DRAFT
A MANUAL FOR CONDUCTING FIELD SCREENING
EOR SUBSURFACE GASOLINE CONTAMINATION
US. ENVIRONMENTAL PROTECTION AGENCY
LIBRARY
P.C. BOX 93478
IAS VEGAS, NV 89193-3478
by
Gary Robbins, Robert Bristol, Mark Temple and Brendan Deyo
Department of Geology and Geophysics
University of Connecticut
Storrs, CT 06269
James Stuart and Valerie Roe
Department of Chemistry
University of Connecticut
Storrs, CT 06269
Cooperative Agreement
CR-814542-01-2
Project Officer
Philip Durgin
Advanced Monitoring Systems Division
Environmental Monitoring Systems Laboratory
Office of Research and Development
Las Vegas, Nevada 89114
Environmental Monitoring Systems Laboratory
Office of Research and Development
Las Vegas, Nevada 89114
-------
NOTICE
This document is a preliminary draft. It has not been
formally released by the U.S. Environmental Protection Agency
and should not at this stage be construed to represent Agency
Policy. It is being circulated for comments on its technical
merit and policy implications. Mention of trade names or
commerical products does not constitute endorsement or
recommendation for use.
ii
-------
PREFACE
This manual describes field methods that have been
developed and/or tested as part of a research project sponsored
by the U.S. Environmental Protection Agency at the University
of Connecticut to evaluate factors influencing the detection
and monitoring of subsurface gasoline leakage at service
station sites. The manual was developed as part of a training
fi
program for state regulatory personnel on the use of field
screening methods. The manual is not meant to be a survey of
alternative and available methods that might be used in
screening sites. Rather, the methods described are limited to
those used during the course of our research that have been
found to be useful in detecting and delineating subsurface
gasoline contamination, and in assessing processes and
conditions related to the migration of gasoline constituents in
soil gas and in ground water.
iii
-------
ABSTRACT
This manual, describes field methods for screening soil,
ground water and soil gas for detecting and delineating
subsurface gasoline leakage at service station sites. The
methods described in the manual are: a polyethylene bag
headspace method for screening soil and ground water using
total organic vapor detectors; a VOA vial headspace method for
screening soil and ground water for use with portable gas
chromatographs; a soil gas survey method using total organic
vapor detectors that entails gas bag sampling and performance
of serial dilutions; a method for determining relative soil air
permeability while conducting a soil gas survey; and the
measurement of soil gas oxygen and carbon dioxide using
portable meters as an indirect method for detecting subsurface
gasoline leakage. The manual also covers the use of field
measurements for inorganic ground water quality parameters
affected by the biodegradation of hydrocarbons as an indirect
means of detecting and delineating ground water contamination.
Inorganic water quality parameters include dissolved oxygen,
dissolved carbon dioxide, electrode-' potential (Eh), pH,
electrical conductance, alkalinity, chloride, sulfate, iron and
managanese.
iv
-------
CONTENTS
Preface Hi
Abstract iv
Figures vii
Tables . viii
Acknowledgement ix
1. Introduction 1
2. A Summary of Gasoline Characteristics and Factors
Affecting its Detection in the Subsurface 5
Gasoline liquid composition . . 5
Factors affecting the composition of
gasoline in the subsurface 6
Sampling and sample preservation 18
Instrumentation . . . . . . 27
3. Polyethlyene Bag Headspace Screening Method .... 31
Method overview ................ 31
Theory 32
Sampling equipment ..... 39
Sampling procedures ..... 41
Calibration options, testing field equipment,
and assessing field performance of personnel . 43
Advantages and disadvantages 44
4. VOA Vial Headspace Screening Method using
Portable Gas Chromatography . . 46
Method overview 46
Theory 47
Sampling equipment 51
Sampling procedures 52
Calculations 55
Advantages and disadvantages . 56
5. Soil Gas Surveying with Total Organic Vapor
Detection Instruments 57
Method overview 57
Theory 59
Sampling equipment . 65*
Sampling procedures 66
Advantages and disadvantages 69
v
-------
6. Soil Gas Surveying for Oxygen and Carbon Dioxide
as Indirect Indicators of Subsurface Gasoline
Leakage 72
Method overview ................ 72
Theory .72
Sampling equipment and procedures ....... 74
Advantages and disadvantages .... 75
7. Field Measurements of Inorganic Ground Water
Quality Attributes of Gasoline Contamination .... 77
Overview 77
Method summary 77
References 80
vi
-------
FIGURES
Number Paga
2.1 Soil gas sampling system 24
3.1 Polyethylene bag headspace sampling system . . . .40
5.1 Possible forms of serial dilution curves . .... 64
5.2 Example of a flow rate-vacuum pressure curve
for- determining relative soil air permeability . i-^Tl
vii
-------
TABLES
Number Page
1.1 Desirable attributes of field screening
methods ..................... 2
1.2 Potential uses of field screening methods in
contamination assessments ............ 3
2.1 Gasoline constituents . . . . . . . . . . . ... 7
2.2 Gasoline vapor constituents . . .... . . . . .9
2.3 Henry's law constants for gasoline constituents . 14
2.4 Inorganic ground water quality parameters
that may be indicative of gasoline
contamination . . . . . • • • • • • • • • • • • •
2.5 Operating characteristics arid conditions that
may influence the response of total organic
vapor detectors .... i ............ 29
3.1 Sampling system equipment ...... 39
4.1 VOA method equipment ...... 52
7.1 Alternative field methods for measuring
inorganic water quality attributes ....... 78
viii
-------
ACKNOWLEDGEMENT
The authors wish to extend their appreciation to the staff
of the Connecticut Department of Environmental Protection and
members of the U.S. Environmental Protection Agency underground
storage tank program for providing comments on early drafte of
the manual. We are especially indebted to Peter Zack, John
Goldman and Frank Bartolomeo of the LOST Trust Program of the
Connecticut Department of Environmental Protection for their
efforts in conducting additional field tests of the headspace
methods presented in the manual. We would like to thank Dr.
Thomas;;Spittler, Director, 0. S. Environmental Protection Agency
Region 1 Laboratory, Lexington, Mass. for being a constant
source of inspiration for our students and for sharing his
id^as on headspace analysis, amongst other things.
ix
-------
SECTION 1
INTRODUCTION
Field screening refers to measurements that can be
performed in the field to obtain real-time information on
subsurface contamination. The techniques included in this
manual have been developed and/or tested to screen for
subsurface gasoline leakage. The techniques include: a
pqlyethylene bag headspace technique to screen soil or ground
water using commonly employed total organic vapor detectors; a
VOA vial headspace technique to screen soil and ground water
using portable gas chromatographs; a serial dilution technique
to screen soil gas for gasoline vapors using total organic
vapor detectors; techniques for screening soil gas oxygen and
carbon dioxide levels as indirect measures of gasoline
contamination using portable gas instruments; and a: description
of inorganic ground water quality attributes resulting from the
biodegradation of hydrocarbons that may be useful as indirect
indicators of ground water contamination by gasoline.
Desirable attributes of field screening methods are listed
in Table 1.1. These desirable attributes provide a yardstick
against which alternative field screening techniques may be
evaluated. In this regard, the methods presented in the manual-
have been found to possess many of the attributes listed in
1
-------
Table 1.1, based on theorectlcal evaluations, laboratory
TABLE 1.1 DESIRABLB ATTRIBUTES OF FIELD SCREENING METHODS
* Reliable
- well founded theoretical bases
- verifiable procedures and results
* Capable of calibrating measurements to relative or absolute
amounts of contamination
* Capable of obtaining correlations with laboratory analyses
* Easy to learn and use
* 'Rapid results
* Versatile in application
* Inexpensive
testing and field use at sites exhibiting subsurface gasoline
contamination. However, in choosing a field screening method
for application to a specific site, it is important to consider
the method's underlying theory in relation to specific site
characteristics. Not all methods may be applicable to an
individual site, and some methods may be preferred over others.
In addition to applicability in relation to site
characteristics, the choice of a particular method depends on
the level of detail desired or required to evaluate a problem.
Potential uses of field screening methods in contamination
assessments are listed in Table 1.2. The different methods
presented provide different levels of detail in performing the
li6ted assessments. To assist the reader in the choice of a
2
-------
method, subsections are provided on the advantages and
disadvantages of the methods. Further, the reader should be
familiar with properties of gasoline, and subsurface processes
and conditions, that pertain to its detection in the
subsurface. These are covered in the next section.
TABLE 1.2 POTENTIAL OSES OF FIELD SCREENING METHODS IN
CONTAMINATION ASSESSMENTS
SITE INVESTIGATION
* Screen soils during drilling to assess
4.
^-presence of contamination
-depth of contamination
-source of contamination
-subsurface migration pathways
-maximum drilling depth
-need for additional drilling locations
-samples for laboratory analysis
-amounts of contaminated soil
-disposal options of cuttings
* Screen soil6 during excavation/remediation to assess
-lateral and vertical extent of excavation
-storage and disposal options
-in situ remediation effectiveness
-emergency response actions related to spills
* Screen groundwater to
-determine if ground water is contaminated during
exploratory drilling
-monitor well purging prior to sample collection
-screen monitoring wells to map extent of
contamination
-assess whether nearby drinking water.wells are impacted
-increase the frequency of sampling monitoring wells
around facilities
-monitor production well water quality
-monitor the progress of recovery well remediation
-monitor effluent from treatment facilities
3
-------
TABLE 2. continued
* Screen surface water to
-determine contaminated ground water discharge locations
4
-------
SECTION 2
A SUMMARY OF GASOLINE CHARACTERISTICS AND FACTORS
AFFECTING ITS DETECTION IN THE SUBSURFACE
This section provides a brief overview of factors relating
to the detection of subsurface gasoline leakage using the field
screening methods. Pertinent factors include: the composition
of the gasoline; the extent to which gasoline may volatilize
and the compositon of the evolving vapors; the rate of
"si;
fi .
migration of the vapors from trapped free product; the
solubilization and volatilization of gasoline constituents in
ground water; the extent to which - gasoline constituents are
biodegraded; sampling procedures; sample preservation; and
instrumentation.
GASOLINE LIQUID COMPOSITION
Gasoline is a mixture of hydrocarbon constituents. The
composition of gasoline liquid may depend on the source of the
initial petroleum, the refining process, and the addition of
additives. Although C4 through C13 hydrocarbon compounds are
the dominant constituents in gasoline, there is no exacting
composition to gasoline. Hence, the composition of gasoline may
vary from service station to station, as well as throughout the
year at a single station, depending on the above factors. If
5
-------
subsurface gasoline leakage occurs over an extended period of
time, its initial composition may somewhat vary. This
compositional variation may influence the relative abundance of
constituents detected in the subsurface, and may even influence
the extent to which subsurface leakage may be detected.
As a representive listing of constituents in gasoline,
consider Table 2.1. The Table summarizes data from Hoag, Bruell
and Marley (1984). In that study, they identified more than 50
hydrocarbon components in three types of gasoline. Only the
dominant constituents (weight percent > 4%) are included on the
Talble. Also included are the aromatic constituents which tend
il
to be the focus of environmental assessments. The table
readily reveals that gasoline is dominated by relatively
volatile and insoluble constituents. The exception to this
being methyl tertiary butyl ether (MTBE), an octane enhancing
additive (the reader is referred to Garrett et al. (1986) and
Garret (1987) for further information on MTBE). Although
volatile, the aromatic constituents tend to have higher
solubility than the alkane constituents.
FACTORS AFFECTING THE COMPOSITION OF GASOLINE IN THE
SUBSURFACE .
Upon leakage of gasoline into the subsurface,
volatilization, solubilization into vaaose and ground water,
and biodegradation may affect its composition. These processes"
are treated below.
