2USGS v/EPA
science for a changing world
Guidance on the Use of Passive-
Vapor-Diffusion Samplers to Detect
Volatile Organic Compounds in
Ground-Water-Discharge Areas, and
Example Applications in New England
Water-Resources Investigations Report 02-4186
REPORTING LIMIT
U.S. Department of the Interior
U.S. Geological Survey
-------
U.S. Department of the Interior
U.S. Geological Survey
Guidance on the Use of Passive-
Vapor-Diffusion Samplers to Detect
Volatile Organic Compounds in
Ground-Water-Discharge Areas, and
Example Applications in New England
By PETER E. CHURCH, DON A. VROBLESKY, and FOREST P. LYFORD,
U.S. Geological Survey, and RICHARD E. WILLEY,
U.S. Environmental Protection Agency
Water-Resources Investigations Report 02-4186
In cooperation with the
U.S. ENVIRONMENTAL PROTECTION AGENCY
Monitoring and Measurement forthe 21^ Century Initiative
Northborough, Massachusetts
2002
-------
U.S. DEPARTMENT OF THE INTERIOR
GALE A. NORTON, Secretary
U.S. GEOLOGICAL SURVEY
Charles G. Groat, Director
For additional information write to:
Chief, Massachusetts-Rhode Island District
U.S. Geological Survey
10 Bearfoot Road
Northborough, MA 01532
Copies of this report can be purchased from:
U.S. Geological Survey
Branch of Information Services
Box 25286
Denver, CO 80225-0286
or visit our Web site at
http://ma.water.usgs.gov
-------
CONTENTS
Abstract 1
Introduction 1
Advantages and Limitations of Passive-Vapor-Diffusion Samplers 3
Advantages 3
Limitations 4
PART 1. Guidance on the Use of Passive-Vapor-Diffusion Samplers 5
By Don A. Vroblesky
Assembly of Samplers 5
Deployment of Samplers 9
Recovery of Samplers 13
Factors Affecting Deployment of Samplers and Data Interpretation 16
Quality Control and Assurance 17
PART 2. Example Applications in New England 18
Eastern Surplus Company Superfund Site, Meddybemps, Maine 20
By Forest P. Lyford and Edward M. Hathaway
Description of Study Area 20
Purpose and Design of Sampling 23
Results 23
McKin Company Superfund Site, Gray, Maine 25
By Forest P. Lyford, Terrence R. Connelly, and Laura E. Flight
Description of Study Area 25
Purpose and Design of Sampling 25
Results 28
Nutmeg Valley Road Superfund Site, Wolcott and Waterbury, Connecticut 28
By John R. Mullaney, Peter E. Church, and Carolyn J. Pina-Springer
Description of Study Area 28
Purpose and Design of Sampling 31
Results 31
Baird & McGuire Superfund Site, Holbrook, Massachusetts 34
By Jennifer G. Savoie and Melissa G. Taylor
Description of Study Area 34
Purpose and Design of Sampling 34
Results 34
Allen Harbor Landfill, Davisville Naval Construction Battalion Center Superfund Site, North
Kingstown, Rhode Island 37
By Forest P. Lyford, William C. Brandon, and Christine A. P. Williams
Description of Study Area 37
Purpose and Design of Sampling 37
Results 40
Calf Pasture Point, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown,
Rhode Island 40
By Forest P. Lyford, Christine A. P. Williams, and William C. Brandon
Description of Study Area 40
Purpose and Design of Sampling 43
Results 43
Otis Air National Guard/Camp Edwards Superfund Site, Johns Pond, Falmouth, Massachusetts 43
By Jennifer G. Savoie and Denis R. LeBlanc
Description of Study Area 43
Contents III
-------
Purpose and Design of Sampling 45
Results 45
Nyanza Chemical Waste Dump Superfund Site, Ashland, Massachusetts 48
By Forest P. Lyford, Richard E. Willey, and Sharon M. Hayes
Description of Study Area 48
Purpose and Design of Sampling 51
Results 51
Centredale Manor Restoration Project Superfund Site, North Providence, Rhode Island 52
By Peter E. Church, Forest P. Lyford, and Anna F. Krasko
Description of Study Area 52
Purpose and Design of Sampling 54
Results 54
Quality-Assurance Procedures 54
Summary and Conclusions 56
References Cited 58
Appendix 1. Laboratory and Field Testing of Passive-Vapor-Diffusion Sampler Equilibration Times, Temperature
Effects, and Sample Stability 63
By Don A. Vroblesky
Equilibration Times and Temperature Effects 65
Sample Stability 71
Appendix 2. Field Screening of Volatile Organic Compounds Collected with Passive-Vapor-Diffusion Samplers
with a Gas Chromatograph 73
By Scott Clifford
FIGURES
1-9. Photographs showing:
1. Glass vial in two layers of polyethylene tubing 2
2. Passive-vapor-diffusion samplers with (A) vial and screw cap, (B) uncapped glass vial sealed in
polyethylene tubing and secured to wire surveyor flag, and (C) glass vial sealed in polyethylene
sandwich bags and secured to wire surveyor flag 6
3. Heat sealing of glass vial in polyethylene tubing 7
4. Glass vial positioned in sandwich bag so that a single layer of low-density polyethylene is tight
across the opening and the self-locking nylon tie does not interfere with capping 8
5. Installation method for passive-vapor-diffusion samplers in water 0 to 2 feet deep 10
6. Drive-point assembly for installation of passive-vapor-diffusion sampler in water 2 to 4 feet deep
in clayey silt to coarse sand and gravel sediments 11
7. Drive-point method for installation of passive-vapor-diffusion sampler in water 2 to 4 feet deep
in clayey silt to coarse sand and gravel sediments 12
8. Screwing a septated cap onto a glass vial encased in the inner low-density polyethylene tubing 14
9. (A) Attaching and (B) crimping a septated cap onto a glass vial encased in the inner low-density
polyethylene tubing 15
10-21 Maps showing:
10. Locations of sites in New England where passive-vapor-diffusion samplers have been used to
detect and delineate discharge areas of ground water contaminated by volatile organic compounds
into surface-water bodies 19
11. Location of the Eastern Surplus Superfund Site and study area, Meddybemps, Maine 21
12. Potentiometric surfaces and generalized ground-water-flow directions for the surficial and
bedrock aquifers, Eastern Surplus Superfund Site, Meddybemps, Maine, April 30, 1997 22
13. Concentrations of tetrachloroethene (PCE) in passive-vapor-diffusion samplers installed
in river-bottom sediments on the western edge of Dennys River, Meddybemps, Maine,
October 1996 24
14. Location of the McKin Superfund Site and study area, potentiometric surface contours for
the surficial aquifer, and extent of trichloroethene in ground water, Gray, Maine 26
IV Contents
-------
15. Locations of passive-vapor-diffusion samplers installed in river-bottom sediment along and
near the Royal River in September and October 1997, and extent of trichloroethene in ground
water, Gray, Maine 27
16. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in river-bottom
sediments near Boiling Springs, Gray, Maine, September and October 1997 29
17. Location of the Nutmeg Valley Road Superfund Site and study area, Nutmeg Valley, Wolcott
and Waterbury, Connecticut 30
18. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers
installed in river-bottom sediments of the Mad River, Old Tannery Brook, and an unnamed stream,
Nutmeg Valley, Wolcott and Waterbury, Connecticut, My 1997 32
19. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers
installed in river-bottom sediments of the Mad River, Old Tannery Brook, and an unnamed stream,
Nutmeg Valley, Wolcott and Waterbury, Connecticut, November 1997 33
20. Location of the Baird & McGuire Superfund Site and study area, Holbrook, Massachusetts 35
21. Concentrations of trichloroethene plus tetrachloroethene and petroleum compounds in
passive-vapor-diffusion samplers installed in river-bottom sediments of the Cochato River,
Baird & McGuire Superfund Site, Holbrook, Massachusetts, March and April 1998 36
22. Aerial photograph showing locations of the Allen Harbor Landfill and Calf Pasture Point study areas,
Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island 38
23-33. Maps showing:
23. Directions of ground-water flow in the shallow and deep surficial aquifers, concentrations
of volatile organic compounds in ground water beneath the Allen Harbor Landfill, December 1995,
and concentration of trichloroethene in passive-vapor-diffusion samplers installed in tidal-zone
sediments along the shoreline of Allen Harbor Landfill, April 1998, Davisville Naval Construction
Battalion Center Superfund Site, North Kingstown, Rhode Island 39
24. Potentiometric surfaces and generalized ground-water-flow directions for the shallow and deep
surficial aquifers, Calf Pasture Point, Davisville Naval Construction Battalion Center Superfund
Site, North Kingstown, Rhode Island, December 1995 41
25. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in the tidal-zone
sediments along the shoreline and in wetland-bottom sediments near the shoreline, Calf Pasture
Point, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode
Island, March and April 1998 42
26. Locations of the Johns Pond study area and Storm Drain-5 contaminant plume, and the altitude of
water table (March 1993), Cape Cod, Massachusetts 44
27. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers
installed in pond-bottom sediments adjacent to the Storm Drain-5 contaminant plume, Johns Pond,
Cape Cod, Massachusetts, August 1998 46
28. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in pond-bottom
sediments in the zones where high concentrations (greater than 10,000 part per billion by volume)
of trichloroethene were detected with passive-vapor-diffusion samplers in August 1998, Johns Pond,
Cape Cod, Massachusetts, December 1998 47
29. Discharge areas delineated with passive-vapor-diffusion samplers, August and December 1998,
and ground-water pathways of the Storm Drain-5 plume and trichloroethene plumes, Johns Pond,
Cape Cod, Massachusetts 48
30. Location of Nyanza Chemical Waste Dump Superfund Site, passive-vapor-diffusion sampler
locations, potentiometric-surface contours for the surficial aquifer, and directions of ground-water
flow, Ashland, Massachusetts 49
31. The extent of contaminants in ground water and concentrations of chlorobenzene and
trichloroethene detected in passive-vapor-diffusion samples, Nyanza Chemical Waste Dump
Superfund Site, Ashland, Massachusetts, February 1999 50
32. Locations of the Centredale Manor Restoration Project Superfund Site and study area, North
Providence, Rhode Island 52
Contents V
-------
33. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers
installed in channel-bottom sediments of the Woonasquatucket River, a former mill raceway, and
a cross channel, Centredale Manor Restoration Project Superfund Site, North Providence, Rhode
Island, September 1999 53
1A-1D. Graphs showing:
IA. Changes in concentrations of volatile organic compounds in passive-vapor-diffusion samples over
time at 21 degrees Celsius under laboratory conditions in a mixed solution of volatile organic
compounds with aqueous concentrations less than 100 micrograms per liter 68
IB. Changes in concentrations of volatile organic compounds in passive-vapor-diffusion samples over
time at 10 degrees Celsius under laboratory conditions in a mixed solution of volatile organic
compounds with aqueous concentrations ranging from 430 to 570 micrograms per liter 69
IC. Ratio over time of (A) toluene and (B) tetrachloroethene gas concentrations by volume (parts
per billion) in passive-vapor-diffusion samplers to aqueous concentrations by mass (210 to
310 micrograms per liter of toluene and 110 to 340 micrograms per liter of tetrachloroethene)
in a test solution containing the diffusion samplers in 1.9-liter jars at average temperatures of
22.4, 9.5, and 1.4 degrees Celsius 70
ID. Changes in trichloroethene concentrations over time in passive-vapor-diffusion samples from
contaminated ground-water discharge areas in South Carolina in (Site 1) Coastal Plain sediments
and (Site 2) Piedmont sediments with differing sediment types and vertical hydraulic gradients 71
TABLES
1. Volatile organic compounds detected under field conditions with passive-vapor-diffusion samplers
at contaminated ground-water-discharge areas in New England and South Carolina and the range of
minimum reporting limits for these compounds at the nine New England sites 3
2. Superfund sites in New England where passive-vapor-diffusion samplers were used to detect and
delineate volatile organic compounds in bottom sediment of surface-water bodies, hydrologic setting,
principal compounds detected, and maximum vapor concentration measured 20
3. Number and distribution of duplicate samples from the nine study sites in New England 55
4. Relative percent differences of volatile organic compound (VOC) concentrations in duplicate samples
where a VOC was detected above the reporting limit in both samples from the nine study sites in
New England 56
IA. Average concentrations and standard deviations of volatile organic compounds in passive-vapor-diffusion
samplers over time at 21 degrees Celsius under laboratory conditions in 480-milliliter test jars with
spiked concentrations less thanlOO micrograms per liter 66
IB. Average concentrations and standard deviation of volatile organic compounds in passive-vapor-diffusion
samplers over time at 10 degrees Celsius under laboratory conditions in 1.9-liter test jars with spiked
concentrations ranging from 430 to 570 micrograms per liter 66
IC. Ratio of concentrations from passive-vapor-diffusion samplers to aqueous-phase concentrations for
toluene and tetrachloroethene over time at various temperatures under laboratory conditions in 1.9-liter
test jars containing 210 to 310 micrograms per liter of toluene and 110 to 340 micrograms per liter of
tetrachloroethene 67
ID. Average concentrations and standard deviations of trichloroethene over time in passive-vapor-diffusion
samples in bottom sediment of streams at contaminated ground-water discharge areas in the Coastal Plain
(site 1) and the Piedmont (site 2) of South Carolina, 1998 72
2A. Typical achievable reporting limits for volatile organic compounds commonly detected in passive-vapor-
diffusion samplers from a gas chromatograph equipped with a photoionization detector and an electron
capture detector 76
2B. Vapor concentrations of volatile organic compounds commonly detected with passive-vapor-diffusion
samplers in the head space of a 10 micrograms per liter aqueous standard at approximately
0 to 1 degree Celsius 77
2C. Quality controls, acceptance criteria, and corrective actions 79
VI Contents
-------
CONVERSION FACTORS, VERTICAL DATUM, AND ABBREVIATIONS
CONVERSION FACTORS
Multiply
By
To obtain
acre
0.4047
hectare
feet (ft)
0.3048
meters
feet per day (ft/d)
0.3048
meters per day
inches (in.)
2.54
centimeters
mil
0.0254
millimeters
square feet (ft2)
0.0929
square meters
Temperature in degrees Celsius (°C) can be converted to degrees Fahrenheit (°F) as follows:
°F = 1.8°C+32
VERTICAL DATUM
Sea Level: In this report, "sea level" refers to the National Geodetic Vertical Datum of 1929 (NGVD
of 1929), a geodetic datum derived from a general adjustment of the first-order level nets of the United
States and Canada, formerly called the Sea Level Datum of 1929.
ABBREVIATIONS
Jj,g/L micrograms per liter
mg/L milligrams per liter
mL milliliter
mol/m3 moles per cubic meter
PCE tetrachloroethene
ppb v parts per billion by volume
PVC polyvinylchloride
PVD sampler passive-vapor-diffusion sampler
SVOCs semi-volatile organic compounds
TCE trichloroethene
VOCs volatile organic compounds
ds-DCE cw-l^-dichloroethene
Concentration of chemical constituents in air are given in parts per billion by volume (ppb v).
Concentration of chemical constituents in water are given in micrograms per liter (|J,g/L).
Contents VII
-------
Guidance on the Use of Passive-
Vapor-Diffusion Samplers to Detect
Volatile Organic Compounds in
Ground-Water-Discharge Areas, and
Example Applications in New England
By Peter E. Church, Don A. Vroblesky, Forest P. Lyford, and Richard E. Willey
Abstract
Polyethylene-membrane passive-vapor-
diffusion samplers, or PVD samplers, have been
shown to be an effective and economical recon-
naissance tool for detecting and identifying vola-
tile organic compounds (VOCs) in bottom
sediments of surface-water bodies in areas of
ground-water discharge. The PVD samplers con-
sist of an empty glass vial enclosed in two layers
of polyethylene membrane tubing. When samplers
are placed in contaminated sediments, the air in
the vial equilibrates with VOCs in pore water.
Analysis of the vapor indicates the presence or
absence of VOCs and the likely magnitude of
concentrations in pore water.
Examples of applications at nine hazardous-
waste sites in New England demonstrate the utility
of PVD samplers in a variety of hydrologic set-
tings, including rivers, streams, ponds, wetlands,
and coastal shorelines. Results of PVD sampling
at these sites have confirmed the presence and
refined the extent of VOC-contaminated ground-
water-discharge areas where contaminated ground
water is known, and identified areas of VOC-
contaminated ground-water discharge where
ground-water contamination was previously
unknown. The principal VOCs detected were
chlorinated and petroleum hydrocarbons. Vapor
concentrations in samplers range from not
detected to more than 1,000,000 parts per billion
by volume. These results provided insights about
contaminant distributions and ground-water-flow
patterns in discharge areas, and have guided the
design of focused characterization activities.
INTRODUCTION
Passive-vapor-diffusion (PVD) samplers are
designed and primarily used as a reconnaissance
tool to detect and identify volatile organic compound
(VOC) contaminated ground water discharging into
surface-waters bodies at and near hazardous-waste
sites (Vroblesky and others, 1996; Vroblesky and
Robertson, 1996; Vroblesky and Hyde, 1997). Deter-
mining the location of discharging contaminated
ground water is important for plume mapping, evaluat-
ing risk potential to human health and the environment,
and designing focused site-characterization and moni-
toring activities. Applications of PVD samplers at and
near nine hazardous-waste sites in New England dem-
onstrated the samplers' effectiveness in detecting and
delineating VOCs in a variety of hydrologic settings
including rivers, streams, ponds, wetlands, and coastal
shorelines. The PVD samplers also have been used suc-
cessfully as passive-soil-gas samplers in unsaturated
zones to map ground-water contamination (Vroblesky
and others, 1992).
Introduction 1
-------
The PVD samplers, developed by the U.S.
Geological Survey (USGS) (Vroblesky and others,
1996) consist of an empty, uncapped, glass vial
enclosed in two layers of low-density polyethylene
membrane tubing (fig. 1) that are permeable to many
VOCs of environmental interest, such as petroleum and
chlorinated compounds (table 1), but not permeable to
water (Vroblesky and others, 1991). When samplers
are buried in VOC-contaminated pore water in the
bottom sediment of surface-water bodies, an equilib-
rium begins to develop between VOC concentrations in
water and the air in the vial. Equilibrium times, which
are dependent on many factors such as hydraulic con-
ductivity of the sediment and temperature of the pore
water, generally range from 1 to 3 weeks. During sam-
pler recovery the outer tubing, which is used to prevent
sediment, which may be contaminated, from coming
into contact with the inner tubing and the opened
vial, is removed. A cap is then screwed on to the vial,
thereby securing the inner tubing tight against the vial
opening to prevent loss of VOCs vapor in equilibrium
with water concentrations at the sampler deployment
point.
Concentrations of VOCs detected in air in a sam-
pler indicate vapor-phase concentrations in sediment-
pore water. The relative concentration partitioning into
the air and water varies among VOCs and is described
by Henrys' law constant for the particular VOC. A
compound with a relatively high vapor pressure and
low solubility will tend to become more concentrated
in the vapor phase than in the water phase. Several of
VOCs with lower vapor pressure tend to be more con-
centrated in the water phase, but still maintain a vapor-
phase signature. Because the low-density polyethlene
membrane tubing is not a major barrier to VOC diffu-
sion over time, PVD samplers provide a vapor phase
into which VOCs can diffuse from the aqueous phase.
The VOCs in water near PVD samplers diffuse through
the tubing into the air within the glass vial. Vapor-phase
concentrations in the vial are typically reported in parts
per billion by volume (ppb v).
In theory, these vapor concentrations can be con-
verted to concentrations in water through Henry's Law
and Henry's Law constants for specific chemicals. In
practice, however, uncertainties about Henry's Law
constants, pore-water temperatures, equilibration times
for various types of sediments, and analytical preci-
sion, limit this application. If concentrations in water
are needed, however, a modification of this approach,
which is a single layer, water-filled membrane tubing
sampler, will provide aqueous-phase VOC concentra-
tions in a ground-water-discharge area (Vroblesky and
others, 1999).
The ease of constructing, deploying, and retriev-
ing PVD samplers renders this method well-suited for
reconnaissance of VOC plumes discharging to surface
waters. A large amount of spatial data can be collected
in a short period with PVD samplers. For example,
Figure 1. Glass vial in two layers of polyethylene tubing.
2 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Table 1. Volatile organic compounds detected under field
conditions with passive-vapor-diffusion samplers at
contaminated ground-water-discharge areas in New England
and South Carolina and the range of minimum reporting
limits for these compounds at the nine New England sites
Range of minimum
, reporting limits,
Volatile organic compound *
in parts per billion
by volume
Benzene 6 to 25
Ethylbenzene 40 to 90
meta/para-xylene 40 to 90
orf/zo-xylene 60 to 100
Toluene 20 to 40
Tetrachloroethene 5 to 25
Trichloroethene 5 to 25
Chlorobenzene 40 to 70
ds-l^-Dichloroethene 25 (a target compound at only
one site)
1,1,1 -Trichloroethane 8 (a target compound at only
one site)
methyl tert butyl ether Not a target compound at the
New England sites
from experience with PVD samplers in New England,
construction of samplers, deployment of samplers in
streambed sediments at 50 ft intervals along a 2,000 ft
reach, retrieval of samplers after equilibration is
reached (about two weeks), and on-site chemical analy-
sis, may be accomplished in a total of three days.
Results from this sampling may guide the placement of
well points in shallow waters and monitoring wells or
the selection of sediment-sampling locations where
VOCs in water and sediments are detected. Once vapor
concentrations of VOCs are detected with the PVD
samplers, then water samples can be collected with
other methods.
This report describes advantages and limitations
of PVD samplers, offers guidance on the use of PVD
samplers, and summarizes results from nine sites in
New England where PVD samplers have provided
useful information about VOC plumes in ground water.
The report is designed mainly for personnel who are
designing characterization studies that may include
the use of PVD samplers and for personnel who will
be constructing and installing the samplers. This
report was prepared by the U.S. Geological Survey in
cooperation with the U.S. Environmental Protection
Agency (USEPA) Technical Innovation Office (TIO),
and USEPA's Region I.
The authors acknowledge the following individ-
uals, who are members of agencies within the Federal
Remediation Technologies Roundtable (FRTR), for
reviews of this report: Kathryn Davies, Vincent Malott,
Katherine Baylor, Jeffrey Johnson, Richard Muza,
and Ernest Waterman, USEPA, Ground Water Forum;
Deborah Sherer and Kristie Dymond, USEPA, Office
of Solid Waste and Emergency Response; Dominic C.
DiGiulio and Timothy Canfield, USEPA, National
Risk Management Research Laboratory, Ada, OK;
Charles Porfert, USEPA, New England Office of
Environmental Measurements and Evaluation; Robert
Lien, USEPA, National Risk Management Research
Laboratory, Cincinnati, OH; Chung-Rei Mao, Stephen
White, and Jeffrey L. Breckenridge, U.S. Army Corps
of Engineers, Center of Expertise; and Doug Zillmer,
U.S. Naval Facilities Engineering Command Service
Center. In addition, we also acknowledge George H.
Nicholas, New Jersey Department of Environmental
Protection on behalf of the Interstate Technology and
Regulatory Cooperation (ITRC), Diffusion Sampler
Team, for his review of this report.
ADVANTAGES AND
LIMITATIONS OF PASSIVE-
VAPOR-DIFFUSION SAMPLERS
Advantages and limitations are presented here,
before the details of PVD-sampler assembly, deploy-
ment, and recovery are described, to ensure that project
design personnel who may be considering or are plan-
ning to use PVD samplers understand capabilities of
this method. These advantages and limitations may
also be useful for technicians who assemble, deploy,
and recover these samplers.
Advantages
1. The PVD method takes advantage of converging
ground-water-flow lines and upward hydraulic
gradients at ground-water-discharge areas to
bring the target contaminants into contact with
the samplers.
2. The method has been effective in delineating
VOC-contamination-discharge areas beneath
surface-water bodies.
3. PVD samplers can be areally and vertically
distributed to gain information on contaminant-
discharge heterogeneity.
Advantages and Limitations of Passive-Vapor-Diffusion Samplers 3
-------
4. The samplers are inexpensive. A low-cost sampler
can be made from grocery-store sandwich bags
and empty glass vials.
5. In many situations, the samplers are easy to
deploy and recover.
6. Sampler recovery is rapid. The data can be ana-
lyzed on site with field gas chromatography,
or the capped samples can be stored for later
analysis. A stability test with 40-mL volatile
organic analysis (VOA) vials showed that VOC
concentrations in unpunctured, sealed vials did
not substantially decrease in over 121 hours
(Vroblesky and others, 1996).
7. Because the pore size of low density polyethylene
tubing is about 10 angstroms or less, sediment
does not pass through the membrane into the
bag. The outer bag, therefore, effectively pre-
vents contaminated soil from contacting the
inner bag.
