4>EPA
United States
Environmental Protection
Agency
d  Water Issue
    An Approach for Developing Site-Specific Lateral and Vertical Inclusion
     Zones within  which Structures Should be Evaluated for Petroleum Vapor
    Intrusion due to Releases of Motor Fuel from Underground Storage Tanks
                         John T. Wilson1, James W. Weaver2, Hal White3
 Abstract
 1.0 Introduction
 2.0 The Lateral Inclusion Zone
  2.1 Process to Define the Lateral Inclusion Zone
  2.2 Dissolved Contaminant Plumes in the Lateral Inclusion Zone
  2.3 Steps to Apply a Lateral Inclusion Zone
    2.3.1 Map and Estimate the Extent of Contamination
    2.3.2 Define an Inclusion Zone
      2.3.2.1 A Definition that Does Not Consider Ground water
       Flow
      2.3.2.2 A Definition that Considers Ground water Flow
       2.3.2.2.1 Find the Average Direction of Ground water
       Flow
       2.3.2.2.2 Assign a Weight to the Extent of the Inclusion
       Zone for the Direction of Ground water Flow
    2.3.3 Determine if Additional Monitoring Points Would
       Reduce the Extent of the Refined Inclusion Zone
    2.3.4 Test the Inclusion Zone against Simple Transport
       Calculations
 3.0 The Vertical Inclusion Zone
  3.1 Steps to Apply a Vertical Separation Distance to Core Samples
    3.1.1 Acquire Core Samples for Screening
    3.1.2 Screen Core Samples for Subsequent Laboratory
       Analysis
    3.1.3 Compare the Distribution of Contamination in
       Sediment to the Vertical Separation Criteria
  3.2 Steps to Apply a Vertical Separation Distance to Core Samples
 4.0 Next Steps
 5.0 Summary
 Notice
 6.0 References
 Appendix A. Recommendations for Sampling and Analysis
 Appendix B. Quality Assurance
 Appendix C. Equations for Steady State Plume Calculations


 'U.S. EPAORD, wilson.johnt@epa.gov
 2U.S. EPA ORD, weaver.jim@epa.gov
 3 U.S. EPA OUST, white.hal@epa.gov
                          ABSTRACT
                          Buildings may be at risk from Petroleum Vapor Intrusion (PVI)
                          when they overlie petroleum hydrocarbon contamination in the
                          unsaturated zone or dissolved contamination in ground water.
                          The U.S. EPA Office of Underground Storage Tanks (OUST) is
                          preparing Guidance for Addressing Petroleum Vapor Intrusion
                          at Leaking Underground Storage Tank Sites. The OUST
                          guidance provides general screening criteria that can  be used
                          to identify structures that are at risk from PVI.  The criteria are
                          used to determine if a structure is included within a lateral or
                          vertical zone where proximity to the contaminant might make the
                          building vulnerable to  PVI.  If the structure is within  a lateral or
                          vertical inclusion zone, then additional investigation is necessary
                          to evaluate and manage exposure to the vapors.
                          This Issue Paper contains technical suggestions and
                          recommendations proposed by the U.S. EPA Office of Research
                          and Development for applying the criteria provided  in the OUST
                          guidance.  The Issue paper provides a graphical approach
                          to define a lateral inclusion zone based on the proximity of a
                          structure to the presumed maximum extent of contamination.
                          The presumed maximum extent of contamination is defined by a
                          perimeter of clean monitoring locations that are arranged around
                          the known source of contamination.  The lateral inclusion zone is
                          extended past the presumed maximum extent of contamination
                          to allow for uncertainty of the concentrations of contaminants
                          in the space between  monitoring locations.  The Issue Paper
                          provides instructions and suggestions to use knowledge of
                          ground water flow to refine the lateral exclusion zone, and
                          reduce the area where additional investigation is necessary.
                          The Issue Paper provides recommendations on collecting and
                          analyzing core samples to determine the vertical extent of
                          contamination in the unsaturated zone, and water samples to
                          determine the extent of contamination in ground water. The
                          Issue Paper provides  illustrations of the appropriate comparison
                          of the field data to the criteria in the OUST Guidance.   In
                          combination, definition of lateral and vertical inclusion zones
                          makes the best use of site characterization data for assessing
                          the risk of PVI to structures  at a LUST site.  The procedures

-------
outlined in this Issue Paper provide a realistic data-
driven approach to screen buildings for vulnerability
to PVI.

1.0 INTRODUCTION
Vapor intrusion is a process whereby vapors of
hazardous substances move through unsaturated
soil and enter buildings.  Occupants of the buildings
are exposed to the hazardous substances as
vapors in indoor air. The vapors may originate
from contaminated ground water or from light non-
aqueous phase liquids (LNAPLs).  Underground
storage tanks (USTs) are regulated under Subtitle
I of the Solid Waste Disposal Act. Most USTs are
used to store motor fuel (e.g., gasoline, diesel
fuel) that is composed primarily of petroleum
hydrocarbons (PHCs). Releases of motor fuel from
a leaking UST may result in generation of PHC
vapors and can result in  petroleum vapor intrusion
(PVI).
The U.S. EPA is developing Guidance for
Addressing Petroleum Vapor Intrusion at Leaking
Underground Storage Tank Sites (U.S. EPA,
2013a). The guidance provides general screening
criteria that can be used  to identify structures that
are at risk from PVI. In general, structures are at
risk from PVI when they overlie masses of residual
LNAPL in the unsaturated zone, accumulations
of liquid LNAPLs at the water table, or petroleum
contamination  dissolved  in ground water at levels
that have the potential to pose a risk to receptors
through the vapor intrusion pathway.
The potential for human exposure from PVI may
be limited because of the biodegradability of
PHCs.  The PVI Guidance provides recommended
screening levels for petroleum constituents above
which the potential for PVI should be considered. If
the available data on the distribution of petroleum
components in soil and ground water suggest
a reasonable possibility that PVI may impact
a structure, that structure is considered to be
contained within an inclusion zone, which implies
that additional  investigation is necessary to evaluate
and manage exposure to the vapors.
As discussed in detail later in this document, the
inclusion zone considers both lateral and vertical
proximity to the vapor source (i.e., mobile LNAPL,
residual LNAPL, and dissolved contamination). All
structures in the immediate vicinity of the source
             area are first evaluated to determine if they are
             within the lateral inclusion zone.  This approach
             logically follows the typical site investigation as it
             progresses over time from the source area outward
             in the  direction of ground water flow to the edges
             of the  dissolved plume. As more site-specific
             information is compiled, the extent of the inclusion
             zone may change.  If any structure is within the
             lateral inclusion zone, then it is further evaluated to
             determine  if it is in the vertical inclusion zone.
             The lateral inclusion zone is discussed in  Section 2.
             The vertical inclusion zone is  discussed in
             Section 3.  As described  and illustrated in these
             sections, it may be necessary to acquire additional
             site characterization data before this approach can
             be used with confidence  to screen structures and
             determine whether they are within the  inclusion
             zone for PVI.
             Both lateral and vertical inclusion zones should be
             delineated using site-specific  data. A conceptual
             site model (CSM) that integrates all available data
             and information about a particular site  should be
             developed and continually refined as new data
             become available. Especially  near the beginning
             of an investigation at a leaking UST site, there is
             typically much uncertainty due to the lack of site-
             specific data and information. To compensate
             for uncertainty due to lack of data, the  screening
             criteria produce a larger inclusion zone. As more
             data are integrated into the CSM, the degree
             of uncertainty progressively diminishes. Thus,
             the extent of the lateral inclusion zone can often
             be reduced. However, improved understanding
             necessarily takes time and resources.
             If inhabited buildings or sites for future buildings
             are not located within one or the other of these
             inclusion zones, the vapor intrusion pathway may
             be considered to be incomplete and no further
             consideration of the pathway should be necessary
             for these buildings.  This assumes that there are  no
             preferential pathways for contaminant  migration at
             the site.  This also assumes that conditions at the
             site do not change. Factors to consider in deciding
             whether to exclude sites  from further evaluation
             of PVI may include future land use, construction
             of utility trenches through or near previous
             contamination, increased ground water usage that
             might  change the direction of ground water flow,
             and additional releases of contaminants.
 Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 This Issue Paper contains technical suggestions
 and recommendations proposed by the U.S. EPA
 Office of Research and Development for applying
 the criteria provided in OUST's PVI Guidance (U.S.
 EPA, 2013a).  The material in this Issue Paper is not
 guidance from the U.S. EPA Office of Underground
 Storage Tanks (OUST).
 The federal UST program delegates authority to
 implement an UST program to the states. Most
 of the state agencies use a risk  based approach
 to manage vapor intrusion of PHCs and other fuel
 constituents (U.S. EPA 1995, 2002). The staff of the
 state agencies or the Indian nations that implement
 the UST program may choose to implement another
 approach to apply screening criteria recommended
 by U.S. EPA (2013a).  If they choose to implement
 this approach, they may modify  this approach to
 make it more appropriate to their particular needs.

 2.0 THE LATERAL INCLUSION ZONE
 This section discusses methods to determine
 whether proximity of a structure to a source of
 contamination puts the structure at risk for PVI.
 Contamination can be  mobile LNAPL, residual
 LNAPL, or a dissolved plume. It is important
 to define a lateral inclusion zone based on the
 separation distance between the structure and
 monitoring locations that are known to be clean
 instead of the distance from known contamination.
 This is especially critical if the extent of subsurface
 contamination is not well-defined, as there is no way
 to know how far the contaminated material actually
 extends from the source of contamination toward
 the receptor.
 Typically at the beginning of a leaking UST
 investigation the full extent and  location of
 contamination and the direction  of ground water
 flow are not well-defined. An illustration of these
 uncertainties is presented in Figure 1. Here a
 leaking UST has impacted the five  monitoring
 wells initially installed  to assess the extent of
 contamination. Because all of the wells are
 contaminated, the actual extent of contamination
 cannot be determined. Because sufficient ground
 water monitoring data have not yet been collected,
 the direction of ground water flow has not been
 determined. Given the uncertainty in the direction
 of ground water flow, a contaminant plume could
conceivably migrate away from the source in a
variety of directions as shown.
The procedure described to define the lateral
inclusion zone is based on the assumption that
the closer together the monitoring points, the less
uncertainty there is about the extent and location of
contamination. Conversely, with fewer monitoring
points spaced farther apart the uncertainty is
greater. As  monitoring points are placed closer
together and additional  monitoring points are
installed to  fill in the gaps in the monitoring network,
the extent of contamination is determined more
accurately.  This concept is depicted schematically
in Figure 2, which shows a simplified relationship
between the location of clean monitoring points
and the extent of contamination. In this example,
contamination extends from leaking USTs in the
direction of a potential receptor, which has been
established by determining the ground water flow
direction. The extent of contamination is bounded
laterally by  two clean monitoring points, but no well
is available to provide a boundary to the plume
in the direction of ground water flow.  In Figure
2(a), contamination extends between two clean
monitoring  points for an unknown distance and
may, therefore, impact a down-gradient dwelling.
This scenario may occur even if the clean wells
are closer together, as shown in Figure 2(b).  In
Figure  2(c), an additional monitoring location has
been installed and determined to be clean, which
eliminates the illustrated building from consideration
for additional PVI investigation assuming that there
are no  preferential transport pathways present
that could lead to PVI. This example illustrates
that ground water flow directions and monitoring
well locations should be carefully considered when
defining the lateral inclusion zone. Section 2.3.2.2
provides a methodology to account for ground water
flow direction and locating monitoring wells.
Extending the inclusion  zone by a distance equal
to the distance between monitoring wells is an
arbitrary choice.  This ratio is recommended as a
starting point.  If a caseworker has local knowledge
that justifies either a greater or lesser ratio, that
local knowledge should be applied and the ratio
adjusted accordingly. The ratio should be based on
local regulatory policy and the distribution of existing
and potential receptors  around the release site.
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue  3

