PB91-181776
                                   EPA/600/4-90/028
                                   May 1991
  FIELD COMPARISON OF GROUND-WATER SAMPLING
        DEVICES FOR HAZARDOUS WASTE SITES:
AN EVALUATION USING VOLATILE ORGANIC COMPOUNDS
                         by

                   Karl F. Pohlmann
                   Ronald P. Blegen
                     John W. Hess
                 Water Resources Center
                 Desert Research Institute
                 Las Vegas, Nevada 89120
                 Cooperative Agreement
                     CR 812713-01
                   Technical Monitor

                     Jane E. Denne
            Advanced Monitoring Systems Division
         Environmental Monitoring Systems Laboratory
               Las Vegas, Nevada 89193-3478
         This study was conducted in cooperation with
                 Water Resources Center
                 Desert Research Institute
                 Las Vegas, Nevada 89120
 ENVIRONMENTAL MONITORING SYSTEMS LABORATORY
       OFFICE OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL PROTECTION AGENCY
             LAS VEGAS, NEVADA 89193-3478

-------
                                    TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse before compleri-
1. REPORT NO.
 EPA/600/4-90/02S
                              2.
«. TITLE AND SUBTITLE

 FIELD  COMPARISON OF GROUND-WATER SAMPLING DEVICES FOR
 HAZARDOUS WASTE SITES:  AN EVALUATION"USING VOLATILE
 ORGANIC COMPOUNDS                   	•
                         PB91-181776
                   s. REPORT DATE
                       May 1991
                   6. PERFORMING ORGANIZATION CODE
 , AUTHOR(S)
 Karl F. Pohlmann,  Ronald P. Blegen, and John W. Hess
                                                             8. PERFORMING ORGANIZATION REPORT NO.
J. PERFORMING ORGANIZATION NAME AND ADDRESS
 Water Resources Center,  Desert Research
 2505  E.  Chandler,  Suite  1
 Las Vegas,  NV 89120
Institute
                   10. PROGRAM ELEMENT NO.
                         D109  (115)
                   11. CONTRACT/GRANT NO.
                                                                   CR 812713-01
12. SPONSORING AGENCY NAME AND ADDRESS
 Environmental Monitoring Systems Laboratory- LV, NV
 Office of  Research  and Development
 U.S. Environmental Protection Agency
 Las Veeas. NV  89114
                   13. TYPE OF REPORT AND PERIOD COVERED
                   Project Rpt.  & Summary l/87-9/9€
                   1*. SPONSORING AGENCY CODE
                      EPA/600/07
IS. SUPPLEMENTARY NOTES
16. ABSTRACT	'	~~

    To determine whether ground-water contamination has  occurred or remediation efforts
 have been  effective,  it is necessary to collect ground-water samples in such a way that
 the samples  are representative of ground-water conditions.   Unfortunately, formation of
 stagnant water within conventional monitoring wells requires that.these wells be purged
 prior to'sampling, a  procedure that may introduce significant bias into the determination
 of concentrations of  sensitive constituents such as volatile organic compounds (VOCs).
 The use of in situ ground-water sampling devices, which minimize or eliminate the need
 for well purging, may help alleviate some of the difficulties associated with sampling
 ground water at hazardous waste sites.  In this study,  several ground-water sampling
 devices, including two in situ systems, were field-tested to determine their capability
 for yielding representative. VOC data.  Sampling devices included a bladder pump, a
 bladder pump below an inflatable : packer, a bailer, a bailer with a bottom-emptying
 device, an in situ Westbay MP System..*,. two In situ BAT* devices, and a prototype BAT
 well probe.   The devices were field-tested at a site contaminated by a VOC plume, and
 the comparison was based on the ability of the devices to  recover representative
 concentrations of the VOCs.  The results of this study  indicate that the, tested in situ
 devices may  eliminate the need for veil purging prior to  sample collection and that the
 resulting  samples are at least as representative as those collected with a bladder pump
 in a conventional monitoring well. r\    	_______________
17.
                                 KEY WORDS AND OOCUMgMT ANALYSIS
                  DESCRIPTORS
                                               b.lOENTIFIERS/OPEN ENOEO TERMS  C.  COSATI FJeUlGfOUp
18. DISTRIBUTION STATEMENT

    RELEASE TO PUBLIC
      19. SECURITY CLASS (Tha Report]
       UNCLASSIFIED
113
     20. SECURITY CLASS (Thit p»te-
       UNCLASSIFIED
                                 22. PRICE
EPA F«f» 2220-1 (R««. 4-77)   »*KVIOUS COITION is OMOI.CTC

-------
                                           NOTICE

      The information in this document has been funded wholly or in part by the United States Environmental
Protection Agency under Cooperative Agreement Number CR812713-01 to the Vfater Resources Center of the
Desert Research Institute. It has been subjected to the Agency's peer and administrative review, and it has been
approved for publication as an EPA document. Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.

-------
                                           ABSTRACT

      To determine whether ground-water contamination has occurred or remediation efforts have been
effective, it is necessary to collect ground-water samples in such a way that the samples are representative of
ground-water conditions. Unfortunately, the formation of stagnant water within conventional monitoring wells
requires that these wells be purged prior to sampling, a procedure that may introduce significant bias into the
determination of concentrations of sensitive constituents such as volatile organic compounds (VOCs). The use of
in situ ground-water sampling devices, which minimize or eliminate the need for well purging, may help alleviate
some of the difficulties associated with sampling ground water at hazardous waste sites. In this study, several
ground-water sampling devices, including two in situ systems, were field-tested to determine their capability for
yielding representative VOC data.