8. A variation of a PVD sampler can be used as a
soil-gas sampler in unsaturated sediment to
delineate shallow VOC plumes (Vroblesky and
others, 1992). Vapor-tilled polyethylene
samplers also can be used with a sorbent to
allow the samplers to accumulate VOC concen-
trations over the deployment period (Vroblesky
and others, 1991). Still another variation con-
sists of water-tilled, low-density polyethylene
tubing sampler deployed in bottom sediments
at ground-water-discharge areas to yield aque-
ous concentrations of VOCs (Vroblesky and
others, 1999).
Limitations
1. Because the change in VOC concentrations within
PVD samplers in response to changes in ambi-
ent concentrations typically takes 24 hours or
longer, VOC concentrations within the sam-
plers represent an integration of concentrations
from the most recent part of the deployment
period until the samplers attain equilibrium.
The equilibration time depends on several fac-
tors, including the temperature and the rate of
water movement past the sampler. Under labo-
ratory conditions, equilibration times in static
water ranged from about 24 hours at 21 °C to
about 102 hours at 10°C. Under field condi-
tions, equilibration times can range from as
little as 12 hours in a rapidly discharging
unconsolidated sand, to three weeks or more in
colder, less permeable sediment. Suggested
PVD deployment periods are typically two
weeks, but may vary depending on site-specific
temperature and hydraulic conditions. The
required equilibration time is a disadvantage
over some types of real-time sampling meth-
ods, such as extracting water from a core or
pumping water from a small-diameter probe.
Unlike these methods, however, the PVD sam-
plers can provide an undisturbed sample, which
minimizes the risk of short-term concentration
changes from sediment disturbance and reduces
the uncertainty associated with the source of
water from a sample obtained from pumping.
2. The PVD samplers are appropriate only for
volatile compounds.
3. Analysis of the samples requires a gas
chromatograph.
4. Deployment of the samplers in shallow waters
typically is a simple task; however, deployment
in deep waters may require the services of
SCUBA divers or other installation methods.
5. In some streams, the source of detected VOCs
may not be readily determined without further
work because of complexities in hydraulics and
sediment heterogeneity that lead to unusual
contaminant-discharge distributions. For exam-
ple, in some streams of the Rocky Mountains
region, where ground-water-flow direction is
approximately parallel to streamflow, locations
of ground-water-discharge areas can change
with time. Furthermore, if the samplers are
deployed in an area of VOC-contaminated
bottom sediment derived from sediment trans-
port along the stream, then the VOC concentra-
tions in the PVD samplers may reflect
contaminant concentrations in the sediment
rather than in discharging ground water. Conse-
quently, an effort should be made to ensure that
the sampling location is a gaining reach.
6. The samplers must be deployed in an area where
ground water is discharging to surface water to
adequately reflect ground-water concentrations.
In areas where the water in contact with the
PVD samplers is largely infiltrated surface
water, the concentrations detected by the PVD
4 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
samplers probably will represent surface-water
or sediment-contaminant concentrations rather
than ground-water concentrations.
7. Transient flow of ground water, such as bank
storage after a flood wave or tidal cycle, may
cause temporary or cyclic changes in concen-
trations of VOCs that could affect interpreta-
tions about the extent and concentration level of
VOCs in ground water.
8. Follow-up studies that use other sampling meth-
ods are needed to determine actual concentra-
tions of VOCs in water, if that is a goal of the
study. The PVD samplers provide VOC concen-
trations as a gas. Because the partition between
aqueous and vapor phases depends on several
factors, such as temperature and pressure,
which may vary from site to site and are not
always known, calculations of aqueous-phase
concentrations from PVD samplers should be
considered estimates. An alternative approach
to obtaining aqueous-phase VOC concentra-
tions in a ground-water-discharge area is to use
a water-filled, low-density polyethyene sampler
as described elsewhere (Vroblesky and others,
1999).
9. Caution should be used when deploying PVD
samplers in streambeds subject to rapid ero-
sion, because the samplers may be washed
away. Samplers also may be difficult to find
when the surveyor flags are submerged in high
flows in streams or buried beneath sediment as
flows recede.
PART 1. GUIDANCE ON THE
USE OF PASSIVE-VAPOR-
DIFFUSION SAMPLERS
By Don A. Vroblesky
This section of the report provides guidance on
PVD-sampler assembly, deployment, and recovery to
detect volatile organic compounds in ground-water-
discharge areas. As an aid to ensure proper use of this
method and interpretation of the data collected, factors
affecting PVD-sampler deployment, data interpreta-
tion, and quality control and assurance, are also
discussed.
ASSEMBLY OF SAMPLERS
Several approaches may be used to construct a
PVD sampler. A vial may be enclosed in "lay-flat,"
low-density polyethylene (LDPE) tubing, flexible
tubing that is laid flat and wound-up in a roll, and then
heat-sealed at both ends, or the vial may be enclosed in
zipper-type sealable sandwich bags (fig. 2). The vials
should be glass and sealable with a septated cap. Typi-
cal vials used include 20-mL crimp-top glass vials and
40-ml volatile organic analysis (VOA) glass screw-top
vials. Vials used in the studies summarized in this
report are 40-mL VOA screw-top vials. When the
20-mL crimp-top glass vials are used, 2-, 3-, or 4-mil
LDPE may be used for the bag material. When 40-mL
VOA vials are used, 2- or 3-mil LDPE for the inner
material is advised, because the cap is difficult to screw
onto 4-mil LDPE. The septum for the cap should be
Teflon or Teflon coated. The effort for constructing
PVD samplers is approximately equivalent for the two
methods. Also, the material costs differ little between
methods, except for the initial purchase of a heat sealer
if constructing samplers with tubing. The LDPE tub-
ing, however, is more resilient to punctures and abra-
sion during placement and retrieval, and is therefore
preferred to construction with sandwich bags, particu-
larly for placement in coarse gravels.
To construct a PVD sampler from lay-flat LDPE
tubing and a heat sealer, the following supplies are
needed: 2-in. wide (approximately 1.5-in. diameter),
2-, 3-, or 4-mil lay-flat LDPE tubing; a glass vial, a
wire surveyor flag, self-locking nylon ties, and a heat
sealer. The following approach describes a typical
PVD-construction sequence.
1. Cut an 8-in. length of 2-in. wide lay-flat LDPE
tubing. The inner layer of tubing should be 2-
or 3-mil thick when VOA vials are used and 2-,
3-, or 4-mil when crimp-top serum vials are
used. For the studies summarized in this report,
40-mL VOA screw-top vials were used.
2. Heat-seal one end. The heat sealer should be
adjusted to provide a uniform seal without
melting through the LDPE. Multiple seals may
be required with some heat sealers with less
than 400 watts of impulse power.
3. Insert an uncapped empty 40-mL glass VOA vial
or a 20-mL crimp-top serum vial into the tube.
Store the cap in a clean environment away
from the PVD samplers until the sample is
recovered.
Assembly of Samplers 5
-------
B.
C.
Figure 2. Passive-vapor-diffusion samplers with (A) vial arid screw cap, (B) uncapped glass vial sealed in polyethylene
tubing and secured to wire surveyor flag, and (C) glass vial sealed in polyethylene sandwich bags and secured to wire
surveyor flag.
4, Remove the excess air space from the LDPE tub-
ing. This can be accomplished by squeezing the
LDPE tubing tightly against the vial or by
twisting the tubing to tighten it against the vial.
5a. Position the unsealed end of the bag across the
sealing element of the heat sealer, so that the
sealing element is as close as practical to the
mouth of the enclosed vial without stretching
the LDPE across the opening (fig. 3). The
LDPE should not be folded or wrinkled where
it crosses the heating element of the heat sealer.
Seal the bag. Once the bag is sealed, trim off
the excess LDPE tubing. In this method, the
LDPE is not necessarily tight across the vial
opening.
5b. An optional method to arranging the heat-sealed
end of the inner LDPE tubing across the vial
opening is to, after both ends of the tubing are
sealed, pull the tubing over the vial opening and
fold the heat-sealed end of the tubing against
the glass vial. After folding, secure the folded
tubing to the vial with a self-locking nylon tie
in a place where it will not interfere with
attaching a vial cap during sampler recovery. In
this method, the LDPE should be tight against
the vial opening. This method improves sam-
pler integrity over the first method because it
reduces the probability of accidentally cutting
the inner LDPE while removing the outer
LDPE during sampler recovery.
6 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Figure 3. Heat sealing of glass vial in polyethylene tubing.
6. Once the inner LDPE tubing is secure, cut another
8-in. length of 2-in. wide lay-flat LDPE tubing.
This tubing will constitute the outer layer,
which will be removed during sampler recovery
to prevent sediment from interfering with
capping the vials.
7. I-Ieat-seal one end. The heat sealer should be
adjusted to provide a uniform seal without
melting through the LDPE. Multiple seals may
be required with some heat sealers with less
than 400 watts of impulse power.
8. Place the glass vial enclosed in the inner layer of
LDPE into the new LDPE tube.
9. Press the LDPE against the vial to remove the
excess air space from the outer LDPE tubing.
The lack of air space will reduce buoyancy and
maintain a consistent vapor volume.
10. Position the unsealed end of the bag across the
sealing element of the heat sealer so that the
sealing element is as close as practical to the
mouth of the enclosed vial. No folds or wrin-
kles should be present where the LDPE tubing
crosses the heating element of the heat sealer.
Seal the bag. Once the bag is sealed, trim off
the excess LDPE tubing.
11. Attach a wire surveyor flag to the PVD sampler to
aid in sampler recovery. A practical method of
attachment is to use self-locking nylon ties. The
ties are attached tightly enough so that the sur-
veyor flag does not pull free from the sampler
during sampler recovery. The vial is attached so
that the vial opening is in the opposite direction
of the surveyor flag (fig. 2), One approach is to
allow approximately 2 in. of wire extending
beyond the nylon tie. The 2 in. of wire is then
bent back 180 degrees over the nylon tie and
laid adjacent to the wire above the tie. Add
another nylon tie to secure the bent part of the
wire tight against the vial (fig. 2). By this
method, the surveyor flag does not pull free
from the sampler during retrieval. The widest
practical spacing between nylon ties will
reduce a tendency for samplers to rotate on
the wire and become wedged in sediment
during retrieval.
The following method details how to construct a
sampler with sandwich bags instead of lay-flat tubing.
1. Place an uncapped empty glass vial in a zipper-
type polyethylene sandwich bag. Store the cap
in a clean environment away from the PVD
samplers until sampler recovery.
Assembly of Samplers 7
-------
2. Remove the excess air space from the bag to
reduce buoyancy of the sampler when
deployed. This can be accomplished by squeez-
ing or rolling the bag around the vial. Seal the
bag with the zipper.
3. Secure the bag around the vial opening by tight-
ening a self-locking nylon tie below the vial
opening so that the tie and the excess bag will
not interfere with capping the vial. A single
layer of LDPE should now be tight across the
vial opening (fig. 4).
4. Place the glass vial enclosed in the LDPE bag into
another sandwich bag. Remove excess air from
the outer bag and seal it with the zipper.
5. Attach a wire surveyor flag to the PVD sampler to
aid in sampler recovery. A practical method of
attachment is to use self-locking nylon ties. The
ties are attached tightly enough so that the sur-
veyor flag does not pull free from the sampler
during sampler recovery. The vial is attached so
that the vial opening is in the opposite direction
of the surveyor flag (fig. 2). One approach is to
allow approximately 2 in. of wire extending
beyond the nylon tie. The 2 in. of wire is then
bent back 180 degrees over the nylon tie and
laid adjacent to the wire above the tie. Add
another nylon tie to secure the bent part of the
wire tight against the vial (fig. 2). By this
method, the surveyor flag does not pull free
from the sampler during retrieval. The widest
practical spacing between nylon ties will
reduce a tendency for samplers to rotate on the
wire and become wedged in sediment during
retrieval.
Figure 4. Glass vial positioned in sandwich bag so that a single layer of low-density
polyethylene is tight across the opening and the self-locking nylon tie does not interfere
with capping.
8 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
DEPLOYMENT OF SAMPLERS
Effective use of the samplers to delineate dis-
charging contaminated ground water requires that sam-
plers be deployed in ground-water-discharge areas.
Many approaches have been used to identify gaining
reaches of streams (zones where ground water dis-
charges to surface water). To verify that the target sec-
tion of the surface water is a gaining reach, install a
streambed piezometer. In its simplest form, a stream-
bed piezometer can be a pipe driven a few feet into the
bed sediment. A bolt loosely fitted into the bottom of
the pipe before installation can prevent sediment from
moving into the pipe. After deployment a narrow rod
then can be used to drive out the bolt to allow water to
enter the pipe. After stabilization, comparison of
ground-water-head measurements within the piezome-
ter to surface-water stage outside the piezometer can
indicate whether the head gradient is upward (gaining
reach) or downward (losing reach). With this method,
care should be taken to avoid a clogged pipe or a pipe
with a leaky connection to surface water along the
annular space between pipe and streambed sediments.
Many other methods have been used to identify
areas of upwelling ground water beneath surface water.
Near-shore discharge through a lakebed has been esti-
mated with seepage devices (Lee, 1977) and hydraulic
potentiomanometers (Winter and others, 1988). In
areas where temperature differences between ground
water and surface water are greater than normally
expected, surface-water-temperature measurements
and aerial infrared photography have been used to
identify areas of ground-water discharge to streams
(Silliman and Booth, 1993), lakes, and wetlands
(Olafsson, 1979; Lee, 1985; Lee andTracey, 1984;
Baskin, 1998). Discharge areas of ground water to
lakes sometimes can be located by towing temperature
and specific-conductance probes from a boat (Lee,
1985). Researchers also have used the distribution of
aquatic plant species as indicators of ground-water
discharge to fens (Glaser and others, 1981; 1990;
Verhoeven and others, 1988; Wassen and others, 1989),
to saline wetlands (Swanson and others, 1984), and to a
lake (Rosenberry and others, 2000).
Even within gaining reaches of a stream,
the distribution of contaminant discharge can be
complex. After storms, ground-water discharge may
be dominated by release of bank storage. This
transient flow may temporarily mask contaminated
ground-water discharge.
Deployment of PVD samplers involves burying
of samplers in the bottom sediment of a surface-water
body. Ideally, the samplers should be buried at the
bottom of the transition zone from surface water to
ground water to ensure that the sample collected repre-
sents VOCs in ground water. Delineating the transition
zone, however, often is difficult, and holes dug beneath
the water tend to rapidly refill with sediment. Samplers
placed at shallow depths (for example, 0 to 0.5 ft) may
be within this transition zone and samples may be
affected by surface water. Samplers placed at shallow
depths may also become dislodged. Samplers placed
at greater depths (for example, greater than 1.5 ft) may
be below the transition zone, but also may be difficult
to retrieve. The most effective depths of sampler
deployment may vary spatially and with time, and are
dependent on many factors, including hydraulic con-
ductivities of the sediments and hydrologic conditions.
With the deployment method described, a practical
target depth is between 0.5 and 1.5 ft. Deployment
depths are described in detail more in the section
"Factors Affecting Deployment of Samplers and Data
Interpretation."
In shallow waters, waters up to 2 ft deep, where
the samplers can be installed with hand augers or shov-
els (fig. 5), one approach to digging the hole is to
shovel the sediment until the likelihood of hole col-
lapse makes further digging impractical. At that point,
insert the shovel into the sediment and push forward
to create an opening between the back of the shovel
and the sediment. In more cohesive sediment, the hole
can be excavated with a hand auger. Exercise care
during insertion of the PVD sampler into the hole to
prevent rupturing of the polyethylene membrane cover-
ing the vial opening. Backfill the hole with the inserted
PVD sampler with the sediment removed from the
hole. Ensure that the hole has been adequately back-
filled above the sampler to minimize entrainment of the
top-most sediment layer above the PVD sampler to the
bottom of the hole. To reduce the potential for contami-
nation from sample-labeling pens, label the surveyor
flag either several days before PVD-sampler construc-
tion (to allow vapors from water-proof markers to
dissipate) or after the sampler is buried.
Deployment of Samplers 9
-------
«
0
<
1
fi>
3
¦D
-------
In water deeper than about 2 ft, manual
insertion of a sampler in a hand-augered hole or
behind a shovel blade is not practical in conven-
tional wading gear (hip or chest waders). A
drive-point assembly has been effective in water
2 to 4 ft deep where chest waders are needed.
The drive-point: assembly has also been used
from a row boat (Campbell and others, 2002)
and through an ice cover (Lyford and others,
2000; Church and others, 2002) to install sam-
plers in soft bottom sediment at depths to about
7 ft. Greater depths of installation may be possi-
ble from a boat, barge, or through ice, but has
not been attempted. The drive-point assembly
and its application are illustrated in figures 6
and 7.
Drive-point assemblies can be con-
structed at various lengths to suit different
needs. The drive-point assembly most com-
monly used in New England studies consists of
a 72-in. long, 1 3/4-in. outside diameter (01)).
1 5/8-in. inside diameter (ID) electrical conduit
outer pipe; a74-in. long, 1 1/2-in. OD, 1 3/8-in.
ID electrical conduit inner pipe; and an 80-in.
long, 7/8-in. OD, 3/4-in. ID polyvinylchloride
(PVC) pipe. A 2-in. OD pipe cap is attached to
the top of the 1 I/2-in. inner pipe, and a 2-in.
long steel point is flush-mounted to the bottom.
The outer and inner pipes are driven into the
bottom sediments by striking the pipe cap on
the inner pipe with a sledge or slide hammer. At
the desired depth of installation, the inner pipe
is removed, leaving a hole in the sediment
extending about 2-in. deeper than the bottom of
the outer pipe. The surveyor flag end of the
PVD sampler is then pushed into the PVC pipe
to where it is stopped by the tubing of the sam-
pler. The PVC pipe and attached sampler are
then inserted and pushed through the outer
pipe into the sediment. The sampler is held in
place by the PVC pipe as the outer pipe is
removed. Sediments then collapse around the
sampler, and the PVC pipe is removed from the
sediment.
Figure 6. Drive-point assembly for installation of passive-vapor-
diffusion sampler in water 2 to 4 feet deep in clayey silt to coarse
sand and gravel sediments. (A) slide hammer, (B) 1 5/8-inch inside
diameter steel electrical conduit, (C) 1 3/8-inch inside diameter
steel electrical conduit with machined point, (D) polyvinylchloride
sampler-insertion pipe.
Deployment of Samplers 11
-------
D. E. F.
Figure 7. Drive-point method for installation of passive-vapor-diffusion sampler in water 2 to 4 feet deep in clayey silt to coarse sand
and gravel sediments. (A) 1 5/8-inch inside diameter (ID) steel electrical conduit with 1 3/8-inch ID steel insert conduit with 2-inch
point driven into pond-bottom sediment with slide hammer, (B) insert pipe removed after driven to desired depth, (C) insertion of
passive-vapor-diffusion sampler to polyvinylchloride (PVC) pipe (surveyor flag and wire inserted into pipe with sampler exposed),
(D) insertion of PVC pipe with sampler into sediment through 1 5/8-inch ID steel conduit, (E) removal of 1 5/8-inch ID steel conduit,
and (F) removal of PVC pipe leaving sampler installed in sediment.
12 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
At depths greater than 4 ft, divers may be needed
to install PVD samplers. Divers have been used to
install samplers in pond-bottom sediment to pond
depths of 30 ft (Savoie and others, 2000). Divers
inserted PVD samplers in soft sediment by hand and
used a trowel in gravelly sediments.
For studies in New England, the samplers were
found to be resistant to removal by high flows if
buried to depths of 8 in. or greater into the sediment.
Less than five percent of the approximate 1,250 sam-
plers installed in the New England sites were lost. In
a few instances, firmly planted samplers were not
found when the surveyor wire flag broke free from
the self-locking nylon ties upon retrieval, or when
stream-channel sediment buried the flag.
RECOVERY OF SAMPLERS
The amount of time PVD samplers must remain
deployed in the sediment before recovery is, in part,
based on the data-quality objective of the study. If the
objective is to identify the presence or absence of
VOC(s), samplers could be recovered within a few
days. Field studies suggest that this can be accom-
plished after a deployment period of 8 days or less,
and as little as 24 hours in some environments. If the
objective is to estimate the concentration(s) of VOC(s),
samplers cannot be recovered until enough time for
equilibrium has elapsed. As discussed earlier, several
factors, such as hydraulic conductivity of streambed
sediments, hydraulic gradients, and water tempera-
tures, affect the amount of equilibration time needed.
Field evidence, discussed in Appendix I, suggests that
an equilibrium period of approximately 2 weeks is ade-
quate for most investigations in sandy formations.
Longer or shorter periods may be appropriate depend-
ing on water temperatures and hydraulic conditions. It
is important to remember, however, that PVD samplers
typically are deployed in sediments as a reconnaissance
tool to locate areas where ground water contaminated
with VOCs is discharging. For this use, determining the
presence or absence of target VOCs may be sufficient
to meet the data-quality objectives of the sampling. A
recent study showed that within 24 hours in four sepa-
rate streams, the recovered PVD samplers contained
chlorinated aliphatic compounds from discharging
ground water at concentrations well above detection
limits, although the samplers had not yet equilibrated
with the ground water concentrations at three of the
sites (Vroblesky and Campbell, 2001).
Recovery of PVD samplers can be accomplished
relatively rapidly. A 2-member team is needed; one
person with "dirty hands" who retrieves samplers and
touches the outer tubing only; the other person with
"clean hands" who caps and stores samplers and
touches the inner tubing only. A second clean hands
person may be needed in situations where sets of sam-
plers (10 or 20 samplers) are delivered to an on-site
portable laboratory several times a day, and to assist in
labeling and note taking. The specific recovery steps
are listed below:
1. Pull the surveyor flag or excavate the sediment to
remove the PVD sampler. Pull with a steady
tension rather than a sudden forceful extraction
that can cause nylon ties to break. Examine the
sampler for integrity. Record unusual features,
such as discoloring or water inside the outer
bag. Discard or quickly cap and record a
sampler with a ruptured inner seal.
2. Cut and remove the outer tubing or bag from
around the vial opening. Do not pierce or cut
the inner tubing or bag, because this can allow
trapped vapors within the vial to escape or
allow ambient air to enter, resulting in incor-
rectly low VOC concentrations within the vial.
An alternative approach to removing the outer
tubing that reduces the chance of puncturing
the inner bag is to cut the outer nylon ties. After
cutting the ties, use scissors to cut the end of
the outer tubing adjacent the vial opening, and
then push on the opposite end of the outer
tubing to slide the sampler into the hands of the
clean-hands person. Diagonal cutters (electri-
cian's pliers) are effective for cutting nylon ties
that attach samplers to surveyor flags. Inspect
the sampler and record any unusual features,
such as discoloring or water inside the the inner
bag. Discard or quickly cap and record a
sampler with a ruptured inner seal.
3. Cap the vial by screwing (fig. 8) or crimping
(fig. 9) a cap onto the vial to seal the inner
tubing or bag over the vial opening. Use caps
that have a Teflon or Teflon-lined septum to
allow sampling by syringe.
Recovery of Samplers 13
-------
4. Label the vial after capping if the vial is removed from the labeled surveyor
flag. The inner tubing or bag may be cut around the bottom of the cap and
removed to facilitate labeling.
5. Store the PVD samplers away from any potential VOC-contaminant sources
and in a chilled environment (4°C; ice or refrigerator), to reduce VOC
leakage. If the vapor in the sampler is to be analyzed immediately, chilled
storage is not needed.
The vapor sample obtained from the recovered PVD sampler can be ana-
lyzed on site with a gas chromatograph. Guidance on the use of a gas
chromatograph is described in Appendix 2.
Figure 8. Screwing a septated cap onto a glass vial encased in the inner low-density
polyethylene tubing.
14 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Figure 9. (A) Attaching and (B) crimping a septated cap onto a glass vial encased in the
inner low-density polyethylene tubing.
Recovery of Samplers 15
-------
FACTORS AFFECTING
DEPLOYMENT OF SAMPLERS AND
DATA INTERPRETATION
The effectiveness of PVD samplers as a recon-
naissance tool to detect discharge areas of VOC-
contaminated ground water depends on a variety of fac-
tors affecting contaminated ground-water discharge.
Understanding these factors will help in selecting opti-
mal sampler-deployment locations and will aid in data
interpretation.
Aquifer and streambed lithologic heterogeneity
affects discharge complexity and the optimal sampler-
deployment locations. Using PVD samplers, Lyford
and others (1999a) found that an irregular pattern of
trichloroethene discharge to the Royal River, Maine,
was related to the lithologic heterogeneity of the aqui-
fer and riverbed sediments, and that significantly
higher concentrations of contaminants discharged at
sand boils. Conant (2000) found that contaminant dis-
charge to a river in Ontario was predominantly associ-
ated with local gaps in the semi-confining unit beneath
the river. Discharge of ground water beneath lakebeds
can be particularly difficult to investigate because the
water can discharge at low rates over a large area, and
both rate and area can change with time. Winter (1976,
1978) has shown that the discharge is controlled pre-
dominantly by the spatial distribution of heads and
hydraulic conductivity in the aquifer, as well as the
bathymetry and sediment type of the lake bottom.