6
-------
TABLE 2.1 GASOLINE CONSTITUENTS
A. Major Gasoline Constituents*
weight %
constituent
leaded
unleaded
superunleaded
n-hexane
11.0
11.1
12.9
n-& isopentane
8.3
8.9
6.1
2-pentane
8.9
6.1
2.4
2-methyl pentane
6.3
5.4
4.3
methyl tert. butyl
ether
4.0
1.9
1.4
isoctane
2.0
1.8
8.7
unknown
1.4
1.0
5.8
aromatics
15.6
16.1
19.3
total
58
52
61
B. Aromatic Composition
weight
% (relative
aromatic %)
constituent
leaded
unleaded
superunleaded
benzene
3.9 (26)
3.2 (20)
4.4 (23)
toluene
4.5 (30)
4.8 (30)
6.0 (31)
ethylbenzene
1.2 ( 8)
1.4 ( 9)
1.5 ( 8)
total xylenes
5.6 (37)
6.7 (42)
7.4 (38)
total
15.6
16.1
19.3
data source: Hoag, Bruell and Marley (1984)
* only constituents having a weight % > 4%
in any one of the gasolines is listed.
Volatilizatlon
Upon leakage of gasoline into soil, the gasoline may
migrate under a positive pressure head due to the fluid level
in a leaking tank or fluid pressure in a leaking pipeline,
7
-------
gravity and capillarity. Where the gasoline is in contact with
soil air (i.e., air in the pore spaces between soil grains)
along the fringe of the free product, its constituents car
volatilize. The different constituents in gasoline will have
different tendencies to volatilize depending on the overall
composition of the gasoline, the fraction of a particular
constituent in the gasoline, and a constituent's vapor
pressure. Further, the extent to which volatilization takes
place will depend on subsurface temperatures and the degree to
which evolved vapors may migrate by diffusion or advective flow
away from the gasoline-soil air contact. The result of
volatilization will be a time varying ohange in the composition
of the gasoline liquid as well as a time varying change in soil
vapor composition. Table 2.2 provides some insight as to how
the composition of gasoline may change as a result of
volatilization and what the vapor composition may be like. The
Table reveals that' fresh gasoline vapor will be dominated by
alkane constituents. As such, one may expect a depletion in
these constituents in subsurface gasoline. In turn, one might
expect an enrichment in the relative composition of aromatics
in subsurface gasoline. Since volatilization is a time and
spatial (dependent. upon properties of the soil) varying
process, one should keep in mind that th'e relative abundance of
gasoline constituents may vary depending on the age of gasoline
in a soil sample and the location of the samples. This is
particularly important in conducting field screening using
8
-------
TABLE 2.2 GASOLINE VAPOR CONSTITUENTS
Constituent
Weight %
Cumulative Weight %
isopentane
35.8
35.8
n-butane
19.1
54.9
isobutane
15.2
70.1
n-pentane
13.1
83.2
2-methylpentane
6.3
89.5
n-hexane
3.2
92.7
3-methylpentane
3.1
95.8
propane
1.8
97 . 6
aromatics
2.4
100.0
data source: Wadden, Ono and Wakamatsu (1986).
Values are averages from measurements of the
vapor composition in the headspace of tanks
containing regular and unleaded gasoline, Tokyo.
total organic vapor instruments whose response depends on both
the absolute and relative abundances of detectable
constituents.
Table 2.2 indicates that despite the high concentration of
aromatics in fresh gasoline, their vapor concentrations
relative to other constituents may be very low (a consequence
of the aromatic compounds having relatively low vapor
pressures). This is important to keep in mind in relation to
soil gas "surveying and in screening soil samples using
headspace techniques. The data in the Table would imply one may
more readily detect subsurface gasoline leakage by focusing on
alkanes or using instruments with high.-alkane sensitivity. The
dominance of alkane compounds in gasoline vapor can also lead
to disparities between field screening measurements on soil
samples using total organic vapor detection instruments (which
9
-------
essentially detect the alkane compounds) and laboratory
analyses which focus on aromatic compounds.
Vapor Migration
As noted above the extent to which vapors migrate in the
soil away from the gasoline will have a feedback on the extent
to which the gasoline will volatilize. If vapor migration is
restricted, a local vapor/liquid equilibria may be achieved
which can result in inhibiting volatile losses and changes in
the initial gasoline composition. Vapor migration would likely
be restricted in fine grain and moist soil. If vapor migration
in the soil is restricted, it results in very high, vapor
gradients in the soil gas composition. This in turn can
confound the detection and delineation of subsurface gasoline
leakage using soil gas techniques. That is, appreciable vapors
might only be detected in the near proximity of subsurface
gasoline.
There are a number of factors that may influence gasoline
vapor migration. Firstly, gasoline vapors may migrate by
advective processes, i.e., due to air pressure gradients.
Advection may be induced by barometric pressure changes or
density differences in soil air. One might expect these
conditions to be active very near to the surface or very close
to the source of vapors. Advective migration will depend on the
air permeability of soil. Soils having relatively high air
permeability (i.e., coarse, fractured or dry soil) have a
10
-------
higher potential for vapor migration due to advection.
Secondly, temperature gradients in the subsurface can
induced vapor migration. For example, a variation in subsurface
temperature from 0 to 30 °C, will bring about a 5 fold increase
in the vapor pressure of benzene. Likewise, the tendency of
other volatlles in gasoline will increase with temperature.
Hence, increases in soil temperature can cause increases in
the rate of volatilization of gasoline in the subsurface. In-
turn, this can promote vapor migration. One may expect vapors
to migrate from areas of higher temperature to lower
Vi
temperature in soil. Temperature induced vapor migration would
likely be significant in situations where air and ground water
temperature vary significantly on an annual basis, and where
subsurface gasoline leakage is relatively shallow. In
conducting soil gas surveying in such situations, a survey may
be more effective during warmer weather.
Thirdly, and, perhaps most importantly, vapor migration
in the subsurface will occur due to diffusion, i.e., in
response to vapor concentration gradients. Soils that are
course and dry will tend to conduct vapors more readily by
diffusion. This implys that gasoline vapors will be spread over
larger areas which can be more readily detected by soil gas
surveying. In evaluating soil gas survey data, one should keep
4
in mind that the distribution of vapore will not only depend on.
source conditions but also on soil properties (e.g., lithology
and moisture content). Importantly, the rate of diffusion of
11
-------
different gasoline vapor constituents may vary, owing tc
different tendencies of constituents to adsorb onto soil
particles, to adsorb onto organic matter in the soil and tc
absorb into soil pore water or ground water. As a generality,
one might expect the alkane constituents of gasoline vapors tc
migrate by diffusion more readily than the aromatic
constituents. Along with their higher tendency to volatilize
from gasoline, their higher tendency for diffusion would make
the alkane constituents more sensitive indicators for detecting
subsurface gasoline leakage by soil gas surveying.
Solubilization
Another factor which may influence the subsurface
composition of gasoline and its vapor is the degree to which
constituents may dissolve in soil pore water and ground water.
In direct contact with infiltrating water or ground water,
different constituents in the gasoline will have different
tendencies to partition into the water. The tendency for a
constituent to dissolve in the water will depend on its
concentration in the gasoline and its solubility. To some
degree many gasoline constituents will dissolve in subsurface
water, although the aromatic compounds and MTBE have higher
solubilities. This becomes apparent ifi conducting headspace
analysis for aromatic compounds on ground water samples using
the VOA vial method with a portable gas chromatograph. In the
near-field of a contaminant source, chromatograms exhibit many
12
-------
peaks and not just those of the more soluble constituents. This
has a positive attribute in recognizing that a well may be near
a source of free product. It has a negative attribute in that
the peaks on a chromatogram for aromatic constituents may
exhibit co-elution, making it difficult to identify the peak or
resulting in an overestimation of the concentration of an
aromatic compound. The fact that many volatile constituents are
present in ground water also increases the detectablity of
gasoline contamination using the polyethylene bag sampling
method with total organic vapor detectors.
Once gasoline constituents are dissolved in ground water,
they may subsequently volatilize from the ground water table.
The tendency of this to happen is described by a constituents
Henry's law constant. Table 2.3 lists Henry's law constants
for pertinent constituents of gasoline. The higher the
constant the higher the tendency for a constituent to partition
out of the water. The values of Henry's law constants listed
are for 25<>C. In addition to being constituent specific,
Henry's law constants will increase with temperature and the
salinity of the ground water (so-called "salting out effect").
The temperature dependence of partitioning would imply that
soil gas concentrations over a plume may vary temporally, if
the ground water temperature varys throughout the year. Also,
it would suggest increased sensitivity to soil gas detection,
during summer months (i.e., when the ground water temperature
13
-------
Table 2.3 HENRY'S LAW CONSTANTS FOR GASOLINE CONSTITUENTS
Cconstituent
Henry's law constant*
(dimenionless)
isopentane
n-butane
isobutane
n-pentane
2-methylpentane
n-hexane
3-methylpentane
propane
methane
MTBE2
aromatics
<0.01
0.22-0.33
57
39
49
51
70
70
71
29
27
daita sources: Mackay and Shiu (1981)
1 Values for Henry's law constants are expressed here as
the equilibrium vapor conc./water conc.
2. MTBE Henry's law constant value from our studies.
is higher). Further, the higher Henry's law constants for the
alkane compounds relative to that of the aromatics implies that
a ground water plume may be more readily detected using soil
gas methods by sensing for alkanes rather than aromatics. As in
detecting gasoline in soil, alkanes would be the prinicpal
components of vapors emanating from ground water contaminated
by gasoline. This would imply that in conducting headspace
screening on ground water samples and in conducting soil gas
surveying using total organic vapor instruments, alkanes are
the prinicipal constituents being sensed':
Because of the higher tendency of alkanes to volatilize
relative to aromatics from the ground water table, one may
observe in conducting headspace analysis on ground water
14
-------
samples a decrease in the ratio of alkanes to aromatics as a
function of distance from a source of free product. This may be
useful in attempting to delineate locations of subsurface free
product.
Biodegrariation
Following subsurface leakage, the hydrocarbons in
gasoline may be subject to aerobic and anerobic degradation by
naturally occurring micro-organisms, principally bacterial The
extent to which this occurs depends on a host of factors, a
<
detailed discussion of which is beyond the scope of this
manual (the reader is referred to the following references for
futher information: McNabb and Dunlap (1975); Schwille (1976);
Atlas (1981); Barker and Patrick (1985); Wilson and Rees (1985)
Wilson et al. (1986)). Of primary interest here are effects of
biodegradation that may be useful in detecting and delineating
subsurface gasoline leakage.
Soil Gas Effects
Aerobic degradation of gasoline in the subsurface can
result in marked decreases in soil air oxygen and increases in
carbon dioxide levels. Marrin (1985) and Marrin (1987) have
reported 6uch effects at sites exhibiting subsurface gasoline
contamination. In field testing using portable carbon dioxide
and oxygen meters, we have found that soil gas carbon dioxide
and oxygen levels can be successfully used a6 indirect measures
15
-------
of subsurface leakage. We have observed good correspondence
between soil gas surveys using total organic vapor detection
instruments and measurements of soil air oxygen (postive
correlations) and carbon dioxide (negative correlations).
The depletion of soil air oxygen levels in the vicinity of
subsurface gasoline can result in extinguishing the flame of
portable flame-ionization detectors, commonly employed in
conducting surveys, when conducting direct-probe sampling or
analyzing sampling bags. To circumvent this problem as well as
several others associated with using total organic vapor
*
detectors, we have developed a serial dilution technique.
Anerobic degradation of hydrocarbons in the subsurface can
result in the generation of methane, along with carbon
dioxide. In the cited Marrin studies, as well as in our own
study, we have found methane to be useful in detecting
subsurface gasoline contamination.
The biodegradation of subsurface gasoline can result in
changes in the absolute and relative abundances of gasoline
constituents in soil and soil gas. We have found that
aromatic.levels in soil samples over a few day period following
sample collection can undergo dramatic decreases owing to
biodegradation. Biodegradation has also been cited to explain
difficulties in detecting aromatic constituents in soil gas
when probing just a few feet above contaminated ground water
(Thompson and Marrin, 1987). We have al6o observed that
aromatic detection in soil gas i6 restricted to the near-field
16
-------
of the source.
