EPA/600/R-14/318
December 2014
xvEPA
United States
Environmental Protection
Agency
Issue
Ground Water Issue Paper: An Approach that Uses the
Concentrations of Hydrocarbon Compounds in Soil Gas at
the Source of Contamination to Evaluate the Potential for
Intrusion of Petroleum Vapors into Buildings (PVI)
John T. Wilson, Kenneth Jewell, Cherri Adoir, Cindy Paul,
Christopher Ruybal, George DeVaull, James W. Weaver*
CONTENTS
1.0 Introduction 1
2.0 Conceptual Model of Aerobic
Biodegradation of Petroleum Vapors 5
3.0 Models Available to Evaluate Petroleum
Vapor Intrusion 11
4.0 Data Requirements to Forecast
Petroleum Vapor Intrusion 30
5.0 Application of the Approach to a Site
with Basements 46
6.0 Application of the Approach to a Site
with High Concentrations of Methane
and Slab-on-Grade Construction 56
7.0 Implementing the Approach 61
8.0 Comparison of Soil Gas Samples
Collected from Vapor Probes and
Water Wells 68
9.0 References 76
10.0 Quality Assurance 79
Appendix A
Kinetics of Stabilization during Field Sampling 115
Appendix B
Recommended Practice for Collecting
Soil Gas Samples 120
"Corresponding Author:
USEPA- ORD - National Risk Management Research Laboratory - Ground Water and
Ecosystems Restoration Division, 919 Kerr Research Drive, Ada OK 74820 USA
Tel: 580.436.8550 Email: weaver.jim@epa.gov
1.0 INTRODUCTION
The federal underground storage tank program
was originally created in 1984 (U.S. EPA,
2014a). It was designed to protect ground water
used as drinking water from contamination by
releases of motor fuel from underground storage
tanks. In the early years, the focus was clearly
on groundwater with little emphasis on human
exposure to vapors of petroleum hydrocarbons
that might enter buildings.
In 1983, the National Academy of Science
published their report titled Risk Assessment in
the Federal Government: Managing the Process
(NRC, 1983). As one response to the NRC
report, the U.S. EPA published its Guidelines
for Exposure Assessment (U.S. EPA, 1992).
The Guidelines emphasize a comprehensive
evaluation of all possible routes of exposure. In
the Guidelines, inhalation is as important a route
of exposure as is ingestion. As a consequence,
exposure to petroleum vapors that might intrude
into a building became an exposure scenario of
concern at sites where there was a release of
motor fuel from an underground storage tank.
-------
1.1 Purpose of this Issue Paper
This Issue Paper offers an alternative approach to
screening that could be applied before a decision
is made to do sub-slab monitoring or indoor air
monitoring. The approach in this Issue Paper is
based on three factors: (1) the concentrations of
hydrocarbons in soil gas at the source of the vapors,
(2) the separation distance between the receptor and
the source of the vapors, and (3) a presumption that
aerobic biodegradation will reduce the concentrations
of hydrocarbons in the unsaturated zone.
Results from monitoring indoor air can be influenced
by ambient air quality and sources and sinks in the
building; all of which are variable in time. Results
from monitoring sub-slab air can be variable in both
time and in location below the slab. This increases
the chances that monitoring will fail to detect vapor
intrusion. Concentrations in soil gas at the source
of the hydrocarbons are much less variable. This
approach can also be used in parallel with sub-slab
monitoring or side-slab monitoring to support the
findings of the near-slab monitoring.
The U.S. EPA Office of Underground Storage Tanks
(OUST) is currently developing guidance to evaluate
the risks from vapor intrusion of petroleum compounds
from fuel spills at underground storage tank sites. This
document, Technical Guide For Addressing Petroleum
Vapor Intrusion At Leaking Underground Storage Tank
Sites, focuses on evaluating the potential for petroleum
vapor intrusion from underground storage tanks
regulated under 40 CFR280 (see http://www.epa.gov/
oust/fedlaws/index.htm).
The OUST guidance organizes site characterization at
a fuel release site into a sequence of decision points.
In the first tier, any building of concern is evaluated
for immediate threats such as fires or explosions. If
no immediate threats are present, the potential for
petroleum vapor intrusion is evaluated in the second
tier.
The location of any building of concern is compared
to the location of known petroleum contamination
in soil and sediment or in ground water. If it cannot
be shown that the building is separated from known
contamination by clean sediment or ground water
(in the horizontal plane), the building is considered
to be in a lateral inclusion zone and the exposure to
hydrocarbon contaminants must be characterized
further. The next step is to compare the vertical
separation of the building from hydrocarbons in soil
and sediment or in ground water. If it cannot be
shown that there is an adequate vertical extent of
clean unsaturated sediment between the building and
known contamination in sediment or ground water,
the building is considered to be in a vertical inclusion
zone and the exposure to hydrocarbon contaminants
must be characterized further. Wilson et al. (2012a)
provides technical recommendations to define lateral
and vertical inclusion zones at a site.
If a building of concern is in the vertical inclusion
zone, either remediation is necessary, further
characterization is necessary, or both are necessary.
This Issue Paper provides recommendations for
additional characterization that might be conducted
before samples are acquired below the slab of a
building or from the indoor air of a building.
If it is necessary to remediate a site by active soil
venting or some other remedy that removes or destroys
hydrocarbons, the approach in this Issue Paper may
be a useful technique to characterize the efficacy of
the remedy and determine whether the remedy was
adequate to manage the risk for PVI.
This Issue Paper provides technical recommendations.
This Issue Paper is not guidance provided by U.S. EPA
OSWER or U.S. EPA OUST.
1.2 Evolution of U.S. EPA Guidance and
Recommendations on Petroleum Vapor
Intrusion
In response to the 1983 National Academy of Science
report (NRC, 1983), U.S. EPA has taken two distinct
approaches to evaluate vapor intrusion: one based
on shallow soil gas samples, and a second based on
vertical separation distances.
7.2.7 Guidance based on shallow soil gas
samples
The U.S. EPA's first approach to providing specific
technical guidance for evaluation of vapor intrusion
was the OSWER Draft Guidance for Evaluating
the Vapor Intrusion to Indoor Air Pathway from
GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Groundwater and Soils [Subsurface Vapor Intrusion
Guidance] (U.S. EPA, 2002). The guidance was
built around generic attenuation factors (a) where the
factor is the concentration in indoor air divided by the
concentration in a soil gas sample. To evaluate risk, a
sample of soil gas is acquired and analyzed; then the
concentration of the compound of concern in the soil
gas is multiplied by the attenuation factor to predict the
concentration in indoor air. Examples are presented
in Table 1.1. The values of the attenuation factor are
based on professional judgment and on representative
values of the attenuation factor that were measured in
buildings in case studies.
The guidance provided in U.S. EPA (2002) has been
updated. U.S. EPA now provides a Vapor Intrusion
Screening Level (VISL) Calculator and associated
User's Guide (U.S. EPA, 2014e, 2014f). Screening
levels for benzene and hexane at 25 °C are provided in
Table 1.2.
The OSWER Draft Guidance and the Vapor
Intrusion Screening Level (VISL) Calculator were
developed to describe the intrusion of vapors of
chlorinated solvents, in particular solvents such as
trichloroethylene (TCE). The attenuation factors are
most appropriate for chemicals that behave like TCE.
Table 1.1. Example Generic Screening Levels corresponding to a lifetime risk of 1 x 104. Values are from pages 57 and 58
of U.S. EPA (2002).
Chemical
Benzene
Hexane
Target
Indoor Air
Concentration1
(H9/m3)
31
200
Target
Shallow
Soil Gas
Concentration2
(H9/m3)
310
2,000
Target
Deep
Soil Gas
Concentration3
(H9/m3)
3100
20,000
Target
Groundwater
Concentration4
(H9/L)
140
2.9
1 Target Indoor Air Concentration to Satisfy Both the Prescribed Risk Level and the Target Hazard Index [R=10-4, Hl=1]
2 Target Shallow Soil Gas Concentration Corresponding to Target Indoor Air Concentration Where the Soil Gas to Indoor Air Attenuation Factor=0.1.
3 Target Deep Soil Gas Concentration Corresponding to Target Indoor Air Concentration Where the Soil Gas to Indoor Air Attenuation Factor=0.01.
4 Target Groundwater Concentration Corresponding to Target Indoor Air Concentration Where the Soil Gas to Indoor Air Attenuation Factor = 0.001 and
Partitioning Across the Water Table Obeys Henry's Law.
Table 1.2. Example Generic Screening Levels corresponding to a lifetime risk of 1 x 10~4. Values are from the Vapor Intrusion
Screening Level (VISL) Calculator (U.S. EPA, 2014e).
Chemical
Benzene
Hexane
Target
Indoor Air
Concentration1
(H9/m3)
31
730
Target Sub-slab and Exterior Soil Gas
Concentration2
(jig/m3)
310
7,300
Target
Groundwater
Concentration3
(H9/L)
140
9.9
1 Target Indoor Air Concentration to Satisfy Both the Target Cancer Risk (TCR) and the Target Hazard Quotient (THQ) where TCR=10-4, and THQ=1.
2 Target Sub-slab and Exterior Soil Gas Concentration Corresponding to Target Indoor Air Concentration Where the Soil Gas to Indoor Air Attenuation
Factor=0.1.
3 Target Groundwater Concentration Corresponding to Target Indoor Air Concentration Where the Soil Gas to Indoor Air Attenuation Factor = 0.001 and
Partitioning Across the Water Table Obeys Henry's Law.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 3
-------
7.2.2 Guidance Based on Vertical Separation
Distances
The OSWER Draft Guidance opted to not address
intrusion of petroleum vapors from releases of motor
fuel from underground storage tanks. As stated on
page 2 of the OSWER Draft Guidance:
The draft guidance is suggested for use at
RCRA Corrective Action, CERCLA
(National Priorities List and Superfund
Alternative Sites), and Brownfields sites,
but is not re commended for use at Subtitle I
Underground Storage Tank (UST) sites at this
time. The draft guidance recommends
certain conservative assumptions that may
not be appropriate at a majority of the current
145,000 petroleum releases from USTs. As such,
the draft guidance is unlikely to provide an
appropriate mechanism for screening the vapor
pathway at UST sites.
As is described in the U.S. EPA publication Petroleum
Hydrocarbons And Chlorinated Solvents Differ In
Their Potential For Vapor Intrusion (U.S. EPA, 2011),
there is a fundamental difference in the behavior
of chlorinated solvents such as TCE and petroleum
hydrocarbons such as benzene. Because TCE is not
biologically degraded in the presence of molecular
oxygen, it is usually not degraded in the unsaturated
zone. As a consequence, the potential for vapor
intrusion is most sensitive to the rate of diffusion of
TCE along a vertical concentration gradient in the soil
gas. In contrast, the potential for vapor intrusion of
benzene and other petroleum hydrocarbons is most
sensitive to biological degradation of the hydrocarbon
as it diffuses along the concentration gradient. Over
the same separation distance, the concentration of
benzene that might enter a building is much lower than
the concentration of TCE.
Robin V. Davis with the Utah Department of
Environmental Quality had collected and collated data
on the concentrations of petroleum hydrocarbons in
ground water and in soil gas at sites with a release
of motor fuel from an underground storage tank. In
May 2011, Robin Davis provided her database to the
U.S. EPA to provide a basis for establishing separation
distances that can distinguish sites with a significant
risk of petroleum vapor intrusion from sites with no
significant risk. Peter Eremita (Maine Department
of Environmental Protection) and Jackie Wright
(Environmental Risk Sciences Pty Ltd, Carlingford,
New South Wales, Australia) also provided significant
data.
U.S. EPA evaluated the empirical database on the
distribution of petroleum hydrocarbons in soil gas
above contaminated ground water or sediment (U.S.
EPA, 2013a). The report identified vertical separation
distances from contaminated ground water that were
adequate to allow natural aerobic biodegradation
to reduce the concentration of fuel hydrocarbons
in soil gas to acceptable levels. The report also
identified vertical separation distances from sediment
contaminated with residual fuel hydrocarbons. See
Evaluation Of Empirical Data To Support Soil
Vapor Intrusion Screening Criteria For Petroleum
Hydrocarbon Compounds (U.S. EPA, 2013a). See
"PVI Database" at http://www.epa.gov/oust/cat/pvi/
index.htm.
These empirically derived separation distances were
used to support guidance on separation criteria in
the OUST draft guidance (U.S. EPA, 2013b). The
provisional recommended vertical separation distances
are presented in Table 1.3.
The vertical separation distance represents the
thickness of clean, biologically active soil between
the source of petroleum hydrocarbon vapors and
the lowest (deepest) point of a receptor (building
foundation, basement, or slab). For this purpose,
clean is defined as having a concentration of Total
Petroleum Hydrocarbons (TPH) in soil or sediment
<250 mg/kg for Diesel Fuel or Weathered Gasoline or
< 100 mg/kg for Fresh Gasoline. The source of vapors
can be motor fuel present as a light nonaqueous-phase
liquid (LNAPL), as residual LNAPL, or dissolved in
ground water.
In 2010, a team of industry scientists, consultants,
and a state regulator (McHugh et al., 2010) proposed
that three meters (9.8 feet) was a sufficient vertical
separation distance above hydrocarbon contamination
in ground water and that ten meters (32.8 feet) was
a sufficient distance above a LNAPL source. These
distances are significantly greater than the provisional
U.S. EPA recommended vertical separation distances
in Table 1.3. This difference is because the McHugh
GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 1.3. Recommended vertical separation distance between contamination and building foundation, basement, or slab.
Media
Soil
(mg/kg)
Groundwater
(mg/L)
Benzene
<10
>10
<5
>5
TPH
< 250 for Diesel Fuel or
Weathered Gasoline
< 100 for Fresh Gasoline
> 250 for Diesel Fuel or
Weathered Gasoline
> 100 for Fresh Gasoline
<30
>30
Indication
of NAPL
No
Yes
No
Yes
Vertical Separation
Distance (feet)
>6
>15
>6
>15
et al., (2010) analysis was not able to resolve some
anomalies (Robin Davis, personal communication).
In 2013, a different team of industry scientists,
consultants, and a state regulator (Lahvis et al., 2013)
independently analyzed the data set used to extract the
provisional U.S. EPA recommended vertical separation
distances. From their analysis, they found that there
is a > 95% chance that the concentration of benzene
in soil gas will be <30 (ig/m3 whenever the separation
distance from LNAPL was >13 feet. In their analysis,
they found that there was a >95% chance that the
concentration of benzene in soil gas will be
<30 (ig/m3 at any separation distance above benzene
dissolved in ground water. Compared to Lahvis et
al., (2013), the provisional U.S. EPA recommended
vertical separation distances for contaminated
groundwater are conservative. However, there is little
practical difference in the separation distances for
LNAPL.
Wilson et al. (2012a) provides technical
recommendations to implement the provisional
separation criteria in U.S. EPA (2013b). If application
of the separation criteria determines that a building
is in the vertical inclusion zone and further action is
necessary, this Issue Paper provides one approach that
could be used to further screen the vertical distribution
of hydrocarbon contamination at the site and determine
whether it is necessary to collect soil gas samples
from beside or below the foundation of the building or
samples of indoor air.
2.0 CONCEPTUAL MODEL OF
AEROBIC BIODEGRADATION OF
PETROLEUM VAPORS
The approach taken in this document is based on
measurements of hydrocarbons in soil gas at the source
of the vapors. It is built around a conceptual model
for biodegradation of hydrocarbons in soil gas that is
widely accepted (Abreu et al., 2009a,b; DeVaull, 2007;
Lavis et al., 2013; Ostendorf and Kampbell, 1991;
Roggemans et al., 2001; U.S. EPA, 2012; U.S. EPA,
2013a,b).
2.1 Interaction of Diffusion of Oxygen and
Diffusion Petroleum Vapors
U.S. EPA (2011) provided a generalized conceptual
model of the interaction of oxygen and petroleum
vapors in the unsaturated zone (Figure 2.1).
Hydrocarbons enter soil gas by either partitioning
from dissolved hydrocarbons in ground water or
by volatilization of residual hydrocarbons in the
unsaturated zone and nonaqueous phase hydrocarbons
floating on the ground water. The hydrocarbons
diffuse upward from their source.
Aerobic biodegradation of petroleum hydrocarbons
consumes oxygen which diffuses down from the
surface. The rate of aerobic biodegradation is limited
by the supply of oxygen. The bulk of biodegradation
occurs at a front where the rate of diffusion of oxygen
GW Issue 5
-------
Land Surface
Oxygenated
Soil
Oxygen
Flux
•PHOCHj
Flux
Impacted
Soil
Increasing Concentration
Figure 2.1. The vertical distribution of hydrocarbon and
oxygen in soil (reprinted from Figure 3 of U.S. EPA, 2011).
from the surface matches the stoichiometric oxygen
demand of the hydrocarbons diffusing upward from
their source. However, as the residual hydrocarbon
vapors continue to diffuse upward to the surface, they
are further degraded.
Roggemans et al. (2001) compared the distribution of
LNAPL and hydrocarbons and oxygen in soil gas at
28 sites. They were able to categorize the behavior
of oxygen and hydrocarbons into four patterns as
described in Table 2.1.
Roggemans et al. (2001) noted that pattern C can be
explained by a failure to replace the oxygen that was
consumed to degrade hydrocarbons with oxygen from
the atmosphere. At the three sites that followed pattern
D, the soil gas was isolated from the atmosphere
by some restriction such as a layer of wet clay with
negligible air-filled porosity, by the foundation of a
building or by pavement. However, the presence of
wet clay or the foundation of a building or pavement
does not necessarily cause a site to follow pattern C.
At sites that fall into pattern C, there is no detectable
biodegradation of hydrocarbon vapors, and the
intrusion of hydrocarbons would be expected to follow
the same pattern as chlorinated solvents.
Table 2.1. Four patterns for the distribution of oxygen and petroleum hydrocarbons in soil gas at gasoline spill sites as
described by Roggemans et al. (2001).
Pattern
A
B
C
D
Number of Sites
with Pattern
16 of 28
5 of 28
3 of 28
4 of 28
Distribution of
Hydrocarbons
Hydrocarbons in soil
gas.
Hydrocarbons in soil
gas.
Hydrocarbons in soil
gas.
No hydrocarbons
detected in soil gas.
Distribution of Oxygen
throughout the Vadose Zone
Adequate concentrations of oxygen
in shallow soil gas, but not enough
oxygen in deep soil gas to support
biodegradation of hydrocarbons.
Concentrations of oxygen are
adequate to support degradation of
hydrocarbons.
Concentrations of oxygen are not
adequate to support degradation of
hydrocarbons.
Concentrations of oxygen are
adequate to support degradation of
hydrocarbons.
Consequences of Pattern
Depth distribution of hydrocarbons
matches Figure 2.1. Conceptual model
of this approach is appropriate to site.
Conceptual model of this approach is
appropriate to site.
Conceptual model of this approach is
not appropriate to site. The approach
should not be used.
Not enough information to apply the
approach. If detection limits are below
screening levels, no reason to apply the
approach.
GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
If a site matches patterns A or B, the concentrations of
a hydrocarbon of concern that leave the subsurface and
can intrude into a building are controlled by a number
of factors. The primary control is the rate constant for
biodegradation of the hydrocarbon in the aerobic zone.
The extent of biodegradation is a function of the rate
constant and the residence time of the hydrocarbon in
the aerobic zone. The residence time is controlled, in
turn, by the length of the diffusion path in the aerobic
zone. The length of the diffusion path is controlled
by the depth of the reaction zone. The depth of the
reaction zone is controlled by the balance between the
flux of oxygen from the surface and the flux of oxygen
demand associated with the hydrocarbons. As a result,
the depth of the reaction zone is controlled by the
separation distance between the source of the vapors
and the upper boundary on the soil and by the strength
of the oxygen demand. Finally, the strength of the
oxygen demand is controlled by the concentrations of
the hydrocarbons in soil gas at the source.
DeVaull (2007) developed algebraic equations that
described the behavior of hydrocarbons as depicted in
Figure 2.1. His equations locate an intermediate point
between the source of vapors and the atmosphere (or
the receptor), where the diffusion of oxygen to that
point is balanced by the diffusion of oxygen demand
from the source of hydrocarbon vapors. In soil gas
above this intermediate point, the concentrations
of oxygen should be adequate to support aerobic
biodegradation of petroleum hydrocarbons.
The actual concentration of an individual hydrocarbon
will be controlled by the interaction between diffusion
of the hydrocarbon along a concentration gradient
from the source to the receptor and degradation of the
hydrocarbon at each point along the concentration
gradient. This interaction leads to complexity.
Degradation changes the concentration gradient. The
gradient controls residence time along the diffusion
flow path, which, in turn, controls the extent of
degradation.
Figure 2.2 compares the distribution of concentrations
of benzene in soil gas in aerobic sediment to the
distribution that would be expected if there was no
biodegradation and the distribution was only controlled
by diffusion along the concentration gradient. Data are
taken from the VW-10 location at the Hal's Chevron
Site in Green River, Utah (U.S. EPA, 2014b; Wilson
et al., 2009). All the soil gas samples had adequate
oxygen to support aerobic biodegradation.
TPHg in Sediment (mg/kg)
0 100 200 300 400 500 600 700 800 900 1000
1
t
8
o
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
—TPG (GRO)
• Benzene in Soil Gas
Measured
—Benzene in Soil Gas
Predicted Diffusion
200
400 600 800 1000 1200 1400 1600
Benzene in Soil Gas (mg/m3)
Figure 2.2. Comparison of benzene in soil gas to the
distribution expected from diffusion along a concentration
gradient.
The expected redistribution of benzene from diffusion
follows a straight line. The actual concentrations
of benzene were substantially lower and followed a
curved distribution.
As a general observation, the distribution of
concentrations of individual hydrocarbons along a
diffusion path in aerobic soil or sediment follows
a first order rate law of concentration on distance
(Johnson, et al., 1999). This interaction is illustrated
in Figure 2.3. Data are taken from the Coachella Site
COA-2 as evaluated by Ririe et al. (2002).
The concentrations of oxygen declined as depth
increased and the concentrations of oxygen followed
a linear distribution with depth. In contrast, the
concentrations of benzene increased as depth
increased. The logarithm of the concentration of
benzene followed a linear relationship with depth. As
the benzene diffuses toward the surface, for a given
length in the diffusion path, a constant fraction of
the benzene vapors that are present at that depth are
removed.
If Z0 is the concentration of benzene at the deepest
depth interval where the concentrations of oxygen are
adequate for biodegradation and Z is some shallower
GW Issue 7
-------
Oxygen in Soil Gas {%)
10 15 20
25
0.5 -
1.5 •
2.0-
2.5 •
3.0 -
•} e, .
m »
• *
• « * Oxygen
m • Benzene
* '
* •
*
• u.u
• 0.5
- 1.0 •£
• 1.5 |
• 2.0 f
•2.5 °
• 3.0
. •* ^
100
10000
Benzene (pg/m3)
1000000
Figure 2.3. Inverse distribution of oxygen and benzene in
soil gas (Coachella Site COA-2 in Ririe et al., 2002).
depth, then the length of the diffusion path in the
aerobic zone is Z minus Zo. If the concentration
of benzene in soil gas at depth Z0 is C0 and the
concentration at depth Z is C, then the attenuation in
concentration of benzene between Z0 and Z is C/C0.
If attenuation in concentration with depth follows a
first order law, then:
(Z — Zo) is proportional to In (— 1.
Therefore,
or
_£_
^Co^
where Z is a proportionality constant (Johnson et al..
1998; Johnson et al., 1999; DeVaul, 2011).
The data in Figure 2.3 are re-plotted in Figure 2.4.
Because there was adequate oxygen in the deepest
sample, the deepest sample is Z0. The depth of the
sample is expressed as the distance above Z0.
The value of Z was estimated as the slope of a linear
regression of the length of the diffusion path length on
the natural logarithm of concentration of benzene in
soil gas. In this case, the value of Z is -0.29 meters.
Because the value of Z is negative, the concentration
of benzene declines as the distance above Z0 increases.
For every increase in distance above Zo equal to Z, the
concentration of benzene declines by a factor of 2.7.
0 -1 -2 -3-4-5-6 -7 -8 -9 -10
Natural Logarithm of Concentration of Benzene (jjg/m3)
Figure 2.4. Relationship between the length of the
diffusion path in the aerobic zone and the reduction
in concentrations of benzene in soil gas (data from
Coachella Site COA-2 in Ririe et al., 2002).
If data are available on the distribution of
concentrations of hydrocarbons with depth, a fitted
value for the proportionality constant Z can be used to
extrapolate a concentration that would be expected in
soil gas in contact with the foundation of a building.
However, such data are rarely available.
DeVaull (2007) defines the proportionality constant
Z as an aerobic diffusive reaction length (L^). By
making reasonable assumptions, DeVaull showed that:
Equation 2.1
where Deffis the effective diffusion coefficient, H is
the Henry's Law Constant (concentration in air divided
by concentration in water), Kw is the first order rate
constant for biodegradation of the hydrocarbon of
interest in the water phase, and 6W is the water filled
porosity. DeVaull used the relationship of Jury et al.
(1983) to estimate Defffmm the molecular diffusion
coefficients in air and in water, the Henry's Law
Constant and the total and water filled porosity.
Values for the aerobic diffusive reaction length
are particularly sensitive to the air filled porosity. As
part of a sensitivity analysis, DeVaull (2007) compared
values for aerobic diffusive reaction length (Lf>) in
8 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
soil with high and low total porosity and high and low
water content. See Table 2.2.
Values for LR varied over an order of magnitude.
A smaller value of LR corresponds to more rapid
attenuation. The upper range of the calculated LR
for benzene (26 cm) is in reasonable agreement
with the value of L for benzene that was fit to the
distribution of benzene in soil gas in Figure 2.4
(29 cm).
Although the alkylbenzenes and aliphatic
hydrocarbons differ greatly in their physical
properties and the rate constant for their
biodegradation, the calculated values for LR are
similar. This can be explained by the off-setting
influence of these properties.
The rate constant for biodegradation of the
alkylbenzenes in the pore water of the sediment is
one hundred fold slower than the rate constant for
the aliphatic hydrocarbons. However, degradation
can only happen to hydrocarbons that are dissolved
in water and the Henry's Law constant for aliphatic
hydrocarbons is approximately two hundred fold
higher. If there were equivalent amounts of the
compounds in the sediment, the concentrations
of aliphatic hydrocarbons in water would be two
hundred fold lower. As a result, the estimated
aerobic diffusive reaction lengths (Lf>) for aliphatic
hydrocarbons are only about 70% longer than for
aromatic hydrocarbons. It is reasonable to expect the
aromatic hydrocarbons and aliphatic hydrocarbons
to be degraded concomitantly as they diffuse upward
from the source of contamination.
This simple conceptual model makes it possible to
estimate the concentration of a particular hydrocarbon
that would leave the upper boundary of the soil and
be available to intrude into a building knowing only
(1) the separation distance between the source and the
receptor at the upper boundary, (2) the concentration
of the particular hydrocarbon of concern at the source,
(3) the concentration of all of the other hydrocarbon
vapors in the soil gas at the source, and (4) the
rate constant for biodegradation of the particular
hydrocarbon in the aerobic zone.
Table 2.2. Range of values expected for the aerobic diffusive reaction length (Lp). Data from DeVaull (2007).
Compound
Benzene
Toluene
Ethyl benzene
Xylenes
EC* > 5-6 aliphatic hydrocarbons
EC > 6-7 aliphatic hydrocarbons
EC > 7-8 aliphatic hydrocarbons
EC > 8-9 aliphatic hydrocarbons
Deff
cm2 sec~1
0.00097-0.026
0.00097-0.026
0.00083-0.022
0.00079-0.021
0.0011-0.029
0.0011-0.029
0.0011-0.029
0.0019-0.051
H
dimensionless
0.23
0.28
0.33
0.22
51
54
56
59
Mean Kw
hr1
0.79
0.79
0.79
0.79
71
71
71
71
LR
cm
2.1-26
2.3 - 29
2.4 - 29
1.9-23
3.6 - 44
3.7 - 45
3.8 - 46
3.9 - 47
*Equivalent Carbon number based on boiling point.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue
-------
2.2 Rates of Biodegradation in Aerobic
Geological Material
DeVaull (2007) collected and collated first order rate
constants for biodegradation of fuel hydrocarbons in
water. Recently, DeVaull (2011) provided an expanded
collection of rate constants. See Figure 2.5. Notice
the vertical orange bar. Most of the rate constants
for aerobic biodegradation of aromatic hydrocarbons
in groundwater are within an order of magnitude of
each other. The rate constants for degradation of the
alkanes were higher.
The average rates in the expanded data set in DeVaull
(2011) are approximately two fold lower than the rates
in Devaull (2007). However, the general relationships
hold. The first order rate constants for degradation of
the aromatic hydrocarbons in water were consistent
with each other. There was useful agreement between
rates for degradation of benzene. The lowest rate
constant of 41 rate constants was only an order of
magnitude lower than the geometric mean and median
of the rate constants. The rate constants for alkanes
are one to two orders of magnitude higher than the
rates of degradation of the aromatic hydrocarbons.
Results: Aerobic Petroleum Biodegradation Rates in Soil
M-31 "-HB3R"—
N.JO — - — "O •• 'A*
N-10 «0&
N«4 -IA--
N>7 •- •lEHB A -"
ALKANES
propane — -^
n-butane • • • •
n-pentane
n-hexane • • •<
melhykycluhexane *B3B
trimethylpentane -HEB
n-octane —
n-nonane
ivdecane
ivdodecane
AROMATICS
benzene
-toluene
ethylbenzene
- xylenes
trimethylbenzene
cumene
• naphthalene
jf— N.W
&\" N'lt
Q N'2
> A — W-9
— N-6
••DH* — • N*17
•ft N-4
Aromatic Hydrocarbons
kw = 0.48 /hr (0.08 to 3.0)
kw = 40 /hr (7.8 to 205)
Aliphatic Hydrocarbons
0 geometric mean * data values
^median Aari(hmetic mean
- | ' - data ranges: 50%, 68% (2 ffj, 100%
0.01 0.1 1 10 100 1000 10000
first-order water phase rate, !<„, (1/hrs)
Figure 2.5. Range, median, geometric mean and arithmetic mean of rate constants
for biodegradation of fuel hydrocarbons in soil and sediment.
10 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
2.3 Effect of Building Size
on the Distribution of Oxygen
in Unsaturated Zone
If a large building has a foundation that is slab-
on-grade, the foundation may reduce the access of
oxygen from the atmosphere to the unsaturated zone,
and reduce aerobic biodegradation of hydrocarbon
vapors in the unsaturated zone. Any area below the
slab with concentrations of oxygen that are too low to
support biodegradation is in an oxygen shadow of the
slab. The restriction of the supply of oxygen would
increase the possibility of petroleum vapor intrusion.
Under a contract with U.S. EPA, Abreu et al. (2013)
used a three dimensional computer model to evaluate
the importance of the size of the slab on the vertical
separation distance that was necessary adequate
oxygen to support aerobic biodegradation of petroleum
hydrocarbons. The following is from page 40 of their
report.
At the highest vapor concentration modeled
in this report (10,000,000 jug/m3):
• An oxygen shadow developed within one
year beneath a small building 33 fix 33ft
(10 mx 10m) with a shallow 5ft (1.6 m)
vadose zone
• An oxygen shadow developed within one
year beneath a medium size building
98ft x 98ft (30 mx30m) with a moderate
thickness vadose zone 15 ft (4.6 m)
• An oxygen shadow did not develop under
a building with dimensions of 66 ft x 66ft
(20 mx 20 m) with a moderate thickness
vadose zone 15 ft (4.6 m) even after a
simulated transport time of 20 years
A vapor concentration of 10,000,000 /ug/m3 is high.
This concentration of total hydrocarbons would
only be expected in soil gas in contact with NAPL
hydrocarbons or soil gas above ground water in contact
with NAPL. The recommended vertical separation
distance above NAPL in the [Draft] Technical Guide
For Addressing Petroleum Vapor Intrusion At Leaking
Underground Storage Tank Sites (U.S. EPA, 2013b) is
15 feet. Based on the modeling of Abreu et al. (2013),
the recommended separation distance is appropriate for
a conventional residential house or a small commercial
building with the same dimensions.
The approach taken in this Issue Paper assumes that
oxygen will be available for aerobic biodegradation of
hydrocarbons in the unsaturated zone below a building.
The approach is intended for a building of the size of a
typical residential house (66 ft x 66 ft or 20 m by
20 m). It is not intended for large commercial
buildings.
The model results suggest a potential concern with
large buildings. However, model results have not been
verified against a reasonable number of case studies.
The models may not adequately account for processes
that allow oxygen transport across large foundations.
Because there is little field data to evaluate how large
buildings affect oxygen distribution, it would not be
appropriate to apply the evaluation approach to large
buildings at this time.
3.0 MODELS AVAILABLE TO
EVALUATE PETROLEUM VAPOR
INTRUSION
Environmental models are based on the application
of mass conservation principles to transport and
transformation of quantities in the environment.
Generally, all environmental models are based on a
two-part conceptualization: an empirically-determined
principle relating chemical, physical and biological
quantities, and empirical coefficients. Taken
together, these two components have the potential for
representing transport and transformation of petroleum
vapors in the vadose zone below a building.
Although vapor intrusion models may represent
important processes, the ability to determine
definitively that there are no vapor impacts to buildings
("screen for PVI") also depends on application-related
factors. These factors include the degree to which
the site conceptual model matches the structure of the
mathematical model, the inherent limitations imposed
by the assumptions in the mathematical model, the
values chosen for input parameters, and the ability to
calibrate the mathematical model to site conditions.
In the approach taken in this Issue Paper, the
concentration of a chemical is measured in soil gas at
GW Issue ] ]
-------
the source of contamination. Then a mathematical
model is used to forecast a concentration of the
chemical in air in a building. To complete the
evaluation, the predicted concentration is compared to
a regulatory standard.
The regulatory standard is a fixed number that is
specified by the appropriate regulatory authority.
However, there is uncertainty associated with the
forecast of the model. There is uncertainty in the
chemical analysis of the soil gas; however, this
uncertainty is usually not important. There is
uncertainty in whether the sample that was collected
and analyzed was representative of the soil gas
at the source of vapors. Section 4 discusses this
contribution to uncertainty in some detail. Finally,
there is uncertainty in the assumptions made in
the mathematical model compared to the real, but
unknown, situation. This is the greatest source of
uncertainty. One way to deal with this uncertainty is to
perform an uncertainty analysis on the assumptions in
the model.
The most commonly used mathematical models for
vapor intrusion are modifications of the Johnson
and Ettinger model (Johnson and Ettinger, 1991).
However, these models do not explicitly evaluate the
contribution of aerobic biodegradation of compounds
in soil gas. These models were not written around
the conceptual model presented in Section 2. This
section describes three mathematical models that
were designed to predict the biodegradation of
petroleum hydrocarbons in soil gas, and incorporate
the contribution of biodegradation into a prediction of
vapor intrusion. These models are the Biovapor model
(API, 2012), the PVIScreen model (U.S. EPA, 2014g)
and the Abreu Three Dimensional Model (Abreu et al.,
2009a,b).
The BioVapor model is set up from default parameters.
If the user has site specific data that is different from
the default assumptions, the user can update the model.
The user can also change the assumed value for
parameters for which there are no data, and conduct a
sensitivity analysis. The user must run each simulation
in the sensitivity analysis separately.
The PVIScreen model shares a similar theoretical
structure with BioVapor. However, instead of being
set up with a discrete value for each parameter, it is set
up with a range of values or a frequency distribution of
values for many of the parameters. Individual values
for the parameters are selected at random in a Monte
Carlo simulation. One thousand separate simulations
are performed. The PVIScreen model automatically
performs the sensitivity analysis and reports the most
probable concentration of a chemical in indoor air
and the percent of the simulations that exceed the
regulatory standard that is specified by the user.
The BioVapor model and the PVIScreen model are
publically available at no cost to the user. They are
supported by a user's manual. A user can learn to run
the models in a few hours to a day. Each model run
takes at most a few seconds.
BioVapor or PVIScreen can only simulate the
behavior of hydrocarbons in the vertical dimension.
The Abreu Three Dimensional Model (Abreu et al.,
2009a,b) simulates the effects and interactions in three
dimensions. As a result, it can simulate the effect of a
building on the access of oxygen from the atmosphere
to the hydrocarbons in the subsurface below the
building. Although the Abreu Three Dimensional
Model is more capable than the BioVapor model or the
PVIScreen model, it is more complex than BioVapor
or PVIScreen. It is not intended for users other than its
developers and it is not publically available.
