematic Random Sampling	  12
               2.3.3    Grab versus Composite Sample Types	  12
       2.4      SAMPLING PLAN	  12
       2.5      EXAMPLE SITE	  13
               2.5.1    Background  	  13
               2.5.2    Site History and Reconnaissance	  15
               2.5.3    Identification of Parameters of Concern	  15
               2.5.4    Sampling Objectives 	  16
               2.5.5    Selection of Sampling Approaches	  16
               2.5.6    Sampling Plan  	  16
                                              IV

-------
3.0     FIELD ANALYTICAL SCREENING, SAMPLING EQUIPMENT, AND GEOPHYSICAL
       TECHNIQUES	  18

       3.1     FIELD ANALYTICAL SCREENING	  18
               3.1.1   Flame lonization Detector 	  18
               3.1.2   Photoionization Detector	  18
               3.1.3   Gas Chromatograph	  19
               3.1.4   Geoprobe	  19
               3.1.5   Soil Gas Technique 	  19
               3.1.6   Field Parameter Instruments	  19
               3.1.7   X-Ray Fluorescence	  19
       3.2     GROUND-WATER SAMPLING EQUIPMENT 	  20
               3.2.1   Bailer	  20
               3.2.2   Hydraulic Probe	  20
               3.2.3   Air-Lift Pump  	  20
               3.2.4   Bladder Pump  	  20
               3.2.5   Rotary Pump 	  21
               3.2.6   Peristaltic Pump	  21
               3.2.7   Packer Pump 	  21
               3.2.8   Syringe Sampler	  21
               3.2.9   Ground-Water Sampling Equipment Selection Factors	  21
       3.3     GEOPHYSICAL METHODS	  22
               3.3.1   Surface Geophysics 	  22
..

                      Borehole Geophysics ................................................  23
               3.3.3   Geophysical Techniques for Ground- Water Investigations ....................  25
       3.4     EXAMPLE SITE [[[  26
               3.4.1   Selection of Field Analytical Screening Techniques .........................  26
               3.4.2   Selection of Sampling Equipment  ......................................  26
               3.4.3   Selection of Geophysical Methods ......................................  26

4.0     GROUND-WATER SAMPLE COLLECTION AND PREPARATION .......................  31

       4.1     INTRODUCTION [[[  31
       4.2     STATIC WATER LEVEL [[[  31
       4.3     WELL PURGING [[[  31
               4.3.1   Stabilization Purging Techniques .......................................  32
               4.3.2   Wells that Purge Dry ................................................  32
       4.4     GROUND-WATER SAMPLE COLLECTION ...................................  32
       4.5     GROUND-WATER SAMPLE PREPARATION ..................................  33
               4.5.1   Filtering [[[  34
               4.5.2   Homogenizing/Aliquotting ............................................  34
               4.5.3   Splitting [[[  34
               4.5.4   Final Preparation [[[  34
       4.6     EXAMPLE SITE [[[  34
               4.6. 1   Sample Collection ..................................................  34
               4.6.2   Sample Preparation  .................................................  37

5.0     QUALITY ASSURANCE/QUALITY CONTROL (QA/QC)  ...............................  38

       5. 1     INTRODUCTION [[[  38

-------
       5.4     QA/QC SAMPLES 	  39
              5.4.1   Field Replicate Samples	  39
              5.4.2   Background Samples	  39
              5.4.3   Rinsate Blank Samples	  40
              5.4.4   Field Blank Samples  	  40
              5.4.5   Trip Blank Samples	  40
              5.4.6   Performance Evaluation/Laboratory Control Samples 	  40
              5.4.7   Matrix Spike/Matrix Spike Duplicate Samples	  40
              5.4.8   Laboratory Duplicate Samples	  41
       5.5     EVALUATION OF ANALYTICAL ERROR	  41
       5.6     CORRELATION BETWEEN FIELD SCREENING RESULTS AND DEFINITIVE
              LABORATORY RESULTS 	  41
       5.7     EXAMPLE SITE	  42
              5.7.1   Data Categories 	  42
              5.7.2   Sources of Error	  42
              5.7.3   Field QA/QC Samples	  42
              5.7.4   Laboratory QA/QC 	  42

6.0     DATA PRESENTATION AND ANALYSIS	  43

       6.1     INTRODUCTION	  43
       6.2     DATA POSTING 	  43
       6.3     CROSS-SECTION/FENCE DIAGRAMS	  43
       6.4     CONTOUR MAPPING	  43
       6.5     WELL LOCATION MAP	  43
       6.6     STATISTICAL GRAPHICS	  43
       6.7     RECOMMENDED DATA INTERPRETATION METHODS 	  44
       6.8     EXAMPLE SITE	  44

Appendix A    EXAMPLE OF FLOW DIAGRAM FOR CONCEPTUAL SITE MODEL	  45

References  	  48
                                             VI

-------
                                         List of Figures






1       Conceptual Site Model 	  4




2       Site Sketch: ABC Plating Site  	  14




3       Soil Boring/Monitoring Well Completion Log  	  36




A-l     Migration Routes of a Gas Contaminant  	  45




A-2     Migration Routes of a Liquid Contaminant	  46




A-3     Migration Routes of a Solid Contaminant	  47







                                         List of Tables






1       Applicability of Surface Geophysical Techniques to Ground-Water Investigations	  27




2       Advantages and Disadvantages of Surface Geophysical Techniques to Ground-Water Investigations .  28




3       Applicability of Borehole Geophysical Techniques to Ground-Water Investigations	  29




4       Advantages and Disadvantages of Borehole Geophysical Techniques to Ground-Water Investigations  30
                                                 vn

-------
                                     1.0   INTRODUCTION
1.1     OBJECTIVE AND SCOPE

This is Part  II of the fifth volume in a series of
guidance documents that assist  Superfund Program
Site  Managers,  On-Scene  Coordinators  (OSCs),
Remedial Project Managers (RPMs), and other field
staff in obtaining representative samples at Superfund
sites.  The objective of representative sampling is to
ensure that a sample or a group of samples accurately
characterizes  site conditions.

Most hazardous waste site investigations utilize some
form  of a ground-water  sampling or monitoring
program to fully characterize the  nature and extent of
contamination. Because site conditions may  differ,
experienced hydrogeologists and geochemists should
be consulted  to establish the most suitable types of
sampling and monitoring for each site.

The  purpose of  this  document  is  to address
representative  ground-water sampling.  Ground-water
modeling and monitoring well installation are briefly
introduced  but are not  addressed in detail in  this
document. References on these topics are provided in
Section 1.6

The representative ground-water sampling principles
discussed in this document are applicable throughout
the Superfund Program.  The following chapters will
help field personnel to assess available information,
select an appropriate sampling approach and design,
select   and  utilize  field  analytical/geophysical
screening  methods  and   sampling  equipment,
incorporate suitable types and  numbers of quality
assurance/quality  control  (QA/QC)  samples,  and
interpret and present the  site analytical data.

As the  Superfund Program has developed,   the
emphasis of the response action has expanded beyond
addressing  emergency  response  and  short-term
cleanups.   Each  planned  response  action must
consider a variety of sampling objectives, including
identifying    threat,   delineating   sources   of
contamination, and  confirming  the achievement of
clean-up standards.  Because many important  and
potentially costly decisions are based on the sampling
data, Site Managers and other field personnel must
characterize site conditions accurately. To that end,
this document emphasizes the use of cost-effective
field analytical and geophysical screening techniques
to characterize the site and aid in the selection of
sampling locations.
1.2    UNIQUE   CHARACTERISTICS
        OF GROUND WATER

The following are media-specific variables of ground
water that  should be  considered when performing
representative ground-water sampling:

•   Homogeneity -  Ground water,  as  a medium, is
    usually homogeneous, especially when compared
    to other media such as soil, air, or waste.

•   Seasonal and Localized  Variation  in Flow  -
    Seasonal and localized variations in  ground-water
    flow should be considered when  developing a
    ground-water assessment program.   Seasonal
    variations are generally controlled by weather.
    Surface  streams  gain or  lose water  to  the
    subsurface when flood or drought conditions are
    present. Localized variations in flow are caused
    by  nearby, outside  influences,  as when a
    production well creates a cone of  depression in
    the water table.

•   Inaccessibility for Investigation - Ground water
    is often inaccessible  to standard grab sampling
    techniques.  Because ground water is subsurface,
    wells must often  be drilled and completed for
    sampling if no  existing wells are available.
    Sampling  ground water  is   generally   more
    complicated,  labor-intensive,  time-consuming,
    and expensive than sampling other media.

•   Natural Background Composition - Knowledge of
    the natural background composition is necessary
    in order to determine the effects of a site on the
    ground water.  Background or control monitoring
    wells  are  necessary  to  determine  ambient
    composition.

•   Water  Treatment  - Ground-water samples  are
    often  extracted  from  existing residential  or
    commercial wells that have been treated with
    softeners or have been filtered or altered in other
    ways.   Sampling  (times,  parameters, methods,
    preservatives, etc.) may have  to  be altered in
    order  to compensate  for or  avoid treatment
    variables.

•   Reproducibility of Sampling Results - Ground
    water is a flowing water body below the earth's
    surface.  Physical  and chemical characteristics
    may vary over  time  and  space because of the

-------
    factors listed above  (e.g., seasonal variation).
    Contaminants tend to flow through ground water
    in a plume  or  plug  of varying  concentration;
    contamination sources may discharge in pulses or
    as a continuous flow; and contaminants may react
    with ground water to  chemically transform over
    time.     Because  of   this   flowing   nature,
    contaminant or natural constituent concentrations
    can vary.  This variation could affect duplicating
    sample results  over  an extended time period.
    Contaminants will most often continue to be
    detected   in   ground   water,  but   sample
    concentration ranges may be altered, either by an
    increase  or a  decrease,  or  contaminant by-
    products may be detected.
1.3    REPRESENTATIVE SAMPLING

Representative ground-water sampling ensures that a
sample or group of samples accurately reflects the
concentration of the contaminant(s) of concern at a
given time and  location.   Analytical  results  from
representative  samples reflect  the   variation  in
pollutant presence and concentration throughout a site.

In addition to the  variables introduced due to the
characteristics of the sample media (as discussed in
Section 1.2), this document concentrates on those that
are introduced in the field.  These latter variables
relate to site-specific conditions, the sampling design
approach, and  the techniques  for collection  and
preparation of samples.   The following variables
affect  the  representativeness   of  samples   and
subsequent measurements:

•   Media variability - The physical  and chemical
    characteristics of ground water.

•   Contaminant    concentration  variability
    Variations in  the  contaminant  concentrations
    throughout the  site and/or variables affecting the
    release of site contaminants into ground water on
    or away from the site.

•   Collection   and   preparation  variability
    Deviations in analytical results attributable to bias
    introduced during sample collection, preparation,
    and transportation (for analysis).

•   Analytical variability - Deviations in analytical
    results attributable to  the manner in which the
    sample was stored, prepared, and analyzed by the
    on-site or off-site laboratory. Although analytical
    variability   cannot   be   corrected   through
    representative sampling, it can lead to the false
    conclusion that error is due to sample collection
    and handling procedures.
1.4    REPRESENTATIVE SAMPLING
        OBJECTIVES

Representative sampling objectives for ground water
include the following:

•   Identify the presence of contamination, including
    source,   composition,   and   characteristics.
    Determine if it is hazardous.

•   Establish  the  existence  of  an imminent or
    substantial threat to public health or welfare or to
    the environment.

•   Establish  the  existence  of  potential  threat
    requiring long-term actions.

•   Develop containment and control strategies.

•   Evaluate treatment options.

Note:  Clean-up goals are generally established for
ground water  and are not considered  a sampling
objective.

1.4.1  Identify Contamination and
        Determine Hazard

One of the first objectives during a response action at
a site is  to determine the presence, identity,  and
potential threat of any hazardous materials.  Field
screening  techniques can be used for rapid detection
of contaminants.  Upon confirming the presence of
hazardous materials, sample and/or continue screening
to identify their compositions and determine their
concentrations.

1.4.2  Establish Imminent or
        Substantial Threat

Establishing threat to the public or environment is a
primary objective during any  response action.  The
data obtained from characterizing the contaminants
will help the Site Manager to  determine whether an
imminent  or substantial threat exists and whether a
response action is necessary. The type and degree of
threat determines the rate at which a response action
is taken.

-------
1.4.3  Determine Long-Term Threat

Site conditions may support a long-term threat that is
not imminent or substantial.  Characterization of the
contaminants can assist the Site Manager to determine
the need for long-term  remediation and  response.
Samples should be collected in a manner that enables
their use for evaluating the  site under the Hazard
Ranking System.

1.4.4  Develop  Containment and
        Control Strategies

Once the chemical constituents and threat have been
determined,   many   strategies  for  ground-water
containment and control are available. Analytical data
indicating the presence of chemical hazards are not in
themselves sufficient to select a containment  or
control strategy.  Site reconnaissance and historical
site research provide  information on site conditions
and the physical  state of the contaminant sources;
containment  and  control  strategies  are  largely
determined by  this  information.   For  example,
trenching and pump  and treat  systems can prevent
spread of contamination in an aquifer.

1.4.5  Evaluate Treatment Options

The contaminants should be identified, quantified, and
compared  to  action   levels  (e.g.,   maximum
contaminant  levels  (MCLs)  for  drinking water).
Where regulatory action levels do not exist, site-
specific clean-up levels are determined by the Region
(often in consultation with the Agency for Toxic
Substances and Disease Registry  (ATSDR)) or by
State  identification of Applicable or Relevant and
Appropriate Requirements (ARARs). If action levels
are exceeded, a  series of chemical and physical tests
may be  required to evaluate possible  treatment
options.
1.5    CONCEPTUAL SITE MODEL

A conceptual site model is a useful tool for selecting
sampling locations.  It helps ensure that sources,
migration pathways, and receptors throughout the site
have been considered before sampling locations are
chosen.   The conceptual model assists the Site
Manager in evaluating the interaction  of different site
features. Risk assessors  use conceptual models to
help plan for risk assessment activities. Frequently, a
conceptual model is created as a site map (see Figure
1) or it may be developed as a flow  diagram which
describes potential migration of contaminants to site
receptors (See Appendix A).

A conceptual site model follows contaminants from
their sources, through migration pathways (e.g., air,
ground  water), and  eventually to the assessment
endpoints.  Consider the following when creating a
conceptual site model:

•   The state(s) of each contaminant and its potential
    mobility

•   Site topographic features

•   Meteorological   conditions   (e.g.,    wind
    direction/speed,     average     precipitation,
    temperature, humidity)

•   Human/wildlife activities on or near the site

The conceptual site model in Figure 1 is an example
created  for this document.  The model  assists in
identifying the following site characteristics:

    Potential Sources:  Site (waste pile, lagoon);
    drum dump; agricultural activities.

    Potential Migration Pathway (Ground Water):
    Leachate  from the  waste pile, lagoon,  drum
    dump, or agricultural activities.

    Potential Migration Routes:  Ingestion or direct
    contact  with  water from the aquifer  (e.g.,
    ingestion of drinking water, direct contact when
    showering).

    Potential Receptors of Concern:

        Human  Population  (Residents/Workers):
        Ingestion or direct contact with contaminated
        water from the aquifer.

Preliminary   site   information  may  provide  the
identification of the contaminant(s) of concern and the
level(s)  of the contamination.  Develop a sampling
plan based upon  the receptors of concern and the
suspected sources  and pathways.   The model may
assist in  the selection of on-site and off-site sampling
locations.

-------

-------
1.6    OVERVIEW OF GROUND-
        WATER MONITORING WELL
        INSTALLATION AND
        GROUND-WATER MODELING

Ground-water  monitoring  well  installation   and
ground-water modeling are complex issues which fall
outside the scope of this document. Many standard
operating procedures (SOPs) covering ground-water
monitoring well  installation techniques have been
published.  Monitoring well installation and ground-
water modeling  are  briefly introduced  here with
several  specific  items for consideration.  Refer to
existing SOPs and other reference documents for more
in-depth study.

1.6.1  Ground-Water Monitoring Well
        Installation

For most Superfund response actions where ground-
water sampling is performed, existing ground-water
production wells (commercial or residential) are used,
if available,  to  obtain  samples.   Chemical  data
obtained from this type  of well depict the general
quality  of water that is being delivered to the user
community. Ground water is usually a composite of
multiple aquifer strata which may mask the presence
of narrow or small contaminant plumes from a single
stratum.  For this reason, production  wells are  not
suitable for detailed  source,  case-preparation,  or
research types  of  monitoring.   Such  detailed
monitoring efforts require wells designed to determine
the  geologic  and hydrologic quality  at  specific
locations and depths.  The following items must be
considered  for    ground-water   sampling  from
monitoring wells:

•   Drilling method
•   Monitoring well components
•   Monitoring well location
•   Well diameter
•   Well depth
•   Well screen location

Refer to the U.S. EPA A Compendium of Superfund
Field     Operations     Methods,      OSWER
Directive   9355.0-14;  Compendium   of  ERT
Ground-water  Sampling  Procedures,   OSWER
Directive    9360.4-06;   RCRA   Ground-Water
Monitoring  Technical   Enforcement   Guidance
Document, OSWER Directive 9950.1;  and RCRA
Ground-Water  Monitoring:     Draft   Technical
Guidance, EPA/530-R-93-001, for specific details on
monitoring  well  installation.    The  latter  two
documents should be referenced for information on
locating, installing, and developing monitoring wells.

Locating Monitoring Wells

Often, one  well  is  sited  near the  center  of the
contaminant plume  just  downgradient  from  the
contamination  source.  Another well is installed
downgradient of the contaminant source, outside the
limits of the plume.  For background data, one well
may be placed outside of the  contaminant  plume,
upgradient of the contaminant source.  Additional
wells may  be  installed to  track the  amount  of
contaminant dispersion taking place.

Determining the  depth to  sample  is critical for
successful ground-water monitoring. Sampling depth
depends  on the contaminant density, the  aquifer
characteristics,  and the slope of the  water table or
potentiometric  surface.    The  number   of wells
necessary to monitor  ground water varies  depending
on many factors.  For example, if an impoundment
contamination source is higher than the surrounding
landscape,  leachate may  flow locally in all four
downgradient directions.  In this case, at least four
wells are needed to monitor plume movement, plus a
background well may be desired in an unaffected area.
In addition, some wells may be installed at more than
one depth in a contaminant plume to verify vertical
flow or spread of contamination at different depths.

See Driscoll, 1986, pp. 715-16 for more information
on locating monitoring wells.

Well Casing and Well Screen

Select a well casing material based on water quality,
well  depth,  cost,  borehole  diameter,  drilling
procedure, and Federal, state, and local regulations.
Types of casing materials include: steel, poly vinyl
chloride (PVC), fiberglass, and Teflonฎ.   Common
well casing diameters range from 2 inches to  12
inches or greater, and  depend on well type, well size,
well depth, and subsurface geology.  Often a series of
progressively smaller-diameter well casings are used
from the ground surface to the well depth.

A well  screen  is a filtering device which permits
water to enter the well from the saturated  aquifer
while preventing sediment from entering the well.  A
well screen has slots  or perforations and attaches to
the well casing.  It  can be constructed of metal,
plastic,  or other  material.   Important criteria for
selecting a well screen include: a large percentage of

-------
open area, nonclogging slots, resistance to corrosion,
and a sufficient column and collapse strength.

SeeDriscoll, 1986, pp. 413-431, and Fetter, 1993, pp.
339-344 for more information regarding well casing.
See Driscoll, 1986, pp. 395-405, and Fetter, 1993, pp.
345-346 for more information regarding well screens.
See U.S. EPA, November 1992, pp. 6-16 - 6-38 for
advantages and disadvantages of selecting well casing
and screen materials.