Ground Water Effects
As with soil gas, biodegradation of hydrocarbons can
result in increases in dissolved carbon dioxide and methane,
and decreases in dissolve oxygen in ground water. We have found
these effects to be useful indirect field indicators of
#
gasoline contamination, especially in helping to locate
sources of free product.
• In association with changes in dissolved gas levels, we
haive observed changes in inorganic ground water quality
parameters associated with gasoline contamination. Most of
these parameters may be measured with field instruments, and
hence, are useful as indirect indicators of contamination.
Table 2.4 lists parameters and observations on how they change
in relation to contamination level, and potential causes for
the observed changes. In addition to these observations, we
have observed that the ground water in the most heavily
contaminated areas has a greenish tint and has a high
suspended solids content, likely due to high levels of iron.
Exposure of this water to air results in a rapid color change
to bright orange, reflecting the oxidation of iron. In this
regard, we have observed that bright orange coloration of
stream water and on debris in streams is a good indicator of
locations where contaminated ground water discharges to the
surface.
17
-------
TABLE 2.4 INORGANIC GROUND WATER QUALITY PARAMETERS
THAT MAY BE INDICATIVE OF GASOLINE
CONTAMINATION
Parameter
Dissolved
oxygen
Electrode
potential
(Eh)
Iron and
manganese
Dissolved
carbon
dioxide
Alkalinity
Specific
conductance
Chloride
Sulfate
PH
Changs1
decrease
more negative
increase
increase
increase
increase
increase
decrease
increase
Qaiias.
microbial consumption
oxygen consumption
creating reducing
conditions
solubilization of
metals from soil or tanks
in response to
reducing conditions
microbial respiration,
hydrocarbon oxidation
or reduction
dissolved carbon
dioxide increases
bicarbonate alkalinity
increases in metals and
and bicarbonate
solubilization of salts
and silicates
reduction to sulfide
hydrogen ion depletion
resulting from reduction
reactions and formation
of hydrogen sulfide
1. refers to
increase.
how the parameter changes as BTEX levels
BTEX Effects
Biodegradation not only reduces the absolute abundances of
aromatic constituents in ground water, but also brings about
18
-------
changes in their relative abundances. That is, the different
aromatic constituents biodegrade at different rates. In the
near-field of contaminant sources we have found the relative
abundance of aromatic constituents to be similiar to that
listed in Table 2.1. That is, toluene and the xylenes are the
dominant aromatic constituents in the ground water. Outside of
the near-*field zone, we have observed the concentration of
toluene to greatly diminish. Here, ethylbenzene and the xylenes
become the dominant constituents. In the far-field, benzene
and/or ortho-xylene become the dominant constituents. Further,
when MTBE is present, its spatial distribution is more
extensive than that of the aromatics, owing to its high
solubility, relatively low volatility and lack of
biodegradation. Often, in the far-field, MTBE is observed
whereas the aromatic constituents are absent. These
observations may be quite useful when conducting the VOA vial
headspace screening method in locating the source of
contamination, distingushing contributions to ground water
contamination when several service stations may be involved,
evaluating the relative age of a problem, and in assessing the
effectiveness of remediating free product.
SAMPLING AND SAMPLE PRESERVATION
In conducting field screening and in obtaining samples for
laboratory analysis, one should follow protocols that maintain
the integrity of samples. The following are recommended common
19
-------
sense practices that should be followed and kept In mind in
conducting field work. They are meant to supplement and not to
replace more rigorous protocols that may have to be followed as
set forth by regulatory agencies.
Soil Sampling
Cross contamination amongst samples should be minimized by
using clean sampling tools. If reused, tools should also be
cleaned between samples. One should minimize exposing soil
samples to air during handling to prevent volatile loses. In
taking samples, one should avoid sampling soil that has been
exposed to air, i.e., dig into the sides of excavations and
soil piles, use the center of core material. Our experience
would indicate that the aromatic constituents in soil samples
collected in typical sampling jars, and stored under
refrigeration may undergo significant levels of biodegradation
within Just several days. Hence, it is recommended that
laboratory analysis be conducted with 48 hours of sample
collection. This is particularly important in attempting
correlations between field screening results and laboratory
analyses.
Ground Water Sampling
Ground water samples are generally taken from monitoring"
wells having screen sections that extend from some level above
to some partially penetrating depth below the water table. Our
20
-------
studies would indicate that the concentration of constituents
obtained from typical wells will exhibit a mass continuity bias
(i.e., an averaging bias) that depends on the screen depth, the
properties of the sand pack, vertical concentration gradients
of constituents within the formation outside the well, the
amount of water purged from the well prior to sampling and how
the well is purged (i.e., water levels achieved in the well
during purging). To obtain a consistent, although biased, data
set, it is recommended to use consistent procedures amongst
wells in terms of how they are purged and how samples are
collected. It is also advised to conduct multilevel sampling to
obtain information on the vertical distribution of
contamination. The reader is referred to Robbins (1989),
Robbins et al. (1989a), and Robbins et al. (1989b) for details
on this subject.
As with soi1 sampling, ground water samples should be
collected in cleaned containers. In conducting field screening,
containers may be reused throughout the course of field work.
It is suggested containers be cleaned between samples and
rinsed several times with the water to be analyzed, prior to
collecting a sample to be screened. Also, to avoid cross
contamination it is a good idea to sample wells that have lower
levels of contamination first, if sampling containers are to be
reuse. Between sample collection and field analysis, one should
attempt to minimize volatile losses from samples by minimizing
6ample expose to air and sample agitation. The latter is also
-------
important in analyzing for dissolved gases. Further, water
sampling containers should be insulated to minimize temperature
changes. This can be important in screening for inorganic
attributes of contamination, such as specific conductance. In
screening water from a tap at a service station or from nearby
homes, aerators should be removed to help prevent volatile
loss. If a service station has a ground water well, water from
the tap should be screened.
Our studies would indicate that VOA vial samples taken for
laboratory analyses of aromatics may undergo biodegradation if
this samples are not preserved (see Roe et al., 1989). This is
particularly important in comparing field screening results
with laboratory analyses. Hence, the use of preservatives in
VOA vials is recommended (see Roe et al. (1989) and DSEPA
(1982) for preservation techniques).
Soil Gas Sampling
With respect to the soil gas sampling techniques- discussed
in this manual, sampling considerations and recommendations are
listed below:
(1) Probe Type
Probes should be chemically inert with respect to gasoline
vapors. They should be cleaned if they are to be reused. They
should be designed to facilitate cleaning, placement, depending
on soil conditions, and extraction from soil. They should be
placed in a manner to assure a good soil/probe seal and to
22
-------
assure they are not clogged. Probes may be tested for clogging
by monitoring vacuum pressure during sampling. Sealing at the
surface may be facilitated by compacting moist bentonite
around the probe. Sealing may be tested by squirting a
detectable (but not a constituent of gasoline) vapor around
the probe. Probes should have small bore to minimize the volume
of air requiring purging.
(2) Installation
How probes are installed depends on the depth of probing
and the type of soil at a site. Probes may be driven in soft
soil, driven or vibrated into sandy soil, or drilled and driven
into gravelly soil. One should keep in mind, the items
described above in relation to specific site conditions. If
probes are to be placed at 6igificant depth, one should have
confidence that the method chosen results in good sealing at
depth. This is difficult to test. Further, if air permeability
is to be tested, probes should be placed in a manner to
minimize soil disturbance near the probe inlet.
(3) Sampling
Figure 2.1 shows a suggested design of a soil ga3 sampling
system. As a whole, the system should be constructed of
materials that are chemically inert to "'"gasoline vapors, in a
manner that facilitates field cleaning and purging, and with
minimum volume to minimize purging. The particulate filter is
23
-------
PUMP
PRESSURE GAUGE
0-15
in. HP
VALVE
FLOW RATE
GAUGE
GASSAMPLE
COLLECTION OR
DIRECT MEASUREMENT
STAINLESS
• STEEL
TUBING
PARTICULATE
FILTER
I
1
777777777777"
77
Figure 2.1 Soil Gas Sampling System
24
-------
suggested to minimize system clogging. The gauges permit
monitoring sampling flow rate and vacuum pressure, useful in
monitoring purging, assessing whether the system is well
sealed or clogged, in conducting soil air permeability testing
during soil gas sampling and in performing serial dilution
tests. The value assists in system leak testing, and conducting
air permeability testing. The pump, in addition to being made
of inert components, should be leak tight. It is recommended
that soil gas samples to be analyzed with total organic vapor
detectors be collected in chemically inert sampling bags- (see
Robbins et al. (1990a) and Robbins et al. (1990b). That is, one
should avoid direct-probe sampling with these types of
instruments. Samples should be analyzed immediately after
sample collection. To be representative of a given depth, the
amount of air purged from a probe and collected should be
minimized. It is suggested that 5 probe and system volumes be
purged prior to sample collection. Sample collection should be
conducted at low flow rates and vacuum gauge pressures to aid
in minimizing system leakage, to minimize entraining fine
particles- into the system, and to facilitate purging.
(4) Location of Probes
The exact location of probes will always be site specific.
To roughly screen a site, one should evaluate a site for places
that would permit rapid subsurface screening. This could
include: manhole covers for utilities, underground storage
vaults for utilities, manhole covers for sewers and runoff
25
-------
drains, facility drains, soil around the manways of gasoline
tanks, soil beneath gasoline pumps, grassy areas, the headspace
of wells, or septic tanks. Beyond rapid screening, one should
consider probing around on-site utilities, the backfills of
which may conduct vapors. Obviously, caution is advised here.
One 6hould avoid probing into wet areas. Further in this
regard, it is recommended not to conduct soil gas surveying
during and for a few days following precipitation. If soils are
subject to frost heave (fine grain soils), they should onlyibe
probed during times when they are thawed. Tank backfills should
be11 probed early on in an investigation. Beyond the immediate
tank area(s), one should consider subsurface gasoline and
gasoline vapor migration pathways in deciding on probe
locations. Here, one may want to probe the backfills of
underground utilities that may trap free product (e.g., sewer
lines), and conduct surveying in a fashion that is oriented
with respect to ground water flow. Also, it is recommended to
concentrate probes in areas having high permeability (either
for water or air). It is common practice to establish probe
locations on a grid. It is suggested orienting the grid with
respect to the ground water flow direction. The spacing of a
grid will be somewhat site specific. Our experience would
indicate in fine sandy soil with appreciable silt content
(glacial till) probing should be conducted on no more than 10
foot centers in the immediate vicinity of the site.
26
-------
(5) Probe Dept.h
Depth.profiling of soil vapors should be conducted early
in an investigation to determine an appropriate sampling depth,
especially if one is interested in tracking ground water
contamination. Depth profiling is also useful in assessing the
vertical extent of subsurface leakage and to evaluate vapor and
gasoline migration pathways. It is recommended that soil gas
probing be conducted at a minimum depth of 1 foot below
-------
instrument, field personnel should practice using an
instrument before conducting field.screening at sites.
This manual assumes a certain level of instrument
familiarity on the part of the user. However, it is worthwhile
to point out several aspects of good field practice. Field
instruments should be calibrated at least once a day during
operation. During operation, calibration should be periodically
checked with calibration standards. If battery powered, battery
checks should be conducted periodically during operation, and
instruments should not be operated under low battery
conditions. In performing calibrations, the concentration of
calibration standards should be within an order of magnitude of
that expected for samples. The development of calibration
plots using several standards not only aids in improving
accuracy but also provides a means to evaluate the performance
of field personnel in conducting measurements. During the
course of field screening, occasionally, splits from the same
sample should be screened in triplicate to provide an idea of
measurement error.
The field screening methods included in the manual entail,
the use of different kinds of instruments. Each of these
instruments have ranges of operation in terms of concentration,
temperature, and other conditions. It is suggested that the
user check the manual and contact the manufacturer regarding
factors that might influence an instrument's response. The soil
sas and polyethylene bag headspace methods presented in the
28
-------
manual entail the use of total organic vapor detectors. Table
2.5 summarizes pertinent operating characteristics and
conditions that may influence the response of commonly
employeed total organic vapor detectors, as determined from our
research in using these instruments in screening for gasoline
contamination.
