THE USE OF INDUSTRIAL HYGIENE SAMPLERS
FOR SOIL-GAS MEASUREMENT
-------
EPA 600/4-89/008
PB89-166359
THE USE OF INDUSTRIAL HYGIENE
SAMPLERS FOR SOIL-GAS MEASUREMENT
bv
H B Kerfoot and C L Maver
Lockheed Engineering and Management Services Companv
Eiivnonmental Programs
Las Vegas Nevada S9I 19
Coniraci No 68-03-1249
Proiect Officer
P B Durgin
Advanced Monitoring Division
Environmental Monitoring Svstems Laboratory
Las Vegas Nevada 89114
ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
LAS \ EGAS. NEVADA 89114
-------
NOTICE
The information in ihis document has been funded whollv or in pan bv the United States
Environmental Protection Agencv under contract number 68-03-3249 to Lockheed Engineering and
Management Services Companv Incorporated It has been subject 10 the Agencv s peer and
administrative review and it has been approved for publication as an EPA document Mention of
trade names or commercial products does not constitute endorsement or recommendation for use
ii
-------
ABSTRACT
The Lockheed Passive Sampling Svstem (LPSS) for soil-gas measurement was field tested at
the Pittman Lateral near Henderson. Nevada The svstem uses n sampler consisting ot an
industrial hvgiene organic vapor monitor inside a metal sampling manifold buried at a depth of
approximately 0 3 meters (I foot) Samplers are analyzed off site bv NIOSH Method P&CAM 127
¦\t the Pittman Lateral unconfirmed ground water occurs at 1 5 to 4 3 meters l5 to 14 feet)
heneath the ground surface Two distinct ground-water volatile organic contaminant plumes exist
ni the site one is primarily chloroform and the other benzene and chlorobenzene Both
plumes have been delineated bv repeated sampling and analvsis of ground water from monuoiing
•lells located at b4-meter i200-foot) intervals along a line perpendicular to the direction ot
giound water flow In a previous studv funded bv the U S EPA Environmental Monitoring Svstenis
Laboratory Las Vegas Nevada (EMSL-LV) a grab-sample/on-site analvsis technique and a
different passive-sample/remote analvsis method were field-iesied at ihis same sue
The LPSS technique successfully delineated a chloroform soil-gas plume above ground water
contaminated wuh chloroform at concentrations up to 800 fig per liter The chloroform soil-gas
concentrations at a depth ol 0 3 meters (I foot) measured bv LPSS correlate strongly with
soil-gas concentrations at I 3 meters (4 feet) measured in the earlier grab-sample/on-site
analvsis studv The results of the LPSS method also correlate well with ground-water chloroform
concentrations The variability among results separated bv short distances tea I meter) was
less than that tor grab sampling wuh on-site analvsis and was between 12 and 22 percent
iclative standard deviation The vnnabilitv among 10 samplers over a 27-foot distance parallel
10 die direction of ground water (low was 12 percent ielative standard deviation The precision
of the LPSS results used to estimate an empirical relationship for the spatial resolution
lor the technique ai this sue This spatial icsolution was between 10 and 100 feet depending
on the concentration ol chloroform in the soil gas The efficiency of the sampler is seen to
\arv wuh chloroform concentration suggesting a possible concentration-dependent hunuditv
effect LPSS measurements were also made at four locations in a I-meter (3-foot) square pattern
above ground water containing approximately 5000 £tg per liter benzene and 4000 /Jg per liter
Lhlorobenzene Although the samplers were buried for three months, no delectable amounts of
benzene or chlorobenzene were measured upon analvsis Aldehvdes esters and ethylene dibromtde
were detected on the samplers These results agree with results from the earlier studv and are
eonsisteni wuh subsurface oxidation eliminating benzene and chlorobenzene Irom the vndose zone
lii
-------
TABLE OF CONTENTS
Page
Abstract M,
Figuies M
Tables vn
Acknowledgement
Introduction I
Objectives and Goals |
Background I
Conclusions and Recommendations 9
Experimental 10
Site Description 10
Apparatus and Procedures 12
Evaluations 15
Results and Discussion |7
Performance of the LPSS Above the Chloroform
Ground-water Plume |7
Performance of the LPSS Above the Benzene/
Chlorobenzene Ground-water Plume 24
Qualirv Control 24
References 25
V
-------
FIGURES
Page
Figure I VOC Transport troni the Water Table to the Atmosphere 1
Figure 2 Cross-Section of a Tvpical Diffusional Sampler 7
Figure 3 Subsurface at the Pittman Lateral 10
Figure 4 Lockheed Passive Sampling Svstem (LPSS) 12
Figure 5 Sampling Locations at the Studv Site 14
Figure b Scatter Plot of Passive Sampling Versus Giab-Sampling
Results 20
Tigure 7 Grab Sampling and LPSS Passive Sampling Soil-Gas
Chloiotorm Concentrations and Ground-Water Chloroform
Concentrations 21
VI
-------
FIGURES
Page
Figure I VOC Transport from ihe Water Table to (he Atmosphere J
Figure 2 Cross-Section of a Tvpical Diffusional Sampler 7
Figure 3 Subsurface at the Pittman Lateral 10
Figure 4 Lockheed Passive Sampling Svstem (LPSS) 12
Figure 5 Sampling Locations at the Studv Sue 14
Figure t> Scatter Plot of Passive Sampling Versus Oiab-Sampling
Results 20
rigure 7 Grab Sampling and LPSS Passive Sampling Soil-Gas
Chloiolorm Concentrations and Ground-Water Chloroform
Concentrations 21
vi
-------
TABLES
Page
Table I Vtost Frequerulv Identified Substances at 546 Superfund Sites 2
Table 2 Concentrations ot Chloroform Benzene and Chlorobenzene (yg/L) in
Ground-Water Samples Collected from Wells Along the Pittman
Lateral I I
Table 3 Performance ol the 3M 3500 Organic Vapor Monitor 13
Table 4 Results ol Analvses ol Organic Vapor Monitors for Chloroform 18
Table 5 Comparison ol Soil-Gas Passive and Grab Sampling Results with
Giound-Wnter Analvses 19
Table t> Comparison of Results from Passive and Grab Sampling Methods 19
Table 7 Results ol Closek Spaced Passive Soil-Gas Samplers 22
Table 8 Spatial Resolution of Soil-Gas Measurements bv Passive Sampling 22
Table 9 Sampler Efficient Studv Results 23
-------
ACKNOWLEDGEMENT
We acknowledge Mohammed J Miah of Lockheed-EMSCO for help wnh statistical evaluation of
the data The assistance of R L Pieper and the Corporation in providing organic \apor
moniiors and analytical support is appreciaied
viii
-------
INTRODUCTION
OBJECTIVES AND GOALS
This report describes a field evaluation of a passive-sampling technique for soil-gas
surveving The Field evaluation was performed at a site where other soil-gas measurement
methods have been evaluated (I) The objectives of the evaluation were to determine the degree
ot correlation of results with those from a previous grab-sample survev at the site, to assess
the correlation of results with ground-water volatile organic compound (VOC) concentrations to
evaluate the precision of results from closelv spaced samples as a function of VOC concentra-