Representative simulations made with the Abreu Three
Dimensional Model are published in Abreu et al.
(2009a) and U.S. EPA (2012). Figures and data tables
provided in Abreu et al. (2009a) are provided in this
Issue Paper as figures. A user can compare the figures
to identify the particular simulation that most closely
describes the conditions at the user's site. The figures
allow the user to predict a value for the attenuation
factor between soil gas and indoor air knowing only
the separation distance between the building and the
source of vapors and the concentration of hydrocarbon
vapors in the soil gas at the source of contamination.
To complete the exposure evaluation, the user
multiplies the measured concentration of benzene or
other chemical in soil gas by the attenuation factor
to estimate the concentration in indoor air. It is not
necessary to know anything about the model or the
process of implementing the model to use the figures.
However, there is no mechanism to evaluate the
uncertainty in the model simulations.
12 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
3.1 BioVapor
BioVapor is a spreadsheet model that runs in EXCEL®.
It is based on the conceptual model described in
Section 2 and in particular the equations of DeVaull
(2007). The model provides a simulation of a steady
state one-dimensional distribution of hydrocarbons
from the source of hydrocarbon vapors to indoor air
in a building sited above the source of hydrocarbon
vapors. BioVapor does not directly account for
spatial or temporal variation in the input parameters.
As a result, one run of the model with one set of
input parameters should not be expected to provide
an accurate prediction of the actual behavior of
petroleum vapors at a site. The user is expected to
perform a sensitivity analysis, to set up the model for
a reasonable range of input parameters and compare
the range of predicted indoor air concentrations to the
appropriate regulatory standards.
The model has input screens for Environmental
Factors, for Chemicals, for Chemical
Concentrations, and for a Chemical Database
(Figure 3.1).
Environmental Factors include assumptions about
the supply of oxygen, exposure and risk factors,
building parameters and vadose zone (unsaturated
zone) parameters. The user can describe the supply
of oxygen three ways. The user can specify (1) a
concrete slab-on-grade foundation or a basement with
a concrete slab in contact with the soil, or (2) specify
bare soil as would be the case with a pier-and-beam
foundation or a basement with an earthen floor, or (3)
the depth of the aerobic zone below a building.
The user can accept defaults or input particular values
for Exposure and Risk Factors for inhabitants of
a building, including the target hazard quotient for
BioVapor
A 1-D vapor intrusion Model:
with Oxygen-Limited Aerobic Biodegradation
Abreu parameters
Hal's Site Cafe VW-5 simulation
1) PROJECT INFORMATION
Site ID #:
Address
Completed by
Date:
Job ID
BtoVapor Version 2.1 bets
2) INPUT SCREENS
Environmental
Factors
2)
Chemicals
Chemical
Concentrations
Chemical
Database
3) RESULTS SCREENS
1} VI Risk
2)
3)
Subsurface
Profile
Detailed
Results
Print Report
Figure 3.1. Screen capture of the opening screen of BioVapor.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 13
-------
individual chemicals, the target excess individual
lifetime cancer risk, the carcinogen averaging time,
the non-carcinogenic averaging time, the body weight
of an adult, the exposure duration, the exposure
frequency and the indoor inhalation rate exposure
adjustment.
The user can accept default parameters or input
particular values for Building Parameters that
include the indoor mixing height, the air exchange
rate, the foundation thickness, the foundation area, the
foundation crack fraction, the total porosity
of soil-filled cracks, the water-filled porosity of
soil-filled cracks and the airflow through the basement
foundation. These parameters are further described
and defined in the User's Guide (API, 2012).
Finally, the user can accept default parameters
or define values for Vadose Zone Parameters
including soil porosity, soil water content, soil
organic carbon fraction, soil bulk density, air flow
under the foundation, depth of the aerobic zone under
the foundation, oxygen concentration under the
foundation, annual mean temperature, baseline rate of
soil oxygen respiration, the depth to the source from
the bottom of the foundation and the minimum oxygen
concentration required for aerobic respiration. At other
places in this Issue Paper, the depth of the source from
the bottom of the foundation is termed the separation
distance.
The input screen for Chemicals requires the user to
identify the chemicals in ground water or chemicals
in soil gas that act as the source of vapors. The user
identifies the chemicals that are potential risk drivers,
such as benzene, that will be compared to standards.
The user also identifies other individual chemicals,
such as pentane, that are not regulated compounds
but which contribute to the oxygen demand of the
chemicals in the soil gas. Finally, the user identifies
classes of chemicals that might contribute to oxygen
demand, such as, total petroleum hydrocarbons in the
range of gasoline (TPH-GRO C6-C10).
The input screen for Chemical Concentrations
requires the measured concentration of each individual
chemical or class of chemicals that was specified in
input screen Chemicals.
Calculations in BioVapor draw from a Chemical
Database. The user can edit or modify individual
properties for a chemical such as the first-order rate
constant for biodegradation or the Henry's Law
constant.
BioVapor has an output screen for Vapor Intrusion VI
Risk (Figure 3.2). The output includes the Source-
to-Indoor Air Attenuation Factor, the Predicted
Indoor Air Concentration, the Hazard Quotient, and
the Risk Level for exposure to indoor air containing
the chemical of concern.
Commands and Opti
Home Print
SoilGastolndoofAir
Mentation Factor
Figure 3.2. Screen capture of the VI Risk output screen of BioVapor.
14 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
There is also an output screen that provides a
Subsurface Profile comparing the concentrations
oxygen and the chemicals with depth (Figure 3.3).
Chemical and Oxygen
Concentrations vs. Depth
(198)
Norm allied Concentration
Figure 3.3. Screen capture of the Subsurface Profile
output screen of BioVapor.
There is also a screen with Detailed Results including
the depth of the aerobic zone, the indoor air attenuation
factor to be expected when there is no biodegradation,
and the flux of the chemical into the building.
As applied in this Issue Paper, BioVapor simulations
were set up using the Residential Default Values on
the Environmental Factors input screen with the
following expectations. Airflow under Foundation
(Qf) was set to be equal to Air Flow Through
Basement Foundation (Qs). The value was 83 cm3 of
air per second. The Air Exchange Rate was set to 12
per day. The default rate constant for biodegradation
of benzene in BioVapor is 0.79 per hour. The rate
constant in the Chemical Database (row 14, column
U) was changed to 0.079 per hour. These parameters
were altered from the default values to produce
simulations that were conservative forecasts of the
expected concentrations of benzene in indoor air.
Simulations made with these values for the parameters
are called BioVapor Generic simulations. If the values
for water content or content of soil organic matter in
the Environmental Factors input screen were altered
to reflect site specific knowledge, the simulations were
termed BioVapor Site Specific.
The list of Environmental Factors illustrates both
the strength and weakness of BioVapor or any other
mathematical model that attempts to forecast the
concentration of a vapor from soil gas in indoor
air. The list of parameters is comprehensive and
includes many factors that are known to influence
PVI. However, site specific values for many of these
parameters will not be available at any particular site.
The models must be set up with a large number of
assumed values.
3.2 PVIScreen
The PVIScreen model was created by staff of
U.S. EPA/ORD to address uncertainty in model
parameters. Because uncertainties in parameters are
unavoidable, PVIScreen is designed to always perform
an uncertainty analysis. The PVIScreen model
incorporates the uncertainty in the output from a model
into the process of comparing the output from a model
to a regulatory standard.
As of this writing (March 2014), PVIScreen is in the
process of peer review and clearance. When it is
cleared for distribution, it will be available on an EPA
website.
PVIScreen extends the concepts of BioVapor by
• implementing an automated uncertainty analysis,
• linking directly to a fuel leaching model,
• providing the capability to use a flexible unit
conversion system,
• displaying key risk outputs directly with model
results, and
• providing an automatically-generated model
application report.
PVI GW Issue 15
-------
3.2.7 Uncertainty Analysis in PVIScreen
Uncertainty analysis, as used here, includes the impact
of the inherent sensitivity of the model to changes in
parameter values and the actual magnitude of those
changes. The method used in PVIScreen is to presume
that selected parameters of the model are uncertain and
then use Monte Carlo simulation to combine different
values for parameters in multiple runs of the model.
To accomplish this task, the model is provided with
ranges of values for each parameter or a frequency
distribution of values for each parameter. For each set
of inputs, PVIScreen runs 1,000 simulations and then
builds frequency distributions of the model outputs.
Prime examples of unmeasured parameters are the
biodegradation rates, the building air exchange rate,
the flow of air from the soil into the building, the
foundation crack width, and soil moisture. PVIScreen
allows these parameters to be defined by a range
(minimum and maximum) or a statistical distribution
(Table 3.1). Parameters indicated as following a
uniform distribution increase uniformly from the
minimum value (cumulative frequency of 0) to the
maximum value (cumulative frequency of 1). The
moisture content for example ranges from 0.05 to 0.20
for these examples. Frequency distributions defined by
more than two points are used for the biodegradation
rates which were obtained from DeVaull (2007and
2011).
The PVIScreen model was set up using a combination
of site specific values for the examples that follow and
ranges for parameters which would reasonably not be
expected to be determined for the site (Table 3.1). In
particular the values for the air exchange rate, crack
width, and advective flow into the building ("Qsoil")
are taken from previous OSWER guides. The other
parameters appearing in the table are based on site-
specific measurements for the field cases.
The biodegradation rates are automatically considered
to follow the distributions developed by DeVaull (2007
and 2011). Although the actual distributions used by
PVIScreen consist of multiple points, the rate constants
range from 0.028 per hour to 3.0 per hour for benzene,
7.1 per hour to 710 per hour for TPH by TO-15, and
0.31 per hour to 190 per hour for methane. The data
on the rate constants for biodegradation of benzene,
TPH and methane are presented in Figure 3.4.
i -
0.9 -
0.8 -
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
-•-benzene
-•-TPH by Modified TO-15
^-Methane
0.01 0.1 1 10 100
First Order Degradation Rate (1/hr)
Figure 3.4. Cumulative frequency distributions for the
first order degradation rate constants.
16 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
Table 3.1. Parameters in PVIScreen that are constants or are uniformly-distributed parameters.
Parameter
MoistureContent
Porosity
FractionOrganicCarbon
SoilTemperature
AirExchangeRate
CeilingHeight
Width
Length
FoundationDepthBelowGrade
FoundationThickness
dirt floor
CrackWidth
Qsoil
DiffusionlnAir
DiffusionlnWater
SurfaceConcentration
MinimumBiodegradationConcentration
Distribution
Uniform
Uniform
Uniform
Uniform
Uniform
Uniform
Constant
Constant
Constant
Uniform
no
Uniform
Uniform
Constant
Constant
Constant
Constant
Values
0.05
0.2
0.3
0.35
0.005
0.015
10
15
0.1
1.5
8
12
30
30
8.5
6
8
0.5
5
1
10
0.175
1.70E-05
2.89E+05
1.38E-KM
Unit
dimensionless
dimensionless
dimensionless
dimensionless
dimensionless
dimensionless
c
c
1/hr
1/hr
ft
ft
ft
ft
ft
in
in
mm
mm
L/min
L/min
cnf/s
cnf/s
mg/m3
mg/m3
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 17
-------
3.2.2 Running PVIScreen
In PVIScreen, the building, vadose zone and aquifer
are denned in a layout which relates the bottom of
the foundation to a zone of petroleum contamination.
Typically, the source of contamination is a region that
contains contaminated soil gas or a separate-phase
hydrocarbon (NAPL- or non-aqueous phase liquid).
Input parameters describe the size and characteristics
of each component in the model and PVIScreen
adjusts and annotates the site schematic depending on
the inputs (Figure 3.5).
The Opening Screen of PVIScreen (Figure 3.6) has
buttons that direct the user to Select Input, View/Edit
Input, Run PVIScreen, Schematic, Results, Report,
About, and Exit. The buttons are enabled in sequence
to direct the user through the necessary steps for
running the model.
1
| 8.377 (ft)
\_/ ^^^^^^^^^^^^i
8.000 (ft)
11. 00 (ft)
19.00 (ft)
11.00 (ft)
Figure 3.5. PVIScreen site
schematic for a NAPL source
directly beneath a building.
iiiiiiftiuiiniiiifiufiiiiHitiiiiiiittiiiiiiiiiiHii iHuiiiinir
* E'stipglrcfl Sstedlnpu! VuwCdit Input Run PVlScrem About Ex*
Existing input file named: EPAIssuePaperExample.PVIScreen.csv
W«kcm«ta PVISoten
Figure 3.6. Screen capture of PVIScreen, showing the tabs that give the user access to set up the model
and tabs to access the output of the model.
18 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
After completing all required runs of the model, the
results are processed into output probability curves for
each chemical specified to be in the soil gas. These
output probability curves, along with risk levels, are
the main outputs of the model. There is a tab for each
chemical in the soil gas.
Figure 3.7 provides the output of the predicted
concentrations of benzene in indoor air from an
example PVIScreen simulation. The example
simulation used concentrations of benzene and other
hydrocarbons that would be expected to be in soil
gas in contact with NAPL. The chart is in the tab
that presents the output for benzene concentrations
in indoor air. The x axis is the base-10 logarithm
of the concentration of benzene in indoor air, where
concentration is in (ig/m3. The y axis is the relative
frequency of that concentration.
Notice the grey line at the bottom of the figure. This
is the probability density function of the predicted
concentrations of benzene. It is analogous to the
familiar bell curve that describes the normal or
Gaussian distribution. The concentration that
occurs at the highest frequency is the most probable
concentration. This is the concentration at the top
of the curve that is marked with the "M". The value
of that most probable result is 0.2371 (ig/m3. The
base-10 logarithm of 0.2371 is -0.625, as is plotted
on the x-axis in the figure. As the concentrations
become progressively higher than 0.2371 (ig/m3 or
progressively lower than 0.2371 (ig/m3, they are less
frequent.
The multicolored line in Figure 3.7 is the cumulative
frequency distribution. It is the frequency of all
concentrations that are equal to or less than the
concentration that is plotted on the x- axis. The
cumulative frequency distribution starts low and
increases as more values are added. PVIScreen
truncates the curve below an indoor air concentration
of 10'4 (ig/m3. When all the values are included the
frequency is 1.0 or 100%. The cumulative probability
curve is color coded for the probability that the
• EPAPVEcmn
« bating Input Select Input
Statistics results plotted
i f. --1-* tolu*n« ctnytol
VwvJEM Input Schematic Run PVIScreen
sult for benzene indoor air concentration
1.09 Indcw A* Conctfttmcii (Logic ug/m3)
Wrie Report At»ui
b
0.2U» Mm* Rn* Id 5 Uwl 12WO ug/mJI
•tr indKiMi iptofwd uncxt ink lew!
OKAbonMuirdQuotMfitaf 10{»DO ugymil
' H" mdtuuH ipeatMd huird quotient
Probability Thrt ChOMn Itoh L»«l<«) Ai» ti
H»]lwi|xglMbil4yal>«
Modvile piotib* fy ol
^ Lower protuMity of nc
I 07371 iig/m]
« rrtn) by J l» % rf mnulMoii)
*M" tfKlit jtn nwrmun pfrjtuMily tnuK
Figure 3.7. Screen capture of PVIScreen, showing results for benzene. "M" indicates the most probable
PVIScreen result, "C" and "H" indicate the specified cancer risk and non-cancer hazard levels, respectively.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 19
-------
selected value of C or H will be exceeded in the
cumulative distribution. The probability of exceeding
the value of C or H is the probability of error in
saying that there is no risk of death from cancer or
acute disease. If C or H falls into the portion that is
coded green, the probability of error is low. If C or
H falls into the portion coded yellow, the probability
is moderate, and if C or H falls into the red, the
probability is high.
When this particular PVISceen simulation was set
up, the acceptable risk of dying from cancer due
to exposure to benzene in indoor air was set at 1 in
100,000 (l.OE-05 or ten to the minus five risk level).
This corresponds to a benzene concentration of
2.9 (ig/m3, which is the mid-point of the inhalation
exposure for carcinogenic risk specified in the EPA's
Integrated Risk Information System (IRIS). This
concentration is marked with a green vertical line
segment on the cumulative frequency distribution.
The segment is labeled C. Under the circumstances
defined in the input to PVIScreen for this simulation,
0.2625% of the Monte Carlo runs exceeded 2.9 (ig/m3.
The uncertainty analysis indicates that there should
not be a risk of cancer from benzene in indoor air (The
Most Probable Result of 0.2371 (ig/m3 is less than the
acceptable risk level of 2.9 (ig/m3). The chance of
error in that determination is 0.3%.
The vertical green line labeled H is the concentration
of benzene (30 (ig/m3) that corresponds to the Hazard
Quotient that was specified in the input. As noted
in the legend, none of the Monte Carlo simulations
exceeded 30 (ig/m3. The uncertainty analysis indicates
that there should not be a risk of an acute hazard
from benzene in indoor air. Because 1000 runs were
included in the simulation, the chance of error in
that determination is less than 0.1%. Because the
most probable concentration (the vertical blue line
mentioned above) is well below both the cancer and
hazard levels, there is additional confidence in the
determination that this case presents low risk.
PVIScreen automatically produces a report which
• describes the model,
• shows the physical layout for the simulation,
• ranks the input chemicals for cancer and non-
cancer risks, and
• repeats the input data.
Figure 3.8 presents a screenshot of a portion of
the report that is automatically generated for each
simulation. In addition to the sections pictured in
the figure on PVIScreen Background and Run
Identification , the report also contains sections on
Site Description, Risk Results, Input Data, and
Chemical Input Data.
The report is written in standard hypertext markup
language (HTML) and is displayed in a browser.
PVIScreen displays the report if possible; otherwise it
directs the user to the location of the report. Browser
names and their directories can be loaded into an input
file to assure PVIScreen can display the report.
3.3 Abreu Three-Dimensional Model
Lilian Abreu created a complex three-dimensional
computer model of petroleum vapor intrusion that
can be used to evaluate the role and contribution of
the important properties of the building, as well as
the behavior of the hydrocarbons in soil gas beneath
and beside the building (Abreu and Johnson, 2005;
Abreu and Johnson, 2006; Abreu et al., 2009a; Abreu
et al., 2009b; Abreu et al., 2013). The model is not
available to the public, and does not have a name.
For convenience, it will be called the Abreu Three-
Dimensional Model in this Issue Paper. Abreu et
al. (2009b) is the most extensive presentation of
simulations using the model.
The model is a numerical code that solves equations
for transport and reaction of oxygen and hydrocarbons
in three-dimensional space. It considers advective
flow of air and diffusion in soil gas and soil water.
The model considers the dimensions of the foundation
of building, the depth of foundation below land
surface, and the size and location of cracks in the
foundation. It considers the volume of air in the
building and the average turnover time of the air. It
considers critical soil properties including the soil
texture and the water-filled porosity of the soil. It also
considers aerobic degradation of the hydrocarbons in
soil gas and the oxygen consumption resulting from
biodegradation of the hydrocarbons.
20 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
')\S] C\UMB\)im\Woite|M«\PVEtrMn- P - 0 |1 QOUh«s\Jim\Worte(ii«\(>...
PVIScreen Model Report
PVIScreen Background
PVIScreen is a model for assessing impacts from petroleum vapors on residences. PVIScreen was designed for automatic
uncertainty analysis using Monte Carlo simulation. The main result from the model is a probabiliy curve for indoor air
concentration for each simulated chemical. Both cancer and non-cancer risk levels are indicated on the probability curves.
PVIScreen is based on the BioVapor model (Devaull, 2007; API, 2010). PVIScreen extends the capabilities of BloVapor by
including automatic uncertainty analysis, flexible unit selection, and direct inclusion of liquid gasoline (NAPL). Major
assumptions of the model include:
• Oxygen supply permits/limits biodegradation of petroleum vapors
• Multiple components of fuel contribute to oxygen demand
• Soil Respiration contributes to oxygen demand
• Homogeneous vadose zone
• Steady state conditions
Run Identification
Site: Basement House
Location: Broadway Blvd
City: Ada
State: OK
Analyst: JWW
Figure 3.8. Screen capture of the first lines of a report generated with PVIScreen.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 21
-------
3.3.7 Effect of Depth of the Source of
Hydrocarbon Vapors
Three simulations from the Abreu Three-Dimensional
Model are depicted in Figure 3.9. The charts represent
a cross section through the three-dimensional space
simulated by the model. The charts start at the center
of the building and extend past the building into open
land.
The projections of the model are consistent with
the conceptual model discussed in Section 2.0. The
concentrations of oxygen decrease with depth until
they can no longer sustain aerobic biodegradation.
Compare the lower left hand panel of Figure 3.9.
The source of hydrocarbon vapors is 9 meters deep.
Under open land, the concentrations of oxygen are
not adequate for aerobic biodegradation at a depth
near 6 meters below land surface. At depth intervals
less than 6 meters, the concentration of hydrocarbons
decline as a logarithmic function of decreasing depth.
For each meter of approach to the land surface, the
concentration decreases approximately one-hundred
fold.
Underneath the building, the relationship is more
complex. The building tends to shield or shadow
the soil from the atmosphere. As a result, the
concentrations of oxygen are lower under the building,
and the concentrations of hydrocarbons are higher.
The highest concentrations are near the center of
the building. In this situation, the vulnerability of
a building to vapor intrusion would depend on the
location of the cracks with respect to the dimensions of
the building. Cracks in the center would produce more
vapor intrusion than cracks at the periphery.
The Abreu Three-Dimensional Model is capable of
integrating the different contribution from various
locations into an overall estimate of the attenuation
factor (a). This is an important feature that is
not shared by the other screening models that are
available.
Hydrocarbon
Oxygen
Figure 3.9. Simulations of
the distribution of oxygen and
hydrocarbon vapors near and
beneath a building after transport
and reaction processes come to a
steady state. The total concentration
of hydrocarbons in soil gas at the
source was 1.14E+08|jg/m3. The
rate constant for biodegradation
of hydrocarbons in soil water was
0.79 per hour. The contours of
concentrations of hydrocarbons
are normalized to the source. The
contours of oxygen are normalized
to the atmosphere. Figure 5 in
Abreu et al. (2009a). Reprinted
from Groundwater Monitoring &
Remediation with permission of the
National Ground Water Association.
Copyright 2009.
22 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Compare the panels depicting hydrocarbons in Figure
3.9. Notice that the estimated values for the attenuation
factor are a sensitive function of the separation
distance between the source of vapors and the bottom
of the receptor. If the separation distance is 7 meters,
the estimated value of a is 3.6E-12. If the separation
distance is 5 meters, the estimated value of a is
7.8E-08. If the separation distance is only 1 meter, the
estimated value is almost ten thousand fold lower (a
is 8.7E-04). Changes of only a few meters will change
the estimated value of a by ten-thousand fold.
3.3.2 Effect of a Basement Compared
to Slab-on-Grade
The Abreu Three-Dimensional Model predicts a
difference in the behavior of buildings that are
constructed with a slab-on-grade foundation, or that
have a basement (Figure 3.10). If the depth to the
source of vapors from land surface is the same, the
building with a basement is more at risk.
Because it is complex and because it is not publicly
available, the Abreu Three-Dimensional Model may
not be the most accessible or convenient model to
screen a particular site. However, the forecasts of the
model compare well to actual field data. Because it is
so detailed, it should provide a robust estimate of the
steady-state behavior of under specified conditions.
The model is most sensitive to the concentration of
hydrocarbons in soil gas at the source of the vapors,
to the separation between the source of vapors and
the building acting as a receptor and to the rate of
biodegradation of the vapors. The model has been run
over a realistic range of these three parameters (Abreu
et al., 2009a). The conditions at a particular site can be
matched to a preexisting model simulation to provide
a convenient forecast of the value of the attenuation
factor (a).
Hydrocarbon
Oxygen
Figure 3.10. Comparison of
the distribution of oxygen and
hydrocarbon vapors in soil gas
beneath or beside a building
with a basement compared to a
building that is constructed slab-
on-grade. Reprinted from Figure 7
in Abreu et al. (2009a). Reprinted
from Groundwater Monitoring &
Remediation with permission of the
National Ground Water Association.
Copyright 2009.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 23
-------
3.4 Summary Figures of Abreu
Three-Dimensional Modelling
Abreu et al. (2009a) ran multiple simulations for two
buildings. One building had a basement with the
bottom of the slab 2.0 meters below land surface.
The other house was slab-on-grade with the slab 0.2
meters below land surface. The houses were 10 meters
long and 10 meters wide. The thickness of the slab
was 0.15 meters. The air mixed to a height of 2.4 m
(ceilings were 8 feet high). As a result, the volume of
indoor air was 244 m3. The air in the house was turned
over every 0.5 hours. The only crack in the slab was
along the perimeter of the foundation. The total crack
length was 39 meters. The width of the crack was
0.1 cm.
The soil was a homogeneous sandy soil. The volume
of total pore space in the soil was 37.5% of the total
volume of the soil. The volume of water in the soil
was 5.4% of the total volume. The water content of
the simulation is relatively low. Because the bacteria
that degrade the hydrocarbons inhabit the soil water,
the rate of biodegradation as predicted by the model
is directly proportional to the water content. To
provide a conservative estimate, a soil type and water
content were selected that are associated with less
attenuation due to biodegradation.
3.4.1 Steps in an Exposure Evaluation of PVI
for Benzene
For the convenience of the reader, Abreu et al.
(2009a) organized some of their simulations into a
simple figure that plots the simulated value of the
attenuation factor (a) against the concentration of
the hydrocarbons at the source for five different
separation distances. See Figure 3.11. A semi-site
specific estimate for the attenuation factor (a) is
extracted from the figure by the following process. A
sample of soil gas is acquired and analyzed for total
hydrocarbons. This includes petroleum hydrocarbons
and methane. The theoretical oxygen demand of the
total hydrocarbons is expressed as the concentration of
benzene with an equivalent demand. In their example,
the concentration of total hydrocarbons was equivalent
to 10 mg/L benzene. Then the separation distance
between the depth of the gas sample and bottom of the
receptor is determined. In the example, this distance is
2 meters. A line is projected up from the concentration
of hydrocarbons to the separation distance and then
across to the estimate of a. In this case, the estimate is
l.OE-07.
The concentration of benzene in indoor air is estimated
by multiplying the measured concentration of benzene
in the sample of soil gas by the estimate of a. The
final step is to select an acceptable concentration of
benzene and compare the estimated concentration of
benzene in indoor air to the acceptable concentration.
1.E-02
Dissolved phase
NAPL
1.E-10
0.1
1 MO) 100
Source Vapor Concentration (mg/L)
1000
L= 1 m
L = 2 m
•L = 3m
L = 5 m
— *— L= 10m
— No Biodegradation, L = 1 m
-- No Biodegradation, L =10 m
Figure 3.11. Simulated values of the attenuation
factor as predicted from the concentration of
hydrocarbon vapors at the source and the separation
distance. Reprinted from Figure 10 in Abreu et al.
(2009a). Reprinted from Groundwater Monitoring &
Remediation with permission of the National Ground
Water Association. Copyright 2009.
24 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
3.4.2 Figures to Estimate a to Screen
for PVI of Benzene
Abreu et al. (2009a) provided additional estimates of
the attenuation factor (a) as tables. Estimates were
provided if there was no biodegradation, if the rate
constant for aerobic biodegradation in soil water was
0.079 per hour, and if the rate constant was 0.79 per
hour. A rate constant of 0.79 per hour was selected
because it was the geometric mean of the rates of
degradation of aromatic hydrocarbons as reported in
DeVaull (2007). A rate constant of 0.079 was selected
as a reasonable lower boundary on the rate constant.
DeVaull (2011) reported an updated collection of rate
constants. The range of 31 rate constants for benzene
biodegradation was 0.028 to 3 per hour. The geometric
mean for benzene biodegradation was 0.3 per hour.
The geometric mean divided by or multiplied by the
geometric standard deviation was 0.079 to 1.2 per
hour. A rate constant of 0.079 per hour would include
the higher 84% of published rates. A rate constant of
0.79 would include the higher 24% of rates in DeVaull
(2011).
Estimates were also provided in the tables for a
building built slab-on-grade, as well as a building
with a basement. Figures 3.12 through 3.17 present
the estimates in the tables of Abreu et al. (2009a) as
figures similar to Figure 3.11. Figures 3.13, 3.15 and
3.17 are for buildings with basements. Figures 3.12,
3.14 and 3.16 are for buildings built slab-on-grade.
Figures 3.12 and 3.13 forecast an attenuation factor
(a) where there is no biodegradation. Figures 3.14 and
3.15 make the forecast with a rate constant of 0.079
per hour, and Figures 3.16 and 3.17 make the forecast
with a rate constant of 0.79 per hour.
The unit used in Figure 3.11 for the concentration of
hydrocarbons in soil gas is not something that can be
measured directly. It must be calculated from reported
concentrations. When soil gas is analyzed for the
concentration of total petroleum hydrocarbons (TPHg),
the conventional unit is (ig/m3 of a hydrocarbon with a
molecular weight of 100 g/mole (CyH16). To facilitate
direct comparison of analytical data to the figures,
Figures 3.12 through 3.17 express the theoretical
oxygen demand of the hydrocarbon vapors as the
equivalent demand of CyH16 in (ig/m3 and not the
equivalent demand of benzene in mg/L.
The red and yellow regions in Figures 3.12 through
3.17 are projections of the values of the attenuation
factor (a) that are necessary to attain the acceptable
indoor air concentrations for benzene if the
benzene content in the soil gas were the maximum
concentration of benzene that could be expected in
soil air above gasoline. If the soil temperature were
26 °C (summer in Florida), the vapor pressure of
benzene would be 100 mm Hg or 0.13 atmospheres.
Before the initiation of Reformulated Gasoline in
1995, the benzene content of gasoline could be as
much as 2.5% (Kirchstetter et al., 1999). If the vapor
pressure of benzene above gasoline is proportional
to the mole fraction of benzene in gasoline, the
concentration of benzene in soil gas would be 0.0042
atmospheres or 1.4E+07 (ig/m3.
This value of 1.4E+07 (ig/m3 will be used as a
plausible maximum value for benzene in soil gas in
contact with NAPL. At most older spills, weathering
processes will reduce the equilibrium concentration
of benzene many fold. If the acceptable indoor air
concentrations at the 10~6, 10~5 and 10~4 risk levels are
0.3, 3 and 30 (ig/m3, then the necessary attenuation
factors (a) would be 2E-08, 2E-07 and 2E-06,
respectively.
The upper boundary on the red region in Figures 3.12
through 3.17 is the attenuation factor needed to reach
the 10"4 risk level. The lower boundary of the red
region is the attenuation needed to reach the 10~5 risk
level. The lower boundary of the yellow region is the
attenuation needed to reach the 10~6 risk level. At the
right boundary of the red and yellow regions, benzene
would be 2% of the hydrocarbons in soil gas. The
regions make the very conservative assumption that
the concentration of benzene in soil gas in contact with
the source of hydrocarbon vapors does not change as
the concentration of total hydrocarbons goes down.
The regions assume that the concentration of benzene
at the source only goes down when benzene is 100% of
the total hydrocarbons. That is the reason for the break
in the regions at a concentration of 1.4E+07 (ig/m3.
If a particular separation distance and a particular
value of total hydrocarbons plots above the colored
region, the predicted value of a is not adequate to
prevent PVI of benzene. If the predicted value is
below the colored region, the value is adequate.
GW Issue 25
-------
1.0E-02
1.0E-03
1.0E-04
1.0E-05
e 1.0E-06
O
1.0E-07
1.0E-08
1.0E-09
1.0E-10
A = 0.0 per hour
Slab on Grade Construction
I II » i LU
-Separation 1 m
-•-Separation 2 m
Separation 3 m
— Separation 4 m
—Separation 5 m
Separation? m
Separation 10 m
1.0E+05 1.0E+06 1.0E+07 1.0E+-08 1.0E+09
Source Vapor Concentration as TPH plus Methane (ug/m3)
Figure 3.12. Forecasts of a assuming slab-on-grade construction and no biodegradation.
1.0E-02
A = 0.0 per hour
Basement
-Separation 1 m
-•-Separation 2 m
Separation 3 m
-Separation 4 m
-Separation 5 m
Separation? m
Separation 10m
1.0E-10
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (pg/m3)
Figure 3.13. Forecasts of a assuming a basement and no biodegradation.
26 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
?
o
'*-»
3
A = 0.079 per hour
Slab on Grade Construction
, u m ^
.
-Separation 1 m
-•-Separation 2 m
Separations m
-Separation 4 m
—^-Separation 5 m
Separation? m
Separation 10m
1.0E-10
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (ug/mj)
Figure 3.14. Forecasts of a assuming slab-on-grade construction and a low rate of biodegradation.
A = 0 079 per hour
Basement
-Separation 1 m
-•-Separation 2 m
Separation 3 m
-~~ Separation 4 m
-Separation 5 m
Separation 7 m
Separation 10m
1.0E-10
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (jjg/nv)
Figure 3.15. Forecasts of a assuming a basement and a low rate of biodegradation.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 27
-------
1.0E-02
1.0E-03
re
3
E 1
.OE-04
.OE-05
.OE-06
.OE-07
OE-08
OE-09
1
1
1.0E-10
A = 0 79 per hour
Slab on Grade Construction
-•-Separation 1 m
-•-Separation 2 m
-Separation 3 m
-Separation^ m
—Separation 5 m
Separation 7 m
Separation 10m
II
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (ug/m3)
Figure 3.16. Forecasts of a assuming slab-on-grade construction and a medium rate of biodegradation.
1.0E-02
1.0E-03 i
A = 0.79 per hour
Basement
-•-Separation 1 m
-•-Separation 2 m
Separation 3 m
-Separation 4 m
—Separation 5 m
Separation? m
Separation 10m
1.0E-10
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (ug/m3)
Figure 3.17. Forecasts of a assuming a basement and a medium rate of biodegradation.
28 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Compare Figures 3.12 and 3.13. If the petroleum
hydrocarbons do not degrade, there is little influence
of separation distance or presence of a basement on the
simulated values for the attenuation factors (a). In all
cases, they cluster around l.OE-03. The actual value
of the attenuation factor would be largely controlled by
properties of the building such as size and location of
cracks, the volume of air in the building and the rate of
exchange of air in the building. In every simulation,
the values of a are not adequate to meet indoor air
standards when the concentration of benzene in soil
gas is the plausible maximum concentration.
Figures 3.14 and 3.15 present simulations with a
relatively low contribution of biodegradation. In
general, the lower the concentration of hydrocarbons
in soil gas, the lower the attenuation factor (a). Below
concentrations near 1E+07 (ig/m3, the curves flatten
out. This is because the concentration of hydrocarbon
is too low to bring the concentration of oxygen below
the concentration needed for aerobic biodegradation.
All of the soil profile is available for aerobic
biodegradation. At high concentrations of
total hydrocarbons in soil gas at the source, the
predicted values of a converge on the values
predicted in the absence of biodegradation. This
is because the oxygen demand exerted by the
hydrocarbons has made the concentration of
oxygen too low to support aerobic biodegradation
throughout most of the separation distance between
the source of vapors and the building.
Even with a relatively low contribution of
biodegradation, there are wide variations in the
estimated value of the attenuation factor (a) for various
values of the concentration of hydrocarbon in soil gas
and various values for the separation distance. The
risk of PVI is largely influenced by factors operating in
the soil and sediment and less so by factors related to
the building itself.
As mentioned previously, the predicted values of
the attenuation factor (a) that fall above the colored
shapes are not adequate to protect indoor air from
PVI when the concentration of benzene in soil gas is
the maximum concentration of benzene that could be
expected in soil air above gasoline. If a prediction
falls into this region of a chart, it is necessary to
compare the actual measured concentration of
benzene in soil gas to predict a. The actual measured
concentration of benzene will depend on the fuel that
was spilled, on the age of the spill and on the extent of
weathering of the fuel in the time since it was spilled.
Figures 3.14 and 3.15 will be used to screen sites in the
case studies presented in Sections 5, 6 and 8.
Figure 3.18 compares the distribution of concentrations
of benzene in all the samples of soil gas that were
evaluated in U.S. EPA (2013a). The concentration
in only a few percent of the samples approached
1.4E+07 (ig/m3, which is the plausible maximum
value for benzene in soil gas in contact with NAPL.
The predicted concentration of benzene in indoor
air will be sensitive to both the predicted value of
the attenuation factor (a) and the measured value of
benzene in soil gas. Section 4 and Section 7 discuss
methods to acquire samples of soil gas and determine
the concentration of TPH and benzene.
0.25 n
3.2 to 32 3.2 to 32 3.2 to 32 3.2 to 32 3.2 to 32 3.2 to 32 3.2 to 32
E+00 E+01 E+02 E+03 E+04 E+05 E+06
Concentration of Benzene in Soil Gas (jjg/m3)
Figure 3.18. Distribution of Concentrations of Benzene
in Soil Gas Samples that were evaluated in U.S.EPA
(2013a).