1.6.2  Ground-Water Modeling

Ground-water models, like conceptual site models,
can be useful when selecting sampling approaches,
objectives,  and locations.   Ground-water models
developed for Superfund sites attempt to provide an
estimation  of how the actual ground-water system
functions.

There  are  many  types  of ground-water models
available (e.g., physical, analog, mathematical). The
International   Ground-Water  Modeling   Center
(IGWMC)  has developed  a  ground-water model
definition  which   emphasizes  the importance  of
describing  a ground-water system mathematically.
The IGWMC defines a ground-water model as " a non-
unique,  simplified, mathematical description of an
existing   ground-water   system,  coded  in   a
programming language, together with a quantification
of the ground-water system the code simulates in the
form of boundary conditions, system parameters, and
system stresses."
A ground-water model may be useful throughout site
investigation activities because it can be adjusted as
conditions in the actual ground-water system become
better defined. The data which are generated by the
model can be used to refine sampling approaches and
locations as necessary.  Typically, a ground-water
modeling report will include data (results), along with
a discussion of activities such as model calibration
and conceptual model development.  A suggested
format for  a ground-water modeling report can be
found   in   U.S.  EPA  Ground-Water  Issue:
Fundamentals   of   Ground-Water   Modeling
(EPA/540/S-92/005).
1.7    EXAMPLE SITE

An example site, presented at the end of each chapter,
illustrates the development of a representative ground-
water sampling plan that meets Superfund Program
objectives for early actions or emergency responses.

-------
                     2.0   GROUND-WATER SAMPLING DESIGN
2.1     INTRODUCTION

The purpose of ground-water sampling is to provide
technical  information  relative  to  the  nature and
condition of subsurface water resources  at a specific
time and place.   Designs to monitor the  status of
ground water range from the studies  of  naturally
occurring geochemical constituents to the detection or
assessment of contamination within a ground-water
system.

Ground-water sampling objectives include identifying
threats,  delineating   sources  and  extent  of
contamination, determining  treatment and disposal
options, and  confirming the attainment of targeted
clean-up levels.  Representative sampling designs are
developed  to  most  accurately  characterize  the
hydrogeologic system  and its interaction  with  the
environment.    Sampling protocols must  integrate
detailed  sampling  methodology,  techniques  and
practices to  ensure valid  assessment.    Sampling
methodology  and practice may be the most common
source of assessment error.  Consequently,  sampling
methodology and practice collectively demand careful
preparation, execution, and evaluation to accurately
characterize  the  hydrogeologic   system  or  its
subsystems.  (For additional information see: U.S.
EPA  Ground Water,  Volume  II:   Methodology,
EPA/625/6-90/016b; and Palmer,  Christopher M,
Principles of Contaminant Hydrogeology.)

There  are many methods and types of equipment
useful for site  characterization and sample collection.
Selection of these factors is a critical component of a
site-specific sampling design.

A properly developed ground-water sampling design
defines the sampling purpose, protects  site worker
health and safety, effectively utilizes  resources, and
minimizes  errors.   The sampling design will vary
according to  the characteristics of the  site.  When
developing a sampling design, consider:

•   Prior actions at the site (e.g., sampling events,
    compliance inspections)
•   Regional    ground-water    properties    and
    characteristics
•   Potential    on-site   waste   sources   (e.g.,
    impoundments, waste piles, drums)
•   Topographic,   geologic,   hydrologic,   and
    meteorologic conditions of the site
•   Flora, fauna, and human populations in the area
2.1.1  Pre-Sampling  Plan
        Investigation

The  pre-sampling plan investigation  provides the
planner with information critical to the development
of a sound ground-water sampling design.  Integration
of all pertinent facts regarding the site history, the
population(s)  affected,   and  concentrations  of
substances on a site must be reviewed.  After all of the
pertinent  information has  been processed  and
incorporated into  a thorough site pre-evaluation, the
sampling plan can be developed.  Considerations for
sampling plan modification  should be reviewed as
necessary in light of the complex  nature of ground-
water resource dynamics.

Site History

Review of the site's history helps  assess the natural
and  man-made impacts  on  a  site.   Geographic,
geologic, tax, and fire insurance maps can indicate the
status of the site. These maps can usually be found at
local and collegiate libraries or municipal and county
tax  offices.   Aerial photographs are  helpful in
reviewing operational use of the site.  Archival aerial
photographs may show changes in operation and site
condition  over time.   This information can  be
correlated   with   information   from   potentially
responsible parties.

Hydrogeologic information is critical to developing a
sampling plan.   A  ground-water system is  site
specific,  depending upon  local geology, land and
subsurface use, precipitation and water use, proximity
to water  bodies, and  hydrogeologic  parameters
affecting contaminant  transport.  Hydrologic  and
hydrogeologic information can be found in libraries or
requested from the U.S. Geological Survey (USGS),
Water Resources Division, or  state  geological
agencies  and  their  water  branches.    Inspection
histories can be used  to determine prior health status
of the site in view of possible trends. Local, state, and
federal agencies dealing with health or environmental
inspection can provide such historical information
about a site.

Affected Populations

Human population statistics for the selected area can
establish the number of  people  threatened by the
contaminant exposure.  Include populations affected

-------
by projected migration of contaminants within the
ground-water system. Knowing the interaction of the
contaminant within a ground-water system and the
potential  regional  populations  exposed  to  the
contaminant will focus the sampling plan to the source
and possible pathways of the contaminant.  Wildlife
populations in  the area must be  studied as well.
Wildlife in ponds, lakes, streams, rivers, and bays is
often affected by contaminants transported by ground
water discharging into  surface water.  Information
regarding   regional   wildlife  populations  and
susceptibility to hazardous substances can be obtained
from  federal and state wildlife and conservation
agencies.

Detection Levels versus Maximum
Contaminant Levels

Sampling  plan  development must also address the
concentration level of  the contaminant within the
ground-water system in relation to the maximum
contaminant levels (MCLs) allowed within a public
water system.  Refer to the Federal Register for the
levels requiring enforceable action. Knowledge of the
chemical contaminant interaction within the ground-
water system can add  insight into the fate of the
contaminant (soluble or insoluble  in water;  less or
more dense than water; the nature of reactivity with
sediment or geology of the subsurface). Correlate the
concentration level  versus the location  of these
concentrations.  A sequence  of order can  then be
applied  to the  locations.  Ideally, a  pattern may
develop that can be related to the ground-water system
and its dynamics.  In the case of a single location,
investigate potential sources in the surrounding area
either by  working backwards from  an identified
contaminant spot  to a  potential source, or from  a
potential source to an identified contaminant spot.
Also consider source-to-current-location pathways and
projected pathways when developing a sampling plan.

2.1.2  Types  of Information
        Provided by Ground-Water
        Sampling  Assessment

There are several types of information that a ground-
water sampling  assessment provides. These include
but are  not limited  to:  measure  of ground-water
quality,  contaminant concentrations  compared  to
action levels, selection of the appropriate  response
action, and determination of ground-water flow and
contaminant plume movement.
Measure of Ground-Water Quality

Ground-water   sampling   assessments   provide
information concerning measure  of ground-water
quality of a site or region. Water quality is classified
according to many categories and its intended use.
Drinking water is especially subject to guidelines. A
sampling assessment of ground water can determine
whether the quality of the water has been maintained,
upgraded, or allowed to degrade.  The natural  and
artificially  induced characteristics of ground water
from a specific site or region can be established by
ground-water sampling assessments — specifically, the
chemical, biological, and physical characteristics of
the ground water.

Contaminant Concentrations Compared
to Action Levels

Ground-water sampling assessments provide a single
contamination level for a particular sampling location,
or a set of contamination levels for several sampling
locations within a site.  Comparison to action levels in
ARARs determines the basis for further action.  Thus,
sampling can evaluate potential hazards and represent
a  condition of ground-water  character  requiring
enforceable action procedures.

Selection of Appropriate Response
Action

The level of contaminant concentration as determined
through sampling  assessments is a critical factor in
selecting a site response action.  Depending upon the
degree or level  of  contaminant  concentration,
contaminant frequency, or  number of  locations
established as contaminated, and the site's potential
threat to human health or the environment, a rapid or
extensive clean-up program can be formulated, as well
as temporary or short-term responses (e.g., provision
of bottled water).

A sampling assessment may not always  indicate
contamination  of the site.  Careful examination of
sampling  protocol  must  consider  the  range  of
explanations.  A miscalculation of suspected source
sites;  gross procedure error in sampling,  laboratory
analysis,  or documentation; or error at many other
points in sampling protocol could be the source of
assessment error.  These errors are addressed more
extensively in Chapter 5.

If  quality  assurance/quality   control  (QA/QC)
procedures have been  followed for ground-water
sampling assessment, then it is possible that sources

-------
of contamination may originate from above ground
systems (e.g., lead entering tap water in the proximity
of the facility).  In any case, a sampling assessment at
the least can characterize  the natural ground-water
conditions, which  can be  used as a  control  or
comparison.

Determination of Ground- Water Flow and
Contaminant Plume Movement

Knowing  the  direction of ground-water flow  is
important when evaluating a contaminated aquifer.
When  contamination enters the ground at a  higher
head (gradient) than exists  at nearby  shallow wells,
these wells may become contaminated.  Ground water
flows from higher head to lower head. The direction
of water movement may be determined using water-
elevation data from a minimum of three wells.  See
Driscoll,  1986, pp. 79-85  and Freeze and Cherry,
1979, pp.  168-236  for more information regarding
ground-water flow.

Ground-water tracers, such as dye or salt may be used
to track ground-water flow velocities and contaminant
plume movement.  A tracer is placed in one well and
the time of its  arrival in a second well downgradient
from the first well is noted.  The dilution of the tracer
detected  in  the   second   well  can  indicate the
contaminant dilution rate  and help  determine the
contaminant source concentration as well as the width,
depth,  and spreading velocity of the plume. Tracers
also may be used to help determine aquifer porosity,
hydraulic conductivity, and dispersivity.
The tracer selected  must be detectable in extremely
low concentrations and must not react chemically  or
physically  with   the   ground-water  or  aquifer
composition.  See Driscoll, 1986, pp. 84-85 for more
information regarding ground-water tracers.

2.1.3  Site Reconnaissance

A site reconnaissance can be conducted at an earlier
date or immediately prior  to sampling activities.  It
allows  field personnel  to assess  actual, current site
conditions, evaluate areas of potential contamination,
evaluate potential hazards associated with sampling,
and finalize a sampling plan.  Site  reconnaissance
activities for  a ground-water  assessment include:
observing and photographing  the site; noting site
access  and potential  evacuation  routes;   noting
potential safety hazards; inventorying and recording
label   information   from  drums, tanks,  or  other
containers; mapping process and waste disposal areas
such as landfills, impoundments, and effluent pipes;
mapping potential contaminant migration routes such
as drainage, streams, and irrigation ditches; noting the
condition  of  animals  and/or  vegetation;  noting
topographic and/or structural features;  noting and
mapping existing ground-water monitoring or other
types of wells for potential sampling; and  siting
potential  locations for new monitoring  wells if
necessary. Field personnel  should use  appropriate
personal protective equipment when engaged in any
on-site activities.  Consider the following site-specific
factors while performing a site reconnaissance:

•   Sampling Objectives  - Sampling is conducted
    typically  to  comply  with regulations  for  the
    detection    or   assessment   of   suspected
    contamination  within the   subsurface.    The
    information gathered aids in the identification of
    known and unknown  substances present within
    the  site and the level and extent of contamination
    of the environment. The information is used to
    document the condition of  the  ground-water
    system  as an initial  assessment,  a record  of
    development,  or  as  evidence  of remediation
    efficiency and compliance.

•   Sample Collection and Toxicity  - The samples
    collected are intended to document the absence or
    measure  the presence of contaminants.   The
    measure of acute or chronic toxicity is evaluated
    by  assessing the site's extent of contamination,
    the  time period in relation  to the extent, and
    health hazards associated with the contaminant
    exposure time frame.

•   Statistical Concerns - A site visit will familiarize
    the  sampling planner with the environment to be
    sampled.  Conspicuous indicators of potential
    contamination sources or contamination effects
    may suggest use of a judgmental or bias sampling
    design. A geostatistical sampling method can be
    cost-effective and  time-efficient when compared
    to   strictly   random  or   random-stratified
    procedures.  When using less random methods,
    the  choice  of sampling locations should be
    documented and  justified.   Employ random
    sampling in addition to  bias sampling and include
    background  or control samples for  a thorough
    representation of  the ground-water character.
    (See  Section 2.3  for  a discussion of sampling
    approaches.)  (For additional information  see
    Keith, Lawrence H., Principles of Environmental
    Sampling.)

•   Timing of the Response -  Consider seasonal
    variation when evaluating a site.  Predictions of
    bad weather can influence technique and design.
    The urgency  of  the  action weighed  against

-------
seasonal constraints may  dictate  the options
available within the targeted budget.

Site-Specific Factors  Affecting  Ground-Water
Flow Many factors of a site control the path or
direction of ground-water flow.  A combination of
geologic  survey  information  with  the  site
reconnaissance can better familiarize the planner
with the dynamics of the hydrogeologic system.
The local geology of a site can determine the
direction and rate of ground-water movement by
means of its orientation and composition (e.g.,
horizontal,  tilted  or  vertical structures,  and
confining clay versus unconfining  sand  and
gravel).  The degree of development of a site and
its local topography can affect the ground-water
flow (e.g., parking lot runoff disproportionally
delivers water quantities to the subsurface  and
greater  slopes  afford less infiltration of water to
the  subsurface).    The  extent  and type  of
vegetation can affect the amount  of rainfall that
actually recharges an  aquifer system.  Dense
vegetation  and high  evapo-transpiration from
vegetation allows very little water to descend to
the subsurface.  Seasonal variations  can cause
reversal of ground-water flow direction.  This is
usually  associated with water bodies such as
streams, rivers, ponds, and lakes.  Water may
flow  to or  from  streams depending upon its
surface  elevation in relation  to adjacent water
table surfaces.  During flood conditions, water
usually flows from rivers toward the surrounding
subsurface. During drought, water moves toward
the lower level of the stream surface from higher
ground-water  surfaces.    (Consult  U.S. EPA
Handbook, Ground Water, EPA/625/6-87/016,
Chapter 4: Basic Hydrogeology.)

Analytical Parameters - The site reconnaissance
can help develop the list of analytical parameters.
For example, a reconnaissance may indicate the
presence of battery casings.  Lead would then be
a substance of concern. The site may contain
constraints that may or may not allow a variety of
tests to be performed.  The cost-effectiveness of
testing  within the site's constraints can lead to
limited options available to properly analyze the
ground-water system.  Testing methods may vary
within one site (e.g., monitoring well sampling,
hydroprobe extraction, etc.) in order to evaluate
multiple criteria vital to the site assessment.

Degradation for Transformation') Products - Sites
may  contain  degradation  (or transformation)
products, or by-products, of the contaminant that
are detectable and potentially as hazardous as the
    contaminant itself. Sampling for the product can
    lead to clues of the source substance location and
    its reactive status within the subsurface.

    Sampling Order -  The sampling plan  should
    address a specific order of sampling locations
    (and depths at a single location) to be developed.
    In order to use equipment efficiently, the plan
    should attempt to sample from "clean" to "dirty"
    locations, reducing the potential for contaminants
    to affect relatively less contaminated locations.
    Typically, the background  or "clean" location of
    a  site is  hydrologically  upgradient from the
    suspected contaminant "hot spot."  Depending
    upon  the nature of the  contaminant  (e.g., a
    "sinker"  or "floater"), the  sampling  at different
    depths within a column of water in a monitoring
    well should also  follow a sequence.
2.2    PARAMETERS OF CONCERN,
        DATA QUALITY OBJECTIVES,
        AND QUALITY ASSURANCE
        MEASURES

2.2.1  Parameters of Concern

Drinking  water  populations,  contaminants,  and
migration pathways are additional parameters that
should be considered  when  developing a sampling
plan.    Often, ground-water contamination goes
undetected because it is not directly visible. Drinking
water odor or taste complaints by residents close to
the site are usually the initial indication of ground-
water contamination and potential health hazards.
The sampling data  should accurately delineate  the
extent of ground-water contamination, determine the
impact on drinking  water populations, and indicate
potential migration pathways to such populations. It
is important to design the sampling plan to determine
where contaminants are most highly concentrated, and
to locate areas of decreasing detectable  concentrations
and those not yet contaminated.

2.2.2  Data Quality Objectives

Data  quality  objectives (DQOs) state the level  of
uncertainty that is  acceptable  for data collection
activities and define the certainty of the  data necessary
to  make decisions.  The overall goal of DQOs for a
representative ground-water  sampling  plan  are  to
acquire  thorough and accurate  information about
subsurface water conditions  at a site.   DQOs  are
unique  and site specific  and should  address  the
                                               10

-------
contaminant's interaction with the immediate  site
environment.    When  establishing  DQOs  for  a
particular project, consider:

•   Decision(s)  to be  made  or  question(s)  to  be
    answered
•   Why analytical data are  needed and how the
    results will be used
•   Time and resource constraints on data collection
•   Descriptions of the analytical data to be collected
•   Applicable model or data interpretation method
    used to arrive at a conclusion
•   Detection limits for analytes of concern
•   Sampling and analytical error

2.2.3  Quality Assurance Measures

To ensure that analytical samples are representative of
site conditions, quality assurance measures must be
associated with each sampling and analysis  event.
The sampling plan must specify QA measures,  which
include, but are not  limited to, sample collection,
laboratory  SOPs,  sample  container  preparation,
equipment decontamination,  field blanks, replicate
samples, performance evaluation  samples, sample
preservation  and handling,  and  chain-of-custody
requirements.  Quality  assurance components are
defined as follows:

•   Precision - Measurement of variability in the data
    collection process

•   Accuracy (bias)  - Measurement of bias  in the
    analytical process; the term "bias" throughout this
    document   refers   to   (QA/QC)   accuracy
    measurement

•   Completeness   -  Percentage  of   sampling
    measurements which are judged to be valid

•   Representativeness - Degree to which sample
    data accurately  and  precisely  represent  the
    characteristics   and   concentrations   of  the
    source/site contaminants

•   Comparability - Evaluation of the  similarity of
    conditions   (e.g.,   sample   depth,   sample
    homogeneity) under which separate sets of data
    are produced

Refer to Chapter 5, Quality Assurance/Quality Control
(QA/QC), for more detailed ground-water QA/QC
information.
2.3    REPRESENTATIVE  GROUND-
        WATER SAMPLING
        APPROACHES AND SAMPLE
        TYPES

Judgmental  sampling  is the primary representative
sampling approach used for ground water.  Other
representative sampling approaches for ground water
such  as random, systematic grid,   and  systematic
random  sampling  are described  below.    For
information   on  the   other types  of  sampling
approaches, refer to U.S. EPA, Superfund Program
Representative Sampling Guidance, Volume 1 — Soil,
OSWER Directive 9360.4-10.

2.3.1  Judgmental Sampling

Judgmental  sampling  is the biased selection of
sampling locations at a site,  based on  historical
information, visual inspection, sampling objectives,
and professional judgment. A judgmental approach is
best  used   when  knowledge  of   the   suspected
contaminant(s) or its origins is available.  Judgmental
sampling includes no randomization  in the sampling
strategy, precluding statistical interpretation of the
sampling results.  Criteria for  selecting sampling
locations are dependent  on the particular site and level
of contamination expected.