non to estimate the spatial resolution of the method and to test (he sampler/manifold
assembly tor its etficiencv in trapping VOCs in the field
BACKGROUND
General
Sampling and analvsis for evaluation of contamination of soil and ground water bv organic
compounds is tradmonallv performed bv drilling boreholes and collecting and analvzing soil
cores and ground-water samples Because of the high cost and slow turnaround time of analvses
workers in the area of site-characterization have turned to alternative rapid inexpensive
preliminary field-reconnaissance techniques for guidance in selecting the most appropriate
locations tor ground-water sampling For highlv conductive dissolved ground-water contaminants
telectrolvtes) such as salts and strong acids or bases measurement of the conductivity of
ground water hv surtace geophvsical techniques has been successful (2) However for organic
contaminants such techniques are not applicable since these compounds rarelv a Iter the
electrical properties of ground water For organic compounds which are volatile soil-gas
measurement has emerged as a cost-effective reconnaissance technique for preliminary site
characterizations (13 4)
Organic compounds are a major threat to aquifers because of their potential long residence
times (5) Due to their high activirv VOCs are more mobile than other compounds Table I
shows a list of the 25 most frequently identified substances regulated under the Comprehensive
Environmental Response Compensation and Liabilitv Act (CERCLA or Superfund) of ihese 15 are
VOCs amenable to soil-gas surveving In addition studies at the U S EPA Environmental
Monitoring Laboratory m Las Vegas. Nevada (EMSL-LV) have shown that VOCs are rvpicallv the
first detected when ground water is being contaminated hv leaching from a hazardous-waste
disposal site (6) Majot components of gasolme and jet fuels aie also VOCs so that sotl-gas
measurement is applicable to mapping the areal extent of contamination from such mmuies (7)
and can be the basis of leak-detection systems lor underground storage tanks (8) The hehavioi
ot VOCs in ihe subsurface and the principles behind soil-gas ctnveving are discussed below
Volatile Organic Compounds in the Subsurface
When VOCs are present in the subsuitace eithei in an oigamc liquid phase oi dissolved m
giound water several processes under both thermodynamic and kinetic conttol can take place
In the saturated zone VOCs can undergo sorption to organic matter in the soil liquid diffusion
I
-------
TABLE 1. MOST FREQUENTLY IDENTIFIED SUBSTANCES AT 546 SUPERFUND SITES'
Rank
Substance
Henry's Law
Constant0
Percent of Sites
t
Tnchloroeihvlene11
0 38
33
2
Lead and compounds'5
NA
30
Toluene
0 23
28
-4
Benzene
0 24
26
5
Polvchlorinated biphenvls (PCBs)
-0
22
b
Chloroform1-
0 15
20
7
Teirachloroethvlene*
0 59
16
8
Phenol
-0
15
<)
Arsenic and compounds'3
NA
15
10
Cadmium and compounds'3
ma
15
1 1
Chromium and compounds'3
na
15
12
I l I-Trichloroethanee
0 67
14
1 "i
Zinc and compounds'3
MA
14
14
Ethvlbenzene *¦
0 28
13
15
Kvlene*-
0 20
1 "i
16
Methvlene chloride11
0 08
12
17
fcurs-l 2-Dichloroeihvlene
0 33
1 1
IS
Mercury and compoundsb
MA
10
19
Copper and compounds'3
ma
9
:o
Cyanides (soluble salts)b
ma
S
21
Vinvl chloride*
47 (I0°C)
8
22
1 2-Dichloroethane*
0 044
8
23
Chlorobenzene*
0 14
8
2-1
1 I-Dichloroeihane1'
0 23
8
25
Carbon tetrachloride1"
1 0
7
'ComponiHls dnicnnhle fo deieuion h\ sotl i»ck
3
VilHpitii tmni RtrleierKe !
b
This Uriss 01 compound; is rilmoM imniidhtv non-vnlnhle
Hcnii i Lhw consi,inl is ihe rtiiiilifrriuin 1M' lontt-niMiuni (/it'Ll i'l ,i u>l
Wt/L) .ii 25"C
ihli i oiii[>oiiml th\ iilul ¦¦
s i Oil; i nullum mi n IK 1
NA Nni
2
-------
vertical and horizontal mixing due to dispersion biotransformations, and chemical reactions
(5) tn the \adose (unsaturated) zone the vapors Irom these VOCs can undergo gas-phase
diffusion vapor/sorption equilibria vapor/solution equilibria biotransformations, and
chemical reactions In the intermediate zone (capiliarv fringe) the potential can exist for anv
ol the above processes to occur Bv definition the capiliarv tringe is fullv saturated The
water there is held bv capiliarv forces or soil suction and so is under less than atmospheric
pressure In actualitv heterogeneirv in the size and shape of \adose zone materials can result
in the presence of adjacent air-filled and water-filled channels in the capillary tringe
Although there are manv processes that can affect VOCs in the subsurface, the list ol
processes that are important where soil-gas surveving is applicable is extrenielv limited In
order to detect VOCs in soil gases above contamination, the rate of introduction of VOCs into
the soil atmosphere must be sufficient to maintain a measureable VOC concentration there That
situation will occur only when the rate-limiting step in vertical transport of VOCs from the
Loniannnation (source) to the atmosphere (sink) is diffusion through the vadose zone Undei
such conditions the VOC \apor concentiation above the capiliarv fnnge/vadose zone inteiface
will he at a concentration controlled bv a dvnamic vapor/solution equilibrium while VOC
(.oncentrations in soil gases between there and the soil surface will be determined bv the rate
ot diffusion of the gases through the soil (9) Figure I depicts idealized VOC transport from
ground water to the atmosphere
SATURATED ZONE
FIGURE I. VOC TRANSPORT FROM THE WATER TABLE TO THE ATMOSPHERE
Tick s Laws describe the process of diffusion (10) these equations have been modified
exiensivelv to describe diffusive flow through porous media (II 12) Pick s Law foi VOC
diffusion in air through a unit area as coirected for available an-filled pore space i«
wlieie is the gas-phase diffusion coefficient P is the an-filled poiosits L is ihe
gas-phase VOC concentration at the capiliarv fringe C is the concentration at the soil
surface and Z is the depth from the suiface to the capiliarv fiinge Millington and Qunk have
* o 3 °Oo O 0 HP ArC 1 O •• „ 0 6^0 • rOC
£°P.'J>tofo-^o °o(
>:§l\ b* VADOSE ZONE^r°0o°o-°5.f-0^
v , r»« °i-a °o •. ,0 Lv° ••o • • • ortVv 0 o
bf Q 0 »• 0 - 0 oOQ.d 0
fKroo-;;;
Vv CAPILLARY FR5NGE
KINETIC
> CONTROL
(DIFFUSION)
DYNAMIC EQUILIBRIUM
(DISSOLVED ^ GAS)
mass How = -D P (C -C ) Z
a s a
I I)
3
-------
derived an expression based on Fick s Law tor diffusion of gases through soils that is
modified for the torttiositv (non-linearirv) ot diffusion paths through porous media ( H) In
ihat desctiption ot diffusion in soils the gas-phase diffusion coefficient Da of the diffusing
compound is modified for tortuositv bv multiplying it bv P the air-filled porosirv of the soil
laised to the 4/3 power Equation 1 when modified for tortuositv (13) becomes
mass now = -D Pi/3 P (C -C ), Z (2)
a s a
for drv soil where P1/3 is the lerm incorporated to describe the tortuosirv of porous media.