Concentrations of HC Compounds in Soil Gas at Source of C
GW Issue 29
-------
3.4.3 Including Methane in Total
Hydrocarbons
The screening approach using the simulations of Abreu
et al. (2009a) requires that methane be included in the
hydrocarbons. Usually, methane data are acquired by
a different analysis. If the data are reported in ppb
(volume/volume), multiply ppb by 0.662 to get
Hg/m3. Methane has relatively more hydrogen than
the higher alkanes and as a result it has a higher
theoretical oxygen demand on a mass basis. To correct
for the differences in theoretical oxygen demand.
multiply the measured concentration of methane by
1.14 before adding the concentration of methane to the
concentration of TPHg to calculate total hydrocarbons
in soil gas.
3.4.4 Application of Exposure Evaluation of
PVI to other Compounds
Abreu et al. (2009a,b) simulated the behavior of pure
benzene in soil gas to the behavior of a mixture of
hydrocarbons that represented a weathered gasoline.
The Abreu Three-Dimensional Model predicts
equivalent distributions of oxygen. The distribution of
other TEX compounds was similar to the distribution
of benzene; however, the distribution of aliphatic
hydrocarbons differed from the distribution of benzene
alone (Abreu et al., 2009b). One consequence is that
Figures 3.14 through 3.17 can be used to estimate an
attenuation factor (a) for other aromatic hydrocarbons,
but they should not be used for aliphatic hydrocarbons
such as hexane.
4.0 DATA REQUIREMENTS TO
FORECAST PETROLEUM VAPOR
INTRUSION
The Interstate Technology & Regulatory Council
has produced a Guidance Document that describes
a process for screening, investigating, and
managing sites for PVI (ITRC, 2014). The ITRC
Guidance Document has detailed specifications and
recommendations on methods to sample and analyze
soil gas.
To make a robust forecast of the contribution of
petroleum vapor intrusion to concentrations of fuel
components in indoor air, it is necessary to have
information on the properties of the building and
the properties of the subsurface environment. If
the approach uses the figures in Section 3.4, the
exposure assessment assumes generic properties for
the building. The exposure assessment is influenced
by the properties of the soil gas at the source of the
vapors, and the separation distance between the source
of vapors and the building. These properties are
described in this section.
If information on the properties of the building is
available, the exposure assessment can be refined using
either the BioVapor Model as described in Section
3.1 or the PVIScreen Model described in Section 3.2.
Consult the user's guides of the models for instructions
on the properties of the building being described. This
Issue Paper does not consider the properties of the
building.
4.1 Conversions of Concentrations from
Units of ppb-v to [*g/m3
Data on concentrations of organic compounds in gas
samples are conventionally reported on a volume basis
in units of parts per million by volume (ppm-v) or
parts per billion by volume (ppb-v). This is because
the working standards used to calibrate the instruments
are created by diluting standard gases. However, the
oxygen demand of a hydrocarbon in gas is related to
the chemical formula of the hydrocarbon, as well as its
concentration on a volume basis. If concentrations of
hydrocarbons are to be used to estimate their impact
on oxygen demand, they should be expressed in units
of (ig/m3. Standards for acceptable concentrations of
individual organic compounds in soil gas are often
expressed in units of (ig/m3.
Concentrations in ppb (v/v) can be converted to units
of (ig/m3 using the following formula:
= ppb * molecular weight * 0.0413
The formula applies for conditions at sea level and
room temperature.
If you are interested, the formula is derived as follows.
If you are not interested, skip to Section 4.2.
30 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
The ideal gas law states: Pi*V = Ui*R*T
where Pj is the absolute partial pressure of the individual hydrocarbon (i) in the gas, Vis the volume of the
sample of gas containing the hydrocarbon, «z- is the number of moles of the individual hydrocarbon in the
sample, R is the universal gas constant, and Tis the absolute temperature in degrees Kelvin[K],
The value of R is 0.08205 (L*atm)/(mole*T). The value of T [K] equals the temperature in degrees Celsius
plus 273. At room temperature T is near 295 K. The molecular weight of a compound is denned as the
number of grams (g) per mole («), so n = m/ molecular weight where m is the mass in grams. Substituting
m/molecular weight for n and rearranging for concentration as m/V, then
m
—. = p. * -molecular weight/(R * T)
At sea level, a gas present at a concentration of 1 ppb has a partial pressure of one billionth of the
atmospheric pressure (Patm>- The equation above becomes:
m _ ppb * molecular weight * Patm
~V~~ R*T
When Patm is 1.0 atmosphere:
Patm _ 1
R*T 0.08205*(273+22)
= 0.0413
4.2 Analysis of Compounds of Concern
To apply the approach to a particular building, it is
necessary to know the concentration of compounds
of concern (COCs) in the soil gas at the source of
the vapors. The separation distance between the
source of vapors and the building and the theoretical
oxygen demand of all the hydrocarbons in the soil
gas are used to estimate attenuation factors (a) for
the concentrations of COCs between the source of
vapors and indoor air in the building. To estimate the
concentration of the COCs in indoor air, the estimate
of a is multiplied by the measured concentration of
the COC in the soil gas. To complete the exposure
evaluation, the predicted indoor air concentration is
compared to some reference standard concentration
that is specified by the appropriate regulatory authority.
The most straightforward approach to determine
the concentration of a COC in soil gas is to use an
analytical procedure that would be used to measure
the concentrations of the compound in indoor air. This
ensures that the analyses will meet all the requirements
for data quality that are imposed by the regulatory
authority.
4.2.1 Individual Petroleum Hydrocarbons
The U.S. EPA identifies four methods that are
intended to determine the concentrations of individual
petroleum hydrocarbons in samples of ambient air
or indoor air. They are Method TO-15 (U.S. EPA,
1999a), Method TO-14A (U.S. EPA, 1999b), Method
TO-3 (U.S. EPA, 1984) and Method TO-17 (U.S. EPA,
1999c). These methods can also be applied to soil
gas. Because their maximum reporting concentration
is near 10,000 (ig/m3, it is often necessary to dilute
samples of soil gas before they can be analyzed.
Compendium Method TO-15 is widely used (U.S.
EPA, 1999a). Gas samples can be collected into an
evacuated stainless steel canister that has gone through
a special process to make the interior of the canister
chemically inert. After the gas sample is collected,
Concentrations of HC Compounds in Soil Gas at Source of C
GW Issue 31
-------
the canister is sealed and shipped to the laboratory for
analysis by high resolution gas chromatography using
a mass spectrometer as the detector.
Compendium Method TO-14A is an update of an
older method and is similar to Method TO-15. The
sample is collected into a stainless steel canister and it
is analyzed using gas chromatography, but the method
does not necessarily use a mass spectrometer as the
detector. Detectors can include a flame ionization
detector, an electron capture detector, or a photo-
ionization detector. A gas chromatograph with a
mass spectrometer detector is more expensive and
requires more training to operate. As a result, the
costs per sample for Compendium Method TO-15
may be greater than Method TO-14A. However, the
mass spectrometer detector can distinguish between
organic compounds that are not fully separated on
the chromatography column. It is less likely to
misidentify a compound.
Under routine conditions, Method TO-14A and TO-
15 can detect individual petroleum hydrocarbons at
concentrations above 0.5 ppb on a volume basis. This
detection limit corresponds to a benzene concentration
of 1.6 (ig/m3.
Method TO-3 uses a gas chromatograph with a flame
ionization detector or an electron capture detector,
or both detectors. The method does not specify a
container for the gas sample. It can be used with a
sample that is collected into a flexible plastic bag (such
as a Tedlar® Gas Sampling Bag).
Method TO-17 requires that the gas sample be
collected onto a special sorbent trap. The trap is then
shipped to the laboratory for analysis. Methods TO-
14A and TO-15 require approximately one liter or
six liters of sample to fill the steel container. Method
TO-17 requires as little as 50 mL. The smaller sample
can be collected from a vapor point with a plastic
syringe and loaded onto the trap in the field. This is a
significant advantage when the pneumatic conductivity
is low and it is difficult to acquire one or more liters of
soil gas.
Method TO-17 using a Tenax TAtrap allows for
analysis of high molecular weight hydrocarbons as
would be found in jet fuel or diesel fuel.
4.2.2 Fuel Oxygenates
Method TO-14A was developed for non-polar
compounds such as hydrocarbons. However, Method
TO-15 was developed to allow analysis of more polar
compounds such as the fuel oxygenates. Analytes
amendable to Method TO-15 include ethanol, methyl
fert-butyl ether, ethyl tert-butyl ether, fert-amyl methyl
ether and diisopropyl ether.
4.2.3 Lead Scavengers (EDB and 1,2-DCA)
Ethylene dibromide or EDB (1,2-dibromoethane)
and 1,2-DCA (1,2-dichloroethane) can be analyzed
using Method TO-15. However, the detection limit is
high compared to acceptable concentrations in indoor
air. A detection limit of 0.5 ppb on a volume basis
corresponds to a concentration of 3.9 (ig/m3 for EDB
and of 2.1 ng/m3 for 1,2-DCA.
According to the EPA Integrated Risk Assessment
Information System (U.S. EPA, 2014c), the acceptable
concentration of EDB in indoor air at the 1 in 10,000
risk level is 0.2 (ig/m3. The acceptable concentration
at the 1 in 1,000,000 risk level is 0.002 (ig/m3.
Because an attenuation factor will be applied to
concentrations in soil gas, Method TO-15 has adequate
sensitivity to measure the lead scavengers in soil gas.
However, to measure EDB in indoor air at adequate
sensitivity, it will probably be necessary to use Method
TO-14A with an electron capture detector.
4.2.4 Total Petroleum Hydrocarbons
in Soil Gas
An analysis of Total Petroleum Hydrocarbons is
generally represented as TPH. If the analysis only
considers the range of hydrocarbons that form the bulk
of gasoline, the analysis is termed TPH-g. Generally
TPH-g includes compounds with six to ten carbon
atoms. If the analysis only considers the range of
hydrocarbons that form the bulk of diesel fuel, the
analysis is termed TPH-d. Generally TPH-d includes
compounds with ten to twenty-eight carbon atoms.
Computer models such as BioVapor, PVIScreen or the
Abreu Three-Dimensional Model actually consider the
composite oxygen demand of the hydrocarbons. This
hypothetical parameter would be the equivalent of the
32 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
biological oxygen demand of waste water. Abreu et
al. (2009a) expresses the composite oxygen demand
as the concentration of benzene with an equivalent
demand. Figures 3.4 through 3.9 express the composite
demand as the concentration of CyH16 with an
equivalent demand. The molecular weight of C7H16
(100 g/mole) is near the mean molecular weight of
gasoline.
The computer models require a value for the composite
oxygen demand. BioVapor and PVIScreen will
calculate this value internally if the concentrations of
methane and all the individual petroleum hydrocarbons
in soil gas are provided. The models can also be
provided the Total Petroleum Hydrocarbons expressed
as an equivalent hydrocarbon. Figures 3.14 through
3.17 also require a value for the composite oxygen
demand expressed as the concentration of C7H16 with
the same oxygen demand.
Methods TO-3, TO-14A and TO-15 were designed for
individual compounds. However, some vendors also
offer an analysis of Total Petroleum Hydrocarbons
in the range expected in gasoline (TPHg). One such
vendor uses the following protocol:
TPH-g by EPA TO-15 is calibrated using
a single point gasoline calibration standard
analyzed with each analytical batch. The
TPH-g is determined in each sample by
summing the area of the total ion
chromatogram of the GC/MS run and
subtracting non-petroleum related
components from the total area. This total
area (approximately C3 to C12 range) and
the response factor of gasoline are used to
calculate the TPH result.
The TPH-g is expressed as the equivalent of a
hydrocarbon with a molecular weight of 100 Daltons.
The value of TPH-g can be entered in BioVapor or
PVISceen as the concentration of heptane. If a sample
is analyzed by Method TO-15, one sample of soil
gas can be used to determine both TPH-g and the
compounds of concern.
Methods TO-3, TO-14A, TO-15 and TO-17 were
designed to determine the concentrations of Toxic
Organic compounds of regulatory concern. They
are not calibrated for many of the compounds in
petroleum motor fuel that are relatively benign.
These compounds that are not included in Methods
TO-3, TO-14A, TO-15 and TO-17 include propane,
the butanes, the pentanes, and several heavier
hydrocarbons that are important components of soil
gas in contact with motor gasoline. As a result, it
is not possible to estimate TPH-g by adding up the
concentrations of the individual compounds in a TO-3,
TO-14A, TO-15 or TO-17 analysis.
4.2.5 Methane, Oxygen, Carbon Dioxide,
Nitrogen
Methods TO-3, TO-14A, TO-15 or TO-17 do not
provide an analysis for methane. It is necessary to
have an analysis for methane to correctly describe the
composite oxygen demand of the hydrocarbons in soil
gas (Abreu et al., 2009a; DeVaull, 2007; Jewell and
Wilson, 2011). Methane can be determined by EPA
Method 3C (U.S. EPA, 2014d) or by ASTM D-1945
(ASTM, 2010). The methods also determine the
concentrations of oxygen, carbon dioxide and nitrogen.
If a modification of ASTM D-1945 is used to
determine the concentrations of ethane, ethylene,
acetylene, propane, n-butane, /so-butane,
n-pentane, /'so-pentane, and a composite of C6 to
C7 hydrocarbons, it is possible to determine the
concentrations of the petroleum hydrocarbons and the
concentrations of methane, oxygen, carbon dioxide,
and nitrogen in the same analytical run.
As will be described in Section 4.5, the concentration
of oxygen in a sample of soil gas can be used to correct
for dilution or leakage as the sample was acquired.
The concentration of carbon dioxide can be used to
distinguish a sample of soil gas from a sample of the
atmosphere. The concentration of nitrogen can be used
to calculate the mass balance of all the permanent
gases in the sample and determine the accuracy of the
analyses for the other permanent gases.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 33
-------
4.3 Correcting Concentrations of
Contaminants for Leakage or Dilution
Figure 4.1 depicts two monitoring tools that can be
used to collect soil gas from the source of hydrocarbon
contamination. The device on the left is a typical
probe built for the purpose of collecting soil gas. The
device on the right is a conventional ground water
monitoring well. Many of the ground water wells at
fuel release sites were also designed to detect NAPL.
As a result, they have a portion of the screen above the
water table. In many respects, both the vapor probe
and ground water well are similar. They are installed
in a borehole. The sampling point is surrounded with
a sand pack. The sand pack and sampling point are
protected from contamination from the surface with a
seal of grout or bentonite clay. The point of extraction
of the sample is protected by a vault or cap.
Seal
Figure 4.1. Depiction of a vapor probe set near the
LNAPL source of vapors and a monitoring well screened
across the water table.
If the water table is within the screened interval, a
conventional monitoring well resembles a large vapor
probe. There are only three important differences.
The vapor probe has a much smaller diameter and
much less void volume than the screen and riser of
the ground water well, the vapor probe terminates
above the water table, and the vapor probe is generally
constructed of stainless steel or nylon tubing instead
of PVC plastic. Jewell and Wilson (2011) were able to
use ground water monitoring wells to sample soil gas
at a number of UST fuel spill sites in Oklahoma.
Any real sample is subject to dilution of the analytes
by leaks of air into the sampling train, and mixing of
the soil gas in the unsaturated zone. When soil gas is
sampled for chlorinated solvents, the general approach
to control for leaks is to use tracer compounds to
document any contribution of air from atmosphere
above the sampling point to the sample. If the tracer
compound is found in the sample above a certain
concentration, the sample is compromised by the leaks.
In addition to leaks, there is another process that can
dilute the sample. If the transition from oxygenated
soil gas to anoxic soil gas occurs in the screened
interval of the monitoring well, then the source
vapors will be diluted with air that has much lower
concentrations of benzene and TPH (Figure 4.2). The
flow paths with less contaminated soil gas will dilute
the gas in flow paths that are directly in contact with
the source of petroleum hydrocarbons. Leak testing
at the surface can only reveal the contribution of leaks
in the sampling train. It cannot evaluate the effect of
mixing in the soil profile.
If the casing of adjacent monitoring wells or vapor
probes are not properly sealed with a well cap, air can
enter through the screen of an adjacent monitoring well
or vapor probe, and dilute the soil gas in the vicinity of
the well or vapor probe that is being sampled.
Figure 4.2. Dilution of soil gas samples by dilution with
cleaner soil gas.
34 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
The approach taken in this Issue Paper is a screening
level evaluation. If screening indicates that there
is a reasonable possibility for PVI, then further
characterization or site remediation is required.
As a result, it is not necessary to know the exact
concentrations of benzene or TPH in soil gas at the
source. An upper boundary on the concentration is
useful for the evaluation.
Air in contact with LNAPL is often devoid of oxygen.
Davis et al. (2009) compared the concentrations of
hydrocarbons and oxygen in soil gas from seven field
sites in Australia. With very few exceptions, when
oxygen was present, hydrocarbons were absent and
oxygen was absent when hydrocarbons were present.
The conceptual models discussed in Section 2 assume
that the air at the source is devoid of oxygen. To put
a conservative boundary on the true concentration
of benzene, TPH and methane at the source of
contamination, we will assume that all the
benzene, TPH or methane came from the source of
contaminated vapors, which did not have oxygen.
We will assume that all the oxygen in the sample
came from a leak or came from clean air in the
unsaturated zone.
A simple factor can correct for the effects of leaks as
a sample is collected or dilution of the sample in the
soil gas. The correction factor is the concentration of
oxygen in the atmosphere divided by the difference
in the concentration oxygen between the atmosphere
and the sample of soil gas. If you are interested in
a derivation of the correction factor, the derivation
is provided below. If you are not interested, skip to
Figure 4.3. This figure plots the correction factor
for concentrations of benzene, TPH or methane from
the measured concentration of oxygen, assuming the
measured concentration of oxygen in the atmosphere
is 21%.
The correction factor is derived as follows. The mass of benzene, TPH or methane in the sample of gas
that was analyzed is the same mass of benzene, TPH or methane that was extracted from the soil gas in
contact with the NAPL. The concentration of benzene, TPH or methane in the source gas collected into
the sample and the concentration in the gas actually analyzed are related as follows:
Cone. Source Gas * Vol. Source Gas =
Cone. Gas Analyzed * Vol. Gas Analyzed
Cone. Source Gas _ Volume Gas Analyzed
Cone. Sample Analyzed Volume Source Gas
The fraction of the sample that was analyzed that is represented by the air from the vapor source is:
Volume Source Gas _ O-, in Atmosphere - O, in Soil Gas
Volume Gas Sampled O2 in Atmosphere
Rearranging, and solving for the concentration in the source gas:
Cone. Source Gas =
O7 in Atm.
Cone. Sample Analyzed -
, in Atm.-O, in Soil Gas
Concentrations of HC Compounds in Soil Gas at Source of C
GW Issue 35
-------
4 6 8 10 12 14 16
Oxygen in Soil Gas Sample (% Wv)
16
Figure 4.3. A correction factor for concentrations of
hydrocarbons in soil gas near the source of hydrocarbons
based on the concentration of oxygen.
When the concentration of oxygen is above 19% (v/v),
the correction based on oxygen becomes uncertain.
Do not apply the correction factor for concentrations
of oxygen above 19%. The correction factor can be
calculated for oxygen measured by ASTM D-1945
(ASTM, 2010), or it can be measured with a meter
in the field. If it is measured with a meter in the
field, determine the concentration of oxygen in the
atmosphere immediately before and immediately after
determining the concentration in the soil gas sample.
If the concentration of oxygen is measured in the
laboratory, collect a sample of the atmosphere and
include it for analysis.
Wilson et al. (2012b) reported the concentration of the
major of components of soil gas from three wells at a
fuel spill site in Antlers, Oklahoma (Table 4.1). See
Section 6 for details of the site.
Table 4.1. Major components of soil gas at a motor fuel spill site.
Nitrogen
Methane
Carbon Dioxide
Oxygen
Gasoline Hydrocarbons
Hydrogen
Sum
Ratio of Oxygen to Nitrogen
% (v/v)
% (v/v)
% (v/v)
% (v/v)
% (v/v)
% (v/v)
% (v/v)
Concentration
(Atmospheres)
78.0
0.0002
0.04
21.0
NA
0.00006
99.04
0.27
MW-2B
6.6
71.0
19.6
1.23
1.90
< 0.001
100.33
0.19
MW-9
19.9
61.2
18.4
1.02
0.16
< 0.001
100.68
0.05
GMW-1B
24.1
60.1
11.7
3.92
0.59
< 0.001
100.41
0.16
36 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
As described in Wilson et al. (2012b), the
concentrations of nitrogen, methane, carbon dioxide,
oxygen, gasoline hydrocarbons and hydrogen
were determined using an Agilent Micro 3000 gas
chromatograph (GC). The instrument is configured
with four miniaturized GC systems in a modular
format. All four modules contain a micro-Wheatstone
bridge thermal conductivity detector.
The concentrations of methane and carbon dioxide
were high compared to the atmosphere, and the
concentrations of nitrogen and oxygen were low. At
this site, much of the methane and carbon dioxide was
produced by fermentation of ethanol in the gasoline
that was spilled. Methane and carbon dioxide in the
ground water became super saturated, bubbled out of
the ground water, and displaced the nitrogen in the
soil gas. In monitoring wells MW-2B and GMW-1B,
the ratio of oxygen to nitrogen in the gas sample was
similar to the ratio in the atmosphere. This indicates
that most of the nitrogen and oxygen in the sample
came from a leak in the sampling process. In well
MW-9, oxygen was depleted with respect to nitrogen.
This indicates that most of the nitrogen in the sample
actually came from soil gas, and much of the oxygen
in the sample came from a leak.
Notice that the mass balance on the sum of the gases
was within a percent. Some vendors will measure
the other permanent gases and estimate nitrogen by
difference. If nitrogen is actually measured, then the
mass balance provides a check on the data quality of
the analyses.
If the concentration of oxygen in a sample is high,
this may indicate that a leak had a major impact on
the concentrations of hydrocarbons in the soil gas. It
may also indicate that the rate of biodegradation of
the hydrocarbons may have been too slow to consume
much of the oxygen in the soil gas. The concentration
of carbon dioxide can distinguish between these
two possibilities. The background concentration
of carbon dioxide in the atmosphere is near 0.04%.
The background concentration of carbon dioxide in
normal soil gas is near 3% and the concentration in
contaminated soil gas can be even higher.
4.4 Field Screening of Vapor Samples to
Determine When to Sample
Chemical analyses are a significant cost of site
characterization. The U.S. EPA data quality process
requires that a sample be representative of the
environmental medium being sampled. This issue
is particularly important when the volume of the
sample submitted for analysis is small with respect
to the internal volume in the sampling instrument.
This is the reason that ground water monitoring wells
are often purged before water samples are taken for
analysis. There are two approaches for collecting a
representative groundwater sample from a well. The
well can be purged for some specified volume, such
as three casing volumes, or the well can be purged
until sensitive parameters, such as the concentration of
dissolved oxygen or the oxidation/reduction potential,
come to equilibrium.
Jewell and Wilson (2011) sampled soil gas from
conventional ground water monitoring wells. They
purged the gas from the well at a flow rate of 10
liters per minute for a minimum of twenty minutes
before they took a sample. This portion of the Issue
Paper discusses the second approach: monitoring
concentrations of hydrocarbons, oxygen and carbon
dioxide in soil gas until they come to equilibrium.
This discussion does not apply to vapor probes, which
usually have a small internal volume. The internal
volume of a vapor probe may be 0.1 liter, while the
volume of the sample is 1.0 or 6.0 liter. McAlary et al.
(2009) offer useful suggestions for collecting samples
from vapor probes, particularly when the pneumatic
conductivity of the geological material is low. They
recommend purging one internal volume, and then
collecting the sample using conditions that impose a
vacuum of no more than 100 inches of water on the
soil gas.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 37
-------
4.4.7 Oxygen and Carbon Dioxide to Monitor
Stabilization of Samples
A variety of field instruments are available to
determine the concentrations of oxygen and carbon
dioxide in gas. Field meters typically report
concentrations of oxygen and carbon dioxide with
increments of 0.1%. The concentration of carbon
dioxide in the atmosphere (0.04%) is below the typical
reporting limit of the meter.
The oxygen detectors sense an electrical current
that is produced by a chemical reaction between the
hydrocarbon and oxygen in the gas. The current is
proportional to the concentration of hydrocarbon.
Material in the detector is consumed as they operate.
As a result, the detectors have a shelf life. When the
detectors are no longer within specifications, they
must be replaced. The carbon dioxide detector uses a
spectrophotometer to measure the infrared absorbance
of the gas at a wavelength characteristic of carbon
dioxide. The carbon dioxide detectors do not degrade
overtime.
Generally, the oxygen detectors and the carbon
dioxide detectors are not subject to interference from
other compounds. To our knowledge, there are no
compounds that might reasonably be expected in soil
gas at gasoline spill sites that may produce a false
reading for oxygen or carbon dioxide, or that might
degrade the performance of the detector. However,
it is always good to check the user's guide and the
manufacture's literature for potential interferences.
4.4.2 Stabilization of Oxygen and Carbon
Dioxide in Soil Gas Samples from Water
Monitoring Wells
Soil gas samples that are acquired from groundwater
monitoring wells behave much like ground water
samples. The concentrations can change as the soil
gas is sampled. These concerns do not apply to small
samples of soil gas taken from vapor probes. The
following discussion applies only to soil gas samples
that are acquired from groundwater monitoring wells.
During the course of sampling, the concentrations
of oxygen typically decrease over time and the
concentrations of carbon dioxide typically increase.
The time taken for concentrations of oxygen and
carbon dioxide to equilibrate depends on the rate of
extraction of air from the monitoring well and the
internal volume of the well, and other factors that have
not been defined.
Figure 4.4 and Figure 4.5 are examples of a well
that equilibrated quickly and a well that equilibrated
more slowly, relative to the total purge volume. To
allow direct comparisons, the volume of air that was
extracted was normalized to the total internal gas
volume in the well. The total volume includes the
sampling train and an estimate of the gas filled volume
of the sand pack. Extraction of air is reported in purge
volumes removed. Data for an additional eight sites
are provided in Appendix A.
The data used in these case studies 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
or the Montana Department of Environmental
Quality.
The concentrations of oxygen and carbon dioxide
in Well M-2 at a site in Antlers, OK (Figure 4.4)
equilibrated in as little as 1.4 purge volumes. In
contrast, oxygen and carbon dioxide in Well MW-4 at
a site in Helena, MT (Figure 4.5) required 14 purge
volumes to equilibrate. This order of magnitude
discrepancy in the kinetics of equilibration is a good
argument for the use of field meters to recognize stable
conditions before samples are taken for laboratory
analysis.
38 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
1E+09
1E+08
1E+Q7
1E+06
1E+05
QL
1E+D4
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
•> 20
O
I
ra
8
15
10
:
n Total
Hydrocarbons
Benzene
100
90
BO
70
60
50
40
30
20
10
0
0.0 0.2 0.4 0.6 0.8 1.0
Purge Vol (-)
1.2 1.4
•C02
-02
-CH4
Figure 4.4. Kinetics of equilibration of oxygen,
carbon dioxide, methane and benzene in soil gas
from well MW-2 at the EZ Go service station in
Antlers, Oklahoma.
1E+09
1E+08
„ 1E+07
1E+06
1E+05
QL
1E+04
I
20
15
10
0 2 4 6 8 10 12 14 16
4—i—I—i—I—i—I—i—I—i—I—i—I—i—H
10 min 20 min 30 min 40 min
_-fl B ° "
••? Total
Hydrocarbons
- Benzene
0.8
0.6 |
*
0.4 I*
0,2
-- C02
— O2
— CH4
00
0 2 4 6 8 10 12 14 16
Purge Vol (-)
Figure 4.5. Kinetics of equilibration of oxygen,
carbon dioxide, methane and benzene in soil gas
from well MW-4 at the Kev's Auto site in
Helena, Montana.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 39
-------
4.4.3 Stabilization of Methane and Petroleum
Hydrocarbons in Soil Gas Samples from Water
Monitoring Wells
In the field studies described in Figures 4.4 and 4.5 and
Appendix A, the true equilibration of the wells was
determined by measuring the concentrations of total
hydrocarbons and the concentrations of benzene over
time in the soil gas pumped from the well. Samples
were collected after 1, 10, 20, 30 and 40 minutes
of pumping. Petroleum hydrocarbons and methane
were determined using a modification of ASTM
D-1945 as described by Jewell and Wilson (2011).
Total hydrocarbons concentration was calculated
from the concentrations of all the individual gasoline
hydrocarbons and methane. The concentrations of
benzene and total hydrocarbons stabilized as the
concentrations of oxygen and carbon dioxide stabilized
(Figure 4.4 and 4.5). Data for an additional eight sites
are provided in Appendix A.
At five of the ten sites, the concentrations of benzene
and total hydrocarbons increased as the well stabilized.
At four of the ten sites, the concentrations decreased as
the well stabilized. At one site there was no significant
change in the concentrations of benzene or total
hydrocarbons as the well stabilized.
At many of the wells, the kinetics of stabilization
of concentrations of benzene or total hydrocarbons
in air samples was more rapid than the kinetics of
stabilization of the meter readings for oxygen or
carbon dioxide. The air samples were collected from a
common outlet from the sampling train that also served
as the inlet to the field meter. The difference in the rate
of equilibration of the field meter and the air samples
can only be explained as hysteresis in the readings
from the field meter. If field meters for oxygen and
carbon dioxide have stabilized, it is reasonable to
conclude that the concentrations of benzene and
petroleum hydrocarbons have stabilized.
4.5 Vertical Separation Distance between
Source and Receptor
Figure 4.6 depicts the relationship between the bottom
of a building that is the receptor of hydrocarbon vapors
and the source of the vapors in the subsurface (taken
from Figure 6 in U.S. EPA, 2013b).
U.S. EPA recommends a minimum separation distance
from LNAPL in the unsaturated zone of 15 feet and
a minimum separation distance from contamination
dissolved in ground water of six feet.
The depth of the unsaturated zone changes with the
elevation of the water table. Examine the center panel
of Figure 4.6. When the water table is lower than the
top of the LNAPL and the NAPL is exposed to the soil
gas, the concentrations of hydrocarbons in soil gas will
be the concentrations that would be expected for soil
gas in contact with NAPL.
If the concentration of NAPL has reached residual
saturation, the NAPL will not rise in elevation with
a rise in the water table. Examine the bottom panel
of Figure 4.6. Occasionally, the water table at a
site may be high enough to inundate the residual
NAPL. When that is the case, the soil gas will not
be in direct contact with LNAPL. There will be less
transfer of hydrocarbons to soil gas and the measured
concentrations of hydrocarbons and compounds of
concern in soil gas will be typical of the much lower
concentrations that are seen above contaminated
ground water. When the water table is high, it is
also likely that NAPL would not accumulate in
a monitoring well. When the water table drops
and again exposes the NAPL to the soil gas, the
concentrations of hydrocarbons in soil gas return to the
higher concentrations expected for soil gas in contact
with NAPL.
If the site characterization is done when the water table
has inundated the NAPL, there is a danger of applying
the criterion for separation from contaminated ground
water, when the criterion for separation from NAPL is
more appropriate.
To appropriately apply the U.S. EPA minimum
screening distance for NAPL, it is necessary to know
the distribution of NAPL to a depth of 15 feet below
the bottom of the receptor, regardless of the position
of the water table. Do not apply a minimum separation
distance of six feet above contaminated ground water
unless core samples have been acquired that extend
at least 15 feet below the bottom of the receptor. If
NAPL is present in the first 15 feet of core samples,
then the vertical separation distance to hydrocarbon
contamination is less than the recommended vertical
40 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Vadose Zone
Recommended Vertical
Separation Distance from Purely
Dissolved-Phase Hydrocarbons
in Ground Water _
Water F3
Table
Vadose Zone
Vadose Zone
D D
Recommended Vertical
15 ft Separation Distance from NAPL
in Ground Water or the
Unsaturated Zone.
Recommended Vertical
15 ft Separation Distance from NAPL
in Ground Water or the
Unsaturated Zone.
Y
Figure 4.6. Recommended vertical separation between sources of hydrocarbon vapors and a building.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 41
-------
separation distance. See Wilson et al. (2012a) for
recommendations on collecting and analyzing core
samples to determine the vertical separation distance.
The driller's log for a vapor probe or ground water
monitoring well is often a useful source of information
to infer a separation distance. Most logs provide
data on field screening of core samples for NAPL.
Subsamples of the cores are enclosed in a plastic
bag, and the air inside the bag is allowed to come
to equilibrium with the NAPL in the core sample.
Then the concentration of hydrocarbons in the air is
determined with a field organic vapor analysis (OVA)
meter. The meters can use a photoionization detector
(PID), a flame ionization detector (FID) or a catalytic
combustion cell (explosimeter). Meters that use a PID
detector are sometimes called an organic vapor meter
(OVM).
Figure 4.7 provides example data from a spill of motor
gasoline into a sandy aquifer at a site near Madison.
Wisconsin. The field screening was conducted with
a Scott TLV Sniffer OVA meter that was calibrated
to hexane (Scott Instruments, Exton, PA). U.S.
EPA (2013b), as referenced in Wilson et al. (2012a),
defines clean sediment for the purposes of screening
for PVI as having less than 250 mg/kg TPH. Wilson
et al. (2012a) recommends that any core sample that
screens with more than 100 ppm organic vapors be
analyzed for TPH. In the sediment depicted in Figure
4.7, whenever the OVA readings were less than 100
ppm, the TPH was less than 250 mg/kg.
The data used in this 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
Wisconsin Department of Natural
Resources.
250
TPH (mg/kg)
750 1000 1250 1500 1750 2000
12
500 1000 1500 2000 2500 3000 3500 4000
Organic Vapor Analysis (ppm)
Figure 4.7. Relationship between the
vertical distribution of TPH and field
screening of core samples with an
organic vapor analyzer (OVA).
42 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
Another, even more conservative criterion, is the
concentration of TPH in soil gas where the TPH
increases conspicuously above the background
concentration. The depth interval where there is sharp
increase in the OVA meter readings can be considered
the first depth interval that has evidence of NAPL.
The sample just above the increase is the deepest depth
interval where there is no evidence of NAPL. The
separation distance extends from the bottom of the
foundation of the building to the deepest interval with
no evidence of NAPL.
Compare Figure 4.8. This was a driller's log for vapor
probes that were installed below the basement of
the Oasis Hotel at the Former Hal's Chevron Site in
Green River, Utah. Field screening revealed that OVA
readings at depths of 3, 4, 5 and 6 feet below the
basement were 5.5, 6.7, 8.6, and 8.2 ppm using a PID
detector. At a depth of 7 feet, the OVA reading was
191 ppm. The field log indicates "smear streaks at 7
feet". The true separation distance is greater than 6
feet and less than 7 feet. A separation distance of 6
feet would be appropriate for evaluation of PVI using
BioVapor or PVIScreen or by using Figures 3.14
through 3.17 in this Issue Paper.
The data used in this 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
Utah Department of Environmental
Quality.
Jfttfr
^jjj
S E C 0 R
SECOR International Inc.
308 E. 4500 S. Suite 100
Murray, Utah 84107
Project NO.: 26CH.3430000 i_Og of Borehole: VW-4
Site: Hal's Chevron Logged By.1 T. Clark
Address: Green River, Utah Client: ChevronTexaco
Drilling Contractor: EARTHPROBE Top of Casing Elev.:
Method/Equipment: Hand Auger Surface Elev.:
Date Drilled: 10/08/03 GrounOwater Elevation:
Well Completed: 10/08/03 Total Depth: 7.25'
SUBSURFACE PROFILE
|
*n-
•
•
-
s-
-
10-
,
t
0
JO
I
;> **
9
%$%
^^
1
<
Lithobgic Description .».•/
r^Ar
Ground Surface /"
^CONCRETE
Brown clayey SILT
Reddish brown SILT with dayey SILT
beds
Smear streaks ai T
PID from inside hole 0.2 ppm
SAMPLE
METHOD
RECOVERY
MOISTURE
Moist
O
1
ML
SAMPLE
BLOWS
6"/6-/6"
INTERVAL
PID (ppm)
Well Completion Details
Well Diameter: 2.0"
Borehole Diameter: 3.5"
WEIL VAULT "\^
5.5
,;;
fB.2
S 101
codcsere— -^ fai fe ^
MT.VArro {r&Wj, 'WK
BCTTOMTE mm ^
HZTPttf jj H
SrwiLBSSlrraL "" .tf- >'•'. -•V:
1W»S*JCASWO— ^^ "tol
BENTOMTE
OZTPW.T ^
ETHnBETUflHO ^ EgMi ^
*e&M^t' ~- j-si: --:;|
Figure 4.8. An example of a driller's log with information on the vertical extent of NAPL.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 43
-------
As a second line of evidence in an exposure
assessment, the approach in this Issue Paper might be
applied to a site that is not in the U.S. EPA vertical
inclusion zone (U.S. EPA, 2013b; Wilson et al.,
2012a). In this situation, the separation distance
between NAPL hydrocarbons and the building will be
greater than 15 feet. An example is provided in Figure
4.9. The site is the former Noon's Store in Helena,
Montana. Core samples were screened every five
feet to a total depth of 50 feet. To a depth of 30 feet,
there is no evidence that the TPH would be expected to
exceed 250 mg/kg. The reasonable lower boundary on
the separation distance is 30 feet.