Once a contaminant has been detected in the ground
water, the source and extent must be identified.  To do
this,  an   understanding    of   the  contaminant
characteristics   and    the   local   geologic   and
hydrogeologic conditions is needed.  Characteristics
of the contaminant and any daughter (degradation)
products must be known in  order to understand how
the material  may be transported  (both vertically and
laterally) from the contamination  source.  Knowledge
of the local  hydrogeology  is  needed in  order to
identify  areas   or  zones   that would  facilitate
contaminant migration, such as water  bodies and
gravelly or  sandy  soils.  The permeability of the
underlying rock type should be analyzed, as well as its
depth, which will  help  to narrow the potential
sampling  area.   For   example,  if  the underlying
bedrock strikes northeast to southwest, then sampling
of an aquifer should also be in this direction, unless
cross-contamination between aquifers has already
been identified.

When appropriate (based  on sampling objectives,
availability, sampling parameters, and budget), sample
available  local  residential  or  commercial  wells
following a relatively systematic pattern based on the
                                                  11

-------
geology of the area. In the example given in the
paragraph above, wells would be sampled along a line
northeast to  southwest.  If the number of wells
available is not sufficient to adequately  identify the
extent of contamination, then  additional monitoring
wells could be installed.

During a ground-water assessment, the  selection of
locations for monitoring well installation is done with
a judgmental approach.  This is generally because
monitoring wells are complex, expensive, and time-
consuming to install. In order to best determine the
nature of a suspected contaminant plume, monitoring
wells need to be  placed in  areas most likely to
intercept the plume.  Using a random, systematic grid
or a systematic random approach would  likely result
in too many wells that miss the contaminant plume.
Even placement of background or control monitoring
wells favors a judgmental approach. Locations are
selected based on  the  site reconnaissance and the
planner's knowledge of the suspected contaminants,
site geology, and hydrology.

2.3.2  Random, Systematic Grid, and
        Systematic Random Sampling

Random, systematic grid, and  systematic random
sampling are generally  not used for ground-water
sampling because sampling points are pre-determined
from either existing wells or monitoring wells which
are placed by  judgment.  However, these approaches
may be useful  for soil gas testing to assist  in the siting
of new monitoring wells. They can also be useful for
conducting Geoprobeฎ sampling, if necessary.  For
additional information on these sampling  approaches,
refer   to  U.S.   EPA,   Superfund    Program
Representative Sampling Guidance, Volume 1 — Soil,
OSWER Directive  9360.4-10.
2.3.3  Grab     versus
        Sample Types
Composite
Grab samples are essentially the only type of samples
collected for ground water.  Unlike surface water,
ground water is not composited.  Each ground-water
sample is representative of a discrete  location and
horizon in the subsurface.
Site Location - The location of the site will often
influence the size  of the sampling area and
whether sampling should be conducted on or off
site or a combination of both.

Local Geology and Hydrology - Local  geology
and hydrology can determine whether off-site
sampling is necessary and defines ground-water
sampling boundaries and locations. For example,
if an aquifer is very deep or there is a confining
layer between the  ground surface and the aquifer,
then sampling within a small area may be all that
is necessary in order to determine the extent of
contamination within that aquifer.

Topography - Topography  will  control  the
direction of surface runoff and may give clues to
subsurface  conditions.  For  example, wells in
valleys may not be of the same aquifer  as wells
on a hill.

Analytical  Parameters  - If  contaminants are
initially unknown, then a broad  spectrum  of
analytical parameters is  usually collected.  As
more information about  the  site becomes
available  (through  screening  or laboratory
analysis),  the  number  of parameters  can be
streamlined or altered in order to more effectively
characterize the  site.   If the  contaminant is
known,  then concentrate on sampling for it and
its degradation products.

Sampling Budget  - Budget constraints inevitably
affect operations. A  combination of screening
and analytical  techniques minimizes expenses
while still  providing  an  acceptable  level  of
quality for the sampling data.

Physiochemical    Nature    of   Suspected
Contamination  When designing the sampling
plan, take into account the physical and chemical
nature of the suspected contaminants, then design
the sampling plan to facilitate efficient detection
of   the  contaminants   through  sampling
methodology, equipment,  and  analyses.  For
example, the water density  or  solubility  of  a
contaminant may provide  an indication of the
contaminant's physical location within the water
column.
2.4    SAMPLING PLAN

To develop a successful and practical representative
ground-water  sampling plan,  the  following  site-
specific information is required:
                          Water has  a  specific  gravity of one.  Some
                          chemical compounds,  such as many complex
                          petrochemicals, have a specific gravity of greater
                          than one, and are therefore more  dense  than
                          water. These substances tend to sink and include
                          chlorinated solvents, wood preservatives, other
                                                  12

-------
    coal tar wastes, and pesticides. These compounds
    are referred  to  as  dense  nonaqueous  phase
    liquids (DNAPLs), or "sinkers".  On the other
    hand, a specific gravity of less than one will
    allow a contaminant to float on or near the water
    table, and includes many fuel oil products and
    byproducts (e.g., gasoline, benzene, toluene, ethyl
    benzene, xylene (BTEX), and other straight chain
    hydrocarbons). These compounds are referred to
    as light nonaqueous phase liquids (LNAPLs), or
    "floaters". Nonaqueous phase liquids (NAPLs)
    tend to exist as separate layers  in the water
    column.  A substance with a specific  gravity
    value near to or equal to one will generally
    dissolve  in  the  water column  (e.g.,  acetone,
    phenols, and creosote). Because of the potential
    stratification in the water column due to  NAPL
    substances, sampling location with respect to the
    suspected contaminant location  within the well
    should always be considered.

    LNAPLs commonly occupy the  capillary fringe
    zone above the water table. In a confined aquifer,
    these compounds are found along the  upper
    surface of the permeable unit and also within the
    overlying confining layer.

    DNAPLs   cause   additional   representative
    sampling concerns.   These compounds move
    downward under the influence of gravity until
    reaching a less permeable formation where they
    may either accumulate, move downslope along
    the bedrock, or penetrate  fractures.  Special
    precautions  should be taken during drilling at
    sites suspected of DNAPL contamination;  ensure
    that  the drilling does not induce the spread of
    free-phase DNAPL  contaminants.  Monitoring
    well installation should be suspended when a
    DNAPL  or  low permeability  lithogic  unit is
    encountered.  Fine-grained aquitards (e.g., silt or
    clay) should be assumed to permit downward
    DNAPL migration.  For guidance on sites with
    potential DNAPL contamination, see U.S. EPA
    Estimating  the Potential for  Occurrence of
    DNAPL at Superfund Sites, OSWER Directive
    9355.4-07.

Additional elements which should be addressed in a
representative ground-water sampling plan include:

•   Sample Number -   The number  of samples
    collected depends  on the  number of  sample
    locations. Normally one sample  is taken at each
    location, except for  QA/QC requirements (e.g.,
    replicates,   and  matrix  spike/matrix   spike
    duplicates).   If there  are  multiple, discreet
    aquifers at the site, then samples of each may be
    necessary.   Splitting samples  also requires an
    increase in the number of samples.

•   Sample  Volume -  The sample volume  is
    dependent on the analytical parameters.  It is also
    dependent on whether the contaminant is known
    or unknown.   A greater volume is generally
    needed  when  the  contaminant  is  unknown
    because a larger suite of parameters is usually
    selected.

•   Sample Location - Sample location is generally
    dictated   by   the   availability   of  existing
    monitoring,  residential, or commercial wells.
    New monitoring wells are located by judgmental
    methods.

•   Sample Depth - Sampling depth is typically the
    bottom or screened  zone of a  well.  However,
    there may be times when certain stratigraphic
    horizons within the water column may need to be
    discreetly sampled (e.g., capturing "floaters" or
    "sinkers").  (Procedures for addressing stratified
    samples are  discussed in Section 4.4.)

•   Sample Order - Sampling order is from the least
    contaminated to the  most contaminated wells or
    areas (if known).

2.5    EXAMPLE SITE

2.5.1  Background

The ABC Plating  Site  is  located in northeastern
Pennsylvania approximately 1.5  miles north of the
town of Jonesville. Figure 2 provides a layout sketch
of the site and surrounding area.  The site covers
approximately four acres and operated as a multi-
purpose specialty electroplating facility from 1947 to
1982.  During its years  of operation,  the company
plated automobile and airplane parts with chromium,
nickel, and copper.  Cyanide solutions were used in
the  plating  process.    ABC   Plating deposited
electroplating wastes into two unlined shallow surface
settling lagoons in the northwest portion of the site.
Pennsylvania Department of Environmental Resources
(PADER) personnel cited the owner/operator for the
operation of  an  unpermitted  treatment system and
ordered the owner to submit a remediation plan for
state approval. Before PADER could follow up on the
order, the lagoons were partially backfilled with the
wastes in place.   The process building was later
destroyed by  a fire of suspicious origin. The owner
abandoned the facility and could not be located by
                                                  13

-------
                        Figure 2: Site Sketch
                          ABC Plating Site
                                       A
                                     TREELINE
                                                    A
                       A
A
                             SUSPECTED
                             - LAGOONS
                                    SUSPECTED
                                     TRENCH
                                  HOUSE
                                 TRAILER
        SCALE IN FEET
 100    50
            I
                      100
                                LEGEND
                       DAMAGED
                       BUILDING
                         AREA
— -   SURFACE FLOW

---   SITE BOUNDARY
                               14

-------
enforcement  and legal authorities.   Several  vats,
drums,  and  containers were left  unsecured  and
exposed to the elements. The state contacted EPA for
an assessment  of the  site  for a possible federally
funded   response   action;   an  EPA   On-Scene
Coordinator (OSC) was assigned to the task.

2.5.2  Site History and
        Reconnaissance

The EPA OSC reviewed the PADER site file. In 1974
the owner was cited for violating the Clean Streams
Act  and for storing  and  treating industrial  waste
without a permit. The owner was ordered to file a site
closure  plan  and to remediate the settling lagoons.
The  owner, however, continued operations and was
then ordered to begin remediation in 90 days or be
issued a cease and desist order. Soon after, a follow -
The OSC obtained copies of aerial photographs of the
site  area from the local district office up inspection
revealed that the lagoons had been backfilled without
removing the waste.

The OSC and a sampling contractor (Team) arrived on
site  to  interview local and  county  officials,  fire
department officers,  neighboring residents (including
a  former   facility   employee),   and  PADER
representatives regarding site operating practices and
other site details.   The former employee sketched
facility  process features on a  map copied from state
files. The features included two settling lagoons and
a feeder trench which transported plating wastes from
the process  building to  the lagoons.   The  OSC
obtained copies of aerial photographs of the site area
from  the local district  office  of the  U.S.  Soil
Conservation Service.  The state provided the OSC
with  copies of all historical site and violation reports.
These sources indicated the  possible presence and
locations of  chromium,  copper,  and zinc plating
process  areas.

The Team mobilized to the site with all the equipment
needed to perform multi-media sampling.  The OSC
and  Team  made a  site entry, utilizing  appropriate
personal protective equipment and instrumentation, to
survey the general site conditions.  They observed 12
vats, likely containing plating solutions, on a concrete
pad where the original facility process building once
stood.   Measurements of pH ranged from 1 to 11.
Fifty drums and numerous smaller containers (some
on the concrete pad, others  sitting directly on the
ground) were leaking and bulging because of the fire.
Some rooms of the process  building could not be
entered due to unsafe structural conditions caused by
the fire.   The Team noted many areas of stained soil,
which  indicated container  leakage,  poor  waste
handling practices, and possible illegal dumping of
wastes.

2.5.3  Identification of  Parameters of
        Concern

During the site entry, the OSC and Team noted that
several areas were devoid of vegetation, threatening
wind erosion which could transport heavy metal- and
cyanide-contaminated soil particulates off site.  These
particulates could be deposited  on residential property
downwind or be inhaled by nearby residents.

Erosion gullies located on site indicated surface soil
erosion and stormwater transport. Surface drainage
gradient  was toward the  west and northwest. The
Team observed stressed, discolored, and necrotic
vegetation immediately off site  along  the surface
drainage route.  Surface drainage of heavy metals and
cyanide was a direct contact hazard to local residents.
Surface water systems were also potentially affected.
Further   downgradient,  site  runoff  entered  an
intermittent tributary of Little Creek, which in turn
feeds Barker Reservoir.  This reservoir is the primary
water  supply  for   the   City  of  Jonesville and
neighboring communities, which are located 2.5 miles
downgradient of the site.

The site  entry team  observed that the  site was not
secure and there were signs of trespass (confirming a
neighbor's claim that children play  at  the facility).
These activities could lead to  direct contact with
cyanide and heavy metal contaminants, in addition to
the potential for chemical burns from direct contact
with strong acids and bases  as might be found in
leaking or unsecured drums or containers.

After interviewing residents,  it  was  established that
the homes located to the south and nearest to the site
rely  upon private wells for their primary drinking
water supply. Ground water is also utilized by several
small community production systems  which have
wells located within 2 miles of the site.  The on-site
settling lagoons were unlined and therefore posed a
threat to ground water, as did precipitation percolating
through contaminated soils.  Contamination  might
have  entered   shallow  or  deeper aquifers and
potentially migrated to off-site drinking water wells.

During Phase  1  sampling activities,  full  priority
pollutant  metals  and total cyanide  analyses were
conducted on all soil and  ground-water  samples sent
to the laboratory.  These parameters were initially
selected  based  on  a study  of plating chemistry:
                                                   15

-------
plating facilities generally use either an acid or basic
cyanide bath to achieve the desired coating on their
metal products.  Since Phase 1 samples were collected
from the  areas of  highest  suspected contaminant
concentration (i.e., sources and drainage pathways),
Phase 2 samples (all media types) were analyzed for
total chromium, hexavalent chromium (in water only),
and cyanide, the only  analytes detected consistently
during the Phase  1  analyses. During Phase 3,  the
samples sent to the laboratory for definitive analysis
were analyzed for total chromium and cyanide.

2.5.4  Sampling  Objectives

The  OSC  initiated  an assessment with  a specific
sampling objective,  as follows:

•   Phase 1 — Determine whether a threat to public
    health,  welfare, and  the environment exists.
    Identify sources of contamination  to support an
    immediate   CERCLA-funded   activation   for
    containment of contaminants and security fencing
    (site  stabilization  strategies) to reduce direct
    contact concerns  on site.  Sample  the nearby
    drinking water wells for immediate  human health
    concerns.

Once CERCLA funding was obtained and the site was
stabilized:

•   Phase 2 — Define the extent of contamination at
    the site and  adjacent  residential  properties.
    Estimate the costs for early  action options and
    review  any  potential   long-term  remediation
    objectives.  For example, install and sample soil
    borings and monitoring wells on site to evaluate
    potential impact on subsurface soils and ground
    water.

•   Phase  3 — After  early  actions  are  completed,
    document the attainment of goals.  Assess that the
    response action was completed to the selected
    level and is suitable for long-term goals.

2.5.5  Selection  of Sampling
        Approaches

The OSC, Team, and PADER reviewed all available
information to formulate a sampling plan.  The OSC
selected a judgmental sampling approach for Phase 1.
Judgmental sampling supports the immediate action
process by best defining on-site contaminants in the
worst-case scenario  in order  to evaluate the threat to
human health, welfare, and the environment. Threat
is  typically established  using   a  relatively small
number of samples (fewer than 20) collected from
source areas or suspected contaminated areas based on
the historical data review and site reconnaissance.  For
this  site,  containerized wastes  were  screened to
categorize the contents and to establish a worst-case
waste volume, while soil samples were collected to
demonstrate whether a release had already occurred,
and  nearby  residential drinking water wells were
sampled for immediate human health concerns.

For Phase 2, a stratified systematic grid design was
selected to define the extent of contamination in soils.
The grid could accommodate analytical screening and
geophysical surveys.   Based  on search  sampling
conducted at sites similar to ABC Plating, a block grid
with a 50-foot grid spacing was selected.  This grid
size  ensured a  10 percent or  less probability of
missing a "hot spot" of 45 feet by 20 feet.  The grid
was extended to adjacent residential properties when
contaminated soil was identified at grid points near
the boundary of the site.

Based on the results of soil sampling and geophysical
surveys,  a judgmental approach was used to select
locations for installation of 15 monitoring wells: at
"hot spots";  along  the perimeter of the suspected
plume  established  from  analytical   results   and
geophysical survey plots; and at background ("clean")
locations. Subsurface soil and ground- water samples
were collected from each of the 15 monitoring well
locations for laboratory  analysis to  establish  the
presence  and,  if   applicable,  the  degree   of
contamination at depth.

2.5.6  Sampling Plan

During Phase 1, containerized wastes were evaluated
using field analytical screening techniques.  Phase 1
wastes-screening indicated the  presence of strong
acids and bases and the absence of volatile organic
compounds.  The Team collected a total of 12 surface
soil samples (0-3 inches) and 3 ground-water samples
during this phase and sent them to a laboratory for
analysis. The soil sampling locations included stained
soil  areas, erosion channels, and  soil adjacent to
leaking containers.  Background samples  were not
collected  during  Phase   1  because  they  were
unnecessary for activating immediate action response
funding.  Ground-water samples were collected from
three nearby residential wells.  Based on  Phase  1
analytical results, chromium was selected as the target
compound   for   determination   of   extent   of
contamination in soil and ground water.
                                                   16

-------
During Phase 2 sampling activities, the OSC used a
transportable X-ray fluorescence (XRF) unit installed
in an on-site trailer to screen soil samples for total
chromium in order to limit the number of samples to
be sent for laboratory  analysis.  Soil sampling was
performed at all grid nodes at the surface (0-4 inches)
and subsurface (36-40 inches). The 36-40 inch depth
was selected based on information obtained from state
reports  and local  interviews, which  indicated that
lagoon  wastes  were  approximately  3 feet below
ground   surface.     Once   grid  nodes  with  a
contamination level  greater  than  a  selected target
action level were located, composite samples were
collected from each adjoining  grid cell.  Based on the
XRF  data,  each  adjoining  grid cell was either
identified  as  "clean"  (below   action  level)   or
designated for response consideration (at or above
action level).

Also  during Phase   2,  the  OSC  oversaw   the
performance of ground penetrating radar (GPR) and
electromagnetic   conductivity  (EM)  geophysical
surveys to help delineate the buried trench  and lagoon
areas, any conductive ground-water plume, and any
other  waste  burial  areas.     The   GPR   and
comprehensive EM surveys were conducted over the
original  grid.   Several  structural discontinuities,
defining possible disturbed areas, were detected.  One
GPR  anomaly corresponded with  the suspected
location and orientation of the feeder trench.  The EM
survey identified several high conductivity anomalies:
the suspected feeder trench location, part of the lagoon
area,  and a small area  west of the process building,
which may have been an illegal waste dumping area.
(Field analytical screening and geophysical techniques
are further discussed in Chapter 3.)

Using the data obtained during soil sampling and the
geophysical surveys, a ground-water assessment plan
for Phase 2 was prepared. The Team collected depth
soundings and water level measurements of the nearby
residential wells to assess aquifer usage and location
(depth).  With these data and the analytical results
from  Phase  1, a  work  plan for monitoring well
installation and testing on site was developed.  The
plan consisted of:
•   Installation of overburden, bedrock contact and
    bedrock  (open borehole) monitoring  wells  in
    order to  evaluate the shallow water table and
    aquifer conditions

•   Analysis  of  subsurface  soils retrieved during
    borehole/well drilling in order to evaluate the
    extent of contamination in overlying soils

•   Collection of depth soundings and water level
    measurements of the newly installed monitoring
    wells to map  aquifer and water table gradients

•   Collection of ground-water samples from each
    monitoring well

•   Performance  of hydraulic tests  in order  to
    evaluate aquifer characteristics

The monitoring wells were located in areas shown,
during soil sampling, to be  heavily contaminated;
along the outer perimeter of a contaminant plume
based  on  soil XRF results  and the  geophysical
surveys;  and  an  apparent upgradient location for
background conditions  comparison.   Fifteen wells
were located at grid nodes corresponding to the above
results.    (Section  4.6.1  provides details  on the
performance of well installation (drilling), testing and
surveying,  and ground-water sampling procedures.)