-------
                           (a) A plume produced by a
                           constant ground water flow
                           direction in uniform porous
                           media
                            A plausible extent
                            of contamination
II
o
^

^
Leaking
USTs
Contaminated
Monitoring Point
. Potential
Receptor
Subsurface
Contamination
Direction
k Ground Water
Flow
                                                                                       An equally
                                                                                     plausible extent
                                                                                    of contamination
                        (b) A plume produced under the same
                        conditions as in (a) above, except the
                        direction of ground water flow differs by
                        90 degrees.
                                                              (c) Potential extent of
                                                              contamination due to variable
                                                              ground water flow conditions.
Figure 1.   Examples of Plausible Extent of Contamination for Hypothetical Petroleum Release
 Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
                          MP-1
        MP-3
                                                   -i      (b)
                                                   -2
                  Lea king USTs

           Q    Monitoring Point

          ^^  }  Zone of Potential Contamination

                  Potential Receptor


                  Direction Ground Water Flow
MP-3
                  MP-2
 Figure 2.    Effect of the distance between clean monitoring points on the extent of the plausible zone of potential
             contamination.
 The ratio of one-to-one in the judgment of
 the authors is a good point of departure for
 unconsolidated media.  In fractured consolidated
 media, particularly if the hydraulic gradient is
 aligned with fracture orientation, a larger ratio would
 be appropriate.
 Strictly speaking, no matter how close together
 they are, the contaminant concentration between
 two monitoring points is never known with absolute
 certainty; it can only be extrapolated. Because there
 is a practical limit to the number of monitoring points
 that can be installed, there will always be some
 degree of uncertainty. The techniques described in
 this Issue Paper recognize the uncertainty inherent
 in the site investigation process and represent
 one approach for balancing between being
 overly protective and not sufficiently protective.
 Site-specific data regarding the actual extent of
 contamination and its potential for migration are
 necessary for defining the lateral inclusion zones.

 2. ] Process to Define the Lateral Inclusion Zone
 Figure 3 illustrates the process of defining the
 inclusion zone. In this example, a first round
 of sampling showed that the LIST resulted in
 contamination of all five wells surrounding the
 leaking LIST (red circles, e.g.,  representing borehole
 locations). New monitoring locations were installed
 to establish the extent of contamination (blue
 circles). Soil samples  and ground water samples
 from the new location were found to be clean. In
 this case, the maximum extent of contamination
 may be presumed to be defined  by the smoothed
 shape bounding the clean monitoring points
 (Figure 3(a)). EPA recommends  that dwellings (e.g.,
 House A) within the area of presumed maximum
 extent of contamination are to be evaluated  for
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                                 round Water Issue 5

-------
                                                (a) The presumed maximum
                                                extent of contamination is the
                                                area bounded by clean
                                                monitoring points: all
                                                contaminated monitoring
                                                points should be contained
                                                within this area.
                                       Presumed maximum extent of
                                       known contamination
                                               II     Leaking
                                                       USTs
                                               O       Clean
                                                   Monitoring Point

                                               p   Contaminated
                                                   Monitoring Point

                                                      Potential
                                                      Receptor
                                                       Lateral
                                                    Inclusion zone
                         (b) Draw straight lines around
                         the perimeter of contamination.
                         These lines connect each
                         adjacent clean monitoring point.
                         Then draw perpendicular lines
                         (e.g., x', and y'), that are the
                         same length as the lines
                         connecting adjacent wells.
                                                                (c) To identify the lateral
                                                                inclusion zone, draw a line
                                                                connecting the endpoints of
                                                                the perpendicular lines and
                                                                the clean  monitoring points,
           Lateral
           Inclusion Zone
D
Figure 3.    Determination of lateral inclusion distance based on separation distance between clean monitoring
             points
 Ground We
  for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 potential PVI impacts. While Houses B, C, and
 D are outside the presumed maximum extent of
 contamination, there is uncertainty about the extent
 of contamination between monitoring locations,
 particularly where monitoring points are separated
 by a large distance.
 The uncertainty in the presumed maximum
 extent of contamination may be accounted for by
 extending the inclusion zone beyond the presumed
 maximum extent of contamination (defined by
 the blue line in Figure 3(a)). This concept is
 illustrated in Figures 3(b) and 3(c). Any building
 within the inclusion zone defined in this manner is
 recommended for further evaluation. If any portion
 of a structure falls within the inclusion zone,  all of
 the structure is considered to be within the inclusion
 zone. With this concept of inclusion zone, Houses B
 and C, in addition to House A, are recommended to
 be investigated for potential vapor intrusion impacts.
 This example illustrates that more closely spaced
 monitoring locations allow for greater certainty in
 defining the areas likely to be impacted by vapor
 intrusion and, generally, will reduce the areal extent
 of the inclusion zone. This example also illustrates
 that it is important to carefully consider the
 placement of monitoring points relative to receptors,
 so that portions of a  building are not unnecessarily
 included in the inclusion zone.
 The lateral inclusion zone is defined by bounding
 the plume with clean monitoring points.  However,
 defining the boundary of the plume is less important
 in those parts of the  site with no occupied buildings.
 To minimize expense, monitoring points should be
 located so they provide the most usable information
 for both the initial site characterization effort and
 any follow-up assessment of vapor intrusion. Be
 sure to place monitoring points between the source
 of contamination and any potentially impacted
 buildings. This approach is followed in the example
 presented below in Section 2.3. In the example, a
 new well is placed in front of buildings that might
 be down gradient of the source, but where the
 edge of the plume is not well defined.  In contrast,
 no additional work is suggested in areas that were
 upgradient of the source, or that did not have
 structures that would be vulnerable to PVI. For
 a new case the selection of the initial monitoring
 locations should be related to the locations of
buildings.  These locations can be chosen to
minimize the number of monitoring points installed.

2.2 Dissolved Contaminant Plumes in the Lateral
    Inclusion Zone
Contaminant plumes are dynamic features and
generally necessitate three-dimensional monitoring
to assess the transient behavior of ground
water flow and the transport of contaminants. In
unconsolidated deposits, the contaminant plume
should extend down gradient in the direction
of ground water flow. However, a variety of
hydrological phenomena can change the direction
of ground water flow, including aquifer recharge
following rainfall or snow melt, changes in the
pumping of ground water, and tides or changes
in the stage of a nearby river.  Heterogeneity of
geologic materials comprising the upper-most water
bearing zone  also may influence the direction of
migration and extent of contaminant plumes. Plume
behavior in heterogeneous materials may be quite
different from  that anticipated for homogeneous
materials.  In some cases plumes may be either
narrower or broader, or bifurcated with lobes moving
in different directions.
Changes in the direction of ground water flow are
common at leaking LIST sites (see Goode and
Konikow, 1990; Mace et al., 1997; Wlson, 2003;
Wlson et al. 2005a;  Wilson et al., 2005b).  Figure 4
illustrates variability of ground water flow directions
at two leaking LIST sites. In Figure 4(a), the flow
direction as indicated by the cluster of arrows varies
by more than  90 degrees. The fluctuation of ground
water flow directions in Figure 4(b) ranges over
nearly 180 degrees.  Determination of flow direction
may require periodic sampling over more than
one annual cycle to understand the ground water
flow regime at a given site. As the plume migrates,
appropriate adjustments to the sampling plan
should be  made to ensure that potential receptors
continue to be protected.
EPA recommends that ground water elevations be
measured when the  wells are sampled so that the
direction of ground water flow can be determined for
that particular sample round.
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue  "J

-------
                                               Sampling Point
                          Approximate Scale in Meters
                                                                      (a) An MTBE site in Elizabeth City,
                                                                      NC. The arrows represent the
                                                                      distance that water would move
                                                                      in one year, based on the
                                                                      direction and hydraulic gradient
                                                                      present in a particular round of
                                                                      sampling. The origin of the
                                                                      arrows is the center of the LNAPL
                                                                      source area. The black dots are
                                                                      locations of monitoring wells.
                                                                      The shaded area includes all the
                                                                      monitoring wells with
                                                                      concentrations of MTBE above 20
                                                                      ug/l. Reprinted from  Figure 3.1.
                                                                      of Wilson etal. (2005a).
                MW-12
               MW-11
            MW-10 •

              MW-6
        MW-9«
                             20 meters
• MW-7    «MW-13
    xJVlW-1
     "    Tank Pit
         MW-4
        MW-3
                                           0.00 0.01 0.02 0.03 0.04
                                             I    I I  I   I   I I
                                                  Hydraulic
                                                  Gradient
(b) A leaking UST site in New
Jersey. The arrows represent the
distance that water would move
in one year, based on the
direction and hydraulic gradient
present in a particular round of
sampling. The origin of the
arrows is the center of the LNAPL
source area (shaded gray).
(Wilson etal., 2005b)
Figure 4.    Relationship between the distribution of contamination in ground water and the variation in direction and
             magnitude of ground water flow
 Ground We
            for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
2.3 Steps to Apply a Lateral Inclusion Zone
There are four general steps in defining a lateral
inclusion zone:
 IMap and estimate the extent of contamination at
   the site with existing monitoring points.
 2. Define an inclusion zone. Consider ground water
   flow direction.
 3. Determine if additional monitoring points could
   be used to reduce the extent of the inclusion
   zone.
 4. If information is available, test the inclusion zone
   against simple transport calculations, and adjust
   the inclusion zone as required.
Srinivasan et al. (2004) used a  site in South
Carolina  as a case study to illustrate the
implementation of a software application that can be
used to identify the optimum locations of monitoring
wells. The Optimal Well Locator (OWL) is further
described in Section 2.3.2.2.1.  This Issue Paper will
use the same site as  a case study to define a lateral
inclusion zone for ground water contamination.
  The data used in the case study are provided
 as an illustration.  They do not necessarily
 reflect current conditions at the site, and have no
 bearing on past or current regulatory action taken
 by the South Carolina  Department of Health and
 Environmental Control.
 v	/

2.3.1 Map and Estimate the Extent of
    Contamination
The first step is to obtain a map showing the
distribution of contamination and the location of
potential  receptors at the site. Figure 5 in this
paper is a reproduction of Figure 5 originally
presented by Srinivasan et al. (2004). The source
of contamination is located in a commercial area
extending along an arterial highway.  On the other
side of the contaminated area are four residential
houses.  The contours on the map showing the
general distribution of contamination do not include
the houses that may potentially be impacted by
PVI; however, there are no clean wells between
the source of contamination and these potential
receptors.
Close examination of the contours shows that
the boundaries of the plume, even if based on an
                                                   interpolation scheme, are arbitrary; the location
                                                   of the 10 ug/L, 100 ug/L and 1,000 ug/L contours
                                                   are unsupported by data over most of their length.
                                                   There are no wells that bound the lateral extent of
                                                   contamination between the 10 ug/L contour and
                                                   the houses. The location of the toe of the plume
                                                   (i.e., the longitudinal extent of the plume) beneath
                                                   Circus Donuts is similarly unsupported by data by
                                                   any wells that define the longitudinal extent of the
                                                   plume.  The contours present a highly subjective
                                                   depiction of the extent of contamination, limiting it to
                                                   the commercial area without justification based on
                                                   the data. As a result, the available data for this site
                                                   does not support understanding of potential impacts
                                                   to the neighboring houses.