The data used in these case studies 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
Montana Department of Environmental Quality.
The driller's logs that are depicted in Figures 4.8
and 4.9 have three features in common. The driller
attempted to collect a continuous core from the surface
to several feet below the water table. Cores were
acquired across the entire vertical interval, not just the
material at the water table and five feet above the water
table. Sub-cores were screened for NAPL across the
entire vertical interval. Starting with the most shallow
depth interval, several of the depth intervals that were
screened had undetectable or trivial concentrations
of hydrocarbons in air in equilibrium with the core
samples.
SVE-SE
Separation
Distance
30 feet
-1
SVE-2
PID 0-200 ppm
Hi-Noon Petroleum-Noon's #43)
Boring SVE-2
Clqr
S.ndyCl.y
BMl
Figure 4.9. A second example of a driller's log with information on the vertical extent of NAPL.
44 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Compare Figure 4.10. This log was taken at a former
Gasmat Site in Helena, Montana. The first sample
for screening was taken at a depth of five feet. The
screening indicated a potential for NAPL at this
depth. The first depth interval that was screened had
an unacceptable PID reading. There is no information
that can be used to interpret a separation distance.
It would not be appropriate to apply the approach in
this Issue Paper to soil gas data from this location.
In situations where it is impossible to interpret a
separation distance, it is more appropriate to apply the
generic attenuation factor for shallow soil gas as is
provided in guidance from the U.S. EPA or from the
appropriate state regulatory agency.
GROUNDWATER WELL INSTALLATION REPORT
Separation
Distance
<5 feet
Top of
Screen
6.8 feet
Project
Project No,
Method o' irs:-i'c'.ion
Installed
Well No..
localionjj.
-**)•*•)** Tirr.g. 09DO
Figure 4.10. A third example of a driller's log with information on the vertical extent of NAPL.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 45
-------
5.0 APPLICATION OF THE APPROACH
TO A SITE WITH BASEMENTS
(GREEN RIVER, UT)
The data used in this 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 Petroleum Storage Tank Trust Fund
of the Utah Department of Environmental Quality.
A spill of motor fuel from underground storage tanks
at the former Hal's Chevron in Green River, UT
created a zone of LNAPL that extended underneath a
motel and cafe that adjoined the service station. The
motel has an office with a basement and the cafe has a
basement. See Figure 5.1.
Figure 5.1. Aerial photograph of the Hal's Chevron Site in
Green River, Utah. The photograph was downloaded from
Google Maps®.
The Utah Department of Environmental Quality
(DEQ) installed a series of monitoring wells to define
the extent of contaminated ground water and the extent
of LNAPL underneath the motel and cafe. They also
installed sets of multi-depth vapor monitoring points.
Figure 5.2 compares the location of the monitoring
points and buildings to the known extent of NAPL at
the site. The extent of NAPL was defined by Robin
Davis and other staff of the Utah DEQ. They used a
large number of wells and core sampling points that
are not depicted in Figure 5.2.
The Utah DEQ installed multi-depth vapor monitoring
points in the basement of the motel office (VW-4 in
Figure 5.2) and in the basement of the cafe (VW-5).
All the other monitoring points were installed outside
of the buildings. The vapor monitoring points had an
inner diameter of 0.17 inch. The ground water wells
had an inner diameter of 2 inches.
The release was in a series of silty fine sands, silts,
clayey silts and silty clay (Figure 5.3). The monitoring
wells were screened across the water table and
the depth interval with NAPL. The deepest vapor
monitoring point in each cluster was screened in the
depth interval with NAPL.
5.1 Determining the Separation Distance
from the Source to the Basement of the
Cafe
Vapor point VW-7 is just south of the cafe (Figure
5.2). A series of core samples were acquired next
to VW-7. A paste sampler was used to acquire a
10 cm3 sub sample at various depth intervals along
the core samples. The sub samples were extracted
into methanol and analyzed for TPH. The peak
concentration of TPH was at a depth of 14 feet (Figure
5.3). From depths of three feet to 12 feet, TPH was
<20 mg/kg. The floor of the basement in the cafe was
eight feet below land surface. Based on the TPH in
core samples, the separation distance was at least
12-8 = 4 feet, but not greater than 14-8 = 6 feet.
Figure 5.4 is the driller's log for the vapor monitoring
points below the basement of the cafe. To a depth of
seven feet below the basement floor or 15 feet below
land surface, the PID screening did not indicate the
presence of NAPL. Allowing 0.5 feet for the depth of
the concrete in the floor, the separation distance is at
least 7 - 0.5 = 6.5 feet.
The two estimates of the separation distance are in
acceptable agreement. A separation distance of 6.5
feet will be used for purposes of screening. Because
BioVapor and PVIScreen allow input of the separation
46 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
distance, a value equivalent to 6.5 feet was used in
the models. The figures based on the Abreu Three-
Dimensional Model plot the separation distance in
meters. A distance of 6.5 feet is 1.98 meters. The line
corresponding to a separation distance of 2 meters will
be used to estimate a value of the attenuation factor.
Edge of LNAPL
>0.01 ft thick in well
Motel with
Basement
North
50 Feet
(M) Conventional Ground Water Monitoring Well
© Multi-Depth Vapor Monitoring Point
Figure 5.2. Relationship
between the extent of
NAPL, the location of
receptors and the location
of monitoring points at the
Hal's Chevron Site.
f
-------
>?
NV
S E C O R
SECOR International Inc..
,' 308 E. 45DU £. Suite- UK:
Murray. Ulsr, 84107
Project No.: 26CH.3430001
Site: Hal's Chevron
Address: Green River. Utal.
D/illing Contractor: EAnTHPROBE
Method/Equipment: Ksntl Auiicr
Date Drilled: 10/OB/C?
Well Completed: W!0b&:-
[_og of Borehole: VW-5
Client: ChevronTetaco
Top of Casing Elev.:
Surface Elev.:
Groundttiater Elevation
Iota/ Depth: 7,25'
SUBSURFACE PROFILE
SAMPLE
(1-
— -
5-
-
in-
n
1
-' 1
&
&
& ,yjty ^^
\ \wfyr
•^TfTiiTTI Ji-^rtaii.
-" CONCRETE
Brown olsypy SILT
j
Reddish brown SILT
1
IPID from tnside hole at completion
t.1 ppm
ujG
-i O
a T
r* f~
<£
n £
>-
I
0
u
UJ
cr
in
K
C
• •••
O
ML
Moist
ML
w,
^ i£
Ob
(E 10
<
**
IT
uJ
r-
t
c^
0
S
-UMNO"
6.5
2.6
13.2
Well Completion Details
Well Diameter 0.25'
Boring Diameler 3.E'
Figure 5.4. Driller's Log for vapor monitoring points at location VW-5 in the basement of the cafe.
5.2 Forecast of the Indoor Air
Concentration in the Basement
of the Cafe
Concentrations of benzene and petroleum
hydrocarbons in the vapor monitoring points were
determined by a modification of TO-15(U.S. EPA,
1999a). Methane was determined by a modification
of ASTM D-1945 (ASTM, 2010). Soil gas from the
ground water monitoring wells was collected and
analyzed as described in Jewell and Wilson (2011).
The primary data used to calculate the concentration
of source vapors are presented in Table 5.1. At this
site, the concentration of methane in soil gas was low,
and methane made no appreciable contribution to
the total hydrocarbons. The data from conventional
monitoring wells in Table 5.1 and subsequent tables
are corrected for leaks and dilution using the reported
concentration of oxygen, as described in Section 4.3.
The sample from the vapor probes comprised a little
more than one liter soil gas. Because the samples
were so small, they can effectively be considered point
samples. There should be no significant dilution of the
sample as the sample is collected, and no correction
was applied to the vapor probe samples.
When the correction for dilution was applied to the
sample from the monitoring well, the concentrations
were in reasonable agreement between monitoring
points VW-7 and MW-47.
The total hydrocarbons in the vapor monitoring point
and the soil gas from the ground water well are plotted
as vertical red lines in Figure 5.5. Values are plotted
for monitoring point VW-7 without correction and
MW-47 with correction. Figure 5.5 reprints Figure
3.15 showing only the 2 meter line.
Source Vapor Concentrations are plotted to one
significant figure. To be conservative, a figure was
used that was based on a first order rate constant for
biodegradation of 0.079 per hour.
48 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Applying the correction made no difference in the
prediction. The concentrations of total hydrocarbons
in the two monitoring points predict a value of the
attenuation factor (a) of 8E-05.
The concentration of benzene in indoor air was
estimated by multiplying the attenuation factor (a) by
the measured concentration of benzene in the soil gas
in VW-7 and the corrected concentration of benzene in
MW-47. Results are presented in Table 5.2.
u
ra
e
o
I
1.0E-02 i
1 OE-04 i
1.0E-05 ,
1.0E-06
1 OE 07
1.0E-08
1.0E-09
1 OE-10 -
A = 0.079 per hour
Basement
^^-—
-•-Separation 2 m
Figure 5.5. Estimates of the
attenuation factor in soil gas
below the cafe based on
concentrations of vapor at the
source.
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane (MS/™3)
Table 5.1. Concentrations of TPH-g and methane in soil gas near the cafe, and the calculated total concentrations of
hydrocarbons in soil gas.
Monitoring Point
VW-7
MW-47
Oxygen
%
Not Measured
17.2
Correction Factor
No Correction
5.65
TPH-g
ug/m3
3.10E+06*
1.02E+07
Methane
ug/m3
9.30E+03
5.54E+03
Total Hydrocarbons
ug/m3 as C7H16
3.1E+06
1.0E+07
* Values are in scientific notation. The notation 3.10E+06 means 3.10 * 10+6.
Table 5.2. Estimated concentration of benzene in indoor air in the basement of the cafe based on the Abreu
Three-Dimensional Model, on BioVapor and on PVIScreen.
Location
VW-7
MW-47
Benzene
ug/m3
4.6E+03
8.5E+04
a
Abreu
Figure
8E-05
8E-05
Estimated Concentration of Benzene in Indoor Air*
Abreu
Figure
ug/m3
4E-01
7E+00
Generic
BioVapor
ug/m3
3E-01
7E+00
Site Specific
BioVapor
ug/m3
8E-07
2E-05
PVIScreen
Most Probable Result
ug/m3
2E-04
4E-03
' The Utah Indoor Air Guidance (commercial) for benzene is 5E-01 ug/m3.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 49
-------
The concentration of benzene in indoor air was
also predicted using the BioVapor model set up two
different ways. The Generic BioVapor simulations
were set up using the Residential Default Values on
the Environmental Factors input screen with the
following expectations. Airflow under Foundation
(Qf) was set to be equal to Air Flow Through
Basement Foundation (Qs). The value was 83 cm3 of
air per second. To be consistent with the assumptions
of the Abreu Three-Dimensional Model, the Air
Exchange Rate was set to 12 per day. The default rate
constant for biodegradation of benzene in BioVapor is
0.79 per hour.
The rate constant in the Chemical Database (row
14, column U) was changed to 0.079 per hour. These
parameters were altered from the default values to
produce simulations that were conservative forecasts
of the expected concentrations of benzene in indoor
air. Simulations made with these values for the
parameters are called Generic Biovapor simulations.
See Figure 5.3. The water content of core samples
in the interval from 2 feet to 10 feet below grade
averaged 0.20 cm3 water/cm3 soil. This is the
interval that supported aerobic biodegradation of the
hydrocarbon vapors. This water content is four times
higher than the default assumption. The average mass
fraction of organic carbon in the interval from 2 feet
to 10 feet below grade was 0.006 g organic carbon/ g
soil. The Site specific Biovapor simulations used the
average quantity of soil water and organic carbon in
core samples from the site.
There was no practical difference in the estimated
concentration of benzene in indoor air based on the
Abreu Three-Dimensional Model and the estimate
provided by the Generic BioVapor simulation (Table
5.2). Both of these estimates are conservative
estimates. The true attenuation was most likely higher,
which will produce lower concentrations of benzene in
indoor air.
A higher content of water allows more living space
for bacteria that degrade the hydrocarbons. The
Site Specific BioVapor simulation had four times
the quantity of water in the profile. At a given rate
constant for biodegradation, the modeled rate of
degradation is four times faster. At the same time,
the space available for diffusion of hydrocarbons is
reduced 0.33 cm3/cm3 to 0.18 cm3/cm3. The reduction
in concentrations in indoor air vary as the exponent
of the first order rate constant for degradation. The
estimate of the concentration of benzene in indoor
air that was provided by Site Specific BioVapor
simulation was five orders of magnitude lower than the
Generic BioVapor simulation (Table 5.2).
The PVIScreen model was set up using one site
specific value (Table 3.1 in Section 3). The moisture
content was set at 0.20. The Most Probable Result
from the PVISceen simulations was intermediate
between the predictions of the Site Specific BioVapor
simulation and the Generic BioVapor simulation (Table
5.2).
The distribution of the PVIScreen simulations is
presented in Table 5.3. For vapor well VW-7, the Most
Probable Result was 2E-04 (ig/m3, a concentration
which was exceeded by 12.2% of the simulations.
However, none of the 1000 simulations exceeded the
acceptable cancer risk level of 0.52 (ig/m3. Likewise,
no simulations exceeded the hazard quotient of 1.0
(at 30 (ig/m3). The indoor air results for monitoring
well MW-47 were similar, but the Most Probable
Result was an order of magnitude higher. Even so, no
simulations exceeded either the cancer or non-cancer
risk levels.
5.3 Validation of the Forecast for the
Basement of the Cafe
The Utah Indoor Air Guidance (commercial) for
benzene is 0.5 (ig/m3. Because the estimates are near
this value, it is necessary to further characterize the
exposure. In the normal course of events, the next step
would be to collect shallow soil gas near or below the
foundation. The Utah DEQ anticipated the need for
this data, and collected soil gas from the shallow and
deep monitoring points on five dates extending from
October 2003 to September 2009 (Figure 5.6). The
samples described in Tables 5.1 and 5.2 were collected
in September, 2011.
The soil gas had adequate concentrations of oxygen to
support aerobic biodegradation. The concentrations
of benzene steadily declined overtime in the deep soil
gas samples. The concentrations of benzene in the
50 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 5.3. Distribution of the simulations in PVIScreen of the concentration of benzene in indoor air in the cafe.
Location
VW-7
MW-47
PVI Screen Most
Probable Result
Hg/m3
2E-04
4E-03
% Exceeding
Most Probable
Result
12.2
2.0
Acceptable Cancer Risk
(0.52 iig/m3)
0.000
0.000
Acceptable Non-Cancer Hazard
at Hazard Quotient =1.0
(30 jig/m3)
0.000
0.000
shallow soil gas samples also declined over time, but
the variability was much greater.
The highest concentration of benzene at a depth
of three feet below the basement was 140 (ig/m3.
Following the guidance in Table 1.2, and allowing for
a generic attenuation factor from shallow soil gas into
the indoor air of 0.1, this highest concentration would
produce indoor air concentrations of 14 (ig/m3. On
the other sampling dates, the indoor air concentrations
would have been no more than 2.2 (ig/m3. On the last
sampling date, the predicted indoor air concentration
would be 0.15 (ig/m3.
The indoor air concentration predicted from shallow
soil gas on the last sampling date (0.15 (ig/m3) was
lower than the predictions from the simulations of
the Abreu Three Dimensional Model or the Generic
BioVapor Model (7E+00 (ig/m3, see Table 5.2).
VW-5
Basement of Oasis
Cafe
3 feet below floor
11 feet below land surface
7 feet below floor
15 feet below land surface
10/03 11/03
4/06
8/06
9/09
140
Benzene (ug/m3) - O2 (% v/v)
3.0 7.4(17%) 22(18%) 1.5(17)
1000 <35,000 190(9,6%) 87(11%) 4.8(20%)
Figure 5.6. Concentrations of benzene and oxygen in soil gas beneath the cafe.
Original graphic created by Robin Davis, Utah DEQ.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 51
-------
The prediction made from shallow soil gas (0.15
(ig/m3) was greater than the predictions from the Site
Specific Bio Vapor model (2E-05 (ig/m3) or PVIScreen
(4E-03 (ig/m3).
The simulations of the Abreu Three Dimensional
Model or the Generic BioVapor Model were the
best match to the prediction from shallow soil gas;
however, they were roughly an order of magnitude
higher than the prediction from shallow soil gas. The
prediction based on shallow soil gas did not include
any reduction in concentrations due to biodegradation.
Compared to the prediction based on concentration in
shallow soil gas and an attenuation factor of 0.10, the
simulations of the Abreu Three Dimensional Model or
the Generic BioVapor Model were conservative and
over-predicted the concentration in indoor air.
5.4 Determining the Separation Distance
from the Source to the Basement of the
Office
The same approach will be applied to one other
location at the Hal's Chevron site. The office of
the motel was immediately adjacent to the tank
pits for the underground storage tanks (Figure 5.2).
The concentrations of petroleum hydrocarbons and
methane in the soil gas are provided in Table 5.4.
When the correction for dilution was applied to
the ground water monitoring wells, the corrected
concentrations were in reasonable agreement between
monitoring points.
Examine Figure 5.7. Based on field screening with a
PID detector, there is no evidence of petroleum NAPL
contamination to a depth of six feet below the floor
of the basement. Allowing 0.5 foot for the concrete
floor, the separation distance is 5.5 feet. The closest
equivalent in the figures based on Abreu et al. (2009a)
is a separation distance of 2 meters.
Figure 5.8 reprints Figure 3.15 showing only the 2
meter line. The calculated total concentrations of
hydrocarbons from Table 5.3 are plotted in Figure
5.8. They predict a value of the attenuation factor
(a) of 4E-04 for MW-2; a factor of 6E-04 for VW-1
and MW-51; and a factor of 8E-04 for VW-2 and
VW-3. These attenuation factors are multiplied by
the measured concentration in soil gas to predict the
measured concentration in indoor air (Table 5.5).
Table 5.4. Concentrations of TPH-g and methane in soil adjacent to the office of the motel, and the calculated total
concentrations of hydrocarbons in soil gas.
Monitoring Point
VW-1
VW-2
MW-2
MW-51
VW-3
Oxygen
%
Not Measured
Not Measured
1.36
18.2
Not Measured
Correction Factor
No Correction
No Correction
1.07
7.74
No Correction
TPH-g
ug/m3
2.2E+08
2.8E+08
1.3E+08
1.5E+08
2.8E+08
Methane
ug/m3
2.3E+06
1.9E+06
2.0E+06
1.4E+06
5.3E+06
Total Hydrocarbons
ug/m3asC7H16
2.2E+08
2.8E+08
1.3E+08
1.5E+08
2.9E+08
52 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
*
S E C 0 R
SECOR International Inc.
308 E. 4500 S. Suite 100
Murray, Utah 841 07
Project No.: 26CH.3430000
S/fe: Hal's Chevron
Address: Green River, Utah
Drilling Contractor: EARTHPROBE
Method/Equipment: Hand Auger
Date Drilled: 10/08/03
Welt Completed: 10/08/03
SUBSURFACE PROFILE
Q.
i^Tn
5-
10-
.7
I
1
•
Lithologic Description ^t-S
r\l/tf"
Ground Surface S
^CONCRETE
Brown clayey SILT
Reddish brown SILT with clayey SILT
beds
Smear streaks at T
PID from inside hole 0.2 ppm
SAMPLE
METHOD
RECOVERY
MOISTURE
Moist
o
ML
Loo of Borehole: VW-4
Client: ChevronTexaco
Top of Casing Elev.:
Surface Elev.:
Groundwater Elevation:
Total Depth: 7. 25'
SAMPLE
BLOWS
676'/6"
INTERVAL
S
PID (ppm)
IV
5.5
6.7
8.6
191
Well Completion Details
Well Diameter: 2.0"
Borehole Diameter 3.5"
waiwuu s^
tfUSS CAP RTTNGS ON
BWOFMCOHKCTWO ffl ' ~7]
TDfeJMIMCAMSTEH j K '• .
COMCWTC"*^"^ tea te gs&
KTOMTSl M
KNTOMTE
CHP6 ^Tj
KKEN ;'•'.-":-
inoHJOIMO ;
MOVOSD W
KKTDNnC
015-°^
— T»--P "
Figure 5.7. Driller's log for vapor monitoring points at location VW-4 in the basement of the office for the motel.
§
I
1 .OE-02
1.0E-03
1.0E-04
1.0E-05
1 .OE-06
1.0E-07
1.0E-08
1.0E-09
1.0E-10
= 0.079 per hour
Basement
••-Separation 2 m
1.0E+05 1.0E+06 1.0E+07 1.0E+08 1.0E+09
Source Vapor Concentration as TPH plus Methane
Figure 5.8. Estimates of the attenuation factor in soil gas below the
office of the motel based on concentrations of vapor at the source.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 53
-------
5.5 Forecast of the Indoor Air
Concentration in the Basement of the
Office
Compared to the basement of the cafe, the
concentrations of total hydrocarbons and benzene in
soil gas adjacent to the basement of the office are ten
to one-hundred fold higher (compare Tables 5.1 and
5.2 to Tables 5.3 and 5.4). It is not surprising that the
forecasts of benzene concentrations in the basement
of the office are from ten to one hundred fold higher
(compare Table 5.2 to Table 5.5).
As was the case earlier, the estimated concentrations of
benzene in indoor air based on Abreu et al. (2009a) are
in reasonable agreement with the estimates provided
by the Generic BioVapor simulations. The estimates
based on the Site Specific BioVapor simulation are
approximately one-hundred fold lower.
The distribution of the PVIScreen simulations
is presented in Table 5.6. The most probable
concentrations ranged from 7E+01 u.g/m3 to 7E+02
u.g/m3. These concentrations were exceeded by
between 10.6% and 25.3% of the simulations. The
cancer screening level of 0.52 u.g/m3 was exceeded by
between 43.0% and 53.0% of the simulations. The
exceedance for the non-cancer hazard was between
29.0% and 47.8% of the simulations.
Table 5.5. Estimated concentration of benzene in indoor air in the basement of the office of the motel based on the Abreu
Three-Dimensional Model, on BioVapor and on PVIScreen.
Location
VW-1
VW-2
MW-2
MW-51
VW-3
Benzene
ug/m3
1.1E+06
2.3E+06
4.4E+06
8.5E+06
3.6E+06
a
Abreu
Figure
6E-04
8E-04
4E-04
4E-04
8E-04
Estimated Concentration of Benzene in Indoor Air
Abreu
Figure
ug/m3
7E+02
2E+03
2E+03
3E+02
3E+03
Generic
BioVapor
ug/m3
1E+03
3E+03
4E+03
9E+03
4E+03
Site Specific
BioVapor
ug/m3
4E-01
3E+00
1E-01
5E-01
5E+00
PVIScreen
Most Probable Result
ug/m3
7E+01
4E+02
4E+02
7E+02
7E+02
Table 5.6. Distribution of the simulations in PVIScreen of the concentration of benzene in indoor air in the basement of the
office of the motel.
Location
VW-1
VW-2
MW-2
MW-51
VW-3
PVIScreen Most
Probable Result
Hg/m3
7E+01
4E+02
4E+02
7E+02
7E+02
% Exceeding
Most Probable
Result
25.3
13.0
10.6
12.1
10.8
Acceptable Cancer Risk
(0.52 jig/m3)
54.0
60.8
43.0
46.3
60.4
Acceptable Non-Cancer Hazard
at Hazard Quotient = 1.0
(30 iig/m3)
36.9
46.8
29.0
33.9
47.8
54 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
5.6 Validation of the Forecast for the
Basement of the Office
As mentioned earlier, the Utah Indoor Air Guidance
(commercial) for benzene is 0.5 (ig/m3.
All of the estimates are near or are much higher
than that guidance. Following U.S. EPA (2013b), it
is necessary to further evaluate the exposure. The
Utah DEQ had sampled shallow soil gas beneath the
basement of the office. Data are provided in Figure
5.9. The shallow soil gas had adequate concentrations
of oxygen to support aerobic biodegradation of the
hydrocarbons. The concentrations of benzene are
lower in the shallow soil gas. Applying a generic
attenuation factor (a) of 0.1, the predicted indoor
air concentration at the last sampling date would be
5E+00 (ig/m3.
At this location, the estimates based on Abreu et
al. (2009a), on the generic Biovapor model and on
PVIScreen were higher than the estimates based on
measured concentrations of benzene in shallow soil
gas. The estimates provided by the Site Specific
BioVapor simulation were in reasonable agreement
with the estimate from shallow soil gas based on a
generic attenuation factor (a) of 0.1.
All the sampling locations for soil gas were located
between the tank pits and the motel. This would have
biased the screening models if the concentrations
under the basement of the office were actually lower
than the soil gas that was sampled for screening of
PVI. The concentrations of benzene in soil gas
adjacent to the office of the motel varied from 1E+06
to 9E+06 (ig/m3 (Table 5.4). The concentrations of
benzene at an equivalent depth underneath the building
were no more than 3E+03 (ig/m3 (Figure 5.9).
It is probably impossible to avoid sample bias based
on the location of the vapor points or monitoring wells.
However, it is important to control for the bias. Select
sampling points that are between the source and the
receptor.
VW-4
West Side of Oasis
Motel Basement
3 feet below floor
11 feet below land surface
7 feet below floor
15 feet below land surface
10/03
11/03
4/06
8/06
Benzene ((jg/m3) - O2 (% v/v)
72
980
36(12%) 51(9,5%)
200
3,400 410(5.4%) 570(4.1%)
Figure 5.9. Concentrations of benzene and oxygen in soil gas beneath the office of the motel.
Original graphic created by Robin Davis, Utah DEQ.
Concentrations of HC Compounds in Soil Gas at Source of C
GW Issue 55
-------
6.0 APPLICATION OF THE
APPROACH TO A SITE WITH
HIGH CONCENTRATIONS OF
METHANE AND SLAB-ON-GRADE
CONSTRUCTION (ANTLERS, OK)
The data used in these case studies 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.
This case study is a release of E10 gasoline at the EZ
Go service station in Antlers, OK. The release was
associated with underground storage tanks in a tank pit
and associated distribution lines. Figure 6.1 depicts
the relationship between the known extent of NAPL
contamination and ground water monitoring wells, the
convenience store that was the potential receptor, and
sub-slab vapor probes installed near and beneath the
slab of the convenience store.
North
t
50 feet
MW-
VMW-5
Benzene in sediment
> 1 mg/kg
Conventional
Groundwater Well
Sub-slab
Vapor Probe
VMW-2
Convenience Store VMW'3
Figure 6.1. Relationship between location of the building
that might act as a receptor, the known location of
petroleum NAPL, the location of monitoring wells, and
the location of vapor monitoring points.
Figure 6.2 describes the lithology of the unsaturated
material and the top of the aquifer. The log for MW-2
indicated that the Sandy Silt Loam at 10 to 11 feet had
tree roots and may represent the original land surface.
There is a tendency at this site for water to perch at
a level near ten feet below grade. The geological
material was largely a combination of clay, sandy clay
and clayey sand.
Figure 6.2 also depicts the depth intervals sampled by
the wells. Both wells had long screened intervals in
contact with the unsaturated zone.
3 -
Q.
OJ 1n
Q 10
12
14
16
18
20
Well MW-2
N
Grout
Bentonite
Seal
Sand
Pack
Groundwater
Well MW-9
Grout
Bentonite
Seal
Groundwater
uses
1 Asphalt and Fill and
Concrete
uses
Sandy Clay, Firm
Sandy Clay
Sandy Silt Loam
Sandy Clay
Sandy Clay
Clayey Sand
Sandy Clay
Clayey Sand
Concrete and Fill sand
Clayey Sand
Clay
Sand, Very Fine to Fine
Sandy Clay, Moist
Sand, Very Fine to Fine
Sandy Clay, Moist
No Sample
Sandy Clay, Moist
Sandy Clay, Small Gravel
Clay
No Sample
Sandy Clay
Shaley Clay
Sandy Clay
No Sample
Figure 6.2. Comparison of the driller's logs
to construction details of the ground water
monitoring wells.
56 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
6.1 Determining the Separation Distance
from the Source to the Slab of the
Convenience Store
Figure 6.3 presents the results from the OVA screening
of core samples to determine the vertical distribution
of NAPL. Data are from the contractor's report
to the Oklahoma Corporation Commission on the
installation of the wells. At MW-2, in the center of
the NAPL object, the OVA readings start to increase at
depths below 10 feet and reach high levels. The OVA
readings from samples at MW-9 are lower; however.
their distribution is consistent with a separation
distance often feet.
There was floating LNAPL in well MW-2. However.
no floating LNAPL developed in MW-9. Core samples
were sent for analysis of TPH and benzene in the
sediment. At MW-2, samples from 15.5 feet and 21
feet had 388 and 855 mg/kg TPH-g and 8.4 and
16.5 mg/kg benzene. At MW-9, a sample at 20 feet
had < 1 mg/kg TPH-g and < 5 (ig/kg benzene.
0 i
10
1 15 5
a
25
30
«MW-2
iMW-9
500 1000 1500 2000
OVA Reading (ppm)
2500
3000
Figure 6.3. Results of field screening for NAPL
in core samples from the two locations.
6.2 Forecast of the Indoor Air
Concentration in the Convenience Store
Gas samples were collected from well MW-9 and
MW-2 following the protocol of Jewell and Wilson
(2011). Gas samples were collected into 165 ml serum
bottles and analyzed for fixed gases including oxygen
and methane and for selected hydrocarbons including
benzene and the light alkanes. The concentrations of
individual hydrocarbons plus C6+ hydrocarbons were
summed to calculate the total petroleum hydrocarbons
in soil gas.
Results are presented in Table 6.1. Concentrations
were reported by the analysts in ppm (v/v). The
concentrations of oxygen and the formula in Section
4.3 were used to correct the reported concentrations for
dilution or leakage and then the formula in Section 4. 1
was used to convert the corrected concentrations to
The concentrations of TPH-g and benzene were very
similar in the soil gas that was collected from MW-9
and MW-2 (Table 6.1). The concentrations of TPH in
soil gas from MW-2 and MW-9 were consistent with
soil gas in contact with weathered gasoline.
This is reasonable for soil gas from MW-2. The OVA
meter readings on core samples were high and the
concentrations of TPH-g and benzene in core samples
were within the range that would be expected with
NAPL hydrocarbons.
This is not what would be expected from soil gas from
MW-2. The OVA meter readings on core samples
were much lower and TPH-g and benzene were not
detected in the core samples. The concentrations of
benzene in ground water were also very low (data not
shown). It is likely that the soil gas that was sampled
from well MW-9 originated from the NAPL that is
associated with well MW-2 and the tank pit.
The concentration of methane that was corrected for
dilution and leakage was corrected again as discussed
in Section 3.4.3 to express the concentration in units
of THP-g with an equivalent oxygen demand (Table
6.2). The corrected concentration of methane and
the concentration of TPH-g were added to calculate
the concentrations of total hydrocarbons. The
concentrations of total hydrocarbons were used in
Figure 6.4 to estimate a value for the attenuation factor
(a).
Concentrations of HC Compounds in Soil Gas at Source of C
GW Issue 57
-------
Table 6.1. Measured concentrations of oxygen, methane, and benzene in soil gas; calculated concentrations of total
petroleum hydrocarbons; and corrections of measured concentrations for dilution and leakage during sampling. Samples
were collected January 4, 2011.
Sample
Oxygen
%
Methane
%
pg/m3
Benzene
ppm-v
pg/m3
Total Petroleum Hydrocarbons
ppm-v
|jg/m3
Near the Convenience Store, TPH and Benzene Not Detected in Soil
Well MW-9
Well MW-9
Corrected
2.3
40
45
2.6E+08
43
48
1.5E+05
7.8E+03
8.7E+03
3.0E+07
High TPH and Benzene in Soil
Well MW-2
Well MW-2
Corrected
1.9
63
69
4.1E+08
180
190
6.3E+05
1.3E+04
1.5E+04
4.6E+08
Table 6.2. Concentrations of TPH-g and methane in soil gas near the convenience store, and the calculated total
concentrations of hydrocarbons in soil gas.
Sample
Well MW-9
Methane
pg/m3
2.6E+08
Methane-Corrected
|jg/m3
3.0E+08
TPH-g
|jg/m3
3.0E+07
Total Hydrocarbons
|jg/m3as C/H-ie
3.3E+08
1.0E-02 ,
1.0E-03
1 .OE-04
3 1.0E-05
CO
§ 1.0E-06
-------
A concentration of 3E+08u.g/m3 total hydrocarbons
(as C7H16) in soil gas from Well MW-9 predicted an
attenuation factor (a) of 2E-04. The concentration of
benzene in soil gas (corrected for dilution and leakage)
was multiplied by the attenuation factor (a) to predict
the concentration of benzene in indoor air of the
convenience store (Table 6.3).
The concentration of benzene in indoor air was also
predicted using the BioVapor model. The model
was set up using the Residential Default Values on
the Environmental Factors input screen with the
following expectations. Airflow Under Foundation
(Qf) was set to be equal to Air Flow Through
Basement Foundation (Qs). The value was 83 cm3 of
air per second. To be consistent with the assumptions
of the Abreu Three-Dimensional Model, the air
exchange rate was set to 12 per day. Core samples
were moist clay and sandy clay. The water content
was set to 0.19 cm3 water/cm3 soil. This is four times
higher than the default assumption. The default first
order rate constant for biodegradation of benzene
was 0.79 per day. To allow comparisons to Figures
3.6 and 3.7, which are based on the Abreu Three-
Dimensional Model with a first order rate constant of
0.079 per day, the rate constant for biodegradation of
benzene in BioVapor was changed to 0.079 per day.
This will provide higher estimates of the concentration
of benzene in indoor air from the simulation provided
by the BioVapor model. Predictions are provided in
Table 6.3.
In the Risk Information System (IRIS), the 10~5 risk
level for exposure to benzene in air is 1.3 to 4.5
u.g/m3 (U.S. EPA 2014c). A contractor for the
Oklahoma Corporation Commission used parameters
from the Oklahoma Risk Based Corrective Action
model and the U.S. EPA modification of the Johnson
and Ettinger model to generate a site specific Target
Indoor Air Concentration for benzene at a commercial
facility. The site specific target was 0.5 u.g/m3.
The predicted concentrations of the Abreu Three-
Dimensional Model, the Generic BioVapor simulation,
the Site Specific BioVapor Simulation and the Most
Probable Result from the PVIScreen simulations are
near the 10~5 risk level and exceed the site specific
Target Indoor Air Concentration.
Table 6.4 presents the distribution of the PVIScreen
simulations; 94.4% of the simulations exceeded the
cancer risk level, and 23.95% of the simulations
exceeded a Hazard Quotient of 1.
Following the draft guidance in U.S. EPA (2013) it is
necessary to further characterize the exposure.
Table 6.3. Estimated concentration of benzene in indoor air in the convenience store based on the Abreu Three-Dimensional
Model, on BioVapor and on PVIScreen.
Location
MW-9
Benzene
ug/m3
1.5E-K)5
a
Abreu
Figure
2E-04
Estimated Concentration of Benzene in Indoor Air
Abreu
Figure
ug/m3
30
Generic
BioVapor
ug/m3
100
Site Specific
BioVapor
ug/m3
9
PVIScreen
Most Probable Result
ug/m3
20
Table 6.4. Distribution of the simulations in PVIScreen of the concentration of benzene in indoor air in the convenience store.
Location
MW-9
PVIScreen Most
Probable Result
Hg/m3
20
% Exceeding
Most Probable
Result
30.3
Acceptable Cancer Risk
(0.52 jig/m3)
94.4
Acceptable Non-Cancer Hazard
at Hazard Quotient = 1.0
(30 jig/m3)
23.9
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 59
-------
6.3 Validation of the Forecast for the
Convenience Store
Well MW-9 was sampled during a survey of gasoline
spill sites in Oklahoma (Jewell and Wilson, 2011).