Upon monitoring well installation and  sampling, a
hydraulic  (pumping) test  was completed of the
bedrock monitoring wells to gather information about
aquifer  characteristics.    These  data characterize
contaminant  transport  through  the  ground-water
aquifer.  The hydraulic test provided transmissibility,
hydraulic  conductivity,  and   storativity   values.
Utilizing these values with ground-water level data,
the estimated  vertical and horizontal ground-water
gradient  and  velocity  could be  calculated.   All
monitoring wells installed were surveyed for elevation
above mean sea level, needed to determine accurate
depth to  ground  water  (piezometric surface)  and
relative gradients.

Phase 3 activities  are discussed in Section 6.8.
                                                    17

-------
      3.0  FIELD ANALYTICAL SCREENING, SAMPLING EQUIPMENT,
                          AND GEOPHYSICAL TECHNIQUES
3.1     FIELD ANALYTICAL
        SCREENING

Field analytical screening techniques can  provide
valuable information in ground-water sampling.  Field
analytical screening for ground water is used primarily
as a tool for siting monitoring wells and for on-site
health  and safety assessment during  well  drilling
activities. When used correctly, screening techniques
can  help to  limit  the  number  of  "non-detect"
monitoring wells installed.  Some of the commonly
used screening methods for ground-water assessment
are presented in this chapter in the general order that
they would initially  be used at a site, although site-
specific conditions may  mandate a different sequence.
For more information on ground-water field screening
devices, refer to the  U.S. EPA Compendium ofERT
Field Analytical Procedures,  OSWER Directive
9360.4-04, and Compendium of ERT Ground-Water
Sampling Procedures, OSWER Directive 9360.4-06.
Refer to Standard Operating Safety Guides for each
instrument, and the  U.S. Department of Health and
Human Services Occupational Safety and Health
Guidance Manual  for  Hazardous   Waste  Site
Activities (NIOSH  Pub.  85-115) for  site  entry
information.

3.1.1  Flame lonization  Detector

The  flame ionization detector  (FID)  detects and
measures the level of total organic compounds
(including methane) in the ambient air in proximity to
a well or in a container headspace.   The FID uses the
principle of hydrogen flame ionization for detection
and  measurement.   It is especially effective  as  an
ethane/methane detector when used  with an activated
charcoal filter because  most organic vapors  are
absorbed as the  sample passes  through the  filter,
leaving only  ethane and methane to be measured.

The  FID operates in one of two modes:  the survey
mode, or the  gas chromatography (GC) mode.  In the
survey mode, the FID provides an approximate total
concentration of all detectable organic vapors and
gases measured relative to the calibration gas (usually
methane). The GC mode identifies and measures
specific components, some with detection limits as
low  as a few parts per million (ppm), using known
standards analyzed concurrently in the field. Since the
GC  mode requires standards to  identify classes of
compounds, it is necessary to have an idea of which
compounds might be present on site before sampling.
Advantages  of  the FID  are that it  is portable,
relatively rugged, and provides real-time results.

During a ground-water assessment, the FID is used in
the survey mode for monitoring the borehole during
drilling and in the survey or GC mode for health and
safety screening.

The FID does not respond to inorganic substances. It
has positive or  negative response factors for each
compound depending on the selected calibration  gas
standard.  Ambient air temperatures of less than 40
degrees Fahrenheit will  cause  slower responses;
relative humidity of greater than 95 percent can cause
inaccurate and unstable responses.  Interpretation of
readings (especially in the GC mode) requires training
and experience with the instrument.

3.1.2  Photoionization Detector

Another portable air monitoring instrument frequently
used  for  field screening  during  ground-water
assessments is the  photoionization detector  (PID).
Like the FID, the PID provides data for real-time total
organic  vapor measurements, identifying potential
sample locations and extent  of contamination, and
supporting health and safety  decisions. The PID is
useful in performing  soil  gas screening, health and
safety monitoring during well drilling activities, and
headspace screening analysis.  The PID works on the
principle of photoionization. Unlike the FID, the PID
can be  used to detect gross organic and some
inorganic  vapors,  depending on the substance's
ionization potential (IP) and the selected probe energy.
It  is portable and  relatively easy  to  operate  and
maintain in the field.

The PID  detects  total concentrations and  is  not
generally used to quantify  specific substances. PIDs
cannot detect methane; however,  methane is  an
ultraviolet (UV) light absorber,  and false negative
instrument  readings may  register in  methane-rich
environments. The PID cannot detect substances with
IPs greater than  that  of the  UV  light  source.
(Interchangeable UV lamps are available.) Readings
can be  affected by  high wind speeds,  humidity,
condensation, dust,  power lines, and portable radios.
Dust particles and water droplets (humidity) in  the
sample may collect  on the light source  and absorb or
                                                 18

-------
deflect UV energy, causing erratic responses in PIDs
not equipped with dust and moisture filters.

3.1.3  Gas Chromatograph

Although many FIDs are equipped with a GC mode,
an independent, portable GC (gas chromatograph) can
also  be used on site to provide a chromatographic
profile of the occurrence  and intensity of unknown
volatile organic compounds (VOCs) in ground water.
The  GC is  useful as a soil gas screening tool to
determine   "hot   spots"    or   plumes,   potential
interferences, and semi-quantitation of VOCs and
semi-volatile organic compounds (semi-VOCs) in
ground-water samples. For example, when installing
a monitoring well, the GC might be used to  screen
water samples  during drilling in  order to indicate
when  a  target  contaminated  aquifer  zone  is
encountered.

Compounds with high response  factors,  such as
benzene and toluene, produce large response peaks at
low  concentrations, and can mask the presence of
compounds with lower response factors.  However,
recent improvements in GCs, such as pre-concentrator
devices   for  lower  concentrations,   pre-column
detection  with  back-flush capability  for   rapid
analytical time, and the multi-detector (PID, FID, and
electron capture detector (BCD)), all enable better
compound detection.  The GC is highly temperature-
sensitive. It requires set-up time, many standards, and
operation by trained personnel.

3.1.4  Hydraulic Probe

The  hydraulic probe  (Geoprobeฎ is one brand) is a
truck-mounted  device  used  to  collect  screening
ground-water, soil, and soil gas samples at relatively
shallow depths.  The probe is mounted on the back of
a small  truck or van and is operated hydraulically
using the vehicle's engine.  Small diameter hardened
steel probes are driven  to depths of up to 40 feet or
more, depending on soil conditions. Soil gas samples
can then be collected using a vacuum pump. Soil or
water samples can also be  collected using a  small-
diameter shelby tube or slotted well point and foot
valve pump.

The  hydraulic probe can be used in ground-water
investigations to assess vertical  and horizontal extent
of contamination.  Shallow samples can be collected
relatively quickly and  easily. It is useful in a ground-
water assessment to assist in siting monitoring wells
and to install shallow wells  if necessary.  It can also
collect undisturbed ground-water samples without
installing wells. The hydraulic probe is only effective
in unconsolidated geologic materials, however.  In
general,  probing  is  possible  under  conditions
amenable to hollow stem auger drilling.

3.1.5  Soil Gas Technique

Soil gas testing is a quick method of site evaluation.
For ground-water assessments, soil gas testing is used
to  track   contaminant   plumes   and  determine
appropriate locations for installing monitoring wells.
For this technique, a  thin  stainless steel probe is
inserted into  a hole made in the soil with a special
slam bar.  The hole is sealed around the probe and a
sampling  pump  is attached.   Samples  are  then
collected  in  Tedlar bags, sorbent cartridges,  or
SUMMA canisters.  The samples are analyzed using
an FID, PID, or GC. A disadvantage of the soil gas
technique  is  that its ability to detect contaminants
diminishes the further it is  from the source (as
contaminant concentration diminishes).

3.1.6  Field  Parameter Instruments

Field  parameters  measured  during ground-water
sampling   include  pH,   specific  conductivity,
temperature, salinity, and dissolved oxygen.  Specific
conductivity, pH, and temperature are often used as
standard indicators of water quality.  Instruments that
measure these three indicators are used during ground-
water assessments to determine if  a well has been
purged sufficiently (stabilized) prior to sampling (see
Section 4.3).

3.1.7  X-Ray Fluorescence

Field  analytical screening using X-ray fluorescence
(XRF) is a cost-effective and time-saving method to
detect and classify lead and other heavy metals in a
sample. XRF screening  provides immediate semi-
quantitative results. The principle behind XRF is the
detection  and measurement of the X-rays released
from an atom when it is excited by the absorption of
source X-rays. The energy released (fluorescent X-
rays) are characteristic of the atoms present.

Results of XRF analysis help determine the presence
of metals  and are often used to assess the extent of
soil  contamination at  a  site.  For ground-water
assessment, XRF may be used on subsurface soil
samples collected during drilling or with surface soils
when   selecting   locations  for  monitoring  well
installation. XRF use requires a trained operator and
may  require numerous   site-specific  calibration
samples.
                                                  19

-------
3.2    GROUND-WATER  SAMPLING
        EQUIPMENT

Conducting representative  ground-water sampling
requires an understanding of the capabilities of the
equipment used  for  sampling,  since  the  use  of
inappropriate equipment may result in biased samples.
Select appropriate sampling equipment based on the
sampling  objectives, the analytical parameters, the
type of well being sampled (e.g., monitoring well or
drinking  water  well),  and   other   site-specific
conditions.  Follow SOPs for the proper use and
decontamination of sampling equipment.  This section
presents  various  types  of ground-water sampling
equipment and information to assist  in selecting
appropriate materials.

The ground-water sampling devices discussed below
are covered in greater  detail  in  many SOPs and
references on the  various types of available ground-
water  sampling devices.   Refer to U.S. EPA A
Compendium   of  Superfund  Field  Operations
Methods,   OSWER  Directive   9355.0-14,  and
Compendium  of ERT  Ground-Water  Sampling
Procedures, OSWER Directive 9360.4-06, for details
on  the equipment listed.   Also refer  to Driscoll,
Fletcher G., Ground-Water and Wells, 2nd ed., and
the  1985  "Proceedings  of  the  Fifth National
Symposium and Exposition on Aquifer Restoration
and Ground-Water  Monitoring,"  for additional
comparisons  of  the  various  types of sampling
equipment.

3.2.1  Bailer

A bailer is a simple purging device for collecting
samples from monitoring wells.  It usually consists of
a rigid length of tube with a ball check-valve at the
bottom.  A line is used to mechanically lower the
bailer  into the  well to retrieve a  volume of water.
Because bailers are portable and inexpensive, they can
be  dedicated to  monitoring wells  at  a site,  thus
avoiding the need to use a bailer for sampling more
than one  well (and avoiding cross-contamination).
Bailers are available  in a variety  of sizes and
construction materials (e.g., polyvinyl chloride (PVC),
Teflonฎ, and stainless steel).

Bailers are best suited for purging shallow or narrow
diameter monitoring wells. Deeper, larger diameter,
and water supply wells generally require mechanical
pumps to evacuate a large volume of water.

For VOC  analysis, a positive-displacement volatile
sampling bailer is most effective. Bottom-fill bailers,
which are more commonly used, are suitable provided
that care is taken to preserve volatile constituents. Fill
sample  containers directly from the bailer, filling
samples for VOC analysis first.

3.2.2  Hydraulic Probe

The hydraulic probe can be used to collect  shallow
(generally 40 feet or less) ground-water samples using
a mill-slotted well point or retractable screen drive
point.  After the well point is driven to the desired
depth, the probe rod is connected to a vacuum pump
for purging.   (Since ground water is sampled in situ
and is not exposed to  the  atmosphere,  extensive
purging is not required.)

Water   samples  are  collected  using   dedicated
polypropylene tubing fitted with  a small diameter
foot-valve pump.  Samples are collected in 40-ml vials
or other containers for laboratory analysis.   See
Section 3.1.4 for more information on the hydraulic
probe.

3.2.3  Air-Lift Pump

An air-lift pump operates by releasing compressed air
via an air line lowered into the well.  The air mixes
with the water in the well to  reduce the  specific
gravity of the water column and lift the water to the
surface.

Air-lift pumping is used in  well development and for
preliminary testing. For sampling, air-lift pumping is
less efficient than other pumping methods which
follow; it may be selected for use when aeration is
needed to remove gas or corrosive water which can be
destructive to a well pump.  Because an air-lift pump
aerates the water, it is not applicable for VOC sample
collection.

3.2.4  Bladder Pump

A bladder pump consists of a stainless steel or
Teflonฎ housing that encloses  a Teflonฎ bladder.
The bladder pump is operated using a compressed gas
source (bottled gas or an  air compressor).   Water
enters  the bladder through  a lower check valve;
compressed  gas  moves the water through an upper
check valve and into a  discharge line.  The upper
check valve prevents back flow into the bladder.

The bladder pump can be used to  purge and sample to
a depth of approximately 100 feet. It is recommended
for VOC sampling because   it   causes  minimal
alteration of sample integrity as compared with other
                                                  20

-------
ground-water sample methods.  The bladder pump
requires a power supply and a compressed gas supply
or air compressor. The pump is somewhat difficult to
decontaminate and should thus be dedicated to a well
(or dedicated tubing should be used).

3.2.5  Rotary  Pump

A rotary pump is a positive displacement pump which
discharges the same volume of water regardless of the
water pressure.  The rate of discharge is the same at
both low and high pressure, but the input power varies
in direct proportion to the pressure.  The rotary pump
consists of a housing with inlet and outlet ports and
rotating gears or vanes.  As water is discharged from
the  pump, a replacement supply of equal volume is
taken in.

Rotary pumps are useful for well purging and general
sample collection at shallow to deep sampling depths.
Because of water agitation, they may  not be suitable
for  sampling  VOCs,  and they  are   difficult  to
decontaminate between sampling stations.

3.2.6  Peristaltic Pump

A peristaltic pump is a suction lift pump consisting of
a rotor with ball-bearing rollers. Dedicated Teflonฎ
tubing is threaded  around the  rotor.   Additional
lengths of dedicated Teflonฎ tubing  are attached to
both ends of the rotor tubing: one end is inserted into
the  well; the other end is  a discharge tube.  The
sample makes contact with the tubing only, not with
the pump. The tubing should be equipped with a foot
valve to avoid having aerated water from the tubing
fall  back into the well.

A peristaltic pump is suitable for sampling small
diameter wells (e.g., 2 inches). Cross-contamination
is not of concern because dedicated tubing is used and
the sample does not come into contact with the pump
or other equipment.  The peristaltic pump has a depth
limitation of 25 feet  and its use can result in  a
potential  loss of the volatile fraction due to  sample
aeration.

3.2.7  Packer Pump

A packer pump is used to isolate portions of a well or
water column for sampling.  The pump consists of two
expandable parts that isolate a sampling unit between
them.  The parts deflate for vertical movement within
the well and inflate when the desired sampling depth
is reached. The packers are constructed of rubber and
can be used with various types of pumps.
An advantage of the packer pump is it allows the
isolation of a portion of the water column in order to
sample at  a discrete depth.  Disadvantages relate to
the rubber construction of the packers which may
deteriorate over time allowing cross contamination.
The   rubber  also  poses  potential   contaminant
compatibility concerns.  A packer pump should not be
used if the contaminants are unknown,  or where well
casing or contaminant characteristics interfere  or
interact with the pump construction materials.

3.2.8  Syringe Sampler

Syringe samplers  are a relatively new  and less
commonly available  sampling device.    Syringe
samplers were developed by research groups to obtain
ground water samples over a period of time.  The
device consists of a syringe (15 to 1500 ml in volume)
which is lowered into the well to the desired sampling
depth. The syringe plunger is then pulled open by a
remote  method,  either mechanical or  pneumatic,
allowing the syringe to fill.

The  remote operation allows  the collection  of  a
sample at a discrete depth.  In addition,  the interior of
the sampler (i.e., the syringe) is not exposed to the
water column.  Disadvantages to this device include
the small volume of sample that can be collected,  it
cannot be adapted for evacuation/purging uses, and it
is not readily commercially available.

3.2.9  Ground-Water Sampling
        Equipment  Selection Factors

The  following factors should be  considered  when
selecting ground-water sampling equipment.

•   Composition  -  Select the composition of the
    sampling  equipment based  on  the  sampling
    parameters and objectives.  For  example, use
    samplers made of Teflonฎ, glass, or stainless
    steel instead of PVC when sampling for VOCs.
    Consider  well composition  when  selecting
    sampling equipment.   For example, select  a
    stainless steel  bailer when bailing a well with
    stainless steel casing to avoid the introduction of
    organic constituents.  When sampling a PVC-
    cased well,  PVC, stainless steel, or Teflonฎ
    bailers may be used.

•   Physical Constraints - Physical constraints of the
    monitoring well location, power availability, and
    topography are factors that affect selection of
    ground-water sampling equipment.  For example,
    a small diameter or particularly deep well may
                                                 21

-------
    require the use of different purging and sampling
    equipment than that used for other wells at the
    site.  Site accessibility may hinder  the use of
    large or vehicle-mounted equipment.

    Sample Analysis - Equipment should be chosen
    based on its impact on the samples. For example,
    sampling equipment selected for collecting VOCs
    should agitate the water as little as possible.  This
    is not as critical for metals  or other non-volatile
    analyses.

    Ease of Use  - Generally, the more complicated
    the sampling equipment is, the greater the chance
    for some form of failure in  the field.  Utilize the
    simplest effective  sampling devices available.
    Adequate training in equipment safely and use is
    critical to  personnel safety  as well as to sample
    representativeness.       Consider   ease   of
    decontamination  when  using  non-dedicated
    equipment.
3.3    GEOPHYSICAL  METHODS

Geophysical methods can be useful in  conjunction
with  screening  and  sampling activities  to help
delineate   subsurface   features  and   boundaries,
contaminant plumes, and bedrock types.  Geophysical
data can be obtained relatively rapidly, often without
disturbing the site.  The data are helpful in selecting
well locations and screen depths.   The following
sections discuss surface and borehole geophysics and
preferable geophysical techniques for ground-water
investigations.

3.3.1  Surface Geophysics

The following surface geophysical techniques may be
useful in ground-water investigations. As implied by
the name,  these techniques  are  performed  above
ground.  For more detailed information on each of
these techniques (with the exception of gravimetric
surveys), see ERT  SOP #2159 and Driscoll, 1986.
For more information on gravimetric  surveys, see
Driscoll,  1986.

•   Ground Penetrating Radar (GPR) - Uses  a high
    frequency transmitter that emits radar pulses into
    the subsurface.  These waves  are  scattered at
    points of change in the dielectric permittivity of
    the subsurface material and are reflected back to
    an antenna. (Dielectric permittivity is a function
    of bulk density, clay content, and water content of
    the subsurface.) The returning energy wave is
then plotted as a function of time on an analog
plot.  Interpretation of the analog plot identifies
anomalies, clay layers, and water content in the
substrate.

GPR works best in dry,  sandy  soil above the
water table, and  at depths  between 1 and  10
meters (although the full instrument depth range
is less than one meter to tens of meters).  When
properly  interpreted, GPR  data can  indicate
changes in soil horizons, fractures, and other
geological      features,      water-insoluble
contaminants,  man-made buried objects, and
hydrologic features such  as water table depth.
Uneven ground surfaces or cultural  noise affect
GPR results.