                                                   2.3.2 Define an Inclusion Zone
                                                   It is not necessary for the first definition of the
                                                   inclusion zone to consider the direction of ground
                                                   water flow. At recent petroleum release sites,
                                                   this information may not be available. The most
                                                   conservative assumption is that contamination
                                                   can move in any direction, and that movement in
                                                   any particular direction is equally plausible. This
                                                   approach to define the inclusion zone is described
                                                   in Section 2.3.2.1.
                                                   If data are available that can be used to infer the
                                                   direction and magnitude of ground water flow,
                                                   then information on ground water flow can be
                                                   used to refine the Inclusion Zone. Approaches to
                                                   accomplish this are described in Section 2.3.2.2. In
                                                   addition, it may be necessary to install additional
                                                   monitoring wells to adequately define the
                                                   lateral exclusion zone. Approaches for selecting
                                                   appropriate well locations are described in
                                                   Section 2.3.3.

                                                   2.3.2.1 A Definition That Does Not Consider
                                                      Ground water Flow
                                                   The general approach was illustrated schematically
                                                   in Figure 3.  Clean monitoring locations are used
                                                   to establish a boundary around the presumed
                                                   maximum extent of contamination.  Then segments
                                                   are drawn that extend the lateral inclusion zone past
                                                   the presumed maximum extent of contamination.
                                                   The extension of the inclusion zone compensates
                                                   for the uncertainty in the true limit of contamination
                                                   in the space between the monitoring points. The
                                                   approach is applied to the case study beginning
An Approach for Developing Site-Specific Lateral and Vertical Incli
                                                                                 Ground Water Issue  9

-------
                                                         DORCHESTER    ., Q Bnnl
                                                         PAWN & GOLD    I1 9.800J
                                                                               NOT SAMPLED
                                                                               WELLN
                                                                                     3T FOUND


                                                                                         ..
                                                                               o
                                                                               cc
                                                                               cc
                                                                               LU
                                                                               LU
                                                                               I
                                                                               O
                                                                               cc
                                                                               o
                                                                               o
                                                   LJ
                                                   CO
                                                   3
                                                   UJ
                                                   §
                                                   O

                                                   CO
                                                   UJ
                                                                                          o
   Figure 5.   Distribution of benzene in ground water at a LIST release in South Carolina. The red arrows are the
               distance that ground water would be expected to move in three years based on the hydraulic conductiv-
               ity and porosity of the aquifer and the hydraulic gradient that pertained in a particular round of sampling.
               The heavy blue arrow is the distance water would move under average conditions in five years. Circled
               wells have concentrations of benzene less than the detection limit.
I 0  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 with Figure 6, which shows segments connecting
 clean monitoring points to establish the maximum
 presumed extent of contamination.
 The figures in this Issue Paper were created
 using an accompanying EXCEL spreadsheet titled
 Inclusion Zone Calculations. The spreadsheet
 can facilitate the calculations necessary to apply the
 approach to other sites. The spreadsheet contains
 two tabs that facilitate finding the angle between
 well pairs. Use the following process to define an
 inclusion zone on a map of a site. Using M.S. Word,
 PowerPoint or some similar computer application,
 "insert" a straight line over the line between two
 monitoring wells on the map of the site. Then cut
 the line and paste it onto the chart in the tab Angle
 Comparison. Select the line segment and move it
 around on the chart until the axis of rotation of the
 line segment passes through the point (0,0). Then
 open the tab Data Angle Comparison, and change
the value for the direction of a test angle (Cell D21)
by trial and error until the line in Angle Comparison
labelled "test angle" converges with the line pasted
into Angle Comparison. The value of the angle
where the lines converge is the direction of the line
segment.
Evaluation continues for well pairs moving clockwise
around the perimeter as defined by the clean wells.
See Table 1 and Figure 7. The  direction of the line
segment between wells in Table 1 is presented in
Degrees from North with  the first well named in  the
line segment as the axis of rotation.  A clockwise
rotation is a positive direction and a counter
clockwise rotation is a negative direction.  The
direction of the new line segment associated with
each well pair is simply 90° less than the direction
of the segment between wells.   The resulting lateral
inclusion zone is depicted in Figure 8.
 Figure 6.   Area enclosed by the perimeter of clean monitoring wells (shaded red).
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                               round Water Issue ] ]

-------
   Table 1.   Calculations to correct the length of a new line segment for the probability that ground water will flow in
             that direction. See Figure 6 for the line segments.
Line
between
Wells

MW-1 3 to
MW-16
MW-1 6 to
MW-1 7
MW-1 7 to
MW-1 4
MW-1 4 to
MW-1 3
Direction
Line Segment
between
Wells
Degrees right of
North
129
240
291
33
Direction of
New Line
Segment
Degrees right
of North
39
150
201
-57
Distance
between
Clean Wells
Feet
444
260
328
344
Weight on
New Line
Segment

0.0000
0.9451
0.1408
0.0000
Ratio New Line
Segment to
Distance Between
Wells

1
1
1
1
Length of
New Line
Segment
Feet
0
246
46
0
                         -57° = 33°-90°
                                                             39° = 129° - 90°
                               MW-14
                                                                   150° = 240° - 90°
   Figure 7.    The area enclosed by the perimeter of clean monitoring wells (shaded red) with angles of line segments
               that connect the monitoring wells measured clockwise from North. For the line from MW-14 to MW-13,
               the angle is 33P past a complete circle. Extensions of the inclusion zone are directed 90° from the lines
               connecting the monitoring wells. For example, between MW-13 and MW-16 the outward extension is
               129° - 90° = 39°, and between MW-14 and MW-13 the outward extension is 33°- 90° = - 57°.
1 2  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
  t
North
  I
                          inclusion zone
                         based on existing
                         clean wells.
                                  MW-14
                                                                     MW-16
 Figure 8. Lateral inclusion zone defined without using information on ground water flow directions.
 2.3.2.2 A Definition that Considers Ground water
    Flow
 At sites where the flow field is primarily
 unidirectional and the aquifer can be said to
 be homogeneous and isotropic, contaminant
 plumes tend to be elongated in the down gradient
 (longitudinal) direction and extend to a smaller
 degree in the lateral (transverse) direction.  If
 historical ground water monitoring data are
 sufficient to provide a high degree of confidence
 in defining the extent of the plume, then it may be
 reasonable to reduce the extent of the inclusion
 zone in the lateral direction in proportion to the ratio
 of the longitudinal to the transverse extension of the
 plume. To make the  comparisons between lateral
 and transverse extension of a plume, it is best to
 have data describing the seasonal variability in flow
 direction and velocity, and  data from wet years and
 dry years.  Note: this information is not typically
 available at the beginning of an investigation of a
 leaking LIST. Therefore,  more conservative criteria
 are generally used, which results in a larger lateral
 inclusion zone to compensate for the uncertainty
 and variability in the ground water flow direction.
 Panel (a) of Figure 9 depicts a situation  in which
 the plume is roughly circular, with extension in the
                         longitudinal direction (x) equal to extension in the
                         transverse direction (y). Though a circular plume
                         is not common, this situation may be encountered
                         when the ground water flow field is highly variable
                         throughout the year or when a ground water mound
                         forms beneath a tank excavation. In such a case,
                         the inclusion zone could extend outward from
                         clean monitoring  points to the same distance as
                         the spacing between the monitoring points. Note
                         that the inclusion zone may also extend some
                         distance in a direction that may later (after sufficient
                         data have been collected) be considered to be
                         upgradient from the source.
                         Panel (b) of Figure 9 depicts a plume which
                         extends twice as  far in the longitudinal direction as
                         it does  in the transverse direction (or, to state this
                         differently, the plume only extends half as far in the
                         transverse direction as it does in the longitudinal
                         direction). In this  situation, the lateral inclusion zone
                         could reasonably be extended in the transverse
                         direction half the  distance of the spacing between
                         monitoring points along the sides of the plume.
                         In the longitudinal direction, the inclusion zone
                         would extend outward the same distance as the
                         spacing between clean monitoring points. Panel  (c)
                         of Figure 9 is similar to Panel (b) except that the
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                                                         round Water Issue 13

-------
        (a)Y = X
(b) Y = % X
                                                                         (c) Y = % X
                                   %T\
                                                           (d) Adjustment in Lateral
                                                               Inclusion Zone
                                                               when Y = % X
  Figure 9.   Adjusting the Lateral Inclusion Zone to compensate for variations in flow directions.
  longitudinal extension is four times greater than
  extension in the transverse direction.
  Panel (d) of Figure 9 applies this concept to define
  an inclusion zone for flow in one predominant
  direction, where the longitudinal extension is four
  times greater than extension in the transverse
  direction.  The inclusion zone in the transverse
  direction would extend outward only one-quarter
  of the distance between the clean monitoring
  locations.  These adjustments to the lateral inclusion
  zone can be made for a real  site if additional ground
  water monitoring data are available on the changes
  in the hydraulic gradient and the flow direction for
  several rounds of sampling.
  The transverse extension of a plume is generally
  presumed to be a consequence of transverse
  dispersion in flowing ground water.  Because
  transverse dispersion coefficients are low (Gelhar
  et al., 1992), as a practical matter, the transverse
  extension of a plume more likely results from
  variations in ground water flow direction over
  time (Wilson et al., 2005a).  Mace et al. (1997)
  collected data on the variation in flow direction at
  132 gasoline stations in Texas. The median of the
  standard deviation of the direction of ground water
                  flow was 36 degrees.  This extent in variation in the
                  direction of ground water flow can easily account
                  for the transverse extension of most plumes.  At a
                  site in North Carolina, Wilson et al.  (2005a) used
                  the elevation of water in wells to calculate the
                  direction of ground water flow for thirteen separate
                  monitoring events. The space occupied by the
                  plume of contaminated ground water was the same
                  as the space swept out by the variation in ground
                  water flow direction (see Figure 4(a)).

                   2.3.2.2.1 Find the Average Direction of Ground
                      water Flow.
                  The U.S.  EPA provides a software application
                  that can be used to estimate ground water flow
                  directions. It was originally intended to guide the
                  placement of additional monitoring wells at a site
                  (Srinivasan et al., 2004). The Optimal Well Locator
                  (OWL) uses linear regression to fit a plane to the
                  elevation  of ground water in wells during a particular
                  round of sampling. The slope of the plane provides
                  the best estimate of the overall hydraulic gradient
                  and direction of ground water flow during that round
                  of sampling.  The OWL software is available at no
                  cost on an EPA web site (see http://www.epa.gov/
                  ada/csmos/models/owl. html).
1 4  Ground We
   for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
The OWL computer application (Srinivasan et al.,
2004) was used to analyze data on water elevations
and fit a slope to the water table in each of seven
rounds of sampling. Data were available on the
hydraulic conductivity of ground water at the site,
and a value for the aquifer porosity was estimated
(Srinivasan et al., 2004). This information was used
to estimate how far and which direction ground
water would  move under the conditions observed
during each particular round of samples.