TABLE 2.5 OPERATING CHARACTERISTICS AND CONDITIONS THAT
MAY INFLUENCE THE RESPONSE OF TOTAL ORGANIC
VAPOR DETECTORS
KxpIorA meter
4.
* Has low sensitivity for soil gas surveying (% combustible
;ga3), limiting its use to areas very close to free product
or where high levels of methane have been generated from
biodegradation of hydrocarbons.
* Has linear response throughout its operating range.
* In the low scale mode, response requires oxygen. If soil gas
oxygen level is too low, vapor readings can not be
discerned.
* In the low scale mode, response is dampened if the flow rate
of the instrument is curtailed during sampling.
F1ame-lonlnation Detector
* Has high sensitivity to gasoline vapors (Vppm), especially
sensitive to alkane3 but also sensitive to other
constituents.
* Has high sensitivity to methane (positive attribute with
respect to detecting methane from the biodegradation of
hydrocarbons; negative attribute with respect to false
positive detections if other sources of methane are
present).
* Has linear response throughout its operating range.
* Response is dampened if instrument flow rate is curtailed
during sampling.
29
-------
TABLE 2.5 continued
* Flame will extinguish if gas sample has less than 10 to i3%
oxygen.
* Response depends on both absolute and relative abundances of
detectable constituents.
Photo Ion 1 nation Detector
* Has high sensitivity (Vppm) to alkenes and aromatics, but
negligible sensitivity to the major alkane constituents in
gasoline vapors.
* Response is dampened by relative humidity., high levels of
carbon dioxide, and presence of alkanes and methane.
* May exhibit nonlinear response to gasoline vapors when vapor
concentrations exceed about 125 Vppm.
*^Response depends on both absolute and relative abundances of
detectable constituents.
* Response is dampened if instrument flow rate is curtailed
during sampling.
* Response does not appear to be influenced by oxygen level in
gas sample.
sources: Robbins and Temple (1988); Robbins et al. (1989);
Rbbbins et .al. (1990a); Robbins et al. (1990b)
30
-------
SECTION 3
POLYETHYLENE BAG HEADSPACE SCREENING METHOD
METHOD OVERVIEW
The polethylene bag headspace screening method was
originally presented in Robbins et. al. (1989c). The method
provides "a means to perform semiquantitative and relative
sdreening of soil and water for gasoline contamination using
l-i
total organic vapor detection instruments (e.g.. flame-
ionization and photoionization detectors). In general the
method entails the following: a three way valve is connected to
a polyethylene reclosable bag (e.g., Ziploc 1 quart freezer
bag); a water sample is placed in the bag and the bag is
zipped closed; the bag is blown up with air until just taught
using a hand pump; . the bag is agitated to release vapor from
the water sample to the headpsace of the bag; and following
headspace-water transfer equilibrium, the vapor concentration
in the headspace is measured using a total organic vapor
detection instrument. Soils may also be screened by placing a
soil sample in a bag containing distilled water. In this
manner, volatile constituents are partitioned from the soil
sample into the water then into the headspace. The
polyethylene bag provides an environment to control the vapor
31
-------
partitioning, and to sample the evolved vapors in the
headspace in a manner that does not affect the operating flow
rate of a detection instrument (i.e., as air is drawn out of
the bag into the sampling instrument, the bag collapses and
thus maintains a one atmosphere internal pressure). Also,
adding soil to distilled water and agitating results in the
disruption of the soil matrix whose properties (lithology,
porosity, and moisture content) could otherwise cause
variations in vapor partitioning into the headspace.
THEORY
i i
Bag Sampling of Volatile Constituents
Total organic vapor detection instruments generally take
two forms—flame ionization detectors and photo-ionization
detectors (see Table 2.5 for instrument characteristics). Prior
to vapor sample .. measurement these instruments are calibrated
using some calibration gas. Thus, vapor concentration
measurements subsequently made are relative to the calibration
gas concentration and calibration conditions. Importantly, in
gas sampling, care must be taken not to curtail the instruments
flow rate. Curtailment of the flow rate as in the case of
inducing a vacuum, can effect the instruments response. Also,
in the case of the photo-ionization detector, moisture or
changes in relative humidity can effect the instruments
response.
Consider the measurement of the vapor concentration of a
32
-------
single volatile constituent within an inert and collapsible bag
that is sealed and connected to ail organic vapor detection
instrument. As the instrument draws air from the bag during
sampling, the bag collapses and maintains a constant internal
pressure. Therefore, the concentration measured may be
described by
Cmi = RiCi, (3—1)
where Cmi equals the measured vapor concentration for. a
constituent i, Ri equals the instrument's response factor for
constituent i relative to a calibration gas and calibration
conditions, and Ci equals the actual vapor cqncentration.
Equation 3-1 predicts that for a single constituent the
measured concentration will be linearly related to the actual
vapor concentration.
Expanding equation 3-1 for the case where multiple vapor
constituents are present results in
n n n
CraT = Y Cmi = Y (RiCi) = 7 (RiCi/CT)CT, (3-2)
i=l i=l i=l
where CmT is the measured total concentration, Ct is the actual
total concentration, and n equals the number of detectable
constituents. Equation 3-2 predicts tfi'at the measured total
concentration of a mixture of organic vapors from sample to
sample will be linearly proportional to the actual total
concentration if only one constituent's concentration varies,
33
-------
or if the concentration of all constituents vary in proportion
(i.e., Ci/Cx remains constant for each constituent from sample
to sample).
Vapor Partitioning between Water and Headspace
Consider sealing a water sample contaminated with a single
volatile constituent into the bag. If the sample is well
agitated and the volume of headspace remains constant during
agitation, the measured headspace concentration variation with
time may be described by a first order transfer function
(Tchobanoglous and Schroeder, 1985), expressed as
Crni(t) = RiCi(t) = RiCi e[l-exp(-kit)], (3-3)
where Cmi(t) and Ci(t) are the measured and actual vapor
concentrations at time t, respectively, Cie is the actual
equilibrium concentration, and ki is an effective mass transfer
coefficient. Equation 3-3 predicts an exponential achievement
of an equilibrium concentration with time that depends on the
magnitude of the mass transfer coefficient. The mass transfer
coefficient is a function of the individual constituent,
temperature, degree of sample agitation, and the contact area
between the water and the headspace. Mass transfer coefficients
for commonly found volatile contaminants tend to be relatively
large (Mackay and Leinonen 1975), and as has been found in
laboratory studies, headspace concentration equilibrium tends
to be achieved within minutes in well agitated sampling vessels
34
-------
(Griffith et al. 1988).
The equilibrium headspace concentration in equation 3-3
may be related to the equilibrium water concentration by
Henry's law, expressed as
Cie = HiCiwe, (3-4)
where Hi is a dimensionless Henry's law constant and Ciwa is
the equilibrium concentration in the water. Dimensionless
Henry's law constants for gasoline constituents are-listed in
Table 2-3. Equation 3-4 indicates that the higher the Henry's
la^f coefficient, the higher will be the vapor concentration in
equilibrium with the water concentration. It should be noted
that the Henry's law coefficient is sensitive to temperature,
and will generally increase with temperature. Thus, variations
in temperature . will result in variations in vapor-water
concentration equilibrium.
c
The measured Equilibrium headspace concentration may be
related to the dissolved concentration of the constituent when
the water sample was first placed in the bag (i.e., prior to
volatilization) by the following derivation. Prior to
volatilization, all the mass of a constituent is in the water.
Subsequently, the original mass is distributed between the
water and the headspace. This may be expressed by
mio=mive+miwe) (3-5)
where mio equals the initial mass of i in the water, mive
35
-------
equals the mass of i in the headspace at equilibrium, and miwe
equals the mass of i in the water at equilibrium. Dividing all
niass terms by the volume of water (Vw) and multiplying the
vapor mass term by Vhs/Vhs (where Vhs equals the volume of
headspace) results in
Ciwo - Cia(Vhs/Vw) + Ciwa, (3-6)
where Civo equals the initial water concentration, Cie equals
the equilibrium vapor concentration, and Ciwa equals the
equilibrium water concentration. Substituing equation 3-4 into
3-6, rearranging terms and substituing equation 3-1 for the
actual vapor concentration yields,
Cm i e = RiCie = RiHiCiwe = [Ri/(1/Hi + Vhs/Vw)JCiwo, (3-7)
Equation 3-7 indicates that for a single volatile constituent
the measured headspace concentration will be linearly
proportional to the concentration of the constituent in the
water sample prior to volatilization. Also, consistency from
sample to sample requires control of the volume of headspace to
water ratio." With respect to temperature influences, this will
depend on how 1/Hi varies in relation to the Vhs/Vw ratio. If
the volume ratio is large relative to 1/Hi, than temperature
influences on the measured vapor concent-ration will tend to be
dampened out. On the other hand, a larger Vhs/Vw will result
in lower vapor concentrations, and thus decrease detection
sensitivity.
36
-------
The above equations describing water to headspace
partitioning can be readily expanded by summation to treat
multiple constituents, as in the case of expanding equation 3-1
to equation 3-2. Linearity between total equilibrium headspace
concentration and total concentration in the water sample will
depend on the same factors as previously mentioned, and in
addition, the achievement of equilibrium headspace
concentrations by all constituents at the time of measurement.
Vapor Partitioning From Soil To Headapacp.
%
Griffith et al. (1988) demonstrated the effectiveness in
partitioning aromatic compounds from a soil matrix, through
disaggregation and agitation in water. Orice partitioned from
the soil matrix, headspace - equilibrium was shown to be
described by the above theory for partitioning between water
and the headspace. For a single constituent on soil, an
equation may be derived in the same fashion as in the
derivation of equation 3-6 that describes the relation between
soil concentration and vapor concentration. This may be
expressed as
Cm ie — RiCie
= Ri{[Ms/Vw]/[((Ksw + 1)/Hi),;+ Vh3/Vv]}C3, (3-8)
where Ms is the mass of soil added to the bag, Ksw is a-
soil/water partition coefficient, Cs is the concentration of a
constituent in the soil on a mass/mass basis, and other terms
37
-------
are as previously defined. As with the other equations,
providing that the terms to the left of Ca are kept constant
among analyses, the measured headspace concentration ie
predicted to be a linear function of the soil concentration-:'
Griffith et al. (1988) showed that soil/water partition
coefficients for aromatics for quartz rich soil are essentially
zero. Low coefficients would also be expected for other
constituents in gasoline for saturated soils devoid of organic
matter. Nonetheless, equation 3-8 would indicate that even with
incomplete desorption of a constituent from the soil, linearity
of the headspace to soil sample concentration still holds
providing Ksw is a constant from sample" to sample. For nonpolar
organic constituents, Ksw is known to be sensitive to soil
organic content (Chiou et al. 1983). Thus, variations in
organic content from sample to sample would introduce a
nonlinear variability in the headspace to soil concentration
relation. Also, if the salinity of the water in the bag varies
substantially between samples, this could result in non-
linearity in that Hi is a function of salinity.
As "in the previous developments, equation 3-8 can be
readily expanded for the multiple constituent case. Again,
provided that variables as previously mentioned are constant or
controlled, total measured concentration would be predicted to
r
be a linear function of total soil concentration.