       Sampling devices included a bladder pump, a bladder pump below an inflatable packer, a bailer, a bailer
with a bottom-emptying device, a Westbay MP System,  two in situ BAT* devices, and a BAT well probe. The
devices were field-tested at a site contaminated by a VOC plume, and the comparison was based on the ability of
the devices to recover samples containing representative concentrations of the VOCs. The devices were installed
on 6-m centers in a rectangular grid pattern, with the sampling zones of each installation placed at a depth of
about 6 m. The well spacing was chosen to minimize geologic variability over the site while maintaining hydraulic
isolation between installations. The first experimental phase consisted of eight sampling rounds over a 19-week
period and involved seven of the devices. Results revealed considerable variation in the data between sampling
devices and sample rounds. The second phase involved only the bladder pump, the bailer, one in situ BAT device,
and the in situ Westbay MP System and was designed to reduce the possibility of temporal variability by collecting
multiple replicate  samples with each device during a single sampling event. Phase two results suggested that
samples collected with either in situ device yielded VOC concentrations which were as accurate as those collected
with the bladder pump. The bailer yielded the least accurate data. The final phase of the study was designed to
study long-term VOC recovery patterns of six of the devices:  the four devices used in the second  phase, an
additional in situ BAT device, and a bailer with a bottom-emptying device. Four sample rounds were conducted at
 12-week intervals with the results suggesting that two of the in situ devices were as accurate and precise as the
bladder pump. The two bailers utilized during this phase of the study were the least accurate and precise. No
operational problems were encountered with any of the six devices tested during the final phase of the study. The
results of this study indicate that the tested in situ devices may eliminate the need for well purging prior to sample
collection and that the resulting samples are at least as representative as those collected with a bladder pump in a
conventional monitoring  well.

        This report  was submitted in  fulfillment of Cooperative Agreement CR812713-Q1 by the Desert
 Research Institute under the partial sponsorship of the U.S. Environmental Protection Agency. It covers the
 period from July 1,1986, through June 30,1990, and work was completed as of September 30,1990.
                                                 111

-------
                                    ACKNOWLEDGEMENTS

      The authors wish to acknowledge the contributions of the U.S. EPA Environmental Monitoring Systems
Laboratory (Las Vegas, Nevada) and the Desert Research Institute, Water Resources Center. Mr. Joseph J.
DLugosz contributed to the project scope. Dr. Robert Kinnison and Dr. Forest Miller provided assistance in the
experimental design and statistical evaluation of the results. Dr. Michael Whitbeck and Dr. Richard McArthur
provided comments and assistance as the project progressed and critically reviewed the report. Ms. Debora
Noack produced many of the figures and tables and put together the final document. We also wish to thank L'Eggs
Products Inc. and Mrs. Vivian Carpin for their cooperation and willingness to allow these experiments to take
place on their property.
                                                IV

-------
                                         CONTENTS

Abstract  	  m
Acknowledgements	  iy
Figures  	•	•	^
Tables  	«

1.  Introduction	1
       Objectives and Approach  	2

2.  Description of Sampling Devices	4
       Bailers	4
       Bladder Pumps  	6
       BAT Ground-Water Monitoring Systems  	9
       Westbay MP System 	12
3.  Site Description  	16
       Site Location and History  	16
       Geology	16
           Quaternary Alluvium  	18
           Muddy Creek Formation  	19
       Hydrogeology	20
       Hydrochemistry 	23
       Monitoring Well Installation  	25
4.  Experiment A 	28
       Experimental Design  	28
       Results  	31
           Benzene and Chlorobenzene  	32
           Total Organic Carbon  	34
           Week 19 Experiments  	35
       Discussion   	36
5.  Experiment B 	38
       Experimental Design  	38
       Results  	38
       Discussion	40
6.  Experiment C	43
       Experimental Design  	43
       Results  	44

-------
    Discussion  	46
7.   Summary and Conclusions  	50

References  	52

Appendices
    A.  Well Construction Diagrams	55
    B.  Laboratory Analysis Methodology 	61
    C.  Analytical Results 	63
    D.  Data Plots and Statistical Analyses Results 	89
                                                VI

-------
                                            FIGURES

Number                                                                                   Eagfi

    1    Typical bailer design	     5
    2    Typical bladder pump design	     7
    3    BAT filter tips	    10
    4    BAT in-line filter adapter  	    11
    5    Westbay MP sampling probe operation 	    13
    6    Westbay MP VGA bottle holder 	    14
    7    Study site location, eastern Las Vegas Valley 	    17
    8    Pittman Lateral study site  	    18
    9    Configuration and altitude of the top of the Muddy Creek Formation,
         Henderson, Nevada 	    19
    10   Hydrogeologic cross-section of the Pittman Lateral site  	    20
    11   Water table elevation map, Henderson, Nevada 	    22
    12   Contour map of total dissolved solids concentrations (mg/L)
         in ground water, Henderson, Nevada  	    24
    13   Map of benzene concentrations (ppm) in ground water,
         Henderson, Nevada 	    25
    14   Monitoring well layout, Pittman Lateral site 	    26
    15   Plot of benzene concentration (ng/L)  for each of five sample
         splits from wells 1, 4, and 6  	    29
    16   Plot of chlorobenzene concentration Gig/L) for each of five sample
         splits from wells 1, 4, and 6  	    29
    17   Plot of TOC concentration (mg/L) for each of five sample splits
         from wells 1, 4, and 6	    30
    18   Benzene concentration (jig/L) recovered by each sampling
         device as a function of time 	    33
    19   Chlorobenzene concentration (ng/L) recovered by each sampling
         device as a function of time 	    33
    20   TOC concentration (mg/L) recovered by each sampling
         device as a function of time 	    35
    A.1  Well  1 construction diagram and lithologic log   	    55
    A.2  Well 2 construction diagram and lithologic log   	    56
    A3  Well 3 construction diagram and lithologic log   	    57
                                                vu