Electromagnetic Conductivity (END - Relies  on
the detection of induced electrical  current flow
through geologic strata. This method measures
bulk  conductivity  (inverse of  resistivity)  of
subsurface materials below  the transmitter and
receiver. EM is commonly used in the detection
of ground-water pollution, as well  as to  locate
pipes, utility lines, cables, trenches, buried steel
drums, and other buried waste.

EM has limited applications in areas of cultural
noise, including above-ground power lines and
metal  fences,  and lateral geologic variations
which might be misinterpreted as  contaminant
plumes.

Electrical  Resistivity - Used to map subsurface
structures  through differences in their resistance
to  electrical current.  Material resistivities are
measured  as functions of porosity, permeability,
water solution, and concentrations of dissolved
solids in pore fluids.  Bulk resistivity  is measured
in the subsurface by measuring electrical currents
injected through electrodes placed in the soil.

Electrical  resistivity  surveys are  limited  by
electrical noise, such as occur in industrial areas.
Resistivity surveys  should ideally be conducted
in  areas removed from pipelines and grounded
metallic structures such  as metal  fences and
railroad tracks. This requirement precludes use
of electrical resistivity surveys on  many sites.
Resistivity can often be used off site to map area
stratigraphy.    Resistivity  surveys  are  labor
intensive, requiring ground setup and removal of
electrodes for  each station measurement.  Use
extreme   care  during  rain or  wet  ground
conditions.
                                                   22

-------
•   Seismic  Investigations  -  Conducted by  two
    methods:  refractive  and  reflective.   In  the
    refractive method,  the  travel time of acoustic
    waves is measured as they move through and are
    refracted along an  interface  of the subsurface.
    The  reflective method measures travel time of
    acoustic  waves as they  are reflected off  an
    interface.   Seismic  refraction is typically used
    when bedrock is  within 500  feet of the ground
    surface.

    Seismic refraction is useful for mapping discrete
    stratigraphic  layers and therefore can help in
    selecting monitoring well locations and depths.
    A   seismic  refraction  survey  can   provide
    subsurface  stratigraphic and  structural data in
    areas  between  existing  wells  or  boreholes.
    Seismic reflection is used less often in ground-
    water investigations, but is more commonly used
    for deeper and larger-scale stratigraphic mapping
    (e.g., petroleum exploration).

•   Magnetic Investigations - Rely  on local variations
    in the earth's magnetic field to detect ferrous or
    magnetic objects. By mapping variations in the
    concentrations  of the  local  magnetic  fields,
    detection of buried objects such as drums or tanks
    may be accomplished.   Magnetic  surveys  are
    limited by  cultural  noise such as power lines,
    utilities, and metal structures.

•   Gravimetric Surveys - Measure small  localized
    differences in the earth's gravity field caused by
    subsurface  density  variations,  which may  be
    produced by  changes  in rock  type (porosity  and
    grain type), saturation, fault zones,  and varying
    thickness of unconsolidated sediments overlying
    bedrock.  This method is useful in identifying
    buried valleys, particularly in glaciated areas.

    Gravimetric surveys use a portable gravity meter
    which can survey  a large area relatively quickly.
    The  accuracy of the readings  is dependent upon
    the  accuracy  of the elevation determination of
    each station.  (Most altimeters are accurate only
    to plus or minus 2 ft (0.6 m),  so gravity stations
    should be surveyed.)  A gravimetric survey  can
    provide a quick preliminary screening of an area.
    Other geophysical methods or test drilling  can
    then be used to help identify stratigraphy  and
    aquifer characteristics.

Table 1 illustrates the applicability of various surface
geophysical    techniques     to    ground-water
investigations.  Table 2 lists some advantages  and
disadvantages of surface geophysical techniques to
ground-water investigations.

3.3.2  Borehole Geophysics

The following borehole geophysical techniques may
be useful  in ground-water investigations.  Borehole
geophysics may  be used  alone  or  to supplement
surface  geophysical techniques.  Site terrain is  an
important  factor   when    conducting  borehole
geophysical surveys.  Much of the equipment is
mounted or housed inside a truck but can be carried to
well locations if necessary.  Some borehole logs can
be run in a cased as well as open hole.

Often several of the following tests  are run at the
same  time for comparative  purposes.   Borehole
geophysical logs can be interpreted to determine the
lithology, geometry, resistivity, formation factor, bulk
density,  porosity, permeability, moisture content, and
specific yield of water-bearing formations as well as
the chemical  and physical characteristics  of ground
water.   The  operating  principles of the  various
borehole geophysical techniques are similar. A sonde
(a cylindrical tool containing one or more sensors) is
lowered to the bottom of the borehole, activated, and
slowly  withdrawn.   Signals or measurements  at
various   depths  are  recorded   at   the  surface.
Instruments vary from hand-held portable  gear to
truck-mounted, power-driven equipment.  For more
detailed information on each of these techniques, see
Driscoll, 1986.

•   Resistance Logs - Electric logs measuring the
    apparent  resistivity  of  the  rock  and  fluid
    surrounding a well. They are good indicators of
    subsurface  stratigraphy  and   water  quality.
    Electric  current is measured as it flows from
    electrodes in the probe to other electrodes in the
    probe or on the ground surface.

    Resistance  logs  have  a  small  radius  of
    investigation   and  are   very    sensitive  to
    conductivity  of borehole fluid and changes in
    borehole  diameter.    Increases  in  formation
    resistivity  produce corresponding increases in
    resistance measurements on the log. Deflections
    on the log are interpreted as changes in lithology.
    Because  of its excellent  response to lithology
    changes,  the resistance  log is very useful for
    geological correlation.    Formation fluids are
    perhaps   the  most   important  variable  in
    interpreting resistance  logs.  For  example, dry
    sands and clays have high resistivities, but their
    resistivities  decrease  with  water  saturation.
                                                    23

-------
Dissolved minerals also affect resistivity.  (Fresh
water is a poor conductor whereas salt water is a
good conductor; water in saturated clays contains
dissolved minerals from the clay, which results in
high conductivities.)

A limitation of resistance logs is that they can be
run only in uncased boreholes that are filled with
drilling fluid and water. Resistance logging is
therefore most appropriately conducted  before
monitoring well completion.

Spontaneous Potential  CSP) Logs -  Used  in
conjunction with  resistivity  logs  to  show the
naturally  occurring electric  potentials  of the
chemical  and  physical changes   at  contacts
between differing types of geologic  materials.
The  electric  current is  measured between an
electrode placed in an uncased borehole and one
placed at the surface.

SP response is due to small voltage differences
caused  by  chemical   and  physical  contacts
between the borehole fluid and the surrounding
formation.     Voltage  differences  appear  at
lithology changes or bed boundaries and  their
response is used to quantitatively determine bed
thickness  or   formation   fluid   resistivity.
Qualitative interpretation of the data can be used
to identify permeable beds.

Buried cables, pipelines, magnetic storms, and the
flow of ground water can  all cause anomalous
readings. Caution must be exercised when using
SP data in a quantitative fashion. Mathematical
formulas are structured for oil well logging and
incorporate assumptions which may not apply to
fresh water wells.  As with resistance logs, SP
logs  can be run  only  in uncased, liquid-filled
boreholes.

Gamma Logs - Measure the naturally occurring
gamma radiation  emitted  from the  decay  of
radioisotopes normally  found in the  substrate.
Elements that emit natural  gamma radiation are
potassium-40  and daughter  products of  the
uranium and thorium decay series.  Changes in
radiation levels are commonly associated  with
differences in substrate composition.

Gamma  logs  can be  run in open  or  cased
boreholes filled with water or air.  The sensing
device  can be part  of the  same sonde that
conducts SP and resistance logs.  Gamma rays or
photons are measured and plotted as counts per
minute. This method is useful in identifying clay
layers  or other  naturally radioactive geologic
units.

Gamma logging is used to identify the lithology
of detrital sediments,  where the finer-grained
units have higher gamma intensity.  (Fine-grained
materials also tend to  have lower permeability
and effective porosity, important for evaluating
aquifer zones.)  A  limitation with gamma and
other nuclear logs  is that they are affected by
changes in borehole diameter and borehole media
(e.g., air,  water, or mud). Gamma logs record the
sum of the radiation emitted from the formation
and  do  not  distinguish  between  radioactive
elements.  For use in stratigraphic correlation
however, specific element identification is not
critical. Interpretation of gamma logs is difficult
where sandstone and other strata contain volcanic
rock fragments with radioactive minerals (e.g.,
rhyolite).   Interpretation  is also  difficult  in
sandstone  containing   a  large  proportion  of
feldspar  (which contains radioactive potassium-
40).

Gamma-Gamma Logs - Similar to gamma logs
except that  a radioactive  gamma source  is
attached to the  gamma sonde  and the  gamma
particles   reflected  back  from  the  geologic
formation are measured.  Gamma-gamma logs
measure  the differing bulk densities of geologic
materials. They can be used to identify lithology
and  also  to calculate porosity  when  fluid and
grain density are known.

Neutron Logs - Also utilize a radiation source in
the sonde.  The neutron source is a europium-
activated,  lithium  iodide  crystal enriched  in
lithium-6.  The neutron logging tool bombards
the formation with neutrons and measures the
returning radiation.  Neutrons, when ejected from
a nucleus, have great penetrating power and may
travel  through  several  feet  of  subsurface
formation.  All free  neutrons are  eventually
captured by the nuclei of some element.  Neutron
logs respond primarily to hydrogen density.  The
high energy neutrons from the source are slowed
by collision with hydrogen ions in the formation.
This response to hydrogen  ion content is  then
cross-calibrated to porosities for water-saturated
rocks.  Neutron logs respond to  the  hydrogen
content  in  the  borehole  and  surrounding
formation and indicate the porosity of the various
geologic units in the survey.  Neutron logs can be
run in cased or open  holes which are dry or filled
with fluid.
                                               24

-------
Neutron  logs  are  typically  used to determine
moisture content above the water table and total
porosity below the water table.  Neutron logs are
effective for identifying perched water tables.
Neutron  log information  can  also  be used to
determine  lithology  and  conduct  stratigraphic
correlation of aquifers and associated formations
as well as to help determine the effective porosity
and specific yield of unconfined aquifers.

Acoustic CSonic') Logs - Measure the travel time
and attenuation of an acoustic  signal created by
an electromechanical source in the borehole. A
transmitter in the borehole converts the electrical
energy to acoustic  (sound) energy which travels
through the formation as an acoustic pulse to one
or more receivers.  The acoustic energy is then
converted back to electrical energy,  which is
measured at the surface.   The  acoustic wave
velocity  is  affected  by the type  of  material
through which it passes (rock or sediment is more
conductive than is pore fluid), hence it is useful in
determining porosity.

Acoustic logs can help determine fracture patterns
within   semiconsolidated   and   consolidated
bedrock  such  as sandstone, conglomerate,  and
igneous rocks.  Knowledge of fracture patterns in
an aquifer is helpful in estimating ground-water
flow, and thereby  estimating the rate  of plume
movement.  Acoustic logs can be used to locate
the static water level and to detect perched water
tables.

Temperature Logs - Used to measure the thermal
gradient  of the borehole fluid.   The  sonde
measures changes in temperature of the fluid
surrounding it, and the log records resistivity  as a
function  of  temperature.    Borehole  fluid
temperature is influenced by fluid movement in
the borehole and adjacent strata.  In general, the
temperature   gradient  is  greater  in   low
permeability rocks than in high  permeability
rocks,   likely  due   to  ground-water  flow.
Temperature logs provide information regarding
ground-water   movement   and  water  table
elevation.   Temperature  logs are useful  for
detecting  seasonal  recharge  and  subsurface
infiltration of irrigation and industrial wastewater
runoff,   and   quantitative  interpretation   of
resistivity logs.

Temperature logs  are designed to be operated
from the top to the bottom of the borehole, in
order to channel water past the sensor.  Repeat
temperature logs should  be delayed  until the
    borehole fluid has had time to reach thermal
    equilibrium.

Table 3 illustrates the applicability of various borehole
geophysical    techniques     to     ground-water
investigations.  Table  4 lists some advantages and
disadvantages of borehole  geophysical techniques to
ground-water investigations.

3.3.3  Geophysical  Techniques  for
        Ground-Water Investigations

The   following   situations   illustrate   uses   for
geophysical techniques in ground-water assessment.

•   To define the location,  extent, and the movement
    of a  contaminant plume,  several geophysical
    techniques may  be  utilized,  including  EM,
    electrical  resistivity,  and   possibly   GPR.
    Resistivity and spontaneous potential (SP) logs
    could also be utilized as borehole geophysical
    methods.

•   To locate faults and  fracture  systems, seismic
    refraction and reflection and EM are the preferred
    methods,  but  GPR,  electrical resistivity  and
    acoustic logs could also be used.

•   The   mapping of  grain  size  distribution  in
    unconsolidated sediments is not  possible with
    any  geophysical  technique.    It  is  possible,
    however, to  identify different soil  types  of
    different grain sizes (e.g., sand, silt, and clay).
    Seismic  reflection and  refraction, GPR,  and
    gravimetric surveys may be  used to  identify
    differing   formations.     Several   borehole
    geophysical techniques could also be utilized in
    this type of analysis, including gamma, gamma-
    gamma, neutron porosity, resistivity, and SP logs.

•   Definition of lithologic  boundaries may  be
    accomplished  with  seismic   reflection  and
    refraction and with GPR techniques. When using
    borehole geophysics, resistivity, SP, and acoustic
    logs are useful.

•   For mapping  water tables, GPR  and  electrical
    resistivity are preferred but seismic refraction and
    reflection and gravimetric surveys may also be
    used.   If using  borehole geophysics,  direct
    measurement  or temperature logs would be  the
    method of choice.  Resistivity and  SP logs could
    also be used.

-------
    To  define  the bedrock  topography,  seismic
    refraction and  reflection, GPR  and gravimetric
    surveys may be used.

    For delineating stratigraphic layers or subsurface
    features,  such as  buried  stream  channels  and
    lenses, seismic refraction and reflection, electrical
    resistivity, gravimetric surveys, and possibly GPR
    could be used.
collected directly from the residential taps into sample
containers. For Phase 2, soils were collected from the
near surface (0-4 inches) and at depth. Stainless steel
trowels  were used to retrieve shallow soil samples.
Subsurface samples were collected by  advancing
boreholes using a hand-operated power auger to just
above the sampling zone  and then using  a stainless
steel split spoon to retrieve the soil. The split spoon
was advanced with a manual hammer attachment.
3.4    EXAMPLE SITE

3.4.1  Selection  of  Field  Analytical
        Screening Techniques

Phase 1 sampling identified the sources and types of
on-site contaminants in order to establish a threat.
Hazard  categorization techniques,  organic  vapor
detecting instruments (FID and PID), and radiation
and cyanide monitors  were  utilized to tentatively
identify containerized liquid wastestreams in order to
select  initial judgmental soil  sampling locations.
During Phase 2 sampling, a portable XRF unit was
used to determine the extent of soil contamination and
to identify  additional "hot spots." A FID and PID
continued to be utilized throughout all field activities
for health and  safety monitoring during Phases  1
through 3.

The portable XRF  for soil screening was also used
during monitoring well installation. Continuous split
spoon samples were collected  during advancement of
the boreholes. Each spoon was sampled and screened
in the field using the XRF unit.  Selected samples (one
per borehole  location)  were  submitted  to  the
laboratory for confirmation analysis.  One off-site
sample was selected by the field geologist based on
field observations and professional judgment.

Ground-water samples were screened in the field for
pH, specific conductivity, and temperature using a
three-in-one monitoring instrument.  The instrument
probe was placed into a clean glass jar containing an
aliquot of the ground-water sample.  The instrument
was decontaminated prior to and after each sample
screening.
3.4.2  Selection of Sampling
        Equipment

Dedicated plastic scoops were used for Phase 1  soil
sampling.    Phase  1  ground-water samples were
Monitoring wells were installed using a dual-tube, air
percussion drill rig.   Borehole soil samples were
retrieved  using 2-foot stainless steel split spoon
samplers.  Soil from the split spoons was transferred
to sample containers using disposable plastic scoops.
Monitoring well installation is described further in
Section 4.6.1.

Ground water was  sampled in Phase 2  from the
monitoring wells installed on site.  First, monitoring
wells were purged using a 1.5 gallon per minute (gpm)
submersible rotary pump with flexible PVC outflow
hose and  safety cable.  The pump and hose were
decontaminated between well locations by pumping
deionized water through the system.  A similar pump
and hose system was used to perform the  hydraulic
(pumping) test. The pumps are operated  by a gas-
powered generator placed near the well location.

The  ground-water samples were  obtained  using
dedicated bottom-fill Teflonฎ bailers.  The bailer was
attached to nylon  rope, which was selected because
less material would be adsorbed onto the nylon and
brought out  of the well.  Residential ground-water
samples were collected directly  into the  sample
containers from the kitchen sink tap.  Water level and
depth measurements were obtained from monitoring
wells using decontaminated  electronic measuring
equipment.