38
-------
SAMPLING EQUIPMENT
Components of the polyethylene bag sampling system are
listed in Table 3.1. Figure 1, from Robbins et al. (1989c) is
TABLE 3.1 SAMPLING SYSTEM EQUIPMENT
Water Screening
* Valve system
1 three way ball valve with 1/8"NPT fittings
3 1/8"NPT to 1/4" connector fittings
2 metal tubes 1.5"in length, 1/4"0D and 1/8"ID
2 1/8"NPT nuts
n-buna gaskets(1/16" thickness, 3/8"ID, 11/16'OD)
2" length of Tygon tubing
* Paper hole punch
* Quart size reclosable free2er bags
* 200 ml volumetric flask
* Standard constituents) and syringes for preparing
standards for calibration
* Distilled water
* Stopwatch
* Bicycle air pump (tire valve stem cut off and tubing
connected to 2" length of Tygon tubing)
* Ring stand
* Optional—magnetic stirrer, stirrer bar and tool to
retrieve stirrer bar (for agitation)
* 1000 ml volumetric flask for preparing aqueous standards
Soli Screening
* Above materials for water
* Field balance
* Weighing cups or vials
a diagram of the assembled reclosable polyethylene bag sampling
system. The system consists of a one
-------
3-Way Ball Valve
\
X
JO
rK
\ ¦ s
Connectors
>
&
•Nuts
Gaskets
Clamp
Tubing
Quart Size
Reclosable
Polyethylene
Bag
Figure 3.1 Polyethylene Bag Headspace Sampling System
(from Robbins et al. 1989CJ
40
-------
of a detector instrument to the valve system using Tygon
tubing.
SAMPLING PROCEDURES
Test procedures entailed the following steps.
1. Using the hole punch, make a hole in one side of the bag,
aproximately 2" down from the reclosable end of the bag.
2. Insert the 1/8" NPT connector fitting of the valve system
through the hole, and seal the fitting onto the bag using N-
Buna gaskets and nuts on both sides of the hole.
3. Close the reclosable end of .the bag. Attach the air pump
ontoi the valve system, inflate the bag until just taught, and
by closing off the valve isolate the inflated bag. Observe the
bag for leakage. If bag leakage is observed, check the
reclosable seal. If the bag still leaks, discard the bag and
start again. Robbins et al. (1989c) found the inflated bags to
have an average volume of 1391 cc with a 1.4 % standard
deviation.
4. If proven leak tight, zip the bag open for sample
introduction. Aqueous samples are introduced into the bag using
a volumetric flask. It is recommended that you use 100 ml of
water. You may use more water up to 300 cc to increase
sensitivity. Amounts of water beyond 300 cc, have been found to
be difficult to agitate and sample. For soil samples, first
place about a 100 cc of distilled water in the bag. Then weigh
41
-------
out about 25 g of soil. The exact amount is not as important as
along as you are consistent. Your detection sensitivity may be
increased or decreased by using different amounts of soil
sample (recall.equation 3-8). Also, try to measure out the soil
as quickly as possible to avoid volatile loses. Add the soil
sample to the distilled water in the bag. After sample
introduction, seal the bag.
r
5. Immediately after sample introduction inflate the bag until
just taut, and then isolate the bag by closing the valve.
Following isolation the bag should be agitated by hand using a
rocking motion or using a magnetic stirrer. For gasoline
constituents, we have found that 4 minutes of agitation is
adequate to achieve a stable vapor concentration.
6. Following agitation, the detection instrument is connected
to the valve system, the valve is then openned to the bag and
the vapor concentration is then measured. To facilitate
measurement and prevent the accidental introduction of water
into the detector, securing the valve and bag onto a ring stand
with clamps, prior to connecting your detector, is highly
recommended.
7. Following each analysis the bag and.its contents should be
properly disposed. However, we have found that the buna-n
gaskets may be reused many times without sample carryover. The
valve system should then be purged by pumping air through the
system. To check the effectiveness of purging, the valve system
42
-------
should be connected to the detector instrument. If carryover is
detected after rigorous purging, replace the Tygon tubing and
recheck the system. We have found little carryover when
samplingAgasoline and ^axomatic contaminated .water, and soil.
8. To conduct further testing, repeat 1 through 7.
CALIBRATION OPTIONS, TESTING FIELD EQUIPMENT AND
ASSESSING FIELD PERFORMANCE OF PERSONNEL
Robbins et al. (1989c) described a variety of ways of
performing calibrations between headspace readings and
contaminant concentration. For water, calibrations may be
performed by using a single aromatic constituent as a spike for
water (vapor readings would be interpreted as single
constituent equivalent), serially diluting a ground water
sample from a site and having the sample analyzed in the
•laboratory for BTEX, and conducting headspace 'analysis on
several well samples and having the samples also analyzed in
the laboratory for BT-EX. One can conduct similiar calibrations
for soil.
As an alternative approach to calibration, clean ground
water or 6oil from a site may be spiked with fresh gasoline
over a range of concentrations and a calibration curve (plot of
gasoline concentration vs. headspace Treading) developed. This
curve may in turn be used to approximate the amount of gasoline
contamination for actual samples. If the amount of BTEX in the
gasoline is analyzed, the calibration curve can then be used to
43
-------
estimate BTEX levels in the ground water and soil. The
development of a gasoline vs. headspace curve also permits
evaluating the field performance of personnel, and the linear
range of operation of a total organic vapor detector which is
especially important if a photoionization detector is used in
screening. Also, it permits assessing the sensitivity of the
method in detecting contamination which is soil and detector
instrument specific. Based on the development of many
calibration curves using fresh gasoline in performing research
and in conducting training classes, personnel performing the
I .
meithod should be able to obtain a good linear correlation
between gasoline and headspace concentrations (providing vapor
concentrations are in the linear range of the detector and your
instrument is functioning properly) . It is suggested that a
calibration curve be developed each day at the start of field
work using three or four spiked samples.
ADVANTAGES AND DISADVANTAGES
Advantages
1. Relatively high sensitivity--tens of ppb range for water,
and fractions of a mg/kg (ppm) range for soil.
2. Method, given large volume of headspace to water ratio
(e.g., 1 quart bag/100 ml water = Vhs/Vw = 13), is
relatively insensitive to temperature variations.
3. Well controlled physical system.
4. Permits field calibration and performance testing of field
personnel.
5. Relatively simple procedures.
44
-------
•- \K:
id.^^QBTOas^iffcagg^^gacid^^.fo he
®--'¦Mips*}&.T$^i*QV-Xo^&Ti^L^£a& -?xrtpan:sw^:^.owgfct -.abou* b;.
.•*Jkb arela^W^.^D&s titsteoit tx)ndeatr2Et^nsi><.
f ^i^p^eailo ^ooditeql; al voil)
;±n?:iMrter."lapd iaonrrlix^aT^.lJistTgaeiytr-xifeiaiponBee
c Concentrations
•*'? ¦5r. . ' -i
>£lity w£tlkiii Kfcr.to H5X.
-------
SECTION 4
VOA VIAL HEADSPACE SCREENING METHOD USING
PORTABLE GAS CHROMATOGRAPHY
METHOD OVERVIEW
Headspace analysis by gas chromatography (GC) has been
widely used as a laboratory technique for quantitatively
determining volatile contaminant levels in soil and water. With
the advent of portable gas chromatographs, the quantification
of specific volatile contaminants in soil and water may be.
performed in the field. Clay and Spittler (1982) introduced a
headspace analysis technique using a VOA vial which is used in
collecting ground water samples for laboratory analyses. The
VOA vial headspace method described in this section is an
extension of their method and is directed at determining
aromatic constituents (benzene, toluene, ethylbenzene, and
xylenes, collectively referred to as BTEX) in soil and ground
water samples contaminated by gasoline. Further details on the
methods cited here may be found in Giffith et al. (1988) and
Roe et al. (1989).
With respect to water analysesthe method involves:
partially filling a VOA vial with a contaminated water sample,
temperature equilibration of the sample, agitation of the
46
-------
sample to release vapor from the water to the headspace of the
vial, sampling the vial headspace with a syringe,: and
measurement of the vapor concentrations of •BTEX relative to
that of a vial containing aqueous BTEX standards using a GC.
The measured vapor concentrations can then be related to the
initial aqueous concentrations of the sample. Soils may also be
screened by placing a soil sample in a VOA vial that is
partially filled with distilled water. Through agitation,
volatile constituents are partitioned from the soil sample into
the water then into the headspace. Soil concentration may then
be ^determined by relating the vapor concentration of the sample
.to that of an aqueous standard.
THEORY
Water Analysis
The underlying theory of the method is identical to that
presented in Section 3; therefore, only certain aspects of the
theory will be treated here. Consider filling a VOA vial
(typically having a 40 cc volume) partially with water
containing dissolved BTEX (sample or standard). Following
temperature equilibration, and agitation to achieve equilibrium
partitioning between volatile constituents in the water and in
the headspace, the vapor concentration of a constituent in the
vial may be related to the original water concentration (prior
to volatilzation) by equation 3-7, restated here as,
47
-------
Cia = [l/(1/Hi + Vhs/Vw)]Ciwo,
(4-1)
where Cie 1b the equilibrium vapor concentration of constituent
i, Hi is the dimensionless Henry's law constant, Vha is the
volume of headspace in the vial, Vw is the volume of water in
the vial, and Ciwo is the original dissolved concentration of
constituent i.
Equation 4-1 reveals that headspace concentrations will
ideally be linearly related to the aqueous concentrations;of
the samples but only if the terms in parentheses are kept
constant. Recall Hi is a function of temperature and to a much
lesser degree Overall water quality. Hence, to get consistent
results one must maintain samples and standards at a constant
temperature. Further, the volume _ ratio either 'must be
maintained constant amongst samples and standards or must be
known. The equation also reveals that the sensitivity of the
method may be adjusted by changes in the volume ratio (lowering
the ratio increases the vapor concentration) or changes in
temperature (increasing temperature increases Hi : and. the
headspace concentration).
Taking the ratio of an equation 4-1 for an unknown and one
for a standard, gives
Cio Ciwo/(l/Hi + Vhs/Vw)
Cies Ciwoa/(l/Hi + Vha/Vw)s
where the subscript "a" refers to the parameters for the
standard. If the terms in parentheses are constant, then the
48
-------
vapor concentration ratio is directly related to the aqueous
concentration ratio. Hence, the concentration of the unknown is
given by
Ciwo = [Cie/Cias]Ciwo8. (4-3)
With respect to GC analysis of a sample, the vapor
concentration ratio in brackets in the equation would simply be
replaced by a peak area or peak height ratio.
Equation 4-2 may also be used to determine Hi, following
the method of Gossett (1987). Here, an aqueous standard or a
%
ground water sample is put into two VOA bottles having
different volume of headspace to water ratios. Equation 4-6
may then be rearranged find expressed as
Hi = l-[Ci a 1/Ci e 2]
(4-4)
[Ci8l/Cie2][Vh8/Vw]i - [Vha/ Vw ] 2
where subscript numbers refer to the two different YOA vials.
Soil Analvfilfl.
Griffith et al. (1988) demonstrated the effectiveness in
partitioning aromatic compounds from a soil matrix, through
disaggregation and agitation in water. Once partitioned from
the soil matrix, headspace equilibrium was shown to be
described by the above theory for partitioning between water
and the headspace.
Consider a soil sample contaminated with a volatile
-------
constituent at a concentration C> on a mass of constituent/mass
of soil basis. We then add an amount of soil of mass Ms to a
VOA vial partially filled with distilled water. The vial is
then sealed, temperature equilibrated, and agitated. As before
we then determine the equilibrium vapor concentration in the
headspace. As in the water analysis, the theory for soil in
Section 3 is applicable here. Equation 3-8 may be restated as
Cie = {[Ma/Vv]/[((Kav + 1)/Hi) + Vha/Vv]}Ca, .(4-5)
where Kav is a soil/water partition coefficient (ratio of
*
4
constituent mass left on the soil to mass left in the water
when the mass between phases achieves equilibrium during
agitation). It will be assumed at this point that Kav is
negligible once soils are emersed in the water in the vial for
BTEX. This is a reasonable assumption for BTEX in soils
composed chiefly of quartz and other silicate minerals but may
result in underestimations of Ca for organic rich soil. This
assumption may be readily tested for a specific soil by
introducing uncontaminated soil into an aqueous BTEX standard
and performing headspace analysis.