 ^ An EXCEL file titled Inclusion Zone
  Calculations is supplied with this issue
  paper.  The vectors are presented in the tab
  Flow Vectors.  The file contains the additional
  calculations used for the case study. The file can
  be used as a template to apply the calculations to
  another site. Data entry is in the tab data Flow
  Vectors.
 •*.	x1

The seven vectors estimate the distance that
ground water would move at the site if it moved
for three years following the hydraulic gradient
in each of the seven rounds of sampling. The
seven flow vectors are presented in red (see
Figure 5); the average is represented by the blue
vector. In general, ground water flow was not
toward the residential houses, but some of the
vectors indicated that there might be a concern that
contamination might reach some of the houses.
Notice that the flow vectors vary in both direction
and length. Simply taking the mean and standard
deviation of the flow directions would give equal
                                                   weight to short vectors and long vectors.  If we
                                                   assume that the variation in flow direction at the
                                                   site is random, we can use the normal frequency
                                                   distribution to estimate the fraction of the time that
                                                   flow might be in a particular direction.  To do that,
                                                   we need to scale the variation in flow direction to
                                                   the  probability distribution.  As an approximation,
                                                   the  flow direction was weighted by the lengths of the
                                                   vectors.
                                                   The magnitude of the hydraulic gradient at the site
                                                   varied from 0.01184 on 12/12/1995 to 0.02818 on
                                                   1/11/1999 (Table 2). Calculations of the Weighting
                                                   Multiplier for each sampling period are presented
                                                   under tab  Weight Multipliers the Excel file
                                                   Inclusion Zone Calculations.  The gradient in
                                                   each sampling period was divided by the smallest
                                                   gradient, and then the quotient was multiplied by ten
                                                   to calculate the Weighting Multiplier (expressed to
                                                   the  nearest whole number). Results are presented
                                                   in Table 2.
                                                   The weighting is accomplished in the tab Weight
                                                   Calculator in the Excel file Inclusion Zone
                                                   Calculations. The flow direction for each particular
                                                   round of samples was entered multiple times into
                                                   a column of data. The number of times a direction
                                                   was entered  was proportionate to the magnitude
                                                   of the hydraulic gradient on that date.  The number
                                                   of times a flow direction is entered becomes the
                                                   weight assigned to the data from that particular
                                                   sampling date. The mean of all of the multiple
                                                   entries of flow direction is an estimate of the
                                                   average direction of flow, and the standard deviation
Table 2.   Hydraulic gradients and flow directions were extracted for each round of sampling using OWL (red arrows
          in Figure 4). For each round of sampling the hydraulic gradient was used to select a weighting multiplier
          to be used to calculate an average flow direction and the standard deviation of the flow direction.  The
          weighting multipliers are the number of times a direction was entered in tab Weight Calculator of the
          EXCEL file Inclusion Zone Calculations.
Date
1/25/1994
12/12/1995
10/30/1998
12/4/1998
12/21/1998
1/11/1999
3/29/1999
Gradient
0.0153
0.01184
0.01335
0.01186
0.02156
0.02818
0.0198
Gradient/Smallest
Gradient
1.29
1.00
1.13
1.00
1.82
2.38
1.67
Weighting Multiplier
(number of times to enter value in spreadsheet)
13
10
11
10
18
24
17
An Approach for Developing Site-Specific Lateral and Vertical Incli
                                                                                  Ground Water Issue ] 5

-------
  of all of the multiple entries of flow direction is an
  estimate of the variability in the direction of flow.
  The proportionality factor is arbitrary.  However, to
  make the calculated mean and standard deviation
  a reasonable approximation of the "true" mean and
  standard deviation, the smallest hydraulic gradient
  should be entered at least ten times.
  If many values are available for the magnitude and
  direction of ground water flow, entering the weighted
  values of flow direction into the Weight Calculator
  tab can be tedious.  The tab Weight Calculator (2)
  automates the process to some extent. Enter data
  on the magnitude of ground water flow starting with
  cell A33 and data on the direction of flow in cell B33.
  Sort the entered data from the smallest value of
  magnitude of flow to the largest value for magnitude
  of flow. Click on cell C33, and select the  box at
  lower right with the mouse, then pull down to extend
  the formula in rowC across all the cells.  Multiply
  the ratio of the gradient to the weakest gradient by
  ten, and then enter the nearest whole number that
  corresponds to the ratio in the corresponding cells
  starting with D33.
  The spreadsheet uses nested IF statements
  to populate the weighted flow directions (X) in
  column G. The spread sheet then calculates the
  square of the weighted flow directions (X2) in
  column H.  Copy the numbers that are greater than
  zero from cells in column G and H, select paste
  special, and paste them into cells in columns I and
  J as values.  Excel 7.0 only allows seven nested
  IF statements. If there are data available from
  more than seven dates, insert the data from  the
  first seven dates, copy and paste the data from
  columns G and H into columns I and J. Then erase
  the data in columns G and H and insert the data
  from the second seven dates. Copy and  paste the
  numbers from columns G and H into columns I and
  J, inserting the new numbers below the previous
  numbers. Continue the process until columns I
  and J contain the weighted flow directions (X) and
  the square of the weighted flow directions (X2) that
  correspond to all available values for the  magnitude
  and direction of ground water flow.
  The angles extracted using OWL were then
  entered into tab  Weight Calculator in the Excel file
  Inclusion Zone Calculations. The flow  direction
  on 12/12/1995 was entered 10 times and the
           flow direction on 1 /11 /1999 was entered 24 times
           (weighting multiplier in Table 2).  Similar entries
           were made for the other dates. By following this
           procedure all the multiple entries  for all of the
           dates were used to calculate an overall mean and
           standard deviation.  For this data set, the overall
           flow direction was 157 degrees clockwise from
           North, with a standard deviation of 22 degrees
           (cells H11 and H12 of tab Weight Calculator).
           To find the weight for a particular  direction, enter
           the direction in cell H7. The weight relative to the
           average direction of ground water flow appears in
           cell H9.

           2.3.2.2.2 Assign a Weight to the Extent of the
               Inclusion Zone for the Direction of Ground
               Water Flow
           The probability that ground water will flow in a
           particular direction is taken as the solution to the
           probability density function <|)(z).
                                   1
                                  •V/27I
 -V
, 2
           The value of a particular direction of flow is
           entered in cell H7 of tab Weight Calculator. The
           spreadsheet calculates a z score for that particular
           direction by subtracting the particular direction from
           the mean direction, then dividing the difference by
           the standard deviation. The z score is reported
           in cell H15.  For the value of z, the spreadsheet
           calculates a value of the probability density function,
           
-------
 is entered into cell C1 of the calculator in the tab
 Weight Calculator it returns a weight of 0.0000.
 The probability that water will move upgradient
 across the line segment between well MW-13 and
 MW-16 is so small that it can be ignored (weighting
 factor less than 0.01). Weighting factors for these
 line segments are presented in the fifth column of
 Table 1. The weighting factor for the line segment
 between MW-13 and MW-14 is also 0.0000, thus
 for both of these segments, it is not necessary
 to extend the inclusion zone. For the segments
 between MW-16 and MW-17 and between MW-17
 and MW-14, the inclusion zone extends outward,
 but in both of these cases the distance is less than
 that separating the monitoring wells. Figure 10
 shows the  reduced inclusion zone.
 Tab New Line Segment of the Inclusion Zone
 Calculations spreadsheet uses the distance
 between the clean monitoring wells and the ratio
 between the length of the new line segment and the
 distance between the clean wells to calculate the
 length of the new line segment.
2.3.3 Determine if Additional Monitoring Points
   Would Reduce the Extent of the Refined
   Inclusion Zone
Wth the information on the direction and length
of the new line segments between well pairs,
draw a new perimeter that connects the clean
wells and the ends of the line segments that are
projected from the mid-points between clean well
(see Figure 10). Compare Figure 8 and Figure 10.
Although the inclusion zone is much reduced, the
four houses that are immediately to the West of
the contaminated area are still in the inclusion
zone.  There may be benefit in installing additional
monitoring points.
In Figure 11, a hypothetical new well is located
approximately half way between the region with
known contamination and the houses under
consideration.  If the well is clean, for the cost of
one monitoring well, the inclusion zone can be
redefined and no longer includes the four houses
under consideration. Selecting the best location
for a new well involves a trade-off.  If the new
well is located too close to existing contaminated
                                      MW-13
                    MW-14
                                                                       MW-16
 Figure 10.  A Lateral Inclusion Zone defined using information on ground water flow
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                               round Water Issue 17

-------
                                      MW-13
                                                                         MW-14
  Figure 11.   One possible outcome of the evaluation after a new well is installed to better define the Lateral Inclusion
              Zone.
  wells, there is a good chance that it will also
  be contaminated and will not help to refine the
  inclusion zone.  If a new well is located too close to
  a structure (e.g., directly adjacent to the structure
  of concern), it is possible that some portion of
  the footprint of the structure will be in the lateral
  inclusion zone,  even if the well is clean.
  After assessing the need for additional wells, install
  those that are needed and sample and analyze
  ground water to redefine the space assigned to the
  inclusion zone.  If a structure is contained within
  a lateral inclusion zone, then the structure should
  be evaluated to determine if it is within the vertical
  inclusion zone as described  in Section 3.0
  The above discussion presumed that the initial
  site characterization was conducted without
  consideration of a lateral  inclusion zone (or
  petroleum vapor intrusion). Thus the lateral
  inclusion zone extent is being added to the existing
  site conceptual model.  If the definition of the
  lateral inclusion zone is planned initially as a part
  of the site assessment, then some effort may be
  minimized. For example, monitoring wells could be
                located initially to assess building impacts, as was
                done with the additional well placed in Figure 11.

                2.3.4 Test the Inclusion Zone Against Simple
                   Transport Calculations
                The contaminant transport equation provides a
                means to forecast the distance that a contaminant
                might travel with flowing ground water.   Because
                choices must be made for parameters whose true
                values are unknown or uncertain, the forecasts
                from the transport equation are rough estimates
                rather than definitive guides.   However, the rough
                estimates provide a second line of evidence that
                can be used to evaluate the inclusion zones.
                Equations for a one-dimensional, steady-state
                transport equation solution are given in Appendix C.
                U.S. EPA provides a calculator to forecast plume
                length with these equations at http://www.epa.
                Qov/athens/learn2model/part-two/onsite/lenath.
                html. The calculations are also provided under tab
                Plume Lengths in the EXCEL file Inclusion Zone
                Calculations.
                For three of the monitoring wells in the case
                study, an estimate of hydraulic conductivity (Ks)
1 8  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 was available from rising head slug tests in the
 wells. Values were input into Column C of Plume
 Lengths.  The initial concentration of benzene along
 the flow path was assumed to be the concentration
 in the well.  Values were input in Column K. The
 final concentration along the flow path was taken
 to be 0.14 mg/L. This is the target ground water
 concentration corresponding to the target indoor air
 concentration when the indoor air attenuation factor
 is 0.001 (U.S. EPA, 2002).  This value was input
 into Column L.  The average hydraulic gradient
 was 0.0174 (Cell D11 in tab data Flow Vectors).
 This value was input into Column E of tab Plume
 Lengths.