3.4.3  Selection of Geophysical
        Methods

The  GPR instrument  delineated buried trench  and
lagoon  boundaries.    The EM   meter  detected
subsurface conductivity changes, thereby identifying
buried metal containers and contaminants.  The EM-
3 ID, a  shallower-surveying instrument than the EM-
34,  was  selected  because  of the  instrument's
maneuverability and ease  of use, and because the
expected contaminant depth was less than 10 feet.
                                                  26

-------
Table 1: Applicability of Surface Geophysical Techniques to Ground-Water Investigations

Contaminant
Plume Delineation
Faults/Fracture
Detection
Lithologic
Boundary
Delineation
Bedrock
Topography
Delineation
Stratigraphic
Mapping
Water Table
Mapping
Soil Type of
Unconsolidated
Sediments
Metallic Detection
Non-Metallic
Detection
Seepage Detection
Buried Structure
Detection
Seismic
Reflection


P

P

P

P
A

P







Seismic
Refraction




P

P

P
A

P







Electromagnetic
Conductivity
P

P










P


A


Magnetic
Investigations













p



A

Ground
Penetrating
Radar
A

A

A

A

A
P

P

P


P
A

Electrical
Resistivity
P

A





P
P



A
P

A




A



A

P
A

P







P - Preferred Method A - Applicable Method (in most cases)
                                        27

-------
 Table 2:  Advantages and Disadvantages of Surface Geophysical Techniques to Ground-Water Investigations
                                       Advantages
Seismic Reflection
1 Ability to discern discrete layers
1 Less offset space is required than for
 refraction
Velocities 10-20% of true velocities
Data collection and interpretation are
more labor intensive and complex than
for refraction
Depth data not as precise as refraction
Signal enhancement needed to identify
reflected waves
Seismic Refraction
1 Relatively precise depth can be
 determined
1 Provide subsurface data between
 boreholes
1 Ability to map water table and top of
 bedrock
Data collection can be labor intensive
Large geophone line lengths needed
Electromagnetic
Conductivity
1 Lightweight, portable equipment
1 Continuous or quick scan survey
1 Rapid data collection
Interference from cultural noise and
surface metal objects
Limited use where geology varies
laterally
Magnetic
Investigations
1 Can survey large area quickly and cost
 effectively
1 Little site preparation needed
Interference from cultural noise, and
large metal objects
Unable to differentiate between steel
anomalies
Ground Penetrating
Radar
 Can survey large area quickly
 Continuous real-time data display
 Quick data processing
Interference from cultural noise,
uneven terrain, and vegetation
Clay content and shallow water table
inhibit radar penetration	
Gravimetric Surveys
1 Can survey large area quickly
1 Little site preparation
Accurate elevations require surveying
Should be used only as preliminary
screening tool	
Electrical
Resistivity
 Quantitative modeling can estimate
 depth, thickness, and resistivity of
 subsurface layers
Interference from cultural noise,
surface metal objects, and industry
A minimum of two to three crew
members is required
Surveys are labor intensive
                                                      28

-------
Table 3: Applicability of Borehole Geophysical Techniques to Ground-Water Investigations

Contaminant
Plume Delineation
Faults/Fracture
Detection
Lithologic Boundary
Delineation
Bedrock
Topography
Delineation
Stratigraphic
Mapping
Water Table
Mapping
Soil Type of
Unconsolidated
Sediments
Resistance
Logs
P



P
P

P
A

P

Spontaneous
Potential
Logs
P



P
P

P
A

P

Gamma
Logs
P



P
P

P


P

Gamma-
Gamma
Logs




P
P

P


P

Neutron
Logs




A
A

P
P

P

Temperature
Logs
P







P





P

A
P

P
A



P - Preferred Method A - Applicable Method (in most cases)
                                        29

-------
 Table 4:  Advantages and Disadvantages of Borehole Geophysical Techniques to Ground-Water Investigations
                                      Advantages
Resistance Logs
 Indicates lithologic changes
 Indicates amount and type of subsurface
 fluid (water quality)	
Can only be run in uncased borehole
Difficult to interpret lithology when
using drilling fluid with clay additives
Spontaneous
Potential Logs
 Can be run in conjunction with resistance
 log
 Indicates lithologic changes and
 permeable beds	
Can only be run in uncased borehole
Interpretation for water well often
more difficult than for oil well
Gamma Logs
1 Easy to operate
1 Can be run in open or cased borehole
1 Qualitative guide for stratigraphic
 correlation and permeability	
Affected by changes in borehole
diameter and borehole media
Feldspar and volcanic rock fragments
make interpretation difficult	
Gamma-Gamma
Logs
 Can identify lithology and calculate
 porosity when fluid and grain density are
 known
Porosity readings of low density
materials can be erroneously high
Neutron Logs
 Can determine total porosity in saturated
 zone
 Can determine moisture content in
 unsaturated zones
 Can be run in open or cased borehole
Radioactive source requires special
handling by trained personnel
Logging can be somewhat complex
Acoustic Logs
1 Useful for determining relative porosity
1 Indicates fracture patterns in aquifer
1 Can indicate static water level and
 perched water tables	
 Clays may distort readings
Temperature Logs
1 Can indirectly measure permeability
1 Provides information regarding
 ground-water movement and water table
 elevation
Delay repeat logs until borehole fluid
reaches thermal equilibrium
                                                     30

-------
   4.0   GROUND-WATER SAMPLE COLLECTION AND PREPARATION
4.1     INTRODUCTION

During  a response  action, proper  field sample
collection and preparation is as important  as proper
sampling equipment  selection.  Sample collection
refers to the physical removal of an aliquot of ground
water from its source (i.e., aquifer) for the purpose of
either screening or laboratory analysis. Ground-water
sample collection  procedures should be selected so
that  the resultant sample  is representative of the
aquifer or particular water zone being sampled. Field
sample preparation refers to all aspects of sample
handling from collection to the time the sample is
received by  the laboratory.  This chapter provides
information on sample collection and preparation for
ground water.

The  representativeness of a ground-water sample is
greatly influenced by the sampling device used and
the manner in which the sample is collected. Proper
training and use of SOPs  will limit variables  and
enhance sample representativeness.    Selection of
ground-water sampling devices such as bailers  and
pumps should be site-specific and dependent on well
diameter, yield, lift capacity, and the analytes being
sampled. Excessive aeration should be minimized to
preserve volatile constituents.  Where possible, the
bailer or pump used should be compatible with the
analyte(s) of concern.
4.2    STATIC WATER LEVEL

Prior to sampling, the static water level elevation in
each well should be measured.  All measurements
should be completed prior to the sampling event so
that static water levels will not be affected.  The water
level measurements are necessary  to establish well
purging volumes. These measurements can also be
used to construct water table or potentiometric surface
maps  and hence determine local ground-water flow
gradient.  Measure the depth to standing water and the
total depth of the well to calculate volume of stagnant
water in the well for purging.  See ERT SOP #2151
for  detail  on  collecting  static  water   level
measurements.
4.3    WELL PURGING

There is little or no vertical mixing of water in a
nonpumping well, therefore stratification occurs. The
well water in the screened section mixes with the
ground water due to normal flow patterns, but the well
water above the screened section will remain isolated
and become stagnant. The stagnant water may contain
foreign   material  inadvertently  or  deliberately
introduced  from   the   surface,  resulting   in
unrepresentative data. Adequate well purging prior to
sample withdrawal  will safeguard against collecting
nonrepresentative stagnant water samples.

Well purging techniques are specific to the following
well types.

•   Residential.  Commercial,  and Public Supply
    Wells -  Sample residential,  commercial, and
    public supply  wells as near to the wellhead as
    possible and at a point before treatment, such as
    filtering and water  softening units, whenever
    possible.  Open the tap to a moderate flow and
    purge for approximately 15 minutes.  If this is not
    possible,  a  5-minute  purge  is considered  a
    minimum.   As an  alternative to  a minimum
    volume, purging can be conducted until the field
    parameters  pH,  temperature,  and  specific
    conductivity have stabilized (see Section 4.3.1).

•   Monitoring Wells - To obtain a representative
    sample from a monitoring well, it is necessary to
    evacuate the standing  water in the well casing
    prior to sampling. The minimum recommended
    amount that should be purged from a monitoring
    well  is  one casing  volume, but three to five
    casing volumes  of  standing  water should  be
    evacuated where possible in order to obtain a
    ground-water  sample  representative  of  the
    aquifer. In a high yield aquifer where there is no
    standing water  above the screened section of the
    well casing, purging three volumes is not  as
    critical as in lower yield aquifers.   (The  faster
    recharge rate limits the amount of time that the
    water has to interact with the atmosphere and
    casing materials.)  If the well is purged dry, it
    should be considered  sufficiently  purged for
    sampling (refer to  Section 4.3.2 for additional
    information).
                                                 31

-------
The amount of purging a well receives prior to sample
collection depends on the intent of the sampling as
well as the hydrogeologic conditions.   When the
sampling objective is to assess overall water resource
quality, long pumping periods may be required to
obtain a sample  that is representative  of  a  large
volume of the aquifer.    The pumped  volume is
determined prior to sampling, or the well is pumped
until  the  stabilization   of parameters  such  as
temperature,  specific  conductivity,  and pH  has
occurred.

Monitoring to define a contaminant plume requires a
representative sample of  a small volume  of the
aquifer.  These circumstances require that the well be
pumped enough to remove the stagnant water but not
enough to induce flow from other areas.  Generally,
three   well   volumes  are  considered   effective.
Otherwise, the appropriate volume to be removed
prior to sampling can be calculated, based on aquifer
parameters and well dimensions.

Well purging devices include  bailers, submersible
pumps (rotary-type), non-gas contact bladder pumps,
suction pumps, and hand pumps.  See  ERT  SOP
#2007 for specific guidelines on purging wells  prior to
sampling and for more detail on each purging device.

4.3.1  Stabilization        Purging
        Technique

The stabilization technique is an alternative to  volume
purging.   This  method  requires that several  field
parameters be continuously monitored during purging.
When these parameters stabilize, begin sampling. The
parameters used for this method are pH, temperature,
and specific  conductivity.   Stabilization of  these
parameters indicates that the standing water in the
monitoring  well  has been removed and  that  a
representative sample of the aquifer water may now
be collected.  This method of purging is useful in
situations where it is not feasible  to evacuate  three
casing volumes from the well prior to sampling  (e.g.,
large  casing diameter, extremely  deep,  and active
supply wells).  See ERT  SOP #2007 for specific
volume and stabilization  purging techniques.

4.3.2  Wells that Purge Dry

A well that is purged  dry  should be evacuated and
allowed to recover prior to sample withdrawal. If the
recovery  rate is  fairly  rapid and  time  allows,
evacuation of more than  one  volume of water is
desirable.   If the  recovery rate is  slow, the first
recharge can  be considered suitable  for  sample
collection.
4.4    GROUND-WATER SAMPLE
        COLLECTION

In  order  to  maintain sample  representativeness,
dedicated samplers should be used for each well
whenever possible.  When not possible, the sampler
should be decontaminated after each sample collection
and  sufficient  QA/QC blank samples should  be
collected to assess potential cross-contamination.

After  well   purging   is  complete,  collect and
containerize samples in the order of most volatile to
least volatile, such as:

•   Volatile organic analytes (VOAs)
•   Purgeable organic carbon (POC)
•   Purgeable organic halogens (POX)
•   Total organic halogens (TOX)
•   Total organic carbon (TOC)
•   Extractable organic compounds
•   Total metals
•   Dissolved metals
•   Phenols
•   Cyanide
•   Sulfate and chloride
•   Turbidity
•   Nitrate and ammonia
•   Radionucliides

See ERT SOP #2007 for specific detail on  filling
sample  containers,  with special  considerations  for
VOA sampling.

If the contaminants in the water column are stratified
(e.g.,  DNAPLs, LNAPLs),  be  certain to use  an
appropriate sampling device. Modify, where possible,
standard sampling procedures to  collect the sample
from the suspected depth for the contaminant layer. It
may be necessary to  lower the bailer used for sample
collection to a particular depth in the well, or to use a
point-source bailer or other discrete-depth sampling
device.

After a monitoring well  is initially constructed, it
should be developed and purged to remove invaded
water. The well should sit  idle for at least two  weeks
to allow the  water  level  to fully stabilize and the
suspected stratified layers to settle  out. Measurement
of the thickness of a floating (LNAPL) layer may be
accomplished  in several ways.  An indicator gel, chalk
or paste may be applied to an incremented steel tape.
                                                  32

-------
The indicator changes color in the presence of water
or the immiscible layer, depending on the specific use
of the indicator compound. For example, water-level
indicator gel is used to determine the depth to the
water surface.   A  weighted float is then used to
determine the depth to the top of the LNAPL layer.
The  difference between these two readings is the
thickness of the floating layer.

An electronic monitoring device called an interface
probe is  also  available for the LNAPL  layer
measurement. This device, like an electric water-level
sounder, is lowered  into the well along an electronic
wire/line.  When the probe contacts the surface (the
LNAPL layer) a sound is generated. As the sampler
continues  to lower  the probe, a  different electronic
sound is emitted when the water surface, or water/oil
interface,  is reached.   The  line of the device is
incremented, like a water-level sounder, so the layer
thickness can be determined. Standard electric water-
level  sounding  devices, however, will  not  work
properly for these measurements.  The interface probe
is a  specialized instrument  which is commonly
available  and  used   at   fuel  oil/ground-water
contamination sites.

A sample of a floating layer may  be obtained using a
bottom-fill bailer. Care should be taken to lower the
bailer just  through  the  floating  layer,  but not
significantly down into the underlying ground water.
(A clear bailer is preferable for this activity.)

For sampling sinking layers, a discrete-depth-capable
sampling device, such  as a packer pump or syringe
sampler, is  best suited.  When these specialized
devices are not available, depending on the sampling
parameters,  standard  devices may be used.   For
example,  samples at the bottom of the screen or at
some intermediate location may also be obtained with
a standard bailer and  a second well casing. In order to
avoid  mixing  the  waters,   a separate  casing is
temporarily lowered inside the permanent well casing.
The  temporary  casing  is  equipped with  an easily
removed cap on the bottom so that no fluid enters the
casing until it has reached the desired sampling depth.
The cap is then  freed from the bottom of the  inner
casing, allowing water to enter to  be sampled by a
bailer.  At significant depths  below the nonaqueous
layer, several bailers full of  water may need  to be
withdrawn  and  discarded before  the  sample is
obtained from a fresh formation sample.

When a temporary  casing and all other specialized
equipment is unavailable, a standard bailer alone may
be used. Collect a  water sample from the well and
transfer it to  the sample container. Allow the sample
to settle in the sample container into the  separate
stratified layers.  The analytical laboratory may then
decant, as appropriate, to obtain  a sample of the
desired layer. More commonly, the parameters of
concern in the stratified layers are simply included in
the laboratory analysis of the sample as  a whole
without the need to separate into unique  layers.  In
this last example,  care must be taken to allow the
bailer to reach the desired depth in the water column
to insure collecting any dense layers at the bottom of
the well.  (See Section 2.4 for additional discussion on
sampling concerns and the physiochemical nature of
contaminants.)
4.5    GROUND-WATER SAMPLE
        PREPARATION

This  section  addresses  appropriate  ground-water
sample preparation and handling techniques. Proper
sample  preparation and  handling maintain sample
integrity.   Improper handling  can render samples
nonrepresentative and unsuitable for analysis.

The analyses for which a sample is being collected
determines the type of bottles, preservatives, holding
times, and filtering requirements. Samples should be
collected directly into appropriate containers that have
been cleaned to EPA or other required standards.
Check to  see that a Teflonฎ liner is present in the
sample bottle cap, if required.

Samples  should  be  labeled, logged, and handled
correctly,  including appropriate  chain-of-custody
documentation.   Place  samples  in  coolers to  be
maintained at 4EC.  Ship samples to arrive at the
designated analytical laboratory  well  before their
holding times  are expired.   It  is preferable that
samples be shipped or delivered daily to the analytical
laboratory  in order to maximize the time available for
the laboratory to do the analysis.

Certain  conditions  may  require  special handling
techniques.  For example, treatment of a sample for
VOAs with sodium thiosulfate preservative is required
if there is residual chlorine in the water (e.g., a public
water   supply)   that  could  cause  free  radical
chlorination and change the identity of the original
contaminants. (The preservative should not be used if
there is no chlorine in the water.) All such special
requirements must be determined prior to conducting
fieldwork.

Sample preparation for ground water may  include, but
is not limited to:
                                                   33

-------
•   Filtering
•   Homogenizing/Aliquotting
•   Splitting
•   Final Preparation

4.5.1  Filtering

Samples may require filtering, such as for total metals
analysis.  Samples collected for  organic analyses
should not be filtered.  Two types of filters may be
used, which must be decontaminated prior to use. A
barrel filter works with a bicycle pump, which builds
up positive pressure in the  chamber containing the
sample and then forces it through the filter into a
container placed underneath.  A vacuum filter has two
chambers; the upper chamber contains the sample, and
a filter divides the chambers. Using a hand pump or
a Gilianฎ-type pressure pump, a vacuum is created in
the upper chamber and the sample is filtered into the
lower  chamber.   Preservation of the  sample, if
necessary, should be done after filtering.

See ERT SOP #2007, Section 2.7.5, for more detail on
filtering ground-water samples.

4.5.2  Homogenizing/Aliquotting

Homogenizing,  or  aliquotting, is the  mixing or
blending of a grab sample to distribute contaminants
uniformly.  Ideally, proper homogenizing ensures that
all portions of the sample are equal or identical in
composition and are representative of the total sample
collected.  Incomplete homogenizing  can introduce
sampling error. Homogenizing disturbs the ground-
water  sample,  so  it is not appropriate for VOC
sampling.

Homogenizing is done during only one sampling event
per well location, and  only after the VOC sample
portions have first been filled. It may be utilized for
wells  with  extremely  low yield  and  potentially
insufficient  sample  volume  to  fill  all  sample
containers provided by the laboratory.  In some low
yielding wells, the percentage of suspended material
in a bailer-full of sample will increase as sampling
proceeds.   Homogenizing  ensures that  at least  a
minimum  volume  is   aliquotted per  analytical
parameter, and the percentage of suspended material
is equitably divided among all containers  (excluding
VOCs).

4.5.3  Splitting

Split samples are created when the samples have to be
separated  into  two or more  equivalent parts  and
analyzed separately.  Split samples  are most often
collected in enforcement actions to compare sample
results obtained by EPA with those obtained by the
potentially responsible  party.   Split samples also
provide measures of sample variability and analytical
error.  Fill two sample collection jars simultaneously,
alternating the sample stream  or bailer full of sample
between them.

4.5.4  Final Preparation

Final preparation includes preserving,  packaging, and
shipping samples.

Sample preservation is used  to  retard  chemical
breakdown of the sample.  Preservation of ground-
water samples includes controlling pH with chemical
preservatives, refrigerating samples,  and protecting
samples from light.

Select sample containers  on the basis of compatibility
with  the  material  being  sampled,  resistance  to
breakage, and capacity. Appropriate sample volumes
and containers will vary according to the parameters
being analyzed.  Actual sample volumes, appropriate
containers, and holding times  are specified in the
U.S.  EPA Compendium of ERT  Ground-Water
Sampling Procedures, OSWER Directive 9360.4-06.
Package  all  samples in compliance with  current
International Air Transport Association (IATA) or
U.S.   Department   of   Transportation   (DOT)
requirements, as applicable.  Packaging should  be
performed by  someone trained in current DOT
shipping procedures.

See ERT SOP #2007, Section 2.3 for more detail on
ground-water sample preparation.
4.6    EXAMPLE SITE

4.6.1  Sample Collection

During  Phase 1  and  Phase 2, surface  soil samples
were collected from shallow locations.  The samples
were collected as grab samples.  The sample locations
were  cleared of  surface debris,  then samples were
retrieved with disposable plastic scoops and placed
directly into sample  containers.  During Phase 2,
subsurface  soil samples were collected at the  soil
boring/well installation locations, using stainless steel
split spoon samplers.  The split spoon samples were
collected using a hand-held power auger to advance
the hole.  A 2-foot stainless steel split spoon sampler
with hammer attachment was then  pushed into the
                                                  34

-------
hole.  The soil sample was retrieved from the split
spoon sampler using  a disposable plastic scoop to
transfer the soil into a stainless steel bowl.  Several
scoopfuls were collected along the length of the split
spoon sampler and composited  in the bowl.  The
composite sample was then transferred directly into
the sample container using the disposable  plastic
scoop.

Phase 1  and  Phase 2  residential well  ground-water
samples were  collected directly from the kitchen taps
of homes using private wells near to the site.  The
configuration of the residential system was noted in
the logbook  prior to  sampling.   If present,  water
softeners were taken off line. Any screen or filter was
first removed  from the tap, which was allowed to run
for a minimum of five minutes prior to sampling. The
samples  were  collected  directly into the sample
containers.

Fifteen monitoring wells were installed at the site at
locations described in  Section 2.5.6.  The wells were
drilled with a dual-tube, air  percussion rig.  Each
boring was completed to a 9.5-inch diameter.  After
completion of the boring, 4-inch Schedule 40 PVC
casing and 0.010 slot screen were installed in lengths
appropriate to each well.  Shallow wells were drilled
to approximately 40 feet below grade surface (BGS)
and   bedrock  contact   wells  were   drilled   to
approximately 55 to 60 feet BGS.  Continuous split
spoon sampling was performed at each well location
from 4 feet BGS  to  well completion  depth.  The
boreholes were grouted from the bottom to the top of
the lower confining layer, then 10 feet of screen were
set above the grouted portion. PVC casing was set
above the screen to above the ground surface. Casing
was extended to accommodate a 2-foot stick-up above
grade, and then capped.   A 6-inch diameter metal
outer casing with locking cover was installed over the
well casing  stick-up  and secured  2  feet BGS in
concrete.  A concrete  spill pad was then constructed
around each well outer casing to prevent re-infiltration
at the well point.   Upon completion, all monitoring
wells  were   developed   by   purging  using   a
decontaminated  rotary  pump  and  flexible  PVC
disposable hose.

A Team geologist supervising the monitoring well
installation logged each borehole soil lithology from
the retrieved  split  spoon samplers collected during
drilling of the  boreholes.  The geologist scanned each
sampler with a PID immediately upon opening (into
halves) for health and safety monitoring. All logging
was   accomplished  utilizing  the  Unified  Soils
Classification System standard  method.   Figure  3
provides an example of a soil boring and monitoring
well completion log.