Dividing equation 4-5 (where Raw = 0 ) by equation 4-1,
for an aqueous standard, and rearranging, yields
Ca — [Cie/Ciea] * Ciwoa * (1/Hi + Vha/Vw)
f (4_6)
(1/Hi + V hs/Vw)a * [Ma/Vw]
where the subscript "s" refers to parameters for the standard.
50
-------
Use of equation 4-6 permits determining the soil concentration
of aromatic constituents. As in the water analysis, the vapor
concentration ratio in equation 4-6 may be replaced by a. peak
area or a peak height ratio obtained from GC analyses. Values
of Hi may be obtained as previously noted. In equation 4-6, the
headspace volume term for the vial containing soil is equal to
the total vial volume 'minus the sum of water and soil particle
volumes. The volume of the soil particles can be approximated
as Ma/Da, where D8 is the density of the soil particles, and is
about 2.65 g/cc. To be more precise, the density of the soil
particles would have to be determined. Ma would be the mass of
soil dried and Vv for the soil vial would include the initial
moisture content of the soil sample.
SAMPLING EQUIPMENT
Sam-pl Ing
Conventional means of sampling water and soil may be us&d.
Importantly, care should be taken to minimize exposuring the
sample to air. With regard to filling the VOA vial, it should
be filled in a manner to avoid air bubbles. A plastic 12 cc
syringe with the front end cut off may be used to sample soil
t
and readily introduce the sample to a VOA vial with the syringe
plunger.
VOA Vial Preparation
For increased precision and convenience, the weight and
51
-------
volumes of clean vials should be obtained before going to the
field. The vials should be numbered and the volume and weights
then recorded in a notebook. The volume of the vials may be
obtained by filling the vials with water, and by weighing the
vials with septum and caps before and after filling. With
respect to vials used in soil analysis, these should be
prefilled with a known amount of distilled water (about 25 cc).
Supplies ' '
Supplies needed for conducting analyses are listed in
Table 4.1.
TABLE 4.1 VOA METHOD EQUIPMENT
Water Analysis
* 40 cc clean VOA vials with hole caps and septa
* Gastight syringes (50 uL to 500 uL sizes)
* Luer lock syringe (10 to 30 mL size)
* Two luer lock needles (large bore)
* Field balance
* Distilled water
* Temperature bath
* Standard solutions of the desired analytes
* GC equipment and associated supplies
Soil Ana1vs1h
* Above materials for water
* Spatula or other method for obtaining soil sample
SAMPLING PROCEDURES
Steps in the analysis should in general include:
1. Verification of equilibration time — using a sample
52
-------
suspected of being contaminated.
2. Evaluation of vapor injection reproducibility.
3. Sample and standard analysis, with occasional replicates.
4. Vials should all be brought to the same temperature in a
temperature bath.
Water Analysis Procedures
1. Turn on the GC to allow column to condition and detector
to warm up.
2 .f Prepare a 40 cc vial of standard as follows:
a. Insert one luer lock needle (without syringe) through the
septa of a vial completely filled with the aqueous standards.
You should use caution here and'keep the end of the needle
aimed away from you (water may squirt out of the needle).
b. Insert the second needle, attached to the large syringe, and
withdraw 10 cc of liquid.
» i
c. Shake the vial vigorously for one minute.
d. Let the vial sit for at . least one minute.
3. Insert one of the gastight syringes through the septum of
the prepared standard vial and into the headspace above the
liquid. Do not allow the needle to get liquid in it at any
time.
(4) Flush the syringe barrel with the headspace vapors several"
times by moving the plunger up and down.
53
-------
(5) Remove an appropriate sized aliquot of vapors from the via]
and inject it onto the GC. The "appropriate size" will
obviously depend upon the concentration of the standard, the
instrument used, and its sensitivity to the particular
compounds.
6. Do a second injection of headspace from the standard tc
check instrument stability and repeatability.
7. When you are sure that the instrument is operating properly,
prepare your samples as in Steps 2a-d.
(8) Repeat the above steps to analyze the samples. Such a small
fraction of the total analyte is removed when sampling the
t
headpsace that numerous injections can be run from the same
vial with good reproducibility.
Soil Analysis Procedures
The above procedures for water should be conducted with
the aqueous standards used in soil analysis. Immediately
following soil sampling, the soil should be added to the vial
partially filled with distilled water. The vial should then be
reweighed to- determine the mass of soil added. We commonly use
25 cc of distilled water and 5 to 10 g of soil, although these
quantities may be adjusted up or down depending on the level of
contamination suspected. After the soil is introduced into,the
vial, it should be shaken to disaggregate the soil matrix, then
the vial should be placed in the water bath for temperature
54
-------
equilibration. The rest of the procedures are identical to that
for water starting at step 2c.
It is assumed in analyzing water and soil that free
product is not present in a vial following temperature
equilibration and agitation. The presence of free product will
invalidate this analysis. Hence, vials should be carefully
checked for free product.
CALCULATIONS
Water Concentration
*
Ose equation 4-3, if the headspace to water ratios of the
aqueous standard and samples are the same, otherwise use
equation 4-2. Sample concentration will be in the same units as
that of the aqueous standard.
Henrv's Law Constant
Use equation 4-4. The constant will be dimensionless.
Soil Concentration
Ose equation 4-6. If the aqueous concentration of the
standard is in micrograms per liter, soil mass is in grams,
volumes are in cubic centimeters' (milliliters), then the soil
concentration will be in micrograms per gram (parts per
million) which is equivalent to milligrams per kilogram.
55
-------
ADVANTAGES AND DISADVANTAGES
Advantages
1. Relatively high sensitivity—ug/1 range for water, and
fractions of a mg/kg range for soil (instrument dependent).
Sensitivity may be adjusted by changing headspace to water
ratios, mass of soil sample and equilibration temperature.
2. Method provides quantitative and constituent specific
data (see Section 2 on the use of chromatogram patterns for
source delineation).
3. Relatively rapid and reduces potential for constituent loss
due to biodegradation, sample shipment, etc.-
Disadvantages
1. Requires a significant degree of proficiency with GC
operation.
2. Requires temperature control of vials, hence, the needsvf orM#
temperature baths in the field. "
3. May be subject to errors brought about by salting-out
effects and soil adsorption.
4. Multicomponent mixture of gasoline vapors emanating
from either soil or ground water samples may cause
co-elution problems (e.g., inability to identify a peak of
interest, may identify the ..wrong peak, may overestimate
concentrations due to peak overlap).
5. Method is invalid if free product is present following
temperature equilibration and agitation of vials.
56
-------
SECTION 5
SOIL GAS SURVEYING WITH TOTAL ORGANIC VAPOR
DETECTION INSTRUMENTS
METHOD OVERVIEW
This section describes a serial dilution technique for use
with total organic vapor detectors when conducting soil gas
surveying, and a method for determining relative variations in
soil air permeability during the course of a survey. The
methods where originally described in Robbins and Temple
(1988), Robbins et al. (1990a) and Robbins et al. (1990b).
As discussed in Section 2, portable total organic vapor
detection instruments (e.g, explosimeters (ED), flame-
ionization detectors (FID), and photoionization detectors
(PID)) should not1 be used to sample soil gas probes directly.
The response of these instruments depends on their flow rate.
Because soil has a finite air permeability, direct-probe
sampling-will result in a vacuum being generated in the ground
which will curtail the operating flow rate of a total organic
vapor detector (which generally have internal air pumps or
fans). Variations in soil air permeability across a site can
introduce variable detector response to organic vapors, if
direct-probe sampling is conducted, and complicate, if not
57
-------
confound, interpreting survey results.
To circumvent the flow rate curtailment problem, soil gas
samples should be collected in air sampling bags for
measurement with total organic vapor instruments. Even then, as
noted in Table 2.5, the composition of the soil gas and organic
vapors in the gas can cause problematic detector response
during the course of a survey and complicate data
interpretation. If the soil gas oxygen level is below 10 to
13% (resulting from the- biodegradation of hydrocarbons in the
subsurface), the flame of a FID will extinguish when sampling a
bag. Also, one can not obtain soil vapor measurements with the
ED in its low scale mode (the ED will exhibit an initial upward
meter deflection followed by a drop-off in meter reading).
High levels of subsurface carbon dioxide, relative humidity,
and organic vapors can cause the response of the PID to become
non-linear and to be quenched. To circumvent these problems,
and to determine total organic vapor concentration in a bag
sample when the concentration exceeds the range of an
instrument, the serial dilution technique was developed.
The serial dilution technique entails performing a
systematic dilution of a gas bag sample. By taking gas bag
measurements during the dilution, and by extrapolating these
measurments on a semilog plot of concentration vs. dilution
increment, one can obtain a soil gas concentration measurement
for a stable instrument response.
Soil gas surveys are generally conducted prior to having
58
-------
detailed information on soil conditions at a site. As noted in
section 2, soil conditions can influence the volatilization of
gasoline that has leaked into the soil, the migration of
gasoline vapor in the soil and the detection of the vapor in
conducting soil gas surveying. In the absence of information
on soil conditions at a site, there is a significant level of
uncertainty associated with interpreting 6oil gas survey
results (concentration contour maps and vertical profiles) In this
section, a simple means of determining relative soil air
• <
permeability during the course of- surveying is presented to aid
in interpreting survey results. The technique involves taking a
series of flow rate and vacuum pressure measurements while
collecting a soil gas sample, and determining the slope of a
flow rate-vacuum pressure curve.
THEORY
Serial_P.i.lutiQn.
Consider collecting a soil gas sample, having detectable
gasoline vapor constituents, in an evacuated sampling bag using
the system shown in Figure 2.1 (p. 24). The sampling bag would
be attached to the exhaust port of the vacuum pump for sample
collection. Following sample collection a known volume of gas
is then removed and transferred to another evacuated Tedlar
bag. This second bag is then analyzed with a total organic
vapor detector. The initial bag is then refilled with a volume
59
-------
of clean air (i.e., air with no detectable constituents) equal
to that which was first removed. These steps are then repeated
keeping the volume of gas removed and clean air added equal and
constant at each Increment. In this manner, the initial volume
of the gas bag is restored after each refilling increment.
Using the above procedure, if the response factor is
assumed constant (see Section 3, p. 32-34, regarding response
factors), the measured concentration of the detectable
constituent in the sampling bag after each refilling increment
may be described by the following mass and volume continuity
where Cm(j) equals the measured concentration in the sampling
bag after the jth refilling increment, Cm(j-l) equals the
concentration in the sampling bag just prior to the jth
refilling increment, Vo is the initial bag volume, and Vr is
the volume of clean air added at each refilling increment.
Expanding Equation 5-1 from j = 1 to N, where N is the number
of refilling (i.e., dilution) increments, and factoring (Vo-
Vr)/Vo terms results in
expression
Cm (j ) = (Cm(j-l)Vo - Cm(j-l)Vr)/Vo,
(5-1)
Cm(N) = Cmo [(Vo - Vr)/Vo]l[(Vo - Vr)/Vo]2
. .. [(Vo - V r)/V o]N,
(5-2)
which simplifies to
Cm(N) = Cmo[(Vo - Vr)/Vo]H,
(5-3)
60
-------
where Cmo equals the initial measured concentration in the gas
sample. Taking the logarithm of both sides of Equation 5-3
results in
Log{Cm(N)} = Log{Cmo} + NLog[(Vo - Vr)/Vo]. (5-4)
Equation 5-4 predicts a log-linear relation between Cm(N) and
N. The intercept of the equation is the initial bag
concentration, Cno. The slope of the log-linear curve is the
logarithm of the volume terms. Thus, for log-linearity to hold,
the volumes out and in (Vr) during each increment must remain
constant and equal. It should be noted that volume terms in the
equation may be replaced by the product of air flow rate and
pumping time, and N by the total dilution time divided by the
time for a dilution increment. Since the dilution increment is
a constant, a plot of Cm(N) verses total dilution time will
also be log-linear.