Another factor affecting contaminant discharge
to rivers is the orientation of the river relative to the
flow direction of the ground-water contamination. An
investigation in Greenville, South Carolina, showed
that reaches of a stream that were at a sharp angle to the
axis of contamination-plume migration received
greater contaminant discharge than reaches oriented
approximately parallel to the direction of contaminant
transport (Vroblesky, 2000). In a channel meandering
through a tidally flooded wetland, the highest concen-
trations of discharging contaminants were found where
the meander approached the shoreline that contained
the ground-water contamination (Vroblesky and Lorah,
1991). This area was the most probable contaminant-
discharge area because of its proximity to the ground-
water contamination, and because the stream reach was
oriented approximately perpendicular to the ground-
water-flow path. Particular effort, therefore, should
be exercised during sampler deployment in sediment
beneath a meandering stream to ensure adequate
density of sampling locations in reaches where the
stream is oriented at sharp angles to the contaminant
transport direction. Similarly, care should be exercised
in wetlands to adequately target zones where channel
meanders approach the shoreline that contains the
ground-water contamination.
When deploying samplers in lakes or large
streams, consideration should be given to the depth of
the contaminant plume as it approaches the surface-
water body. Knowing this depth will help estimate the
probable distance of contaminant discharge from the
shoreline. In thick aquifers, there is often a deeper
flow system beneath shallow stream subsystems (Toth,
1963). Thus, contaminant discharge from a distant
source, which travels in the deep aquifer system, will
discharge into the surface-water body at a greater dis-
tance from the shoreline than nearby contaminants that
travel in a shallower flow system. Savoie and others
(2000) found that contaminants emanating from a
source approximately 1.5 miles upgradient from the
shoreline of a kettle pond discharged into the pond
100 to 350 ft offshore. A second previously unknown
plume was detected discharging into the pond at a
distance of 25 to 200 ft offshore (Savoie and others,
2000). Samplers deployed near the shore would not
have detected these plumes.
Similar considerations for placement of PVD
samplers should be given to deployment in streams. In
an ideal gaining stream with homogeneous bottom sed-
iment and similar ground-water hydraulic gradients
on both sides of the channel, ground water moving
beneath a particular shoreline typically discharges to
the stream closer to that shoreline than to the opposite
shoreline. In this case, VOCs detected near a particular
shoreline probably came from ground water derived in
the upgradient direction of that shoreline. The litho-
logic and hydrologic complexities of streams, however,
can create complex discharge pathways, sometimes
making it difficult to select optimum-sampler
placement sites and to identify contaminant-source
directions.
Temporal changes in the locations of discharge
areas also can affect concentrations of VOCs in dis-
charge and affect interpretation of data from PVD
samplers. A study of a small Coastal Plain stream in
South Carolina to which petroleum hydrocarbons were
16 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
discharging showed that the stream contained an
upstream gaining reach and a downstream losing reach.
The boundary between the two reaches migrated
upstream during periods of low ground-water levels
and streamflow and downstream during periods of high
ground-water levels and streamflow (Vroblesky and
others, 1997). Thus, PVD samplers placed near the
gaining/losing boundary in this stream can be expected
to intercept upward moving ground water during part
of the year and downward moving surface water during
a different part of the year. Additional temporal varia-
tions in contaminant discharge through a streambed
have been observed following nearby well construction
(Vroblesky and Robertson, 1996).
Because the VOC concentration within a PVD
sampler represents an equilibrium between the vapor
phase in the sampler and the adjacent aqueous solution,
changing aqueous-contaminant concentrations
produce a corresponding change in the vapor-phase
concentrations. If PVD samplers in a local area are
removed sequentially over time following an equilibra-
tion period, they can be used to track temporal changes
in the contaminant concentrations of discharging
ground water (Vroblesky and Robertson, 1996).
Because PVD samplers are sensitive to temporal
fluctuations, samplers for a particular sampling
event should all be collected sequentially within a
few hours of each other to obtain a "snapshot" of the
contaminant-discharge distribution.
The depth to which the samplers are installed
also may affect the results. The samplers should be
installed at or below the ground-water/surface-water
interface; however, the location of the interface
typically is difficult to delineate. It may be at the
sediment/water interface or at some depth below the
sediment. In some areas, the interface may shift as a
result of daily or seasonal fluctuations in river stage and
ground-water flow. Surface water may enter the sedi-
ment at the head of riffles and dropoffs and re-enter the
river at the upstream edge and base of pools (Vaux,
1968; Boulton, 1993). The surface water also can leave
the channel laterally and travel through the stream-
banks before eventually re-entering the channel down-
stream (Harvey and Bencala, 1993). The movement of
surface water into bed sediments is more pronounced
in high permeability sediment than in low permeability
sediment; therefore, PVD samplers buried in shallow
sandy horizons in these zones may intercept local
surface water rather than discharging ground water.
Contaminated ground water upwelling beneath these
zones may be diverted and discharge farther down-
stream (Conant, 2000). For practical reasons, PVD
samplers often are buried at a uniform depth of approx-
imately 0.5 to 1.5 ft, which may or may not be below
the ground-water/surface-water interface. It is impor-
tant, therefore, to consider the implications of subsur-
face streamflow when interpreting the PVD-sampler
data.
Furthermore, if the samplers are deployed in an
area of VOC-contaminated bottom sediment derived
from sediment transported in the stream, then the VOC
concentration in the PVD samplers may reflect contam-
inant concentrations in the sediment rather than in dis-
charging ground water. It is important, therefore, to
consider the possibility of stream transport of contami-
nated sediment when interpreting PVD data. In some
cases, this situation probably can be resolved by
deploying the samplers beneath such sediment.
An additional factor affecting data interpretation
and sampler deployment is the potential for removal
of the target compounds by micro-organisms in the
sediment. The large diversity of micro-organisms and
oxidation reduction conditions commonly found in
wetland sediments may lower contaminant concentra-
tions locally. Consequently, PVD samplers buried
beneath the organic-rich bed sediments may detect a
substantially higher concentration of VOCs than
samplers placed in the upper part of the organic-rich
sediment. In this situation PVD samplers can help
evaluate VOC loss over a particular interval.
QUALITY CONTROL AND
ASSURANCE
The primary purpose of most studies that use
PVD samplers is to determine or verify the presence of
VOCs. Relative concentrations of VOCs detected at the
site are also of interest. Variability and bias introduced
during sample collection, however, affects the interpre-
tation of the results. Confidence in the detections and
the relative concentrations of VOCs in samples col-
lected with PVD samplers can be evaluated by collect-
ing a series of quality-control (QC) samples, such as
duplicate, trip, and equipment-blank samples.
Quality Control and Assurance 17
-------
Duplicate or co-located samples provide infor-
mation needed to estimate the precision of concentra-
tion values affected by the combination of uncertainties
associated with field variability, sample processing, and
the analytical method. A duplicate PVD sampler con-
sists of two separate samplers deployed adjacent to
each other in the same hole. These samplers are typi-
cally held together side-by-side to the same surveyor
flag with nylon ties to ensure that the open end of the
vials are at the same depth. To account for sampler
variability, at least 10 percent of the samplers should be
duplicates. Examples from studies in New England
show that a VOC was detected in 1 of the duplicate
samples but not in the other in only about 2.5 percent of
the 437 duplicate samples. The relative percent differ-
ence (RPD) between VOC concentrations in the 83
duplicate samples where a VOC was detected in both
samples ranged from 0 to nearly 200 percent. About 75
percent of these RPDs, however, were less than 30 per-
cent, which is a reasonable range for a reconnaissance
tool. Duplicate samplers also can provide a backup in
case one of the samplers is compromised.
Trip blanks are PVD samplers that are prepared
offsite, typically with construction of all the samplers
expected to be used at the site. They are stored and
transported to the sampling location with the other
PVD samplers and capped at the sampling location
when the PVD samplers are deployed. The trip blanks
are then stored with other samples as they are recov-
ered, and analyzed with the recovered samples. A posi-
tive detection in the trip blank means that the PVD
samplers were exposed to specific contaminant(s)
sometime before deployment. To some extent, this
detection imparts a degree of uncertainty to the detec-
tions of that specific compound in the recovered PVD
samples. It should be noted, however, that the samplers
re-equilibrate to their surroundings. If background
samples do not contain the specific contaminant, then
it is highly probable that sufficient deployment time
elapsed to allow concentrations of the specific contami-
nant to re-equilibrate to ambient conditions in all of the
deployed samplers.
Some of the PVD samplers should be deployed
in an area of the surface-water body considered to be
away from potential VOC contamination, such as
upstream in rivers. If contaminants are found in
samplers from a target area, but not in the background
samplers, then this provides increased confidence
that the contaminants are not an artifact of the
methodology.
PART 2. EXAMPLE
APPLICATIONS IN
NEW ENGLAND
During 1996 through 2000, PVD samplers were
used at nine Superfund sites in New England to iden-
tify likely discharge areas for VOCs in ground water
(fig. 10). These sites were selected for study because
contamination of ground water by VOCs was known or
suspected. The sites represent a variety of hydrologic
settings including rivers, streams, ponds, wetlands, and
coastal shorelines (table 2). Samplers, all constructed
by methods described in Vroblesky and others (1996),
were placed in sediments ranging from clayey silt to
cobbles. Vapor concentrations in samplers ranged from
not detected to more than 1,000,000 parts per billion by
volume (ppb v). The principal VOCs detected include
the chlorinated compounds tetrachloroethene (PCE),
trichloroethene (TCE), and chlorobenzene, and the
petroleum compounds benzene, ethylbenzene, meta-
para-xylene, ortho-xylene, and toluene. At all nine
Superfund sites, discharge areas of known ground-
water plumes contaminated with VOCs were confirmed
and refined with PVD samplers. At four of these sites,
results of PVD sampling has lead to the identification
of previously unknown plumes of contaminated ground
water and has helped guide further characterization of
ground water at these sites. The following sections
briefly describe each of the nine study areas, state the
purpose and design of PVD sampling, present sam-
pling results on maps, and summarize findings. Also
included is a summary of the quality control and assur-
ance results for the nine studies. These summaries were
extracted from published reports. Additional detail
about any of these studies can be found in the cited
reports.
18 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
MAINE
Eastern Surplus Company
Superfund Site
CANADA I /
UNTIED STATES
McKin Company
Superfund Site
VERMONT
NEW HAMPSHIRE
Nyanza Chemical Waste
Dump Superfund Site
MASSACHUSETTS •
Baird & McGuire
Superfund Site
Nutmeg Valley Road
Superfund Site
CONNECTICUT
Centredale Manor Restoration
Project Superfund Site
Otis Air National Guard/Camp Edwards
Superfund Site (Johns Pond)
Calf Pasture Point and Allen Harbor Landfill,
Davisville Naval Construction Battalion
Center Superfund Site
Base from U.S. Geological Survey digital data, Albers Equal
Area Conic projection, 1992, Standard parallels 29°30' and
45°30', central meridian-96°,1:250,000 scale
0
100 MILES
100 KILOMETERS
Figure 10. Locations of sites in New England where passive-vapor-diffusion samplers have been used
to detect and delineate discharge areas of ground water contaminated by volatile organic compounds
into surface-water bodies.
Quality Control and Assurance
-------
Table 2. Superfund sites in New England where passive-vapor-diffusion samplers were used to detect and delineate volatile
organic compounds in bottom sediments of surface-water bodies, hydrologic setting, principal compounds detected, and
maximum vapor concentration measured
[Concentrations in parts per billion by volume. PCE, tetrachloroethene; TCE, trichloroethene; petroleum compounds, (benzene, toluene, ethylbenzene,
meta/para-xylene, and o/t/zo-xylene)]
Principal
Maximum vapor
Site
Hydrologic setting
compound(s)
concentration
detected
measured
Eastern Surplus Company Superfund Site, Meddybemps, Maine
Lake and river
PCE
240
TCE
70
Toluene1
2,500
McKin Company Superfund Site, Gray, Maine
River
TCE
30,400
Nutmeg Valley Road Superfund Site, Wolcott and Waterbury, Connecticut
River
TCE
> 30,000
Baird & McGuire Superfund Site, Holbrook, Massachusetts
River
Petroleum
216,000
compounds
TCE + PCE
1,900
Allen Harbor Landfill, Davisville Naval Construction Battalion Center
Coastal shoreline
TCE
340,000
Superfund Site, North Kingstown, Rhode Island
PCE
1,700
Benzene
940
Calf Pasture Point, Davisville Naval Construction Battalion Center
Coastal shoreline and
TCE (shoreline)
1,900
Superfund Site, North Kingstown, Rhode Island
wetland area
TCE (wetland)
14
Otis Air National Guard/Camp Edwards Superfund Site, Johns Pond,
Pond
TCE
47,000
Falmouth, Massachusetts
PCE
667
Nyanza Chemical Waste Dump Superfund Site, Ashland, Massachusetts
River, former mill
Chlorobenzene
5,330
raceway, and pond
TCE
1,910
Centredale Manor Restoration Project Superfund Site, North Providence,
River and former mill
PCE
1,390,000
Rhode Island
raceway
TCE
182,000
*May have been derived from adhesive tape used to secure polyethylene membrane to glass vial.
EASTERN SURPLUS
COMPANY SUPERFUND SITE,
MEDDYBEMPS, MAINE
By Forest P. Lyford and
Edward M. Hathaway
Description of Study Area
The Eastern Surplus Company in Meddybemps,
Maine, was a retailer of surplus and salvage items from
1946 until 1985. Activities at the site caused the release
of chemicals, including VOCs and polychlorinated
biphenyls (PCBs), to the environment (E.M. Hathaway,
U.S. Environmental Protection Agency, written
commun., 1996). The Eastern Surplus Company
Superfund site (Eastern Surplus site) covers about 4
acres adjacent to Meddybemps Lake and the Dennys
River, which flows from Meddybemps Lake (fig. 11).
Surficial materials in an approximately 30-acre study
area that encompasses the site vary widely in texture.
Materials include till at the northern end and on the
eastern side of the Dennys River, coarse sand and
gravel on the western side, silt deposited in a marine
environment in the central part of the site, and marine
clay over coarse materials and till on the southern side.
Surficial materials are generally less than 30 ft thick. In
some areas, these materials lie above the water table for
all or much of the time. Ground water in surficial mate-
rials under the site generally flows toward the Dennys
River (fig. 12). Bedrock includes diorite and granite
where ground water moves in fractures. Ground water
in bedrock under the site generally flows eastward
toward Meddybemps Lake at the northern end of the
site and southward to eastward toward the Dennys
River elsewhere (fig. 12) (Lyford and others, 1998).
Some ground water flows toward a depression in the
bedrock potentiometric surface caused by pumping for
residential use on the southeastern side of the study
area.
20 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
67°22'30"
67°20'
'feet
• '' ' "' I, 1997
M
45°02'30l
s Site
W
1 MILE
1 BOLOMETER
CONTOUR INTERVAL IS 10 FEET
NATIONAL GEODETIC VERTICAL
DATUM OF 1929
Base from U.S. Geological Survey
Meddybemps Lake East and Meddybemps Lake West
quadrangles, 1:24,000, provisional edition 1987
Map modified from Lyford and others, 1998
MAINE
0
h
100 MILES
_l
0 100 KILOMETERS
Figure 11. Location of the Eastern Surplus Superfund Site and study area, Meddybemps, Maine.
Eastern Surplus Company Superfund Site, Meddybumps, Maine 21
-------
Gravel Pit
Top of Sic
45°02'21"
45°02'15"
Base map modified from OEST Assoc. Inc., 1997
EXPLANATION
67°21'35"
67°21'25"
EXPLANATION OF SYMBOLS USED
ON OEST BASE MAP
BUILDING
TREELINE
MEDDYBEMPS LAKE
200 FEET
60 METERS _
TOPOGRAPHIC CONTOUR—Interval 5 feet.
Datum is sea level.
-170 WATER-LEVEL CONTOUR IN SURFICIAL AQUIFER—
Shows altitude at which water would have stood in tightly
cased wells completed in the surficial aquifer, April 30,1997.
Dashed where approximately located.
Contour interval is 2 feet. Datum is sea level.
(Lyford and others,1998)
_i70 WATER-LEVEL CONTOUR IN BEDROCK—Shows
altitude at which water would have stood in tightly cased
wells completed in bedrock, April 30, 1997. Dashed
where approximately located. Contour interval is 5 feet.
EXTENT OF SURFICIAL AQUIFER—Marks
western and northern extent of area where
saturated surficial materials are generally
thicker than 2 feet.
GENERAL DIRECTION OF WATER FLOW
Surficial aquifer
Bedrock aquifer
MONITORING WELL
Figure 12. Potentiometric surfaces and generalized ground-water-flow directions for the surficial and bedrock
aquifers, Eastern Surplus Superfund Site, Meddybemps, Maine, April 30, 1997.
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Purpose and Design of
Sampling
Sampling of four monitoring wells in 1988 iden-
tified one area at the northern end of the site, near well
MW-3B, where PCE and TCE were present in ground
water in fractured bedrock (E.M. Hathaway, written
commun., 1996) (fig. 13). Ground-water sampling
points were limited, and other areas of ground-water
contamination were possible. During early phases of a
study by the USGS in 1996-97, PVD samplers were
placed along the shore of Meddybemps Lake and the
west bank of the Dennys River to determine if other
VOC plumes were present and discharging to surface
water (Lyford and others, 1998).
To our knowledge, the use of PVD samplers at
the Eastern Surplus site was the first application in
New England. Sampler construction included the use
of adhesive tape to hold the polyethylene membranes
firmly to bottles. Samplers were placed about 25 to
50 ft apart along most of the shoreline and river bank
(fig. 13). Very coarse materials consisting of cobbles
and boulders along the shoreline and river bank pre-
cluded the installation of samplers in some areas.
During this early attempt to use PVD samplers, the
most effective method found for installation in the
coarse materials was by manual insertion in a hole
formed behind a shovel driven into bottom materials
and forced forward. The PVD samplers were in place
for about one month before retrieval and on-site
analysis of vapors.
Results
The VOCs detected in PVD samplers included
PCE, TCE, and toluene. The compounds PCE and TCE
were detected near well MW-3B, where they had been
detected in ground water (fig. 13). The sample from
one sampler placed along the Dennys River south of
the site also contained PCE and TCE. Subsequent
installation of monitoring wells and ground-water
sampling in this area identified a previously unknown
plume of VOCs, mainly PCE, in surficial materials and
bedrock. The plume appeared to originate at the south-
ern end of the site (Lyford and others, 1998) (fig. 13).
The extent of this plume, shown on figure 13, is based
on water samples from wells that were installed after
the survey with PVD samplers.
Toluene was detected in several PVD samplers
but not in water from monitoring wells. The toluene
may have been derived from the adhesive tape used to
secure the surveyor flag and tubing to the glass vials.
Similar occurrences of toluene in diffusion samplers
wrapped with adhesive tape were described by
Mullaney and others (1999).
Eastern Surplus Company Superfund Site, Meddybumps, Maine 23
-------
67°21l35" 67*21 *25-
EXPLANATION OF SYMBOLS USED
ON OEST BASE MAP
170 TOPOGRAPHIC CONTOUR—
Interval 5 feet. Datum is
sea level
MEDDYBEMPS LAKE
45°02'21
~ BUILDING
TREELINE
cr
/Top of Slop<
V
200 FEET
60 METERS
45°02'15l
Base map modified from OEST Assoc. Inc., 1997
EXPLANATION
— — — EXTENT OF TETRACHLOROETHENE AND TRICHLOROETHENE IN GROUND
WATER—Shows the extent of concentrations greater than 5 micrograms per liter.
PASSIVE-VAPOR-DIFFUSION SAMPLERS—Shows concentrations of tetrachloroethene and
trichloroethene in parts per billion by volume (ppb v).
O Greater than 100 to 1,000
• Trace to 100
O Not detected above reporting limit of 40 ppb v or at trace level
• MONITORING WELL
Figure 13. Concentrations of tetrachloroethene (PCE) in passive-vapor-diffusion samplers installed in river-
bottom sediments on the western edge of Dennys River, Meddybemps, Maine, October 1996.
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
McKIN COMPANY SUPERFUND
SITE, GRAY, MAINE
By Forest P. Lyford, Terrence R.
Connelly, and Laura E. Flight
Description of Study Area
The McKin Company Superfund Site (Mckin
site) in Gray, Maine, is a former waste-collection,
transfer, and disposal facility that operated from 1965
to 1978. Some of the wastes, including petroleum
hydrocarbons and chlorinated solvents, infiltrated and
contaminated ground water. A contaminant plume in
ground water, consisting mainly of TCE, extends
northward and eastward from the site for several thou-
sand feet (fig. 14). The eastern plume discharges to the
Royal River in an area known locally as Boiling
Springs (Sevee and Maher Engineers, Inc., 1999).
Since at least 1993, concentrations of TCE in most
monitoring wells have been declining gradually by
natural attenuation.
In the Superfund area, TCE in ground water
is present in surficial materials and fractured crystalline
rock. The surficial materials consist mainly of
glacially-derived sand and gravel overlain by a thick
layer of glaciomarine clay in some areas (Lyford and
others, 1999a). The marine clays are absent under the
McKin site, but form a confining layer for ground
water in buried coarse materials under part of the area
between the site and the Royal River. In the area near
Boiling Springs, the Royal River has eroded through
the marine clays and exposed underlying coarse-
grained materials. This area of exposed coarse-grained
materials is the principal discharge area for the ground-
water system to the north and west (fig. 14). Water
samples from monitoring wells also indicate that the
discharge of contaminants is focused in the area near
Boiling Springs (Lyford and others, 1999a). Data from
the monitoring wells also indicate that the TCE plume
in fractured granite is coincident with the plume in
overlying surficial materials. Samples were collected
with PVD samplers during a period of low flow in the
Royal River. Historical water-quality data indicate that
concentrations of TCE in the river near the railroad
trestle typically range from 15 to 20 |ig/L for river
discharges at the time of the study.
Purpose and Design of
Sampling
The quantity of TCE that discharges to the Royal
River, at river flows generally less than 100 ft3/s, is
great enough to cause concentrations in the river to
exceed the State of Maine's water-quality standard
for streams of 2.7 |ig/L (RR. Jaffe, Princeton
University, written commun., 1996). In 1996, the
Maine Department of Environmental Protection
(MEDEP) and the USEPA sought, from the potentially
responsible parties (PRPs) for the McKin site, an eval-
uation of remediation methods to reduce discharge of
TCE into the river. The PRPs estimated that a 1,500-ft
long interception system would be needed to success-
fully capture the TCE plume. Evaluation of remedia-
tion strategies and selection of a remediation program
required an understanding of the configuration of
the TCE plume near the river and the distribution of
TCE concentrations across the width of the plume.
Consequently, PVD samplers were installed along the
Royal River and an unnamed tributary in the autumn of
1997 to determine the width of the TCE plume at the
Royal River and variations in concentrations of TCE
across the width of the plume (Lyford and others,
1999a) (fig. 15). In addition to these two goals, it was
anticipated that results of the study would improve the
understanding of contaminant pathways near the river.
Approximately 150 PVD samplers were placed
along the banks of the Royal River and in several
transects across the width of the river over a 3-day
period in September 1997. Samplers also were placed
along an unnamed tributary stream of the Royal River,
in Boiling Springs, and in a seepage area on the north
side of the river downstream from a river bend. After
about 2 weeks, the samplers were retrieved and sam-
ples were analyzed on site with a portable gas chro-
matograph. The PVD samplers were retrieved at about
the rate that they were analyzed in the field laboratory
(about 50 per day). No samplers were lost during the
study.
McKin Company Superfund Site, Gray, Maine 25
-------
70°17'30" 70°16'30"
43o53,30"
43°53'
• MONITORING WELL 1 , 100 miles
I 1 T 1
0 100 KILOMETERS
Figure 14. Location of the McKin Superfund Site and study area, potentiometric surface contours for the surficial aquifer, and
extent of trichloroethene in ground water, Gray, Maine.
MAINE
EXPLANATION
120 WATER-LEVEL CONTOUR IN SURFICIAL AQUIFER— Shows altitude
at which water would have stood in tightly cased wells completed
in surficial aquifer in May 1998. Contour interval is variable.
Datum is sea level. (Data from Sevee and Maher Engineers, Inc., 1998)
EXTENT OF TRICHLOROETHENE IN SURFICIAL AQUIFER— Shows
the extent of concentrations greater than 2 micrograms per liter.