Based on the high concentration of benzene in soil gas
from MW-9 near the convenience store, the Petroleum
Storage Tank Division of the Oklahoma Corporation
Commission tasked a contractor to install and sample
shallow soil gas probes at the perimeter of the
convenience store and sub-slab vapor probes through
the slab-on-grade foundation. Their data are presented
in Table 6.5.
The contractor in their report had the following
observation:
VMW-1 and VMW-2 yielded the highest
COC [compound of concern] and
LCC [leak check compound] concentrations.
The high COC concentrations could be due
to the sandy fill below the foundation of the
building. It is likely that sandy fill has
a higher porosity than the surrounding soils
resulting in the potential for greater vapor
migration and accumulation of vapors in
the fill below the foundation. The exterior
vapor monitoring ports yielded lower COC
concentrations which could be due to the
clay layer observed in the exterior boreholes.
The clay layer has a very low permeability
and likely minimizes the potential for vapor
accumulation at the deeper depths.
As an aside,given the nature of the facility
and the very low detection limits, should
outside ambient air have infiltrated the
samples, it is also reasonable to conclude
that it would likely increase COC
concentrations in the samples.
The highest concentration of benzene in any
of the shallow soil gas sample was in sub-slab
monitoring point VMW-2 (10 (ig/m3). The indoor
air concentrations that were predicted by applying
a generic attenuation factor (a) of 0.1 to the
concentration in sub-slab monitoring point
VMW-2 was 1.0 (ig/m3. This was lower than the
estimate provided by the Abreu simulations
(30 (ig/m3), the estimate provided by the generic
BioVapor simulation (100 (ig/m3), the estimate
provided by the site specific BioVapor simulation
(9 (ig/m3) or the PVIScreen most probable result
(20 (ig/m3). At this location at this site, the models
predicted higher concentrations of benzene in indoor
air, compared to the estimate based on the maximum
concentration of benzene in shallow soil gas.
There was only one round of soil gas sampling
from the shallow monitoring points. The contractor
recommended to the Petroleum Storage Tank Division
of the Oklahoma Corporation Commission that they
sample the shallow monitoring points a second time.
Table 6.5. Concentrations of benzene in shallow soil gas beneath and beside the foundation of the convenience store.
Sample
VMW-1
VMW-2
VMW-3
VMW-5
Depth
2 inches below slab
2 inches below slab
4 to 5 feet below sidewalk
4 to 5 feet below sidewalk
Benzene in Sample
pg/m3
9.4
10
0.74
1.9
Predicted Benzene in
Indoor Air
a = 0.1
pg/m3
0.94
1.0
0.074
0.19
60 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
6.4 Caveat with Respect to High
Concentrations of Methane
The soil gas at the Antlers site had very high
concentrations of methane. The high concentrations
suggested that methane production was displacing
other gases from the soil gas (Wilson et al., 2013,
see also Wilson et al., 2012b; Ma et al., 2014). This
situation is not considered in the structure of the Abreu
Three-Dimensional Model or BioVapor or PVIScreen.
If methane concentrations at the source of hydrocarbon
vapors are above 10% (after correction for dilution and
leakage), evaluate the concentrations of the compound
of concern shallow soil gas, regardless of the
predictions of the simple screening models discussed
in this Issue Paper. If shallow monitoring points
are not available, install a shallow monitoring point
immediately adjacent to building. Sample the soil gas
once and at least one additional time.
7.0 IMPLEMENTING THE APPROACH
This section describes in detail the sequence of
activities that are involved in implementing the
approach. This section provides a synopsis of material
that is discussed at greater depth in Sections 2, 3 and
4, and summarizes data presented in the detailed case
studies in Sections 5 and 6.
7.1 Sites that Are Appropriate
for the Approach
This approach is designed for sites that have been
previously screened to determine if they are within
the lateral inclusion zone and the vertical inclusion
zone (U.S. EPA, 2013b; Wilson et al., 2012a). In the
most direct application, the approach is only applied
to sites that are within both the lateral and the vertical
inclusion zone. The approach can also be applied as
a second line of evidence for sites that may or may
not be in the vertical inclusion zone. The approach
is intended for a building of the size of a typical
residential house (up to 66 ft x 66 ft or 20 m by 20 m).
The approach includes biodegradation of petroleum
vapors in the unsaturated soil in the assessment of
vapor intrusion. As a result, the approach is only
appropriate for structures that are built on soil or
sediment that can support the activity of bacteria that
degrade hydrocarbons using oxygen. This issue is
discussed in more detail in Section 7.4.7.
The approach is intended for sites that do not have
preferential pathways that are conduits for flow of
hydrocarbon vapors. See U.S. EPA (2011) for further
discussion.
7.2 Number and Position
of Sampling Locations
The number and position of sampling locations is
at the discretion of the individual case worker who
is responsible for the site. At least one sampling
location should be on the side of the building that faces
a source of contamination. Presumably, the extent
of contamination will decrease with distance from
the source. If this is the case, there should be less
contamination under the building being evaluated than
there is at the sampling location.
The sampling location should be near the building
being evaluated, but it is not necessary for the location
to be immediately adjacent to the building. If these
criteria are met, a single sampling location should be
adequate for many sites.
7.3 Frequency of Sampling
The number and frequency of sampling events should
be at the discretion of the individual case worker who
is responsible for the site. Data from the first round
of samples should be evaluated to determine if it is
necessary to start immediately to remediate the site. If
the forecast from the first round of samples indicates
that concentrations of the compound of concern in
indoor air should be acceptable, the location should be
sampled at a minimum one additional time to confirm
that the risk of PVI is managed.
7.4 Information Needed to Implement
the Approach
The approach was created to make the maximum
possible use of site characterization data that is already
available for a site. The first step is to review the case
files and collect the information necessary to
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 61
-------
implement the approach. If any of the necessary
information is not available from the case file, then
the missing information must be acquired before the
approach can be implemented.
As described in Sections 2, 3 and 4, the approach
requires the following information for each particular
location that is being screened for petroleum vapor
intrusion:
1) The vertical separation distance between the
bottom of the receptor and the primary source
of the vapors.
2) The concentration of the particular
hydrocarbons of concern in soil gas at the
source of the vapors.
3) The concentration of the fuel hydrocarbons
(TPG) in soil gas at the source of the vapors.
4) The concentration of methane in soil gas at the
source of the vapors.
5) The concentrations of oxygen, carbon dioxide
and nitrogen in the soil gas sample.
6) The appropriate regulatory standard for
concentrations of the particular hydrocarbon
of concern in indoor air.
7) Knowledge that conditions are appropriate
to support growth of bacteria that degrade
hydrocarbons.
7.4.1 The Separation Distance between the
Receptor and the Source of the Vapors
To know the separation distance, it is necessary to
know the important source of the vapors. If the source
is the ground water, the separation distance is the
distance to water table. This should be determined
from the highest elevation of the water table in the
monitoring record.
If the source is residual non-aqueous phase liquid
hydrocarbons (NAPL) in the unsaturated zone, NAPL
on the water table or NAPL below the water table, the
separation distance is the extent of soil or sediment
that does not contain enough NAPL to influence
degradation of vapors in soil gas, as defined below.
Do not assume the separation is defined by the top of
smear zone that would be produced by variations in
the depth to ground water. The separation distance
must be determined by collecting and evaluating core
samples.
The separation distance to NAPL is not related to the
depth to ground water at the time the core samples
were collected. When the water table is high the
NAPL may be inundated, but the NAPL may be
exposed to soil gas when the water table is lower.
Determine that the well log extends to a depth of 15
feet below the bottom of the building. If core samples
have not extended to 15 feet below the bottom of
the building, it is necessary to acquire and screen
additional core samples. Recommendations are
provided in Wilson et al. (2012a).
There are two common techniques to estimate
NAPL in the subsurface. First, core samples can be
extracted and analyzed for TPH in the laboratory. If
the concentration of TPH is less than 100 mg/kg for
fresh gasoline, or less than 250 mg/kg for weathered
gasoline or diesel, then the NAPL should not be
expected to influence biodegradation.
Second, core samples can be screened in the field
by sealing a plug sample from a core into a plastic
bag, allowing hydrocarbons in the plug sample
to equilibrate with the air in the bag, and then
determining the concentration of TPH in the air in
the bag with a field meter. The readings are often
labelled as PID or OVA on a driller's well log. If the
concentration of TPH in air that is equilibrated with
the plug sample is below 100 ppm-v, then the NAPL
should not influence biodegradation.
The resolution in determining the separation distance
is controlled by the spacing of the core samples. If
a good well log is available, and the log provides
information on PID readings or OVA readings on core
samples, the log should be used as the first line of
evidence to determine the separation distance.
Soil or sediment that has been contaminated by a
release of fuel will often be dark grey or black, and
will smell of gasoline of hydrogen sulfide. Consult
62 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
the driller's log for any notes on the color or odor of
the core samples. This can provide a second line of
evidence.
The separation distance is the distance between the
receptor and the deepest core sample that has less than
100 mg/kg of TPH as gasoline or less than 250 mg/kg
of TPH as diesel fuel or weathered gasoline. It is not
the distance between the receptor and the first sample
that has more than 100 or 250 mg/kg of TPH.
Similarly, the separation distance is the distance
between the receptor and the deepest core sample that
produces less than 100 ppm-v of TPH in air when a
sub-sample of the core is equilibrated with air. It
is not the distance between the receptor and the first
sample that produces more than 100 ppm-v of TPH in
air.
If there is a significant distance between the deepest
sample that has less than 100 mg/kg TPH [or 250 mg/
kg TPH] or produces less than 100 ppm-v of TPH in
air and the shallowest sample has more than 100 mg/
kg TPH [or 250 mg/kg TPH] or produces less than
100 ppm-v of TPH in air, it may be worthwhile to
acquire additional core samples and better define the
separation distance.
7.4.2 The concentration of the particular
hydrocarbons of concern in soil gas at the
source of the vapors
The U.S. EPA identifies four methods that are
intended to determine the concentrations of individual
petroleum hydrocarbons in samples of indoor air or
soil gas. They are Method TO-15 (U.S. EPA, 1999a),
Method TO-14A (U.S. EPA, 1999b), Method TO-3
(U.S. EPA, 1984), and Method TO-17 (U.S. EPA,
1999c).
Compendium Method TO-15 is widely used. Gas
samples are collected into an evacuated stainless
steel canister that has gone through a special process
to make the interior of the canister chemically inert.
After the gas sample is collected, the canister is
sealed and shipped to the laboratory for analysis by
high resolution gas chromatography using a mass
spectrometer as the detector.
The case file should contain records of the analysis
of soil gas for the particular hydrocarbons of concern
using methods that are approved by the appropriate
regulatory authority.
If records are not available, then soil gas samples
should be acquired and analyzed. If vapor monitoring
probes are available, they should be used. If they are
not available, consider using groundwater monitoring
wells to acquire the gas samples. Do not use
temporary push wells to acquire the soil gas samples.
7.4.3 The concentration of the fuel
hydrocarbons in soil gas at the source
of the vapors
The case file may contain an analysis of TPH-g by EPA
TO-15. The TPH-g is expressed as the equivalent of a
hydrocarbon with a molecular weight of 100 Daltons,
which is the weight of heptane. If this analysis is
not available, it is necessary to sample the soil gas
again and determine TPH-g by EPA TO-15. On the
same sample also determine the concentrations of the
compounds of concern.
7.4.4 The concentration of methane in soil gas
at the source of the vapors
The case file may contain an analysis of methane
in soil gas. If this analysis is not available, it is
necessary to sample the soil gas again and determine
methane by EPA Method 3C (U.S. EPA, 2014d) or
by ASTM D-1945 (ASTM, 2010). This analysis can
be performed on the same sample used to determine
the concentrations of compounds of concern or the
concentration of TPH.
Do not rely on the reading for methane provided by an
infra-red detector in a field meter. The detector cannot
distinguish methane from petroleum hydrocarbons.
7.4.5 The concenfraf/ons of oxygen, carbon
dioxide and nitrogen in the soil gas sample
The concentration of oxygen can be used to correct
a field sample for leaks in the sampling train and
for dilution of the soil gas in the vapor probe or
monitoring well (see section 7.5.3). This information
is useful and the determination of oxygen in the
soil gas sample is strongly recommended. The
concentration of carbon dioxide can be used in a
GW Issue 63
-------
qualitative way to distinguish a sample of atmospheric
air from a sample of soil gas where petroleum
hydrocarbons have been extensively biodegraded.
The concentration of nitrogen can be used along
with concentrations of oxygen and carbon dioxide to
calculate a mass balance as a quality control check
on the analyses. The analysis of carbon dioxide and
nitrogen is useful but they are not necessary.
EPA Method 3C (U.S. EPA, 2014d) or ASTM D-1945
(ASTM, 2010) can determine methane, oxygen, carbon
dioxide and nitrogen on the same sample in the same
analytical run. Oxygen and carbon dioxide can also be
determined using field meters.
7.4.6 The appropriate regulatory standard for
concentrations of the particular hydrocarbon
of concern in indoor air
The approach in this Issue Paper is used to make an
exposure evaluation. That exposure evaluation at
some point will be reviewed by the government agency
with regulatory authority for the site. Check with the
agency to insure that the approach is implemented with
the current standards that are appropriate for the site.
7.4.7 Knowledge that conditions are
appropriate to support growth of bacteria
that degrade hydrocarbons
There are three circumstances where conditions may
not be appropriate to degrade hydrocarbons.
The separation distance must be in soil or sediment
that has a significant holding capacity for water, and
provides adequate surface area for the growth of
bacteria. Exclude from the separation distance any
interval that is consolidated rock or is dominated by
course sand or gravel without any fine materials.
The soil or sediment may be too dry to allow growth of
bacteria. The soil water potential (*¥) may be so low
as to inhibit the growth of bacteria. Orchard and Cook
(1983) showed that at a soil water potential of -1.0
megapascals (MPa), the rate of soil respiration was
reduced to approximately one-third of the optimal rate.
At a potential of-0.3 MPa, the rate was approximately
one-half of the optimal rate. After water has been
distributed through soil and reached equilibrium, the
soil water potential is related to the depth to the water
table. At a potential of -0.3 MPa, the depth to the
water table should be 1 12 feet. Soil water potential
may be a problem if the depth to the water table is
more than 100 feet. If the source of vapors is NAPL
that occurs more than 100 feet above the water table,
do not apply the approach.
The third condition is a building, pavement, or a layer
of soil that prevents the transfer of oxygen from the
atmosphere to soil gas. This Issue Paper is designed
to screen a building that is used as a private residence.
For a building with the footprint of a conventional
private building, the transfer of oxygen should not be
a problem. The approach in this Issue Paper should
not be used for a larger building unless it can be shown
that the concentration of oxygen is adequate in air
immediately underneath the building foundation or
adjacent paving.
On first examination, it would seem that a layer of
soil that prevents the transfer of oxygen should be
a problem. The most usual circumstance will be a
layer of wet clay with little secondary porosity. The
clay will inhibit the transport of oxygen to the source
of hydrocarbon vapors, but it will also inhibit the
transport of the hydrocarbon vapors to the receptor.
Under most circumstances, a layer of wet soil will
not cause an appreciable increase in the concentration
of petroleum hydrocarbons in indoor air (U.S. EPA,
2013a).
7.5 Evaluate the Exposure
Once all the necessary information has been gathered,
the next step is to forecast the indoor air concentration
of the compound of concern.
7.5. 1 Convert Units
Convert all the concentration data for compounds of
concern, for total TPH and for methane into units of
Hg/m3. If data are reported in ppm, multiply by 1000
to convert to ppb. If data are reported in percent,
multiply by 10,000 to convert to ppb. Data in ppb can
be converted to (ig/m3 with the following formula.
= ppb * molecular weight * 0.0413
64 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
7.5.2 Calculate Total Hydrocarbons
Add the concentration of methane and the
concentration of TPH to get the concentration of
total hydrocarbons in the soil gas. To correct for
the difference in the theoretical oxygen demand
of methane and TPH, multiply the concentration
of methane by 1.14 before adding the corrected
concentration to the concentration of TPH. See
Section 3.4.3 for details.
7.5.3 Correct for Dilution or Leakage
Use the concentration of oxygen in the sample of
soil gas to calculate a correction factor for dilution
and leakage in the sample. The correction factor can
be calculated with the formula below, or it can be
extracted from Figure 4.3.
Concentration in Source Gas =
Concentration in the Sample Analyzed -
O, in Atm.
in Atm.-O, in Soil Gas)
Multiply the measured concentration by the correction
factor to calculate the concentration at the source of
the hydrocarbons.
7.5.4 Forecast the Concentration of
Compound of Concern in Indoor Air
This Issue Paper offers two approaches to forecast
the concentration of benzene or other compound of
concern in indoor air in a building. One approach
uses the simulations of Abreu et al. (2009a, 2009b).
The other approach uses the computer applications
Bio Vapor (API, 2012) or PVIScreen (U.S. EPA,
2014g).
7.5.4. / Approach Using Simulations
of Abreu ef al.
Abreu et al. (2009a, 2009b) modeled a large number
of representative scenarios and combined the results
of the simulations in simple figures. Figure 3.14
combines their simulations for a typical residential
building that is built slab-on-grade, and Figure 3.15
combines simulations for a building with a basement.
To use the figures, multiply the separation distance
in feet by 0.3048 to calculate the separation distance
in meters. Find the largest separation distance that is
plotted in the figure that is smaller than the separation
distance at the site. Plot the concentration of total
hydrocarbons on the X axis. Extend the concentration
up to intersect the line representing the separation
distance, then extend a line to the left to the Y axis to
estimate the attenuation factor (a). To complete the
evaluation, multiply the attenuation factor (a) by the
corrected concentration of benzene in soil gas. This
calculated value is the forecast of the concentration of
benzene in indoor air in the building.
7.5.4.2 Approach Using Bio Vapor or PVIScreen
The concentration of total petroleum hydrocarbons,
methane and the separation distance are input values
for BioVapor and PVIScreen. Consult the User's
Manual to learn how to use the applications. The
computer applications allow the user to input site-
specific parameters such as the water content of the
vadose zone soil or the content of organic carbon. The
parameters were used in the simulations of Abreu et al.
(2009a, 2009b); however, they were held constant for
all the simulations. BioVapor and PVIScreen allow a
forecast that is based on the conditions that are most
applicable to a particular site.
As is the case with approach of Abreu et al. (2009a,
2009b), BioVapor provides a single value for the
forecast of indoor air concentrations. In contrast,
PVIScreen performs a Monte Carlo simulation and
provides a "most probable result" for the forecast of
indoor air concentration, as well as, the fraction of the
simulations that were above a pre-specified cancer risk
level, and the fraction of simulations that were above a
hazard quotient of 1.0.
7.5.4.3 Comparison of fhe Screening Approaches
In the case studies discussed in Sections 5 and 6, four
separate forecasts were made of the concentration
of benzene in indoor air of a building at the site. A
forecast was made by comparing conditions at the site
to the simulations of Abreu et al. (2009a, 2009b). A
second forecast was made using the BioVapor model
without modification of the generic input values for
water content or for the concentration of soil organic
matter (generic BioVapor). A third forecast was made
using the BioVapor Model with site specific values
GW Issue 65
-------
for water content or organic matter (site specific
Bio Vapor). A fourth forecast was made using the
PVIScreen model with site specific values.
The screening approaches were based on the analysis
of soil gas from a deep monitoring point that collected
gas at the source of the hydrocarbon vapors. In
Sections 5 and 6, the forecasts of the screening
approaches were validated by comparing them to a
prediction made by applying a generic attenuation
factor (a) to the measured concentration of benzene
in soil gas from a second shallow vapor probe or
sub-slab monitoring point. Guidance provided by
U.S. EPA (2014e) sets the value of the generic
attenuation factor (a) at 0.1.
The screening approaches can be evaluated by
comparing to the forecast of the screening approach
to the prediction that was provided by applying the
generic attenuation factor. Table 7.1 identifies the four
pairs of monitoring points where the comparison can
be made.
Figure 7.1 compiles data from the four comparisons
that are described in Table 7.1.
The solid blue line in Figure 7.1 is a line of equivalent
concentrations. If the data point plots above the solid
blue line, the concentration of benzene in indoor air
that was predicted from the screening approach was
greater than the concentration that was predicted based
on the generic attenuation factor. If the data plots
below the line, the concentration of benzene predicted
from the screening approach was less than the
concentration predicted from the generic attenuation
factor.
O
I?
c c
£ E
m o
* Abreu et al. Approach
• BioVapor Generic
o BioVapor Site Specific
* PVIScreen
1.0E-02 1.0E-01 1.0E+00 1.0E+01 1.0E+02
Benzene in Indoor Air Predicted from a=0 1
Figure 7.1. Relationship between the concentrations
of benzene in indoor air that are predicted from a
generic attenuation factor and the concentrations
of benzene in indoor air that are predicted from
a screening approach that includes aerobic
biodegradation of benzene.
Table 7.1. Locations of monitoring points that allow a comparison of the prediction made by applying a generic attenuation
factor to prediction made by applying screening models that include aerobic biodegradation of benzene.
Section of Issue Paper
Site
Section 5
Green River, Utah
Section 5
Green River, Utah
Section 5
Green River, Utah
Section 6
Antlers, Oklahoma
Location
Cafe
Office of Hotel
Office of Hotel
Convenience Store
Shallow Monitoring Point Used
to Sample Soil Gas to Apply a
Generic Attenuation Factor of 0.1
VW-5
VW-4
VW-4
VMW-2
Deep Monitoring Point Used to
Sample Soil Gas to Implement the
Screening Approach
MW-47
MW-51
MW-2
MW-9
66 GW Issue
is at Source of Contamination to Evaluate Potential for PVI
-------
The forecasts that were provided using the simulations
of Abreu et al. (2009a) or the Generic BioVapor model
were very conservative. The concentrations were
consistently from thirty fold to a thousand fold higher
than the predictions based on a generic attenuation
factor (a) of 0.1.
The forecasts provided by the simulations of Abreu et
al. (2009a) or the Generic BioVapor model were even
more conservative than the predictions made using
the generic attenuation factor. If a forecast made from
the simulations of Abreu et al. (2009a) or the Generic
BioVapor model indicates that the concentration of
benzene in indoor is acceptable, there is little benefit
from installing and sampling shallow monitoring
points to further validate the forecast.
In contrast, some of the forecasts made with the
site specific BioVapor model or with PVIScreen had
lower concentrations of benzene compared to the
prediction made from concentrations in shallow soil
gas using an attenuation factor (a) of 0.1. Three of
the four forecasts were closer to the line of equivalent
values, and one of the forecasts was much lower. An
attenuation factor (a) of 0.1 ignores any contribution
from biodegradation. It is intentionally conservative.
The forecasts made by the Site specific Biovapor
model or with PVIScreen may have been a better
description of the true conditions at the site.
7.5.4.4 Recommendations on Screening
Approaches
Presumably, the building was determined to be
within a vertical inclusion zone based on previous
characterization of petroleum hydrocarbons in ground
water or soil cores (U.S. EPA, 2013b; Wilson et al.,
2012a).
7.5.4.4.1 Building with No Vapor Monitoring Points
The following recommendations apply to a building
where no vapor monitoring points have been installed.
Often a ground water monitoring well will be installed
in the borehole used to acquire the core samples. If
a ground water monitoring well is available, the well
should be used to sample for soil gas and gas samples
should be analyzed as described above. If more than
one well is in reasonable proximity to the site, sample
all the wells.
The data should be evaluated by comparison to the
simulations of Abreu et al. (2009a, 2009b) as presented
in Figures 3.14 or 3.15. If forecasts are available from
more than one location, make the risk evaluation based
on the forecast with the highest concentration of the
compound of concern in indoor air. Do not average
the forecasts.
If the forecast indicates that the risk of PVI is
acceptable, one option is to continue to monitor the
soil gas additional times to establish that conditions do
not change.
If the forecast indicates that the risk of PVI is not
acceptable, evaluate the data a second time using
PVIScreen or a site specific implementation of the
BioVapor model. If the forecast of PVIScreen or the
site specific implementation of the BioVapor model
indicates that the risk of PVI is not acceptable, one
option is to install a shallow vapor monitoring point
immediately adjacent to the building. Sample soil gas
from the shallow vapor monitoring point and analyze
the gas for concentrations of compounds of concern.
Apply the appropriate generic attenuation factor to
predict the concentration in indoor air.
If a shallow vapor monitoring point is installed, there
is no need to install a deep vapor monitoring point to
sample petroleum vapors at the source of the vapors.
7.5.4.4.2 Building with Vapor Monitoring Points
If vapor monitoring points have been installed at the
site, sample all the monitoring points and perform the
evaluation for each monitoring point. If monitoring
points are available in the range of three to five feet
below the bottom of the building, sample soil gas from
these monitoring points and apply the appropriate
generic attenuation factor as the first line of evidence
to evaluate the risk of PVI. If deeper monitoring
points are available, sample them and apply the
approach in this Issue Paper as a second line of
evidence.
The data from the deeper vapor monitoring points
should first be evaluated by comparison to the
simulations of Abreu et al. (2009a, 2009b) as presented
in Figures 3.14 or 3.15. If forecasts are available
from more than one vapor monitoring point, make the
risk evaluation based on the forecast with the highest
GW Issue 67
-------
concentration of the compound of concern in indoor
air. Do not average the forecasts. If the forecasts
indicate that the risk of PVI is acceptable, one option is
to continue to monitor the soil gas additional times to
establish that conditions do not change.
If the forecasts from any deeper wells indicate that
the risk of PVI is not acceptable, evaluate the data
a second time using PVIScreen or a site specific
implementation of the BioVapor model. If the forecast
of PVIScreen or the site specific implementation of the
BioVapor model indicates that the risk of PVI is not
acceptable, one option is to install one or more sub-
slab sampling points inside the building. Sample soil
gas from each sub-slab monitoring point and analyze
the gas for concentrations of compounds of concern.
Apply the appropriate generic attenuation factor to
predict the concentration in indoor air.
If the distance from the bottom of the building to
the vapor monitoring point is less than the presumed
distance from the bottom of the building to the source
of vapors, but the true distance to the source of vapors
is not known, then use the distance to the monitoring
point as the separation distance in the evaluation.
8.0 COMPARISON OF SOIL GAS
SAMPLES COLLECTED FROM VAPOR
PROBES AND WATER WELLS
Information on the concentrations of benzene and total
petroleum hydrocarbons in soil gas at the source of
the petroleum vapors is useful in screening sites for
petroleum vapor intrusion. Traditionally, samples of
soil gas are acquired from soil vapor probes that are
built for that purpose. However, if existing ground
water monitoring wells could be used to sample soil
gas, this can reduce the cost of site characterization for
petroleum vapor intrusion (PVI).
8.1 Background
Installation of dedicated soil vapor probes increases
cost for site characterization and creates additional
disruption for the site owner. Most leaking
underground storage tank (LUST) sites have an
existing ground water monitoring well network.
Often, a portion of the screened interval of
conventional ground water monitoring wells is above
the water table. Figure 8.1 shows the similarity
between monitoring wells where the water table is
within the screened interval and vapor probes. If soil
gas samples can be obtained from existing ground
water monitoring wells, this would avoid the cost
of installing vapor probes and reduce cost of site
characterization.
JCap
Bentonite grout
2 inch diameter
schedule-40
PVC riser pipe
•%•'
|jj — Bentonite chips
Washed 20-40 sand
2 inch diameter
schedule-40
PVC slotted
pipe
quick connect
fitting
Bentonite grout
Bentonite chips
Washed 20-40 sand
0.25 inch
diameter steel
or nylon tube
Figure 8.1. Similarities between a conventional ground
water monitoring well (a) and a dedicated vapor probe (b).
68 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
A study was conducted to evaluate whether or not
conventional ground water monitoring wells are
functionally different from dedicated vapor probes
and determine what those differences are. Recently,
Jewell and Wilson (2011) proposed a practical method
for sampling petroleum hydrocarbon soil vapors from
conventional ground water monitoring wells that have
some portion of the screen above the water table. They
obtained samples from 12 gasoline sites in Oklahoma
and results showed their method provided comparable
soil gas concentrations compared to dedicated
vapor probes. This method was used to compare
soil gas samples from dedicated vapor probes with
conventional groundwater monitoring wells at Hal's
Chevron LUST site in Green River, Utah, and at the
EZ Go Service Station in Antlers, Oklahoma.
Hal's Chevron site is located in Green River, Utah.
In 1991, it was discovered that between 40,000 to
130,000 gallons of gasoline had leaked from the
UST system and migrated downward through the
vadose zone to the ground water table at about 18
feet (ft) below ground surface (bgs). From 1993 to
2004, 26,000 gallons of free product was removed.
The gasoline plume migrated in an east/southeast
direction for about 300 ft where it stabilized. The site
is underlain by free phase gasoline which provided
a strong source of petroleum hydrocarbon vapors.
The free-phase gasoline plume underlies a motel,
restaurant, and road at depths ranging from 15 to 18 ft
bgs. The unsaturated zone is composed of interbedded
silts and clays. The subsurface is well-characterized
and the lateral extent of LNAPL is well-defined.
8.2 Comparison of Monitoring Wells and
Vapor Probes at Hal's Chevron,
Green River, UT
The data used in this 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 Utah Department of Environmental
Quality.
Conventional ground water monitoring wells and
dedicated multi-depth vapor probes were installed
by The Utah Department of Environmental Quality
(UDEQ) at the site (Figure 8.2). Five vapor probes
(VW-1, VW-2, VW-3, VW-7, and VW-8) and eight
conventional monitoring wells (MW-2, MW-21,
MW-39, MW-41, MW-42, MW-47, MW-48, and
MW-51) were sampled in this study.
Edge of LNAPL
>0.01 ft thick in well
VW-3
North
Motel
slab-on-grade
Motel with
basement
Cafe with
Basement
VW-5
*J MW-42
L
I MW-21
VW-7
® MW-47
50 Feet
© Conventional Ground Water Monitoring Well
(2) Multi-Depth Vapor Monitoring Point
Figure 8.2. Map of Hal's Chevron site
with location of conventional ground
water monitoring wells (MW) and multi-
depth vapor probes (VW). The brown
area depicts the area with known
distribution of NAPL.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 69
-------
The lithology of the site and the vertical distribution
of petroleum are shown in Figure 8.3. Total petroleum
hydrocarbon (TPH) concentration was determined
as gasoline range organics (GRO). The highest TPH
concentration is seen at 14 ft bgs. Depth intervals less
than 12 ft below land surface have less than
250 mg/kg TPH.
The borehole log and details on the completion of the
multi-depth vapor points at VW-7 are shown in
Figure 8.4. As is consistent with Figure 8.3, screening
of core samples with a PID meter showed clean soil
down to 11 ft.
The vapor probes were comprised of 6-inch stainless-
steel vapor sampling screens with polyethylene tubing
to ground surface (Figure 8.5). The deepest vapor
sample point in each vapor well was located either in
the region with NAPL, or just above the NAPL.
Ground water monitoring wells were constructed
using 2-inch or 4-inch diameter polyvinylchloride
(PVC) pipe. The monitoring wells were fitted with an
Ex-Cap® (Atlantic Screen and Manufacturing, Inc.)
that provided a port to extract vapor samples. When
4-inch wells were sampled, a reducer was used to join
the 2-inch Ex-Cap® to the 4-inch riser of the wells.
500
TPH as GRO (mg/kg)
1000 1500
2000
2500
10
multi-depth vapor
probes
V, <<-Asphalt
-Silty Clay
-Clayey Silt
- Silty Clay
-Clayey Silt
- Silty fine Sand
-Silt
h- Silty fine Sand
H Black Stain
Figure 8.3. Distribution of TPH at Hal's Chevron site.
70 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
S E C O R
SECOR International Inc.
309 E. 450QS Suite 100
Murray. Utah 84107
Proper *to.. 26CH4JM30000
S*r HaTs Chevron
Adams. Green River Utah
Dating CantrKtor. EASJHPRQ9E
-BBOO
10/07(03
kVW/ C>.;
Total Depth- 15.25'
SUBSURFACE PROFILE
SAMPLE
0--0
10-
18-
bthobglc Des^olon
.ASPHALT
\GRAVEL ROAC BASE
Brwn dayey SILT
Brawn c*y«y SILT, med hard. §ome
Brown cayey SILT, ried Hard
RwURfi Mown St.". mM. nard
Reddis.1 Srswn SILT
Clean soil down
to 11 feet.
I
B
I
Moisi
Weil Compiatieo CMails
WtfOiwT^ter.Oir
Bering Dimeter w
1.466
Figure 8.4. Borehole log of VW-7.
Figure 8.5. Vapor probes consist of a soil vapor sampling screen and polyethylene tubing to ground surface.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 71
-------
8.3 Field Sampling
Soil gas samples from ground water monitoring wells
were collected using the method of Jewell and Wilson
(2011). The soil gas sampling train used to sample
ground water wells is shown in Figure 8.6. A vacuum
pump was used to extract soil gas from the screened
interval exposed above the water table. The soil gas
passed through a 1-L glass flask that acts as a trap
for water, then to the pump. The vacuum generated
on the headspace of the well was monitored by a
differential pressure gauge that was referenced to the
atmosphere (range 0-50 inches of water). The flow
rate was measured with a rotometer (flow rate meter)
located on the effluent side of the pump and adjusted
with a needle valve between the pump and rotometer.
The gas on the effluent side of the needle valve was
at atmospheric pressure. The effluent of the rotometer
went to a tee split to create a path that supplied gas to
the field meters and a second path to provide an outlet
for excess gas.
Samples from the vapor probes were collected using
conventional soil gas sampling methods into Summa
canisters (samples collected following U.S. EPA,
1994, 1996; canisters provided by H&P Mobile
Geochemistry, Inc., Carlsbad, CA). Soil gas samples
from vapor probes were analyzed by TO-15 (Air
Toxics, Inc). Soil gas samples from conventional
monitoring wells were analyzed as described in Jewell
and Wilson (2011). They used a micro GC with a
thermal conductivity detector. The detection limit
for benzene was much higher using the micro GC
compared to method TO-15.
Exhaust
Rota meter
(flow rate meter)
Figure 8.6. Sampling train for obtaining soil gas samples from a conventional monitoring well.
72 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
8.4 Comparison of Vapor Probes to
Conventional Monitoring Wells at
Green River, UT
For the conventional ground water monitoring wells,
data on concentrations of oxygen and the equation
in Section 4.3 were used to correct the measured
concentrations of benzene, TPH-g and methane for
dilution and leakage.
Table 8.1 compares concentrations of soil gas in
locations between the service station and the office of
the motel. Figure 8.7 shows the sampling locations.
Well MW-39 is located outside the area with LNAPL.
Wells MW-2 and MW51 and vapor probes VW-1,
VW-2 and VW-3 are located within the area with
NAPL(Figure8.7).
The concentrations of benzene, TPH-g and methane
were much lower in well MW-39, which was not
located in the area with NAPL. Comparing the
monitoring wells and vapor probes that were located
in the area with NAPL, the concentrations of benzene,
TPH-g and methane were very uniform. They agreed
within a factor of eight between any of the wells and
any of the vapor probes. Although the vapor probes
had higher concentrations of TPH-g, the wells had
higher concentrations of benzene.
Table 8.1. Comparison of soil gas data from sampling locations between the service station and the adjacent office
of the motel.
Well
MW-39
MW-2
MW-51
Vapor
Point
VW-1
VW-2
VW-3
Screened
Interval
feet
13-28
16-35
13-28
Depth to
Water
feet
19.6
19.6
19.6
Depth
Vapor Point
feet
16
17
18
LNAPL
no
yes
yes
Benzene
|jg/m3
< 6.5E+03
1.1E+06
2.3E+06
4.4E+06
8.5E+06
3.6E+06
TPH
pg/m3
<5.1E+04
2.2E+08
2.8E+08
1.3E+08
1.5E+08
2.8E+08
Methane
pg/m3
< 9.8E+02
2.3E+06
1.9E+06
2.0E+06
1.4E+06
1.6E+06
Service
Station
50 Feet
s
Edge of LNAPL
>0.01 foot thick
in well
Former
USTs
Former
USTs
MW-39
VW-1
VW-2
(M)MW-
© MW-
VW-3
\
W4
2
51
Figure 8.7. Location of sampling points
in Table 8.1.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 73
-------
Table 8.2 compares concentrations of soil gas in
locations near the cafe. Figure 8.8 shows the sampling
locations. Monitoring wells MW-21 and MW-48 are
located outside the area with NAPL while well MW-47
and vapor probe VW-7 are located within the area with
NAPL (Figure 8.8).