-------
A.4  Well 4 construction diagram and lithologic log  	    58
A.5  Well 5 construction diagram and lithologic log  	    59
A6  Well 6 construction diagram and lithologic log  	    60
D.I  Experiment B - Benzene data	    89
D.2  Experiment B - Chlorobenzene data  	    89
D3  Experiment B - 1,2-Dichlorobenzene data	    90
D.4  Experiment B - 1,3-Dichlorobenzene data	    90
D.5  Experiment B - 1,4-Dichlorobenzene data	    91
D.6  Experiment B - 1,1-Dichloroethane data 	    91
D.7  Experiment B - 1,2-Dichloroethane data 	    92
D.8  Experiment B - Ethylbenzene data  	    92
D.9  Experiment B - Trichloroethene data  	    93
D.10 Experiment C - Benzene data	    94
D.ll Experiment C - Chlorobenzene data  	    94
D.12 Experiment C - 1,2-Dichlorobenzene data	    95
D.13 Experiment C - 1,3-Dichlorobenzene data	    95
D.14 Experiment C - 1,4-Dichlorobenzene data	    96
D.15 Experiment C - 1,1-Dichloroethane data 	    %
D.16 Experiment C - 1,2-Dichloroethane data 	    97
D.17 Experiment C - Ethylbenzene data  	    97
D.18 Experiment C - Trichloroethene data  	    98
                                          viu

-------
                                           TABLES

Number                                                                                Page

    1     Sampling devices utilized in Experiment A  	     31
    2     Tukey test results for Experiment A	     34
    3     Sampling Devices utilized in Experiment B   	     39
    4     VOC concentration means and standard deviations of the four
          sampling devices utilized in Experiment B	     39
    5     Tukey test results for Experiment B 	     41
    6     Sampling devices utilized in Experiment C  	     43
    7     Test of effectiveness of bottom-emptying device, sample round 4	     45
    8     Rankings of devices based on Tukey test results - highest VOC recovery	     47
    9     Rankings of devices based on Tukey test results - lowest VOC recovery	     47
    B.I   Chemical species, referenced method analysis, and
          laboratory analysis equipment used  	     62
    C.I   Survey sampling results, well 1  	     63
    C.2   Survey sampling results, well 4  	     64
    C3   Survey sampling results, well 6  	     65
    C.4   Experiment A -Bladder pump sampling results  	     66
    C.5   Experiment A - Bailer sampling results	     67
    C.6   Experiment A - PTFE filter tip sampling results	     68
    C.7   Experiment A - Multi-port sampling results 	     69
    C.8   Experiment A - Well probe sampling results 	     70
    C.9   Experiment A - HDPE filter tip sampling results 	     71
    C.10  Experiment A - Bladder pump/packer sampling results  	     72
    C.ll  Experiment A - LWWD standard solutions 	     73
    C.12  Experiment A - EPA standard solutions 	     74
    C.13  Experiment A - Equipment blanks - bladder pump  	     75
    C.14  Experiment A - Equipment blanks - bailer  	     76
    C.15  Experiment A - Trip blanks  	     77
    C.16  Experiment A - Week 19 sampling results   	     80
    C.17  Experiment B - VOC analytical results for six replicate samples
          collected with each sampling device  - November 8,1988	     81
    C.18  Experiment B - Quality assurance samples	     82
    C.19  Experiment C - VOC analytical results for sample round 1	     83
                                                IX

-------
C.20  Experiment C - VOC analytical results for sample round 2	     84
C.21  Experiment C - VOC analytical results for sample round 3	     85
C22  Experiment C - VOC analytical results for sample round 4	     86
C23  Experiment C - Quality assurance blanks 	     87
C.24  Duplicate analyses of random samples	     88
D.I   Experiment C - Tukey test results for round 1 	     99
D.2   Experiment C - Tukey test results for round 2 	    100
D3   Experiment C-Tukey test results for round 3 	    101
D.4   Experiment C-Tukey test results for round 4 	    102

-------
                                            SECTION 1
                                         INTRODUCTION

      The ability to collect ground-water samples representative of aquifer hydrochemical conditions is a major
concern in any ground-water investigation effort. Unfortunately, there are many factors in the sampling process
that can introduce variability into determinations of chemical constituent concentrations, greatly influencing the
ability to obtain accurate results. Examples include well drilling method, well design, the materials used in well
construction, well development and purging, sampling device, sample handling and preservation, and analytical
technique.

      This Cooperative Agreement research was initiated to study the effectiveness of various elements of
ground-water sampling procedures for hazardous waste sites. An early product of the study was a report (Blegen
et d., 1987) addressing concerns about well-purging, sample collection, and equipment decontamination at
hazardous waste sites. The report outlined current methods and equipment and discussed their relative
effectiveness. It also contained an annotated bibliography of important work on these subjects and a reference list
of over 300 related documents. A matrix of ground-water sampling devices (Pohlmann and Hess, 1988) was also
produced for this project. This report outlined the suitability of 12 categories of devices for sampling 14 types of
ground-water parameters typically found at hazardous waste sites. Another element of the project was
construction of a laboratory-based, simulated well environment for the purpose of providing a standardized basis
for comparing ground-water samplers (Whitbeck and Williams, 1990). In addition, a laboratory-based study of
the effects of purging techniques on volatile organic compound (VOC)  concentrations in low-yield wells was
conducted (Pohlmann et al., 1990). Finally, this report focuses on a field comparison of ground-water sampling
devices for sampling VOCs.

       Of the many  elements of the sampling process, maintenance of sample integrity from the subsurface
sampling point to the sample container has received considerable interest because of the great potential impact
on sample representativeness possible during this phase of the process. In addition, concern for obtaining
 representative samples of ground water containing VOCs has arisen because VOCs are common contaminants at
 hazardous waste sites (Plumb and Pitchford, 1985). To address these concerns, sampling devices and methods
 have been developed to minimize impact on sample quality.