(Data from Sevee and Maher Engineers, Inc., 1998)
Base from U.S. Geological Survey digital
line graphs from Gray quadrangle, 1:24,000
0 1,000 FEET
1 1 i
0 300 METERS
Contour interval 10 feet
National Geodetic Vertical Datum of 1929
Map modified from Sevee and Maher Engineering, Inc.,
1998, and Lyford and others, 1999a
26 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
70°17' 70°16'40"
200 FEET
Base from U.S. Geological Survey digital
line graphs from Gray quadrangle, 1:24,000
Map modified from Lyford and others, 1999a
50 METERS
CONTOUR INTERVAL 10 FEET
NATIONAL GEODETIC VERTICAL DATUM OF 1929
EXPLANATION
— — — EXTENT OF TRICHLOROETHENE IN SURFICIAL AQUIFER—Shows
the extent of concentrations greater than 2 micrograms per liter.
(Data from Sevee and Maher Engineers, Inc., 1998)
© PASSIVE-VAPOR-DIFFUSION SAMPLERS
Figure 15. Locations of passive-vapor-diffusion samplers installed in river-bottom sediment along and near the Royal
River in September and October 1997, and extent of trichloroethene in ground water, Gray, Maine.
McKin Company Superfund Site, Gray, Maine 27
-------
Results
The compound TCE was detected in most PVD
samplers placed downstream of sampler 14R (fig. 16).
The extent of the plume, as determined with PVD sam-
plers, was generally consistent with plume maps drawn
on the basis of water samples from wells (Sevee and
Maher Engineers, Inc., 1998) (tig. 15). The highest
concentrations, however, were focused in an area
where sand boils were apparent in the riverbed a few
hundred feet downstream from Boiling Springs, indi-
cating a major discharge area for ground water and
contaminants (tig. 16). In this general area, concentra-
tions of TCE also were detected in samplers placed in
the center of the river and on the opposite side, indicat-
ing that contaminated ground water was discharging
across the width of the river and probably flowing
beyond the river and floodplain downstream to points
between the sharp easterly bend in the river and the
railroad trestle (fig. 16). The VOCs detected down-
stream of the railroad trestle in an area outside of the
mapped extent of the plume in ground water may result
from the exchange of contaminated surface water with
water in the bottom sediments. Hydraulic head data for
the sediments and TCE concentrations for the river at
the time of sampler retrieval would have been useful to
test this possibility.
NUTMEG VALLEY ROAD
SUPERFUND SITE,
WOLCOTT AND WATERBURY,
CONNECTICUT
By John R. Mullaney, Peter E.
Church, and Carolyn J. Pina-Springer
Description of Study Area
Ground-water contamination by VOCs was
discovered in the 1980s in the Nutmeg Valley area,
Wolcott and Waterbury, Connecticut, and the area
was classified by the USEPA as a Superfund site in
1989 (U.S. Environmental Protection Agency, 1989)
(fig. 17), where approximately 43 industries and 25
residences use ground water, primarily from the bed-
rock aquifer, for industrial and domestic supply. Past
disposal of industrial chemicals has been implicated in
contamination of water from supply wells sampled by
local, State, and Federal agencies during 1979-95.
Contaminants may also be contributed to ground water
from the City of Waterbury landfill (the North End
Disposal Area), located about 1/2 mi upgradient from
the Nutmeg Valley Road Superfund Site (fig. 17). The
VOCs most commonly detected in supply wells tap-
ping the crystalline-bedrock aquifer included TCE,
PCE, and 1,1,1-trichloroethane (TCA). Concentrations
of TCE were as high as 320 |ig/L in samples collected
from supply wells in 1985 (Mullaney and others,
1999).
Two principal aquifers underlie the Nutmeg
Valley study area—an unconsolidated surficial aquifer
consisting of glacial till, glacial stratified deposits, and
postglacial alluvium, and a fractured crystalline bed-
rock aquifer consisting of well-foliated gneiss and gra-
nofels. Glacial till overlies the bedrock and is at land
surface in most upland parts of the study area. Till is
generally less than 10 ft thick; locally, however, it is
more than 25 ft thick. Glacial derived stratified deposits
in the valley, consisting of poor to well sorted layers of
gravel, sand, silt and clay, range in thickness from 0 to
85 ft over till. A semi-confining layer of fine-grained
deposits within the stratified deposits, 5 to 10 ft thick,
overlies silty, sand and gravel in an area beneath the
lower reaches of the unnamed tributary and Old
Tannery Brook, and short reaches of the Mad River
upstream and downstream of the confluence with Old
Tannery Brook. Postglacial alluvial and swamp depos-
its are generally less than 10 ft thick and overlie glacial
stratified deposits on the floodplain surfaces of the Mad
River and Old Tannery Brook. These streams have
incised deeply into glacial stratified deposits during
postglacial time. The texture of the alluvium beneath
the floodplain ranges widely from gravelly sand depos-
ited in former stream-channel positions to fine sand and
silt with significant amounts of organic material in
overbank deposits laid down during floods (Mullaney
and others, 1999).
28 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
70°16'50"
70°16'45"
14R
Boiling
S{
43°53'10"
O O
ooo
moo
• • •
O o o
•oooofco •<,<*>
•QOO° °
Railroad
trestle
43°53'05"
Unnamed Tributary
O
100
Map modified from Lyford and others, 1999a
EXPLANATION
PASSIVE -VAPOR-DIFFUSION SAMPLERS—Shows concentrations
of trichloroethene in parts per billion by volume (ppb v).
• Greater than 10,000
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
• Trace to 100
O Not detected above reporting limit of 8 ppb v or at trace level
Figure 16. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in river-bottom
sediments near Boiling Springs, Gray, Maine, September and October 1997.
Base from U.S. Geological Survey digital
line graphs from Gray quadrangle, 1:24,000
0 200 FEET
o 60 METERS
CONTOUR INTERVAL 10 FEET
NATIONAL GEODETIC VERTICAL DATUM OF 1929
Nutmeg Valley Road Superfund Site, Wolcott and Waterbury, Connecticut
-------
73°01" 72°59'
Chestnut /
Hill Reservoirf jj|
twn ¦' \ pm
— "V,: ai* V -*>V -'/L--
L_ uik. WS f|,/V^I' ; : '
¦# i' •
5 v /N <-y $0 ¦+-"* ' - s
V, \ ^estnut'
Scoville Reservoir
f.^yu^sp
i' S vj. i Tpwer'l
CONNECTICUT
EXPLANATION
STUDY-AREA BOUNDARY
SUPERFUND-SITE BOUNDARY
NORTH END DISPOSAL
AREA BOUNDARY
Base from U.S. Geological Survey quadrangles Waterbury and Southington, Map modified from Mullaney and others, 1999
1:24,000,1968. Photorevised 1984, Digital Raster Graphics Files,
Projection State Plane Feet, Zone 3526
0 1 MILE
1 1 « 1
0 1 KILOMETER
CONTOUR INTERVAL IS 10 FEET
NATIONAL GEODETIC VERTICAL DATUM OF 1929
41°36'
41°34'
long'1
0 SO MILES
50 KILOMETERS
Figure 17. Location of the Nutmeg Valley Road Superfurid Site arid study area, Nutmeg Valley, Wolcott arid
Waterbury, Connecticut.
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Purpose and Design of Sampling
In 1997, the USGS, in cooperation with the
Town of Wolcott and the USEPA, used PVD samplers
as a reconnaissance tool for detecting and delineating
ground-water discharge of VOCs into the local streams:
Mad River, Old Tannery Brook, and an unnamed tribu-
tary to Old Tannery Brook (Mullaney and others,
1999). Samplers were installed in the river-bottom
sediments at 154 sites along the Mad River, Old
Tannery Brook, and an unnamed tributary of Old
Tannery Brook on May 12-27, 1997, and were
retrieved on July 8-10, 1997. Samplers were placed at
100- to 200-ft intervals in the center of the streams and
in transects across the stream at selected locations
(tig. 18). Because the PVD-sampler technology was
new, a second round of sampling was done for compar-
ison with results obtained from the first round of sam-
pling. On October 23-28, 1997, 128 PVD samplers
were installed at locations similar to those in the first
sampling round. These samplers were retrieved on
November 11-13, 1997 (fig. 19). The USEPA analyzed
both sets of samples on site with a gas chromatograph
calibrated for measurement and identification of TCE,
PCE, and petroleum compounds (Mullaney and others,
1999). In addition, vertical head gradients between the
ground water and surface water were measured in
November 1997 at 30 locations along Old Tannery
Brook, the unnamed tributary to Old Tannery Brook,
and the Mad River to determine if and where ground
water was discharging to the streams.
Results
Results from the first round of sampling show
that the highest concentrations and most frequent
detections of VOCs were in zones along the lower
reach of Old Tannery Brook and in the Mad River at
short distances upstream and downstream from the
confluence with Old Tannery Brook (fig. 18). Concen-
trations of TCE ranged from not detected (less than
5 ppb v) to 4,800 ppb v. Concentrations of PCE ranged
from not detected (less than 5 ppb v) to 781 ppb v.
Adhesive tape used to secure the samplers to the sur-
veyor flags in the first round of sampling was found to
contain petroleum compounds; therefore, detections of
these compounds in PVD samples were not reported.
The second round of samplers were retrieved in
November 1997 to determine if the July 1997 results
were reproducible. Instead of tape, nylon ties were
used to secure surveyor flags to the samplers, to avoid
problems with contamination by petroleum com-
pounds. The spatial pattern of VOCs detected was sim-
ilar to the July 1997 detection pattern (fig. 19), but with
fewer detections of PCE and with additional locations
where trace-levels of cis-DCE were detected. This dif-
ference was due to a small difference in the calibration
of the gas chromatograph between the two sampling
rounds. Concentrations of TCE in vapor ranged from
not detected (less than 25 ppb v) to greater than
30,000 ppb v. Concentrations of PCE ranged from not
detected (less than 25 ppb v) to 390 ppb v. Concentra-
tions of benzene ranged from not detected (less than
25 ppb v) to 51 ppb v.
In both sampling rounds, the highest vapor con-
centrations were detected along the lower reach of Old
Tannery Brook near a known contaminated area on the
western side of the brook that contains primarily TCE,
PCE, vinyl chloride, and cis-DCE in soils and ground
water (Loureiro Engineering Associates, 1998a, b).
Ground-water contaminated by TCE also has been doc-
umented on the eastern side of Old Tannery Brook
(HRP Associates, 1991). This area is underlain by the
fine-grained deposits (semi-confining lacustrine
deposit), which suggests that VOCs discharged to the
brook are from contaminated ground water in the post-
glacial alluvium. The high vapor concentrations of
TCE detected along the Mad River may be from
ground-water-contaminant plumes in surficial deposits
beneath the fine-grained layer, or from the fractured
bedrock, or both.
In both sampling rounds, vapor concentrations of
TCE and PCE were detected in the unnamed tributary,
but were lower than those detected in the Old Tannery
Brook and the Mad River (figs. 18 and 19). In the first
sampling round, the highest TCE and PCE vapor con-
centrations were detected 1,200 ft upstream of the Old
Tannery Brook; concentration of TCE was 73 ppb v,
and PCE was 348 ppb v. The highest vapor concentra-
tions in the second sampling round were detected
1,600 ft upstream of Old Tannery Brook; concentration
of TCE was 104 ppb v and PCE was 101 ppb v.
Variations in vapor concentrations across stream
channels were observed at sites where PVD samplers
were installed at the edges and in the center of the
channel. These variations are probably due to the direc-
tion from which the VOCs originate, and also may
be caused by variations in organic matter, biotic
and abiotic processes, and streambed-hydraulic
conductivity.
Nutmeg Valley Road Superfund Site, Wolcott and Waterbury, Connecticut 31
-------
73°00'30"
72o59'30"
41034'45"
41°34'00"
aJ-
Swiss
Tosun
Base from U.S. Army Corps of Engineers, 1993,1997, Scale 1:2,400 0
Projection: State Plane Feet, Zone 3526 |—
1,000 FEET Map modified from Mullaney and others, 1999
T
300 METERS
EXPLANATION
PASSIVE-VAPOR-DIFFUSION SAMPLERS-Shows
concentrations of trichloroethene and
tetrachloroethene in parts per billion by volume
(ppb v), July 1997.
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
• Trace to 100
O Not detected above reporting limit of 5 ppb v or
at trace level
IS Sampler not found
MONITORING WELLS—Shows concentrations of
trichloroethene and tetrachloroethene in ground water in
micrograms per liter (ug/L). Wells sampled May to
September, 1998 (Mullaney and others, 1999)
® Trace to 100 fXg/L
• Not detected above reporting limit of 1 p.g/L or at
trace level
Figure 18. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers installed in river-
bottom sediments of the Mad River, Old Tannery Brook, and an unnamed stream, Nutmeg Valley, Wolcott and Waterbury,
Connecticut, July 1997.
32 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
73o00'30"
72o59'30"
41°34I45M
41°34l00" -
%
Tosun
Base from U.S. Army Corps of Engineers, 1993,1997, Scale 1:2,400 0
Projection: State Plane Feet, Zone 3526 |—
1,000 FEET Map modified from Mullaney and others, 1999
—r
300 METERS
EXPLANATION
PASSIVE-VAPOR-DIFFUSION SAMPLERS—Shows
concentrations of trichloroethene and
tetrachloroethene in parts per billion by volume
(ppb v), November 1997.
Greater than 10,000
Greater than 1,000 to 10,000
Greater than 100 to 1,000
Trace to 100
Not detected above reporting limit of 25 ppb v or
at trace level
MONITORING WELLS—Shows concentrations of
trichloroethene and tetrachloroethene in ground water in
micrograms per liter ( ng/L). Wells sampled May to
September 1998 (Mullaney and others, 1999)
® Trace to 100 ng/L
• Not detected above reporting limit of 1 (ig/L or at
trace level
® Sampler not found
Figure 19. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers installed in river-
bottom sediments of the Mad River, Old Tannery Brook, and an unnamed stream, Nutmeg Valley, Wolcott and Waterbury,
Connecticut, November 1997.
Nutmeg Valley Road Superfund Site, Wolcott and Waterbury, Connecticut 33
-------
BAIRD & McGUIRE
SUPERFUND SITE,
HOLBROOK, MASSACHUSETTS
By Jennifer G. Savoie and
Melissa G. Taylor
Description of Study Area
A ground-water plume containing volatile and
semi-volatile organic compounds at the Baird &
McGuire Superfund Site in Holbrook, Massachusetts,
extends eastward from a former chemical-processing
plant toward and beneath the Cochato River (tig. 20)
(M.G. Taylor, Environmental Protection Agency,
written commun., 1998). The Cochato River once sup-
plied water to the towns of Holbrook, Randolph, and
Braintree, but use of this source ended after contamina-
tion near the river was discovered in 1983. In 1993, a
ground-water-extraction system began operation to
remove contaminants from a sand and gravel aquifer
below the site and the river and to limit the discharge of
contaminants to the river. From 1995 to 1997, contami-
nated sediments were excavated from the river and
incinerated as part of site remediation. Despite these
remedial actions, the USEPA and residents are con-
cerned that contaminants from the ground-water plume
could discharge to the river (Savoie and others, 1999).
Purpose and Design of
Sampling
In March and April 1998, a network of PVD
samplers was installed along the Cochato River to
determine if VOC-contaminated ground water was dis-
charging through the river-bottom sediments while a
ground-water-extraction system was operating and
after the system had been shut down for 2 weeks
(fig. 21). Drive-point piezometers were installed at four
locations within the riverbed of the Cochato River near
the known extent of the ground-water plume. Water
levels from piezometers were compared to river-stage
measurements to determine if the river was gaining
ground water across the study area and if contaminants
could potentially discharge into the river (Savoie and
others, 1999).
Results
Under pumping and non-pumping conditions,
petroleum compounds (benzene, ethylbenzene,
meta/para-xylene, ortho-xylene, and toluene) were
detected in PVD samplers where the plume passes
beneath the river (fig. 21, showing concentrations
under pumping conditions). Concentrations of total
petroleum compounds ranged from not detected
upriver of plume area, but downriver adjacent to
the plume area concentrations were greater than
200,000 ppb v. Under pumping and non-pumping
conditions, concentrations did not differ significantly.
The compounds TCE, PCE, and cis-DCE also were
detected in PVD samplers more than 200 ft down-
stream of the area where the petroleum compounds
were detected. These detections indicate a different
source for TCE + PCE than for the petroleum
compounds (fig. 21).
Water levels in four piezometers were consis-
tently higher than the river stage, which indicates an
upward hydraulic gradient and ground-water discharge
to the river. This observation in the piezometers and
the presence of contaminants in the pore water of river-
bottom sediments indicate that contaminants from
the Baird & McGuire Superfund Site ground-water
plume were discharging into the Cochato River at the
time of this study for both pumping and non-pumping
conditions.
34 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
MASSACHUSETTS
50 MILES
50 KILOMETERS
71°00'
BRAINTREE Cr^erry
Pond
[OUTH
RANDOLPH
'reat
'ond
iylvan Lake
HOLBROOK
Location of study area near
Baird & McGuire Superfund Site.
Area shown in figure 21.
Holbrl
AVON
42°07'30'
ABINGTON
Trout \
¦leayeland
BROCKTON
o
2 MILES
Map modified from Savoie and others, 1999
0 2 KILOMETERS
Figure 20. Location of the Baird & McGuire Superfund Site and study area, Holbrook,
Massachusetts.
Baird & McGuire Superfund Site, Holbrook, Massachusetts 35
-------
42°09'10"
42°09' -
EXPLANATION
AREA OF EQUAL CONCENTRATION OF PETROLEUM
COMPOUNDS IN GROUND WATER, IN
MICROGRAMS PER LITER (jig/L)—July through
September 1997 (J.M. Osborn, Metcalf & Eddy, Inc.,
written commun., 1998)
¦ Greater than 1,000
Greater than 100 to 1,000
Greater than 10 to 100
5 to 10
PASSIVE-VAPOR-DIFFUSION SAMPLERS—Shows
concentration of petroleum compounds in southern part of
site and trichloroethene and tetrachloroethene
(TCE and PCE) in northern part of site, in parts per billion
by volume (ppb v) for pumping conditions
• Greater than 10,000
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
• Trace to 100
O Not detected above reporting limits of 100 ppb v for
petroleum compounds and 20 ppb v for TCE and PCE
or at trace levels
<8 Sampler not found
0 Drive-point piezometer
• MONITORING WELLS—(E designates extraction well)
•—FENCE LINE ON SUPERFUND SITE BOUNDARY
^5 BUILDING FOOTPRINT
71o01'25"
71°0r35"
Base map derived from AutoCAD file provided by q 200 FEET Map modified from
Metcalf & Eddy, Inc., Original Scale 1:500 I 1 t Savoie and others, 1999
Diffusion samplers mapped by U.S. Geological Survey 0 60 meters
using Global Positioning System and
Arclnfo Geographic Information System
Figure 21. Concentrations of trichloroethene plus tetrachloroethene and petroleum compounds in
passive-vapor-diffusion samplers installed in river-bottom sediments of the Cochato River, Baird &
McGuire Superfund Site, Holbrook, Massachusetts, March and April 1998.
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New
-------
ALLEN HARBOR LANDFILL,
DAVISVILLE NAVAL CONSTRUCTION
BATTALION CENTER SUPERFUND
SITE, NORTH KINGSTOWN,
RHODE ISLAND
By Forest P. Lyford, William C.
Brandon, and Christine A.P. Williams
Description of Study Area
The Allen Harbor Landfill at the former
Davisville Naval Construction Battalion Center in
North Kingstown, Rhode Island (tig. 22), was used by
the U.S. Navy from 1946 until 1974 to dispose of
wastes, including municipal-type waste, construction
debris, paint thinners, degreasers, sewerage sludge, and
fuel oil (EA Engineering, Science, and Technology,
1996). This approximately 15-acre landfill on the
west side of Allen Harbor is bordered on the west by
Sanford Road and a large wetland, and is bordered
on the north and south by small vegetated wetlands
(fig. 23).
Landfill wastes are up to 20 ft thick. Geologic
materials beneath landfill wastes and the shoreline
include a discontinuous layer of fine to very fine sand,
generally less than 15 ft thick, over a layer of silt that is
20 to 50 ft thick. Peat layers are in some locations at
the top of the sand layer. The silt layer overlies a dis-
continuous till layer or bedrock. The altitude of the
bottom of the silt layer ranges from 30 to 50 ft below
sea level (EA Engineering, Science, and Technology,
1996). Monitoring wells, some in clusters, are com-
pleted in the upper sand layer (S following the well
number indicates a shallow screen depth), the silt layer
(I following the well number indicates a intermediate
screen depth), and till (D following the well number
indicates a deep screen depth) (fig. 23).
Shallow ground water in landfill wastes and the
upper sand layer flows eastward from a water-table
mound centered near well MW09-18I (fig. 23) to the
shore and southward toward a mudflat area. During the
wet season, shallow ground water also flows westward
toward a wetland on the west side of Sanford Road.
The mound is not apparent in water levels from wells
screened at greater depths near and below the bottom
of the silt layer. Ground-water flow at depth is predom-
inantly eastward and southeastward toward Allen
Harbor (fig. 23). Vertical hydraulic gradients are down-
ward in the area of the water-table mound and upward
near the shore. Gradients reverse in some wells near
the shore during high tide, but this reversal has not
been consistently observed for all tidal cycles (EA
Engineering, Science, and Technology, 1996).
A variety of VOCs, including petroleum and
chlorinated compounds have been detected in samples
from several monitoring wells within the landfill area
(fig. 23) (EA Engineering, Science, and Technology,
1996). Concentrations of VOCs also have been
detected in water from borings installed in sediments
offshore in the harbor (EA Engineering, Science, and
Technology, 1998a).
Purpose and Design of
Sampling
The high concentrations of VOCs in samples
from monitoring wells prompted the use of PVD
samplers to identify potential discharges of VOC-
contaminated ground water along the shore. Because
PVD samplers had not been used previously in a
coastal setting, a secondary goal of this study and a
companion study in nearby Calf Pasture Point (fig. 22)
was to determine if PVD samplers can yield useful
information about discharge points of VOCs along a
tidally affected shoreline (Lyford and others, 1999b).
The PVD samplers were installed during March
16 through 20, 1998, at locations shown in figure 23
and retrieved on April 1 through 2, 1998. Most sam-
plers (79) were installed during low tide at intervals of
about 25 ft along about 1,700 ft of shoreline. Samplers
also were placed at the high-tide level at 20 locations
for comparison to results from low-tide locations, in 12
seeps where ground water was apparently discharging
near the base of the landfill, and at 4 locations on
mudflats south of the landfill.
Allen Harbor Landfill, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island 37
-------
41°37'30"
41°37'00"
RHODE
ISLAND
STUDY
AREA>
10 MILES
0 10 KILOMETERS
"alf Pasture
Point
Wetland
Map modified from Lyford and others, 1999b
Photographic base (composite) derived from
U.S. Geological Survey digital orthophotos of portions of
Wickford and East Greenwich quadrangles, 1:5,000,
UTM, North American Datum, 1983
200 METERS
Figure 22. Locations of the Allen Harbor Landfill and Calf Pasture Point study areas, Davisville
Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island.
38 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
71 °25'15" 71 °25'00"
B+
ALLEN
HARBOR
MW09-11S
Jetty
41o37'30"
Approximate extent
of wetland area
MW09-18I
MW09-07D
MW09-14D
Allen Harbor Landfill
B+
MW09-09D
H B+
MW09-21D
MW09-201
H CD" "2
GfQ
MO
Approximate outer extent
of mudflat area
200 FEET
60 METERS
Base traced from U.S. Geological Survey digital orthophoto, 1:12,000,
GPS from Trimble Asset Surveyor, 1927 North American Datum.
Map modified from Lyford and others, 1999b
EXPLANATION
APPROXIMATE EXTENT OF LANDFILL
O
O
o
o
o
PASSIVE-VAPOR-DIFFUSION SAMPLERS INSTALLED
IN LOW-TIDE ZONE OR HIGH-TIDE ZONE (H),
NEAR SEEP (S), OR ON MUDFLAT (M)—Shows
concentrations of trichloroethene (TCE) and
tetrachloroethene (PCE) in parts per billion by volume
(ppb v); B, if only benzene is detected; B+, if benzene is
detected with TCE and PCE.
Greater than 10,000
Greater than 1,000 to 10,000
Greater than 100 to 1,000
Trace to 100
Not detected above reporting limits of 20 ppb v
for TCE and PCE and 10 ppb v for Benzene
or a trace levels
MW09-20I
©
1
<)
()
()
SELECTED MONITORING WELL AND NUMBER IF INDICATED—Shows
concentrations of (left side) petroleum compounds and (right side) chlorinated
compounds at single well or highest concentration at well cluster, in micrograms
per liter (Jig/L). Wells sampled in May 1995 (EA Engineering, Science, and
Technology, 1996)
Greater than 10,000
Greater than 1,000 to 10,000
Greater than 100 to 1,000
Trace to 100
Not detected above or below detection limit
OFFSHORE BORINGS—Volatile organic compounds detected in water from
offshore borings (EA Engineering, Science, and Technology, 1998b)
GENERALIZED DIRECTION OF SHALLOW GROUND-WATER FLOW
IN UPPER SAND UNIT
GENERALIZED DIRECTION OF DEEP GROUND-WATER FLOW
Figure 23. Directions of ground-water flow in the shallow and deep surfioial aquifers, concentrations of volatile organic
compounds in ground water beneath the Allen Harbor Landfill, December 1995, and concentration of trichloroethene in
passive-vapor-diffusion samplers installed in tidal-zone sediments along the shoreline of Allen Harbor Landfill, April 1998,
Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island.