Again, the concentrations of benzene, TPH-g and
methane in soil gas were reasonably consistent
between the monitoring wells and the vapor probe.
The concentration of methane was higher in the vapor
probe, but the concentrations of THP-g were higher in
the monitoring wells.
As mentioned above, the analytical protocol used for
the wells had a higher detection limit for benzene. As
a result, benzene was not detected in soil gas from
two of the three wells. In the one well where it was
detected, the concentration of benzene was higher in
the well than in the vapor probe.
Table 8.2. Comparison of soil gas data from sampling locations near the cafe.
Well
MW-47
MW-48
MW-21
Vapor
Point
VW-7
Screened
Interval
feet
13-28
13-28
12-27
Depth to
Water
feet
17.6
15.0
17.1
Depth
Vapor Point
feet
15
LNAPL
yes
no
no
Benzene
|jg/m3
8.5E+04
4.6E+03
<4.1E+04
< 5.9E+04
TPH
pg/m3
1.0E+07
3.1E+06
1.5E+07
1.1E+07
Methane
pg/m3
5.5E+03
9.3E+03
3.5E+04
<8.9E+03
50 Feet
L
Cafe with
Basement
VW-5
Edge of LNAPL
>0.01 foot thick in
well
\
MW-21
Figure 8.8. Location of sampling points
in Table 8.2.
VW-7
MW-47
MW-48
74 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 8.3 compares concentrations of soil gas in
locations near the sleeping rooms of the motel. Figure
8.9 shows the sampling locations. Monitoring wells
MW-41 and MW-42 are within the fringe of the area
with NAPL while vapor probe VW-8 is just outside of
the area with NAPL (Figure 8.9).
The concentrations of methane and TPH-g were much
higher in the soil gas from the monitoring wells.
Benzene was not detected in soil gas from the wells.
It is not possible to compare the recovery of benzene
from the wells and the vapor probe at this location.
Results at the Hal's Site in Green River, UT show that
the vapor concentrations of benzene, total
petroleum hydrocarbons and methane measured
in the conventional monitoring wells were similar
to the concentrations measured in the deep vapor
points. When the conventional ground water wells
and the soil vapor probes were located above
LNAPL, the concentration of benzene and total
petroleum hydrocarbons in soil gas produced from the
conventional monitoring wells and the concentrations
in soil gas produced by the deepest vapor probes were
within an order of magnitude. At this site, the approach
of sampling soil gas from conventional ground water
monitoring wells can be as effective as sampling from
vapor probes.
Table 8.3. Comparison of soil gas data from sampling locations near the sleeping rooms of the motel.
Well
MW-41
MW-42
Vapor
Point
VW-8
Screened
Interval
feet
12-28
13-28
Depth to
Water
feet
14.5
16.9
Depth
Vapor Point
feet
15
LNAPL
yes
yes
Benzene
pg/m3
< 2.5E+04
<7.0E-K)3
4.9E+00
TPH
|jg/m3
7.0E+06
7.3E+06
9.0E+02
Methane
|jg/m3
1.7E+05
6.7E+04
< 9.8E+02
Edge of LNAPL
>0.01 foot thick
in well
(M
MW-41
Figure 8.9. Location of sampling points
in Table 8.3.
VW-8
Cafe with
Basement
MW-42
50 Feet
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 75
-------
9.0 REFERENCES
Abreu, L.D.V., and P.C. Johnson. 2005. Effect of vapor
source-building separation and building construction on
soil vapor intrusion as studied with a three-dimensional
numerical model. Environmental Science and Technology
39(12): 4550-4561.
Abreu, L.D.V., and P.C. Johnson. 2006. Simulating the
effect of aerobic biodegradation on soil vapor intrusion
into buildings: Influence of degradation rate, source
concentration, and depth. Environmental Science and
Technology 40 (7): 2304-2315.
Abreu, L.D.V., R. Ettinger, and T. McAlary. 2009a.
Simulated Soil Vapor Intrusion Attenuation Factors
including Biodegradation for Petroleum Hydrocarbons.
Ground Water Monitoring & Remediation 29(1): 105-117.
Abreu, L.D.V., R. Ettinger, and T. McAlary. 2009b.
Simulating the Effect of Aerobic Biodegradation on Soil
Vapor: Evaluation of Low Strength Sources Associated
with Dissolved Gasoline Plumes Intrusion into Buildings.
The American Petroleum Institute. Publication Number
4775. Available at http://www.api.org/Environment-
Health-and-Safety/Clean-Water/Ground-Water/Vapor-
Intrusion/~/link.aspx?_id=4689DFDA5BFC4DAA9ADDA
CC31011247 l&_z=z
Abreu, L., C.C. Lutes andE.M. Nichols. 2013. 3-D
Modeling of Aerobic Biodegradation of Petroleum Vapors:
Effect of Building Area Size on Oxygen Concentration
below the Slab. EPA-510-R-13-002. Office of Underground
Storage Tanks. Available at http://www.epa.gov/oust/cat/
pvi/building-size-modeling.pdf
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. The American Petroleum
Institute. Publication Number 4741. Available at http://www.
api.org/ehs/groundwater/lnapl/upload/474 Ifinall 11805 .pdf
API. 2012. User's Manual Bio Vapor: A 1-D Vapor Intrusion
Model with Oxygen-Limited Aerobic Biodegradation.
Version 2.1. November 2012. The American Petroleum
Institute. Available at: http://www.api.org/environment-
health-and-safety/clean-water/ground-water/vapor-intrusion/
biovapor-form
ASTM. 2010. Standard Test Method for Analysis of Natural
Gas by Gas Chromatography. American Society for Testing
and Materials Designation D 1945-03(2010). Available at
http ://www. astm. org/Standards/D 1945. htm
CRC CARE. 2013. Petroleum Hydrocarbon Vapor Intrusion
Assessment: Australian Guidance. Cooperative Research
Centre for Contamination Assessment and Remediation
of the Environment, Technical Report no. 23. July 2013.
Available: http://www.crccare.com/823C21AO-F98D-llE2-
B4EA005056B60026
Davis, G.B., B.M. Patterson, andM.G. Trefry. 2009.
Evidence for instantaneous oxygen-limited biodegradation
of petroleum hydrocarbon vapors in the subsurface. Ground
Water Monitoring & Remediation 29(1): 126-137.
DeVaull, G.E. 2007. Indoor vapor intrusion with oxygen-
limited biodegradation for a subsurface gasoline source.
Environmental Science & Technology 41, 3241-3248.
DeVaull, G. E. 2011. Biodegradation Rates for Petroleum
Hydrocarbons in Aerobic soils: A Summary of Measured
Data. D-50, in H. V Rectanus and R. Sirabian (Chairs),
Bioremediation and Sustainable Environmental
Technologies-2011. International Symposium on
Bioremediation and Sustainable Environmental
Technologies, Reno, Nevada, June 27-30. ISBN 978-0-
9819730-4-3. Battelle Memorial Institute, Columbus, OH.
H&P Mobile Geochemistry, Inc., Carlsbad, CA. Instructions
for Collecting Soil Vapor into Summa Canisters. Available
athttp://www.pstif.org/apps/soil_vapor_sampling_handouts.
pdf
ITRC. 2014. Petroleum Vapor Intrusion: Fundamentals of
Screening, Investigation, and Management. PVI-1. Interstate
Technology & Regulatory Council, Washington, D.C.
Petroleum Vapor Intrusion Team. Available at:
www.itrcweb.org/PetroleumVI-Guidance.
Jewell, K.P and J.T Wilson. 2011. A new screening
method for methane in soil gas using existing groundwater
monitoring wells. Ground Water Monitoring & Remediation
31(3): 82-94. Available at http://www.epa.gov/ada/pubs/
methane_screening_soil_gas .pdf
Johnson, P.C. and R.A. Ettinger. 1991. Heuristic Model for
Predicting the Intrusion Rate of Contaminant Vapors into
Buildings. Environmenal Science & Technology 25. 1445-
1452.
Johnson, PC., M.W Kemblowski, andR.L. Johnson. 1998.
Assessing the significance of subsurface contaminant vapor
migration to enclosed spaces: Site-specific alternatives
to generic estimates. The American Petroleum Institute.
Publication Number 4674. Available at http://www.api.org/
environment-health-and-safety/clean-water/ground-water/
vapor-intrusion
76 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Johnson, P.C., M.W. Kemblowski, and R.L. Johnson. 1999.
Assessing the significance of subsurface contaminant vapor
migration to enclosed spaces: Site-specific alternatives
to generic estimates. Journal of Soil Contamination 8(3):
389^21.
Jury, W.A., W.F. Spencer and W.J. Farmer. 1983. Behavior
assessment model for trace organics in soil: I. Model
description. Journal of Environmental Quality 12, 558-564.
Kirchstetter, T.W., B.C. Singer, R.A. Harley, G.R. Kendall
and J.M. Hesson. 1999. Impact of California Reformulated
Gasoline on Motor Vehicle Emissions. 2. Volatile organic
compound speciation and reactivity. Environmental Science
& Technology 33(2):329-336.
Lahvis, M.A., I. Hers, R.V Davis, J. Wright and G.E.
DeVaull. 2013. Vapor Intrusion Screening at Petroleum
UST Sites. Groundwater Monitoring & Remediation 33(2):
53-67.
Ma, J., H. Luo, G.E. DeVaull, W.G. Rixey and P.J. J.
Alvarez. 2014. Numerical Model Investigation for Potential
Methane Explosion and Benzene Vapor Intrusion Associated
with High-Ethanol Blend Releases. Environmental Science
& Technology 48: 474-481.
McAlary, T.A., P. Nicholson, H. Groenevelt, and D.
Bertrand. 2009. A case study of soil-gas sampling in silt and
clay-rich (low permeability) soils. Ground Water Monitoring
and Remediation 29: 144-152.
McHugh, T., R. Davis, G. Devaull, H. Hopkins, J. Menatti,
and T. Peargin. 2010. Evaluation of Vapor Attenuation
at Petroleum Hydrocarbon Sites: Considerations for
Site Screening and Investigation. Soil and Sediment
Contamination 19:725-745.
NRC. 1983. Risk assessment in the federal government:
Managing the process. Committee on the Institutional
Means for Assessment of Risks to Public Health,
Commission on Life Sciences, the National Research
Council. Washington, DC: National Academy Press.
Orchard, V.A., andF.J. Cook. 1983. Relationship between
Soil Respiration and Soil Moisture. Soil Biology and
Biochemistry. 15(4): 447-453.
Ostendorf, D.W, andD.H. Kampbell. 1991. Biodegradation
of hydrocarbon vapors in the unsaturated zone. Water
Resources Research 27(4): 453-462.
Ririe, G.T., R.E. Sweeney, and S.J. Daugherty. 2002. A
comparison of hydrocarbon vapor attenuation in the field
with predictions from vapor diffusion models. Soil and
Sediment Contamination 11, 529-544.
Roggemans, S., C.L. Bruce, PC. Johnson, and R.L. Johnson.
2001. Vadose zone natural attenuation of hydrocarbon
vapors: an empirical assessment of soil gas vertical profile
data. API Soil and Groundwater Research Bulletin 15,
American Petroleum Institute. Washington, D.C. Available
athttp://www.api.org/ehs/groundwater/upload/15_bull.pdf
U.S. EPA. 1984. Method TO-3, Revision 1.0, Method for the
Determination of Volatile Organic Compounds in Ambient
Air using Cryogenic Preconcentration Techniques and
Gas Chromatography with Flame lonization and Electron
Capture Detection. April, 1984. Available at http://www.
epa.gov/ttnamtil/files/ambient/airtox/to-3.pdf
U.S. EPA. 1992. Guidelines for Exposure Assessment.
EPA/600/Z-92/001. Federal Register 7(104):22888-
22938. May 29, 1992. Available at http://www.epa.
gov/raf/publications/pdfs/GUIDELINES_EXPOSURE_
ASSESSMENT.PDF
U.S. EPA. 1994. General Field Sampling Guidelines,
SOP#:2001, REVJ:0.0. Environmental Response Team,
United States Environmental Protection Agency. Available
at: http://www.epa.gov/region6/qa/qadevtools/mod5_sops/
misc_field_procudures/ertsop2001fieldsampling-guide.pdf
U.S. EPA. 1996. Soil Gas Sampling, SOP#:2042,
REV#:0.0. Environmental Response Team, United States
Environmental Protection Agency. Available at: http://www.
dem. ri. gov/pubs/sops/wmsr2042 .pdf
U.S. EPA. 1999a. Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient
Air, Second Edition: Compendium Method TO-15,
Determination Of Volatile Organic Compounds (VOCs) In
Air Collected In Specially-Prepared Canisters And Analyzed
By Gas Chromatography/ Mass Spectrometry (GC/MS).
EPA/625/R-96/010b. Available at http://www.epa.gov/
ttnamti 1/files/ambient/airtox/to-15r.pdf
U.S. EPA. 1999b. Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient
Air, Second Edition: Compendium Method TO-14A,
Determination Of Volatile Organic Compounds (VOCs)
In Ambient Air Using Specially Prepared Canisters With
Subsequent Analysis By GasChromatography. EPA/625/R-
96/0lOb. Available at http://www.airtoxics.com/literature/
methods/Method%20TO-14A.pdf
U.S. EPA. 1999c. Compendium of Methods for the
Determination of Toxic Organic Compounds in Ambient
Air, Second Edition: Compendium Method TO-17,
Determination of Volatile Organic Compounds in Ambient
Air Using Active Sampling Onto Sorbent Tubes. EPA/625/
R-96/010b. Available at http://www.epa.gov/ttnamtiI/files/
ambient/airtox/to- 17r.pdf
GW Issue 77
-------
U.S. EPA. 2001. Soil Gas Sampling. Standard Operating
Procedure 2042. Available at http://www.clu-in.org/
download/ert/2042-ROOxx.pdf
U.S. EPA. 2002. OSWER Draft Guidance for Evaluating
the Vapor Intrusion to Indoor Air Pathway from Ground
water and Soils (Subsurface Vapor Intrusion Guidance) EPA
530-D-02-004. Available at http://www.epa.gov/epawaste/
hazard/correctiveaction/eis/vapor/complete.pdf
U.S. EPA. 2011. Petroleum Hydrocarbons And Chlorinated
Hydrocarbons Differ In Their Potential For Vapor Intrusion.
Office of Underground Storage Tanks, Washington, D.C.
20460, September 2011, www.epa.gov/oust. Available at
http ://www. epa. gov/oust/cat/pvi/pvicvi.pdf
U.S. EPA. 2012. Conceptual Model Scenarios for the
Vapor Intrusion Pathway (EPA 530-R-10-003). Office of
Solid Waste and Emergency Response U.S. Environmental
Protection Agency Washington, DC Available at http://
www.epa.gov/oswer/vaporintrusion/documents/vi-cms-
vllfmal-2-24-2012.pdf
U.S. EPA. 2013a. Evaluation of Empirical Data to Support
Soil Vapor Intrusion Screening Criteria for Petroleum
Hydrocarbon Compounds. EPA-510-R-13-001. Office of
Underground Storage Tanks. Available at http://www.epa.
gov/OUST/cat/pvi/PVI_Database_Report.pdf
U.S. EPA. 2013b. [DRAFT] Technical Guide for Addressing
Petroleum Vapor Intrusion at Leaking Underground Storage
Tank Sites. (EPA-510-R-13-xxx) Office of Underground
Storage Tanks. Available at http://www.epa.gov/oust/cat/
pvi/petroleum-vapor-intrusion-review-draft-04092013 .pdf
U.S. EPA. 2014a. What Is The History of The Federal
Underground Storage Tank Program? Web page Available
at http://www.epa.gov/oust/faqs/genesis 1 .htm
U.S. EPA. 2014b. PVI Database Export. Available at http://
www.epa.gov/oust/cat/pvi/
U.S. EPA. 2014c. Integrated Risk Information System
(IRIS). Available at http://www.epa.gov/iris/
U.S. EPA. 2014d. Method 3C-Determination of Carbon
Dioxide, Methane, Nitrogen, and Oxygen From Stationary
Sources. Available at http://www.caslab.com/EPA-
Methods/PDF/m-03c.pdf
U.S. EPA. 2014e. Vapor Intrusion Screening Level (VISL)
Calculator. United States Environmental Protection Agency,
Office of Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology Innovation.
Available at http://www.epa.gov/oswer/vaporintrusion/
guidance.html
U.S. EPA. 2014f. Vapor Intrusion Screening Level (VISL)
Calculator User's Guide. United States Environmental
Protection Agency, Office of Solid Waste and Emergency
Response, Office of Superfund Remediation and Technology
Innovation. Available at http://www.epa.gov/oswer/
vaporintrusion/documents/VISL-UsersGuide.pdf
U.S. EPA. 2014g. PVIScreen Model. In development,
contact Weaver.Jim@epa.gov.
Wilson, IT., K. Jewell, C. Paul, R. Davis, and J. Menatti.
2009. Biodegradation of Petroleum Hydrocarbon Vapors in
the Vadose Zone. Presented at the 2009 Annual National
Tanks Conference & Exposition, Sacramento, CA, March
30, 2009. Available at: http://www.neiwpcc.org/tanks2009/
presentations/Wilson%20Vapor%20Intrusion%20Mon.pdf
Wilson, J.T, J. W. Weaver, and H. White. 2012a. 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.
EPA/600/R-13/047. Available at http://www.epa.gov/
OUST/cat/pvi/epa600rl3047.pdf
Wilson, J.T, M. Toso, D. Mackay, N. de Sieyes and G.E.
DeVaull. 2012b. What's the Deal with Methane at LUST
Spill Sites? Part 1. LUSTLine Bulletin 72, February 2013,
pages 6-8 and 13.
Wilson, J.T, M. Toso, D. Mackay, N. de Sieyes and G.E.
DeVaull. 2013. What's the Deal with Methane at LUST
Spill Sites? Part 2: Vapor Intrusion. LUSTLine Bulletin 72,
February 2013, pages 5-11 and 21.
78 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
10.0 QUALITY ASSURANCE, METHODS
AND PROCEDURES
Data presented in the manuscript was either obtained
from peer reviewed documents published elsewhere
or under a Ground Water and Ecosystems Restoration
Division Quality Assurance Project Plan (GWERD
QAPP). This section provides some of the details
of experimental procedures and provides the
documentation for quality assurance parameters for
data generated under a GWERD QAPP. Data were
developed under three different Quality Assurance
Project Plans.
The data on chemicals in soil gas (Table 4.1) were
generated under the Project Plan titled: Production of
methane and fatty acids as degradation products of
ethanol and petroleum at UST sites in Oklahoma, third
amendment to task 10013/G-15667. All other data on
chemicals in soil gas were generated under the Project
Plan titled: Production of methane and fatty acids as
degradation products of ethanol and petroleum at UST
sites in Oklahoma, fourth amendment task 10013/G-
15667.
Data on concentrations of total petroleum
hydrocarbons in the range of gasoline (TPH-g),
also called gasoline range organics (GRO), in core
samples were generated under the Project Plan titled:
Retrospective Evaluation of a Surfactant Flush to
Remove Gasoline Contamination from an Aquifer at
Golden, Oklahoma, first Amendment, task 14393.
For convenience, the discussion of data quality is
organized by individual data tables or figures.
10.1 Section 4 and Appendix A
70.7.7 So/7 gas components in Table 4.7
Soil gas was pumped from conventional ground water
monitoring wells using the protocol in Jewell and
Wilson (2011). A schematic of the sampling train
used to collect the samples is presented in Figure 10.1.
Rota meter
(flow rate meter)
Figure 10.1. Sampling train to provide soil gas to a field meter and a gas sampling port.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 79
-------
Table 10.1 describes the performance of the quality
assurance parameters during the analysis of the
samples.
The continuing calibration checks and the second
source standards were within +/- 15% of the nominal
value. Carbon dioxide was detected in the field blanks,
but at low concentrations, equal to 0.06% and 0.09%
by volume. The lowest concentration of carbon
dioxide in Table 4.1 was 11.7%. The carbon dioxide
in the field blanks do not affect the interpretation of
the data, and the data were used as intended. The
concentration in the other blanks were less than
the detection limit. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
The purpose of Table 4.1 was to compare two samples
where a leak was suspected in the sampling train
(MW-2B and GMW-1B) to a sample where the
leak was not evident (MW-9). The relative percent
difference for oxygen in the field duplicates of sample
MW-9 was above the acceptance criteria of+/- 15%,
but was much less than the relative percent difference
in samples MW-2B and GMW-1B. The data quality
were acceptable for the intended purpose, to illustrate
the effect of a leak on the ratio of oxygen to nitrogen.
Hydrogen was included in the sample to allow a
calculation of the balance of total gas. The actual
values of the hydrogen data were not interpreted.
Table 10.1. Quality assurance parameters from laboratory-based analyses of soil gas (applies to data in Table 4.1).
Analyte
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Acetylene
Acetylene
Acetylene
Acetylene
Propane
Propane
Propane
Propane
/so-Butane
/so-Butane
/so-Butane
QA Parameter
ccc1
ss2
GB3
FB4
CCC
SS
GB
FB
CCC
SS
GB
FB
CCC
SS
GB
FB
CCC
SS
GB
FB
CCC
SS
GB
Number of Analyses
3
1
4
2
2
1
4
2
2
1
4
2
2
1
4
2
2
1
4
2
2
1
4
Range of Values
101% to 105% Recovery
105% Recovery
All < 2.0 ppm (v/v)*
All < 2 ppm (v/v)*
100% to 103% Recovery
102% Recovery
All < 3.3 ppm (v/v)*
All < 3.3 ppm (v/v)*
95.9% to 98.7% Recovery
102% Recovery
AIK 1.8 ppm (v/v)*
AIK 1.8 ppm (v/v)*
95.7% to 96% Recovery
98.1% Recovery
All < 2.4 ppm (v/v)*
All < 2.4 ppm (v/v)*
98% to 100% Recovery
101% Recovery
AIK 1.7 ppm (v/v)*
AIK 1.7 ppm (v/v)*
96.5% to 98.3% Recovery
98.7% Recovery
AIK 10 ppm (v/v)**
80 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.1. Quality assurance parameters from laboratory-based analyses of soil gas (applies to data in Table 4.1)
(continued).
Analyte
Butane
Butane
Butane
Butane
/so-Pentane
/so-Pentane
/so-Pentane
Pentane
Pentane
Pentane
C6+
C6+
C6+
Benzene
Benzene
Benzene
Oxygen
Oxygen
Oxygen
Nitrogen
Nitrogen
Nitrogen
Carbon Dioxide
Carbon Dioxide
Carbon Dioxide
Carbon Dioxide
Hydrogen
Hydrogen
Hydrogen
QA Parameter
ccc
ss
GB
FB
CCC
SS
GB
CCC
SS
GB
CCC
SS
GB
CCC
SS
GB
CCC
SS
GB
CCC
SS
GB
CCC
SS
GB
FB
CCC
SS
GB
Number of Analyses
2
1
4
2
2
1
4
2
1
4
2
1
4
2
1
4
2
1
4
2
1
4
3
1
4
2
2
1
4
Range of Values
98.6% to 100% Recovery
100% Recovery
AIK1.44ppm(v/v)*
AIK1.44ppm(v/v)*
96.1% to 98.1% Recovery
98.3% Recovery
All < 2 ppm (v/v) **
95.4% to 101% Recovery
95.1% Recovery
All < 2 ppm (v/v) **
106% to 107% Recovery
110% Recovery
All < 2 ppm (v/v) **
111% to 111% Recovery
11 2% Recovery
All < 2 ppm (v/v) **
101% to 102% Recovery
104% Recovery
AIK1251ppm(v/v)*
98.6% to 99.4% Recovery
101% Recovery
All < 5 ppm (v/v) **
94.4% to 109% Recovery
97.5% Recovery
All < 2.52 ppm (v/v)*
6. 18 and 9. 16 ppm (v/v)
101% to 103% Recovery
106% Recovery
All < 3.03 ppm (v/v)*
*
-------
Table 10.2. Relative percent difference between duplicate samples of laboratory-based analysis of soil gas.
(Applies to data in Table 4.1).
Monitoring Well
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers GMW-1
Antlers GMW-1
Antlers GMW-1
Antlers GMW-1
Antlers GMW-1
Antlers GMW-1
Date Collected
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
2/13/2012
Soil Gas Component
Oxygen
Methane
Hydrogen
Total Petroleum
Hydrocarbons*
Nitrogen
Carbon Dioxide
Oxygen
Methane
Hydrogen
Total Petroleum
Hydrocarbons*
Nitrogen
Carbon Dioxide
Oxygen
Methane
Hydrogen
Total Petroleum
Hydrocarbons*
Nitrogen
Carbon Dioxide
Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Relative Percent
Difference
127
29
NC
44
199**
17.2
27.1
2.1
NC
1.3
6.3
7.3
4.5
2.4
NC
0.75
199***
3.4
NC = Not calculated, one or both of the pair is < MDL
*Sum of concentrations of individual hydrocarbons in ppm (v/v).
^Concentration in MW-2A sample was 31.4 ppm, concentration in MW-2B sample was 66200 ppm, indicating a leak in the MW-2B sample. Data in Table
4.1 from sample MW-2B.
***Concentration in GMW-1 A sample was 38.9 ppm, concentration in GMW-1 B sample was 241000 ppm, indicating a leak in the MW-1B sample. Data in
Table4.1 from sample GMW-1 B.
82 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
70.7.2 Quality Assurance of Carbon Dioxide,
Oxygen and Methane using Field Meter in
Figure 4.4, Figure 4.5 and Appendix A
Oxygen, Carbon Dioxide and Methane in the soil gas
sample were determined using a Landtec GEM2000
gas analyzer (Landtec Inc., 850 S. Via Lata, Suite 112,
Colton, CA 92324).
The Landtec GEM2000 gas analyzer was calibrated
daily prior to use. The gas analyzer was calibrated
again whenever the ambient air temperature changed
more than 20 degrees Fahrenheit. Additionally, the gas
analyzer was checked against the calibration standard
any time the gas analyzer was turned off.
The GEM2000 gas analyzer was calibrated using
standard gas samples that were acquired from LandTec
(850 S. Via Lata, Suite 112, Colton, CA 92324). The
carbon dioxide and methane sensors were calibrated
with LandTec's commercial standard containing 50%
methane, 35% carbon dioxide, and balance nitrogen
gas. Ambient air was used to calibrate the oxygen
sensor. The zero response for the carbon dioxide senor
and the methane sensor was calibrated against a gas
standard that had 4% oxygen and the balance nitrogen.
The zero response for the oxygen sensor was calibrated
against the standard containing 50% methane, 35%
carbon dioxide, and balance nitrogen gas.
The following calibration procedure was located in the
"calibration" window. After all channels were zeroed,
the calibration gas standard mixture containing 50%
methane, 35% carbon dioxide and balance nitrogen is
supplied to the instrument. When readings stabilize,
the "SPAN CH4" channel was selected allowing the
reading to adjust to the input parameter. Then the
"SPAN CO2" channel was selected allowing the
reading to adjust to the input parameter. Next, ambient
oxygen was supplied to the instrument. When readings
stabilize, the "SPAN O2" channel was selected
allowing the reading to adjust to the input parameter.
This procedure completed the GEM2000 calibration.
As mentioned above, calibration was carried out using
a commercial gas standard with 50% methane, 35%
carbon dioxide and the balance nitrogen mixture, and
with ambient air that contains 20.9% oxygen. At some
sites, the calibration was compared against a second
source standard mixture containing 2.5% methane,
20.1% carbon dioxide, and 10.1% oxygen. Calibration
checks, also known as "Bump Cal", were used
periodically throughout the sampling process to ensure
meter calibration was within acceptable parameters.
The acceptable range was +/- 1% of display reading.
Table 10.3 presents the results from the field
calibration of the gas analyzer.
Four of the calibration checks were outside the
acceptable range, but were within +/- 2% of the
nominal value of the standard. In the figures, the meter
readings were plotted to reveal changes over time as
the concentrations of the compounds in the sampled
gas approached equilibration. A range of
+/- 2% of the nominal value of the standard is useful
for this purpose. The quality of the data were
acceptable for the intended purpose, and all the data
were used in the figures.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 83
-------
Table 10.3. Results of field calibration and calibration checks of the Landtec GEM2000 gas analyzer.
Figure
4.4
4.5
A.1
A.2
A.3
Well
MW-2
MW-4
MW-9
MW-4
SVE-NE
Site
EZGo
Kev's Auto
Visocan Site
Visocan Site
Noon's Store
Location
Antlers, OK
Helena, MT
Helena, MT
Helena, MT
Helena, MT
Date
5/30/12
10/26/12
10/25/12
10/26/12
10/25/12
Calibration
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Nominal Value
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
Reading
51.9%*
35.1%
20.9%
49.8%
35.1%
20.9%
50.9%
35.1%
20.0%
2.4%
20.2%
10.7%
50.1%
35.4%
20.9%
2.3%
19.6%
10.6%
50.9%
35.1%
20.0%
2.4%
20.2%
10.7%
50.1%
35.4%
20.9%
2.3%
19.6%
10.6%
84 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.3. Results of field calibration and calibration checks of the Landtec GEM2000 gas analyzer. (Continued).
Figure
A.4
A.5
A.6
A.7
A.8
Well
SVE-SE
HMM-6
GMW-1
MW-9
MW-1
Site
Noon's Store
Noon's Store
EZGo
EZGo
Miller Mart
Location
Helena, MT
Helena, MT
Antlers, OK
Antlers, OK
Wapanucka,
OK
Date
10/24/12
10/25/12
5/30/12
5/30/12
4/29/13
Calibration
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Check
Check
Check
Standard
Standard
Standard
Nominal Value
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
50% CH4
35% C02
20.9% 02
2.51 %CH4
20.1%C02
10.1%02
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
20.9% 02
50% CH4
35% C02
4.0% 02
Reading
50.1%
36.9%*
21.1%
2.4%
19.6%
10.7%
50.1%
35.4%
20.9%
2.3%
19.6%
10.6%
51.9%*
35.1%
20.9%
49.8%
35.1%
20.9%
51.9%*
35.1%
20.9%
49.8%
35.1%
20.9%
WAR**
WAR**
WAR**
" Values outside the acceptable range of +/-1 % of the nominal value of the standard.
"* WAR means Within Acceptable Range. Field notebook indicates that values were within the acceptable range, but a value was not recorded.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 85
-------
70.7.3 Benzene and TPH-g in Figure 4.4,
Figure 4.5 and Appendix A
Soil gas samples were acquired by water displacement
into 165 ml serum vials. The vials did not drain
completely and always retained a few drops of water.
Bacteria in the water in contact with the gas samples
might have degraded the hydrocarbons in the gas
samples. The water contained l%trisodium phosphate
as a preservative.
The vials were returned to the Kerr Center for
analysis by micro gas chromatography with thermal
conductivity detection as is described in Jewell and
Wilson (2011).
Data in Figure 4.4, Figure A.6 and A.I are from wells
that were installed at the direction of the Petroleum
Storage Tank Division of the Oklahoma Corporation
Commission at an EZ Go service station in Antlers,
OK. The quality assurance parameters for data in
Figure 4.4, Figure A.6 and Figure A.7 are presented in
Table 10.4.
The continuing calibration checks and the second
source standards were within +/- 15% of the nominal
value. The concentration in all the blanks were less
than the detection limit. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
Table 10.4. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure 4.4, Figure A.6 and Figure A.7.
Analyte
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Acetylene
Acetylene
Acetylene
Acetylene
Propane
Propane
Propane
Propane
Butane
Butane
Butane
Butane
QA Parameter
ccc1
ss2
LB3
FB4
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number of Analyses
7
1
12
2
6
1
12
2
6
1
12
2
6
1
12
2
6
1
12
2
6
1
12
2
Range of Values
90.6% to 107% Recovery
93.9% Recovery
AIK1.98ppm(v/v)*
AIK1.98ppm(v/v)*
94.6% to 99.9% Recovery
95.5% Recovery
All < 3.29 ppm (v/v)*
All < 3.29 ppm (v/v)*
93.7% to 101% Recovery
92.7% Recovery
AIK 1.75 ppm (v/v)*
AIK 1.75 ppm (v/v)*
90.8% to 98.5% Recovery
95.4% Recovery
All < 2.39 ppm (v/v)*
All < 2.39 ppm (v/v)*
9 1.9% to 107% Recovery
85.7% Recovery
AIK 1.74 ppm (v/v)*
AIK 1.74 ppm (v/v)*
105% to 110% Recovery
98.2% Recovery
AIK 1.44 ppm (v/v)*
AIK 1.44 ppm (v/v)*
*
-------
Table 10.5 presents the relative percent difference
between duplicate laboratory samples for the data in
Figure 4.4, Figure A.6 and Figure A.7.
The relative percent difference of the lab duplicates
were all less than 15%. Each well was sampled four
times at each time period. There were four field
duplicates. The individual data are plotted on the
figures. In Figure A.6, the data at the initial sampling
interval were variable, ranging over as much as a
factor of four. In subsequent sampling periods the
range of values were less than 150% to 50% of their
mean. The figures plot a line connecting the mean
of the four samples at each time period. The figures
were intended to illustrate the tendency to reach
equilibration. The data in the later time periods were
closely clustered about their mean. The data were of
acceptable quality for the intended purpose and all of
the data were used.
Table 10.5. Relative percent difference of laboratory duplicates on samples of soil gas analyzed in the laboratory for data in
Figure 4.4, Figure A.6 and Figure A.7.
Monitoring Well
Antlers MW-1-10min-F in
Figure A.6
Antlers MW-1-10min-F in
Figure A.6
Antlers MW-2-0 min-Ain
Figure 4.4
Antlers MW-2-0 min-Ain
Figure 4.4
Antlers MW-2-30 min-B in
Figure 4.4
Antlers MW-2-30 min-B in
Figure 4.4
Antlers MW-9-10min-E in
Figure A.7
Antlers MW-9-10min-E in
Figure A.7
Antlers MW-9-40 min-F in
Figure A.7
Antlers MW-9-40 min-F in
Figure A.7
Date Collected
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
5/30/12
Soil Gas Component
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Relative Percent
Difference
0.18
0.49
0.43
0.23
0.67
0.54
0.12
0.48
0.78
2.5
*Sum of concentrations of individual petroleum hydrocarbons plus methane in pg/m3.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 87
-------
Data in Figure A.8 are from a well that was installed at The continuing calibration checks and the second
the direction of the Petroleum Storage Tank Division
of the Oklahoma Corporation Commission at the
Miller Mart service station in Wapanucka, OK. The
quality assurance parameters for data in Figure A.8 are
presented in Table 10.6.
source standards were within +/- 15% of the nominal
value. The concentration in all the blanks were less
than the detection limit. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
Table 10.6. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure A.8.
Analyte
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Acetylene
Acetylene
Acetylene
Acetylene
Propane
Propane
Propane
Propane
Propylene
Propylene
Propylene
Propylene
/so-Butane
/so-Butane
/so-Butane
/so-Butane
Butane
Butane
Butane
Butane
QA Parameter
ccc1
ss2
LB3
FB4
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
Range of Values
106% to 111% Recovery
107% Recovery
AIK1.76ppm(v/v)*
<1.76ppm(v/v)*
103% to 105% Recovery
104% Recovery
AIK1.22ppm(v/v)*
<1.22ppm(v/v)*
98.2% to 100% Recovery
100% Recovery
AIK3.1ppm(v/v)*
<3.1 ppm(v/v)*
99% to 101% Recovery
98.6% Recovery
All <2.51 ppm (v/v)*
<2.51 ppm (v/v)*
107% to 109% Recovery
110% Recovery
AIK 1.1 6 ppm (v/v)*
< 1.1 6 ppm (v/v)*
108% to 111% Recovery
110% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
99.7% to 102% Recovery
102% Recovery
AIK 10 ppm (v/v)**
< 10 ppm (v/v)**
104% to 107% Recovery
107% Recovery
AIK 0.72 ppm (v/v)
< 0.72 ppm (v/v)
GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.6. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure A.8. (Continued).
Analyte
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
1-Butene
1-Butene
1-Butene
1-Butene
/'so-Butylene
/'so-Butylene
/'so-Butylene
/'so-Butylene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
/'so-Pentane
/'so-Pentane
/'so-Pentane
/'so-Pentane
Pentane
Pentane
Pentane
Pentane
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
1-Pentene
1-Pentene
1-Pentene
1-Pentene
QA Parameter
ccc
ss
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number of Analyses
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
3
1
6
1
Range of Values
104% to 106% Recovery
106% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
110% to 11 3% Recovery
114% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
11 2% to 11 4% Recovery
115% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
108% to 111% Recovery
110% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
101% to 103% Recovery
103% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
89.3% to 108% Recovery
108% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
98.5% to 101% Recovery
110% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
107% to 110% Recovery
101% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
108% to 110% Recovery
111% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 89
-------
Table 10.6. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure A.8. (Continued).