   The target ground water concentration is
  derived from a target indoor air concentra-
  tion for benzene of 31  ug/m3 (U.S. EPA, 2002
  Table 2a). The air concentration was divided
  by the dimensionless Henry's Law constant
  (0.22 = mg/L in air divided  by mg/L in water)
  to get an equivalent concentration in water,
  and then multiplied by 1000 to allow for at-
  tenuation  of concentrations between benzene
  in soil gas beneath a  building and concentra-
  tions within the building.
  >*^	_^

 Ground water contaminated with petroleum
 hydrocarbons is consistently anaerobic. Suarez
 and Rifai (1999) reported that the mean rate
 constant for anaerobic biodegradation of benzene
 at 45 field studies was 0.003 per day, corresponding
 to a half life of 230 days.  Falta et al. (2012)
 recommends a first order rate constant of 1.1
 per year (equivalent to a half life of 230 days)
 to model anaerobic degradation of benzene at
 gasoline release sites. Data from a variety of
 field and laboratory studies are collated  in the tab
 Rates of Benzene Degradation in the EXCEL file
 Inclusion Zone Calculations. Most of the rates
 were published in Aronson and Howard  (1997).  The
 median half life was 248 days.
 A half life of 230 days was used to make the first
 estimate of  plume length, and as a sensitivity
 analysis, a half life of 693 days was also used to
 estimate plume length. A value for the degradation
 half life of 693 days includes 75% of the half lives
 collated under tab Rates of Benzene Degradation.
Values for half life are input in Column G of tab
Plume Lengths. A sensitivity analysis was also
performed with reasonable values of the effective
porosity.  Values of 0.20 and 0.25 were input into
Column D of tab Plume Lengths.
In Column O of tab Plume Lengths, the
spreadsheet calculates the lengths of the plumes
that are forecast for these specified conditions.
The calculations use a value for the longitudinal
dispersivity (a) that is input in Column I of tab
Plume Lengths. The spreadsheet uses the
formula of Xu and Eckstein (1995) to estimate an
appropriate value of a from the calculated length.
Manually input different values for a into cells in
Column I until the input value in Column  I matches
the calculated value in Column J. When values in
Columns I and J agree within a foot, the value for
the plume length in Column O can be taken as the
forecast of plume length.
Table 3 provides the plume lengths from the
sensitivity analysis. As a worked example, the
forecast of plume length for the plume originating
from MW-11was calculated as follows. Where
the hydraulic gradient (H) is 0.0174 foot per foot,
the hydraulic conductivity (K5) is 1.66 feet per day,
and the effective porosity (6) is 0.25 ft3 per ft3; the
seepage velocity (v) is:
  v =
H*Ks _0.0176* 1.66
   6         0.25
                      = 0.1156feet per day
Where the half life of natural biodegradation
is 230 days, the first order rate constant (A,) is
0.003013 per day. For well MW-11, the initial
concentration of benzene (c0) is 4.5 mg/L (see
Figure 4).  As mentioned above, the acceptable
concentration of benzene (c) is taken to be
0.14 mg/L.  Input of trial values for the longitudinal
dispersivity (a) into Column I of tab Plume Lengths
predicts a value of a of 10 feet, based on the plume
length equation from Appendix C.
     2oc In c/
                 2*12*ln
x =
            5	 _
                        0.14
                        4.5
1-
                         4*0.003013*10
                            0.1156
                                       = 162feet
Figure 12 plots the plurne lengths in Table 3 against
two configurations of the inclusion zone.  The
An Approach for Developing Site-Specific Lateral and Vertical I
                                    ' Water Issue 19

-------
   Table 3.   Maximum plume length forecast from the maximum concentration of benzene in a monitoring well, the
             hydraulic conductivity at that location, and an estimate of effective porosity and degradation half life.
Concentration
(mg/L)

Hydraulic
Conductivity
(feet per day)
Effective Porosity
ft3/ft3
Half Life
(days)
Maximum
One-dimensional Plume
Length
(feet)
MW-9
6.3
6.3
6.3
6.3
0.77
0.77
0.77
0.77
0.20
0.20
0.25
0.25
230
693
230
693
106
300
88
242
MW-11
4.5
4.5
4.5
4.5
1.66
1.66
1.66
1.66
0.20
0.20
0.25
0.25
230
693
230
693
198
560
162
453
MW-6
1.89
1.89
1.89
1.89
2.21
2.21
2.21
2.21
0.20
0.20
0.25
0.25
230
693
230
693
191
546
153
440
                                            MW-13
                            MWM4
                                                                      MW-16
   Figure 12.  Comparison of forecasts of plume lengths to two configurations of the inclusion zone.
20  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 red arrows originating at wells MW-9, MW-6 and
 MW-11 in Figure 12 correspond to the forecasts
 associated with a porosity of 0.25 and a half life of
 230 days.  The blue arrows correspond to a half
 life of 693 days. The arrows extend in the average
 direction of ground water flow.
 In the red-colored inclusion zone, the ratio of the
 distance that the inclusion zone extends past the
 clean wells to the distance between the clean
 wells is set at 1.0.  The lengths of the plumes that
 are predicted from the average rate of benzene
 biodegradation in ground water (red arrows) are
 contained within the red inclusion zone. For
 average conditions, there is no evidence from the
 forecast that the inclusion zone is not protective.
 However, this is not the case for  plume lengths
 that are based on a rate of degradation that would
 include 75% of rates in the literature (the blue
 arrows). The forecast plume lengths from wells
 MW-6 and MW-11 extend past the red inclusion
 zone. To make the inclusion zone conform to the
 forecast for well MW-6, it was necessary to set the
 ratio at 2.0 (blue-colored inclusion zone.  It is not
 possible to adjust the inclusion zone to include  the
 forecast from well MW-11  with any reasonable ratio
 of distances.
 This process should be repeated for every well
 within the area enclosed by clean monitoring wells
 using well-specific input parameters. The forecasts
 have the most value to understand the expected
 locations of the plume where no  monitoring data
 are available (such as the forecast from well
 MW-6).  The forecasts  have less value for regions
 that are represented by real monitoring data (such
 as the forecast from well MW-11  compared to the
 measurement at well MW-17).
 Although the inclusion zone seems to be greatly
 expanded by the forecast, it must be recalled
 that there are no monitoring data to support the
 assumed location of the toe of the plume (Figure 5).
 Adding a monitoring well in the primary direction
 of ground water flow would greatly increase the
 credibility of the site assessment, and very likely
 reduce the size of the inclusion zone.
 If information is available about the flow of ground
 water at the site, this information can be used to
 adjust the configuration of the inclusion zone.  If
 information about the flow of ground water is not
available, then the configuration of the inclusion
zone must be determined by professional judgment
or by local policy.  Over time as information is
collected on actual impacts to residences and
the impact that was predicted by a particular
configuration of the inclusion zone, it will be
possible to optimize this screening process.

3.0 THE VERTICAL INCLUSION ZONE
After characterizing the extent of contamination and
defining a lateral inclusion zone, there still may be
a number of residences potentially at risk for vapor
intrusion.  At this point, the vertical separation
criteria should be applied. Table 4 provides example
vertical separation-distances based on Davis (2009)
and Cal EPA (2012). The separation distance for
ground water contamination is the distance between
the lowest part of the structure of concern and the
highest historical elevation of the water table. The
separation distance for LNAPL is the minimum
extent of clean soil that is required between the
contaminated sample and the receptor. It is not
the separation distance between the contaminated
sample and the receptor.  There may be additional
contamination in soil  between the sample and the
receptor. In addition,  data on the stratigraphy at the
site, which should be incorporated into the CSM,
should be considered in determining whether there
is sufficient oxygen in the subsurface to promote
aerobic biodegradation or whether relatively
impermeable layers may prevent the intrusion of
vapors into overlying  buildings.
The limits on the vertical separation distance
that would cause a structure to be included in a
vertical inclusion zone are based on experience
with biodegradation of vapors of petroleum
hydrocarbons in the unsaturated zone (Lahvis et al.,
1999; API, 2000; DeVaull, 2007; Davis, 2009; Cal
EPA, 2012). There are two important assumptions
in applying the vertical separation distance: that
the soil is "clean" and that there is adequate
moisture in the soil to support biodegradation of the
hydrocarbon vapors.
For a PVI investigation, "clean" soil does not
necessarily mean that it is contaminant-free, but
rather that the level of any contamination present is
low enough so that the biological activity of the soil
is not diminished and the subsurface environment
will support sufficient populations of microorganisms
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue 21

-------
  Table 4.   Example conditions for a structure to be included in the Vertical Inclusion Zone. If any condition applies,
             the structure is in the Vertical Inclusion Zone.
             These conditions are provisional and are for illustration purposes only. They are based on Davis (2009)
             and Cal EPA (2012). At such time as U.S. EPA Office of Underground Storage Tanks (OUST) issues the
             Guidance for Addressing Petroleum Vapor Intrusion at Leaking Underground Storage Tank Sites, the con-
             ditions for vertical separation in the Guidance will supersede the conditions in this table.
Media
Soil
(mg/kg)
Groundwater
(M9/L)
Benzene Concentration
<10
>10 (LNAPL)
<5,000
>5,000 (LNAPL)
TPH Concentration
<250
>250 (LNAPL)
<30,000
>30,000 (LNAPL)
Vertical Separation Distance*
(feet)
<6
<15
<6
<15
    The thresholds for LNAPL indicated in this table are indirect evidence of the presence of LNAPL. These thresholds
    may vary depending on site-specific conditions (e.g., soil type, LNAPL source). Investigators may have different
    experiences with LNAPL indicators and may use them as appropriate. Direct indicators of LNAPL also apply; these
    include measurable accumulations of free product, oily sheens, and saturated bulk soil samples. For more informa-
    tion, see API (2000).
    The vertical separation distance represents the thickness of clean (TPH < 250 mg/kg), biologically active soil
    between the source of PHC vapors (LNAPL, residual  LNAPL, or dissolved PHCs) and the lowest (deepest) point
    of a receptor (building foundation, basement, or slab). EPA recommends that sub-slab sampling be conducted to
    evaluate the risk of PVI whenever contamination above the specified threshold is present in any sample and the
    distance between the contamination and an overlying building is less than these vertical distances.
  to aerobically biodegrade PHC vapors. As a point
  of departure, soil with less than 250 mg/kg TPH can
  be considered "clean." State agencies may choose
  a different definition based on their local conditions
  and circumstances. The California Environmental
  Protection Agency uses a value of 100 mg/kg (Cal
  EPA,  2012).
  Establishing that there is adequate moisture to
  support growth of bacteria is a substantial challenge
  in desert climates. U.S. EPA (2013a) notes that
  soil moisture content greater than 2% is adequate
  to support biodegradation activity (Leeson and
  Hinchee,  1996). However, biodegradation is limited
  when the moisture content is at or below the
  permanent wilting point (Zwicketal., 1995; Holden,
  Halverson, and Firestone, 1997).  Adequate soil
  moisture is indicated if the landscape supports the
  growth of indigenous vegetation (Riser-Roberts,
  1992). Agencies in states with desert landscapes
  may wish to take advantage of their local knowledge
  and apply local criteria.
             3.1 Steps to Apply a Vertical Separation
                 Distance to Core Samples
             There are five steps to defining a vertical inclusion
             distance:
              1. Acquire core samples or a series of core samples
                that represent the entire interval from the
                receptor to the lowest potential location of the
                water table.
              2. Screen the core samples in the field  with an
                Organic Vapor Meter (OVM) to determine if
                samples should be acquired for laboratory
                analysis.  If contamination is detected by the
                OVM screening, analyze the sediment samples
                for the concentrations of Total Petroleum
                Hydrocarbons and for Benzene.
              3. Compare depths and concentrations of
                contaminants in core samples to the Vertical
                Separation Distance Criteria.
              4. Acquire a sample of ground water and analyze
                for concentrations of TPH and Benzene.
2 2  Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
  5. Compare depths to ground water and
   concentrations of contaminants in ground water
   to the Vertical Separation Distance Criteria.