        Several published studies have focused on the effects sampling devices have on the physical and chemical
 integrity of the samples they are used to collect. Comparative studies, based either in the laboratory or in  the
 field, are a common means of determining which sampling mechanisms or procedures function best under a given
 set of conditions. Laboratory studies by Unwin (1984), Barcelona  et al. (1984), and Stolzenburg and Nichols
 (1985), and field studiesby Houghton and Berger (1984), Muskaet al. (1986), Imbrigiottaef a/. (1986), Pearsall and

-------
Eckhardt (1987), and Yeskiser al. (1988) compared some of the more commonly utilized ground-water sampling
methodologies and devices.

      In general,  the devices tested during these studies are used  in conventional monitoring  wells and
therefore require the removal of stagnant water from the well prior to sampling. While well purging is generally
considered necessary in order to collect a representative ground-water sample, the purging process may also
introduce considerable bias in the sampling. Gillham et al. (1985) note several potential problems associated with
well purging, including:

        -   There is no assurance that all of the stagnant water has been removed from the well,
            or that the resulting samples have not been contaminated by stagnant water remaining
            in the well.

        -   A large drawdown, induced by the removal of large volumes of water during well
            purging, may alter ground-water chemistry (degassing, volatilization), or may draw in to
            the well unrepresentative ground water from different hydrogeologic zones.

            Well purging can be time-consuming and require extra equipment, thereby increasing
            sampling costs.

        -   Waste water disposal may become a problem if the volume is large and/or if the water
            is contaminated.

       Among the newer commercially available sampling devices are two devices that need to have little or no
water purged from the system prior to sampling. These devices are not used in conventional monitoring wells but
are  stand-alone systems  installed  directly into the  subsurface, similar to conventional  monitoring well
installations. Because of the nature of their design and installation, these devices collect samples almost directly
from the formation, so are often referred to as in situ systems. To date, few studies comparing these devices to the
more conventional ground-water sampling methods have been published. However, the existence of these new
devices raises three important questions:

        -   Do sampling methods exist which  may eliminate the need for well purging prior to
            sample collection?

        -   How valid or representative is the  resulting sample?

            Are the proposed in situ monitoring techniques inherently invalid because of the
            necessity of well purging?

       This study was initiated to address the concerns expressed by these questions. In other words, can in situ
 monitoring techniques yield data  representative of subsurface geochemical conditions, or will these data be
 biased due to the inherent need to purge the well bore prior to sampling?

 OBJECTIVES AND APPROACH

       The primary objective of this study was to conduct a field comparison between several ground-water
 sampling devices, including two in situ devices that require little or no pre-sample purging. Seven  different
 sampling methods were evaluated to determine if the non-pumping methods yielded representative data, and to

-------
compare the accuracy and precision of the various sampling devices. The comparison was based on the ability of
each device to deliver a representative ground-water sample from the subsurface environment to the ground
surface and into an appropriate sample container. Because VOCs are common contaminants at hazardous waste
sites, emphasis was placed on the effect these devices had on the recovery of VOCs present in ground water at the
study site.

       The ideal approach to a study of this type would be to conduct the comparison in such a manner that each
sampling device was subjected to identical sampling conditions. Under these conditions, the only sampling
variable would be the device itself. However, the very nature of a field study implies variation in subsurface
geologic and hydrochemical conditions. This problem was addressed by locating, designing, and constructing the
wells in such a way that hydrogeologic and geochemical conditions were as similar as possible in a field setting.

       The investigation was conducted in three separate phases (experiments A, B, and C) with the approach
including the following major procedures:

            Six ground-water monitoring wells were installed close to each other at a site in
            southeastern  Las Vegas Valley.  Geologic and  geophysical logs were  utilized  to
            delineate geologic variability at the site.

        -   An initial set of multiple replicate samples was collected from three wells to investigate
            variability in hydrochemistry between the sampling points. Laboratory results were
            also utilized to estimate expected laboratory error.

        -   Ground-water samples for experiment  A were collected with each device on a
            regularly scheduled basis during nine sampling rounds over a 19-week period. Samples
            were analyzed in the  laboratory for a variety of organic and inorganic chemical
            parameters.

         -  Laboratory results were subjected to statistical analysis, including a two-way analysis
            of variance designed to determine if significant chemical variations could be attributed
            to sampling device and/or sampling time. Emphasis was placed on sampling device
            effects on VOC concentration.

         -   A second experiment, experiment B, was developed based on the experiences and
             results of the first phase. Experiment B  included replicate sampling during a single
             sampling period. Samples were analyzed for VOC concentration only.

         -   Statistical analysis of the new data included an analysis of variance and multiple
             comparison tests. Results of the analyses were evaluated to determine which sampling
             devices yielded the most accurate and precise VOC concentrations.

         -   Experiment C added the element of time by conducting four sample rounds over a
             one-year period. Patterns of VOC recovery over  time and long-term  device
             performance were of primary interest.

-------
                                           SECTION!

                             DESCRIPTION OF SAMPLING DEVICES


      Seven ground-water sampling  devices were utilized in this comparative  study, all of which  are
commercially available. The devices may be categorized as: grab samplers (bailers), positive displacement
mechanisms (bladder pomp, bladder pump below an inflatable packer), and in sou devices (Westbay MP System9,
two BAT*  Ground-Water Monitoring Systems, and a BAT well probe). To minimize the potential for bias
resulting from contact with the materials of which the samplers were constructed, the devices were selected with
the most chemically inert materials available from the manufacturers.

BAILERS

      Bailers are perhaps the simplest of all ground-water sampling devices. They may be easily constructed
from most tubular materials and are relatively inexpensive and simple to use. The bask: design of most bailers
consists of a long tube (usually of rigid PVC, Teflon*, or stainless steel) open at the upper end where a haul line is
attached and with a simple ball-and-seat check valve at the bottom end (Figure 1).

      The bailer is lowered into a well by the haul line. Contact with the water surface dislodges the check valve
ball, which  allows water to flow through the main body of the bailer as it is lowered through the water column.
When the descent of the bailer is stopped at the desired sampling depth, the weight of the overlying water column
closes the check valve, thus trapping a sample inside the bailer. The bailer is then withdrawn from the well and the
sample transferred to appropriate sample containers.