Analyte
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
Benzene
Benzene
Benzene
Benzene
QA Parameter
ccc
ss
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number of Analyses
3
1
6
1
3
1
6
1
3
1
6
1
Range of Values
108% to 110% Recovery
111% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
92% to 97.8% Recovery
9 1.4% Recovery
All < 2 ppm (v/v) **
< 2 ppm (v/v) **
97.4% to 97.7% Recovery
100% Recovery
All < 2 ppm (v/v) **
2.02 ppm (v/v)
*
-------
Table 10.7 presents the relative percent difference
between duplicate laboratory samples for the data in
Figure A.8.
The relative percent difference of the lab duplicates
were all less than 5%. Each well was sampled four
times at each time period. There were four field
duplicates. The individual data are plotted on the
figures. In all sampling periods the range of values
were less than 150% to 50% of their mean. The figure
plots a line connecting the mean of the four samples at
each time period. The figure was intended to illustrate
the tendency to reach equilibration. The data in the
later time periods were closely clustered about their
mean. The data were of acceptable quality for the
intended purpose and all of the data were used.
Table 10.7. Relative percent difference of laboratory duplicates on samples of soil gas analyzed in the laboratory for data in
Figure A.8.
Monitoring Well
Wapanucka MW-1-20 min-F
in Figure A.8
Wapanucka MW-1-20 min-F
in Figure A.8
Wapanucka MW-1-60 min-F
in Figure A.8
Wapanucka MW-1-60 min-F
in Figure A.8
Date Collected
4/29/2013
4/29/2013
4/29/2013
4/29/2013
Soil Gas Component
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Relative Percent
Difference
0.0
0.64
0.40
0.35
*Sum of concentrations of individual petroleum hydrocarbons plus methane in pg/m3.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 91
-------
Data in Figure 4.5, Figure A. 1, Figure A. 2, Figure
A.3, Figure A.4 and Figure A.5 are from conventional
ground water wells that were installed at the direction
of the Petroleum Release Section of the Montana
Department of Environmental Quality at three sites in
Helena, MT. Soil gas samples were acquired by water
displacement into 165 ml serum vials. The vials were
returned to the Kerr Center for analysis by micro gas
chromatography with thermal conductivity detection as
was described in Jewell and Wilson (2011).
The quality assurance parameters for data in Figure
4.5, Figure A.I, Figure A.2, Figure A.3, Figure A.4 and
Figure A.5 are presented in Table 10.8.
The continuing calibration checks and the second
source standards were within +/- 15% of the nominal
value. The concentration in all the blanks were less
than the detection limit. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
Table 10.8. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure 4.5, Figure A. 1,
Figure A.2, Figure A.3, Figure A.4 and Figure A.5.
Analyte
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Acetylene
Acetylene
Acetylene
Acetylene
Propane
Propane
Propane
Propane
Propylene
Propylene
Propylene
Propylene
/so-Butane
/'so-Butane
/'so-Butane
/so-Butane
QA Parameter
CCC1
ss2
LB3
FB4
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
Range of Values
101% to 107% Recovery
102% to 107% Recovery
AIK1.98ppm(v/v)*
AIK1.98ppm(v/v)*
96.6% to 101% Recovery
96.1% to 101% Recovery
All < 3.29 ppm (v/v)*
All < 3.29 ppm (v/v)*
92.7% to 96.7% Recovery
92.6% to 97.1% Recovery
AIK 1.75 ppm (v/v)*
AIK 1.75 ppm (v/v)*
93.6% to 97.7% Recovery
92.8% to 95.6% Recovery
All < 2.39 ppm (v/v)*
All < 2.39 ppm (v/v)*
98.2% to 101% Recovery
100% to 101% Recovery
AIK 1.74 ppm (v/v)*
AIK 1.74 ppm (v/v)*
97.7% to 101% Recovery
98% to 98.7% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
99.7% to 103% Recovery
101% to 102% Recovery
AIK 10 ppm (v/v)**
AIK 10 ppm (v/v)**
92 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.8. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure 4.5, Figure A.1,
Figure A.2, Figure A.3, Figure A.4 and Figure A.5. (Continued).
Analyte
Butane
Butane
Butane
Butane
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
1-Butene
1-Butene
1-Butene
1-Butene
/'so-Butylene
/'so-Butylene
/so-Butylene
/so-Butylene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
/so-Pentane
/so-Pentane
/so-Pentane
/so-Pentane
Pentane
Pentane
Pentane
Pentane
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
QA Parameter
ccc
ss
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
Range of Values
94.2% to 96.4% Recovery
94.6% to 95.4% Recovery
All < 1 .44 ppm (v/v)
All < 1 .44 ppm (v/v)
96.2% to 98.5% Recovery
96.9% to 97.9% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
100% to 104% Recovery
101% to 103% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
98.8% to 102% Recovery
99.4% to 100% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
97.1% to 100% Recovery
97.5% to 99% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
97.3% to 99.9% Recovery
98.1% to 98.6% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
85.9% to 107% Recovery
85.3% to 101% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
97.5% to 101% Recovery
107% to 109% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
96.3% to 99.8% Recovery
89.6% to 90.3% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 93
-------
Table 10.8. Quality assurance parameters from laboratory-based analyses of soil gas for data in Figure 4.5, Figure A.1,
Figure A.2, Figure A.3, Figure A.4 and Figure A.5. (Continued).
Analyte
1-Pentene
1-Pentene
1-Pentene
1-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
Benzene
Benzene
Benzene
Benzene
QA Parameter
ccc
ss
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
CCC
SS
LB
FB
Number
15
4
30
2
15
4
30
2
15
4
30
2
15
4
30
2
Range of Values
98.8% to 103% Recovery
99.8% to 101% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
98.7% to 102% Recovery
99.4% to 101% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
97.6% to 107% Recovery
91. 2% to 102% Recovery
All < 2 ppm (v/v) **
All < 2 ppm (v/v) **
96% to 97.8% Recovery
96% to 96.9% Recovery
All <2 ppm (v/v) **
All <2 ppm (v/v) **
*
-------
Table 10.9 presents the relative percent difference
between duplicate laboratory samples for the data in
Figure 4.5, Figure A.I, Figure A.2, Figure A.3, Figure
A.4 and Figure A.5.
The relative percent difference of the lab duplicates
were all less than 15%. Each well was sampled four
times at each time period. The individual data are
plotted on the figures. In Figure A.5, the data at the
initial sampling interval were variable, ranging over
as much as a factor often. In subsequent sampling
periods the range of values were less than 150% to
50% of their means. The figures plot a line connecting
the mean of the four samples at each time period. The
figures were intended to illustrate the tendency to reach
equilibration. The data in the later time periods were
closely clustered about their mean. The data were of
acceptable quality for the intended purpose and all of
the data were used.
Table 10.9. Relative percent difference of laboratory duplicates on samples of soil gas analyzed in the laboratory for data in
Figure 4.5, Figure A. 1, Figure A.2, Figure A.3, Figure A.4 and Figure A.5.
Monitoring Well
Kev'sAutoMW-4-20
min-F in Figure 4.5
Kev'sAutoMW-4-20
min-F in Figure 4.5
Visocan Site MW-9-0
min-F in Figure A. 1
Visocan Site MW-9-0
min-F in Figure A. 1
Visocan Site MW-9-40
min-Ain Figure A. 1
Visocan Site MW-4-10
min-B in Figure A.2
Visocan Site MW-4-40
min-E in Figure A.2
Noon's Store HMM-6-10
min-Ain Figure A.5
Noon's Store HMM-6-10
min-Ain Figure A.5
Noon's Store HMM-6-40
min-B in Figure A.5
Noon's Store HMM-6-40
min-B in Figure A.5
Date Collected
10/26/2012
10/26/2012
10/25/2012
10/25/2012
10/25/2012
10/26/2012
10/26/2012
10/25/2012
10/25/2012
10/25/2012
10/25/2012
Soil Gas Component
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Total Hydrocarbons*
Total Hydrocarbons*
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Benzene
Total Hydrocarbons*
Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Relative Percent
Difference
0.27
0.18
0.05
0.43
1.6
1.0
0.16
0.20
0.34
0.0
1.2
*Sum of concentrations of individual petroleum hydrocarbons plus methane in pg/m3.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 95
-------
70.7.4 So/7 gas components in Tables 6.7, 6.2
and 6.3
The quality assurance parameters for data in Tables
6.1, 6.2 and 6.3 are presented in Table 10.10.
The continuing calibration checks and the second
source standards were within +/- 15% of the nominal
value. However, the maximum concentration of
methane in any blank was 384 ppm. The measured
concentrations of methane in wells MW-9 and MW-2
was 400,000 and 630,000 ppm. The methane in the
blank was not a significant fraction of the methane in
the samples.
The maximum measured concentrations of /so-butane,
/'so-pentane, pentane and C6+ compounds in any
blank were 37, 2.1, 3.3 and 470 ppm respectively.
These would correspond to a concentration of TPH
of 2.1E+06 (ig/L. The measured concentrations of
TPH in well MW-9 was 3.0E+07 (ig/L and in well
MW-2 was 4.6E+08 (ig/L (Table 6.1). The measured
concentration of TPH in well MW-9 (3.0E+07 (ig/L)
was then added to the concentration of methane in well
MW-9 (3.0E+08 (ig/L) to calculate a concentration of
total hydrocarbons (3.3E+08 (ig/L), see table 6.2. The
concentration of total hydrocarbons was fit to Figure
6.4 with a precision of one significant digit to estimate
the attenuation factor of benzene. The concentrations
of hydrocarbons in the blanks would not affect the
estimate of the attenuation factor.
The concentrations of oxygen in the blanks was a high
as 3.9% by volume. Using the equation in Section 4.3,
this would require a correction of the hydrocarbon data
by multiplying by a factor of 1.2. The concentration
of total hydrocarbons was fit to Figure 6.4 with a
precision of one significant digit to estimate the
attenuation factor of benzene. The concentrations of
oxygen in the blanks would not affect the estimate of
the attenuation factor. The concentration in all the
other blanks were less than the detection limit.
The quality of the data were acceptable for the
intended purpose, and all the data were used.
Table 10.10. Quality assurance parameters from laboratory-based analyses of soil gas for data in Table 6.1, 6.2 and 6.3.
Analyte
Methane
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Ethane
QA Parameter
ccc1
ss2
LB3
FB4
TB5
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
Number of Analyses
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
Range of Values
92.3% to 98.2% Recovery
96.2% Recovery
All < 0.57 ppm (v/v)*
< 0.57 to 194 ppm (v/v)
159 to 384 ppm (v/v)
97.1% to 102% Recovery
99.1% Recovery
AIK 1.44 ppm (v/v)*
AIK 1.44 ppm (v/v)*
AIK 1.44 ppm (v/v)*
96.5% to 101% Recovery
98.2% Recovery
All < 0.70 ppm (v/v)*
All < 0.70 ppm (v/v)*
All < 0.70 ppm (v/v)*
96 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.10. Quality assurance parameters from laboratory-based analyses of soil gas for data in Table 6.1, 6.2 and 6.3.
(Continued).
Analyte
Acetylene
Acetylene
Acetylene
Acetylene
Acetylene
Propane
Propane
Propane
Propane
Propane
Propylene
Propylene
Propylene
Propylene
Propylene
/so-Butane
/so-Butane
/so-Butane
/so-Butane
/so-Butane
Butane
Butane
Butane
Butane
Butane
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
1-Butene
1-Butene
1-Butene
1-Butene
1-Butene
QA Parameter
ccc
ss
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
Number of Analyses
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
Range of Values
98.1% to 104% Recovery
101% Recovery
AIK1.2ppm(v/v)*
AIK1.2ppm(v/v)*
AIK1.2ppm(v/v)*
97.3% to 101% Recovery
96.8% Recovery
All < 0.60 ppm (v/v)*
All < 0.60 ppm (v/v)*
All < 0.60 ppm (v/v)*
96.5% to 101% Recovery
96.5% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
89.1% to 92.9% Recovery
89.7% Recovery
AIK10ppm(v/v)**
< 10 to 6.5 ppm (v/v)**
AIK10ppm(v/v)**
94.7% to 100% Recovery
95.7% Recovery
All < 0.44 ppm (v/v)*
< 0.44 to 37 ppm (v/v)*
All < 0.44 ppm (v/v)*
94.1% to 98.7% Recovery
94.8% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
95.6% to 101% Recovery
97.1% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 97
-------
Table 10.10. Quality assurance parameters from laboratory-based analyses of soil gas for data in Table 6.1, 6.2 and 6.3.
(Continued).
Analyte
/so-Butylene
/so-Butylene
/so-Butylene
/so-Butylene
/so-Butylene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
/so-Pentane
/so-Pentane
/so-Pentane
/so-Pentane
/so-Pentane
Pentane
Pentane
Pentane
Pentane
Pentane
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
2-Methyl-2-Butene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
1-Pentene
1-Pentene
1-Pentene
1-Pentene
1-Pentene
QA Parameter
ccc
ss
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
Number of Analyses
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
Range of Values
96.4% to 103% Recovery
99.4% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
93.8% to 100% Recovery
96.3% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
94% to 97.6% Recovery
95.3% Recovery
All < 2 ppm (v/v)**
< 2 to 2.1 ppm (v/v)**
All < 2 ppm (v/v)**
95.3% to 102% Recovery
97.3% Recovery
All < 2 ppm (v/v)**
< 2 to 3.3 ppm (v/v)**
All < 2 ppm (v/v)**
94.0% to 96.0% Recovery
94.0% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
97.1% to 101% Recovery
94.8% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
96.5% to 97.8% Recovery
94.5% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
98 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.10. Quality assurance parameters from laboratory-based analyses of soil gas for data in Table 6.1, 6.2 and 6.3.
(Continued).
Analyte
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
C6+ Hydrocarbons
Benzene
Benzene
Benzene
Benzene
Benzene
Oxygen
Oxygen
QA Parameter
ccc
ss
LB
FB
TB
CCC
SS
LB
FB
TB
CCC
SS
LB
FB
TB
FB
TB
Number of Analyses
3
1
4
4
3
3
1
4
4
3
3
1
4
4
3
4
3
Range of Values
96.1% to 98.1% Recovery
94.2% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
91. 7% to 104% Recovery
98.6% Recovery
All < 2 ppm (v/v)**
< 2 to 470 ppm (v/v)**
< 2 to 68.9 ppm (v/v)**
88.6% to 92.4% Recovery
92.7% Recovery
All < 2 ppm (v/v)**
< 2 to 4.9 ppm (v/v)**
All < 2 ppm (v/v)**
4, 130 to 34,800 ppm (v/v)
28,000 to 39,300 ppm (v/v)
*
-------
The quality assurance parameters for data in Tables
6.1, 6.2 and 6.3 are presented in Table 10.11.
The relative percent difference of the field and lab
duplicates were all less than 15%. The data were of
acceptable quality for the intended purpose and all of
the data were used.
Table 10.11. Relative percent difference of field and laboratory duplicates on samples of soil gas analyzed in the laboratory
for data in Table 6.1.
Monitoring Well
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Date Collected
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
Soil Gas Component
Oxygen
Methane
i-Butane
i-Pentane
C6+
Benzene
Methane
i-Butane
i-Pentane
C6+
Benzene
Oxygen
Methane
i-Butane
i-Pentane
C6+
Benzene
Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Relative Percent
Difference
7.7
4.7
2.4
2.0
9.8
5.8
4.6
0.33
0.90
0.67
1.8
7.0
1.2
3.2
2.3
0.7
2.4
100 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
70.7.5 TPH-g in Core Samples Figure 4.7
Core samples were acquired from a spill of motor
gasoline at the former A-1 Movers site in Madison,
Wisconsin. Samples were acquired in May, 2006.
The samples were acquired using the GeoProbe
Macrocore® system. The cores were cut open with
a cleaned or decontaminated hacksaw, plug samples
were taken from the face of the exposed core and then
extracted in the field into a 50% methanol in water
solution. To determine the wet weight of the sample
that was extracted, the vials were weighed and then
weighed again after they received the sample. The
weight of the plug sample was the final weight minus
the initial weight. The wet weight of the plug samples
varied from 38 to 44 grams.
The methanol extracts were spiked into water
and analyzed at Kerr Center for Total Petroleum
Hydrocarbons in the gasoline range (TPH-g)
using purge and trap/gas chromatography. The
concentrations of TPH-g in the extracts were reported
in (ig/L of extract. The method detection limit was
309 (ig/L. The density of methanol is 791 g/L. The
density of the 50% solution of methanol in water used
to extract the core samples was assumed to be 896 g/L.
The mass of the extracting solution was calculated by
weighing each vial empty and again after it received
the extracting solution. The mass of the extracting
solution in grams was divided by 896 g/L to determine
the liters of extracting solution. The volume of the
extracting solution in liters was multiplied by the
concentration in (ig/Lto calculate the mass of TPH-g
in the extract in (ig.
Separate plug samples taken from the same depth
interval as the samples that were extracted in methanol
were taken back to the laboratory and air dried. The
loss on drying was used to calculate the water content
of the plug samples. The moisture of the samples
ranged from 12.5% to 18.4% of the dry weight. The
average water content and the wet weight of the plug
samples were used to calculate the dry weight of
the plug samples that were extracted into the 50%
methanol and water solution.
For each plug sample, the mass of TPH-g (jig) was
divided by the dry weight of the sample (g) to calculate
the concentration of TPH-g in the sample (mg/kg).
The quality assurance parameters for TPH-g data in
Figure 4.7 are presented in Table 10.12.
The continuing calibration checks and the second
source standards were within +/- 15% of the nominal
value. The concentration in all the blanks were less
than the detection limit. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
Table 10.12. Quality assurance parameters from analyses of core extracts for data in Figure 4.7.
Analyte
TPH-g
TPH-g
TPH-g
TPH-g
TPH-g
TPH-g
QA Parameter
ccc1
ss2
LB3
LCS4
TB5
Lab Duplicate
Number of Analyses
15
1
19
3
1
1
Range of Values
96.6% to 11 3% Recovery
96.0% Recovery
All <309 |jg/L in methanol extract
or 0.2 mg/kg soil extracted*
101% to 105% Recovery
<309 |jg/L in methanol extract
or 0.2 mg/kg soil extracted*
RPD=3.52%
* < the limit of quantitation.
1CCC is Continuing Calibration Check.
2SS is a Second Source Standard and is a measure of accuracy.
3LB is a laboratory method blank.
4LCS is laboratory control spike.
5TB is a trip blank.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 101
-------
10.2 Section 5
70.2.7 TPH on Core Samples, Figure 5.3
Core samples were acquired using the GeoProbe
Macrocore® system. The cores were cut open with a
hacksaw, plug samples was taken from the core, and
extracted in the field into 50% methanol in water.
To determine the wet weight of the sample that was
extracted, the vials were weighed and then weighed
again after they received the sample. The weight of
the plug sample was the final weight minus the initial
weight. The wet weight of the plug samples varied
from 38 to 44 grams.
The methanol extracts were spiked into water
and analyzed at Kerr Center for Total Petroleum
Hydrocarbons in the gasoline range (TPH-g) using
purge and trap/gas chromatography and the Modified
Wisconsin Method for Determining Gasoline Range
Organics by Purge and Trap Gas Chromatography-
OI. Modifications to the method included utilizing a
composite Gasoline Standard (Restek #30205) and a
baseline to baseline area sum from the end of methanol
(retention time (RT) -11.5 min.) to a final RT of 25.5
min. The final time corresponds to the last peak eluted
from the composite Gasoline standard.
The concentrations of TPH-g in the extracts were
reported in (ig/L of extract. The quantitation limit
was 50 (ig/L. The density of methanol is 791 g/L.
The density of the 50% solution of methanol in water
used to extract the core samples was assumed to be
896 g/L. The mass of the extracting solution was
calculated by weighing each vial empty and again
after it received the extracting solution. The mass of
extracting solution in grams was divided by 896 g/L to
determine the liters of extracting solution. The volume
of extracting solution in liters was multiplied by the
concentration in (ig/Lto calculate the mass of TPH-g
in the extract in (ig.
Separate plug samples taken from the depth interval
of the samples that were extracted in methanol were
taken back to the laboratory and air dried. The loss
on drying was used to calculate the water content of
the plug samples. The moisture of the samples ranged
from 12.5% to 18.4% of the dry weight. The average
water content and the wet weight of the plug samples
were used to calculate the dry weight of the plug
samples that were extracted into the 50% methanol and
water solution.
For each plug sample, the mass of TPH-g (|ig) was
divided by the dry weight of the sample (g) to calculate
the concentration of TPH-g in the sample (mg/kg).
The quality assurance parameters for TPH data in
Figure 5.3 are presented in Table 10.13.
The surrogate spikes, the continuing calibration
checks, the second source standards, and the laboratory
control spikes were within +/- 15% of the nominal
value. The quality of the data were acceptable for the
intended purpose, and all the data were used.
Table 10.13. Quality assurance parameters from analyses of TPH-g in soil extract (applies to Figures 5.3 and 8.3).
Analyte
FB*-reference
BFB**-reference
TPH-gasoline range
TPH-gasoline range
TPH-gasoline range
QA Parameter
Surrogate Spike
Surrogate Spike
Continuing Calibration Check
Second Source Standard
Laboratory Control Spike
Number
22
22
13
1
5
Range of Values
85.6% to 106% Recovery
92.1% to 11 9% Recovery
90.0% to 11 3% Recovery
102% Recovery
94.6% to 108% Recovery
*Fluorobenzene
* *4-Bromofluorobenzene
102 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.14 presents TPH-g sample data compared
to concentrations of TPH-g in the Trip Blank and the
methanol blank used to extract the solvents.
There were detectable concentrations of TPH-g in the
blanks. Borehole EPA-1 was immediately adjacent to
the multi-depth vapor probes at location VW-7. The
purpose of the sampling was to determine the depth
intervals near the water table that were contaminated
with NAPL, and the greatest depth interval that was
not contaminated. The water table was near 15 feet
below land surface. The concentrations in the blanks
were less than 1% of the concentrations from the deep
core samples near the water table and therefore had
negligible impact on that sample data.
Table 10.14. Concentrations of TPH-g in samples depicted in Figure 5.3 compared to the concentrations of TPH-g in the trip
blanks and methanol blanks.
The highest concentration in any blank was 1.2E+04
(ig/L. If the highest concentration in any blank had
come from a core sample, the concentration in that
core sample would have been approximately 9 mg/
kg. This is less than the criterion of 250 mg/kg. In
the deepest samples where the concentration of TPH
was less than the criterion of TPH<250 mg/kg, the
concentration in the samples was <0.04 mg/kg. The
concentrations in the blanks could not have impacted
the identification of the deepest depth interval that met
the criterion of TPH < 250 mg/kg.
Sample Name
VW-7 location
EPA-1 1-2A
EPA-1 2-3A
EPA-1 3-4A
EPA-1 4-5A
EPA-1 5-6A
EPA-1 6-7A
EPA-1 8-9A
EPA-1 9-10A
EPA-1 10-11A
EPA-1 10-11B
EPA-1 12-13A
EPA-1 12-13B
EPA-1 13-14A
EPA-1 13-14B
EPA-1 14-15A
EPA-1 14-15B
EPA-1 15-16A
EPA-1 15-16B
MeOH Blank-1
MeOH Blank-2
Trip Blank-1
Trip Blank-2
Sample Depth
(feet below land surface)
1.5
2.5
3.5
4.5
5.5
6.5
8.5
9.5
10.5
10.5
12.5
12.5
13.5
13.5
14.5
14.5
15.5
15.5
TPH-g in Methanol Extract
(M9/L)
3.4E+05
6.9E+04
2.4E+04
<50
<50
8.0E+02
<50
<50
<50
<50
<50
<50
3.1E+06
1.4E+06
1.4E+06
8.8E+05
2.8E+05
8.4E+04
2.1E+03
<50
1.2E+04
3.2E+03
TPH-g in Core Samples
mg/kg
263
53
19
<0.04
<0.04
1
<0.04
<0.04
<0.04
<0.04
<0.04
<0.04
2400
1084
1084
681
217
65
NC
NC
NC
NC
NC = Not calculated, core sample not extracted
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 103
-------
Table 10.15 presents the relative percent difference
between duplicate field samples for the data in Figure
5.3.
No laboratory duplicates were analyzed. The relative
percent difference of the field duplicates were not
within 15%. The values of the relative percent
differences were large compared to the values of the
Continuing Calibration Checks and Laboratory Control
Spikes in Table 10.13. The large relative percent
differences in Table 10.15 reflect variation in TPH-g
in the samples due to the expected heterogeneity
of the contaminated zone on the scale of the plug
samples that were collected and submitted for analysis.
The purpose of the analyses was to distinguish the
deepest depth interval with TPH less than 250 mg/
kg. That interval was 12 to 13 feet. There was no
further interpretation put on the depth intervals with
concentrations greater than 250 mg/kg. The quality of
the data were acceptable for the intended purpose, and
all the data were used.
Table 10.15. Relative percent difference of field duplicates for analysis of TPH-g on samples of soil extracts (applies to
Figures 5.3 and 8.3).
Vapor Monitoring Point
VW-7, 10 to 11 feet
VW-7, 11 to 12 feet
VW-7, 12 to 13 feet
VW-7, 13 to 14 feet
VW-7, 14 to 15 feet
VW-7, 15 to 16 feet
Date Collected
6/15/2007
6/15/2007
6/15/2007
6/15/2007
6/15/2007
6/15/2007
Soil Component
TPH-g
TPH-g
TPH-g
TPH-g
TPH-g
TPH-g
Duplicate
Field
Field
Field
Field
Field
Field
Relative Percent
Difference
NC*
NC*
NC*
39
22
54
*NC means not calculated because both replicates were less than the detection limit of 0.04 mg/kg.
70.2.2 So;7 Gas Samples, Tables 5.1 through
5.6 and 8.7 through 8.3
10.2.2.1 Total Hydrocarbons and Benzene
in Soil Gas from Vapor Probes
Soil gas samples from vapor monitoring points
installed by the Utah Department of Environmental
Quality were acquired by purging three tubing volumes
with a syringe, and then collecting approximately 1
liter of soil gas into a Summa Canister. The vacuum
in the canister drew in the sample. The samples were
analyzed by Air Toxics LTD. (180 Blue Ravine Road,
Suite B, Folsom, CA 95630). Benzene was analyzed
by EPA Method TO-15. Total petroleum hydrocarbons
were analyzed by modified EPA Method TO-15.
Methane was analyzed by modified ASTM D-1946.
The quality assurance parameters for data in Tables
5.1, 5.2, 5.4, 5.5, 8.1, 8.2 and 8.3 are presented in
Table 10.16.
All of the quality assurance parameters were
within the acceptable range except for the recovery
of the surrogate spike of l,2-dichloroethane-d4.
The recovery of l,2-dichloroethane-d4 exceeded
the nominal value. Presumably, the recovery of
compounds of interest would also exceed the true
value. However, the other two surrogate recoveries
were acceptable and are more representative of the
compounds of interest. Therefore, the impact on
sample data for this one high surrogate recovery
is considered minimal. Because the data are used
to screen the site, there is limited or no harm if the
recoveries exceed the true value by a small factor, in
this case less than a factor of three. The data were
accepted and used in the screening process.
104 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.16. Quality assurance parameters from analyses of soil gas from vapor probes (applies to Tables 5.1, 5.2, 5.4, 5.5,
8.1,8.2 and 8.3).
Analyte
1,2-Dichloroethane-d4
4-Bromofluorobenzene
Toluene-d8
Benzene
Benzene
Benzene
Benzene
TPH ref. to Gasoline
TPH ref. to Gasoline
TPH ref. to Gasoline
TPH ref. to Gasoline
Methane
Methane
Methane
QA Parameter
Surrogate Spike
Surrogate Spike
Surrogate Spike
CCV1
LCS2
LCSD3
Lab Blank
CCV
LCS
LCSD
Lab Blank
LCS
LCSD
Lab Blank
Number of Analyses
26
26
26
4
4
4
4
4
4
1
1
1
Range of Values
78% to 262% Recovery
92% to 109% Recovery
76% to 104% Recovery
95% to 97% Recovery
98% to 106% Recovery
96% to 101% Recovery
AIK1.6|jg/m3 *
100% to 100% Recovery
Not Performed
Not Performed
All < 100 |jg/m3 *
99% Recovery
102% Recovery
< 0.00028%(v/v)*
*
-------
Table 10.17 presents the relative percent difference
between duplicate field samples and duplicate
laboratory samples for the data in Tables 5.1, 5.2, 5.4,
5.5, 8.1,8.2 and 8.3.
The relative percent differences of benzene and TPH
in the field duplicates from VW-3 at 18 feet were not
within 15%. The large relative percent differences
in the field duplicates from VW-3 at 18 feet may
reflect changes in the concentrations of petroleum
hydrocarbons from one sampling day to the next. As
presented in Table 8.2, the concentrations of TPH
and benzene in soil gas from VW-3 at 18 feet were
from two to three fold lower than the concentrations
of benzene and TPH in the groundwater monitoring
wells. The highest concentration of benzene in the
field duplicates from VW-3 at 18 feet was 1.8 fold
higher than the lowest concentration. The highest
concentrations of TPH in the field duplicates from
VW-3 at 18 feet was 2.2 fold higher than the lowest
concentration. The variation within the samples from
VW-3 at 18 feet was not larger than the variation
between VW-3 at 18 feet and the samples from the
groundwater monitoring wells. The conclusion
drawn from the comparison was as follows: Again,
the concentrations of benzene, TPH-g ... in soil gas
were reasonably consistent between the monitoring
wells and the vapor probe. The variability between
replicate samples from VW-3 at 18 feet did not affect
the comparison of data from VW-3 to the groundwater
wells.
The relative percent differences of benzene and TPH
in all the other field duplicates or laboratory duplicates
were within 15%. The quality of the data were
acceptable for the intended purpose, and all the data
were used.
Table 10.17. Relative percent difference of field and laboratory duplicates on samples of soil gas from vapor probes (applies
to Tables 5.1, 5.2, 5.4, 5.5, 8.1, 8.2 and 8.3).
Vapor Monitoring Point
VW-3 at 18 feet
VW-3 at 18 feet
VW-7at15feet
VW-7at15feet
VW-7at15feet
VW-3 at 18 feet
VW-3 at 18 feet
VW-7at15feet
VW-7at15feet
VW-7at15feet
VW-1at8feet
VW-3 at 18 feet
VW-7at15feet
Date Collected
09/08/2011
09/07-08/2011*
09/08/2011
09/07/2011
09/07-08/2011
09/08/2011
09/07-08/2011
09/08/2011
09/07/2011
09/07-08/2011
09/07/2011
09/08/2011
09/07-08/2011
Soil Gas Component
Benzene
Benzene
Benzene
Benzene
Benzene
TPH ref. to Gasoline
TPH ref. to Gasoline
TPH ref. to Gasoline
TPH ref. to Gasoline
TPH ref. to Gasoline
Methane
Methane
Methane
Duplicate
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Field
Field
Relative Percent
Difference
0
28.5
2.2
1.1
0
0
36.6
2.2
3.3
14.8
0
3.1
7.7
*Most of the field duplicates were collected on subsequent days.
106 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
10.2.2.2 Individual Hydrocarbons and Oxygen in
Soil Gas from Monitoring Wells
Soil gas samples were acquired using the protocol of
Jewell and Wilson (2011) from conventional ground
water wells that were installed at the direction of the
Utah Department of Environmental Quality. Soil gas
samples were acquired by water displacement into 165
ml serum vials. The vials were returned to the Kerr
Center for analysis by micro gas chromatography with
thermal conductivity detection. The quality assurance
parameters for data in Tables 5.1, 5.2, 5.4, 5.5, 8.1, 8.2
and 8.3 are presented in Table 10.18.
There was no CCC, SS, or GB for oxygen. For the
other compounds, the continuing calibration checks
and the second source standards were within +/- 15%
of the nominal value. The concentration of oxygen in
the field blanks and trip blank were as much as 3.29%
by volume. Using the equation in Section 4.3, this
would require a correction of the hydrocarbon data
by multiplying by a factor of 1.2. The concentration
of total hydrocarbons was fit to Figure 6.4 with a
precision of one significant digit to estimate the
attenuation factor of benzene. The concentrations of
oxygen in the blanks would not affect the estimate of
the attenuation factor. The concentration in all the
other blanks were less than the detection limit. The
quality of the data were acceptable for the intended
purpose, and all the data were used.
Table 10.18. Quality assurance parameters from analyses of soil gas from ground water wells (applies to Tables 5.1, 5.2, 5.4,
5.5, 8.1,8.2 and 8.3).
Analyte
Methane
Methane
Methane
Methane
Methane
Ethylene
Ethylene
Ethylene
Ethylene
Ethylene
Ethane
Ethane
Ethane
Ethane
Ethane
Acetylene
Acetylene
Acetylene
Acetylene
Acetylene
QA Parameter
CCC1
SS2
GB3
FB4
TB5
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
Number of Analyses
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
Range of Values
98.4% to 106% Recovery
101% to 105% Recovery
AIK1.47ppm(v/v)*
AIK1.47ppm(v/v)*
<1.47ppm(v/v)*
103% to 105% Recovery
104% to 105% Recovery
AIK1.77ppm(v/v)*
AIK1.77ppm(v/v)*
<1.77ppm(v/v)*
98.4% to 101% Recovery
100% to 101% Recovery
All < 2.65 ppm (v/v)*
All < 2.65 ppm (v/v)*
< 2.65 ppm (v/v)*
100% to 102% Recovery
102% to 102% Recovery
All < 2.5 ppm (v/v)*
All < 2.5 ppm (v/v)*
< 2.5 ppm (v/v)*
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 107
-------
Table 10.18. Quality assurance parameters from analyses of soil gas from ground water wells (applies to Tables 5.1, 5.2, 5.4,
5.5, 8.1,8.2 and 8.3). (Continued).
Analyte
Propane
Propane
Propane
Propane
Propane
Propylene
Propylene
Propylene
Propylene
Propylene
/so- Butane
/so- Butane
/so- Butane
/so- Butane
/so- Butane
Butane
Butane
Butane
Butane
Butane
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
frans-2-Butene
1-Butene
1-Butene
1-Butene
1-Butene
1-Butene
/'so-Butylene
/'so-Butylene
/'so-Butylene
/'so-Butylene
/'so-Butylene
QA Parameter
ccc
ss
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
Number of Analyses
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
Range of Values
98.4% to 107% Recovery
102% to 107% Recovery
All < 2.05 ppm (v/v)*
All < 2.05 ppm (v/v)*
< 2.05 ppm (v/v)*
97.3% to 105% Recovery
99% to 104% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
92.1% to 99.9% Recovery
99.9% to 105% Recovery
AIK10ppm(v/v)**
AIK10ppm(v/v)**
< 10 ppm (v/v)**
98.1% to 106% Recovery
101% to 106% Recovery
AIK 1.99 ppm (v/v)*
AIK 1.99 ppm (v/v)*
< 1.99 ppm (v/v)*
98.4% to 101% Recovery
100% to 101% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
97.1% to 108% Recovery
100% to 107% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
96.9% to 106% Recovery
100% to 105% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
108 GW Issue
unds ;n So// Gas at Source of Contamination to Evaluate Potential for PVI
-------
Table 10.18. Quality assurance parameters from analyses of soil gas from ground water wells (applies to Tables 5.1, 5.2, 5.4,
5.5, 8.1, 8.2 and 8.3). (Continued).
Analyte
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
c/s-2-Butene
/so-Pentane
/so-Pentane
/so-Pentane
/so-Pentane
/so-Pentane
Pentane
Pentane
Pentane
Pentane
Pentane
2-Methyl-2-butene
2-Methyl-2-butene
2-Methyl-2-butene
2-Methyl-2-butene
2-Methyl-2-butene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
frans-2-Pentene
1-Pentene
1-Pentene
1-Pentene
1-Pentene
1-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
c/s-2-Pentene
QA Parameter
ccc
ss
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
CCC
SS
GB
FB
TB
Number of Analyses
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
5
2
7
4
1
Range of Values
96.1% to 105% Recovery
98.5% to 104% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
96.1% to 105% Recovery
99.1% to 104% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
91.1% to 99.6% Recovery
90.5% to 104% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
9 1.1% to 105% Recovery
109% to 11 4% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
96.8% to 101% Recovery
94.1% to 9 1.9% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
96.8% to 107% Recovery
100% to 106% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
96.8% to 105% Recovery
101% to 105% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 109
-------
Table 10.18. Quality assurance parameters from analyses of soil gas from ground water wells (applies to Tables 5.1, 5.2, 5.4,
5.5, 8.1, 8.2 and 8.3). (Continued).