Soil samples were then collected in wide-mouth clear
glass jars by transferring a portion of each lithologic
unit in the split spoon with a disposable plastic scoop
and compositing the sample  in the  jar.   At the
completion  of each  borehole,  each sample was
screened in the field using the XRF unit.   Select
samples (one per borehole location) were forwarded to
the laboratory for confirmation analysis.  Split spoon
samplers were decontaminated after each use.

Upon completion and development, the  15 on-site
monitoring wells were  sampled for  ground-water
analysis.  The well caps were brushed and cleaned off
prior to  opening.  Immediately upon removing the
well cap, a PID was operated over the opening to
determine VOC levels, if any,  in the breathing zone.
The VOC monitoring was performed to establish if a
higher level of respiratory protection was required.
Depth to water level measurements were then taken of
each well to the nearest 0.01 ft.  The total depth of the
well was obtained with a depth  sounder.  The volume
of water in the well was then calculated using the
formula below.  For a four-inch well, well volume
would equal 0.632 gallon/ft.:

    Well volume =  B x (radius of well)2 x height
               of water column x 7.48 gallon/ft3

         (conversion factor for ft3 to gallons)

Each monitoring well was purged prior to obtaining a
representative sample.  Wells with sufficient yield
were purged three well volumes. Low-yielding wells
were purged once to  dryness.  (Most wells on site are
low-yielding.)  Purging  was completed using a 1.5
gpm decontaminated submersible (rotary-type) pump
with flexible PVC outflow hose and safety cable. The
pump was slowly lowered to a point approximately 3
feet above the bottom of the well.  With the known
flow rate, length of pumping required was calculated.
Purge water was pumped into 55-gallon steel drums.
(The drums were staged and later disposed of properly
based on the results of  analysis of their contents.)
Low-yielding overburden wells were purged with a
decontaminated stainless steel  bottom-fill bailer and
polypropylene rope until dry.  All wells were allowed
to recover overnightbefore sample collection, or until
sufficient water was present to complete a sample set.

Each monitoring well was sampled after purging and
recovery. Ground-water samples were collected using
dedicated disposable Teflonฎ bailers.  Each bailer
was attached to a clean polypropylene rope and intro-

-------
Figure 3: Soil Boring/Monitoring Well Completion Log
JZ
*-ซ
Q.
(D
Q
-
-


-
1 n
I U
-
15
20 —
-
-
•an
oU
_
35
Ww
40
T^W
—
-
45 —
-
-
en
OU

55
Symbol
Stratigraphy
SP/
SM


SM


SP



SC


SP

SM


SM/
SC




SC

Sample Description

Tan. vf. Sd. w. 10% Sit.

As above w. 30% Sit.

Tan. vf. Sd. coarsening
to med. Sd. at 25'



Tan. vf. Sd. w. 10% Gry.
lumpy Cly.
Iron oxide staining
Tan. F-med. Sd.
As above w. 2% Sit. & Cly.

Tan. vf. Sd. & Sit. w. lumps of
Gry. Cly. w. Iron oxide staining



Brn. vf. Sd. & Sit. w. lumps of
Gry. Cly. w. Iron oxide staining
Cly. 30%
Completion Data
Protective steel casing
with locking cap





Portland
Cement
+ 5%
Bentonite
Grout

41 Diam.
PVP
r V \j
Casing


# 2/16
f\ i
Sand





Bottom
^\ -^
Cap ^
/
/
/
/
/
-^— ,
/

/
/
/
/
/
1



	
















•

\
\
\
\
\
^
\
\
v
v
\\
\
\
\\
\
i



a
"B-

•\





	 g 51
Borehole






Bentonite
Slurry

Lonestar
# 2/12
Sand
0.010
slot
screen
TD56ft
                       36

-------
duced into the well. The bailer was lowered slowly to
the approximate mid-point of the  well.  Once the
sample was collected, care was taken not to  agitate the
water while  pouring  directly  into the appropriate
sample containers.   An  additional ground-water
aliquot was placed into a large wide-mouth glass jar
in order to  obtain conductivity, temperature, and pH
measurements. These measurements were recorded in
the field logbook.

After well sampling, a hydraulic (pumping) test was
performed  to determine aquifer characteristics for
mathematical modeling of potential  contaminant
plume migration.  The hydraulic test was  conducted
using  one  well  as a  pumping well with  three
observation wells.  The pumping well was purged at
a rate of 22  gpm for 30 hours.  All wells  (observation
and pumping) were monitored during pumping and for
4 hours after pumping ceased. Drawdown data from
the wells were used to calculate the characteristics of
the aquifer.

To generate accurate gradient and well location maps,
the 15 newly installed monitoring wells were surveyed
for vertical location using feet above mean sea level
(MSL) units. Vertical elevations were taken at a mark
on the top of the inner casing of each monitoring well,
to establish a permanent location for all future water
level measurements and elevations.  A permanent
benchmark was located near to the site by  the survey
team to determine all the well elevations. Elevations
were then  measured  against  the  benchmark  and
mapped in MSL units.

All non-disposable equipment, including drill rig and
equipment,  stainless steel bailers, pumps, water level
indicators, and depth sounders, were decontaminated
between each location and  prior to the first sampling
event each day.

4.6.2  Sample Preparation

All sample containers were supplied by the  contracted
analytical laboratory. Chemical preservation was also
provided by  the  laboratory through pre-preserved
bottleware.   Sample  containers for ground-water
samples consisted of:
•    1-liter polyethylene bottles for total chromium,
     pre-preserved with  reagent-grade  nitric  acid
     lowering the pH to less than 2 after addition of
     the sample

•    1-liter  polyethylene   bottles  for  hexavalent
     chromium

•    1-liter polyethylene  bottles for cyanide,  pre-
     preserved with sodium hydroxide

Sample containers for soils consisted of 8-ounce glass
jars with Teflonฎ caps for all parameters.

All samples were preserved to 4E C by placing them
in coolers packed with "blue ice" immediately  after
collection and during shipment.  (The laboratory was
responsible for  cooling and refrigeration of samples
upon arrival.)

The samples were packaged in compliance with IATA
requirements for environmental samples.  Chain-of-
custody paperwork was prepared for the samples.
Laboratory paperwork was completed as appropriate
and  the samples were shipped to the predesignated
laboratories for analysis.   Holding times for  total
chromium and cyanide are  less than six months, but
hexavalent chromium has a holding time of less than
24 hours.  This  was coordinated in advance  with the
analytical  laboratory   and  required  daily   ground
delivery of samples to the laboratory.

Because many of the ground-water samples from the
on-site wells were extremely turbid, the non-volatile
portions  of samples were filtered in  the laboratory
prior to analysis.  Filtering was accomplished using a
barrel filtering device with a minimum pore size of
0.45 microns. Samples for chromium analysis were
split and filtered so that dissolved and particulate
chromium  could  be  differentiated.    Dissolved
chromium is  of  concern because of its  ability to be
transported in ground water.
                                                  37

-------
         5.0   QUALITY ASSURANCE/QUALITY CONTROL  (QA/QC)
5.1     INTRODUCTION

The  goal  of representative  sampling is  to  obtain
analytical results that accurately depict site conditions
during a defined time interval.   The goal  of quality
assurance/quality control (QA/QC) is to implement
correct methodologies which limit the introduction of
error into the sampling and analytical procedures, and
ultimately into the analytical data.

QA/QC samples evaluate three types of information:
1) the magnitude of site variation; 2) whether samples
were cross-contaminated during sampling and sample
handling procedures; and 3) whether a discrepancy in
sample results is a result of laboratory handling and
analysis procedures.
5.2    DATA CATEGORIES

EPA has established data quality objectives (DQOs)
which   ensure  that  the  precision,   accuracy,
representativeness, and quality of environmental data
are  appropriate for  their  intended  application.
Superfund  DQO  guidance  defines  two  broad
categories  of  analytical  data:    screening  and
definitive.

Screening data are generated by  rapid, less precise
methods of  analysis with  less  rigorous  sample
preparation than definitive data. Sample preparation
steps may  be restricted to simple procedures such as
dilution with  a  solvent,  rather than  elaborate
extraction/digestion and cleanup. At least 10 percent
of  the  screening  data are confirmed  using the
analytical   methods  and QA/QC procedures  and
criteria  associated with definitive data.  Screening
data without  associated confirmation data  are not
considered to be data of known quality.  To  be
acceptable, screening data must include the following:
chain-of-custody, initial and continuing calibration,
analyte  identification,  and analyte  quantification.
Streamlined  QC  requirements  are  the  defining
characteristic of screening data.

Definitive data are generated using rigorous analytical
methods (e.g., approved EPA  reference methods).
These data are analyte-specific,  with confirmation of
analyte identity and concentration.  Methods produce
tangible raw  data (e.g., chromatograms, spectra,
digital values) in the  form of paper printouts or
computer-generated electronic  files.   Data may  be
generated at the site or at an off-site location, as long
as the QA/QC requirements are satisfied. For the data
to be definitive, either analytical or total measurement
error must be determined. QC measures for definitive
data contain all of the  elements associated with
screening data, but also  may include trip, method, and
rinsate blanks; matrix spikes; performance evaluation
samples;   and  replicate   analyses   for  error
determination.

For further information on these QA/QC objectives,
please refer to U.S. EPA Data Quality Objectives
Process for Superfund, pp. 42-44.
5.3    SOURCES OF ERROR

There  are many potential sources  of data error in
ground-water sampling.   The following is a list of
some of the more common potential sources of error:

    •   Sampling design
    •   Sampling methodology
    •   Analytical procedures
    •   Seasonal variations

See U.S. EPA Data Quality Objectives  Process for
Superfund, pp. 29-36, for more information on error.

5.3.1  Sampling Design

The sampling design should utilize approved SOPs
and previously approved sampling designs to ensure
uniformity and comparability between samples.  The
actual sample collection process should be determined
prior   to  sampling.    Sampling  equipment  and
techniques must be standardized for like sampling
situations.

The sampling design should fulfill sampling and data
quality objectives.  The quality assurance objectives
selected  should be built  into the  sampling design,
including all necessary QA/QC samples.

Sampling design errors for ground water include: well
selection, well  location,  well  construction  and
development,  background   sample  location,  and
equipment (material and type).
                                                  38

-------
5.3.2  Sampling Methodology

Sampling   methodology   and   sample  handling
procedures have possible sources of error, including:
cross-contamination from inappropriate use of sample
collection  equipment;  unclean  sample containers;
improper sampling equipment decontamination; and
improper  shipment  procedures.   Procedures  for
collecting, handling, and shipping samples should be
standardized  to allow  easier identification  of any
source(s) of error,  and to  minimize the potential for
error.  Use approved SOPs to ensure that all given
sampling  techniques  are  performed in the  same
manner, regardless of the sampling team, date, or
location of  sampling activity.   Use field  blanks,
replicate samples,  trip blanks,  and rinsate blanks
(discussed in Section 5.4) to identify errors due to
improper sampling methodology and sample handling
procedures. An example of a sampling methodology
error for ground water is inappropriate purging.

5.3.3  Analytical  Procedures

Analytical procedures  may  introduce  errors  from
laboratory cross-contamination, inefficient extraction,
and  inappropriate  methodology.   Matrix  spike,
laboratory duplicate, performance evaluation, and
laboratory  control  samples  help  to  distinguish
analytical error from sampling error.

5.3.4  Seasonal Variations

Seasonal variations are not controllable  but must be
taken into consideration as a source of  error during
ground-water assessments.  Changes in flow direction
or volume can redistribute contaminants throughout a
site, making assessment  difficult.  Plan sampling
events in order to  minimize  the effects of seasonal
variations, if possible.
5.4    QA/QC SAMPLES

QA/QC samples are collected at the site or prepared
for or by the laboratory.  Analysis  of the QA/QC
samples provides information on the variability and
usability of sampling data, indicates possible field
sampling or laboratory error, and provides a basis for
future validation and usability of the  analytical data.
The most common field QA/QC samples are field
replicate, background, and rinsate, field, and trip blank
samples.   The most common laboratory QA/QC
samples  are performance evaluation  (PE), matrix
spike  (MS),  matrix spike duplicate  (MSD),  and
laboratory duplicate  samples.  QA/QC results may
suggest the need for modifying  sample collection,
preparation, handling, or analytical procedures if the
resultant data  do not  meet  site-specific  quality
assurance objectives.

Ground water is typically characterized by low or
trace   concentrations   of  contaminants,  making
precision and accuracy  more important than for
samples  with  higher  concentrations (e.g., waste).
Frequent field blanks are thus appropriate in ground-
water sampling.

The  following  sections briefly describe the most
common types of QA/QC samples  appropriate for
ground-water sampling.

5.4.1  Field Replicate Samples

Field replicates, also referred to as field duplicates and
split samples, are field samples obtained from one
sampling point,  homogenized  (where  appropriate),
divided into  separate  containers,  and  treated as
separate  samples throughout the remaining sample
handling and analytical  processes.   Use replicate
samples  to  assess  error  associated with  sample
methodology  and  analytical  procedures.   Field
replicates can  also be used when determining total
error   for  critical  samples   with  contamination
concentrations near the action level. In such a case, a
minimum of eight replicate samples is recommended
for valid statistical analysis. Field replicates may be
sent to two or  more laboratories  or  to the same
laboratory  as  unique  samples.    For  total  error
determination, samples should be analyzed by the
same laboratory. Generally, one field replicate per 20
samples per day is recommended.

5.4.2  Background Samples

Defining background  conditions may  be difficult
because  of  natural  variability  and the  physical
characteristics of the site, but it is important in order
to quantify true changes in contaminant concentrations
due  to a source  or  site.   Defining  background
conditions is critical for avoiding false positives and
for enforcement purposes  in  naming  responsible
parties.  Background sampling is often required in
ground-water sampling  to verify plume direction,
ambient  conditions, and  attribution  of  sources.  A
properly collected background  sample serves as the
baseline for the measure of contamination throughout
the site. Ground-water background sample locations
should be chosen carefully, usually upgradient from
the suspected source of contamination where there is
little or no chance of migration of contaminants of
                                                  39

-------
concern.  Collect at least one background sample for
comparison, although additional samples are  often
warranted by  site-specific factors  such as natural
variability  of  local  geology and multiple sources.
Background samples may be collected to evaluate
potential error associated  with sampling  design,
sampling methodology,  and analytical procedures.
Refer to U.S. EPA "Establishing Background Levels"
fact  sheet,  OSWER Directive  9285.7-19FS, for
detailed  discussion  on  the proper  selection and
considerations of a background sample location.

5.4.3  Rinsate  Blank Samples

A rinsate blank, also referred to as  an equipment
blank,  is used to assess  cross-contamination  from
improper  equipment  decontamination procedures.
Rinsate blanks are  samples  obtained by running
analyte-free water over  decontaminated sampling
equipment.  Any residual contamination should appear
in the rinsate sample data. Analyze the rinsate blank
for the same analytical parameters as the field samples
collected that day. Handle and ship the rinsate like a
routine field sample.  Where dedicated sampling
equipment is not utilized, collect one rinsate blank per
type  of sampling device per day.

5.4.4  Field Blank  Samples

Field blanks are samples prepared in the field using
certified clean water (HPLC-grade water (carbon-free)
for organic analyses and deionized or  distilled water
for inorganic analyses) which are then submitted to
the laboratory  for analysis.  A field blank is used to
evaluate  contamination  or  error  associated  with
sampling      methodology,      preservation,
handling/shipping,    and   laboratory   procedures.
Handle, ship, and analyze a field blank like a routine
field sample.  Submit one field blank per day.

5.4.5  Trip Blank Samples

Trip  blanks are samples prepared prior to going into
the field.   They consist of certified  clean water
(HPLC-grade) and are not opened until they reach the
laboratory.  Utilize  trip blanks for volatile  organic
analyses only.  Handle, transport, and analyze trip
blanks in the  same  manner as  the  other  volatile
organic samples collected that day. Trip blanks are
used to evaluate  error associated with shipping and
handling and  analytical procedures.   A trip blank
should be included with each shipment.
5.4.6  Performance  Evaluation/
        Laboratory Control Samples

A performance evaluation (PE) sample evaluates the
overall error contributed by the analytical laboratory
and detects any bias in the analytical method being
used.  PE samples contain known quantities of target
analytes manufactured  under strict quality control.
They  are usually prepared by a third party under an
EPA certification program.  The samples are usually
submitted "blind"  to  analytical  laboratories  (the
sampling team knows the contents of the samples, but
the laboratory does not). Laboratory analytical error
may  be evaluated by  the  percent  recoveries and
correct identification of the  components in the PE
sample.  Note: Even though they are not available for
all   analytes,   analyses  of PE   samples   are
recommended in order  to obtain definitive data.

A blind PE sample may be included in a set of split
samples provided to the potentially responsible party
(PRP). The PE sample  will indicate PRP laboratory
accuracy, which may be critical during enforcement
litigation.

A laboratory control sample (LCS) also  contains
known quantities of target analytes in certified  clean
water. In this case, the laboratory knows the contents
of the sample  (the LCS is usually prepared by the
laboratory).  PE and LCS samples are not affected by
matrix interference, and thus can provide a  clear
measure of laboratory error.

5.4.7  Matrix Spike/Matrix Spike
        Duplicate Samples

Matrix spike and  matrix spike duplicate  samples
(MS/MSDs) are field samples that are spiked in the
laboratory with a known concentration of a target
analyte(s) in order to determine percent recoveries in
sample  extraction.   The percent  recovery  from
MS/MSDs  indicates the degree  to  which matrix
interferences  will affect  the identification  of  a
substance.  MS/MSDs  can also be used to monitor
laboratory performance.  When four or more pairs of
MS/MSDs are analyzed,  the  data  obtained may be
used  to evaluate error due  to laboratory bias and
precision.  Analyze one MS/MSD pair to assess bias
for every 20 samples, and use the average percent
recovery for the pair.  To assess precision, analyze at
least  eight  matrix spike  replicates from the  same
sample, and determine the standard deviation and the
coefficient of variation.  MS/MSDs are recommended
for screening data and are required as one of several
methods for determining  analytical error for definitive
                                                  40

-------
data.  Since the MS/MSDs are spiked field samples,
provide sufficient volume for three separate analyses
(triple volume).  When selecting a well for spiked
samples, choose a well capable of providing steady
volume.

5.4.8  Laboratory Duplicate Samples

A laboratory duplicate is a sample that undergoes
preparation and analysis twice.  The laboratory takes
two  aliquots of one sample and analyses them as
separate samples. Comparison of data from the two
analyses   provides   a  measure   of   analytical
reproducibility within a sample set. Discrepancies in
duplicate analyses may indicate poor homogenization
in the field or other sample preparation error, either in
the field or in the laboratory.
5.5    EVALUATION OF
        ANALYTICAL ERROR

The acceptable level of error in  sampling data is
determined by the intended use of the data and the
sampling objectives, including the degree of threat to
public health, welfare, or the environment; response
action  levels; litigation  concerns;  and  budgetary
constraints.

Error may be determined with replicate samples.  To
evaluate the total error of samples with contaminant
concentrations near the response action level, prepare
and analyze a minimum of eight replicates of the same
sample. Analytical data from replicate samples also
serve as a quick check  on errors associated with
sample  heterogeneity, sampling methodology, and
analytical procedures.  Different  analytical results
from two or more replicate samples could indicate
improper  sample  preparation, or  improper sample
handling, shipment, or analysis.