 3.1.1 Acquire Core Samples for Screening
 Determining the vertical separation distance for
 contamination in the unsaturated zone can be
 challenging. To apply the criteria in Table 4, it is
 necessary to document that the clean soil is in fact
 clean. Exterior bulk soil samples should be collected
 from  near the perimeter of the building in the
 direction of the source of contamination. To avoid
 missing a depth interval that might be contaminated,
 it is necessary to recover a complete profile of core
samples from the land surface to the water table.
If possible, it is better to recover core samples to
a depth equal to the lowest elevation of the water
table over time.
To assure that the core profile is complete, compare
the length of the core that is recovered (including
material in the core retainer and the cutting shoe)
to the depth interval that the core barrel was driven
into the earth.  In some subsurface materials, core
samplers driven two or three feet will recover an
equivalent length of core sample, but core samplers
driven four or five feet will not. Adjust the depth
interval driven in each core if necessary to recover
a complete core sample.
TPHGRO(mg/kg)



15
w
ID
I
CD

I
CL
a


0 5000 10000

2 -
4 •
6 •
8 •
10 •
12 •
14 •
16 •
1C -
1 O
20-
, ,

(a)


15000 20000 25000


-•-TPHGRO
—^—Hydraulic Conductivity
	 Water Table

_»
*~^ ^
_^
5-
Q 20 40

• 2
• 4
• 6
• 8
• 10
• 12
• 14
• 16
. 'I R
1 Q
• 20
60 80 100



"S
iff
m
1
(3
0
m
O.
1


Hydraulic Conductivity (feetperday)
Benzene (mg/kg)


I
₯
5
w
1
"53

S
B
Q





0 50 100
2 •
4 -
6 •
8 ',

10

12
14

16
18 -
20 -
22



(b)



150 200 250

— •— Benzene
—=— Hydraulic Conductivity
	 Water Table


-,
f^- — _
^ 	 ^.

0 20 40

2
4
6
8

- 10

12
• 14

16
- 13
- 20
- 22
60 80 100


'E
£
m
•o
E
O
3
S
m
OD
g
"o.
0)
Q




Hydraulic Conductivity (feetperday)
 Figure 13.  Distribution of TPH (panel a) and benzene (panel b) and hydraulic conductivity with depth below land
            surface at a gasoline release site in Golden, OK.
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                                round Water Issue 23

-------
   On occasion, material with a high concentration of
   TPH will literally be well lubricated, and will fall out
   of the core sampler as it is being recovered.  Do
   not ignore the missing sample. Attempt to collect
   core samples in an adjacent bore hole, starting just
   above the elevation that would correspond to the
   missing sample, and  drive the core sampler the
   maximum interval that will acquire a complete core
   sample.
   Figure 13 compares the vertical distribution of
   Total Petroleum Hydrocarbons (TPH) at a site in
   Oklahoma.  Point estimates of hydraulic conductivity
   at the site were made with a pneumatic slug test
   (Butler et al., 2002).  Notice at the site that the
   greater mass of TPH  was confined to material
   that has low hydraulic conductivity.  Petroleum
   hydrocarbons tend to be held by capillary attraction
   to fine textured materials.  At many gasoline
   service station sites,  the first aquifer to produce
   enough water to allow sampling by a monitoring
   well is effectively a confined aquifer.  Much of the
   time, the free water surface will be up in the fine
   textured material containing the TPH, and much of
   the TPH will be covered in water and not in contact
   with soil gas. In times of drought, the free water
   surface often will drop to the contact between the
   fine textured material containing the TPH and the
   transmissive material that comprises the aquifer
   proper.  During times of drought, more of the TPH  in
   the fine textured material  may be  in contact with soil
   gas.
   If a nearby monitoring well is available, determine
   the depth to the free water surface.  If an
   established monitoring well is not available,
   determine the depth to water in the borehole used
   to acquire the core samples. Examine the texture
   of the core samples taken in the depth interval
   across the  free water surface. If the material has
   a fine texture, and particularly if the borehole stays
   open, continue to acquire core samples  until more
   transmissive material is reached.
   Apply the Soil Media  Criteria in Table 4 to the TPH
   values,  even if the material is below the water table
   at the time  the cores  were acquired.

   3.1.2 Screen Core Samples for Subsequent
      Laboratory Analysis
   In the past, core samples for analysis of TPH were
   often acquired at an arbitrary depth below grade
             or an arbitrary depth above the location of the
             water table at the time of sampling. This sort of
             conventional sampling is illustrated in Figure 14.
             At the site in South Carolina, the depth from land
             surface to the water table varied from six to eight
             feet.  Over this vertical interval up to five samples
             were taken for organic vapor monitoring.
             The OUST guidance applies criteria based on the
             thickness of clean, biologically active soil between
             the top of the contamination and the receptor (U.S.
             EPA. 2013a). To apply the criteria,  it is necessary
             to document that the soil is clean across the entire
             separation distance between the contamination and
             the receptor. To minimize the chance of missing
             a contaminated depth interval, it is good practice
             to screen the core samples with an Organic Vapor
             Monitor (OVM) every 0.5 foot starting at 1.0 foot
             below land surface or 1.0 foot below the bottom
             of the structure of concern. Continue screening
             until the depth of the core samples exceeds the
             lowest possible position of the water table.  If the
             OVM meter reading exceeds  100 ppm, a sample
             should be analyzed in the laboratory for benzene
             and total petroleum hydrocarbons (TPH).  Detailed
             recommendations for extracting and analyzing core
             samples are provided in Appendix A.
             Figure 13 presents the vertical profile of TPH
             resulting from a gasoline release in Golden,
             Oklahoma. The concentration of TPH in the interval
             from 7 feet to 9 feet below grade was < 29 mg/Kg.
             Notice the sharp increase in concentrations of
             TPH and  benzene in core material at a depth that
             is just less than 10 feet below land surface. The
             concentration of TPH at a depth of 9.75 feet was
             21,000 mg/kg and the concentration of benzene
             was 197 mg/Kg.

             3.1.3 Compare the Distribution of
                Contamination in Sediment to the Vertical
                Separation Criteria
             A recent study by EPA (2013b) indicates that for
             an oxygen shadow to form beneath a building,
             and thus appreciably reduce the effectiveness  of
             biodegradation to prevent PVI, three conditions
             must be met: the building must be very large
             (including the surrounding impermeable cover),
             the source of vapors must be highly concentrated,
             and the vapor source must be in relatively close
             proximity to the bottom of the building. For a
2 4  Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 typical single family dwelling, it will generally be
 sufficient to collect exterior soil vapor and bulk
 soil samples from only one location immediately
 adjacent to the structure on the side facing the
 source of contamination.  The screening criteria
 applied will be based on the sample analyses from
 this one location. For larger structures, it may be
 necessary to collect samples and apply the criteria
 at several locations along the building perimeter and
 potentially from locations on all sides of the building.
 The criteria for the vertical separation distances
 are provided in Table 4. If either of these criteria
 for vertical separation is satisfied, this site is in
 the vertical inclusion zone, and requires further
 assessment. As indicated in Figure  13, the site in
 Oklahoma is in the vertical inclusion zone because
 there was less than 15 feet of clean  soil between
 the receptor and the first bulk soil sample with
 >250 mg/L of TPH.  The separation distance to the
 receptor was the land surface because the receptor
 had a pier-and-beam foundation.

 3.2 Steps to Apply a Vertical Separation
    Distance to Ground Water Samples
 Applying the vertical criteria for ground water is
 less challenging.  Install a monitoring well in the
 borehole used to acquire the core samples, and
 sample ground water for analysis of benzene and
 TPH. Measure the elevation of the water table in
 the new well.  If a nearby monitoring well has an
 extensive monitoring record, use the variation in
 water table elevations in the older well to estimate
 the variation in elevation of the water table at
 the new location.  Compare the elevation of the
 bottom of the structure of concern to the highest
 elevation of ground water under the  structure.  The
 vertical separation for ground water  does not make
 allowance for the capillary fringe.  Compare the
 vertical separation to the free water surface.
 At the site in South Carolina as depicted in
 Figure 14, the depth to water at the structure of
 concern is near 8 feet. If there is no residual TPH
 in the unsaturated zone, the inclusion zone is
 based solely on the depth to contaminated ground
 water. A depth of 8 feet is  greater than a separation
 distance of 6 feet as described in Table 4.  The
 structure of concern would not require any further
 investigation if the concentration of benzene in
 ground water is < 5 mg/L and TPH is < 30 mg/L.
At the site in Oklahoma as depicted in Figure 13,
the depth to the free-water surface was 13.2 feet.
However, the aquifer did not yield significant water
until a depth of 17 feet, which is considerably
below the major mass of residual gasoline.  The
concentration of benzene in the ground water
was 823 ug/L and the concentration of TPH was
12,300 ug/L.  Based on the concentration of
benzene or TPH in ground water and the separation
distance, this site would not be in the vertical
inclusion zone, and would not require further action.
However, as the site failed the soil screening (i.e.,
TPH at a depth of 9.75 feet was 21,000 mg/kg),
additional investigation for PVI is recommended.
This example illustrates the importance of acquiring
bulk soil samples for analysis, and not relying on
ground water samples alone.

 s~~                                           ^N.
  The data used in the case study are provided
 as an illustration.  They do not necessarily
 reflect current conditions at the site, and have no
 bearing on past or current regulatory action taken
 by the Oklahoma Corporation Commission.
4.0. NEXT STEPS
Approaches to screen for PVI are not limited to
the approach presented in this Issue Paper.  The
inclusion zones discussed in this Issue Paper are
defined by proximity to contaminated ground water
or to LNAPL hydrocarbons in the unsaturated zone.
If a structure is in the inclusion zone as defined by
benzene or TPH in ground water or TPH in core
samples, one possible next step is to evaluate the
concentrations of hydrocarbons in the soil gas.
Samples of soil gas can be acquired from sub-slab
monitoring points, or vapor probes, and analyzed
for contaminants of concern such as benzene. The
measured concentrations can then be compared to
concentration limits in the OSWER draft guidance
for evaluating vapor intrusion (U.S. EPA 2002).
The possibility of vapor intrusion of petroleum
hydrocarbons is inversely related to the possibility
of aerobic biodegradation of the petroleum vapors
in  the unsaturated zone (DeVaull, 2007).  In turn,
the possibility of biodegradation is related to the
separation distance, the oxygen demand of the all
the hydrocarbons in soil gas at the source of the
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue 25

-------
            Hdtf NOiS3~IHVH3
                                                               m y
                   5 S

                   II *
GROUND WAI
                                                                     5
1  I
P?  S

                                                                            o <
                                     e
                                     z

                                     ^


                                     1
       III
                                     i   ia§
                                        111 — S


EGEND
_i

5 1 <
z g t p
0 £ H 5
i 5= s §
gi f! | SI 5
II fi i H I
Sri " " J
o « o

„
I"
TOTAL BTEX
TPH(rag*g(
Un



I g
t °
£j "
o
I
Q
Z
                                                                                                 CD
                                                                                                 •ku
                                                                                                 CD