The substitution of Equation 3-1 into 5-4 gives Equation
5-5,
Log{Ri.(N)C(N)} = Log{RioCo} + NLog[(Vo - Vr)/Vo]. (5-5)
In Equation 5-5, Ri(N) and C(N) are the response factor and
actual concentration at the Nth dilution increment if a single
detectable constituent where presentj respectively. Their
product is the measured concentration. Likewise, Rio and Co
are the response factor and actual concentration in the initial
gas bag, respectively. Their product would be the initial
61
-------
measured concentration which is the intercept of a plot of
Log{Ri(N)C(N)} versus N. Where multiple constituents are
present, the Ri term would be a composite response factor equal
to the sum term in equation 3-2 and the concentration terms
would refer to total detectable concentrations.
Figure 5-1 illustrates idealized dilution curves (from
Robbins et al., 1990a) described by Equation 5-5. The shapes of
the curves depend on how the response factor varies as the
initial gas in the bag is serially diluted with clean air.
Curve A represents the case where the response factor is
5.
constant over the serial dilution. Curve B illustrates the case
where initially the response factor is low and then it
increases to a constant value. In this case, the proportional
increase in the response factor with dilution is less than the
proportional decrease in concentration. The ratio of
concentrations between curves B and A at any dilution increment
equals the response factor for the gas conditions in the bag at
that increment. To circumvent variable response factor
conditions, the log-linear portion of curve B can be
extrapolated to the concentration intercept. This concentration
would equal the concentration measured in the sampling bag
prior to dilution for the constant response factor condition of
curve A. Curve C represents the case where the response factor
was initially low. to the left of point 1, the proportional
increase in the response factor with dilution is greater than
the proportional decrease in concentration. To the right of
62 i
-------
point 1, the shape of the curve is due to the same conditions
that govern curve B. Thus, by performing a serial dilution on
a bag sample, ploting measured concentration vs. dilution
increment on a semilog plot, and extrapolating the log-linear
portion of the dilution curve to its intercept, one can
circumvent variable response factor effects and determine the
initial bag concentration for constant response conditions.
Relative Soil Air Permeability
The permeability of soil to air will depend on such
factors as the lithology, grain size, grain size distribution,
porosity and moisture content of the soil. Hence, measurement
of the relative air permeability of soil at soil gas probe
locations permits assessing subsurface variation in these
factors. These same factors, of course, influence subsurface
ga6 migration and detection.
Consider installing the same type of soil gas probe in the
same manner throughout a survey, and attaching the sampling
system shown in Figure 2.1, p. 24, to the probes. Shortly after
the pump is turned on a constant flow rate and vacuum pressure
will be achieved. If the sampling system is the same and
probes are installed in an identical manner at each survey
point, the flow rate and vacuum pressure'achieved at each probe
location will reflect the soil air permeability around the
probe. If the vacuum pressure is maintained at low gage
pressure relative to 1 atmosphere (408 inches of water), the
63
-------
c
o
D
s_
C
(1)
U
c-
o
U
O)
o
I
Di lufions
'igure 5.1 Possible Forms of Serial Dilution Curves
(from Robbins et al. (1990a)
64
-------
air may be considered to be incompressible. For this situation,
the theory describing pumping air at a constant flow rate is
analogous to that describing pumping ground water at a constant
rate. Hence, the change in air flow rate with change in vacuum
pressure during sampling will be directly proportional to the
soil air permeability. For the sampling system, the flow rate
and vacuum pressure may be adjusted using the system valve.
Each adjustment of the valve will result in a change in flow
rate and a change in vacuum depending on the soil air
permeability. The slope of a plot of flow rate vs. vacuum
pressure will therefore be proportional to the soil air
permeability. The determination of an absolute value for soil
air permeability will depend on the configuration of the probe
which is beyond the scope of this work.
SAMPLING EQUIPMENT
To perform serial dilutions and relative soil air
permeability tests, a sampling system as shown in Figure 2.1 is
needed. It is suggested that vacuum pressure be maintained
below 15 inches of water for sampling and performing air
permeability tests. With this requirement, the capacity of the
pump will depend on the soil air permeability anticipated and
whether air permeability tests are to be performed. If tests
are not to be performed, a very low capacity pump is suggested.
For silty soils we have found a pump having a maximum capacity
65
-------
of about 3 liters/min. to be adequate to perforin soil air
permeability tests. For sandy soil, one may need a pump having
at least 10 times this capacity. Also, one may need pressure
gages having a lower range than 15 inches of water to resolve
pressure changes.
For gas sampling and performing serial dilutions, you will
need two air sampling bags, a stop watch, semilog graph paper
and a straight edge. We have found Tedlar sampling bags, having
a sliding shut-off valve and a capacity of 3 liters to be
adequate. To facilitate connecting the bags to the pump,
*
polyethylene tubing and Tygon tubing can be connected to the
shut-off valve on the sampling bags. To analyze flow tests for
soil air permeability you will need normal graph paper.
SAMPLING PROCEDURES
Serial Dilution Testing
The following steps are recommeded for performing serial
dilutions.
1. Using the sampling system, evacuate the two clean sampling
bags. Attach the system to a probe, turn on the pump and allow
the system to purge. Connect one bag to the exhaust port of the
pump, turn down the system valve to achieve a low flow rate,
open the bag valve, time how long it takes to fill the bag and
then isolate the bag by closing its shut-off valve.
2. Disconnect the sampling system from the probe and purge it.
66
-------
Connect the initial sampling bag to the tubing used in
connecting the system to the probe. Attach the second bag to
the exhaust port of the pump. Turn on the pump, adjust the flow
rate using the system valve to achieve the initial flow rate
during sampling. Open the initial bag valve to the system, wait
several seconds to allow for purging, then open the second bag
to the system to transfer gas between bags. The total time for
gas transfer starting when the initial bag was open to the
system should be noted. It is suggested that the total time for
gas transfer be 1/3 to 1/2 that used in originally filling the
bag when probe sampling, although you can adjust this to any
desired time. Just before this period is achieved, close the
second bag. When this time is achieved, close the original bag.
3. Sample the second bag. If no reading is obtained, sampling
at this probe is concluded. If a flame out occurs while using
an FID, note this. If an initial rise then fall-off deflection
occurs on an ED meter, note this. If a reading is obtained
while using a PID, FID or ED, record it and plot the reading
on the semi-log paper. The log axis on the paper should be
labelled concentration. The regular coordinate axis should be
labelled dilution,increment. The first sampling is the zero
dilution increment.
4. Remove the original bag from the sampling system and purge
the system. Purge and evacuate the second bag using the
sampling system. Purging may be conducted by repeatedly filling
-------
the bag with clean air and evacuating it using the sampling
system. The bag should be checked for purging completeness by
attaching it, when filled with clean air, to your detector.
5. Purge the sampling system. Attach the original bag to the
exhaust port of the pump. Turn the pump on, adjust the flow
rate to the original sampling rate. Open the original sample
bag valve. Refill the bag to its original volume with clean
air. The pumping time should equal the full transfer time in
step 2.
&
6 Repeat steps 2 through 5. When you observe for at least
three dilution increments where the fraction of concentration
dilution with each increment is to equal the fraction of volume
represented by the . the transfer volume, the procedure is
finished. That is, if the transfer volume is 1/3 the original
bag volume (also equivalent to the fraction of transfer time
relative to the initial filling time), you may stop performing
dilutions when you observe the concentration to change by 1/3
between three successive dilutions.
7. Using the three data points from step 6, construct a best
fit straightline on your graph paper and determine the
intercept (zero dilution increment) concentration.
It is suggested that you practice the procedure using a
gas standard. If you are using a PID, you should practice
performing dilutions using clean, drv air. In the field, you
68
-------
may use clean ambient air with the PID. You can maximize your
sensitivity with the PID in the field if you use clean, dry
air. Also, the dilution procedure should be performed
immediately after sample collection.
Soil Air Permeability Tent Procedure
After the initial bag sample is taken. You can perform the
soil air permeability test by taking several flow rate and
vacuum pressure readings (at least three) using the sampling
system. To obtain readings, just increase or decrease-the
system 'flow rate using the valve on the system. After each
adjustment of the valve, wait a few seconds for the system to
stabilize then record the readings. Plot the readings as shown
of Figure 5.2. Draw a best fit curve through the data. Then
take the slope of the curve. Slope values obtained may then be
contoured or a vertical profile prepared to provide an
indication of the spatial distribution of soil air permeability
at a site.
ADVANTAGES AND DISADVANTAGES
Advantages
1. The serial dilution technique circumvents problematic
responses of total organic vapor dete'ctors in performing
serial dilutions.
2. The serial dilution method permits determining vapor levels
if a detection instrument's maximum measurement reading is
exceeded.
3. With practice, the method takes only a few minutes.
69
-------
4. The relative soil air permeability method provides
information on variations in soil properties across a site,
useful in interpreting the spatial distribution of detected
organic vapors.
Disadvantages
1. The serial dilution technique requires a significant level
of proficiency.
2. The technique will no-c circumvent tne inherent varing
response problem of total: organic vapor: detectors when, and
if, the relative abundance of vapor:constituents ^change
between soil gas probes-.
3.% Soil air permeability results may be influenced by soil
jj disturbance- that can occur when soil gas probes are
installed.
70
-------
' 2500-1
2000
~1500
500-
. 2.0 "4.0 .6.0 . 8.0 .: 1<
Gauae Pressure (inr.h of
Figure 5.2 Example of a flow rate-vacuum pressure
curve for determining relative soil air
permeability (from Robbins and Temple,
1988).
71
-------
SECTION 6
SOIL GAS SURVEYING FOR OXYGEN AND CARBON DIOXIDE
AS INDIRECT INDICATORS OF SUBSURFACE GASOLINE
LEAKAGE
METHOD OVERVIEW
This brief section describes soil gas surveying for oxygen
and carbon dioxide using portable meters as indirect indicators
of- subsurface gasoline leakage. Soil gas carbon dioxide
readings may be determined using a portable infrared meter and
the serial dilution technique described in Section 5. Soil gas
oxygen may be determined using 'a portable electrochemical
galvanic 6ensor, commonly used in monitoring oxygen levels in
confined work spaces. To conduct oxygen measurements, the meter
sensor is sealed in a pipe connected in series with the exhaust
port of the soil gas sampling system shown in Figure 2.1.
THEORY
Ambient air has oxygen and carbon dioxide levels of about
20.9 and 0.05 on a volume (molar) percent basis, respectively.
Natural aerobic and anerobic degradation of organic matter in
soil due to soil organisms, principally bacteria, results in a
depletion in subsurface oxygen and an enrichment in subsurface"
carbon dioxide, relative to that of air. The extent to which
72
-------
this occurs depends on a host of factors including subsurface
temperature, the types and amounts of micro-organisms present,
the exchange of gases between the subsurface and surface, the
type and amount of carbon source present, and the availability
of nutrients for organisms. As discussed in Section 2, the
hydrocarbons in gasoline are a carbon source for subsurface
organisms. The biodegradation of gasoline in the subsurface can
result in the generation of additional carbon dioxide (either
from aerobic or anerobic microbial degradation) and depletion
of subsurface oxygen (consumed in aerobic degradation),
\
relative to ambient conditions. For example, we have found at a
site of subsurface gasoline leakage elevated levels of carbon
dioxide up to 19 % and a drop in subsurface oxygen level down
to about 0.5 % in the area of known subsurface gasoline
contamination. Further, where aerobic degradation occurs,
theory would predict an inverse, one to one, correlation
between the concentration of these gases. That is, for each
mole of oxygen consumed, one mole of carbon dioxide should be
generated. Where anerobic degradation takes place, carbon
dioxide will be generated (along with methane) in the absence
of oxygen consumption. We have observed such effects in
conducting soil gas surveying. Hence, this method provides a
means to indirectly assess whether subsurface gasoline leakage
has occurred by conducting soil gas surveying for oxygen and
carbon dioxide. Analyzing for these gases can also provide
indications as to the extent to which natural degradation may
73
-------
be active at a site, both spatially and temporally. Also, these
parameters have use in monitoring soil clean-up by venting
systems, as well as ground water remediation by pumping or by
in situ bioremediation schemes.