Other modifications of Fick s Law to account for tortuositv have also been formulated (12)
Equation 2 predicts a linear concentration gradient with depth under homogeneous
conditions A strong linear depth dependence of the VOC concentration in soil gas has been
observed at the site of this study (I) and in other field (3) and laboratory (9. 14) studies
Mon-kinetic factors such as equilibrium sorption on soil are not relevant to discussion ol the
kinencallv controlled VOC concentrations in the vadose zone that exist when soil-gas surveving
is applicable
The VOC concentration gradient berween the atmosphere and the soil gas in contact with the
capillary Irmge is the driving torce tor the diffustonal vertical transport of VOCs The
concentration ot VOC in the gas phase in contact with the capillarv fringe is dictated bv a
dvnanuc equilibrium between the dissolved and the gas-phase VOC The phvsical-chemical
expression which describes this equilibrium is Henrv s Law f 15)
The Henrv s Law constant of a VOC can be obtained bv dividing the vapor piessuie ot the
pure VOC bv us solubilirv in distilled water Table I lists the Henrv s Law constants for
several VOCs ol interest however these values should be used onlv for guidance as to the
relative suitabilttv of the VOC for detection bv soil-gas measurement This is because the
ambient pressure organic content of the soil temperature and ionic strength of the water
encountered in the field can modtfv the situation from the conditions used to determine ihe
labulated \alties
So VOC concentrations in soil gases in the vadose zone are determined hv equilibnum
considerations at the capillarv tringe and kinetic factors in the vadose zone Under
circumstances controlled bv equilibrium (actors kinetic considerations such as vaporizanon
'tites are not ol concern In situations where the rate ol processes conirol the situation
i kinetic control) equilibrium considerations are important onlv to indicate the direction ot
change it is important to not confuse these concepts in order to effectively plan and
interpret soil-gas survevs
P;issi\e and Grab Sampling Techniques in Soil-gas Sampling
Soil-gas measurement was originallv developed in the 1920 s for petroleum exploration (16)
Recentlv ihe technology has been used for delineation of the extent of contamination bv VOCs
(4 17) Soil gases are tvpicallv sampled bv one of two approaches grab sampling and pa"i\e
sampling
In grab sampling a hollow pipe is dinen into the ground to a piescnbed depth .mil soil
gases are withdrawn through it Samples aie then analvzed bv gas chromatogiaphv at oi neai ilie
sampling location This method offers the benefit of immediate lesults a leaiuie wliu.lt i^
attractive because it allows the sampling plan to be modified as ilie <=unev piogiesses In
addition preliminary measurements can he peifoimed to allow ime^tigatoic to optimize ien.un
suivev pniameters such as sampling depth -\n advantage ot this appio.uh i* «lint in addition to
oil-sue screening of soil gas analvsis ot soil and giound-uatei samples can be peitoimed The
4
-------
drawback of rhis approach is ihat it is expensive in that it requires sophisticated sampling
and analytical equipment n specialist to operate and maintain it and associated support
systems u-# generators compressed gases)
In passive sampling an absorbent sampler is buried at a shallow depth and allowed to
collect VOCs from the soil atmosphere After a set time (8 hours to several weeks), the samplei
is lemeved sealed immediatelv and transported to a laboratory tor analvsis Analvsis
i esu I is indicate the identirv and concentration of VOCs collected bv the sampler The mam
advantage of passive sampling lies 111 the simplicitv of field operations field support
(supplies personnel and equipment* is much less costlv than for grab sampling, and equipment
problems are virtually non-existent The disadvantages associated with passive sampling aie
that results are not available for davs to weeks and that deep sampling is difficult without
more elaborate equipment
Soil-gas surveving technologies are designed as reconnaissance tools to be used in
conjunction with ground-water qualitv data not to replace it Confirmation of the predictions
of contamination from soil-gas results based on analvsis of soil cores and ground-water
samples will alwavs be necessarv However a preliminary soil-gas survev can help sue
investigators plan a more cost-effective effort
Based on the advantages and disadvantages of the two approaches to soil-gas survev mg Mie
investigators can choose one or the other dependrng on the priorities of the particular task at
hand In general passive sampling will be preferable when there is already some knowledge
ol site conditions so thai sampling locations can be chosen ahead of time and for smaller
sui\e\5 (less than 40 points) Examples of situations where passive sampling would be
pteferable include evaluating the possibility of VOC migration onto or awav from a site
Llarif\mg of the situation berween monitoring wells or in a specific area and for periodic
U' v monthly) monitoring of vapor wells for detection of leakage from underground storage
tanks On the other hand grab sampling techniques are preferable where the situation is verv
poorl\ defined or the effort is exploratory and a high potential for changes in the initial
sampling plan exists Grab sampling is also useful where the availability of analytical
iquipnient is desirable for screening soil and ground-water samples For grab sampling to be
i.ost-etfectiv e the survev must be ot sufficient size to justify the mobilization and operation
losis associated with the use of on sue analytical equipment
\ private firm has developed a passive-sampling device for oil exploration and has also
used it for mapping ground-water VOC contamination (18 19) Allhough the sampler is said to
measure the integrated VOC flux" at sampling locations, no rigorous evaluation of it has been
performed One publication indicating the potential of the technique for application to
ground-water contamination showed that it measured high levels of toluene carbon
tetrachloride and dichloroeihvlene in the headspace above a well However analvsis of
ground-water samples Irom that well did not show detectable concentrations of these compounds
119) In addition the technique failed to detect chloroform although that compound was the
maior containment as measured in ihe gas chromatographv/mass spectrometry (GC/MS) analysis of
ground-water samples in that studv (19) In a studv of the technique performed hv EMM -L.is
Vegas above chloroform-contaminated giound water lesults showed no Lorielanoit with
ground-waier concentrations verv poor precision among results fiom samplers separated hv I 2
meters (3-6 leet) and high levels of blank contamination ( I) Results horn a grab-sampling
method used simultaneously showed good correlation with giound-watei concentianon* and a lueh
level of precision 11)
Because ol the failuie ot this passive-sampling technique wheie a grab-sampling method li.id
succeeded and due to the attractive features of passive sampling we decided to test passive
soil-gas sampling using a diffusional sampler that lias been validated tor workplace atmospheies
5
-------
Theory of Diffusional Sampling
Sampling for airborne contaminants was originallv developed using active-sampling
techniques wherebv sample air is pumped through a device containing a sorbent-for VOCs.
usuallv a glass tube tilled with activated charcoal The mass of the vapor collected is a
direct function of the sampling rate of the pump the ambient concentration and the sampling
time For passive sampling vapors are also collected on a sorbent but transport is bv gaseous
diffusion rather than bv pumping The mass uptake ot vapor is controlled bv the effective
sampling rate of the sampler which is a tunction of the dimensions of the sampler and ilie
diffusion coefficients of each compound collected ambient concentrations and sampler exposure
time OSHA has established general precision and accuracy criteria for air sampling and
analvsis which must be met for results to be considered valid measurements ot workplace
atmospheres (29 CFR 1910) While these criteria were originallv developed for "active"
sampling thev also applv to passive sampling There are several commerciallv available
samplers marketed for industrial hvgiene use that meet the OSHA criteria
Figure 2 shows a schematic of a rvpical diffusional sampler The sorbent used for higher
concentrations of VOCs is charcoal Tenax has been used in a passive sampler developed bv EPA
for indoor air measurements where VOC concentrations are much lower than those ot concern in
soil-gas or workplace atmospheres (20 21) The draft shield shown in Figure 2 is used to ensure
that the sole mechanism ot VOC transport from the atmosphere to the sorbent is gas-phase
diffusion With this assumption Fick s Law can be applied to calculate the rate of VOC
accumulation on the sorbent
Rate = D (A/L) (C -C ) (1)
a a o
where Da is the diffusion coefficient of the VOC in air (cm2/sec) A is the sorbent area (Lin' i
L is the^ diffusion path length (cm) Ca is the VOC concentration outside the draft shield
(ng/cm ) and Cg is the VOC concentration at the sorbent surface (ng/cm 1 For a charcoal
sorbent Cq is assumed to be zero for most VOCs and equation I becomes
Rate = D (A/L) (C ) (2)
a a
thai the rate ot VOC accumulation (ng/sec) is equal to the ambient VOC concentration
multiplied bv the term Da(A/L) The term Da(A/L) can be expressed in units ol cm sec and can
be considered to be the volume (in cm ) scrubbed free of VOCs in one second
Although the variables A (sorbent area) and L (diffusion path length) are not influenced b\
the variables that affect gases (pressure and temperature) the sensttmrv of the diffusion