Allen Harbor Landfill, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island
-------
Results
Concentrations of VOCs were detected in 41
of 115 vapor samplers placed near the Allen Harbor
Landfill. The most common VOC detected in the sam-
plers was TCE. Other VOCs detected included ben-
zene, toluene, and PCE. Vapor concentrations for total
VOCs exceeded 100 ppb v at eight locations (tig. 23).
VOCs were detected at 10 of the 20 high-tide locations
and at 5 of the 12 seeps. Comparison of the sample
results at the 20 locations where high-tide and low-tide
samplers were installed shows that where VOCs were
detected in the 10 high-tide samples, VOCs were
detected in only 3 of the companion low-tide samples;
and where VOCs were not detected in the remaining 10
high-tide samples, VOCs were detected in 2 of the low-
tide samples. The concentrations detected at seeps were
generally near minimum reporting levels, indicating
that the seeps were not major discharge areas for
VOCs. A trace of TCE was detected at one mudflat
location.
Highest VOC detections were in samplers that
were placed near well MW09-20I, where high VOC
concentrations were detected in ground water. The
extent of the area where VOCs were detected near well
MW09-20I indicates a VOC plume in ground water
under the landfill that is at least 300 ft wide at the
shore. The extent of the plume near well MW09-20I
had not been mapped previously.
CALF PASTURE POINT,
DAVISVILLE NAVAL CONSTRUCTION
BATTALION CENTER SUPERFUND
SITE, NORTH KINGSTOWN,
RHODE ISLAND
By Forest P. Lyford, Christine A.P.
Williams, anc/William C. Brandon
Description of Study Area
Calf Pasture Point is an area between Allen
Harbor and Narragansett Bay in North Kingstown,
Rhode Island (fig. 22), and was part of the former
Davisville Naval Construction Battalion Center.
Former waste-disposal activities in the area caused
contamination of ground water by VOCs (EA
Engineering, Science, and Technology, 1998b).
The land-surface altitude in the area shown in
figure 24 ranges from about 14 ft above sea level near
well MW07-14D to sea level at the shore line. Depth to
fractured crystalline bedrock ranges from near land
surface near monitoring well MW07-14D (fig. 24) to a
maximum depth of about 70 ft near Narragansett Bay
at the southeastern corner of figure 24. Bedrock depths
are about 30 to 40 ft near the shore of Allen Harbor and
at the entrance channel to Allen Harbor. An upper sand
layer consisting of fine to very fine sand overlies a layer
of silt, which, in turn, overlies till or bedrock. The silt
layer is absent along a north-south-trending till ridge
between wells MW07-26S and MW07-21S, where the
upper sand overlies till. The upper sand layer includes
materials that were dredged from the harbor or bay and
placed in a former lagoon in the central part of the
study area and also consists of materials formed by
recent sedimentation on the southern end of the study
area (fig. 24) (Church and Brandon, 1999).
Ground water in the upper sand layer flows semi-
radially outward from a topographic high area near
well MW07-14D toward Narragansett Bay, Allen
Harbor, and the entrance channel to Allen Harbor.
Ground-water-flow direction is generally to the south-
east in the deep till aquifer (fig. 24) (Church and
Brandon, 1999). Vertical hydraulic gradients generally
are downward, but near well MW07-21S upward gradi-
ents have been observed, except during high tide when
the gradient is zero or downward (EA Engineering,
Science, and Technology, 1997). Conceptually, shallow
ground water discharges within the intertidal zone or at
shallow depths beyond the intertidal zone, and deep
ground water in till and bedrock discharges in the inter-
tidal zone or further offshore. Data, however, are not
available for confirmation of discharge points. Shallow
ground water also appears to discharge to small wet-
land areas near the entrance channel to Allen Harbor
(fig. 25).
A plume of VOCs in ground water extends
from the general area of monitoring well MW07-14D
towards Narragansett Bay, the entrance channel to
Allen Harbor, and Allen Harbor (fig. 25). The
VOCs detected in the plume include vinyl chloride,
1,2-dichloroethene, PCE, TCE, and 1,1,2,2-tetrachloro-
ethane. Consequently, highest concentrations of VOCs
in ground water are in wells completed below the silt
layer at depths of 30 ft or more in till and bedrock.
In the area of the till ridge where the silt layer is
absent, VOCs are at shallow depths of 25 ft or less
(EA Engineering, Science, and Technology, 1998b).
40 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
41°37'30"
41°37,20"
71°24'45" 71°24'30"
MW07-14D /
n '
v^Calf Pasture Point
Former Dock
ENTRANCE CHANNEL TO
ALLEN HARBOR
Base traced from U.S. Geological Survey 0 200 FEET Map modified from Church and Brandon 1999, and
digital orthophoto, 1:12,000, GPS from | 1 f EA Engineering Services and Technology, 1997
Trimble Asset Surveyor, 1927 q go METERS
North American Datum
EXPLANATION
WATER-LEVEL CONTOUR IN UPPER SAND AND
LOWER TILL—Shows altitudes at which water would
have stood in tightly cased wells completed in the
upper sand and the deep till deposits, December 1,
1995. Contour interval is 1 foot.
1 Shallow wells in sand
2 Deep wells in till
GENERALIZED GROUND-WATER FLOW
DIRECTION IN UPPER SAND
GENERALIZED GROUND-WATER FLOW
DIRECTION IN LOWER TILL
APPROXIMATE LAND-SURFACE CONFIGURATION
OF CALF PASTURE POINT BEFORE FILLING WITH
DREDGED MATERIAL
APPROXIMATE EXTENT OF DREDGED FILL AND
FORMER LAGOON
POST-1966 SEDIMENTATION
W-21S WELL SITES, NUMBER, AND SAMPLING DEPTHS
S
S Shallow (sand)
D Deep (till)
C Cluster (shallow and deep)
Figure 24. Potentiometric surfaces and generalized ground-water-flow directions for the shallow and deep surficial
aquifers, Calf Pasture Point, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode
Island, December 1995.
Calf Pasture Point, Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island
-------
71 °24'45" 71°24'30"
Probable source area of
0 VOCs in ground-water
41°37'30"
MW07-26S O
Calf Pasture Point
Wetland
area
MW07-21S
^ H
Former Dock
ENTRANCE CHANNEL TO
ALLEN HARBOR
Spink Neck
41°37'20"
Base traced from U.S. Geological Survey
orthophoto, 1:12,000, GPS from Trimble Asset
Surveyor, 1927 North American Datum,
plume extent from EA Engineering, Science,
and Technology, 1998b
200 FEET
—I
60 METERS
EXPLANATION
Map modified from Lyford and others, 1999b, and
EA Engineering Services and Technology, 1997
SELECTED MONITORING WELLS AND NUMBERS
IF INDICATED—Shows concentrations of total
chlorinated volatile organic compounds at a single well
or highest concentration at well cluster, in micrograms
per liter (ng/L). Wells sampled in December 1995 and
May 1996 (EA Engineering, Science, and Technology,
1998b).Well numbers ending in S and D indicate screen
depth, shallow and deep, in the surficial aquifer.
Greater than 10,000
Greater than 1,000 to 10,000
Greater than 100 to 1,000
Trace to 100
Not detected above or below detection limit
— APPROXIMATE EXTENT OF TOTAL VOCs IN
GROUND WATER—Shows extent of volatile organic
compounds greater than 10 micrograms per liter.
PASSIVE-VAPOR-DIFFUSION SAMPERS—Shows
concentrations of trichloroethene and tetrachloroethene
in parts per billion by volume (ppb v); Samplers
with no identifier are in low tide and wetland areas;
H, Samplers in high tide areas; S, Samplers in seeps
o Greater than 1,000 to 10,000
O Greater than 100 to 1,000
• Trace to 100
° Not detected above reporting limit of 20 ppb v or at
trace level
Figure 25. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in the tidal-zone
sediments along the shoreline and in wetland-bottom sediments near the shoreline, Calf Pasture Point,
Davisville Naval Construction Battalion Center Superfund Site, North Kingstown, Rhode Island, March and
April 1998.
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New
-------
Purpose and Design of
Sampling
The PVD samplers were used at Calf Pasture
Point to identify possible discharge areas of VOC-
contaminated ground water. Other than the study at
the nearby Allen Harbor Landfill, PVD samplers had
not been used previously in a coastal setting. Conse-
quently, a secondary goal of the study was to determine
if PVD samplers can yield useful information about
areas of VOC discharge along a tidally affected
shoreline (Lyford and others, 1999b).
The PVD samplers were installed along the
shoreline in the area of the VOC plume at low- and
high-tide locations, in two wetland areas inland from
the entrance channel to Allen Harbor, and in four seeps
in or near the intertidal zone. The samplers were
installed during April 7 through 9,1998, at 65 locations
(fig. 25), and retrieved during April 28 through 29,
1998. The shoreline shown on figure 25 is approxi-
mately the extent of water at high tide. Most samplers
were placed during low tide at intervals of 50 ft along
about 1,500 ft of shoreline. Because of the limited
number of samplers available at the time of the study,
the southwestern side of the study area was given a
lower priority, and the distances between samplers was
greater than elsewhere. Samplers were placed at the
high-tide level at seven locations for comparison to
results from low-tide locations.
Results
Concentrations of VOCs were detected in sam-
ples from 7 of 37 PVD samplers placed within the
intertidal zone at Calf Pasture Point and in samples
from 1 of 24 samplers placed in wetland areas. Con-
centrations of VOCs were not detected in seepage
areas. The compound TCE was the only VOC detected,
except for a trace of PCE detected in one sample
(Lyford and others, 1999b). Concentrations of TCE
detected in the PVD samplers ranged from a trace to
1,900 ppb v (fig. 25). The occurrences of VOCs along
the shoreline were later confirmed by sampling from
drive-point wells.
OTIS AIR NATIONAL
GUARD/CAMP EDWARDS
SUPERFUND SITE, JOHNS
POND, FALMOUTH,
MASSACHUSETTS
By Jennifer G. Savoie and
Denis R. LeBlanc
Description of Study Area
A plume of dissolved VOCs in ground water
extends 9,000 ft from the site of a storm drain on the
Otis Air National Guard/Camp Edwards Superfund
Site, also known as the Massachusetts Military
Reservation (MMR), to Johns Pond, Mashpee, in the
Cape Cod area of Massachusetts (Air Force Center for
Environmental Excellence, 1997,1998a) (fig. 26). This
ground-water plume, known as the Storm Drain-5
(SD-5) plume, primarily consists of TCE with concen-
trations as high as 66,000 mg/L. Investigations by the
MMR's Installation Restoration Program (IRP) sug-
gested that the SD-5 plume was discharging to Johns
Pond (Air Force Center for Environmental Excellence,
1998b). The MMR Installation Restoration Program, in
cooperation with the USGS, sought to confirm that the
SD-5 plume was discharging to the pond and to
delineate the extent of the discharge area.
The Cape Cod aquifer near Johns Pond consists
of about 250 ft of glacial outwash sand and gravel. The
sediments are texturally uniform laterally and verti-
cally. Johns Pond is a ground-water flow-through
glacial kettle pond in this sand and gravel outwash
plain. Ground water generally flows into the pond near
its western side and discharges back into the ground
near its eastern side, as indicated by the water-table
contours on figure 26.
Otis Air National Guard/Camp Edwards Superfund Site, Johns Pond, Falmouth, Massachusetts 43
-------
MASSACHUSETTS
50 MILES
50 KILOMETERS
70°32'30'
70°30'
Massachusetts
Military
Reservation
SD-5
55
41°39'
STORM DRAIN-5
PLUME
3
3D
Area shown
in figure 27
ASHUMET
POND
45
JOHNS
POND
41 °37,30l
35
0 1 MILE Map modified from Savoie and others, 2000
1 1 1 1 1
and information from Air Force Center for
Environmental Excellence, 1998a
1 KILOMETER
EXPLANATION
STORM DRAIN-5 PLUME
—35— WATER-TABLE CONTOUR IN SURFICIAL AQUIFER—Shows altitude at
which water would have stood in tightly cased wells completed in the
surficial aquifer, March 1993. Contour interval is 5 feet.
Datum is sea level (Savoie, 1995).
MASSACHUSETTS MILITARY RESERVATION BOUNDARY
Figure 26. Locations of the Johns Pond study area and Storm Drain-5 contaminant
plume, and the altitude of water table (March 1993), Cape Cod, Massachusetts.
44 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Purpose and Design of
Sampling
In this study, PVD samplers were installed in the
bottom sediments of Johns Pond, to confirm that VOCs
from the SD-5 plume emanating from the MMR were
discharging into the pond (Savoie and others, 2000). In
July 1998, an array of 143 PVD samplers was buried
about 0.5 ft below the pond bottom in the presumed
discharge area of the SD-5 plume and left in place for
about 2 weeks to equilibrate (fig. 27). Divers installed
samplers at water depths of 5 to 30 ft in Johns Pond.
The lines of samplers extended a short distance into
an area of fine grained bottom sediments, presumed
to be the extent of most ground-water discharge.
In November 1998, a second more closely spaced
array of 119 PVD samplers was deployed on the basis
of interpretation of data collected in August 1998.
Results
Data from the PVD samplers indicated two areas
of high VOC concentrations. Samples from the first
area contained TCE and PCE with concentrations in
vapor as high as 890 and 667 ppb v, respectively
(fig 27). This discharge area is about 1,000 ft wide,
extends from 100 to 350 ft offshore, and is interpreted
to be the discharge area of the SD-5 plume. Lines of
samplers were long enough, by chance, to define the
shape of the discharge area. Samples from the second
area were closer to shore than the discharge area of the
SD-5 plume and contained vapor concentrations of
TCE as high as 47,000 ppb v. Ground-water samples
collected with a drive-point sampler near this location
confirmed the presence of TCE with concentrations as
high as 1,100 mg/L. The array of PVD samplers
deployed in November 1998 was centered around the
area of high TCE concentrations to map this presumed
separate plume (fig. 28). The discharge area detected
with the PVD samplers retrieved in December 1998
was about 75 ft wide and extended from about 25 to
200 ft offshore. Vapor concentrations of TCE in this
area were as high as 42,800 ppb v. Subsequent drilling
by MMR Installation Restoration Program consultants
confirmed that the TCE plume appears to be another
plume that originates northwest of Johns and Ashumet
Ponds and travels underneath Ashumet Pond (Air Force
Center for Environmental Excellence, 1999, 2001)
(fig. 29). Because of variations in ground-water-flow
patterns laterally and with depth, this plume enters the
area of Johns Pond from a different direction than the
SD-5 plume.
Otis Air National Guard/Camp Edwards Superfund Site, Johns Pond, Falmouth, Massachusetts 45
-------
70°31'28" 70°31 '22"
41°38'02"
41°37'55"
Map modified from Savoie and others, 2000
Figure 27. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers installed in
pond-bottom sediments adjacent to the Storm Drain-5 contaminant plume, Johns Pond, Cape Cod, Massachusetts,
August 1998.
BOAT RAMP
STORM DRAIN-5
PLUME
JOHNS
POND
~r 1—
50 METERS
Area shown
Oq in figure 28
EXPLANATION
PASSIVE-VAPOR-DIFFUSION
SAMPLERS—Shows concentrations of
trichloroethene and tetrachloroethene
in parts per billion by volume (ppb v).
O Greater than 10,000
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
O Trace to 100
O Not detected above reporting limit of 18 ppb v or at
trace level
® Sampler floating above pond bottom or sampler
not found
200 FEET
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
70°31'30"
70°31'28"
STORM DRAIN-5
PLUME
JOHNS POND
Map modified from Savoie and others, 2000
EXPLANATION
PASSIVE-VAPOR-DIFFUSION SAMPLERS—Shows concentrations
of tricholoroethene in parts per billion by volume (ppb v).
# Greater than 10,000
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
0 Trace to 100
O Not detected above reporting limit of 18 ppb v or at trace level
® Sampler floating above pond bottom or sampler not found
Figure 28. Concentrations of trichloroethene in passive-vapor-diffusion samplers installed in pond-bottom sediments in
the zones where high concentrations (greater than 10,000 parts per billion by volume) of trichloroethene were detected
with passive-vapor-diffusion samplers in August 1998, Johns Pond, Cape Cod, Massachusetts, December 1998.
Otis Air National Guard/Camp Edwards Superfund Site, Johns Pond, Falmouth, Massachusetts
-------
70°31' 40"
70 31120"
41 38' 05" —
41 37' 55" —
STORM DRAIN-5
PLUME
(1998)
ASHUMET
POND
NEWLY-FOUND
DISCHARGE AREA
(DECEMBER 1998)
EXPLANATION
STORM DRAIN-5 PLUME AND DISCHARGE AREA
OF TRICHLOROETHENE AND TETRACHLOROETHENE
INTO POND WHERE VAPOR CONCENTRATIONS ARE
GREATER THAN 100 PARTS PER BILLION BY VOLUME
EXTENT OF NEWLY-FOUND (1998) DISCHARGE AREA OF
TRICHLOROETHENE WHERE VAPOR CONCENTRATIONS
ARE GREATER THAN 100 PARTS PER BILLION BY
VOLUME—Plume approximated by subsequent ground-water
monitoring (Air Force Center for Environmental Excellence, 1999)
LOCATION OF PASSIVE-VAPOR-DIFFUSION SAMPLERS
UNKNOWN SOURCE AND DIRECTION
I
STORM DRAIN-5 PLUME
AREA OF DISCHARGE
(AUGUST 1998)
JOHNS
POND
0 200 FEET Map modified from Savoie and others, 2000
1 i i and assumed pathway of plume provided by
' * ' Air Force Center for Environmental Excellence, 1998-99
0 100 METERS
Figure 29. Discharge areas delineated with passive-vapor-diffusion samplers, August and December 1998, and ground-water
pathways of the Storm Drain-5 plume and trichloroethene plumes, Johns Pond, Cape Cod, Massachusetts.
NYANZA CHEMICAL WASTE
DUMP SUPERFUND SITE,
ASHLAND, MASSACHUSETTS
By Forest P. Lyford, Richard E.
Willey, and Sharon M. Hayes
Description of Study Area
The Nyanza property, part of the Nyanza
Chemical Waste Dump Superfund site, is a parcel of
land in Ashland, Massachusetts (fig. 30), where from
1917 to 1978, several textile dye manufacturers dis-
posed of various waste products. Some of the wastes
entered the ground-water system and formed a plume
that extends to the Sudbury River and a nearby former
mill raceway (fig. 31) (Roy F. Weston, Inc., 1998).
The ground-water system includes a surficial
aquifer of glacial lake deposits and till, and a bedrock
aquifer of fractured granite. The glacial lake deposits
range in grain size from silt to coarse sand and gravel.
The thickness of the fine-grained glacial lake sediments
increases eastward, and the depth to bedrock increases
eastward from less than 30 ft in the upstream end of the
mill pond area to about 80 ft near the upstream end of
the mill raceway. Most of the Superfund site is on till-
covered bedrock, and the Sudbury River is on silt, sand,
and gravel (Ebasco Services, Inc., 1991).
48 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
71°28'30" 71o28'00M
42°15'50n
<
£D
3
N
£D
o
3"
0)
q 42°15'30"
c
3
"D
to
c
"D
0)
Q)
3
a
g Base map modified from area site plan, Department of the 0 500 FEET Map modified from Lyford and others, 2000
£j> Army New England District Corps of Engineers, | 1—, 1
0 Concord, Massachusetts, and Weston Managers Designers/Consultants q jqq METERS
0)
O
3"
C
W
(D
s Figure 30. Location of Nyanza Chemical Waste Dump Superfund Site, passive-vapor-diffusion sampler locations, potentiometric-surface contours for the
surficial aquifer, and directions of ground-water flow, Ashland, Massachusetts.
Sudbury river
channel
Asm
R£ILROAD
mEGunko
NYANZA
CHEMICAL
WASTE DUMP
SUPERFUND
SITE
MASSACHUSETTS
50 MILES
0 50 KILOMETERS
MILL POND
str£et
190 —
EXPLANATION
APPROXIMATE BOUNDARY OF THE FORMER
NYANZA PROPERTY
WATER-LEVEL CONTOUR IN SURFICIAL AQUIFER—
Shows altitude at which water would have stood in tightly
cased wells completed in the surficial aquifer, March 1998.
Contour interval is 5 feet. Datum is sea level.
Data from Roy F. Weston, Inc., 1998
GENERAL DIRECTION OF GROUND-WATER FLOW
IN SURFICIAL AQUIFER
PASSIVE-VAPOR-DIFFUSION SAMPLERS
MONITORING WELLS
-------
o
c
a
£D
3
O
¦o
¦o
o"
£D
o"
3
0)
42°15'30"
m
3
(Q
Q)
3
a
Figure 31. The extent of contaminants in ground water and concentrations of chlorobenzene and trichloroethene detected in passive-vapor-diffusion samples,
Nyanza Chemical Waste Dump Superfund Site, Ashland, Massachusetts, February 1999.
MHGUNJCO
MILL
POND
PLEASANT
AILR
Base map modified from area site plan, Department of the
Army New England District Corps of Engineers, Concord,
Massachusetts and Weston Managers Designers/Consultants
500 FEET
100 METERS
Greater than 10,000
Greater than 1,000 to 10,000
Greater than 100 to 1,000
Trace to 100
Not detected above reporting limit of 12 ppb v
for TCE and 40 ppb v for CB or at trace levels
Map modified from
Lyford and others, 2000
(^) Sampler not found
• MONITORING WELLS
EXPLANATION
EXTENT OF GROUND-WATER PLUME WHERE STANDARDS FOR
DRINKING WATER ARE EXCEEDED—Shows area where at least one
chemical in ground water that is believed to originate from Nyanza site
exceeds Massachusetts criteria for drinking water. The location of this
line was determined on the basis of water samples from numerous wells
not shown on this map (Roy F. Weston, Inc., 1999a).
PASSIVE-VAPOR-DIFFUSION SAMPLERS—Shows concentrations of
(left side) trichloroethene (TCE) and (right side) chlorobenzene (CB) in
parts per billion by volume (ppb v).
-------
A potentiometric surface map for the surficial
aquifer (fig. 30) indicates that ground water flows
northward from the Nyanza property to the Sudbury
River and Mill Pond and eastward to the Sudbury River
and former mill raceway downstream from the dam
that forms Mill Pond (Roy F. Weston, Inc., 1998). A
plume of contaminants in the surficial and bedrock
aquifer system follows the ground-water-flow direction
(fig. 31). Contaminants detected in ground-water
monitoring wells near the river include the VOCs
1,1,1- trichloroethane, benzene, chlorobenzene,
d.v-DCE, PCE, TCE, and vinyl chloride. Chloroben-
zene, TCE, and cis-DCE are the VOCs most commonly
detected in ground water in the area of the plume. Also
detected in ground water are mercury and the semi-vol-
atile organic compounds (SVOCs) 1,2,4-trichloroben-
zene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and
1,4-dichlorobenzene (Roy F. Weston, Inc., 1999a).
Contaminants were not detected in monitoring wells
near the downstream segment of the mill pond area,
where water-level data show the pond is a source of
recharge to ground water (Roy F. Weston, 1999b)
(fig. 31).
Purpose and Design of
Sampling
PVD samplers were used near the Nyanza prop-
erty to accomplish three objectives (Lyford and others,
2000):
1. Determine if the distribution and concentrations
of VOCs detected in samples from PVD sam-
plers placed in stream-bottom sediments are
consistent with the previously mapped distribu-
tion of contaminants in ground water near the
river and former mill raceway.
2. Determine the time needed for VOCs in bottom
sediments to re-equilibrate after installation of
the samplers at this site.
3. Determine if PVD samplers might serve as
alternatives to other sediment-pore-water
sampling techniques, specifically seepage
meters and whole sediment samples, to
characterize the occurrence of VOCs in
stream-bottom sediments.
To accomplish the first objective, PVD samplers
were placed at 22 locations along the south bank of the
Sudbury River in the area of Mill Pond and along the
Sudbury River and raceway downstream from Mill
Pond Dam. Samplers were deployed during January 19
through 20, and retrieved on February 16, 1999. To
accomplish the second objective, two clusters of three
PVD samplers each were placed at two sites. One PVD
sampler from each cluster was retrieved at 1-week
intervals and transported to USEPA's Lexington,
Massachusetts, laboratory for analysis. To accomplish
the third objective, PVD samplers were placed at loca-
tions selected by USEPA to compare analytical results
with analyses of sediments and pore water extracted
from seepage meters (Lyford and others, 2000). The
seepage meters were inverted segments of steel drums
equipped with a nozzle on the top for extraction of
water (Roy F. Weston, Inc., 1999a).