      Nielsen and Yeates (1985) and Scalf a til. (1981) list several advantages and disadvantages of using bailers
to sample ground water from wells. Advantages include:

        -   Bailers may be constructed from a wide variety of materials and in  virtually any
            dimensions to accommodate any well diameter and desired sample volume.

        -   Bailers are mechanically simple, easily operated, portable, and require no external
            power source.

        -   Bailers are relatively inexpensive, making it possible to dedicate a separate bailer to
            each well.

        -   Samples may be taken from virtually any depth.

Disadvantages of bailers include:

        -   Purging a well of stagnant water with  a bailer is often impractical, particularly with
            deeper wells. Other devices are often needed to purge the well prior to sampling with
            the bailer.

-------
      Haul line
           Vent
Check valve
                                      T
                                      91cm
                             4.2cm
     Figure 1. - Typical bailer design.

-------
           Haul lines must be of a noncontaminating material, adequately cleaned, and dedicated
           to a single well to prevent cross-contamination.

           Aeration, degassing and turbulence may occur while lowering and raising the bailer
           through the water column and while transferring the sample to appropriate containers.

       -   Sampling personnel may be exposed to any contaminants in the sample.

       -   The sample may not be representative of a specific point within the water column.

       -   Movement of the bailer may dislodge paniculate matter from the casing wall, resulting
           in a turbid and unrepresentative sample.

      One bailer utilized in this project was an all-Teflon closed-top  bailer produced by the  Galtek*
Corporation. The dosed top with side ports design was chosen to minimize the possibility of stagnant water
entering the device as it was retrieved from the well. A Teflon-coated stainless steel haul line was used to lower
and raise the bailer. Sample water was discharged into dean glass beakers by carefully dislodging the check valve
ball against the bottom of the beaker. This bailer will be denoted as the "standard bailer" or "bailer" in this report.

      The second bailer utilized in this project was similar in operating principle but used a  controlled flow
bottom-emptying device (BED) to transfer the sample directly from the bottom of the bailer to a volatile organic
analysis (VOA) vial. This feature was designed by the manufacturer to minimize agitation of the sample that may
occur during transfer to the sample container by decanting from the bailer top or by discharging into a beaker
from the bailer bottom and  then pouring into a VOA vial. This bailer, denoted as the "BED Bailer" in this report,
was manufactured by Norton Performance Plastics. A closed top and "V"-notched orifices on the side of the
bailer at the top were designed to prevent  the contents from mixing with well water during retrieval.

      The bailers were cleaned in the laboratory before being used in the field. The cleaning procedure involved
careful scrubbing with a nonphosphate detergent/tap water solution, followed by rinsing with a 10 percent
acetone solution and distilled water. Equipment blanks were collected following the distilled water rinse. The
haul lines were subjected  to the same cleaning process. When not in use, the bailers were  stored in dean
polyethylene bags.

BLADDER PUMPS

      Bladder pumps, also known as gas-squeeze pumps, are classified as positive-displacement sampling
devices due to the utilization of positive gas pressure to "push" water samples from depth to the ground surface
for collection. While specific designs may vary somewhat, most bladder pumps consist of a flexible membrane (the
"bladder") housed in a long rigid tube, a water intake check valve, a discharge check valve, a tubing line which
allows for gas pressurization of the annular space between the membrane and the pump housing, and another
tube for sample discharge  (Figure 2). Bladder pumps require a supply of compressed gas and an automated
control system which controls gas pressure and gas flow rates to the pump, which in turn dictate water sample
discharge rates (Nielsen and Yeates, 1985).

      Bladder pump operation is relatively simple. When the pump is lowered into the well  and submerged,
water may enter the pump by passing through the intake check valve and into the bladder. Gas pressure applied to
the annular space between the bladder and the rigid outer wall of the pump compresses the bladder. This causes
the intake check valve to dose and forces the water sample through the discharge check valve and up the sample
discharge tube. When this "pump phase" is complete, pressure inside the pump is released and vented at the
ground surface by the control system. The  discharge check valve doses under the weight of the overlying water
column within the discharge tube and prevents water in the tube from flowing back into the  pump. With the
bladder now fully relaxed, water may again enter the pump and fill the bladder. The process is repeated to cycle
water to the ground surface. Adjustments to the automated control system regulate applied pressure and control

-------
Sample Discharge Check Valve-
                   Gas Inlet
                Pump Casing
                     Bladder •
    Water Intake Check Valve •
                                                             102 cm
                                               4.2 cm
                        Figure 2. - Typical bladder pump design.

-------
the frequency with which pressure is applied to and released from the pump, thus controlling sample flow rates
and lift capability as well as optimizing pumping efficiency (Nielsen and Yeates, 1985).

       As noted by Nielsen and Yeates (1985), Scalf et al. (1981) and Gillham et at. (1983), advantages of the
bladder pump include:

           No direct contact occurs between the driving gas and the water sample, thereby
           minimizing concerns about gas stripping.

        -  The pumps may be constructed of a wide range of materials, including stainless steel
           and Teflon.

           A wide range of pumping rates is possible and the rate may be controlled relatively
           easily. This allows the bladder pump to be used for well evacuation at high pumping
           rates as well as for sampling at low flow rates.

        -  Depth capability is controlled primarily by the operating pressure of the pump. Many
           bladder pumps are capable of pumping lifts of over 60 m.

        -  The pumps are easily disassembled for cleaning and repair.

        -  The pumps are commercially available in a variety of lengths and diameters, although
           most are designed for use in 5.1 cm diameter monitoring wells.

Disadvantages include:

        -  Large gas volumes and long cycles are necessary when pumping from deeper wells.
           This increases operating time and expense.