coefficient of the VOC to these parameters must be considered If it is relativelv
insensnne concentrations of VOCs at sampling locations can be calculated directlv from the
amount accumulated and the term D (A/L)
Diffusion coefficients determined empirically are available for manv volatile organic
compounds or can be calculated to within 5°?- (22) bv
D = 4 3 k tO"3 (M )1/2 educed of
air and tfie VOC (M = (M + M )/(M \| ) M = moleculai weielu ot the VOC M = moleculai
r a b a b a - = 1
weight of air (28 97)) P = atmospheric picsmiie tatmi \ = niolai volume ol the \ 0( (cm mob
and V = molar volume of nir (20 I cm" mol)
b
6
-------
DRAFT SHIELD
CHARCOAL
COLLECTION SURFACE
DIFFUSION PATH
~ CURE 2. CROSS-SECTION OF A TYPICAL DIFFUSIONAL SAMPLER
evaluation of each of the terms on ihe riglit-hand side ol Equation 3 we can determine ihe
effect ol changes in pressure or temperature on the diffusion coefficient and thus the term
IJ(ArL) in Equation 2 Equation 3 shows that the diffusion coelficient is inverselv propornonal
io ihe barometric pressure and directly proportional to ihe absolute temperature raised to (he
V2 power However as the pressure increases the decrease in the diffusion coefficient is
balanced bv an equal increase in the concentration so that the sampling rate is independent of
pressure (23) Similarly the increase in due to an increase in temperature is parttnllv
offset bv a concommitani decrease in concentration so that the net effect on the raie of VQC
accumulation Equation 2 is small (24) Even without these partial or complete obvintions of the
clfecis of changes in temperature and pressure in Ds the result on D of a change in temperatuie
Ironi 5'C to J5°C predicted b\ Equation 3 would be approximately l& percem However in outdoor
o aluations under a range of conditions (-15-35°C and 20-80% RH) no significant effects of T and
P on sampling rate occured 120)
Based on Fick s Law ihe average time required for a molecule to diffuse through the static
.in (aver of length L m the sampler would be
t = L2.2D (seconds) (4)
a
For example for a VOC 'Aiih D =0 12 cmJ/sec and a diffusion path length L of 0 65 cm the
average time for the VOC to diffuse from the draft shield to the sorbent is I 76 seconds
Therefore dtffusional sampling occurs rapidlv even though a static air column is used
Sampling bv charcoal lor other sorbent) difftmonal ampler* 15 limited bv rhe capacuv of
the sampler for various VOCs This capacirv must be determined on a compound-specific basi=
In addition studies of the use of these samplers in the atmosphere have shown thai wind
velocirv and angle of incidence have a minimal elfect on (lie sampling late Wmd is not a
factor in soif-gas sampling but the verv fact that the atmosphere being sampled i* statu, could
create a situation where the VOC mass transfer onto the sorbent i« limited bv (slow > VOC
transport from the ^-adose zone to the draft shield ot the momtoi If cik.Ii a situation doe";
occur (lie VOC transport is non-steadv-staie and a consideiahb moie detailed inatlieni.mi.il
description than above is necessary (26) Tins descnption uonld icqmre painnietei5 wlnt.li aie
not readilv available for most sites
7
-------
Ah hough VOC sorption on charcoal is considered irreversible for most compounds, there is
evidence for loss of some weaklv sorhed VOCs during storage between sampling and analvsis (27)
The performance of charcoal-sampling devices sold for industrial-hvgiene applications lias been
studied for various compounds Although recoveries obtained tor certain compounds can varv from
100 percent depending on monitor design thev are reproducible enough for application ol a
correction factor (27) Such a correction is a standard practice 128)
Analvsis of these monitors is performed according to NIOSH method P & CAM 127 this method
involves solvent-desorpnon and analvsis of the resulting solution bv gas chromatography (29)
The method has been evaluated for manv VOCs of interest for soil-gas surveying and the precision
and accuracv are documented (29) The method gives qualitv control procedures and outlines
calculation methods for the analvsis
Evaluation of the application of passive sampling to VOCs in soil gases using technology
that is well-described theoretically and has well-documented performance in sampling VOCs from
nir is a wav to help characterize the soil atmosphere If the technique which works well in air
does not work or works less well in soil gases it will indicate that different factors must be
considered in sampling soil gases than in the sampling atmosphere If the technique does work
¦veil when applied to soil-gas sampling it presents another approach to soil-gas sur\eving
8
-------
CONCLUSIONS AND RECOMMENDATIONS
Results from use of the Lockheed Passive Sampling Svstem (LPSS) for soil-gas surveying
correlated well with ground-water chlorotorm concentrations ot 10 to 1000 ^g per liter above
ground water that occurs at 3 to 4 5 meters (9 to 14 feet) below the ground surface The LPSS
lesults Irom measurements at a depth of 0 3 meters (I foot) correlated verv strongly with
lesults from samples from a depth of 1 3 meters (4 feet) at the same site in an earlier
grab-sample survev (r* =0 95) The precision of measurements using the LPSS technique is
superior to that obtained in an earlier grab-sample soil-gas survev at the sue Studies ot the
sampler efficiency indicate that at low concentrations of chloroform in the soil gas. sorption
efficiency mav be impaired more bv the effects of liumiditv than at higher concentrations An
empirical statistical calculation based on the observed precision of the measurements as a
function of concentration indicated that the distance between siatisticallv signtcantlv
different measurements at that site ranged from approximately 3 meters (10 feet) to 33 meters
1100 feet) with better spatial resolution at lower concentrations
Limned LPSS measurements above a benzene/chlorobenzene groundwater plume agreed with an
eailier stud*. that showed negligible benzene or chlorobenzene to be present in the soil gas
abo^e this highly contaminated ground water Specific compounds identified were consistent with
ihe occurrence ot aerobic biodegradation as postulated in the earlier studv
The LPSS technique offers superior precision low cost, and minimal requirements for
training for field personnel in comparison to grab-sampling methods Further studies
encompassing different hvdrogeologic conditions and other volatile organic compounds aie
lecommended All further studies of the LPSS technique should include for each studv sue
comparison of results to ground truth closelv spaced samplers to estimate overall precision,
and sampling at several depths In addition further studv of the situation in the vadose zone
above ihe benzene and chlorobenzene contaminant plume at the Pirtman sue is recommended Tests
ot microorganisms in the soil there tor acclimation to and metabolism of these compounds should
he neriormed In addition vadose-zone carbon dioxide and oxvgen concentrations should he
measured
9
-------
EXPERIMENTAL
SITE DESCRIPTION
Field tests were conducted at (he Pittman Lateral a site having known ground-water VOC
contamination The Pittman Lateral is a nght-of-way of a major water conduit serving Las
Vegas Nevada It is in undeveloped desert about I I miles (18 km) southeast of Las Vegas This
site has been used since 1982 bv EMSL-Las Vegas to test both surface geophvsical (21 and
soil-gas (I) survey methods
At the Pittman Lateral monitoring wells have been installed at 64-meter (200-foot)
intervals along a line perpendicular to the northward direction of ground-water flow On
several occasions over the past five years, ground-water samples have been collected from these
'veils and analvzed for inorganic and organic contaminants Analvsis bv gas chromatography/mass
spectrometrv of ground-water samples indicates that rwo distinct plumes ot organic and inorganic
contaminants exist at ihe Pittman Lateral One plume has trichloromethane (chlorolorm) as its
primarv VOC component while the other plume has benzene and chlorobenzene as ihe major VOC
contaminants Figure 3 shows the subsurface at the Pittman Lateral and the location of
monitoring wells there Table 2 lists the VOC concentrations measured in ground-water samples
trom monitoring wells at the site
The livdrogeoiogv at the Pittman Lateral is relativelv simple A clav aquiclude undeihes
.in imconfined aquifer composed of moderatelv well-sorted unconsolidated gravel alluvial
deposits with discontinuous caliche cement horizons which varv in thickness from 5 to 25 meters
(15 to 80 feel) (30) The ground surface water table and aquiclude all dip gentlv to the
north where the aquifer davlights at the Las Vegas Wash Along the Pittman Lateral unconfined
ground water occurs at 1 5 to 4 3 meters (5 to 14 feet) below ground surface
BENZENE/CHL0R08ENZENE
t \ |
CHLOROFORM 3
1—505
1660 —i
500
1640
~_J—-r- - r _ T
1620
r
1600 —
>
1580 —
CLAY
1560
610
620
625
635
630
650
645
640
STATIONS
EAST
!fti -ji/Fit ' SCALt IN fEET
USI *tll q 5Q0
1 i SCALE IN MET£RS
o :oo
FIGURE 3. SUBSURFACE AT THE PITTMAN LATERAL
10
-------
1 Mil i; 2 ( ONUJN I R V I'lONS ()l ( III OROI ORM IJkN/I.NI. \NI) ( III.OROIW.NZI.NI. (,/y/l ) IN
GROUND-WAT!- R SAMI'I I S COIIK TI.D I ROM WT.I I S \I()N(. I III I'll I'M \N I \TKR M '
(lilorolorm Concxnirjiioiis )
Will Niiinlicr
1 j.i ii
nl 7
nl 'i
t»: i
'i-3 n^S
r.J'J
Ml
'>33
3'M
.1
< 10
500 430
ihi
1 1
< 10
.1
1 1 /X 1
>
*
>
^ XSO
\
4 .N5
s
s
570 1000
1 ;s
s
s
\
X'X5
¦¦
*>41 73:
s
>
tV.fl/.t:nc
ConciMlrjnon* tug/1 )
Well Nuinlxjr
D.iii
o33
n35
nil
n W
1,11 f)4 3
n) 5
ft 17
fi l*i
o51
f>3 3
3/.S3
< 10
340
3M)0
1700
3:00 3400
JI00
3500
1300
37
< 10
1 I/X4
S
< 10
%
1 >
s
1,^5
S
< 10
ym
11
0
s
n
.VX5
N
•»
N
:5»x
u:o
N
N
s
< 10
Clilorolxn/j-nc Conciniranons Cog/I.)