SAMPLING EQOIPMENT AND PROCEDURES
Carbon Dloxidft
Carbon dioxide in 6oil gas samples may be measured in the
field using a portable infrared (IR) meter. It is suggested
that the meter have the capability to read up to 20% carbon
dioxide. A soil gas sample may be collected in a gas sampling
bag using the sampling system shown in Figure 2.1, by attaching
an evacuated bag to the exhaust port' of the pump. If the carbon
dioxide level in a sampling bag exceeds the maximum reading of
an instrument, the bag may be diluted using the serial dilution
technique described in Section 5 to circumvent this problem.
The reader is referred to Section 5, regarding details on the
use of sampling bags, and procedures for bag and system
purging.
Oxygen
A commonly used portable meter (electrochemical galvanic
cell) for confined space monitoring may be used to conduct soil
gas surveying for oxygen. Meters of this type generally have an
external probe connected by wiring to the meter housing. To
sample soil gas oxygen, the external probe may be placed in
74
-------
series with the exhaust port of the pump. This may be
accomplished by constructing a leak proof housing for the
probe. One that we have used consists of a 10 cm length of 3.17
cm (1.25") schedule 40 PVC pipe, caped at both ends with PYC
end caps. Tapped into the end caps are two quick connect
valves. The valves permit testing for leaks after the probe is
inserted and sealed. One end of the pipe is notched. After the
probe is placed in the pipe, the wire lead from the probe is
positioned into the notch. The caps are then pushed on the pipe
and the pipe and caps are wrapped with Parafilm. With both
quick connect valves open, the pipe is then attached to the
exhaust port of the pump using polyethylene and Tygon tubing.
When sampling a soil gas probe, it may take about a minute for
the oxygen meter to reach a stable reading, depending on your
pumping rate.
ADVANTAGES AND DISADVANTAGES
Advantages
1. The oxygen and carbon dioxide measurements are
straightforward and rapid to perform using inexpensive
equipment.
2. The measurements can provide indirect and supplemental
information on subsurface gasoline leakage and degradation.
Disadvantages
1. The measurements provide indirect indications of subsurface
gasoline leakage requiring confirmation.
2. Variations in subsurface conditions and background gas
75
-------
levels can complicate interpreting gas measurements. This
may be especially the case in soils with abundant organic
matter and carbonate minerals.
76
-------
SECTION 7
FIELD MEASUREMENTS OF INORGANIC GROUND WATER
QUALITY ATTRIBUTES OF GASOLINE CONTAMINATION
OVERVIEW
In Section 2, inorganic ground water quality attributes
arising from biodegradation, of gasoline contamination wher
described. In this section, we briefly provide a summary o
field methods that can be used to measure the water qualit
attributes.
METHOD SUMMARY
In Section 2, Table 2.4, p. 18, lists key inorganic wate
quality parameters that we have observed to provide a
indication of gasoline contamination of ground water. In Tabl
7.1, we have provided a listing of field methods that may b
used to measure the parameters. We have experience using som
of the methods listed. Other alternative methods are include
in the table because commerical field kits are available. I
the ground water is colored or turbid (refer to the discussio
in Section 2), methods that rely on colorlmetry may b
difficult to accurately perform. Also, if the ground water i
colored or turbid, use of a pH meter for monitoring
77
-------
titration, rather than color indicators, is suggested.
TABLE 7.1 ALTERNATIVE FIELD METHODS FOR MEASURING
INORGANIC WATER QUALITY ATTRIBUTES
Parameter
PH
Specific conductance (EC)
Alkalinity
Dissolved 02
Dissolved CO2
J
Electrode Potential ("Eh")
Nitrate
Chloride
Sulfate
Iron
Manganese
Methodology
Field meter
Field meter
Titration
Membrane electrode
Colorimetry
Membrane electrode
Titration
Platinium electrode
Specific ion electrode
Colo'rimetry
Specific ion electrode
Spectrophotometry
Colorimetry
Spectrophotometry
Colorimetry
Colorimetry
Several points should be kept in mind in interpreting
inorganic attribute data. Firstly, these measurements are
indirect indicators of gasoline contamination that require
confirmation. Secondly, background • -variations in these
parameters may occur at a site. Thirdly, the background water
quality and soil mineralogy of a site may serve to mask, or
inhibit, the development of the inorganic attributes. Finally,
78
-------
if determined from monitoring well sampling, the absolute and
relative levels of the parameters may be influenced by mass
continuity bias (refer to section 2).
79
-------
REFERENCES
Atlas, R.M., 1981, Microbial degradation of petroleum
hydrocarbons: an environmental perspective,
Microbiological Reviews, v. 45, no. 1, pp. 180-209.
Barker, J.F., and G.C. Patrick, 1985, Natural attenuation of
aromatic hydrocarbons in a shallow sand aquifer,
Proceedings of Petroleum Hydrocarbons and Organic
S Chemicals in Ground Water: Prevention, Detection and
Restoration Conference. National Water Well Association.
Houston, TX. pp. 150-177.
Chiou, C.T., P.E. Porter and D.,.. u^uiucuuius, iooo, rarGition
equilibria of nonionic organic compounds between soil
organic matter and water. Environ. Sci. and Technol., v.
17, no. 4, ppi: 227-231.
Clay, P., and T.M. Spittler, 1982, The use of portable
instruments in hazardous waste site characterizations.
Proceedings of the National Conference on Management of
Uncontrolled Hazardous Waste Sites. HMCSI, Silver Spring,
MD. pp. 40-44.
Garret, P., 1987, Oxygenates as ground water contaminants,
Proceedings of the Conference on Alcohols and Octane,
San Antonio, TX. p. 1-10.
80
-------
Garrett, P., M. Moreau, and J.D. Lowry, 1986, MTBS as a ground
water contaminant, Proceedings of Petroleum Hydrocarbons
and Organic Chemicals in Ground Water: Prevention,
Detection and Restoration Conference. National Water Well
Association, Houston, TX. 12 p.
Gosset, J.M., 1987, Measurement of Henry's law constants for
C-l and C-2 chlorinated hydrocarbons, Env. Sci. and Tech.,
v. 21, no. 2, pp. 202-208.
Griffith, J.T., G.A. Robbins, and T.M. Spittler, 1988, A
new method for field analysis of soils contaminated with
%
f:
i1 aromatic compounds. Proceedings of the Focus Conference or
Eastern Regional Ground Water Issues. National Water Well
Association, Stamford, CT. pp. 223-248.
Hoag, G.E., C.J. Bruell and M.C. Marley, 1984, A study of the
mechanisms controlling gasoline hydrocarbon partitioning
and transport in groundwater systems, University of
i «
Connecticut, Institute of Water Resources Rept. 3707 06,
51 p.
Kerfoot, H.B., C.L. Mayer, P.B. Durgin, and J.J. D'Lugosz,
J^J.V 1988, Measurement of carbon dioxide in soil gases
for indication of subsurface hydrocarbon contamination,
Ground Water Monitoring Rev. Y.8, No. 2, pp. 67-71.
Mackay, D. and P.J. Leinonen, 1975, Rate of evaporation of
low-solubility contaminants from water bodies to
atmosphere. Environ. Sci. and Technol., v.9, no. 19, pp.
1178-1180.
81
-------
Mackay, D., and W.Y. Shiu. 1981. Critical review of Henry's
law constants for chemicals of environmental interest. J.
Phys. Chem. Ref. Data, v. 10, no. 4, pp. 1175-1199.
Marrin, D.L., 1985, Delineation of gasoline hydrocarbons
in groundwater by soil gas analysis. Proceedings of
HazMat West Conference. Long Beach, CA. pp. 112-119.
Marrin, D.L., 1987, Soil gas analysis of methane and carbon
dioxide: delineating and monitoring petroleum
hydrocarbons. Proceedings of Petroleum Hydrocarbons and
Organic Chemicals in Ground Water: Prevention, Detection
and Restoration Conference. National Water Well
Association, Houston, TX. pp. 357-367.
McNabb, J.Fi, and W.J. Dunlap, 1975,^ Subsurface biological
activity in relation to ground water pollution, Ground
Water Jour., v. 13, no. 1, pp. 33-44.
Robbins, G.A., 1989, Influence of using purged and partially
penetrating monitoring wells on contaminant detection,
mapping, and modeling, Ground Water Jour. v. 27, no. 2,
PP. 155-162.
Robbins, G.A., R.D. Bristol, J.M. Hayden, and J.D. Stuart,
1989a, Mass continuity and distribution implications for
collection of representative ground water samples from
montoring wells, Proceedings of Petroleum Hydrocarbons anc
Organic Chemicals in Ground Water: Prevention, Detection
and Restoration Conference. National Water Well
Association, Houston, TX. pp. 125-139.
82
-------
Robbins, G.A., R.D. Bristol and Y.D. Roe, 1989c, A field
screening method for gasoline contamination using a
polyethylene bag sampling system. Ground Water Monitoring
Review, v. 9, no. 4, pp 87-97.
Robbins, G.A. , B.G. Deyo, M.R. Temple, J.D. Stuart, and M.J.
Lacy, 1990a, Soil gas surveying for subsurface
gasoline contamination using total organic vapor detection
instruments, Part I, theory and laboratory
experimentation, Ground Water Mont. Rev., v. 1U, no. 3;
(in press).
%
Robbins, G.A., B.G. Deyo, M.R. Temple, J.D. Stuart, and M.J.
Lacy, M.J., 1990b, Soil gas surveying for subsurface ,
gasoline contamination using total organic vapor,..detect ion,
instruments, Part II, field experimentation, Ground Water
Mont. Rev., v. 10, no. 4, (in press).
Robbins, G.A., J.M. Hayden, and R.D. Bristol, 1989b, Vertical
dispersion of* ground water contaminants in the near-field
of leaking underground gasoline storage tanks, Proceedings
of Petroleum Hydrocarbons and Organic Chemicals in Ground
Water: Prevention, Detection and Restoration Conference.
National Water Well Association, Houston, TX.
pp. 373-381.
Robbins, G.A., and M.R. Temple, 1988, Detection of subsurface
gasoline contamination in New England glaciated terrain
using soil gas surveying. Proceedings of the Focus
Conference on Regional Ground Water Issues. National Hater
83
-------
Well Association, Stamford, CT. pp. 635-648.
Roe, V.D., M.J. Lacy, J.D. Stuart, and G.A. Bobbins, 1989,
Manual headspace method to analyze for the volatile
aromatics of gasoline in ground water and soil samples,
Anal. Chem., v. 61, pp. 2584-2585.
Schwille, F., 1976, Anthropogenically reduced groundwaters,
Hydrological Sci. Bull., v. XXI, no. 4, pp. 629-645.'
Tchobanoglous, G., and E. D. Schroeder, 1985, Hater Quality.
Addison-Wesley Publishing Company, Reading, MA, 768 pp.
Thompson, G., and D. Marrin, 1987, Soil gas contaminant
| investigations: a dynamic approach. Ground Water
Monitoring Review, v. 7, no. 3, pp. 88-93.
U.S. Environmental Protection Agency, 1982, Test methods for
evaluation solid waste, physical and chemical methods,
SW-846, 2nd ed.
Wadden, R.A., I. Uno and S. Wakamatsu, 1986, Source
discrimination of short-term hydrocarbon samples measure
aloft. Environ. Sci. and Technol., v. 20, no. 5,
pp. 473-483.
Wilson, B.tf., B.E. Bledsoe, D.H. Kambell, J.T. Wilson, J.M.
Armstrong, and J.H. Sammons, J.H., 1986, Biological fate
of hydrocarbons at an aviation gasoline spill site,
Proceedings of Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention, Detection and
Restoration Conference. National Water Well Association,
Houston, TX. pp. 78-90.
-------
Wilson, B.H., and J.F. Rees, 1985, Biotransformation of
gasoline hydrocarbons in methanogenic aquifer material",
Proceedings of Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention, Detection and
Restoration Conference. National Water Well Association
Houston, TX. pp. 128-139.
85
------- |