        -  Although adjustable, pump discharge is intermittent

        -  The pump may subject the sample to turbulent flow and large pressure changes.

           Minimum pumping rates may be higher  than ideal for sampling ground water
           contaminated with VOCs.

        -  Check valves may fail in water with a high suspended solids content.

        -  Commercially available pump units may be relatively  expensive.

       Two all-Teflon Well Wizard* bladder pumps (produced by  Q.ED. Environmental Systems, Inc.) were
utilized for this project. The  gas-supply and sample discharge tubes and all fittings are also made of Teflon.

       One of the two bladder pumps was used in conjunction with an inflatable packer mounted above the
pump. The packer serves to reduce purge volumes by isolating stagnant well water above the pump and
preventing that water from  migrating downward to  the pump intake. The Purge Mizer* inflatable packer
(manufactured by Q.E.D. Environmental Systems, Inc.) consists of a stainless steel body and fittings and a Viton*
bellows. The pump/packer combination was dedicated to a single well, and after installation  it remained
downhole for the duration of the sampling period. The other bladder pump was also dedicated to a single well, but
it was necessary to remove it from the well after purging and sampling to allow for sampling with the bailer.

       The pomps were cleaned in a manner similar to that described for the bailer. Prior to installation in a well
the pumps were placed in cleaning tubes which contained the cleaning solutions. The pump was activated and
allowed to pump approximately 19 L of each solution through the pump and discharge lines. The solutions used
were (a) a nonphosphate detergent/tap water solution, (b) a 10 percent acetone solution, and (c) distilled water.
Equipment blanks were collected during the final distilled water rinse.

-------
BAT GROUND-WATER MONITORING SYSTEMS

      The BAT Ground-Water Monitoring System is a relatively new type of ground-water monitoring system
developed  in  Sweden by BAT Envitech, Inc. Each component of the  system is sealed, and hydraulic
interconnections between components are accomplished through the use of hypodermic needles, flexible seals,
and induced pressure gradients. Ground-water samples are collected in sealed, evacuated glass vials which may
be sent directly to the laboratory. Therefore, the system makes it possible to collect pressurized water and gas
samples without purging large amounts of well water or transferring samples to other containers, a procedure
which often results in a loss of VOCs or external contamination (Torstensson and Petsonk, 1986).

      The primary feature of the system is the BAT filter tip (Figure 3a), the standard configuration of which
consists of a thermoplastic body and a filter of high-density polyethylene. The filter tips can be produced from a
variety of materials and in several design configurations to meet expected installation and sampling conditions.
The filter tip is reinforced with a core of Teflon-coated stainless steel and sealed with a flexible septum of a
resilient material (fluororubber septa were used for this project). The filter tip is threaded onto the bottom of an
extension pipe and additional lengths of pipe are added as needed. Normal installation procedures call for the
filter tip to be pushed into the ground to the desired sampling depth under a static load, although the installation
may be completed in pre-drilled holes (Torstensson and Petsonk, 1986); the latter was done during this study.

       Ground-water samples are obtained by inserting  the pre-sterilized, pre-evacuated sample vial into a
sample container housing, which is then lowered down the extension pipe. At the lower  end of the housing, a
"guide sleeve" assembly contains a double-ended hypodermic needle. The glass sample vials contain a flexible
septum similar to those in the filter tips. Contact between the guide sleeve and the filter tip cap causes the needle
to puncture the septa in both the cap and the sample vial. The vacuum in the vial may then draw ground water
from the formation, through the filter tip and into the vial. When the sample housing  is withdrawn, the
hypodermic needle withdraws from both the sample vial and the filter tip. The guide sleeve mechanism causes the
needle to withdraw from the sample vial first, thus preventing a loss of sample fluid and/or gas (Torstensson and
Petsonk, 1986). The septa in both the filter tip and the vial automatically reseal as the needle is withdrawn.
Pressurized samples may be obtained if the sample vial remains connected to the filter tip long enough for the
pressure inside the vial to equalize with formation water pressure. When the vial is disconnected from the filter
tip, formation pressure is preserved inside the sample vial (Torstensson, 1984). Filtered samples may be obtained
through the use of an in-line filter adapter (Figure 4).

       The in situ design of the BAT system means that the volume of water which needs purging is considerably
less than that in conventional monitoring wells. Theoretically, only the small volume of water contained within
and immediately outside the filter tip itself would need to be removed prior to  sample collection, because this
water is in contact with the sampling device and could be considered stagnant water. The manufacturer suggests
that one 250 to 500-mL sample volume, collected and discarded before collection of samples, may be sufficient to
purge the  system. As discussed  previously,  this small purge volume feature is of primary interest in this
investigation. Other advantages or disadvantages of in situ systems observed during the investigation will be
discussed in subsequent sections of this report.

       The two in situ filter tips used for this study were designed to be threaded onto a 5.1-cm-diameter Teflon
extension pipe, which allows sample sizes of up to 500 mL to be obtained. Both of the in situ filter tips are modified
versions of the standard BAT Mk n filter tip. One of the tips has a body of polytetrafluoroethylene (PTFE, or
Teflon) with a porous PTFE filter (hereafter referred to as the "PTFE filter tip"),  while the other tip has a
polyacetal body with a filter of high-density polyethylene (hereafter referred to as the "HDPE filter tip"). Both in
situ filter tips were installed in pre-drilled boreholes, with the installations completed by backfilling with gravel
pack and cement-bentonite slurry surface seals.