Well Number
1) in
fi33
r>35
(>37
nl')
Ml r» 1 i
h \ ^
h 17
fl I'i
n5l
f>5 1
J/.SJ
< 10
1:0
4000
S 100
} 100 2400
700
Jr»00
2-400
SOO
:i
1 1 'X4
< 10
>
¦> s
s
s
>
s
i/xj
< 10
1 100
7100
n
>
1)
11
h/S5
>>
>
N
At>:\
50o0
N
s
fiO
I b.iinpks In i: P\ Muluul i\2 1 (<»(. Nth) ikltiimn limn - S,,u»'|
il = Not ikin ml
n - Nul sjinpkd
ii ~ Not .iikiK/ciI (hoi.liny liuu i_Mt.iili.il)
I 1
-------
The soil in the vadose zone at the site is a Caliza verv-gravellv sandv loam from a
sandv-skeletal mixed, thermic familv of the tvpic Calciorihidf31) The clav content is 2 to 8
percent decreasing with depth and the permeability is inoderaielv high, at 5 to 15 cm per hour at
depths of 0 to 40 cm and 15 to 50 cm per hour at depths of 40 to 150 cm (29) The average
tainfall in the area is 10 cin with occasional periods of flooding during rainstorms The
shrink-swell potential of the soil is low (31) so thai the permeabilirv remains fairiv constant
when the soil is wetted The soil is characterized bv a pH ot 7 9 to 8 4 and the salmirv is
low with the electrical conductivity below 2 //mhos per cm(3l) The organic content of the soil
is euremelv low
APPARATUS AND PROCEDURES
Apparatus
The Lockheed Passive Sampling Svstem (LPSS) is comprised of an activated carbon organic
wipor monitor (No 3500. 3M St Paul Minnesota) suspended in a sampling manifold made Ironi a
one-quari can Figure 4 shows a diagram of the LPSS assemblv
FIGURE 4. LOCKHEED PASSIN E SAMPLING S\STEiM ILPSS) .SAMPLLK
12
-------
Field Procedures
Samplers were placed in holes 0 3 meters (I foot) deep, the holes were backfilled, and the
•samplers were retrieved and sealed after a 2-^eek exposure period Samplers were placed 6 4
meters (20 feet) to the north south east and west of 6 adjacent groundwater monitoring wells
separated bv 200 feet along a line perpendicular to the northward direction of ground-water
flow At five wells the south sampler served as the northwest corner of four samplers in a
square pattern with I-meter (3-foot) sides while at the sixth (631). one additional sampler was
placed I-meter (3 feet) to the south of the southern sampler A line of 10 samplers at 1-meter
13-foot) intervals was placed parallel to the direction of ground water flow beginning at a
point halfwav between well 623 and 625 as a check on short-range precision along the direction
of ground-water flow A manifold with two monitors in n was placed I-meter to the south of the
sampler to the east of five wells as a test of the sampling efficiency of the svstem or
vertical flux At the sixth well (631) additional samplers were placed 9 6 and 12 8 meters (30
and 40 feetl to ihe east of the well Figure 5 shows the sampling locations around each well
Analysts
Soil-gas samplers were solvent-desorbed and analvzed bv gas chromatography according to the
manufacturer s instructions (32) Pesticide-grade hexane (Burdick & Jackson) was used as a
solvent for chloroform determinations and carbon disulfide was used for the tour samplers from
above the henzene/chlorobenzene plume For chloroform analvsis was performed with a
Hewlen-Packard 5710 gas chromatograph using an electron capture detector for benzene and
Uilorobenzene a flame ionization detector was used
Calculations
From gas chromatographic analvsis of a I -/vL injection of the solution containing the
dcsorhed VOCs the concentration of each VOC can be calculated bv multiplication of the VOC peak
height times the slope (i»g/^L per height) of ihe calibration curve Because 1500 of solvent
was used to desorb the VOCs the VOC concentration calculated was then multiplied bv 1500 to
obtain ihe nanograms of the VOC collected From this value the mean results of analvsis of
Held blanks was subtracted When the resulting value is divided bv the sampling rate of the
organic \apor monitor (in ng/ppb \ hr) the average concentration (in ppb) over the exposuie
period results fable 3 lists ihe sampling rate reco^erv and capacirv of the 3M 3500 Oiganic
\ apor Monitor lor benzene chlorobenzene and chloroform (32)
TABLE 3. PERFORMANCE OF THE 3M 3500 ORGANIC VAPOR MONITOR
= = = = Sampling Rate = = = =
Compound (cm /min) (ng/ppb¦ lir) Recovery(%) Capacity (mg)
Benzene 35 5 ± 0 6* 6 78 102 13
Chlorobenzene 29 3 + 0 6* 8 12 100 >25
Chloroform 33 5 9 7S l)S 7
Inilniks lO'i'illetHC HHiiVtil
Smmir RclurrKt* S2.
13
-------
N
x
i
W x
1Q < •
.0 ft
10 ft
20 ft
X c
E 1
Y ¦
= 30
' E-10 *"
T '
A xal x
3 ft !