Results
The distribution of VOCs in ground water
detected in samples from PVD samplers agrees well
with the distribution of contaminants in ground water
mapped on the basis of samples from monitoring wells
(fig. 31). Low levels of TCE (less than 25 ppb v) in
PVD samples at the location farthest upstream indi-
cates that VOCs in ground water may extend somewhat
further west than the plume shown in figure 31. The
absence of VOCs in samples from samplers placed at
the downstream segment of Mill Pond is consistent
with water-level observations that this section of
the pond acts as a recharge source to ground water.
Although chemical data are limited along the Sudbury
River downstream from the dam, the general absence
of VOCs in PVD samples indicates that this river reach
was not a major discharge area for contaminants in
ground water at the time of the study. The presence of
VOCs in PVD samples along the mill raceway, how-
ever, confirms the mapped extent of the contaminant
plume in ground water and indicates that the raceway is
a discharge area for contaminated ground water.
An evaluation of equilibration time for PVD
samplers in bottom sediments disturbed during installa-
tion of samplers indicates that 3 weeks or more may be
needed in some settings for equilibration. The results
were inconclusive, however, because changes in river
stage and discharge may have affected concentrations
of VOCs. Also, concentrations of VOCs in sediments
may vary over short distances, and sampling and
analytical methods are imprecise.
Nyanza Chemical Waste Dump Superfund Site, Ashland, Massachusetts 51
-------
A comparison of analytical results for PVD
samplers to analytical results for water from seepage
meters indicated that concentrations of chlorobenzene
and TCE correlated well for the two sampling methods.
A comparison of results from PVD samplers to chemi-
cal analyses of sediments indicated that concentrations
of chlorobenzene and TCE correlated poorly for the
two methods. At several locations, PVD samplers
detected VOCs where they were not detected in sedi-
ment samples. The apparent absence of VOCs in sedi-
ment samples may have resulted from high quantitation
limits for the analyses.
CENTREDALE MANOR
RESTORATION PROJECT
SUPERFUND SITE, NORTH
PROVIDENCE, RHODE ISLAND
By Peter E. Church, Forest P.
Lyford, and Anna F. Krasko
41 51'30"
Description of Study Area
At the Centredale Manor Restoration Project
Superfund Site in North Providence, Rhode Island,
the location of a former chemical company and a
drum reclamation company, PCBs, dioxin, SVOCs,
and VOCs have been detected in soils, and VOCs
have been detected in ground water (A.F. Krasko,
U.S. Environmental Protection Agency, written
commun., 1999). The study area is an elongated area
of about 12 acres along the eastern bank of the
Woonasquatucket River just downstream of the U.S.
Route 44 bridge (fig. 32). A former mill raceway that is
about 1,900-ft long and located several hundred feet
east of the river forms the approximate eastern bound-
ary of the study area (fig. 33). The southern boundary
of the study area is 50 ft downstream of the confluence
of the mill raceway and the Woonasquatucket River,
about 2,250 ft downstream from the U.S. Route 44
bridge. A cross channel, about 175-ft long, connects
the river to the mill raceway about 600 ft upstream
from this confluence.
Monitoring wells, installed in the northern part
of the site in March 1999, encountered a top layer of
fill, 3 to 6 ft thick, composed of silt, sand, gravel, and
fragments of bricks, concrete, and wood. The fill is
underlain by 3 to more than 8 ft of sand and gravel,
41 51'15"
Figure 32. Locations of the Centredale Manor
Restoration Project Superfund Site and study area,
North Providence, Rhode Island.
which, in turn, is underlain by silty, gravelly sand,
described as till of unknown thickness (A.F. Krasko,
U.S. Environmental Protection Agency, written com-
mun., 1999). The water table at the time these wells
were drilled ranged from 2.6 to 7.5 ft below land sur-
face. A potentiometric surface map is not available for
this site. From June to November 1999, the USEPA
collected numerous soil samples in the study area and
detected dioxin, PCBs, VOCs, and SVOCs. The occur-
rence and distribution of contaminants in ground water
had not been characterized at the time of this study.
>vidence
RHODE
ISLAND
71 °29'30" 71°29'05'
Area shown
In figure 33
ralSpnna
JOHNSTON
1,000 FEET Map modified from
Church and others, 2000
0 150 METERS
0 10 MILES
1 1—1
0 10 KILOMETERS
NORTH PROVIDENCE
52 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
71WI5B 71fl29'10B
0 FEET
41°51'30"
Bro )k
Village
0 FEET
500 FEET
>— 500 FEET
-1,000 FEET
Ce itred
I lanoi
1,000 FEET-
1,500 FEET
Cross channel
41°5t15M
1500 FEET-
2,000 FEET
200 FEET
50 METERS
——1,900 FEET
2,250 FEET ->
EXPLANATION
PASSIVE-VAPOR-DIFFUSION
SAMPLERS—Shows concentrations of
tetrachloroethene and trichloroethene in parts per
billion by volume (ppb v).
• Greater than 10,000
O Greater than 1,000 to 10,000
O Greater than 100 to 1,000
• Trace to 100
O Not detected above reporting limit of
10 ppb v or at trace level
® Sampler not found
—- DIRECTION OF RTVER FLOW
Map modified from
Church and others, 2000
Figure 33. Concentrations of trichloroethene plus tetrachloroethene in passive-vapor-diffusion samplers
installed in channel-bottom sediments of the Woonasquatucket River, a former mill raceway, and a cross
channel, Centredale Manor Restoration Project Superfund Site, North Providence, Rhode Island, September
1999.
Centredale Manor Restoration Project Superfund Site, North Providence, Rhode Island
-------
Purpose and Design of
Sampling
On September 8 and 9,1999,115 PVD samplers
were installed along the river, raceway, and cross
channel to identify possible discharge areas of VOC-
contaminated ground water and for a preliminary eval-
uation of the distribution and concentrations of VOCs
and contaminant pathways in ground water (Church
and others, 2000). The distance between samplers was
50 ft, except in reaches of the river and cross channel
where the USEPA had indicated possible discharge of
contaminants; in these locations, samplers were placed
at intervals of 25 ft or less.
On September 21 and 22, 1999, the PVD sam-
plers were retrieved. Sixty of the 62 samplers installed
in the Woonasquatucket River were retrieved despite
2 near-flood flows in the 2 weeks after they were
installed, which is most likely when the 2 samplers
were lost. Nine of the 19 samplers installed in the lower
section of the mill raceway were not found. Most likely
these samplers were either washed away or buried in
recently re-worked streambed sediments caused by
high flows. All of the samplers installed in the upper
section of the mill raceway and in the cross channel
were retrieved. Target compounds for analysis, selected
on the basis of soil-sample data, were benzene, ethyl-
benzene, toluene, meta/para-xylene, ortho-xylene,
chlorobenzene, PCE, TCE, and 1,1,1-trichloroethane.
Results
VOCs were detected in 58 of the 60 PVD sam-
plers placed in the river, 10 of the 24 samplers in the
upper mill raceway, 9 of the 10 samplers from the
lower mill raceway, and 9 of the 10 samplers in the
cross channel. The compounds TCE and PCE were
the principal VOCs detected of the nine target com-
pounds, and vapor concentrations of these two com-
pounds were generally less than 100 ppb v (fig. 33).
Higher vapor concentrations, however, were detected
along short reaches of these waterways. Vapor concen-
trations of TCE+PCE in samplers placed in the
Woonasquatucket River about 500 to 600 ft down-
stream of the U.S. Route 44 bridge ranged from about
4,000 to 1,600,000 ppb v (fig. 33). The high vapor con-
centrations in this short reach, compared to vapor con-
centrations in river-bottom sediments upstream and
downstream and in the former mill raceway and cross
channel indicate that this is a major discharge area of
contaminated ground water. The compounds TCE and
PCE were detected in most PVD samplers downstream
of this discharge area to the outlet of the mill raceway,
but concentrations of these samples were much lower.
These concentrations may reflect discharge of less
contaminated ground water, especially in the approxi-
mate 350 ft reach of the river above the mill raceway
outlet, or may represent mixing of contaminated river
water with sediment-pore waters. Concentrations of
TCE+PCE greater than 100 ppb v also were detected in
the lower part of the upper mill raceway, in the lower
mill raceway, and in the eastern part of the cross chan-
nel, indicating possible discharge areas of contami-
nated ground water (fig. 33). Nondetect or trace levels
of VOCs immediately downstream of Route 44 suggest
minimal contributions of VOCs from upgradient
sources.
QUALITY-ASSURANCE
PROCEDURES
Quality-assurance procedures for PVD sampling
in New England were designed to help explain spurious
detections of VOCs or anomalously high or low con-
centrations, if any, that would be difficult to explain on
the basis of available site information. For all studies
reported here, the primary objective was to identify
possible discharge areas for VOCs in ground water. For
this objective, the main quality concern was detections
in PVD samplers where VOCs were not present and
nondetections where VOCs were present. To help eval-
uate the concern that the sampling or analytical tech-
nique might cause detections of VOCs where they were
not present, samplers were placed in areas where con-
taminants were not likely, such as upstream from or
upgradient from mapped contaminant plumes in
ground water. For all studies, numerous samples indi-
cated that VOCs were not present in detectable quanti-
ties where they were unlikely to be present in ground
water.
Duplicate samples were collected to help
address the concern of false VOC detections where
they were not present or false nondetections where they
were present. Analyses of numerous samples that indi-
cated concentrations below or above detection limits
were confirmed with duplicate samples. Of the 437
54 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
duplicate-samples analyses (number of duplicate sam-
ples times the number of target compounds), a VOC
was not detected above reporting limit in either sample
in 343 duplicate samples; a VOC was detected above
the reporting limit in both samples in 83 duplicate sam-
ples; a VOC was detected above the reporting limit in
one sample but not the other in 11 duplicate samples
(table 3). In 6 of the 11 exceptions, where a VOC was
detected in one of the samples, but was not in the other,
the highest concentration was less than 100 ppb v.
The concentrations of VOCs detected in PVD
samplers for some sites provided useful information
about the relative magnitude of concentrations of
VOCs that might be expected at the ground-water and
surface-water interface. For this use of the data, results
from duplicate samples generally provided assurance
that concentrations detected in samples were reason-
able for that point sampled. A relative percent differ-
ence of 30 percent is commonly used to evaluate
measurement performance in screening methods. A
summary of relative percent differences between dupli-
cate samples where a VOC was detected in both sam-
ples is shown in table 4. The relative percent difference
for most duplicate samples analyzed for chlorinated
VOCs (PCE, TCE, and chlorobenzene) were within 30
percent. For unknown reasons, however, a lower per-
centage of samples analyzed for petroleum compounds
were within the 30 percent criteria for relative percent
difference.
Trip blanks and equipment blanks also were used
in some places to determine if contaminants were intro-
duced during transport or during capping of sampler
vials. No VOCs were detected in any of these samples
except for a low concentration of toluene in one sam-
ple. The detection of toluene did not compromise the
result observed for that site.
Table 3. Number and distribution of duplicate samples from the nine study sites in New England
[Number of duplicate sample analyses is the number of duplicate samplers deployed multiplied by the number of target compounds]
Study site
Number of
duplicate
sample
analyses
Concentration
detected above
reporting limit
in both samples
Concentration
not detected
above reporting
limit in both
samples
Concentration
detected above
reporting limit
in only one
sample
Eastern Surplus Company Superfund Site,
Meddybemps, Maine 12 4 8 0
McKin Company Superfund Site, Gray, Maine 12 8 4 0
Nutmeg Valley Road Superfund Site, Wolcott and
Waterbury, Connecticut 56 8 45 3
Baird & McGuire Superfund Site, Elolbrook,
Massachusetts 215 33 175 7
Allen Elarbor Landfill, Davisville Naval Construction
Battalion Center Superfund Site, North Kingstown,
Rhode Island 24 1 23 0
Calf Pasture Point, Davisville Naval Construction
Battalion Center Superfund Site, North Kingstown,
Rhode Island 6 0 6 0
Otis Air National Guard/Camp Edwards Superfund
Site, Johns Pond, Falmouth, Massachusetts 36 12 23 1
Nyanza Chemical Waste Dump Superfund Site,
Ashland, Massachusetts 20 6 14 0
Centredale Manor Restoration Project Superfund
Site, North Providence, Rhode Island 56 11 45 0
Total 437 83 343 11
Quality-Assurance Procedures 55
-------
Table 4. Relative percent differences of volatile organic
compound (VOC) concentrations in duplicate samples where
a VOC was detected above the reporting limit in both
duplicate samples from the nine study sites in New England
Volatile
organic
Number of
duplicate
Percent of relative
percent differences of
concentrations in
duplicate samples
compound
samples
Less than
or equal to
30 percent
Greater than
30 percent
Tetrachloroethene
13
92
8
Trichloroethene
37
81
19
Chlorobenzene
2
100
0
Benzene
10
60
40
Ethylbenzene
6
50
50
meta/para-Xylene
7
43
57
ortho-Xylene
4
75
25
Toluene
4
75
25
SUMMARY AND
CONCLUSIONS
Passive-vapor-diffusion (PVD) samplers are
designed for detecting and delineating areas of VOC-
contaminated ground water discharging into surface-
water bodies. A PVD sampler consists of an empty
glass vial sealed inside a polyethylene membrane
tubing that is permeable to many VOCs of interest,
such as petroleum and chlorinated compounds, but is
not permeable to water. Samplers are buried in the
bottom sediment of surface-water bodies, at or below
the transition zone between surface water and ground
water; and VOCs in the adjacent pore water, if present,
diffuse through the polyethylene tubing and equilibrate
with concentrations of air in the empty vial.
Applications of PVD samplers at and near nine
hazardous-waste sites in New England have demon-
strated the effectiveness of this sampling method in
several hydrologic settings, including rivers and
streams, ponds, wetlands, and coastal shorelines
through a variety of bottom sediments including sand,
silt, clay, organics, gravel, and cobbles. Areas of VOC-
contaminated ground-water discharge from known
ground-water plumes were confirmed or refined
with PVD samplers at all nine sites. Areas of VOC-
contaminated ground-water discharge from previously
unknown ground-water plumes were identified with
PVD samplers at the following Superfund Sites:
Eastern Surplus Company in Maine; Baird & McGuire
and Otis Air National Guard/Camp Edwards (Johns
Pond) in Massachusetts; and Centredale Manor
Restoration Project in Rhode Island.
The samplers should remain in place until suffi-
cient time has elapsed for the pore water to recover
from the environmental disturbances caused by sam-
pler deployment and for the samplers to attain suffi-
cient VOC concentrations to fulfill the data-quality
objectives of the study. If the data-quality objective
is to locate a VOC-contaminated ground-water-
discharge area, then the samplers may be recovered
before they have completely equilibrated, if they have
accumulated sufficient VOC concentration to identify
the contaminant-discharge area. Field studies suggest
that this can be accomplished after a deployment
period of 8 days or less, and in as little as 24 hours
in some environments.
Spacing of samplers and selection of sampler
location are also important in achieving the data-
quality objectives. A sampling strategy for detecting
plumes of VOCs in ground water or refining plume
boundaries typically requires preliminary knowledge
about potential sources of VOCs and a conceptual
model of pathways for contaminants in ground water.
Sampling should extend well upstream and down-
stream from likely discharge areas. The studies in New
England have demonstrated that plumes typically dis-
charge well beyond a riverbank or shoreline on the
plume side of the water body. In fact, samplers placed
at the edge of the water body may not detect a plume.
In small streams a few feet wide, samplers placed along
the center of a stream are typically sufficient to map
plumes in ground water. Most plumes in New England
hydrologic settings are at least 100 ft wide where they
enter a surface-water body, so a spacing of 50 to 100 ft
should be adequate to detect the presence of VOCs. In
large water bodies such as rivers and ponds, a grid of
samplers spaced 50 to 100 feet apart would typically
detect the presence of VOCs. If the sampling goal is to
characterize local variations in concentrations of
VOCs, a shorter spacing may be needed.
The ability of PVD samplers to detect areas
of discharging VOC-contaminated ground water
depends on a variety of factors affecting contaminant
discharge. These factors include the location and
56 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
lithologic heterogeneity of the discharge area, the
orientation of the stream channel relative to the
ground-water contamination, the offshore distance
of the contaminant-discharge area, the potential tempo-
ral nature of the discharge-area location and discharg-
ing concentrations, and potential removal of VOCs by
bed-sediment micro-organisms.
Two studies in New England (Lyford and others,
2000; Savoie and others, 2000) included co-located
PVD and water-tilled polyethylene bags described by
Vroblesky (2001) to determine if concentrations of
VOCs in pore water can be predicted from concentra-
tions observed in PVD samplers. In theory, Henry's
Law can be used to convert concentrations of VOCs in
vapor to concentrations in water. Results of the two
studies cited indicated that the use of PVD samplers for
this purpose might be appropriate. Also needed, how-
ever, are temperature data measured with commercially
available temperature probes for pore water at the sam-
pling depth, and equilibration times of PVD samplers
for the sediments in which the sample is obtained. In
some settings, direct collection of pore-water samples,
such as with small-diameter probes (Henry, 2000), may
be more efficient than use of PVD samplers.
Several other types of information are useful for
interpreting results from PVD sampling. It is helpful
to know concentrations of VOCs in surface water. It
is commonly assumed that the VOCs that discharge
from ground water to surface water quickly volatilize
to the atmosphere. In practice, however, VOCs persist
downstream from a source and can compromise
results of PVD sampling. Sampling of surface water
upstream and downstream of the study area and at sev-
eral points along a stream, depending on the length of
the stream reach, is useful for interpreting PVD data
from sediment-pore water. Surface-water data might
be collected by suspending a sampler in surface water
near the bottom of the water body or collecting water
samples for head-space analysis on site or for standard
laboratory analyses offsite.
It is also helpful to collect hydraulic head data at
piezometers. Concurrent water levels in piezometers
and surface water at the time of sampler deployment or
retrieval will help determine if VOCs are discharged
with ground water or accumulated in sediments from
other sources. Continuous stage and ground-water-
level data are also needed to detect changes in hydro-
logic conditions, such as a flood wave, that may affect
VOC concentrations in sediments. Subsequent sam-
pling may be desirable if hydrologic conditions
changed appreciably during the sampling period.
Knowledge gained and lessons learned from the
nine New England studies may be useful for others
considering the application of PVD samplers. Results
from these studies have provided insights on ground-
water-flow patterns near surface-water bodies. Varia-
tions of VOC concentrations over short distances
within areas of ground-water discharge were detected
at several sites, which suggests local variations in con-
taminants at the contaminant source and possible local
variations in ground-water-flow patterns. Discharge of
VOC-contaminated ground water has been detected
across the widths of rivers and streams. In one case,
continued flow beneath the floodplain at a bend in the
river resulted in VOCs being discharged through the
bottom sediments to the opposite bank further down-
stream. Results at one site reinforced concepts about
gaining and losing reaches in a mill-pond area and
indicated that a former mill raceway was the principal
discharge area for contaminated ground water down-
stream from the mill pond. Numerous detections of
VOCs within a tidal zone near a landfill supported
the concept that much of the shallow ground-water
discharges in the tidal zone. The absence of VOCs in
another tidal zone where VOCs are known to be present
at depth suggests that the deep ground water discharges
further offshore. The effect of surficial geology on
ground-water discharge to surface-water bodies was
observed at one site where the area of ground-water
discharge from a deep surficial aquifer was affected by
the lateral extent of an overlying confining lacustrine
deposit.
Concentrations of VOCs in PVD samplers placed
in river-bottom sediment downstream from the likely
extent of a plume, but also in areas where VOCs were
detected or are likely to be in surface water, may indi-
cate an exchange of water between the surface water
and bottom sediments. In this situation, surface-water
sampling is needed, as well as the use of PVD samplers
in the bottom sediments of the surface-water body. This
additional sampling will help determine if VOCs in
surface water have affected the concentrations of VOCs
detected with the PVD sampler.
The absence of VOCs in PVD samplers does not
exclude the possibility that VOCs are present in ground
water. For at least two studies, VOCs were not detected
Summary and Conclusions 57
-------
in some areas where they were known to be present in
ground water. Local flow conditions may have diverted
the discharge of contaminated ground water in these
areas, or surface water may have been the principal
source of water in bottom sediments. Where VOCs are
suspected but not detected in PVD samplers, ground-
water sampling may be needed for confirmation. The
absence of VOCs at the ground-water/surface-water
interface in areas of known ground-water contamina-
tion may indicate that surface water is a source of
recharge. It may also indicate that discharge points for
contaminated ground water are outside the area of
investigation. This information is useful for further
characterization of ground-water contamination and
contaminant pathways.
Results from PVD samplers provide a qualitative
assessment of VOC concentrations in ground-water
plumes where they enter surface-water bodies. A fair
correlation was observed between concentrations of
VOCs in PVD samples and concentrations in water
from co-located seepage meters. Several uncertainties,
however, preclude estimating actual VOC concentra-
tions in ground water from concentrations in vapor
from PVD samplers. These uncertainties include equil-
ibration times, variations in concentration over short
lateral distances, variations in concentrations vertically
(particularly at shallow depths), variations in tempera-
tures at the ground-water and surface-water interface,
and changing hydrologic conditions that affect ground-
water-flow patterns and flow rates. These uncertainties,
however, are not critical to successful application of the
technique if the data-quality objective is simply to
identify the presence of VOCs.
Quality-assurance procedures for PVD sampling
help explain questionable detections of VOCs or unex-
pected high or low concentrations. For all studies in
New England, numerous samples indicated that VOCs
were not present in detectable quantities where they
were unlikely to be present in ground water. Analysis
of duplicate samples helped address the concern of
false VOC detections where VOCs are not present, or
false nondetections where VOCs are present. Numer-
ous duplicate samples confirmed concentrations below
or above detection limits. Duplicate samples also pro-
vided assurance that concentrations detected were rea-
sonable for the point sampled. The relative percent
difference for most duplicate samples analyzed for
chlorinated VOCs were within 30 percent. For petro-
leum compounds, a lower percentage of samples ana-
lyzed were within the 30 percent criteria for relative
percent difference. In some areas, analysis of trip
blanks and equipment blanks helped determine if con-
taminants were introduced during transport or during
capping of sample vials. VOCs were not detected in
any of these samplers, except for a low concentration
of toluene in one sampler, which did not compromise
the result observed for that site.
Unexplored uses of PVD samplers include evalu-
ation of chemical transformations at the ground-
water/surface-water interface. These transformations
can be evaluated by targeting degradation products
during analysis of vapor or water, and by identification
of areas of high concentrations of chemicals of envi-
ronmental concern, such as semi-volatile organic com-
pounds and metals, that are co-located with VOCs in
ground water. After equilibration times for a particular
setting are determined, PVD samplers also may help
monitor concentrations of VOCs. Deploying PVD sam-
plers could provide an alternative to installing wells,
when a permanent well would present a safety hazard,
such as to boats and swimmers, or where permanent
wells would be subject to damage by vehicles or ice.
Similarly, PVD samplers could be useful for studying
how VOC concentrations change as hydrologic
conditions change.
REFERENCES CITED
Air Force Center for Environmental Excellence, 1997,
SD-5 South pre-design technical memorandum
(Draft): Jacobs Engineering Group, Inc., October
1997, variously paged.
1998a, SD-5 South project execution plan
(Draft): Jacobs Engineering Group, Inc., January
1998, variously paged.
1998b, Technical memorandum, assessment of
SD-5 South-Johns Pond interaction for decision
modification (Draft): Jacobs Engineering Group,
Inc., November 1998, variously paged.
1999, SD-5 South plume and adjacent TCE
plume design and data-gap technical memoran-
dum, (Draft): Jacobs Engineering Group, Inc., July
1999, variously paged.
2001, Final Chemical Spill 10, Remedial
Investigation Report: Jacobs Engineering Group,
Inc., September, 2001, variously paged.
58 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Baskin, R.L., 1998, Locating shoreline and submarine
springs, in Pitman, J.K., and Carroll, A.R., eds.,
Modern and ancient lake systems: Utah
Geological Association Guidebook 26, p. 51-57.
Boulton, J.A., 1993, Stream ecology and surface-
hyporheic hydrologic exchange: implications,
techniques and limitations: Australian Journal of
Marine and Freshwater Research, v. 44,
p. 553-564.
Campbell, J.P., Lyford, F.P., and Willey, R.E., 2002,
Comparison of vapor concentrations of volatile
organic compounds with ground-water
concentrations of selected contaminants in
sediments beneath the Sudbury River, Ashland,
Massachusetts, 2000: U.S. Geological Survey
Open-File Report 02-143, 33 p.
Carignan, R., 1984, Interstitial water sampling by
dialysis, methodological notes: Limnology and
Oceanography, v. 29, p. 667-670.
Church, P.E., and Brandon, W.C., 1999, Distribution of
salinity in ground water from the interpretation of
borehole-geophysical logs and salinity data, Calf
Pasture Point, Davisville, Rhode Island: U.S.