       A third BAT device used for this project is a major modification of the in situ tip and requires completely
 different assembly and installation procedures. However, once installed, it is functionally identical to a standard

-------
           s'lcm
41 cm
                     • Septum of Fluororubber'
                       Cap of PTFE or PVC






k
1
L
^





R
                         Thread connects to.
                         installation adapter

                          Thread connects to
                          5.1 cm FTFE Pipe
     Viton O-ring
                        Body      Body
                         of        of
                        PTFE    Polyacetal
                           PTFE Filter
                     Avg. Pore Size: 35 pm
                     Max. Pore Size: 120 urn
                     Apparent
                     Permeability: 1.7 x 10"4 cm/s
                           HOPE Filter
                     Avg. Pore Size:  15
                     Apparent
                     Permeability: 1.7 x 10~3 cm/s
          V-
             (a)
Tip of PTFE or
Polyacetal
                            (b)
Figure 3. - BAT filter tips: (a) In situ BAT filter tip, (b) BAT well probe (adapted
         from BAT Envitech, Inc., 1987).
                                10

-------
7.6 cm
                                                     Sample container housing
                                                      500 mL glass evacuated
                                                      sample vial
                                                     Hypodermic needle

                                                     Fluororubber septum
                                                 Filter housing
                 0.45 jim filter,  2.5 cm diameter
                                              Luer"  adapter
                                              Fluororubber septum

                                              Septum holder, thickness 5 mm
                                           Millipore®" filter housing, lower part
                                                   Guide sleeve with double-ended
                                                   hypodermic needle
                    K-2.2
cm
               Figure 4. - BAT in-line filter adapter (adapted from BAT
                          Envitech, Inc., 1987).
                                         11

-------
filter tip. Developed by BAT Envitech, Inc., for the Desert Research Institute (DRI), the BAT "well probe"
(shown in Figure 3b) is designed to be poshed down the inside of an existing 5.1-cm-diameter well. The two Viton
O-rings on the probe function as a packer to seal off the well screen from the environment in the casing above.
With the well screen sealed off, the probe functions as a sampling port from which samples from the screened
interval may be extracted. The well probe was installed in a conventionally designed monitoring well following
well development. The well was purged of stagnant water prior to insertion and placement of the probe within the
screened interval

WESTBAYMP SYSTEM

      The Westbay MP System, designed and distributed by Westbay Instruments, Ltd., is a ground-water
monitoring device which allows discrete samples or measurements to be taken at multiple levels within a single
borehole. The system consists of various lengths of casing joined by regular or valved port couplings, and a variety
of specialized tools and probes designed to access the ports and retrieve samples or measurement data from the
environment outside the sealed casing. The modular design allows for the establishment of as many monitoring
zones as desired, and the well design may easily be altered at the time of installation to adjust to unexpected
subsurface conditions. The valved nature of the ports does not allow formation water to enter the MP casing, thus
eliminating the need to purge stagnant water from inside the casing prior to sampling. Sampling probes accessing
a measurement port draw the sample from directly outside the casing (Black et al., 1986).

       The standard Westbay MP casing has an inside diameter of 3.7 cm and is available in several different
lengths. All couplings and end caps are connected to the casing with a flexible shear rod, and O-rings provide a
tight hydraulic seal which prevents water from entering the casing. Regular couplings are used where valved ports
are not required. Pumping ports are used at intervals where it may be desirable to remove large quantities of
water. The port may be  opened and dosed with a specially designed tool by moving an internal sleeve which
exposes (or covers) slots that allow water to flow into the casing. Opening this port makes it possible to develop a
desired monitoring zone, and positioning a pumping port just below a measurement (sampling) port allows for the
development of the region around that port as well. A measurement port coupling is essentially a regular
coupling with a check valve in the coupling wall through which fluid samples may be extracted. Normally dosed,
the valve may be accessed and opened with a sampling probe (Black & al., 1986).

       The procedure for obtaining a fluid sample is depicted in Figure 5. A sample container is attached to the
sampling probe and lowered down the casing to a point below the desired measurement port. The backing shoe is
briefly activated (pneumatically activated, nitrogen-gas driven), releasing the location arm. The probe and
sample bottle are then raised above, then lowered down to, the measurement port where they automatically
position themselves with respect to the port valve. Activating the backing shoe pushes the probe to the wall of the
coupling and simultaneously forces the port valve to open as the face seal of the probe seals around the valve.
Opening the probe sampling valve creates a hydraulic connection between the formation water outside of the
port and the sample bottle. After allowing enough time for the sample container to fill, the sampling valve is
dosed, the backing shoe deactivated, and the probe returned to the surface. The port valve automatically closes
as the backing shoe is deactivated (Black a al., 1986).

       Several sample container configurations are available. The container used to collect VOA samples is
shown schematically in  Figure 6. The container holds a standard 40-mL glass VOA bottle. Two hypodermic
needles penetrate the bottle septum, one of which is hydraulically connected to the sampler probe through a
032-cm-diameter Teflon tube. When the sampler valve is opened, water flows through the tube and needle and
into the bottle. As the bottle fills, the second needle vents gas and excess water into the container housing. This
design allows the bottle to be thoroughly flushed with formation water with only the last 40 mL kept as a sample
(Black era/., 1986).

       The Westbay MP System components chosen for this study were made of stainless steel. Three
measurement ports were installed at various depths, but only data from the lowermost port were used in the
                                                 12

-------
           Casing-
          Sampling
          probe
            Location
            arm-

           Measure-
           ment
           pon valve
            M!
            measure-
            ment port
            coupling
                                                                    To sample
                                                                    container
                                     5  Backing
                                     5/shoe
                                     rl
                                                                                 Sampling
                                                                                 valve
                      a) Probe located
                        at measurement
                        pon coupling.
                        Sampling valve
                        closed.
b) Probe activated.
   Sampling valve
   closed.
c) Probe activated.
   Sampling valve
   open.
                    Figure 5. - Westbay MP sampling probe operation (adapted from
                               Black era/., 1986).
comparative work. A pumping port was installed 0.6 m below the lower measurement port to allow development
of the monitoring zone and a 0.9-m section of screen (0.025-cm slots) was placed around both ports. The well
construction was completed by backfilling with gravel pack around the monitoring zones, separated by bentonite
seals. A standard method of development for Westbay installations, and the method used for this installation, is
                                               13

-------
                                    y   from sampling
                                        probe
                                                        One-way pressure relief valve
    Venting
    hypodermic needle
30 cm
                                                      • Head space container
                                                    0.3 cm Teflon tubing
needle

  Iffl
    Stainless steel
    holder for
    glass sample
    bottle
                                                         Septum
                                                    Filling hypodermic needle
                                                        Glass VOA sample bottle
       Figure 6. - Westbay MP VOA bottle holder (adapted from Black et al., 1986).
                                         14

-------
to open a pumping port and airlift the water from the inside of the casing. The Westbay MP System will be
denoted as the "multi-port" in this report.