                                                                                                 I
                                                                                                 .c

                                                                                                 co
                                                                                                 to



                                                                                                 1

                                                                                                 .0)

                                                                                                 LU
                                                                                                  5
                                                                                                 O
                                                             I
                                                             LU

                                                             I
                                                                                                 0)
                                                                                                 I

                                                                                                 1

                                                                                                 1
                                                                                                 CD
                                                                                                 1
                                                                                                 I
                                                                                                 .05




                                                                                                 •*



                                                                                                 1
                                                                                                 O)
26  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 vapors, and the concentration of benzene at the
 source.
 Another approach is to use these parameters to
 screen sites for PVI. The ratio of the concentration
 of benzene in indoor air to the concentration of
 benzene in vapors at the source is called the
 attenuation factor (U.S.  EPA 2002).  Abreu et
 al., 2009) performed computer simulations that
 predicted the effect of biodegradation on the
 attenuation factor. The results of a large number
 of complex simulations are summarized in a simple
 figure that plots the attenuation factor against the
 total oxygen demand for a variety of separation
 distances. In their approach, the figure is  used
 to predict an attenuation factor that is specific for
 conditions at a particular site. To complete the
 evaluation, the attenuation factor is multiplied by the
 measured concentration of benzene in soil gas at
 the source of the vapors.
 The approach of Abreu et al. (2009) may have
 application at many sites.  However, it is important
 to attain a robust estimate of the total oxygen
 demand.  Jewell and Wilson (2011) applied the
 approach to several gasoline release sites in
 Oklahoma. They took precautions to measure
 methane in the soil gas  as well as concentrations
 of petroleum hydrocarbons.  At three of eleven
 sites, including the contribution of methane to the
 total oxygen demand caused the predicted indoor
 air concentration of benzene to exceed the U.S.
 EPA Generic Screening Level for indoor air (9.8E-
 03 ppm v/v). The sites would not have exceeded
 the Generic Screening Level if the oxygen demand
 was calculated from the concentration of petroleum
 hydrocarbons alone.
 Conventional ground water monitoring wells at
 gasoline service stations are usually screened
 across the water table. This means that monitoring
 wells can often be used to collect soil gas. Jewell
 and Wilson (2011) used conventional wells to
 acquire their soil gas samples. At many sites,  it
 may be possible to use the same wells that were
 previously used to screen ground water to screen
 soil gas.
5.0 SUMMARY
U.S. EPA's Guidance for Addressing Petroleum
Vapor Intrusion at Leaking Underground Storage
Tank Sites (U.S. EPA, 2013a) is intended to provide
general criteria to  identify structures that are at risk
from petroleum vapor intrusion  (PVI).  This issue
paper provides one approach to apply criteria set
forth in U.S. EPA (2013a), but does not represent
U.S. EPA guidance.
An inclusion zone  is used to recognize structures
that may be at risk from PVI. The inclusion zone
generally consists of a lateral zone based on the
delineation of a clean perimeter and a vertical zone
based on the vertical separation distance between
the structure and contamination in the subsurface.
The delineation of the lateral inclusion zone in
this approach recognizes the fact that the lateral
separation distance  between a  residence and
contaminated ground water is dependent on the
identification of the edge of a contaminant mass,
whether it is mobile LNAPL,  residual LNAPL, or a
dissolved plume.  Many sets of site characterization
data do not explicitly define this boundary, but rely
on drawn contours that may be arbitrary.  In this
approach, the lateral inclusion zone depends on
the delineation of a clean perimeter. If monitoring
points at a site are scarce or are widely separated,
there will be uncertainty about the location of
contamination  in the areas between the monitoring
points. A building  may be at risk even though it
is marginally outside the clean perimeter. The
approach provides a reasonable procedure to
extend the lateral inclusion zone based on the
location and spacing of monitoring  points. As site
monitoring data are collected over time, the lateral
inclusion zone may be reduced in its extent.
Once a lateral  inclusion zone is identified, it can  be
further refined  to optimize the screening process
and avoid unnecessary risk characterization within
buildings.  As is shown in the examples in this Issue
Paper, it may be necessary to acquire more data
before the approach can be used with confidence to
screen structures for PVI.  The  lateral  inclusion zone
may present a clear picture of the best locations
for new wells.  Ground water flow directions vary
at most sites,  so data collected  over time on the
direction of ground water flow can be used to
refine the inclusion zone, very possibly shrinking
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue 2 7

-------
  it. If information is available on the hydrological
  characteristics of the site, a simple transport and
  fate model can be used to forecast the lateral extent
  of contaminated ground water from particular wells.
  These forecasts can provide an additional line of
  evidence to evaluate or further refine the lateral
  inclusion zone.
  After identifying a lateral inclusion zone, there still
  may be a  large number of residences potentially
  at risk for vapor intrusion. At some sites it may not
  be possible to define a lateral inclusion zone. At
  this point, the vertical inclusion criteria should be
  applied. This Issue Paper recommends five simple
  steps to determination of the vertical  extent of clean
  soil between the building and the contamination
  below the building, and to compare that extent
  of clean soil to the criteria for vertical separation
  distance in U.S. EPA (2013a).
  In combination, definition of lateral and vertical
  inclusion zones make the best use of site
  characterization data for assessing the risk of
  PVI to structures at a LUST site. Ultimately, a
  useful prediction of the possibility of petroleum
  vapor intrusion in a particular building depends
  on knowledge of contaminant transport and
  transformation, and the site-specific distribution of
  contaminants. The procedures outlined in this Issue
  Paper provide a realistic data-driven  approach to
  screen buildings for vulnerability to PVI.


  NOTICE
  The U.S. Environmental Protection Agency through
  its Office of Research and Development conducted
  the research described here as an in-house effort.
  This Report has been subjected to the Agency's
  peer and administrative review and has been
  approved for publication as an EPA document.


  6.0 REFERENCES
  Abreu, L.  D. V., R.  Ettinger,  and T  McAlary. 2009.
     Simulated  Soil v\Vapor  Intrusion Attenuation
     Factors including Biodegradation for Petroleum
     Hydrocarbons. Ground  Water  Monitoring  &
     Remediation 29(1): 105-117.
  Aronson  D. and   PH. Howard.  1997. Anaerobic
     biodegradation  of organic chemicals  in ground
    water: A summary of field and laboratory studies.
               PreparedfortheAmerican Petroleum Institute, SRC
               TR-97-0223F.
            API. 2000. Non-Aqueous Phase Liquid (NAPL) Mobility
               Limitsin Soil. American Petroleum Institute. Soil and
               Ground water Research Bulletin No. 9.
            API.  2005. Collecting and Interpreting  Soil Gas
               Samples from the Vadose Zone: A Practical Strategy
               for Assessing the Subsurface Vapor-to-indoor Air
               Migration Pathway at Petroleum Hydrocarbon Sites.
               American Petroleum Institute. API Publication 4741.
            Butler, J.J, J.M. Healey, G.W. McCall,  E.J. Garnett,
               and S.P Loheide  II. 2002.  Hydraulic tests with
               direct-push equipment. Ground water40 (1): 25-36.
            Cal  EPA. 2012. Low-Threat Underground Storage
               Tank Case Closure Policy. California Environmental
               Protection Agency, State Water Resources Control
               Board, http://www.swrcb.ca.gov/ust/lt  els plcv.
               shtml.
            Davis,  R.V 2009. Update on  recent studies  and
               proposed screening criteria for the vapor-intrusion
               pathway. LU.S.T.LINE Bulletin 61, pages 1-14.
            DeVaull, G. E. 2007. Indoorvaporintrusion with oxygen-
               limited biodegradation for a subsurface gasoline
               source. Environmental Science and Technology
               41(9): 3241-3248.
            EPA. 2004a. Performance Monitoring of MNARemedies
               for VOCs in Ground Water.  EPA/600/R-04/027.
               Office of Research and Development, National Risk
               Management Research Laboratory, Ada, OK.
            EPA. 2004b. Monitored Natural Attenuation. Chapter IX
               inHowToEvaluateAlternativeCleanupTechnologies
               For Underground Storage Tank Sites: A Guide For
               Corrective Action Plan Reviewers (EPA 510-B-94-
               003; EPA510-B-95-007; and EPA 510-R-04-002).
               Office of Underground Storage Tanks, Washington,
               DC.
            Falta, R.W, A.N.M. Ahsanuzzaman, M.B. Stacy and
               R.C. Earle. 2012. REMFuel: Remediation Evaluation
               Model for Fuel Hydrocarbons, User's Manual Version
               1.0. EPA/600/R-12/028. Available: http://www.eoa.
               gov/nrmrl/awerd/csmos/models/remfuel.html
            Gelhar, L.W, C. Welty, and K.R. Rehfeldt. 1992. A
               critical review of data on field-scale  dispersion in
               aquifers, Water Resources Research 28(7): 1955-
               1974.
            Goode,  D.J.  and L.F.  Konikow. 1990.  Apparent
2 8  Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
   dispersion in transient ground water flow.  Water
   Resources Research. 26(10): 2339-2351.
 Holden, P. A., LJ. Halverson, andM.K. Firestone. 1997.
   Water stress effects on toluene biodegradation by
   Pseudomonas putida. Biodegradation 8(3): 143-151.
 Jewell, K. P., J. T. Wilson. 2011. A New Screening
   Method for Methane  in Soil Gas  Using  Existing
   Groundwater Monitoring Wells.   Ground  Water
   Monitoring & Remediation. 31 (3): 22-94 (2011).
 Lahvis, M., A. Baehr, and R. Baker. 1999. Quantification
   of aerobic biodegradation and volatilization rates of
   gasoline hydrocarbons near the water table under
   natural attenuation conditions. Water Resources
   Research. 35(3):753-765.
 Leeson, A.,  and R.E. Hinchee. 1996. Principles and
   Practices of Bioventing.  Volume  1: Bioventing
   Principles and Volume 2: Bioventing Design. Battelle
   Memorial Institute. September.
 Mace, R.E., R.S. Fisher, D.M. Welch, and S.P  Parra.
   1997. Extent, mass, and duration of hydrocarbon
   plumes from leaking petroleum storage tank sites in
   Texas. Geological Circular97-1, Bureau of Economic
   Geology, The University of Texas, Austin, Texas.
 Minnesota Pollution Control Agency. 2008. Soil Sample
   Collection and Analysis Procedures. Guidance
   Document 4-04. Petroleum Remediation Program.
   Available at: http://www.pca.state.mn.us/index.php/
   view-document. html?gid=3025
 Riser-Roberts, E.  1992. Bioremediation of Petroleum
   Contaminated Sites. Florida: CRC Press,  Inc.
 Srinivasan, P., Pope, D.F., and Striz, E. 2004. Optimal
   Well Locator (OWL): AScreening Tool for Evaluating
   Locations of Monitoring Wells, User's Guide Version
   1.2. EPA 600/C-04/017 February 2004. Available:
   http://www. epa.gov/nrmrl/awerd/csmos/models/
   owl, html
 Suarez, M. P. and H .S. Rifai. 1999. Biodegradation rates
   for fuel hydrocarbons and chlorinated solvents in
   ground water. Bioremediation Journal 3 (4): 337-362.
 U.S.  EPA. 1995. OSWER Directive 9610.17: Use of
   Risk-Based Decision  Making in  UST Corrective
   Action Program. Available: http://www.epa.gov/oust/
   directiv/od961017.htm.
 U.S. EPA. 2002. OSWER DraftGuidanceforEvaluating
   the Vapor Intrusion to  Indoor Air Pathway from
   Ground waterand Soils (Subsurface Vapor Intrusion
  Guidance)  EPA530-D-02-004. Available: htto://
  www. epa. gov/epawaste/hazard/correctiveaction/
  eis/vapor/complete.pdf
U.S. EPA. 2013a. [DRAFT] Guidance for Addressing
  Petroleum Vapor Intrusion at Leaking Underground
  Storage Tank Sites . (EPA-xxx-xx-xx-xxx) Office of
  Underground Storage Tanks.
U.S.  EPA. 2013b.   3-D  Modeling  of Aerobic
  Biodegradation  of Petroleum Vapors: Effect  of
  Building Area Size on Oxygen Concentration below
  the Slab; June 4, 2012 draft report prepared by
  ARCADIS U.S.,  Inc. (EPA-xxx-xx-xx-xxx) Office of
  Underground Storage Tanks.
U.S. EPA. 2013. Evaluation of Empirical Data toSupport
  Soil Vapor Intrusion Screening Criteria for Petroleum
  Hydrocarbon Compounds. (EPA-510-R-13-001)
  Office of Underground Storage Tanks.
van Genuchten, M.T andWJ.AIves, 1982. Analytical
  Solutions of the  One-Dimensional Convective-
  Dispersive Transport  Equation, U.S. Department
  of Agriculture, Agricultural  Research  Service, U.S.
  Salinity Laboratory, Riverside,  California, Technical
  Bulletin 1661.
Wlson, J.T 2003. Fate  and transport of MTBE and
  other gasoline components. In: MTBE Remediation
  Handbook,  Amherst,  MA: Amherst  Scientific
  Publishers,  pp. 19-61.
Wlson, J.T, P.M. Kaiser and C.Adair.2005a. Monitored
  Natural Attenuation of MTBE as a Risk Management
  Option at Leaking Underground Storage Tank Sites.
  EPA600/R04/179.
Wlson, J.T, C. Adair, P.M. Kaiser, and R.  Kolhatkar.
  2005b. Anaerobic biodegradation of MTBE at a
  gasoline spill site. Ground Water Monitoring and
  Remediation 25(3): 103-115.
Xu, M. and Y.  Eckstein. 1995. Use of weighted least-
  squares method  in evaluation of the relationship
  between dispersivity and field  scale. Ground water
  33(6): 905-908.
Zwick, T.C., A. Leeson, R.E. Hinchee, L. Hoeppel,
  and L. Bowling. 1995. Soil Moisture Effects During
  Bioventing in Fuel-Contaminated Arid Soils. Third
  International In-Situ  and On-Site Bioreclamation
  Symposium. In-Situ Aeration, v. 3, Batelle Press,
  San Diego,  CA.
An Approach for Developing Site-Specific Lateral and Vertical Incli
                               Ground Water Issue 29