Geological Survey Water-Resources Investigations
Report 99-4153, 47 p.
Church, P.E., Lyford, F.P, and Clifford, Scott, 2000,
Distribution of selected volatile organic
compounds determined with water-to-vapor
diffusion samplers at the interface between ground
water and surface water, Centredale Manor Site,
North Providence, Rhode Island, September 1999:
U.S. Geological Survey Open-File Report 00-276,
9 p., 1 pi.
2002, Volatile organic compounds, specific
conductance, and temperature in the bottom
sediments of Mill Pond, Ashland, Massachusetts,
April 2001: U.S. Geological Survey Open-File
Report 02-35, 10 p.
Conant, Brewster, Jr., 2000, Ground-water plume
behavior near the ground-water/surface-water
interface of a river, in Proceedings of the Ground
Water/Surface Water Interactions Workshop,
January 26-28, 1999: Denver, CO, U.S.
Environmental Protection Agency, p. 23-30.
EA Engineering, Science, and Technology, 1996,
Revised draft final, IR Program Site 09, Allen
Harbor Landfill, Phase III remedial investigation,
volumes I and II—Technical Report, Naval
Construction Battalion Center, Davisville, Rhode
Island: Sharon, MA, EA Engineering, Science,
and Technology, 308 p.
1997, Draft Final, IR Program Site 07, Calf
Pasture Point, Phase III remedial investigation,
volume I—Technical Report, Naval Construction
Battalion Center, Davisville, Rhode Island:
Bedford, MA, EA Engineering, Science, and
Technology, 282 p.
1998a, Draft, IR Program Site 09, Allen Harbor
Landfill offshore investigation report, Naval
Construction Battalion Center, Davisville, Rhode
Island: Bedford, MA, EA Engineering, Science,
and Technology, 16 p.
1998b, Final, IR Program Site 07, Calf Pasture
Point, Phase III remedial investigation, volume I—
Technical Report, Naval Construction Battalion
Center, Davisville, Rhode Island: Bedford, MA,
EA Engineering, Science, and Technology, 282 p.
Ebasco Services, Inc., 1991, Draft final remedial
investigation report, Nyanza II—Groundwater
Study, Ashland, Massachusetts, volume I: Boston,
MA, Ebasco Services, Inc., variously paged.
Glaser, PH., Janssens, J.A., and Siegel, D.I., 1990,
The response of vegetation to chemical and
hydrological gradients in the Lost River Peatlands,
northern Minnesota: Journal of Ecology, v. 78,
p.1021-1048.
Glaser, PH., Wheeler, G.A., Gorham, E., and Wright,
H.E., Jr., 1981, The patterned mires of the Red
Lake peatland, northern Minnesota—Vegetation,
water chemistry, and landforms: Journal of
Ecology, v. 69, p. 575-599.
Hare, P.W., 2000, Passive diffusion bag samplers
for monitoring chlorinated solvents in ground
water: Monterey, CA, The Second International
Conference on Remediation of Chlorinated and
Recalcitrant Compounds, Battelle, May 22-25,
2000, v. 2, no. l,p. 375-386.
Harvey, J.W., and Bencala, K.E., 1993, The effect of
streambed topography on surface-subsurface water
exchange in mountain catchments: Water
Resources Research, v. 29, no. 1, p. 89-98.
References Cited 59
-------
Henry, M.A., 2000, Appendix D, MHE push point
sampling tools, in Proceeding of the Ground-
Water/Surface-Water Interactions Workshop: U.S.
Environmental Protection Agency EPA/542/R-
00/007, p. 199-200.
HRP Associates, Inc., 1991, Site assessment report, for
Celinda W. Mayo, 76 Wolcott Road, Wolcott,
Connecticut: New Britain, CT, updated 1988,
unpaged.
Huckins, J.N., Petty, J.D., Lebo, J.A., Orazio, C.E., and
Priest, H.F., 1997, Comment on "Accumulation
of organochlorine pesticides and PCBs by
semipermeable membrane devices and Mytilus
edulis in New Bedford Harbor": Environmental
Science & Technology, v. 31, p. 3732-3733.
Lee, D.R., 1977, A device for measuring seepage
flux in lakes and estuaries: Limnology and
Oceanography, v. 22, no. 1, p. 140-147.
1985, Method for locating sediment anomalies in
lakebeds that can be caused by ground water-flow:
Journal of Hydrology, v. 79, p. 187-193.
Lee, D.R., and Tracey, J.P, 1984, Identification of
ground-water discharge locations using thermal
infrared imagery, in Proceedings of the Ninth
Canadian Symposium on Remote Sensing: St.
Johns, Newfoundland, Canada, p. 301-308.
Loureiro Engineering Associates, 1998a, Status
report, supplemental subsurface investigation, the
Highland Manufacturing Company, 1240 Wolcott
Road, Waterbury Connecticut: Plainville, CT,
Loureiro Engineering Associates, February 1998,
unpaged.
1998b, Supplemental subsurface investigation
report, Highland Manufacturing Company,
Waterbury, Connecticut, July 1998: Plainville, CT,
Loureiro Engineering Associates, variously paged.
Lyford, F.P, Flight, L.E., Stone, J.R., and Clifford,
Scott, 1999a, Distribution of trichloroethylene and
geologic controls on contaminant pathways near
the Royal River, McKin Superfund Site Area,
Gray, Maine: U.S. Geological Survey Water
Resources Investigations Report 99-4125, 20 p.
Lyford, F.P, Kliever, J.D., and Clifford, Scott, 1999b,
Volatile organic compounds detected in vapor-
diffusion samplers placed in sediments along and
near the shoreline at Allen Harbor Landfill and
Calf Pasture Point, Davisville, Rhode Island,
March-April 1998: U.S. Geological Survey Open-
File Report 99-74, 9 p.
Lyford, F.P, Stone, J.R., Nielsen, J.P, and Hansen,
B.P, 1998, Geohydrology and ground-water
quality, Eastern Surplus Superfund Site,
Meddybemps, Maine: U.S. Geological Survey
Water-Resources Investigations Report 98-4174,
68 p.
Lyford, F.P, Willey, R.E., and Clifford, Scott, 2000,
Field tests of polyethylene-membrane diffusion
samplers for characterizing volatile organic
compounds in stream-bottom sediments, Nyanza
Chemical Waste Dump Superfund Site, Ashland,
Massachusetts: U.S. Geological Survey Water-
Resources Investigations Report 00-4108, 19 p.
Mullaney, J.R., Mondazzi, R.A., and Stone, J.R., 1999,
Hydrogeology and water quality of the Nutmeg
Valley Area, Wolcott and Waterbury, Connecticut:
U.S. Geological Survey Water-Resources
Investigations Report 99-4081, 90 p.
Nichols, R.L., 1993, Characterization of shallow
ground water at TNX: South Carolina,
Westinghouse Savannah River Company report
WSRC-TR-92-508, Savannah River Site, 50 p.
Nichols, R.L., Hamm, Larry, and Jones, W.F., 1995,
Numerical modeling of the TNX area hybrid
groundwater corrective action: South Carolina,
Westinghouse Savannah River Company Report
RC-RP-95-787, Savannah River Site, 28 p.
Obrien & Gere Engineers, Inc., 1997a, Passive bag
sampling results, JMT Facility, Brockport, New
York: Albany, NY, Consultant's report to General
Electric Company, October 10, 1997, 10 p.
1997b, Passive bag sampling results, JMT
Facility, Brockport, New York: Albany, NY,
Consultant's report to General Electric Company,
December 12, 1997, 10 p.
Olafsson, J., 1979, Physical characteristics of Lake
Myvatn and River Laxa: Oikos, v. 32, p. 38-66.
Phifer, M.A., Nichols. R.L., May, C.P, Pemberton,
B.E., and Sappinton, F.C., 1995, Final report, TNX
test recovery well pump test, TNX unconfined
aquifer, well TRW-1 (U): South Carolina,
Westinghouse Savannah River Company report
WSRC-TR-95-0132, Savannah River Site, 83 p.
Rosenberry, D.O., Striegl, R.G., and Hudson, D.C.,
2000, Plants as indicators of focused ground water
discharge to a northern Minnesota Lake: Ground
Water, v. 38, no. 2, p. 296-303.
60 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Roy F. Weston, Inc., 1998, Groundwater sampling,
Nyanza Chemical Waste Dump Superfund Site,
Ashland, Massachusetts: Manchester, NH,
variably paged.
1999a, Semi-annual groundwater monitoring
report, Spring 1999, Nyanza Chemical Waste
Dump Superfund Site, Ashland, Massachusetts,
draft report: Manchester, NH, variably paged.
1999b, Evaluation of contamination, Nyanza
Chemical Waste Dump Superfund Site, Ashland,
Massachusetts: Manchester, NH, variably paged.
Savoie, Jennifer, 1995, Altitude and configuration of
the water table, western Cape Cod, Massachusetts,
March 1993: U.S. Geological Survey Open-File
Report 94-462, 1 sheet.
Savoie, J.G., LeBlanc, D.R., Blackwood, D.S.,
McCobb, T.D., Rendigs, R.R.,and Scott Clifford,
2000, Delineation of discharge areas of two
contaminant plumes by use of diffusion samplers,
Johns Pond, Cape Cod, Massachusetts, 1998: U.S.
Geological Survey Water-Resources Investigations
Report 00-4017, 30 p.
Savoie, J.G., Lyford, F.R, and Scott Clifford, 1999,
Potential for advection of volatile organic
compounds in ground water to the Cochato River,
Baird & McGuire Superfund Site, Holbrook,
Massachusetts, March and April 1998: U.S.
Geological Survey Water-Resources Investigations
Report 98-4257, 19 p.
Sevee & Mahar Engineering, Inc., 1998, Subsurface
investigations report, McKin Superfund Site/Gray
Depot Area, Gray, Maine: Cumberland Center,
Maine, 19 p.
1999, Data transmitted and site conceptual
model description, McKin Superfund Site:
Cumberland Center, Maine, variously paged.
Silliman, S.E., and Booth, D.F., 1993, Analysis of time-
series measurements of sediment temperature for
identification of gaining vs. losing portions of
Juday Creek, Indiana: Journal of Hydrology,
v. 146, p. 131-148.
Swanson, G.A., Adomaitis, V.A., Lee, F.B., Serie, J.R.,
and Shoesmith, J.A., 1984, Limnological
conditions influencing duckling use of saline lakes
in south-central North Dakota: Journal of Wildlife
Management, v. 48, p. 340-349.
Toth, J., 1963, A theoretical analysis of groundwater
flow in small drainage basins: Journal of
Geophysical Research, v. 68, no. 16, p. 4795-
4812.
Vaux, W.G., 1968, The flow and interchange of water
in a streambed: U.S. Fish and Wildlife Bulletin 66,
p. 479^89.
Verhoeven, J.T.A., Koerselman, W., and Beltman, B.,
1988, The vegetation of fens in relation to their
hydrology and nutrient dynamics, in Symoens, J.J.
ed., A case study, in vegetation of inland waters:
Dordrecht, Kluwere Academic Publishers,
p.249-282.
Vroblesky, D.A., 2000, Influence of stream orientation
on contaminated ground-water discharge; in
Proceedings of the Ground Water/Surface Water
Interactions Workshop, January 26-28, 1999:
Denver, CO, U.S. Environmental Protection
Agency, p. 143-147.
Vroblesky, D.A., 2001, User's guide for polyethylene-
based passive diffusion bag samplers to obtain
volatile organic compound concentrations in
wells—Pari I. Deployment, recovery, data
interpretation, and quality control and assurance:
U.S. Geological Survey Water-Resources
Investigations Report 01-4060, 18 p.
Vroblesky, D.A., and Campbell, T.R., 2001,
Equilibration times, compound selectivity, and
stability of diffusion samplers for collection of
ground-water VOC concentrations: Advances in
Environmental Research, v. 5, p. 1-12.
Vroblesky, D.A., and Hyde, W.T., 1997, Diffusion
samplers as an inexpensive approach to
monitoring VOCs in ground water: Ground Water
Monitoring and Remediation, v. 16, no. 3, p. 177-
184.
Vroblesky, D.A., andLorah, M.M., 1991, Prospecting
for zones of contaminated-ground-water discharge
to streams using bottom-sediment gas bubbles:
Groundwater, v. 29, no. 3, p. 333-340.
Vroblesky, D.A., Lorah, M.M., and Trimble, S.P,
1991, Mapping zones of contaminated ground-
water discharge using creek-bottom-sediment
vapors, Aberdeen Proving Ground, Maryland:
Groundwater, v. 29, no. 1, p. 7-12.
Vroblesky, D.A., Nietch, C.T., Robertson, J.F., Bradley,
P.M., Coates, John, and Morris, J.T., 1999, Natural
attenuation potential of chlorinated volatile
organic compounds in ground water, TNX
floodplain, Savannah River Site, South Carolina:
U.S. Geological Survey Water-Resources
Investigations Report 99-4071, 43 p.
References Cited 61
-------
Vroblesky, D.A., Rhodes, L.C., Robertson, J.F.,
and Harrigan, J.A., 1996, Locating VOC
contamination in a fractured-rock aquifer at the
ground-water/surface-water interface using
passive vapor collectors: Groundwater, v. 34,
no. 2, p. 223-230.
Vroblesky, D.A., and Robertson, J.F., 1996, Temporal
changes in VOC discharge to surface water from a
fractured rock aquifer during well installation and
operation, Greenville, South Carolina: Ground
Water Monitoring and Remediation, v. 16, no. 3,
p. 196-201.
Vroblesky, D.A., Robertson, J.F., Fernandez, Mario,
and Aelion, C.M., 1992, The permeable-
membrane method of passive soil-gas collection;
in Proceedings of the Sixth National Outdoor
Action Conference: Las Vegas, Nev., National
Water Well Association, May 5-13, 1992, p. 3-26.
Vroblesky, D.A., Robertson, J.F., Petkewich, M.D.,
Chapelle, F.H., Bradley, P.M., and Landmeyer,
J.E., 1997b, Remediation of petroleum
hydrocarbons-contaminated ground water in the
vicinity of a jet-fuel tank farm, Hanahan, South
Carolina: U.S. Geological Survey Water-
Resources Investigations Report 96-4251, 61 p.
Wassen, M.J., Barendregt, A., Bootsma, M.C., and
Schot, P.P., 1989, Ground water chemistry and
vegetation of gradients from rich fen to poor fen in
the Naardenmeer (the Netherlands): Vegetatio,
v. 79, p. 117-132.
Winter, T.C., 1976, Numerical simulation analysis of
the interaction of lakes and ground water: U.S.
Geological Survey Professional Paper 1001, 45 p.
Winter T.C., 1978, Numerical simulation of steady
state three-dimensional ground-water flow near
lakes: Water Resources Research, v. 14, no. 2,
p. 245-254.
Winter, T.C., LaBaugh, J.W., and Rosenberry, D.O.,
1988, Design and use of a hydraulic
potentiomanometer for direct measurement of
differences in hydraulic head between ground
water and surface water: Limnology and
Oceanography, v. 33, no. 5, p. 1209-1214.
62 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Appendix 1. Laboratory and Field Testing of
Passive-Vapor-Diffusion Sampler Equilibration
Times, Temperature Effects, and Sample Stability
By Don A. Vroblesky
-------
-------
Laboratory and field tests were conducted to
determine the time required for the vapor in the air-
tilled vials of the PVD samplers to equilibrate with
concentrations of VOCs in the pore water of the adja-
cent sediment, and to determine the stability of VOC
concentrations between the time of sample recovery
and sealing of the sample vials. For the laboratory and
field tests, the PVD samplers consisted of a 20-mL
serum vial enclosed in a heat-sealed, 1.5-in. diameter
LDPE tube. The vial was arranged such that a single
layer of polyethylene was held tightly in place over the
vial opening. The tubing was secured to the vial by a
self-locking nylon tie. The assembly then was placed
inside a second tube and heat sealed, trapping a mini-
mum of air. Samplers used in the field were attached
with self-locking nylon ties to a wire surveyor flag to
mark the sampling site and to facilitate sampler
recovery.
Upon recovery of the PVD samplers, the outer
tube was cut open, leaving the inner tubing intact.
A vial cap with a Teflon-coated stopper then was
crimped onto the vial and inner bag. Vapor samples
were obtained from the PVD samplers by inserting a
syringe needle tip through the Teflon-coated stopper
beneath the vial cap and extracting the vapor with a
gas-tight syringe. Analysis of the the vapor was ana-
lyzed by photoionization detection with a Photovac
lOSPlus gas chromatograph.
Equilibration Times and
Temperature Effects
Equilibration times of PVD samplers under labo-
ratory conditions were examined at 21°C and 10°C. For
the test at 21°C, groups of three PVD samplers were
added to 480-mL water-filled test jars spiked with a
mix of VOCs at concentrations of approximately 20 to
100 |ig/L each of cis-1,2-dichloroethene (ds-DCE),
benzene, TCE, toluene, and PCE, in February 1998.
The jars were stored at a 21°C for the duration of the
test. For the test at 10°C three PVD samplers were
added to each of fourteen 1.9-L test jars spiked with a
mix of benzene (570 |ig/L), toluene (520 |ig/L), TCE
(430 |ig/L), and PCE (500 |ig/L), in June 2001. The test
jars were maintained at approximately 10°C by storing
them in an incubator. Water temperatures in the jars
were measured at each sampling time and ranged from
7.6 to 11.3°C (average of 10.2°C). The water-filled
jars in both tests contained no headspace.
An additional test was done to compare differ-
ences in equilibrium concentrations at various tempera-
tures. Standards of toluene and PCE were added to
water in a 40-L container. The PVD samplers were
placed in 1.9-L jars in groups of three, and the jars
were filled with test water by submerging and capping
the jars underwater in the 40-L container. The jars were
stored at different temperatures. At various times, one
jar from each temperature treatment was opened, the
temperature was measured, the PVD samplers were
capped, and a water sample was collected for toluene
and PCE laboratory analysis. Vapor samples from the
PVD samplers were analyzed by headspace analysis.
Water temperatures ranged from 0.2 to 4.1°C (average
of 1.4°C) in the coldest treatment, from 7.9 to 10.8°C
(average of 9.5°C) in the next coldest treatment, and
from 19.4 to 24.6°C (average of 22.4°C) in the room-
temperature treatment. Concentrations in water ranged
from 210 to 310 |ig/L of toluene and 110 to 340 |ig/L
of PCE.
Recovery of the PVD samplers at each sampling
time consisted of opening the jar, removing the PVD
samplers, cutting open the external LDPE bag, and
sealing the PVD samplers with crimp-type caps with
Teflon-faced seals. The test at 21°C involved 10 sam-
pling times over approximately 59 hours (table 1 A),
the test at 10°C involved 13 sampling times and 1
duplicate over approximately 222 hours (table 15), and
the test at 3 temperatures involved 6 sampling times
and 1 duplicate sample over 456 hours (table 1C).
In addition, the USGS conducted relatively
short-term tests of PVD samplers deployed in stre-
ambed sediments in areas of VOC-contaminated
ground-water discharge to examine sampler equilibra-
tion dynamics under field conditions. For these tests,
the samplers were deployed in different environments.
Sitel was in Coastal Plain sands at the Savannah River
Site, South Carolina, in an area with a relatively high
upward hydraulic gradient (0.3 ft/ft). The sediment at
site 1 had a relatively large hydraulic conductivity of
21 to 65 ft/d (Nichols, 1993; Nichols and others, 1995;
Phifer and others, 1995). Site 2 was in silty saprolite
downgradient from a former waste-drum disposal site
in a part of the stream with a relatively low upward
hydraulic gradient (0.02 ft/ft).
Appendix 1 65
-------
Table 1 A. Average concentrations and standard deviations of volatile organic compounds in passive-vapor-diffusion samplers
over time at 21 degrees Celsius under laboratory conditions in 480-milliliter test jars with spiked concentrations of less than 100
micrograms per liter
[ds-DCE, cis-1,2-dichloroethene; TCE, trichloroethene; ppb, parts per billion]
Time,
in hours
Number of
Average concentrations, in ppb by volume
Standard deviation, in ppb by volume
samplers
cis- DCE
Benzene
TCE
Toluene
cis- DCE
Benzene
TCE
Toluene
0
3
0
0
0
0
0
0
0
0
0.75
3
158
223
118
146
23
77
52
75
1.75
3
431
498
171
110
124
175
73
40
2.75
3
654
939
271
218
159
393
103
119
5.75
3
687
1,054
297
276
192
300
71
50
10.75
3
1,416
2,058
492
565
315
669
209
316
23.75
3
1,970
3,101
813
913
14
64
26
27
31.75
3
1,880
3,044
802
975
68
121
66
122
51.75
3
1,853
2,923
732
929
54
86
29
52
58.75
3
1,705
2,747
695
879
34
58
10
24
TablelB. Average concentrations and standard deviation of volatile organic compounds in passive-vapor-diffusion samplers
over time at 10 degrees Celsius under laboratory conditions in 1.9-liter test jars with spiked concentrations ranging from 430 to
570 micrograms per liter
[PCE, tetrachloroethene; TCE, trichloroethene; D, Duplicate test jar; NA, not applicable; ppb, parts per billion]
Time, in
hours
Number of
Average concentrations, in ppb by volume
Standard deviation, in ppb by volume
samplers
Benzene
TCE
Toluene
PCE
Benzene
TCE
Toluene
PCE
0
3
0
0
0
0
0
0
0
0
2.25
3
150
303
0
0
10
35
0
0
5.25
3
823
1,065
333
69
359
514
90
15
9.25
3
1,880
2,130
1,185
334
1,655
1,451
540
81
14.25
3
6,166
4,999
2,853
700
3,877
2,762
1,177
165
28.75
2
20,265
14,395
9,787
2,432
318
389
896
144
35.25
2
21,180
14,725
9,737
2,347
1,414
2,539
655
166
57.25
1
25,450
24,060
15,110
4,223
NA
NA
NA
NA
78.25
3
31,227
33,523
20,360
6,179
2,177
4,393
767
313
102.25
3
31,990
35,293
22,480
7,054
652
1,979
1,644
612
150.25
2
31,220
34,275
23,320
7,867
438
1,025
184
522
150.25 D
1
32,150
37,140
21,430
7,229
NA
NA
NA
NA
222.25
3
31,543
35,253
21,900
7,465
957
1,303
1,763
584
66 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Table 1C. Ratio of concentrations from passive-vapor-diffusion sampler to aqueous-phase concentrations for toluene and
tetrachloroethene over time at various temperatures under laboratory conditions in 1.9-liter test jars containing 210 to
310 micrograms per liter of toluene and 110 to 340 micrograms per liter of tetrachloroethene
[ppb v to |lg/L, ratio of parts per billion by volume to micrograms per liter; °C, degrees Celsius; —, no samples]
Ratio of passive-vapor-diffusion to aqueous-phase concentrations (ppb v to pg/L)
Hours Toluene Tetrachloroethene
22.4C 9.5C 1.4C 22.4C 9.5C 1.4C
0 0.0 0.0 0.0 0.0 0.0 0.0
24 76.9 11.3 15.5 14.4 3.9 2.2
72 116.1 73.4 19.7 60.1 35.9 6.6
192 107.3 62.7 21.0 67.6 35.9 11.7
336 105.3 60.0 27.9 81.3 39.0 17.3
336 100.3 55.8 - 69.8 46.7
456 103.5 49.8 23.4 68.4 33.9 17.2
Hydraulic gradients at each site were determined
by driving 1-in-diameter steel pipes into the streambed
to depths of 1.5 and 3 ft. A bolt loosely seated into the
downward end of the pipes prevented sediment from
moving up into the pipe as it was driven into the sedi-
ment. The bolt then was driven out of the pipe to allow
water to enter. After a few hours of equilibration, the
water levels in the pipes were measured relative to the
stream stage outside the pipes. These water levels pro-
vided an approximate measurement of the upward
hydraulic gradient. At each field site, approximately 20
to 30 PVD samplers were buried in an area of approxi-
mately 6 ft2. The samplers were recovered in groups of
three at various times over approximately 50 hours.
Equilibration time for PVD samplers under field
conditions depends on the time required by the PVD
sampler to equilibrate with ambient water, and the time
required for the contaminant distribution in pore water,
disturbed by installation of the sampler, to return to
ambient conditions. Laboratory tests presented here
provide information regarding the time required by the
PVD sampler to equilibrate with ambient water. Field
data for this and other investigations involving various
types of in-situ samplers can provide a better under-
standing of the time required for sediment-pore-water
concentrations to equilibrate after disturbances caused
by sampler deployment.