T xS2 XS3
^ MONITORING WELL
x SAMPLING LOCATIONS
* S3. S4 NOT PRESENT AT WELL 631
** E30, E40 ONLY PRESENT AT WELL 631
FIGURE 5. SAMPLING LOCATIONS AT THE STUDY SITE
14
-------
Qualitv Control
Published recoverv coefficients were used for chloroform desorption Field blanks were
processed and laborator\ blanks were also nnalvzed The detection limit was determined as three
nmes the standard deviation of 12 non-consecutive blanks Calibration standards were prepaied
bv serial dilution of stock solutions prepared from neat chloroform and carbon tetrachloride
(Alltech Associates Deerfield Illinois) Calibration was performed daily Multiple I-^L
aliquots ot desorbed solutions from each sampler were analvzed
As a check on contributions 10 error and \ariabilitv of results from souices other than soil
gases monitors were sealed in manifolds tor tour weeks and analvzed for benzene, chlorobenzene.
and chloroform Multiple 1-uL aliquots of each sample were analvzed to estimate the analvsis
<-anance The short-range precision of results was assessed bv closelv spaced samplers
EVALUATIONS
E\aluation of the Performance of the LPSS Above the Chloroform Ground-VValer Plume
correlation with ground-water data—
The mean results from the locanons around each well were compared to the corresponding
ground-water chloroform concentrations 10 evaluate the correlation between those two data sets
Giound-waier analytical results from another EMSL-LV siudv 11) were used for this comparison
(orrelatton with grab sampling data—
The mean LPSS results from locations around each well were compared to grab sampling data
Irom ihe earlier studv
short range variability—
To assess ihe variabilis ot results obtained from samplers separated bv short distances
multiple samplers were buried at ihe sampling location to ihe south of each well At five of
ihe wells <621 023 625 t>27 t>29) ihe location 6 4 meters 120 feet) to ihe south was ihe
northeast corner ol a I-meter i3-loot> square pattern of samplers At the other well (631) a
sampler was placed 7 4 meters (23 feet) south of the well in addition to the one 6 4 meters (20
teen south
Because earlier grab-sample measurements had shown no \ariabilitv among three locations at
I-meter <3-Foot) intervals along a line halfwav between two wells (623 625) and parallel io
ground-water flow ten samplers were buried there ai I-meter (3-foot) intervals along the same
line Results from ihese samplers were used to evaluate the short-range variabiiitv of results
along the direction of contaminant flow
In a studv to evaluate ihe shori-range \ariabiliiv of results perpendicular to the
ground-waier flow direction additional samplers were placed 9 6 meters (10 feet) and 12 8
meters (40 feet) to ihe east (towards higher concentrations) of well 611 Statistical
evaluation of these results will help indicate ihe spatial resoltnion the minimum distance
between samplers to obtain siaticallv stgnificantlv differeni resuls ol ihe technique ai ihis
site
IS
-------
Sampler Efficiency
To determine whether the samplers trapped all of ihe VOCs (hat diffuse upward into the
manifold from the soil 01 actuallv sample the VOC concentration, an additional studv was carried
out At all 6 wells manifolds containing two organic vapor monitors were deploved I metei (3
feet) to the south of the sampler b 4 meters 120 feet) to the east of each well The sums ol
the amounts of chloroform collected bv the t*o monitors in one manifold were compared wnh the
amount collected bv a single-monitor manifold nearbv
Evaluation of tlie Performance of LPSS the Above Benzene/Chlorobenzene Plume
Four samplers were placed in a I-meter square (3-foot) panern at 0 3 meters (I foot) depth
m a location b 4 meters (20 feet) south (upgradient) of the well with the highest
concentrations of benzene and chlorobenzene Because neither compound was detected in
soil gases in the earlier studv these samplers were left in place for three months This
extreme exposure time was provided so that detection of very low concentrations of either
compound would be assured
16
-------
RESULTS AND DISCUSSION
PERFORMANCE OF THE LPSS ABOVE THE CHLOROFORM GROUND-WATER PLUME
Correlation With Ground-Water Data
Results of chloroform analyses of all of the organic vapor monitors from samplers around
wells are listed in Table 4 Table 5 lists the ground-water chloroform concentrations from each
well The mean results from the four symmetrically placed sampling locations surrounding each
well and the soil-gas concentration from samples around that well as measured by a grab-sample
and on-site analysis technique in an earlier study The two soil-gas data sets correlate very
strongly (r =0 95) The soil-gas measurements from this study correlate above the 95 percent
confidence level with the ground-water data This indicates that soil-gas measurement by the
method used in this study can provide a good picture of VOC contamination in underlying ground
water
Correlation With Grab-Sampling Data
Table 6 lists the sampling locations from this study the results obtained and the
chloroform concentrations at a I 3-meter (4-foot) depth at corresponding locations in an earlier
EMSL-Las Vegas study (I) Linear regression of these two data sets indicates that there is a
cot relation of greater than 99 percent significance between them. Figure 6 is a scatter plot of
the results from this study as a function of the results from soil-gas measurements made at a
depth of I 3 meters (4 feet) in an earlier study, using the data in Table 6 It can be seen
that an excellent correlation exists
Figure 7 shows the soil-gas data from this study, and the soil-gas and ground-water
chloroform concentrations from an earlier study (I) plotted as a function of distance across the
plume It can be seen that the soil-gas measurements clearly agree with the location of the
plume as indicated by ground-water analyses
Short-Range Variability
In an effort to assess the overall precision (analytical sampling, and geologic
variability) of the technique the variance of the results fiom the groups of closely spaced
samplers 6 4 meters (20 feet) south of the six wells was calculated Table 7 lists the
individual results and the mean and standard deviation of each group with detectable
concentrations
The standard deviations comprise a linear function of the measured concentration A
composite lunction of the standard deviation of the soil-gas data as a function of concentration
and the concentration as a function of distance has been obtained (33) Using this relation-
ship the minimum distance across the plume from each sampling location where a statistically
significantly different concentration would be observed can be calculated (33) Table 8 lists
the observed concentrations for the southern sampling locations in this study In addition the
table lists the distances east and west from the sampling locations where linear interpolation
predicts that a siaustically significantly different concentration will be measuied This
distance has been called the spatial resolution of soil-gas measurement (33) It is of interest
that the values at wells 623 and 629 are quite different although the observed concentrations
are the same (2 6 ppbv) This could be due to differences in geologic features beneath the two
locations
17
-------
TABLE 4. RESULTS OF ANALYSES OF ORGANIC VAPOR MONITORS FOR CHLOROFORM
Well
S.nnplpr
Mass
Collecietl
IIICI
Correcieti
Mass b
Coll«ml
pph
*21
N
0 2!
0 03
MDh
SI
0 11
NO
s:
0 15
ND
53
0 U
...
ND
S4
0 is
- -
ND
E|a
0 15
ND
E-b
0 10
ND
e:
0 OK
ND
w
0 20
0 02
ND
N
» 73
9 5X
_ (
SI
12 51
12 33
< 4
s:
X 4f1
X 2X
2 3
S3
10 44
10 26
: 8
SJ
7 2X
7 10
1 9
Ela
6 05
5 H7
1 6
"b
3 93
3 75
1 0
E2
: 3n
2 23
0 6
\V
:i )4
20 96
5 7
1
56 92
56 75
15 5
2
52 9X
52 HI
14 4
j
44 S6
44 3X
12 1
64 93
h} 76
1 " 4
*
M 19
<1 01
13 0
<3 "
M 54
14 t>
4 5 S4
45 36
12 4
\
47 X4
47 66
13 0
0
hO 31
nO 13
16 4
10
4X 90
4X 72
13 3
N
"2 4S
- -v *> -
' « - !