Although a  quantified  confidence level may be
desirable, it may  not always be possible.  A 95%
confidence level (5 percent acceptable error) should be
adequate for most Superfund activities. Note that the
use of confidence levels is based on the assumption
that a sample is homogeneous.
5.6    CORRELATION BETWEEN
        FIELD SCREENING RESULTS
        AND DEFINITIVE
        LABORATORY RESULTS

One cost-effective approach for delineating the extent
of site contamination is to correlate inexpensive field
screening  data  and other field  measurements with
definitive laboratory results. The relationship between
the two methods can then be described by a regression
analysis.  The resulting equation  can be used  to
predict laboratory  results based on field  screening
measurements.   In this manner, cost-effective field
screening results may be used in conjunction with off-
site laboratory analysis.

Statistical regression involves developing an equation
that relates two or more variables  at an acceptable
level of correlation.  In this case, the two variables are
field  screening results and  definitive laboratory
results.   The  regression equation  can be used  to
predict a laboratory value based  on  the results of the
screening  device.  The model can also be used  to
place   confidence   limits   around   predictions.
Additional discussion of correlation and regression
can be found in most introductory statistics textbooks.
A simple linear regression equation can be developed
on many calculators or computer databases. Consult
a statistician to check the accuracy of more complex
models.

Evaluation of the accuracy of a model relies in part on
statistical  correlation,  which involves computing an
index called  the  correlation  coefficient (r)  that
indicates the degree  and nature of the relationship
between two or more sets of values.  The correlation
coefficient ranges  from -1.0  (a perfect inverse  or
negative relationship), through 0 (no relationship),  to
+1.0 (a perfect, or positive, relationship). The square
of the correlation coefficient, called the coefficient  of
determination, or simply R2, is an estimate of the
proportion of variance in the dependent variable.  The
value of an acceptable coefficient of variation depends
on the sampling objectives and intended data uses. As
a rule of thumb, statistical relationships should have
an R2 value of at least 0.6 to determine  a reliable
model. However, for health assessment purposes, the
acceptable R2 value may be more stringent (e.g., 0.8).
Analytical calibration regressions have an R2 value  of
0.98 or greater.
                                                  41

-------
Field screening data can be used to predict laboratory
results if there is an acceptable correlation between
them. The predicted values can be located on a base
map and contoured.  These maps can be examined to
evaluate the estimated extent of contamination and the
adequacy of the sampling program.
5.7    EXAMPLE SITE

5.7.1  Data Categories

Screening  data  which  generate  non-definitive,
unconfirmed results were used to select  analytical
parameters and samples to be sent for laboratory
confirmation  analysis.   Samples were sent  to the
analytical laboratory under protocols which provided
definitive  data.  The rigorous  laboratory analyses
provided definitive identification and quantitation of
contaminants.

5.7.2  Sources of Error

All direct reading instruments were maintained and
calibrated  in  accordance  with  their instruction
manuals. Many of these instruments are class-specific
(e.g., volatile organic vapors)  with relative response
rates  that  are  dependent on the calibration gas
selected.   Instrument response to  ambient vapor
concentrations may differ by an order of magnitude
from response to calibration standards.  If compounds
of interest are known, site-specific standards may be
prepared.

The number and location of initial field samples were
based on observation and professional judgment (as
outlined in Section 2.5.5).  Field standard operating
procedures, documented in the site  sampling plan,
established  consistent  screening  and  sampling
procedures among  all sampling personnel, reducing
the chances for variability and  error during  sampling.
Site briefings were conducted prior to all sampling
and screening events to review the use  of  proper
screening and sampling techniques.
Other  steps taken to limit error included proper
sample  preparation,  adherence  to  sample holding
times,   and the  use of  proper  IATA  shipment
procedures.  All off-site  laboratory sample analyses
were  performed  using  approved  EPA  standard
methods and protocols.

5.7.3  Field QA/QC Samples

Field QA/QC samples were collected during soil and
ground-water sampling at the ABC Plating site.  Two
field replicate samples were collected for subsurface
soils; two wells (one overburden and one bedrock)
were selected for replicate collection and analysis of
ground  water.   Rinsate blanks were collected from
split spoon samplers, a bailer, and the submersible
rotary   pump  after  decontamination  by pouring
deionized water  through  the respective piece of
equipment and then into a sample container. The field
replicates and blanks were preserved and prepared as
"regular"  field  samples.  A trip blank  for VOC
analysis and a performance evaluation (PE) sample for
metals were sent to the laboratory. (The PE sample is
not affected by matrix interferences.) The trip blank
was provided  by the  laboratory   (pre-filled  and
preserved) and sent with the sample containers prior
to sample collection. One trip blank per day  was
submitted to the laboratory. Additional volume was
collected and provided to the laboratory for matrix
spike/matrix spike duplicate analyses for one per 20
sample locations for each medium.

5.7.4  Laboratory QA/QC

Instructions on matrices, target compounds,  and
QA/QC  criteria of particular interest were provided to
the laboratory to  help ensure that analytical results
met the required  quality assurance objectives.  The
laboratory analyzed for metals using the methods of
inductively coupled plasma (ICP) spectrometry and
atomic absorption (AA). Two SW-846 methods were
employed for hexavalent chromium analysis: Method
7196, a colorimetric method, and Method 2185, a
chelation method. These two methods were utilized
in an attempt to better quantify hexavalent results.
The presence of cyanide was confirmed in  the
laboratory using total  and amenable cyanide analyses
(colorimetric manual Method 9010).
                                                  42

-------
                    6.0  DA TA PRESENTA TION AND ANAL YSIS
6.1     INTRODUCTION

Data  presentation  and  analysis  techniques  are
performed  with  analytical,  field  screening,  or
geophysical results. The techniques discussed below
can be used to compare analytical values, to evaluate
numerical  distribution  of data, and to reveal the
location of  "hot spots," contaminant plumes, and the
extent of contamination at a site.  The appropriate
methods to present and analyze sample data depend on
the sampling  objectives, the number of samples
collected, the sampling approaches used,  and other
considerations.
6.2    DATA POSTING

Data posting involves placement of sample values on
a site base map or cross-section.  Data posting is
useful for displaying the distribution of sample values,
visually depicting the location of contamination with
associated assessment data.  Data posting  requires
each sample to have a specific location (e.g., x, y, and
sometimes  z  coordinates).   Ideally,  the  sample
coordinates are surveyed values facilitating placement
on a scaled map.  Data posting is useful for depicting
concentration values of  ground-water  and plume
migration.
6.3    CROSS-SECTION/FENCE
        DIAGRAMS

Cross-section diagrams (two-dimensional) and fence
diagrams  (three-dimensional)  depict   subsurface
features such as stratigraphic boundaries, aquifers,
plumes, impermeable layers, etc. Two-dimensional
cross-sections may be used to illustrate  vertical
profiles of ground-water  concentrations  on a  site.
Both cross-sections and fence diagrams can provide
useful   visual   interpretations  of   contaminant
concentrations and migration.
6.4    CONTOUR  MAPPING

Contour maps are useful for depicting ground-water
contaminant concentration values throughout a site.
Contour  mapping requires  an  accurate,  to-scale
basemap of the site. After data posting sample values
on the basemap, insert contour lines (or isopleths) at
a specified  contour  interval, interpolating values
between sample points. Contour lines can be drawn
manually or can be  generated by  computer using
contouring software.  Although the software makes
the  contouring process easier, computer  programs
have a limitation:  as they interpolate between data
points, they attempt to " smooth" the values by fitting
contour intervals to the full range of data values.  This
can result in a contour map that does not accurately
represent general site contaminant trends.  If there is
a big difference in concentration between  a "hot spot"
and the surrounding  area,  the computer contouring
program, using  a contour  interval  that  attempts to
smooth the "hot spots," may eliminate  most of the
subtle site features and general trends.
6.5    WELL LOCATION  MAP

A  well location  map should be prepared  using
surveyed data for all features at the site. This map
serves as a basemap onto which other data may be
plotted  (e.g., data  posting,  contaminant  plume
contours, water elevation  contours).   The map is
drawn to scale  and incorporates all wells located,
installed, and sampled, including residential  and
monitoring wells. The surveyed coordinates for each
monitoring well location could also be posted onto the
map (in feet above mean sea level (msl)) to illustrate
topography and surface gradient.
6.6    STATISTICAL GRAPHICS

The distribution or spread of the data set is important
in determining which statistical techniques to use.
Common statistical analyses, such as the t-test, rely on
normally  distributed  data.    The histogram  is  a
statistical bar graph which displays the distribution of
a data set. A normally distributed data set takes the
shape of a bell curve, with the mean and median close
together  about halfway between the maximum  and
minimum values.    A  probability  plot  depicts
cumulative percent against the concentration of the
contaminant of concern.  A normally distributed data
set, when plotted as a probability plot, appears as a
straight line.  A histogram or probability plot can be
used to see trends and anomalies in the data from a
ground-water site prior to conducting more rigorous
forms of statistical analysis.
                                                  43

-------
6.7    RECOMMENDED DATA
        INTERPRETATION METHODS

The data interpretation methods chosen depend on
project-specific considerations, such as the number of
sampling  locations and their associated  range in
values.  Data which are dissimilar in composition
should not be compared using statistical interpretation
methods. Data posting, screening, and sampling data
sheets,  and cross-section/fence diagrams  may  be
appropriate.  A site feature showing extremely low
data values (e.g.,  non-detects), with  significantly
higher values (e.g., 5,000 ppm) from neighboring "hot
spots"  and little or no concentration  gradient in
between, does not lend itself to contouring software.
6.8    EXAMPLE SITE

A water table contour map was generated with the
water  level  data  for  the  shallow  overburden
monitoring wells.  This  indicated a westward flow
direction, which generally coincides with the surface
topography. The deep bedrock wells lie nearly on a
straight line, and therefore a confident determination
of flow direction was not possible.   A  westward
component of flow direction is evident in the data,
however.   The bedrock  contact  wells  provided
inconsistent water level data, most likely due to the
presence of discontinuous perched water zones at the
well locations.

All ground-water samples were  analyzed for total
chromium and cyanide. Cyanide was not found in any
of the samples above the 50 (ig/1 detection  limit.
Using a detection limit of 50 (ig/1 for chromium, three
filtered samples were found to be contaminated at two
locations  (3OB,  and  6OB/6AW).   Five  of the
unfiltered ground-water samples (Wells 2SA, SOB,
4SA, 6OB, and 6AW) exceeded the detection limit.
These data were posted on a site/well location map to
illustrate well proximities, as well as a map indicating
the contours of contamination.

The  rate  of  chromium contaminant migration in
ground water and the potential long-term impact to
nearby residential wells were estimated  using  a
mathematical model  which  included  worst case
assumptions   and    evaluated   attenuation   of
contaminants through soil and ground water.  The
OSC concluded that the potential for residential well
contamination was minimal.  Removal of  soil, the
source of contamination, was recommended.  This
decision met  the Phase 2 objective of establishing
early action options and consideration of long-term
remediation requirements for ground water.

All containers of wastes were removed from the site.
Soil  treatment/disposal  was  completed using the
existing  grid  design.   Cells were  sampled  and
designated as clean or excavated. Excavated material
was stockpiled while treatment/disposal options were
evaluated.  Excavated cells were filled with stone and
clean soil.  Composite sampling in each cell verified
cleanup, using an action level of 100 mg/kg chromium
in the soil composite.   (The clean-up level was
established based on the earlier mathematical model
and soil attenuation calculations.) The soil response
served as an early action to meet the Phase 3 objective
originally established for the site.
                                                  44

-------
    APPENDIX A -- Example of Flow Diagram For Conceptual Site Model
                        Figure A-1
Migration Routes of a  Gas Contaminant
           from  Origin to Receptor
Change of
Original state Pathway contaminant
of contaminant from state In
of concern" origin pathway
conch
RpQ > A:r
V-^CIO r Mil
solldl
insatlon
> Liquid
_ **
— > Solid
Mcatlon




Final
pathway
to receptor
> SO
^ sw
> so
> AT
>• ^V J.
> sw
^ so
^ sw
Receptor
Human
G,D
G,D
I,D
I,D
G,D
G,D
G,D
Ecological Threat
Terrestrial
G,D
G,D
I,D
I,D
I,D
G,D
G,D
Aquatic
N/A
G,D
N/A
N/A
G,D
N/A
G,D
     *  May be a transformation product
     ** Includes vapors
Receptor Key

D  - Dermal Contact
]  - Inhalation
G  — Ingestlon
N/A - Not Applicable
Pathway Key

Al -Air
SO - Soil
SW = Surface Water
(Including sediments)
GW - Ground Water
                           45

-------
                              Figure A-2
       Migration  Routes of a Liquid  Contaminant
                    from  Origin to Receptor
 Original state
 of contaminant
  of concern*
Liquid
               sw
*  Liquid

->   Gas**
               SO
                  solidification
                    leachate,
                    Infiltration
               AI
  *  May be a transformation product
  ** Includes vapors
                             Solid
   Liquid
     Gas
                                  **
->   so
->   sw

->   GW

->   SO
->   AI

+   sw
                Receptor Key

               D — Dermal Contact
               I - Inhalation
               G • Ingestlon
               N/A - Not Applicable
Receptor
Human
G,D
I,D
G,D
G,D
Ecological Threat
Terrestrial
G,D
I,D
G,D
G,D
Aquatic
G,D
N/A
G,A
G,D
G,D
G,D
G,D
G,D
G,D
N/A
N/A
G,D
N/A
G,D
I,D
G,D
G,D
I,D
G,D
N/A
N/A
G,D
                    Pathway Key

                 AI - Air
                 SO - Soil
                 SW - Surface Water
                    (Including sediments)
                 GW - Ground Water
                                 46

-------
                         Figure A-3
      Migration  Routes of a Solid Contaminant
                from Origin to Receptor
 Original state
 of contaminant
 of concern*
Solid
            AI
               partlculates/
                 dust
            Solid
sw
^  Solid

->  Liquid
            SO
                         Gas
                            **
            Solid
                        Liquid
 * May be a transformation product
 ** Includes vapors
AI
SW
so

sw

sw
Receptor Key
D - Dermal Contact
I - Inhalation
G - Ingestlon
N/A - Not Applicable

Pathway Key
AI . Air
SO - Soil
SW - Surface Water
(Including sediments)
GW - Ground Water
Receptor
Human
I,D
G,D
G,D
Ecological Threat
Terrestrial
I,D
G,D
G,D
Aquatic
N/A
G,D
N/A
G,D
G,D
G,D
G,D
G,D
G,D
so
AI
SW
SO
SO
sw
G3D
I,D
G,D
G,D
G,D
G,D
G,D
G,D
I,D
G,D
G,D
G,D
N/A
G,D
N/A
N/A
G,D
N/A
N/A
N/A
G,D
                            47

-------
                                  References
Aller, Bennett, Hackett, et al.  1989. Handbook of Suggested Practices for the Design ad
   Installation of Ground-Water Monitoring Wells National Water Well Association, Dublin,
   Ohio.

Code of Federal Regulations (CFR), Title 40 -Protection of Environment. Part 136 - Guidelines
   Establishing Test Procedures for the Analysis of Pollutants.

Custodio, Gurgui, Ferreira.  1987. Ground-Water Flow and Quality Modelling D. Reidel
   Publishing Company, Dordrecht, Holland.

Driscoll, Fletcher G. 1986. Ground Water and Wells(2nd edition).

Fetter, C.W., Jr.   1980.  Applied Hydrogeology. Charles E. Merrill Publishing Company,
   Columbus, Ohio.

Garrett, Peter.  1988.  How  to Sample Ground Water and Soils  National  Water Well
   Association, Dublin, Ohio.

Keith, Lawrence, H. 1988. Principles of Environmental Sampling

New Jersey Department of Environmental Protection.  February 1988. Field Sampling
   Procedures Manual

Palmer, Christopher, M. 1992. Principles of Contaminant Hydrogeology

Proceedings of the Fifth National Symposium and Exposition on Aquifer Restoration and
   Ground-Water Monitoring. 1985.

U.S. Department  of Health  and Human Services.   1985.  Occupational Safety and  Healh
   Guidance Manual for Hazardous Waste Site ActivitiesNIOSH Pub. 85-115.

U.S. Environmental Protection Agency.  1995a.  "Establishing Background Levels."  Office of
   Solid Waste and Emergency Response Directive 9285.7-19FS.

U.S. Environmental Protection Agency.  1995b.  SuperfundProgram Representative Sampling
   Guidance, Volume 1 — Soil.  Office of Solid Waste and Emergency Response Directive
   9360.4-10.

U.S. Environmental Protection Agency.  1995c.  Superfund Program Representative Sampling
   Guidance, Volume 4— Waste.  Office of Solid Waste and Emergency Response Directive
   9360.4-14.
                                        48

-------
U.S. Environmental Protection Agency.  September 1993. Data Quality Objectives Process for
   Superfund.  Office of Solid Waste and Emergency Response Directive 9355.9-01.

U.S. Environmental Protection Agency.  1993. Project Summary- Compilation of Ground-Water
   Models. Office of Research and Development. EPA/600/SR-93/118.

U.S. Environmental Protection Agency. November 1992. RCRA Ground-Water Monitoring:
   Draft Technical Guidance EPA/530-R-93-001.

U.S. Environmental  Protection  Agency.  September 1992.  Guidance for Performing Sit
   Inspections Under CERCLA Office of Solid Waste and Emergency Response Directive
   9345.1-05

U.S. Environmental Protection Agency. February 1992. Estimating the Potential for Occurrence
   of DNAPL at Superfund Sites  Office of Solid Waste and Emergency Response Directive
   9355.4-07.

U.S. Environmental Protection Agency. January 1992. Statistical Training Course for Ground-
   Water Monitoring Data Analysis EPA/530-R-93-003.

U.S. Environmental Protection Agency.  1992. Ground-Water Issue- Fundamentals of Ground-
   Water Modeling.  Office of Research  and Development and Office of Solid Waste and
   Emergency Response.  EPA/540/S-92/005.

U.S. Environmental Protection Agency. January 1991a. Compendium of ERT Field Analytical
   Procedures.  Office of Solid Waste and Emergency Response Directive 9360.4-04.

U.S. Environmental Protection Agency. January 1991b. Compendium of ERT Ground-Water
   Sampling Procedures Office of Solid Waste and Emergency Response Directive 9360.4-06.

U.S. Environmental Protection Agency. January 1991c. Compendium of ERTSurface Water and
   Sediment Sampling Procedures. Office of Solid Waste and Emergency Response Directive
   9360.4-03.

U.S. Environmental  Protection  Agency.  1990. Ground Water,  Volume  II:  Methodology
   EP A/625/6-90/016b.

U.S. Environmental  Protection  Agency. April 1989.  Statistical Analysis of Ground-Wate
   Monitoring Data at RCRA Facilities, Interim Final Guidance Office of Solid Waste and
   Emergency Response.

U.S. Environmental Protection Agency. September 1988. Field Screening Methods Catalog —
   User's Guide. Office of Solid Waste and Emergency Response. EPA/540/2-88/005.

U.S. Environmental Protection Agency. December 1987. A Compendium of Superfund Fiei.
   Operations Methods  Office of Solid Waste and Emergency Response Directive 9355.0-14.

                                        49

-------
U.S. Environmental Protection Agency. 1987. Handbook, Ground Water. EPA/625/6-87/016.

U.S. Environmental Protection Agency.  September 1986.  RCRA Ground-Water Monitoriig
   Technical Enforcement Guidance Documen(TEGD). Office of Solid Waste and Emergency
   Response Directive 9950.1.

U.S. Geological Survey.  1991. Concepts and Modeling in Ground-Water Hydrology - A Self-
   Paced Training Course  Open File Report 90-707.
                                       50

-------