Temperature is one primary factor affecting the
equilibration time of VOC movement across a polyeth-
ylene membrane. At 21 °C under laboratory conditions,
the time required for concentrations of cw-DCE, ben-
zene, TCE, toluene, and PCE to stabilize was approxi-
mately 24 hours (fig. 1 A, table 1A). At 10°C under
laboratory conditions, concentrations of benzene, TCE,
toluene, and PCE appeared to require about 102 hours
to equilibrate (fig. IB, table 15). Equilibration times,
therefore, increase as temperature decreases. This
increase in equilibration time with decreasing tempera-
ture is consistent with the general trend previously
shown for sampling major ions and nutrients with dial-
ysis samplers (Carignan, 1984). In these studies, equili-
bration times ranged from approximately 15 days in a
warm (20-25 °C) environment, and approximately 20
days in a cold (4-6 °C) environment.
Appendix 1 67
-------
LU
3
o
>
CO
CO
QC
LU
CL
CO
5,000
4,000
3,000
QC
£ 2,000
< 1,000
LU
o
z
o
o
c/s-1,2-Dichloroethene
Benzene
Trichloroethene
Toulene
HOURS
Figure 1 A. Changes in concentrations of volatile organic compounds in passive-vapor-diffusion
samples over time at 21 degrees Celsius under laboratory conditions in a mixed solution of volatile
organic compounds with aqueous concentrations less than 100 micrograms per liter.
Temperatures also change the equilibrium distri-
bution of VOCs between the aqueous and vapor phase.
At cold temperatures, there is less of a tendency for
volatilization than at warmer temperatures. For the
same aqueous concentration, therefore, higher concen-
trations will be detected in PVD samplers at warm tem-
peratures than at cold temperatures (fig. 1C, table 1C).
In contrast, differences in VOC concentrations
between tests or sites theoretically should not produce
different equilibration times. The concentration gradi-
ent between the inside and outside of the polyethylene
membrane does affect the diffusion flux, as evidenced
by Fick's Law; however, although a steeper concentra-
tion gradient results in an increased diffusion rate, the
required amount of solute transfer is larger than in a
lower concentration gradient. In effect, the equilibra-
tion time is the same for both situations. This fact can
be seen in the formula for halftime of chemical uptake
of organic compounds across a polyethylene mem-
brane in semipermeable membrane devices (SPMDs):
to.5u — -In 0.5 (Kow Vi/Rs),
where to.5u is the halftime to equilibration, K(IW is the
octanol-water coefficient, Vl is the volume of the lipid
sorbent, and RS is the volume of water sampled by the
SPMD per day and is independent of concentration
(Huckins and others, 1997).
Field tests for this investigation demonstrate
that equilibration times can vary among field sites. At
site 1 in South Carolina, the TCE concentrations in the
PVD samplers approximately stabilized after about
12 to 24 hours (fig. ID, table ID). At site 2 in South
Carolina, however, the TCE concentrations still were
increasing at 47 hours (fig. ID, table ID). A field test in
Massachusetts found that PVD samplers in sediments
of the Sudbury River appeared to require 3 weeks or
more to equilibrate (Lyford and others, 2000). These
differences in equilibration times can result from a
variety of factors. At site 1, the rate of water movement
past the samplers was substantially larger than the rate
68 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
50,000
40,000
30,
20,
HOURS
Benzene
Tricholorethene
Toluene
Tetrachloroethene
Figure 1B. Changes in concentrations of volatile organic compounds in passive-vapor-diffusion
samples over time at 10 degrees Celsius under laboratory conditions in a mixed solution of volatile
organic compounds with aqueous concentrations ranging from 430 to 570 micrograms per liter.
past the slower equilibrating samplers at site 2, as evi-
denced by the differences in permeability and hydrau-
lic gradient (fig. ID). The surface water introduced into
the excavated sampler holes would have been more
rapidly replaced by discharging ground water at site
1 than at site 2. Additionally, site 1 was sampled in
early September, and site 2 was sampled two months
later, in November. Colder water temperatures in
November may have contributed to the decrease in
equilibration time. The samplers from the Sudbury
River in Massachusetts were recovered from fine-
grained sediment in February 1999, in water that was
substantially colder than at either of the South Carolina
sites.
An important concept to remember in consider-
ing the length of equilibration time for field deploy-
ment, however, is that PVD samplers typically are used
as a reconnaissance tool in surface-water sediments to
locate areas of discharging ground water contaminated
with VOCs. For this use, the mere presence or absence
of target VOCs in samplers can provide practical
information. At the field locations tested in South
Carolina, the PVD samplers showed considerable con-
centrations of the target VOCs within 24 hours or less
(fig. ID). Despite the apparent lack of equilibration, the
PVD samplers in the Sudbury River showed consider-
able concentrations by the first retrieval (8 days).
Although the PVD samples had not equilibrated at all
of the sites, the data from all of the sites were adequate
to meet the objective of mapping contaminant-dis-
charge areas. These data suggest that a deployment
period of 8 days or less is adequate to provide data suf-
ficient to locate major VOC discharge areas beneath
surface water, and as little as 24 hours is sufficient at
some sites.
If it is important for the data-quality objective
that the PVD samples must have reached equilibrium
at the time of sampler recovery, then the samplers
may need to remain in place longer than 8 days,
depending on the rate of ground-water movement and
Appendix 1 69
-------
A. Toluene
CO
DC
LU
CL
C/)
2
<
DC
LU
o
z
o
o
DC
o
CL
%
DC
LU
DC
LU
CL
W
<
DC
CD
O
DC
O
<
DC
LU
o
z
o
o
W
=>
o
LU
=>
0
<
500
100
B. Tetrachloroethene
100
200 300
HOURS
400
500
TEMPERATURE, IN DEGREES CELSIUS
• 22.4°C
O
9.5°C
1.4°C
Figure 1C. Ratio overtime of (A) toluene and (S) tetrachloroethene gas
concentrations by volume (parts per billion) in passive-vapor-diffusion samplers to
aqueous concentrations by mass (210 to 310 micrograms per liter of toluene and
110 to 340 micrograms per liter of tetrachloroethene) in a test solution containing
the diffusion samplers in 1.9-liter jars at average temperatures of 22.4, 9.5, and
1.4 degrees Celsius.
the ground-water temperature. Some insight into equil-
ibration times can be gained by examining passive-
diffusion-bag (PDB) samplers, which are water-filled
polyethylene diffusion samplers. Under laboratory con-
ditions at 21°C, the PDB samplers equilibrate more
slowly (approximately 48 hours; Vroblesky, 2000) than
PVD samplers (approximately 24 hours); therefore, the
field equilibration times of PVD samplers probably are
no longer than those for PDB samplers under similar
field conditions. One field investigation showed ade-
quate equilibration of PDB samplers to aquifer TCE
and carbon tetrachloride within 2 days in a highly per-
meable aquifer (Vroblesky and others, 1999). In other
investigations, PDB samplers recovered after 14 days
were found to be adequately equilibrated to chlorinated
VOCs (Obrien & Gere Engineers, Inc., 1997a, 1997b;
Hare, 2000); therefore, the equilibration time was 14
days or less under those field conditions. Because it
appears that 2 weeks of equilibration time probably is
adequate for many applications in permeable forma-
tions, a minimum PVD-equilibration time of 2 weeks is
recommended for discharge areas in sandy sediment.
70 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
QC
LU
CL
CO
I—
QC LU
z >-
LU CD
ii
o m
_i
o
QC
8,000
A. Coastal Plain
6.000
4,000
S TE 1. Unconso hdated sands
2.000
0 5 10 15 20 25 30 35 40 45 50 55
B. Piedmont
1 1 1 1 1 1 1 1 1 Ff
1.000
SITE 2. Silty saprolite
lv = 0.02
10 15 20 25 30 35 40 45 50 55
HOURS
Figure 1D. Changes in trichloroethene concentrations overtime in passive-vapor-
diffusion samples from contaminated ground-water discharge areas in South
Carolina in (Site 1) Coastal Plain sediments and (Site 2) Piedmont sediments with
differing sediment types and vertical hydraulic gradients.
Equilibration of PVD samplers in fine-grained
sediment may take longer than 2 weeks, as shown
by the test in the Sudbury River (Lyford and others,
2000). Equilibration times for PVD samplers in poorly
permeable sediment have not yet been determined.
A confounding factor in determining equilibration
times in such conditions is that flow conditions and
possibly plume concentrations can vary with time and
space, and these variations may be reflected in the
PVD-sampler concentrations.
In summary, if the data-quality objective is to
locate areas of VOC-contaminated ground-water
discharge, then a deployment period of 8 days or less is
adequate to provide data sufficient to locate major VOC
discharge areas beneath surface water, and as little as
24 hours is sufficient at some sites. If it is important for
the data-quality objective that the PVD samplers must
have reached equilibrium at the time of sampler recov-
ery, then the samplers may need to remain in place for
about 2 weeks in sandy sediment and possibly longer
than 3 weeks in poorly permeable sediment at cold
temperatures.
Sample Stability
Laboratory tests also were conducted to deter-
mine the diffusion loss of VOCs in PVD samplers
between the time of sampler recovery and sealing of
the sampler vials. The samplers were allowed to equili-
brate for 2 weeks in a water-filled container having
mixed VOCs. The samplers then were removed from
the water and allowed to stand at 21°C for various time
intervals over a period of hours before capping. The
vapor samples then were analyzed by photo-ionization
gas chromatography
Concentrations of benzene, TCE, and toluene in
uncapped PVD samplers at 21°C did not substantially
decrease over 60 minutes between sampler recovery
and capping the sampling vials. These data suggest that
VOC concentrations within the uncapped PVD sam-
plers are relatively stable for at least 1 hour at 21°C
under laboratory conditions. Under field conditions,
however, abrasions and other factors may adversely
affect the membrane. Consequently, it is strongly rec-
ommended that PVD samples be capped and sealed
immediately upon recovery
Appendix 1 71
-------
Table 1D. Average concentrations and standard deviations of trichloroethene over time in passive-vapor-diffusion samplers in
bottom sediment of streams at contaminated ground-water-discharge areas in the Coastal Plain (site 1) and the Piedmont
(site 2) of South Carolina, 1998
[TCE, trichloroethene; ppb, parts per billion; NA, not applicable]
Site 1 Site 2
Hours after
deployment
Number
TCE concentration,
in ppb by volume
Hours after
deployment
Number
TCE concentration,
in ppb by volume
of samples
Average
Standard
deviation
of samples
Average
Standard
deviation
0
1
0
NA
0
1
0
NA
.08
3
5
4
.08
3
0
0
.5
3
307
60
1
3
0
0
1
3
378
146
2
3
6
6
2
3
815
117
3
3
47
19
3
2
1,431
313
4.6
3
92
21
5
3
1,208
270
9
3
136
61
7
3
1,601
516
12.75
3
400
156
10
3
4,710
1,103
17
3
286
120
24
3
5,693
1,300
23
3
586
127
31
3
5,002
1,167
41
3
647
78
47
2
5,019
1,758
50.5
3
894
91
72 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Appendix 2. Field Screening of Volatile Organic
Compounds Collected with Passive-Vapor-Diffusion
Samplers with a Gas Chromatograph
By Scott Clifford, U.S. Environmental Protection Agency
-------
-------
1.0 Summary of Method
Field screening with a portable gas chromato-
graph (GC) provides tentative identification and esti-
mated concentrations of volatile organic compounds
(VOCs) in air samples. This screening technique can
provide quick and reliable results to assist in important
onsite decision making. An aliquot of the air/vapor
sample is injected into a calibrated GC equipped with a
photoionization detector and electron capture detector.
Results indicating VOC detections and estimated con-
centrations are displayed on a chart recorder, a portable
computer screen, or other data-collection system.
Retention times, the time since injection of the sample
into the GC to the time peak responses are recorded,
are used for compound identification. The peak
heights, or areas, are used to estimate concentrations of
the identified compounds. Reporting levels (RL) can
vary depending upon instrument performance and set-
tings, or can be set on the basis of data quality objec-
tives. Typical achievable reporting levels for many
VOCs of interest are shown in table 2A. Concentrations
are in units of parts per billion by volume (ppb v).
2.0 Scope and Application
The procedure described here is applicable
to chemists performing screening for VOCs in air
samples collected with passive-vapor-diffusion (PVD)
samplers.
3.0 Definitions
3.1 FIELD DUPLICATES (FD1 and FD2):
Two separate samples collected at the same time and
place under identical circumstances and treated exactly
the same throughout field and laboratory procedures.
Analyses of FD1 and FD2 give a measure of the preci-
sion associated with sample collection, preservation,
and storage, as well as with laboratory procedures.
3.2 HEADSPACE: Air above water standard in
sample vial.
3.3 LABORATORY DUPLICATE (LD1 and
LD2): Two injections from the same sample. The anal-
yses of LD1 and LD2 give a measure of the precision
associated with the laboratory procedure.
3.4 LABORATORY REAGENT BLANK
(LRB): An aliquot of reagent water or other blank
matrix that is treated exactly as a sample including
exposure to all glassware, equipment, solvents,
reagents, internal standards, and surrogates that are
used with other samples. The LRB is used to determine
if method analytes or other interferences are present
in the laboratory environment, the reagents, or the
analytical apparatus.
3.5 FIELD BLANK: A PVD sampler left out in
the ambient air (for example, attached to a tree) for the
duration of the sample-collection period.
3.6 STOCK STANDARD SOLUTION: A con-
centrated solution containing one or more method ana-
lytes prepared in the laboratory by use of using assayed
reference materials or purchased from a reputable
commercial source.
3.7 WORKING STANDARD SOLUTION: A
solution of several analytes prepared in the laboratory
from stock standard solutions and diluted as needed to
prepare calibration solutions and other needed analyte
solutions.
3.8 SECONDARY STANDARD: A standard
from another vender or a different lot number that is
used to check the primary standard used for estimating
concentrations.
4.0 Health and Safety
Warnings
4.1: The toxicity or carcinogenicity of each
reagent used in this method has not been precisely
determined; therefore, each chemical should be treated
as a potential health hazard. Exposure to these reagents
should be reduced to the lowest possible level. The
laboratory is responsible for maintaining a current
awareness file of the Occupational Safety and Health
Administration (OSHA) regulations regarding the safe
handling of the chemicals specified in this method. A
reference file of data-handling sheets should be made
available to all personnel involved in these analyses.
Use these reagents in a fume hood whenever possible;
and if eye or skin contact occurs, flush with large
volumes of water.
4.2: Always wear safety glasses or a shield for
eye protection, wear protective clothing, and observe
proper mixing when working with these reagents.
Appendix 2 75
-------
Table 2A. Typical achievable reporting limits for volatile
organic compounds commonly detected in passive-vapor-
diffusion samplers from a gas chromatograph equipped with
a photoionization detecter and an electron capture detector
Reporting limit,
Volatile organic compound in parts per billion
by volume
Benzene 10
Ethylbenzene 50
meta/para-x ylenes 50
ortho-x ylene 80
Toluene 40
Chlorobenzene 50
Trichloroethylene 10
Tetrachloroethylene 2
1,1 -Dichloroethene 10
ds-l^-Dichloroethene 15
fraras-l^-Dichloroethene 10
1,1,1-Trichloroethane 6
4.3: Some method analytes have been tentatively
classified as known or suspected human or mammalian
carcinogens. Pure standard materials and stock stan-
dard solutions of these compounds should be handled
with suitable protection to skin and eyes.
5.0 Personnel Qualifications
5.1: The analyst should have at least a 4-year
degree in a physical science.
5.2: The analyst should be trained at least 1 week
and have a working knowledge of this method and
quality control before initiating the procedure.
5.3: All personnel shall be responsible for
complying with all quality-assurance/quality-control
requirements that pertain to their organizational/
technical function.
6.0 Equipment and Supplies
6.1: Gas chromatograph equipped with a photo-
ionization detector (PID) in series with an electron cap-
ture detector (ECD), and an analytical column capable
of separating target analytes.
6.2: Data-collection system (for example, a chart
recorder, intergrator or a portable computer).
6.3: Syringes: steel barrel, volume of 250 |xL to
500 |xL; 2 in., 25-gauge needle.
6.4: Vials: 40 mL volatile organic analysis
(VOA) vials with Teflon lined septum caps.
7.0 Preparation of Air and
Aqueous Standards
7.1: The air standard for each target analyte is
the headspace above a 10 |ig/L aqueous standard at
approximately 0 to 1°C for each target analyte. Aque-
ous standards are prepared with target analyte concen-
trations of 10 |ig/L and are stored on ice in 40 mL VOA
vials with no head space. When ready for use, lOmL of
the aqueous standard are withdrawn from the vials to
produce a headspace from which vapors are analyzed
and serve as air standards. The vials are then placed
into an ice bath in a cooler with the septa side facing
downward, and left to equilibrate for approximately
30 minutes. Because the concentration of the volatile
organic compounds in the head space was calibrated at
approximately 0 to 1°C, the working standards must be
maintained at the same temperature. Vapor-standard
concentrations of volatile organic compounds com-
monly detected with passive-vapor-diffusion samplers
in the headspace of a 10 |ig/L aqueous standard at
approximately 0 to 1°C have been determined through
in-house experimentation and are shown in table 25.
76 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Table 2B. Vapor concentrations of volatile organic
compounds commonly detected with passive-vapor-diffusion
samplers in the head space of a 10 micrograms per liter
aqueous standard at approximately 0 to 1 degree Celsius
Volatile organic compound
Concentration, in parts
per billion by volume
Benzene
Ethylbenzene
meta/para-x ylenes
ortho-x ylene
Toluene
Chlorobenzene
Trichloroethylene
Tetrachloroethylene
1,1 -Dichloroethene
cw-l^-Dichloroethene
trans-1,2-Dichloroethene..
1,1,1-Trichloroethane
151
145
136
112
159
70
142
201
554
90
202
330
7.2: Aqueous standards are made weekly from a
methanol stock solution and volatile organic-free
water, and stored in a 40-mL VOA vial on ice with no
head space.
7.3: The methanol stock standard and secondary
stock standard are replaced every 3 months.
7.4: The aqueous working and secondary stan-
dards are good for 7 days provided these standards are
stored on ice with no headspace.
8.0 Instrument Operation
Gas chromatographs are available from many
commercial sources. Operation of any particular
gas chromatograph should be in accordance with
procedures supplied by the manufacturer.
9.0 Instrument Calibration
Calibration for analysis of vapor samples col-
lected with PVD samplers generally consists of taking
a 200 |xL volume of the headspace in the 10 |ig/L
aqueous, 0 to 1°C standard with a 250 |xL steel-barrel
syringe with a 2 in., 25-gauge needle, and injecting it
into the injection port of the GC. The actual vapor con-
centration is listed in table 2B for selected compounds.
This single point calibration is completed by analyzing
a blank, which is obtained from the headspace above a
volatile organic-free water sample in a 40 mL VOA
vial.
10.0 Analysis
10.1: Insert the syringe needle into the PVD-
sampler septum and use the plunger to flush the syringe
barrel three times. After flushing, pull the plunger up to
the 200 |iL point on the barrel and remove the needle
from PVD sample.
10.2: Insert the syringe needle into the GC
injection port, and push the needle through the septum
until the barrel comes up against the injection port,
immediately push the plunger with a quick action,
remove needle from injection port, and turn on the
data-collection system.
10.3: Record the following information:
1. start of run,
2. sample number,
3. sample volume,
4. attenuation or gain, and
5. other relevant comments.
10.4: The order in which of a group of samples is
analyzed is as follows:
10.4.1: Calibration Standard—Inject a
200-|xL sample of 10-|ig/L standard at 0 to 1°C head-
space into the GC. Keep standard peaks at approxi-
mately 50 percent scale or more, if possible, by
adjusting the attenuation or gain.
10.4.2: Repeat 10-|ig/L standard to check
for reproducibility. Standard chromatograms should
have compound peak heights within 15 percent of each
other and retention times should be identical.
10.4.3: Inject the secondary standard for
confirmation. The acceptance criteria is within 20
percent of the true value.
Appendix 2 77
-------
10.4.4: Clean air blank—Inject a 200-|xL
sample of clean air into the gas chromatograph with the
attenuation set at the same level or lower than the level
at which samples will be run. Blank clean air is taken
from the headspace above volatile organic-free water in
a 40-mL VOA vial.
10.4.5: Samples—Inject 200-|xL sample
volume into the GC at the same attenuation level at, or
lower than the attenuation at which the standard was
run. If contaminant levels on the chromatograms are
off-scale, adjust the attenuation or gain to decrease
instrument response. If the chromatographic peaks
are still off-scale, rerun the samples with a smaller
injection volume.
10.4.6: Repeat 10-|ig/L calibration standard
every 10 samples and at the end of the sample batch to
check the calibration and reproducibility. Standard
chromatograms should have compound peak heights
within 20 percent of each other, and retention times
should be identical.
11.0 Identification and
Estimated Concentration
11.1: Identify compounds present in a sample by
matching retention times of peaks in the sample chro-
matogram to the retention times of standard peaks
determined at site.
11.2: Concentrations are estimated by a peak-
height or peak-area comparison. For example, if the
10-|ig/L aqueous-standard head space had a benzene
peak height of 32 units from a 200-|xL injection with
instrument attenuation at 2, an identified benzene peak
12 units high from a 200-|xL sample injection with
instrument attenuation at 2 would represent a sample
benzene concentration of 57 ppb v.
32 units = 12 units
*151 ppb v Xppbv
X = 57 ppb v Benzene
* See Air Standard Section 7.1 (table IB)
12.0 Interferences
12.1: Method interferences may be caused by
contaminants in solvents, reagents, glassware, and
other sample-processing hardware that lead to discrete
artifacts and (or) elevated baselines in the chromato-
grams. All of these materials must routinely be demon-
strated to be free from interferences under the
conditions of the analysis by running laboratory and
field method blanks.
12.2: Matrix interferences may be caused by
contaminants that coelude with the target compounds.
The extent of matrix interferences will vary consider-
ably from source to source. A different column or
detector may eliminate this interference.
12.3: Contamination by carry-over can occur
whenever high-level and low-level samples are sequen-
tially analyzed. To reduce carry-over, a VOA free water
blank should be analyzed following an unusually con-
centrated sample to assure that the syringe is clean.
13.0 Quality Control
Quality-control procedures and acceptance crite-
ria listed below, along with corrective actions, are
shown in table 2 C.
13.1: A blank and a 1-point standard is used for
instrument calibration. Initially, run a 10-|ig/L standard
(at 0-1 °C) to determine retention times and response
78 Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
-------
Table 2C. Quality controls, acceptance criteria, and corrective actions
[DQO, data-quality objective; RL, reporting limit; RPD, relative percent difference; <, less than]
Quality-control item
Frequency
Acceptance criteria
Corrective action
Initial calibration
Daily, before samples
<15 percent difference
from the first standard1
Inject another standard, check system
Blank
Daily, every batch
< 1/2 of RL1
Repeat blank injection, prepare a new blank, check
system, increase RLs depending on DQOs
Second source standard
Daily, every batch
< 20 percent difference
from true value1
Inject another standard, repeat initial calibration,
system
check
Continuing calibration
Every 10 samples and
< 20 percent difference
Inject another standard, repeat initial calibration,
check
after the last sample
from previous
standard1
system
Field duplicate
Every 20 samples
<30 percent RPD1
Repeat injection
Laboratory duplicate
Daily, every 20 samples
< 20 percent RPD1
Repeat injection, run another duplicate
Field blank
At least one per survey
<1/2 of RL1
Repeat injection. Evaluate data using
technical judgement
Acceptance criteria defined by technical judgement.
factors of instrument. Repeat the 10-|ig/L standard to
check the reproducibility. Acceptance criteria for this
initial calibration is less than 15 percent difference
from the first standard injection.
13.2: Laboratory blanks are analyzed at the ini-
tial calibration and periodically to be sure of no carry
over from previous injections. Technical judgement is
used to determine frequency of blank-sample analysis.
Acceptance criteria is no target-compound peaks
greater than one-half the reporting level.
13.3: A second source standard containing some
compounds of interest is analyzed daily to verify cali-
bration standard. Acceptance criteria is less than 20
percent difference of the true value.
13.4: A standard is run at least every 10 samples
and at the end of the sample batch to update the instru-
ment calibration. Acceptance criteria is less than 20
percent difference from previous standard.
13.5: Run field and laboratory duplicates every
twenty samples. Acceptance criteria is a relative per-
cent difference between the two values of less than
30 percent for field duplicates and 20 percent for
laboratory duplicates.
13.6: Field blanks are analyzed at least once per
survey. Field blanks are PVD samplers left out in ambi-
ent air (for example, attached to a tree) for the duration
of the sample-collection period and are retrieved and
analyzed the same day(s) as samples. Acceptance crite-
ria is no target-compound peaks greater than one-half
the reporting level.
14.0 Data and Records
Management
14.1: All work performed for the analyses of
samples must be entered into a field logbook. These
data are then reviewed and verified for precision, accu-
racy, and representativeness by the project chief or
project manager with analysis of the quality-assurance
data.
14.2: The samples analyzed are logged into a
laboratory-information-management system.
14.3: Chromatograms generated are saved and
filed in a project folder.
Appendix 2 79
-------
Guidance on Use of Vapor-Diffusion Samplers to Detect VOCs in Ground-Water-Discharge Areas, and Applications in New England
------- |