20 3
SI
6X 20
*X 02
IX 6
s:
43 %
43 78
1:0
S3
15 4X
55 30
15 1
S4
31
^2 13
14 2
Ela
86 IX
X6 00
23 5
E:b
70 17
69 99
19 1
E2
75 63
75 45
20 6
w
43 24
43 06
IX 1
^27
N
32 12
31 94
0 0
SI
2X 13
27 95
7 6
S2
26 j:
26 24
- ^
S3
23 XI
23 63
6 5
S-i
19 9 6
19 7X
5 1
Cl»
42 09
4 1 91
1 1 4
E:b
35 51
35 33
T 6
El
35 37
3< 19
0 6
w
15 11
14 93
4 1
r*19
M
10 57
10 39
2 9
St
11 00
10 83
3 0
s:
9 31
0 13
2 5
S3
K 30
X 12
"> >
S4
10 79
10 61
2 9
E1a
5 54
5 36
1 5
E2b
J 4|
4 23
1 :
E2°
4 03
3 85
1 0
w
4 27
4 09
1 2
N
0 20
0 02
ND
SI
0 105
0
ND
s:
0 1 11
0
ND
E30
0 14
0
ND
e-»o
0 26
0 OX
2 v 10 2
Ela
0 10
0
ND
E-b
0 06
0
ND
E2
0 06
0
ND
W
0 12
0
NO
bCoi irciton tot
Dilution lullil
hi,ink = 0 IX nc
= 0 02 |)|>l> i3 \
* 1 <1 III 1 rill 1 vk\ I.IIIOII
i>l hi mk\]
18
-------
TABLE 5. COMPARISON OF SOIL-GAS PASSIVE AND GRAB SAMPLING RESULTS
WITH GROUND-WATER ANALYSES
Well
Ground-Water
Chloroform
Concentration (,ug/L)
Mean Soil-Gas Chloroform Conrcnuaiions ippbv)
Passive Sampling at Giab Sampling at
0 3 Meters (I ftT 13 meiers (4 ft)
621
623
625
627
629
631
28
555
866
175
I I
ND
VD
1 2
19 2
7 8
2 0
VD
I I
40
m
68
23
5
TABLE 6. COMPARISON OF RESULTS FROM PASSIVE AND GRAB SAMPLING METHODS
Mean Soil-Gas Chloiofonn Conccnuanons Ippbvl
Sampling Passive Sampling at Grab Sampling at
Well Location 0 3 Meters (I ft) 13 meters (4 ft)
621
N
E
S
W
ND
NDd
NDa
b
b
li
I I 0
b23
N
E
S
w
2 7
1 6
2 6C
5 1
12
6
27
I 15
624"
14 3
ISO
625
V
E
S
w
20 3
23 5
I 5 0C
IS I
126
U6
51 I
266
627
N
E
S
w
9 0
11
6 7'
4 I
73
125
46
28
629
N
E
S
W
2 9
1 5
2 6C
I 2
25
27
"II
10
531
N
E
S
w
ND
^Dd
MDa
NDa
ND = Nop tlui Mthii ,tL [ma til hi,mk kmiIk
Ni»i vtinplrd
Mean nl Kun samples in vm IK fO ^ t»»»I **> 1 ^
. Mfilft ol lliut pi Mill \ H |MI il1t.ll ^ I Itli l< 1 t nil iloilL I lit H 111 \OHllt 11 III
Mtiin ol k n i>oum m p.n iHtil h* I »ntHi i.uti .» iU-soutlt Imr
19
-------
FIGURE 6. SCATTER PLOT OF PASSIVE-SAMPLING VERSUS GRAB-SAMPLING RESULTS
i
20
SAMPLING nESJL
20
-------
FIGURE 7. GRAB-SAMPLE AND LPSS PASSIVE-SAMPLING SOIL-GAS CHLOROFORM
CONCENTRATIONS AND GROUND-WATER CHLOROFORM CONCENTRATIONS
ppb
Chloroform
370-
J 70-
1 70-
a
a
a <
TO-
10-
1
-NOM«
mi 828 air sis
WELL NUMBER
-1-
Grab Sampling Soil-Gas Chloroform Results
ppb S
Chloroform
i"
629 827 625
WELL NUMBER
823 $21
Passive Sampling Soil-Gas Chloroform Results
ppb
Chloroform
MT 823
WELL NUMBER
1
Ground-Water Chloroform Concentrations
21
-------
TABLE 7. RESULTS OF CLOSELY SPACED PASSIVE SOIL-GAS SAMPLERS
Chloroform Mean
Concenh .ihon Chloroform Standard
Well Sampler (ppbv) Concentration (ppbv) Deviation
623 S-l 3 37
S-2
2 26
S-3
2 80
2 6 0 6
S-4
1 94
625 S-l
18 6
S-2
12 0
S-3
15 1
15 0 2 7
S-4
14 2
627 S-l
7 6
S-2
7 2
S-3
6 5
6 7 10
S-4
5 4
629 S-l
3 0
S-2
2 5
S-3
2 2
2 6 0 4
S-4
2 9
624a 1
15 5
2
14 4
3
12 1
4
17 4
5
13 9
14 3 17
6
14 6
7
12 4
8
13 0
9
16 4
10
13 3
a Location 624 is a north-south line halfway between wells 623 and 625
TABLE 8. SPATIAL RESOLUTION OF SOIL-GAS MEASUREMENTS BY PASSIVE SAMPLING
Mean Chloroform
Concentration
Spatial Resolution
Well
(Standard Deviation)
(Feet)
631
NDa
--
629
2 6 (0 4)
43
627
6 7(1 0)
53
625
15 0 (2 7)
67
623
2 6 (0 6)
9
621
NDa
--
aND = Not detected after subtraction of blank results
22
-------
Sampler Efficiency
The efficiencv of the sampler/manifold assembly was assessed hv comparison of the results
of analvsis of two monitors in the same manifold to the results of analvsis of a single monitor
in a manifold located I meter (J feet) awav The hvpothesis for this comparison was that if the
sum of ihe amount collected bv two monitors is close to twice the amount collected bv a single
monitor in a manifold the technique measures concentration and the sampling rate is not limned
In ihe iaie of diffusion of chloroform from the soil into the manifold On the other hand, if
the sum of the amount of chloroform collected bv the two monitors in a can is essentiallv equal
to the amount collected hv the single-monitor assemblv the sampling rate is determined bv the
iate of diffusion of chloroform into the manifold
The 'efficiencv" value listed in Table 9 is the amount of chloroform collected bv the
single-monitor sampler divided bv the sum ot the amounts collected bv each vapor monitor in the
dual-monttor sampler When the accumulation rate is limited bv the rate of supply of chloroform
liom the vadose zone into the sampler, that value will approach unitv When the rate of
accumulation of chloroform was determined bv the concentration in the manifold, that value will
approach 0 50
TABLE 9. SAMPLER EFFICIENCY STUDY RESULTS
Well Sampler Chloroform Concentrations Efficiencv
623
E-l
1 60
0 98
E-2a
1 02
E-2b
0 61
b25
E-l
23 5
0 59
E-2a
19 1
C-2b
20 6
027
E-l
1 1 4
0 59
E-2a
9 b
E-2b
9 6
b29
E-l
1 5
0 68
E-2a
1 2
E-2b
1 0
It can be noted that the efficiencv listed in Table 9 decreases with decreasing
Loncentrntion This calls into question the applicability of Equation 2 in this situation For
Equation 2 to be applicable the sampling rate of the monitor must not affect C the gas-phase
concentration being sampled This requires n supplv ol VOC more rapid than ihe monitor sampling
rate In the workplace atmosphere where these moniiors are used routinelv tins supplv late i*
ensured bv air currents In soil gases where bulk air movement is verv «low tins inte ol
supplv is equal to the vertical VOC flux In this siudv the vertical flux of clilorotoi 111 is
expected to be directlv proportional to the groundwater chloroform concentration according to
current theorv (4 9) so that the low groundwater chloroform concentrations are expected to
ptoduce comniensuratelv lower vertical (luxes ot clilorotoi m However this vioiild not n I fee t the
applicability of Equation 2 (\ ide nifj/a) because (lie sampling mie of ihe inomioi i« tin c*. il\
proportional to the gas-phase concentration Since the chloioloim concentration at ihe dimpling
23
-------
locations is proportional to the groundwater concentration, as is ihe vertical flux, the ratio
of the chloroform mass accumulation rate to the vertical flux of chloroform should be fairlv
constant given consistent subsurface characteristics (I!)