-------
  APPENDIX A.  RECOMMENDATIONS FOR
      SAMPLING AND ANALYSIS
  The conventional practice to sample for Total
  Petroleum Hydrocarbons in some states is to take
  a bulk sample of sediment into a sealed jar, return
  the jar to the laboratory on ice, and store the jar in
  a refrigerator until a subsample was taken to be
  extracted. This practice can result in considerable
  loss of VOCs and produce erroneous results. The
  authors recommend that sediment samples for
  analysis of TPH and Benzene should be preserved
  in methanol in the field as soon as possible after the
  core samples are acquired.  In the absence of other
  guidance, the authors recommend the procedures
  and requirements as described in Minnesota
  Pollution Control Agency (2008).
  The authors have had good results using the
  following procedure to extract core samples
  into methanol. Plug-samplers were constructed
  before going to the field by cutting the end from
  a 10-ml plastic syringe (Figure A-1 & Figure A-2).
  A sediment core was acquired in an acetate
  liner.  The core was cut through with a saw at the
  depth interval to be sampled (Figure A-3). Then
  a plug-sampler was driven into the exposed face
  of the core sample. The syringe plunger was
  used to provide suction to pull the sample into
  syringe  barrel as the barrel was forced into the
  face of the core sample (Figure A-4).  Each plug
  contained approximately 10 ml of soil and extended
  approximately 2.5 inches into the core. After all
  the necessary plug samples for a particular depth
  interval  were acquired (Figure A-5), the core was
  measured and cut again to present a fresh face at
  the next interval to be sampled.
  The authors have found it to be convenient to
  take all  the samples that might be needed at the
  same time. These include one plug sample for field
  screening with an OVM, duplicate plug samples into
  methanol for analysis of TPH and benzene and a
  sample  taken into a clean empty vial for analysis
  of moisture content. The duplicate plug sample
  for TPH and benzene provides a contingency if a
  sample  is lost, and  provided a field duplicate if one
  is needed for quality assurance purposes. If the
  OVM screening did not reveal contamination, the
  other samples were not analyzed. The samples that
  were extracted into methanol were returned to the
  laboratory and discarded as hazardous waste.
            The plug sample for field screening was sealed
            into a plastic bag containing air. At a later time the
            headspace of the bag was analyzed with an organic
            vapor meter (OVM) (Figure A-6). Our screening
            essentially followed Section A. Headspace Analysis
            of Minnesota Pollution Control Agency (2008).
            Extraction vials were prepared by delivering 10 ml
            of purge-and-trap grade methanol into 40 mL
            Volatile Organic Analysis (VOA) vials. In the field,
            the plug samples were delivered into the vials
            (Figure A-7), the vials were sealed with the screw
            cap, and then the vials were shaken to begin the
            extraction and preserve the samples (Figure A-8).
            In the laboratory, the vials were shaken on a
            mechanical shaker for ten minutes.  If this was
            not adequate to disperse and extract the plug, the
            vials were open and the plug was broken up with
            a spatula, and the vial was put back on the shaker
            for additional extraction. After the sediment was
            extracted, the vials were set out on the counter
            to allow the solids to settle. Then the vials were
            opened and the methanol extracts were taken for
            analysis. The  methanol extract was diluted  into
            distilled water,  and the water was then analyzed by
            EPA Method 8260.
            The final plug sample was used to determine the
            moisture content of the sediment sample. The plug
            was delivered  into a clean empty 40 ml VOA vial.  In
            the laboratory the sample was weighted, then dried
            to constant weight and weighed again.
30  Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 FigureA-1. A sampler was constructed by cutting theendoutof   FigureA-2. Commercial samplers are also available.
     a plastic syringe.
 Figure A-3. A core sample acquired in a plastic sleeve is cut to
     access the core for sub sampling.
Figure A-4. A sampler is inserted into the cut face of the core
    sample to acquire a subsample.
 FigureA-5. Additional samples are acquired from the cut face
     as needed. One sample is transferred to a plastic bag for
     screening of volatile organic hydrocarbons. See Figure A-6.
 Figure A-6. After the volatile hydrocarbons in the subsample
    equilibrated with the air in the sealed plastic bag, the
    concentration of hydrocarbons were measured with an
    organic vapor meter.
An Approach for Developing Site-Specific Lateral and Vertical lnclusi<
                                       jnd Water Issue 31

-------
   Figure A-7. Each subsample for analysis in the laboratory is
      in traduced in to a vial previously prepared with 7 Oml of
      methanol
                  Figure A-8. The vial was sealed and shaken to disperse the
                     subsample and began the extraction. Note that the empty
                     weight of the vial and cap and the weight of the vial and
                     cap plus methanol were recorded in the laboratory when
                     the vials were made up.  When the vial is returned to
                     laboratory it is weighted again to determine the wet weight
                     of the subsample.
32  Ground We
for Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
 APPENDIX B.  QUALITY ASSURANCE
 The following information documents the data
 quality of samples collected and analyzed by
 U.S. EPA from the site in Oklahoma.
 The concentration of benzene in the water sample
 from the site in Oklahoma was analyzed using a
 modification of EPA Methods  5021A(headspace
 sampler) and 8260C (GC/MS).  The method
 detection limit was 0.18 ug/L  and the quantitation
 limit was 0.5 |jg/L. Benzene was not detected in
 the  method blank. The continuing calibration check
 was 101 % of nominal.
 The concentration of benzene and gasoline range
 organics (GRO TPH) in the methanol extracts
 were analyzed by EPA Method 5030B, Revision 2
 (purge and trap followed by GC/FID). The extract
 was diluted 1:50 into water prior to analysis. Some
 samples exceeded the calibration range. These
 samples were diluted 1:500 and analyzed a second
 time.
 The method detection limit for GRO in the extracts
 was 155 ug/L; the limit of quantitation was
 1250 ug/L. The method detection limit for benzene
 in the extracts was 6.5 ug/L; the limit of quantitation
 was 50 ug/L. Neither GRO nor benzene were
 detected in the method blank. The recovery of
 GRO in ten continuing calibration checks ranged
 from 101% to 118%. The recovery of benzene in
 ten continuing calibration checks ranged from 92%
 to 114%.
 After correcting for dilution of the extract into the
 water that was analyzed  by purge and trap, and for
 the average weight of sample that was extracted,
 the  limit of quantitation of GRO  in the sediment was
 9.2  mg/kg and the limit of quantitation of benzene
 was 0.37 mg/kg.  The detection limit of GRO was
 2.1  mg/kg and the detection limit of benzene was
 0.091 mg/kg.
 The relative percent difference between in
 concentrations of benzene in two sets of field
 duplicate samples was 2.3%  and 1.8%.  The relative
 percent difference between in concentrations of
 GRO in two sets of field duplicate samples was
 0.4% and 0.1%.
APPENDIX C. EQUATIONS FOR STEADY STATE
   PLUME CALCULATIONS
At steady state,

                dlc   dc  ,
             D—--v— -Ac = 0
                dx1   dx

where D is the dispersivity, c is concentration, x is
distance, v is seepage velocity, and X is the decay
constant. Using the boundary conditions
                   C(0)=
                   dc_
                   dx
= 0
The solution for the plume length is
                    2a In 9
               x =
                   1-J1+
where a is the dispersivity (D=av), and the
dispersivity is presumed independent of plume
length (van Genuchten and Alves, 1982).  Since
dispersivity is known to be scale dependent
(Gehlar et al., 1992), an implicit calculation can be
substituted linking dispersivity and plume length:
The Xu and Eckstein (1995) regression can give an
indication of the scale dependence of dispersivity

               a= 0.83[logxf414

where a and the plume length, x, are given in
meters.
An Approach for Developing Site-Specific Lateral
                                    Water Issue 33

-------
34  Ground Water Issue
Developing Site-Specific Lateral and Vertical Inclusion Zones

-------
SEPA
      United States
      Environmental Protection
      Agency
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.

If you do not wish to receive these reports CHECK HERED;
detach, or copy this cover, and return to the address in the
upper left-hand corner.
PRESORTED STANDARD
 POSTAGE & FEES PAID
         EPA
   PERMIT No. G-35
       National Risk Management
         Research Laboratory
       Cincinnati, OH 45268


       Official Business
       Penalty for Private Use
       $300


       EPA/600/R-13/047
       December 2012
                                                                        Printed on 100% recycled/recyclable paper
                                                                        with a minimum 50% post-consumer
                                                                        fiber using vegetable-based ink.

-------