Analyte
C6+
C6+
C6+
C6+
C6+
Benzene
Benzene
Benzene
Benzene
Benzene
Oxygen
Oxygen
QA Parameter
ccc
ss
GB
FB
TB
CCC
SS
GB
FB
TB
FB
TB
Number of Analyses
5
2
7
4
1
5
2
7
4
1
2
1
Range of Values
9 1.6% to 102% Recovery
91. 3% to 97.8% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
100% to 109% Recovery
102% to 103% Recovery
All < 2 ppm (v/v)**
All < 2 ppm (v/v)**
< 2 ppm (v/v)**
29,900 and 29,200 (v/v)**
32,900 ppm (v/v)
*
-------
Table 10.19. Relative percent difference of field and laboratory duplicates on soil gas samples from ground water monitoring
wells (applies to Tables 5.1, 5.2, 5.4, 5.5, 8.1, 8.2 and 8.3).
Vapor Monitoring Point
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-2
MW-21
MW-21
MW-21
MW-41
MW-41
MW-41
MW-42
MW-42
MW-42
MW-42
MW-42
MW-42
MW-42
Date Collected
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/07/2011
09/07/2011
09/07/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
Soil Gas Component
Oxygen
Methane
Ethane
Propane
/so- Butane
Butane
frans-2-Butene
1-Butene
/so-Butylene
c/s-3-Butene
/so-Pentane
Pentane
2-Methyl-2-Butene
frans-2-Pentene
1-Pentene
c/s-2-Pentene
C6+
Benzene
Total Petroleum
Hydrocarbons*
Oxygen
C6+
Sum Petroleum
Hydrocarbons
Oxygen
C6+
Total Petroleum
Hydrocarbons*
Oxygen
Methane
Butane
frans-2-Butene
Pentane
C6+
Total Petroleum
Hydrocarbons*
Duplicate
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Relative Percent
Difference
9.7
2.4
0.4
0.3
0.9
0.8
0.6
4.6
1.1
0.6
0.7
0.7
33.7
18.1
32
60
0.2
0.4
7.1
0.5
2.4
4.9
0.7
6.5
85.0
55
6.3
8.8
10.8
24
6.3
44.5
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue ] ] ]
-------
Table 10.19. Relative percent difference of field and laboratory duplicates on soil gas samples from ground water monitoring
wells (applies to Tables 5.1, 5.2, 5.4, 5.5, 8.1, 8.2 and 8.3). (Continued).
Vapor Monitoring Point
MW-47
MW-47
MW-47
MW-47
MW-47
MW-47
MW-47
MW-48
MW-48
MW-48
MW-48
MW-48
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
MW-51
Date Collected
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/09/2011
09/07/2011
09/07/2011
09/07/2011
09/07/2011
09/07/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
09/08/2011
Soil Gas Component
Oxygen
Pentane
C6+
Total Petroleum
Hydrocarbons*
Oxygen
Pentane
C6+
Oxygen
Methane
Pentane
C6+
Total Petroleum
Hydrocarbons*
Oxygen
Methane
Ethane
Propane
/so-Butane
Butane
1-Butene
/'so-Butylene
c/s-3-Butene
/so-Pentane
Pentane
2-Methyl-2-Butene
frans-2-Pentene
1-Pentene
C6+
Benzene
Total Petroleum
Hydrocarbons*
Duplicate
Field
Field
Field
Field
Laboratory
Laboratory
Laboratory
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Relative Percent
Difference
0.0
29
0.6
0.53
1.2
4.8
2.9
1.9
3.4
0.8
0.2
0.53
0.0
2.3
69
61
16.8
2.8
45
25
64
0.2
1.3
14.7
0.1
3.4
0.2
0.4
3.9
*Values of each individual analyte in ppm converted to |jg/m3, and then summed to get Total Petroleum Hydrocarbons.
112 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
The relative percent difference for many of the
individual petroleum hydrocarbons was greater than
15%. The concentrations of the individual petroleum
hydrocarbons are summed to calculate TPH in units of
(ig/L. Table 10.19 also compare the relative percent
difference of the calculated TPH for duplicate field
samples. The relative percent difference for TPH
in samples from MW-41 and MW-42 were 85.0 and
44.5% respectively. The relative percent difference
for samples from the other sampling locations were
less than 15%.
Table 8.3 compares the calculated concentration of
TPH in soil gas from well MW-41 and MW-42 to
measured concentrations of TPH in soil gas from vapor
sampling point VW-8. The data were interpreted to
make the following claim. The concentrations of
methane and TPH-g were much higher in the soil gas
from the monitoring wells. To be conservative in the
comparison of TPH between the vapor sampling point
and the monitoring wells, the samples from MW-41
and MW-42 with the lower calculated concentration
of TPH were used in Figure 8.9. The lower of the two
calculated values for TPH in duplicate field samples
from well MW-42 was 77% of the mean of the
samples, and the lower of the two calculated values for
TPH in duplicate field samples from well MW-41
was 58% of the mean of the samples. These lower
values were almost ten thousand fold higher than the
concentration in gas from vapor sampling point VW-8.
The relative percent difference in the duplicate samples
is small compared to the difference in concentrations
between the gas samples from the monitoring wells
and the gas sample from the vapor sampling point.
The quality of the data were acceptable for the
intended purpose, and all the data were used.
10.3 Section 6
Soil gas samples were acquired using the protocol
of Jewell and Wilson (2011) from conventional
ground water wells that were installed at the direction
of the Oklahoma Corporation Commission. Soil gas
samples were acquired by water displacement into
165 ml serum vials. The vials were returned to the
Kerr Center for analysis by micro gas chromatography
with thermal conductivity detection.
The quality assurance parameters for data in Table 6.1
are presented in Table 10.10 (Section 10.1.4). Table
10.20 presents the relative percent difference between
duplicate field samples and duplicate laboratory
samples for the data in Table 6.1.
Table 10.20. Relative percent difference of field and laboratory duplicates on samples of soil gas analyzed in the laboratory
for data in Table 6.1.
Monitoring Well
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-2
Antlers MW-9
Antlers MW-9
Antlers MW-9
Antlers MW-9
Date Collected
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
1/4/2011
Soil Gas Component
Oxygen
Methane
Benzene
Total Petroleum Hydrocarbons*
Oxygen
Methane
Benzene
Total Petroleum Hydrocarbons*
Oxygen
Methane
Benzene
Total Petroleum Hydrocarbons*
Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Lab Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Field Duplicate
Relative Percent
Difference
3.8
2.4
2.9
2.4
9.6
2.3
0.9
0.2
3.5
0.6
1.2
0.6
*Values of each individual analyte in ppm converted to pg/m3, and then summed to get Total Petroleum Hydrocarbons.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 113
-------
The relative percent difference of the field and lab
duplicates were all less than 15%. The data were of
acceptable quality for the intended purpose and all of
the data were used.
10.4 Section 8
Data in Figure 8.3, Table 8.1, Table 8.2 and Table 8.3
are repeat data discussed in Section 4. The quality
assurance parameters for data in Figure 8.3 are
presented in Table 10.13. Table 10.15 presents the
relative percent difference between duplicate field
samples for the data in Figure 8.3.
The quality assurance parameters for data in Tables
8.1, 8.2 and 8.3 are presented in Table 10.16. Table
10.17 presents the relative percent difference between
duplicate field samples and duplicate laboratory
samples for the data in Tables 8.1, 8.2 and 8.3.
ACKNOWLEDGEMENTS
Technical peer reviews of this document were
provided by Theresa (Terry) Evanson and
David Swimm with the Wisconsin Department
of Natural Resources; Robin Davis with the Utah
Department of Environmental Quality; Tom McHugh
and Lila Beckley with GSI Environmental Inc., in
Houston, Texas; Keith Piontek with TRC solutions in
St. Louis, Missouri; Blayne Hartman with Hartman
Environmental Geoscience, Solana Beach, California;
and Hal White with the United States Protection
Agency, Washington, DC.
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
Issue Paper has been subjected to the Agency's peer
and administrative review and has been approved for
publication as an EPA document.
114 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
APPENDIX A. KINETICS OF
STABILIZATON DURING FIELD
SAMPLING
The data used in these case studies 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 or the Montana Department
of Environmental Quality.
Data are presented from an additional eight sites
related to the kinetics of stabilization of oxygen
and carbon dioxide and the kinetics of stabilization
of benzene and petroleum hydrocarbons in soil gas
samples. All locations, including data presented in
Figures 4.4 and 4.5 in Section 4, used existing ground
water monitoring wells for soil gas collection. Wells
were screened across the water table.
Prior to purging each well, the water level was
measured and well construction logs were reviewed
to determine exposed screen length above the water
table that was available for entry of soil gas. A
modified EX-Cap® was connected to the top of the
well casing to provide a pneumatic port to connect
the sample train. Wells were purged at rates up to 10
L/min as measured using a rotameter. The vacuum
that developed in the well was determined with a
vacuum gauge that read in inches of water. Vacuum
measurements in inches of water directly indicate the
rise of water in the well during purging. Knowing the
exposed screen length, the pump rate was adjusted so
that the vacuum in the well did not exceed the length
in inches of the exposed screen, thereby preventing
water from inundating the entire screened interval.
The outlet from the pump and rotameter was split to
two outlet lines. One supplied the sampled soil gas
to the field meters for CO2 and O2 measurements to
monitor for stable concentrations. The other was
for collection of samples for laboratory analysis of
benzene and Total Hydrocarbons in the sampled soil
gas. Following the procedures of Jewell and Wilson
(2011), samples were collected by water displacement
into glass serum bottles. The bottles were sealed with
butyl rubber Teflon®-faced septum and aluminum
crimp caps. The septa were faced with Teflon® and
had a layer of lead foil presented to the sample. The
water used for displacement contained a 1% solution
of sodium phosphate dodecahydrate to act as a
bactericide and preventing biological degradation of
the samples before sample analysis.
Four samples were collected for Total Hydrocarbons
and Benzene analysis at each time interval. The
greatest variability in concentrations between the four
samples occurred at the initial time, one minute. Later
samples had better agreement in concentration.
1E+09
1E+08
- 1E+07
0123456789
1
1E+06
1E+04
o™
I
8"
15
10
o Total
Hydrocarbons
Benzene
0.8
0.6
0.4 X
U
0.2
0.0
C02
02
CH4
0123456789
Purge Vol (-)
Figure A.1. Location MW-9 at the
Visocan Site at Helena, Ml
See Figure A. 1. Equilibration for CO2 and O2
occurred after 8.5 purge volumes. Benzene was
initially detected above quantitation limits and
decreased below quantitation limits and eventually
method detection limits. Concentrations of Total
Hydrocarbons were stable after 2 purge volumes. Both
Total Hydrocarbons and Benzene decreased from the
first measurement. The lithology around the screen/
sand pack was stratified tan to purple silty clay with
less than 10% sand and clasts up to 1A inch diameter.
GW Issue 115
-------
1E+09
1E+OB
„ 1E+07
1E+05
QL
1E+04
| 2,
5?
1 "
* 1.
Loose Connection to meter
Total
Hydrocarbons
Benzene
1E+09
1E+08
„ 1E+07
1
1E+06
1E+05
QL
1E+04
0.8
M
0.4
0.2
0.0
CO2
O2
CH4
01234567
Purge Vol (-)
Figure A.2. Location MW-4 at the
Visocan Site at Helena, Ml
1
8-
20
,
Total
Hydrocarbons
Benzene
0.8
0.6
0.2
---•CC
02
CH
0.0
01234567
Purge Vol (-)
Figure A.3. Location SVE-NE at the
Former Noon's Store, Helena, Ml
See Figure A.2. Equilibration for CC>2 and C>2
occurred after approximately 6 purge volumes.
Notice a loose connection to the field meter prevented
accurate readings before 2.5 purge volumes. Benzene
was below quantitation limits and eventually non-
detect.
Concentrations of Total Hydrocarbons were stable
after 3 purge volumes. Both Total Hydrocarbons and
Benzene decreased from initial measurements. The
lithology around the screen/sand was stratified green to
purple silty clay with less than 5% sand and clasts up
to !/2 inch diameter.
See Figure A.3. Equilibration for CC>2 and C>2
occurred after 2.5 purge volumes. Benzene was non-
detect and Total Hydrocarbons were non-detect after
the initial sample. This is a classic profile of a clean
well. The lithology around the screen/sand pack was
reddish tan sandy clay, brown sand with gravel, and
brown clayey sand.
116 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
1E+09
1E+08
012345
I
1E+07
1E+Q6
1E+05
QL
1E+04
20
I 15
10
i Total
Hydrocarbons
Benzene
1E+09
1E+08
0.0 0.5 1.0 1.5 2.0 2.5 3,0 3.5
O)
1E+07
1E+06
1E+05
QL
1E+04
CO2
02
20
15
10
2 3
Purge Vol (-)
Total
Hydrocarbons
Benzene
- - C02
— 02
Figure A.4 Location SVE-SE at the
Former Noon's Store, Helena, Ml
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Purge Vol (-)
Figure A.5. Location HMM-6 at the
Former Noon's Store, Helena, Ml
See Figure A.4. Equilibration for CO2 and O2
occurred after 2.5 purge volumes. Benzene was non-
detect and concentrations of Total Hydrocarbons had
varying non-detects and detection below quantifying
limits. The lithology around the screen/sand pack was
reddish tan sandy clay, brown sand with gravel, and
brown clayey sand.
See Figure A.5. Equilibration for CO2 and O2
occurred after 2 purge volumes. Concentrations of
Benzene and Total Hydrocarbons decreased from
initial measurements and stabilized after 2 purge
volumes. The lithology around the sand/screen pack
was a light tan mudstone and limestone.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 117
-------
I
1E+09
1E+08
1E+07
1E+06
1E+05
QL
1E+04
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
o"
1
o™
o
20
15
10
Total
Hydrocarbons
Beniene
1E+09
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
„ 1E+07
I
1E+06
1E+05
1E+04
100
90
SO
70
60 ?
50 £•
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Purge Vol (-)
Figure A.6. Location GMW-1 at the
EZ-Go Service Station at Antlers, OK.
C02
-O2
-CH4
T 20
0~
•g
15
10
D Total
Hydrocarbons
Benzene
100
90
80
70
60 ?
50 SS
40 5
30
20
10
0
C02
02
CH4
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0,7
Purge Vol (-)
Figure A.7. Location MW-9 at the
EZ-Go Service Station at Antlers, OK.
See Figure A.6. Equilibration for CO2 and O2
occurred after 4 purge volumes. Concentrations of
both Benzene and Total Hydrocarbons increased
throughout the purge and approached stability near 4
purge volumes. The lithology around the screen/sand
pack was gravel fill and sandy clay.
See Figure A.7. Equilibration for O2 occurred after
0.7 purge volume. CO2 continued to increase
throughout the entire purge. Concentrations of
Benzene increased from initial measurements and
stabilized after 0.4 purge volumes. Concentrations of
Total Hydrocarbons were stable throughout the purge.
The lithology around the screen/sand pack was brown
clay, red brown clay, and red brown sand clay.
118 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
1E+09
1E+08
„ 1E+07
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
1
1E-MJ6
1E+05
QL
1E+04
1
8~
20
15
10
Loose Connection lo meter
- Total
Hydrocarbons
Benzene
100
90
80
70
60 -
50 i
0.0 0.5
1.0 1.5 2.0 2.5
Purge Vol {-)
3.0 3.5
Figure A.8. Location MW-1 at the Miller
Mart Service Station at Wapanucka.OK.
CO2
O2
CH4
See Figure A.8. Equilibration for CC>2 and Q^
occurred after 1 purge volume. Concentrations
of Benzene and Total Hydrocarbons were stable
throughout the purge. Notice the loose connection to
the field meter that prevented accurate measurements
during the initial stages of the purge. The lithology
around the screen/sand pack was red brown and light
gray clay, silt, sand gravel mix, and some mottling.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 119
-------
APPENDIX B. RECOMMENDED
PRACTICE FOR COLLECTING SOIL
GAS SAMPLES
There are extensive recommendations available for
good practice for collecting soil gas samples from
vapor probes. There are few recommendations for
collecting soil gas from conventional ground water
monitoring wells. This section will provide references
to useful guidance documents for vapor probes and
will provide detailed recommendations for sampling of
soil gas from ground water monitoring wells.
B1. Collecting Soil Gas Samples from
Vapor Probes
U.S. EPA has published a standard operating procedure
for sampling soil gas (U.S. EPA, 2001). The American
Petroleum Institute has produced a publication with
recommendations for collecting soil gas samples from
the vadose zone (API, 2005). Both of these documents
have many useful recommendations, but they describe
the state of practice ten years ago.
Recently, the Cooperative Research Centre for
Contamination Assessment and Remediation of
the Environment in Australia published Australian
guidance for petroleum hydrocarbon vapor intrusion
(CRC CARE, 2013). The guidance is detailed and
easy to read. The technical approach is consistent
with good practice in the USA. The appendices of this
document provide detailed protocols for installation
and sampling of soil gas probes. The document also
provides links to much of the technical and regulatory
guidance that is currently available from state agencies
in the USA.
The Interstate Technology & Regulatory Council
(ITRC) has produced a guidance document on
petroleum vapor intrusion. This document provides
detailed protocols for installation and sampling of soil
gas probes (ITRC, 2014).
B2. Collecting Soil Gas Samples from
Groundwater Monitoring Wells
Jewell and Wilson (2011) describe a procedure to
extract and analyze soil gas from conventional ground
water monitoring wells. This section describes that
procedure in detail and provides recommendations for
good practice.
B.2.7 Equipment Requirements
The sampling train should have the following
components. The flow of air through the sampling
train should be in the order that the various
components of the sampling train are discussed in the
text below.
B.2. /. / A Sampling Cap
Jewell and Wilson (2011) used an EX-Cap® to connect
the sampling train to the monitoring well (Figure B.I).
An EX-Cap® is an adaptation of a conventional J-Plug®
well cap that is commonly used to seal two-inch PVC
wells.
Figure B.1. The cap used by Jewell and Wilson
(2011) to sample soil gas from ground water wells.
120 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
The EX-Cap® was not designed to be air-tight. As a
result, the EX-Cap® as supplied by the manufacturer
may leak air. If the EX-Cap® is to be used as the
sampling cap, it must be taken apart and fitted with
O-rings or other seals to make it air tight. Another
issue is the internal resistance of the compression
fitting to the flow of air. At normal pumping rates,
there is a measureable pressure drop across the EX-
Cap®.
It is possible to fabricate a sampling cap made from
materials that can be acquired at any hardware store.
A nipple with a tube fitting made from metal or PVC
is screwed or glued to a PVC plug. The plug is
connected to the riser of the well with a flexible rubber
collar. Most monitoring wells are either 2 inches or 4
inches in diameter. The diameter of the rubber collar
should match the diameter of the riser. To minimize
contact of the air sample with the flexible collar, the
plug should fit down against the top of the riser.
B.2.1.2 A Trap to Collect Water or NAPL
Accidents happen. On occasion, pumping on the well
may inadvertently yield water or hydrocarbons instead
of soil gas. This can destroy the pump, the rotameter
and any instruments that are connected downstream of
the pump. It is also necessary to keep water out of the
Summa Canister used to collect the sample. The pump
and equipment downstream should be protected by a
water trap or fluid trap. One approach is to insert a 1.0
liter glass vacuum flask between the sampling cap on
the well and the pump (Figure B.2).
Three Way
Valve
Figure B.2. An arrangement of a sampling cap connected to the riser of a well, a water trap
and a sampling port connected to a Summa Canister.
Concentrations of HC Compounds in Soil Gas at Source <
GW Issue 121
-------
Be sure to use a vacuum flask. It will have heavy walls
and a side arm. A tube should run from the sampling
cap through a rubber stopper to the bottom of the flask.
A second line should run from just below the bottom
of the stopper to the pump. The second line can also
be connected to the side arm. If the second line is
plumbed through the stopper, the side arm must be
sealed off.
It is possible to produce a vapor from a ground water
monitoring well that will burn (Jewell and Wilson,
2011). If the vapors in the trap are ignited, the flexible
rubber stopper in the cap should separate from the
flask and allow the gasses to escape. Do not clamp the
stopper down. Do not push it into the flask tightly. The
stopper should be just tight enough to make an air-tight
seal.
The soil gas that is produced from the unsaturated
zone contains water vapor. It can be very close to
being saturated with water vapor. If the sampling
train is cold with respect to the unsaturated zone,
water may condense or freeze out in the sampling
train. Condensation in the rotameter is particularly
problematic. This can be remedied by putting a
desiccant in the trap (Figure B.3).
Figure B.3. A trap containing desiccant to remove water vapor from the soil gas sample.
122 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
B.2.1.3 A Valve fo Collect Gas Samples
Tubing connects the sampling cap to a trap that
protects the pump and downstream equipment. A
three-port valve should be installed to allow the
collection of a sample of the gas from the well into a
Summa Canister for laboratory analysis (Figure B.2).
The Summa Canister should be filled rapidly, over at
most a few minutes. When the Summa Canister is
being filled, position the three-port valve so that the
Summa Canister is open to the monitoring point, but
both the Summa Canister and the monitoring point are
isolated from the vacuum pump. Turn off the vacuum
pump while the Summa Canister is filling.
The three-port valve can be installed just downstream
of the trap. This will protect the Summa Canisters. It
is possible a desiccant could sorb or desorb compounds
of interest. If desiccant is put in the trap, the three-port
valve should be inserted between the sampling cap and
the trap. If the three-port valve is located between
the sampling cap and the trap, the Summa Canister is
not protected from water. Inspect the line from the
monitoring point to the valve as the Summa Canister
fills to see if droplets of water wet the side of the
line. If droplets of water enter a Summa Canister, that
canister is compromised. Sample again with a new
canister. To minimize the risk of pulling water into the
canister, adjust the valve to the canister to lower the
flow rate of soil gas to the canister.
8.2. ].4 A Vacuum Gauge
As will be discussed below, it is important to monitor
the vacuum on the soil gas in the well. A line should
run from a vacuum gauge to the sampling cap or to
the line between the sampling cap and the trap. The
vacuum gauge should be referenced to the atmosphere.
It should display the vacuum in inches of water. This
will alleviate a need to make unit conversions in the
field. The gauge should display a vacuum up to 100
inches of water.
B.2.1.5 A Pump fo Move So/7 Gas
Each 1.0 vertical feet of a 2.0 inch well contains 0.6
L of soil gas. Depending on the depth to water, the
dead air space in the screen and riser of a ground water
monitoring well varies from several liters to tens of
liters. This is too much soil gas to efficiently purge
with the small pumps that are built into conventional
field meters. It is necessary to have a second pump to
produce soil gas from the monitoring well for purging
and for sampling.
The pump should be able to function effectively
against a pressure difference of 100 inches of
water. It is not necessary to have a pump that works
against a higher vacuum. The pump should move
approximately ten liters of soil gas a minute against
a pressure difference of 100 inches of water. The
pump should be constructed with components that
will not sorb petroleum vapors, such as fluoropolymer
elastomer. There are several air pumps on the market
that will meet these specifications. Many of these
pumps are diaphragm pumps.
8.2.1.6 A Rotameter to Measure the Pumping Rate
To monitor the progress of purging, it is necessary
to know how rapidly soil gas is moving through
the sampling train. A rotameter should be installed
downstream of the pump. If the rotameter is
downstream, it operates at atmospheric pressure
instead of in a partial vacuum. If it operates at
atmospheric pressure, there is no need to apply
corrections to the readings from the rotameter. The
rotameter should measure up to 10 standard liters of air
per minute.
B.2.1.7 A Valve fo Control the Flow of Soil Gas
There should be a valve in the sampling train that
allows the operator to have fine control on the flow
rate of soil gas through the pump. This should be a
needle valve, not a ball valve. The needle valve can be
built into the pump, it can be built into the rotameter
or it can be a separate unit that is installed between the
pump and the rotameter. It is more convenient if the
needle valve is built into the rotameter. The needle
value should be installed between the pump and the
rotameter.
The needle valve should allow precise control of the
flow rate of soil gas down to a flow of 1.0 liters per
minute.
Concentrations of HC Compounds in Soil Gas at Source of Contaminatio
GW Issue 123
-------
B.2.1.8 Lines to Supply Soil Gas to a Field Meter
Field meters for methane, oxygen, carbon dioxide
or total hydrocarbons have an internal pump. It
is necessary to supply soil gas to the meter at
atmospheric pressure. This can be done by inserting a
tee between the rotameter and the field meter. One line
from the tee goes to the meter, the other line goes to
exhaust. The meter pumps the soil gas it needs and the
excess is discharged through the exhaust line. Figure
B.4 illustrates this arrangement.
It is necessary to know the pumping rate in the meter
and adjust the sample pump to supply that flow of soil
gas with a comfortable margin. If more soil gas than
is needed by the pump is not delivered to the tee, the
pump in the meter will pull air from the atmosphere
through the exhaust line. The soil gas delivered to
the meters will be diluted with air, which will produce
erroneous readings.
To confirm that things are working as intended, insert
the exhaust line into a beaker of water and check for
bubbles. The presence of bubbles ensures that an
adequate supply of soil gas is being provided to the
meters.
Rotameter and
Needle Valve
Figure B.4. Recommended arrangement of the
lines that supply sampled soil gas to field meters.
124 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
B.2.2 Minimizing the Contamination of
Soil Gas Samples by the Contents of the
Monitoring Well
The intention is to sample oxygen, petroleum
hydrocarbons and methane in soil gas at the source
of the vapors. The sample can be compromised by
interaction of the sampled gas with materials in the
well. If there is floating hydrocarbon in the well,
volatile hydrocarbons released from the floating
hydrocarbon can contribute to the hydrocarbons in the
gas sample. If there are dissolved hydrocarbons in the
water, volatile hydrocarbons that escape the water can
also contribute to the hydrocarbons in the gas sample.
If the level of ground water or floating NAPL
hydrocarbons is above the top of the screen,
the sampling train will produce water or NAPL
hydrocarbons, not soil gas. The well will only produce
soil gas when the initial level of ground water or
floating NAPL hydrocarbons is below the top of
the screen. The vacuum imposed in the well during
purging or sampling will cause the water or NAPL
hydrocarbons to rise inside the well. If the vacuum is
high enough, the rise will inundate the top of the
screen. When the level of well water or floating
NAPL hydrocarbons is above the top of the screen,
the soil gas sample must be pulled through the floating
NAPL hydrocarbons or the well water inside the
well. This greatly increases the chances for transfer of
hydrocarbons from floating NAPL or water to the soil
gas sample, and should be avoided.
The effect of vacuum is illustrated in a laboratory
experiment depicted in Figure B.5. A conventional
two inch PVC well screen and riser was cut in half,
and the sections were attached to a piece of Plexiglas.
Then the Plexiglas was mounted as one side of a steel
box. The box was filled with sand and colored water
was added to create a water table that was below the
top of the screen in both model wells. A ruler was
added to measure the change in water elevations when
soil gas was pumped from Well # 1. A red arrow
identifies the level of water in Well #1, and a second
red arrow identifies the location of the black ball in
the rotameter which measures the rate of purging of
soil gas from Well #1. Figure B.5 represents static
water level conditions, when soil gas has not yet been
pumped from the well.
Figure B.5. Level of water in model monitoring wells with no purging of gas and no vacuum.
Concentrations of HC Compounds in Soil Gas at Source t
GW Issue 125
-------
Figure B.6. Effect of purge rate of soil gas on the vacuum in the well and the elevation of water in the well. Panels A
and B illustrate gas flow rates of 9 and 13 L/min, respectively.
126 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
Figure B.6 compares the effect of two different purge
rates of soil gas from Well #1 on the vacuum produced
and the level of water in the well. Well #2 on the right
side of the figure acts as piezometer. It documents the
free water surface under atmospheric conditions.
In Panel A of Figure B.6, the flow rate of soil gas
was 9 liters per minute from Well #1. The free water
surface in Well # 1 is just below the top of the screen.
The vacuum that developed was near 1.0 inch of water.
In Panel B of Figure B.6, the flow rate of soil gas
was increased to 13 liters per minute. The free water
surface was above the top of the screen. The vacuum
that developed increased to 5 inches of water. The
water was pulled through approximately 5 inches of
well water as it moved from the sand pack around the
well to the inside of the riser. The examples in Figure
B.5 and Figure B.6 are from a more extensive
experiment. All the data are presented in Figure B.7.
It should be noted that only one-half of a well was
used in these experiments. The flow rate in a complete
well would be twice as large.
Without any pumping, almost 2 inches of screen
was exposed to the soil gas. As the rate of pumping
increased, the vacuum increased. The rise of water
in the well was equivalent to the vacuum. Only 1
inch of screen could supply enough soil gas to sustain
pumping at 9 liters per minute. However, at 10 liters
per minute the water rose in the well and inundated the
screen.
At faster flow rates, the rise of water in the well was
greater than the vacuum. This is because the water
above the screen contained bubbles of gas, and thus
had a lower density.
£ "
o .c
i e
O
-------
B.2.3 Sampling Procedures to Prevent Transfer
of Contamination to Soil Gas
Useful information regarding the well can be found
in the well construction log. Take a well log to the
field. From the log, determine the depth to the top of
the screened interval. Determine the depth to water
in the well. Subtract the depth to the top of the screen
from the depth to water. That difference is the depth
interval of screen that is exposed to soil gas. Monitor
the vacuum that develops in the well and adjust the
flow rate as necessary to ensure that the vacuum (in
inches) does not exceed the exposed interval of screen
(in inches).
To save labor costs, it is obviously better to purge and
sample soil gas from a well as rapidly as possible. The
practical rate of pumping is controlled by the depth
interval of the screen that is exposed to soil gas, which
controls the maximum vacuum that should be put on a
well and the permeability of the unsaturated material.
Figure B.8 compares the relationship between the rate
of soil gas extraction and the vacuum that developed
on a number of ground water monitoring wells. These
data were developed using the approach of Jewell and
Wilson (2011). Many wells must be pumped at rates
that are less than 10 liters a minute to avoid developing
vacuums in excess of 50 or 100 inches of water.
S
"re
i>
"o
dj
0
^,
|
0
5
100 -I
90 -
80 -
70 -
60 •
50 -
40 -
30 •
20 -
10 -
n .
n
D
n
D
§s:
A
g* * +
• A ™
*• A + ^ *
1 . ...
*
+ + +
* *
^^ ^
• a£ a sji^*a'
A " » • *
•
6 8 10 12
Pumping Rate (Liters/ Minute)
14
»GWM-1 Antlers
nMW-9 Antlers
AMW-2 Picket
x Antlers MW-2
xMW-3 Picked
oMacalasterT-3
+ 10001 SE29thOKC
-5050 N May
Beggs NE-? well
+MaysvilteTPN-1
• KelchNWTPI
iKelchNWTPIE
xpauls Valley VPW-3
x Pauls Valley VPW-1
• Maysvilie NE Well
+ WapanukaMW-1
Miilerton TP South
• MacDonalds Sinclair TMW-1
-Macdonalds Sinclair TMW-3
•MacDonalds Sinclair TMW-3
» Green River MW-48
Green River MW-21
• Green River MW-2
Green River MW-2-7
e Green River MW 41
• Green River MW-42
Green River MW 47
• Hutchinson KSSW-3
nHutchinsonMP4Deep
Hutchinson Shallow South MP4
• Hutchinson MP 3 Deep
» Hutchinson MP 3 Shallow
Hutchinson MP 2 Deep
• Hutchinson MP 2 Shallow
Hutchinson MP 1 Deep
Hutchinson MP 1 Shallow
Helena SVE-SE
• Helena SVENE
• Helena HMM-6
« Helena MW-9
« Helena MW-7
• Helena MW-4
•••• Helena KevsMW-4
Figure B.8. Relationships between the pumping rate of soil gas and the vacuum that developed on gas in selected
ground water monitoring wells.
128 GW Issue
Gas at Source of Contamination to Evaluate Potential for PVI
-------
B.2.4 Perform a Leak Test
on the Sampling Train
Conduct a leak test on the equipment setup used to
acquire the soil gas samples. Glue a cap on the end of
a section of one inch riser, two inch riser or four inch
riser (as is appropriate for the vapor monitoring point)
to create an air-tight seal. This makes a test section.
Turn off any field meters. Connect the sampling cap to
the test section and run the pump to develop a vacuum
of 50 inches of water, 100 inches of water or whatever
is the maximum vacuum you intend to impose. Then
turn off the pump and position the three port valve
between the rotameter and the exhaust line so as to
isolate the sampling system from the exhaust and from
the field pumps. Watch for one minute to see if the
vacuum readings drop on the gauge. If the vacuum
readings drop, find the leak and correct it.
B3. Summary of ORD Recommendations
for Sampling Soil Gas from Ground Water
Monitoring Wells
At the beginning of each day of sampling, take
equipment blanks of air that has passed through the
sampling train for analysis of benzene and TPH-g.
Take trip and field blanks as required by the quality
assurance project plan.
Zero the vacuum gauge at the beginning of the day.
Check and adjust the zero every time the sampling
train is moved.
Conduct a leak test every time you set up on a new
vapor sampling point.
The field meters should be calibrated as specified by
the manufacturer. Check the calibration by analyzing
the concentration of a standard gas. At a minimum,
the calibration should be checked at the beginning
of a day of field sampling, during the middle of the
day and at the end of the day. The best practice is
to check the calibration when the sampling train is
moved to a new location. The nominal and reported
values of the calibration gas should be recorded in the
field notebook. If the performance of an instrument
is out of the acceptable range, it should be adjusted
or recalibrated to bring its performance into the
acceptable range.
If there is floating NAPL in a well, remove as much
NAPL as possible before sampling gas from the well.
Know the depth interval of well screen exposed to soil
gas and do not apply enough vacuum during sampling
to inundate the screen.
As soil gas is pumped from a monitoring well, the
concentrations of petroleum vapors in the gas can
go up or they can go down (see Appendix A). It is
impossible to know beforehand which pattern a well
will follow. Collect one set of samples for laboratory
analysis approximately one minute after pumping
starts. Monitor soil-gas concentrations during purging
until stable concentrations of oxygen and carbon
dioxide are attained. Then collect a second set of
samples for laboratory analysis. The purge volume
required to reach stable concentrations may be greater
or less than the typically recommended purge volume
(~3 void volumes of air in the well).
Store the samples, trip blanks, field blanks and
equipment blanks away from direct sunlight and at
room temperature.
Determine if the sampling train is contaminated
before moving the sampling train to a new location.
After sampling is concluded at a location, disconnect
the sampling train from the well and pump clean air
through the sampling train for ten minutes. Monitor
the air that passes out of the sampling train with
an OVA meter. If the air is still contaminated after
ten minutes, replace tubing and decontaminate the
equipment in the sampling train. Do not sample a
new location until the OVA meter indicates that the
sampling train is no longer contaminated.
Infrared sensors in an OVA meter are sensitive to radio
frequency interference. Any device that transmits
radio waves can cause the infrared meter readings to
fluctuate. Cell phones are the most common cause
of the problem. Never use cell phones while taking
readings with an infrared sensor. Cell phones should
be turned off or keep at least 20 feet away from an
instrument in active use.
If there is any concern of transfer of contamination
from one location to the next, take equipment blanks
for analysis of benzene and TPH-g before sampling the
new location.
GW Issue 129
-------
Do not use Tygon® tubing. Use nylon or silicone
tubing. Tubing can become contaminated by being in
the vicinity of electrical generators or containers for
gasoline. Store tubing in a protective container or bag.
Store the tubing away from electrical generators or
containers for gasoline.
Where possible, set up the sampling train at a location
that is up-wind from potential sources of airborne
contamination. These sources include electrical
generators, dispenser islands at gasoline service
stations and the exhaust of automobiles.
Isolate electrical generators and fuel cans from all
sampling equipment during storage, transport and field
sampling. Wear disposable gloves when refueling
generators and automobiles. Immediately discard the
gloves after re-fueling equipment.
Change gloves when moving from one sampling
location to the next. This is particularly important if
floating NAPL is present in a well.
When the water table rises, the ground water may
cover up the NAPL. When the water table goes
down, it may expose NAPL to soil gas. Sample the
soil gas more than one time. If possible, one sampling
event should be when the water table is at or near a
seasonal low.
130 GW Issue
unds in Soil Gas at Source of Contamination to Evaluate Potential for PVI
-------
United States
Environmental Protection
Agency
National Risk Management
Research Laboratory
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
EPA/600/R-14/318
December 2014
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
------- |