Because the soil-gas moisture conieni is expected to be relativelv constant among ilie sampling
locations (34) humiditv effects could onlv have caused our observations if the effects ot
humiditv on charcoal sorption elficiencv increase with decreasing chloroform concentration
Sucii a situation has been observed for activated carbon sorption of another chlorinated
compound tnchloroethene (35) The data in thai studv fit the Dubinin-Polvani equation for
sorption
PERFORMANCE OF THE LPSS ABOVE BENZENE/CHLOROBENZENE GROUND-WATER PLUME
Analvsts of all four LPSS samplers from above the benzene/chlorobenzene ground-water plume
did not show detectable amounts of either compound The minimum detectable amount of each
compound on the monitor is 10 milligrams on the basis of the monitor sampling rate and a
?-month exposure time, concentrations ol one part per trillion by volume ot benzene or
chlorobenzene would be measureable
This result confirms the results obtained with the grab-sample studv performed earlier at
this site (I) In that studv. manv compounds were observed in soil gases above the
henzene/chlorobenzene ground-waier plume bur benzene and chlorobenzene were not The absence of
these compounds in the soil-gas was attributed to extensive aerobic biodegradation in the
^hallow vadose zone at the site In that siudv none of the other compounds present in the soil
gases were identified Small amounts (below 10 milligrams) of diethvl ether isopropvl ether
ethylene dtbromide (EDB). and ethvl acrylate were observed The identities of the compounds are
consistent with intermediates in aerobic biodegradation of hvdrocarbons (ethers and esters)
Severn! other (unidentified) compounds were also delected
QUALITY CONTROL
Field nnd laboratory blank monitors showed no contamination However the desorption
solvent thexane) used lor chloroform contained chloroform equi\alent to approximately 0 018
ppbv this was subtracted from all results The detection limn was 0 02 ppbv tor chloroform
and approximately I part per trillion bv volume for benzene and chlorobenzene Monitors sealed
tn manifolds for four weeks showed no detectable benzene chlorobenzene or chloroform
24
-------
REFERENCES
1 Kerfoot HB LJ Barrows Soil-Gas Measurement for Detection of Subsurface O'vaun
Contamination U S EPA Environmental Monitoring Svsienis Laboratorv Las Vegas. 1986
2 Walther E G . D J LaBrecque D D Weber R B Evans J J van Ee. 'Studv of Subsurface
Contamination with Geophvsical Monitoring Methods at Henderson. Nevada.' In Management of
UncowiolL'd Haiaidous Waste Sites Hazardous Materials Control Research Institute. Silver
Spring Md. 1983 pp 28-36
3 Marnn D L G M Thompson 'Investigation of Volatile Contaminants in the Unsaturated Zone
above TCE-Polluted Groundwater ' EPA Project CR81101 8-01-0. U S EPA Ada. Oklahoma, 1984
4 Nadeau R J T S Stone G S Klinger "Sampling Soil Vapors to Detect Subsurface
Contamination A Technique and Case Studv " Proceedings of the Fifth National Confetence on
Management of Uncontrolled Hazardous Waste Sues Washington DC 1984
^ Mackav DM P V Roberts J A Cherrv "Transport of Organic Chemicals in Groundwater
Emuon Set 1'echnoi 19(5) 1985 pp 384-392
Plumh RH AM Pitchford 'Volatile Organic Scans Implications lor Ground Watei
Monitoring." In Pro( eedinys oj the Penoleum Hvdrocatbons and Orqaiuc Cheniuals nt Giound
Water ¦ Prevention Detection and Restoration Nov 13-15 1985 National Water Well
Association Worthington. Ohio pp 207-221
7 Marnn D L "Differences in the Remote Detection of Soil and Groundwater Contamination
Using Soil Gas Analvsis " In Pioceeainqs oj the 22nd Symposium on Engmea uty Geologv mid
Soils Engineering State of Idaho Boise 1986 422-431
5 State of California California Administrative Code Title 23 Chapter 3 Subchapter 16
Underground Tank Regulations Secnons 2641(c)(2) 2641(c)(3). 2646 California Staie
Water Resources Control Board Sacramento California 1985
(> Swallow J A P M Gscliwend Volatilization of Organic Compounds from Urcontined
Aquifers " In Proceedings of the T/nra National Svmposium on Aquifer Resioianon ana
Ground-Water Monitoring National Water Well Association Worthington. Ohio 1983 pp
327-333
10 Alonso M E J Finn Fundamental Universirv Phvstcs Volume II Fields and H'awj
Addison-Weslev. New "iork. 1967 pp 934-941
11 Cunningham RE RJJ Williams Diffusion m Gate* and Porous Media Plenum Mew'iork
1980
12 Dullien F A L Potous Media Fluid lumsijoii tin d I'ow Slimline Academic Pie« \ o< k
1979
13 Millington R J R P Oui'k hans Fiuadm Sot ^7 1200-1207 I9t>l
25
-------
14 Bruell. C J . G E Hoag. In Peiroiewn Hvdrocaibons and Organic Chemicals in Ground Water
National Water Well Association. Worthington Ohio. 1986. In Press
15 Moore WJ Physical Chant *trv ^id ed Prentice-Hall Eneiewood Cliff* New lersev pg
196
16 Horvitz L . 'Geochemical Exploration for Petroleum " Science 229(4716). 30 Aueust 1985
821-827
17 Lappala. E G GM Thompson Detection of Gioundwater Contamination bv Shallow Soil Gas
Sampling in the Vadose Zone. ' In Proceedings ol the Annual Symposium on Cliaiattei nation
and Monitoring of the Vadose Zone National Water Well Association Worthington Ohio
1983
18 Voorhees K J J C Hickev R W Klusman 'Integrative Gas Geochemical Technique for
Petroleum Exploration." American Chemical Socterv Meeting Seattle. Washington American
Chemical Socien/ Washington. D C 1983
19 Voorhees. K J J C Hickev R W Klusman "Analvsis of Groundwater bv a New Static Surface
Trapping/Mass Spectrometry Technique " Analytical Chemistry 56 2604-2607 1984
20 Lewis R G J D Mulik. R W Couiant G W Wooten C R McMillin. 'Thenriallv Desorbable
Passive Sampling Device for Volatile Organic Chemicals in Ambient Air " AnaiMual
Chemistry 1985 (57) 214-219
21 Coutant. R W R G Lewis J D Mulik "Passive Sampling Devices with Re\ersible
Adsorption.'' Ana/vucal Chemtsirv 1985 (57). 219-226
22 OiIIiland E R 'Diffusion Coefficients in Gaseous Svstems ' hid Eng Chem June 1934
23 Palmes ED A F Gunnison J Di Mattio C Tomczvk Personal Sampler for Nitrogen
Dioxide " Am hid Hvg Assoc J 37 570 1976
24 Linch A L Valuation ot Ambient An Quality b\ Pet sound Monuounti Volume I Gases and
Ufpots CRC Press Boca Raton Florida 1981 pg 57
25 Nelson GO Controlled Test Atmospheres-Principles and rcchntques Ann Arbor Science -\nn
Arbor Mich 1971
26 Alonso M J Finn Fundcunenial University Phvsics Volume II Fields and Waves
Addison-Weslev New York. 1967 pg 940
27 Linch. A L Evaluation of Ambient Air Quality bv Personnel Monitoring Volume I Gases and
Vapors CRC Press Boca Raton Florida 1981 pp 58-75
28 Ibid pg 66
29 Tavlor D G RE Kupel J M Brvant Doitmtcntanon ol the NIOSH Validation lev \ U S
Department of Health Education and Welfare Cincinnati Ohio 1977
30 Bingler E C Las Vtyas SE Folio - Grolo^n Map Nevada Bureau of Mines and Geologv
Universitv of Nevada Reno Nevada 1977
26
-------
31 Soil Conservation Service. Soil Survev of Las Vegas Vattev Area. Nevada, p 21. No 184.
US Department of Agriculture. Washington. 1985
J2 3M Pi qantc i 'apoi Monitor Compound Guide INI Corp St Paul Minn 1982
33 Kerfoot. H B M J Miah. "Spatial Resolution of Soil-Gas Measurement " Chemtromcs and
Intelligent LaboratoiM Systems [n Press
^4 Vielson. D R R D Jackson J W Carv D D Evans Eds Soil Water Soil Science Socierv
of America Madison Wisconsin 1984 p 18
35 Werner. M. D Am Ind Hvg Assoc 46(10) 585-590 1985
wps 05/02
27
------- |