       As with the in situ BAT devices, the design of the Westbay MP system is significant in that the volume of
water needing purging from the installation is considerably less than that required in conventional monitoring
wells. Theoretically, only the  small volume of water in contact with the casing outside the measurement port
should need removal before a sample is collected. The manufacturer suggests that, depending on conditions at
the individual site, a 250 mL volume collected and discarded prior to sample collection maybe sufficient to purge
the installation. Again, this small purge volume is a feature that is of primary interest in this study, and other
advantages or disadvantages of in situ devices observed during the investigation will be discussed in subsequent
sections of this report.
                                                 15

-------
                                           SECTION 3

                                      SITE DESCRIPTION
SITE LOCATION AND HISTORY
      The site chosen for this field study is in the southeastern part of Las Vegas Valley, within the city limits
of Henderson, Nevada, and approximately 3 km north of the Basic Management, Inc., (BMI) Complex, a large
facility which houses several major chemical production companies (Figure 7). Directly adjacent to the site is the
Pittman Lateral, a major water conduit supplying Colorado River water to the Las Vegas metropolitan area
(Figure 8). Also adjacent to the study area is a series of 23 ground-water monitoring wells, collectively known
as the Pittman Lateral Transect, which are situated roughly perpendicular to the northward flow of ground water.
The site chosen for installation and testing of the ground-water sampling equipment is near the center of the
Pittman Lateral Transect.

      In the early 1940s, the Federal Government, in response to an increased wartime need for magnesium,
constructed the Basic Magnesium facility in Henderson. After the war, the facility was sold to several private
corporations. BMI was created in 1952 by several of these corporations to manage certain portions of the complex
(JRB Associates, Inc.,  1981).

      From the early 1940s to the late 1970s, unknown quantities and types of liquid and solid wastes were
routinely disposed of in leach beds and unlined evaporation ponds on BMI property. Increased recharge due to
leakage from the unlined ponds had an almost immediate effect on the local shallow ground-water system as
water levels rose rapidly. Flooded basements and cesspools in the community of Pittman, 3 km north of the BMI
complex, were reported as early as 1942 (Geraghty and Miller, Inc., 1980).

      At least one organic plume has been determined to have originated from beneath property within the
BMI complex An accidental spill from an underground storage tank in 1976 released approximately 113,500 L
of benzene, contributing to the extensive organic contamination which exists today (Geraghty and Miller, Inc.,
1980). The downgradient movement of the benzene plume has brought it into contact with a variety of other
organic compounds, which may have been mobilized and transported away from their original disposal areas.

      In the spring of 1983,23 monitoring wells were installed along the Pittman Lateral Transect by the DRI,
Lockheed Engineering and Sciences Company, and the EPA. The primary purpose of the Pittman Lateral wells
was to define the local hydrogeology as well as the chemical character of the contaminant plumes. Since that
time, the site has been used for a number of sou gas studies, fiber optics experiments, and surface and borehole
geophysics experiments.

GEOLOGY

      The near-surface geology of the site is composed of two major geologic units: the unconsolidated sands,
gravels and cobbles of the Quaternary alluvial fan deposits, and the underlying Muddy Creek Formation, here
composed primarily of days, silts and fine sands.
                                                16

-------
                                                                          N
                                                        2  milea
                          101      2 km

                          Contour interval  BO feet
Figure 7. -  Study site location, eastern Las Vegas Valley (adapted from U.S. Bureau of
           Reclamation, 1984).
                                     17

-------
                     Hitman lateral Transect
                                                                                   N
                                                          2000
                3000  feet
                 I
                500
 I
500
1000m
       Figure 8. -  Pittman Lateral Study Site (adapted from U.S. Bureau of Reclamation, 1984).
Quaternary Alluvium

       Quaternary alluvial fan and valley-fill deposits originating in the McCullough Range and  River
Mountains directly overlie the Tertiary Muddy Creek Formation at the study site. The U.S. Bureau of
Reclamation (1982) describes two distinct alluvial deposits near the site. An older fan deposit, consisting of poorly
stratified, weakly indurated, light brown to buff, gravelly sand and cobbles, is usually found as a layer 3-m to
6-m-thick overlying the. Muddy Creek Formation. This unit tends to average 40 to 60 percent sand, 20 to 50
percent igneous gravel and cobbles, and 10 to 20 percent non-plastic fines (U.S. Bureau of Reclamation, 1982;
Fordham et d., 1984). Thickness of these deposits is controlled by the shape of the erosional surface of the Muddy
Creek Formation, with a maximum thickness of 78 m being recorded in a well east of Henderson. Hall (1986)
and U.S. Bureau of Reclamation (1982) report local strongly  caliche-cemented sands and gravels above the
Muddy Creek Formation-alluvium contact.

       Younger alluvial fan and valley-fill deposits are also described by U.S. Bureau of Reclamation (1982) as
poorly sorted, unconsolidated sands and gravels, consisting of 50 to 60 percent sands, 40 to 50 percent fine
                                                 18

-------
subangular gravels, and traces of non-plastic fines. Maximum thickness of these deposits near the study site is
approximately 9 m.

      lithologic logs from observation and monitoring wells drilled in the area reveal several deep and narrow
channels eroded into the top of the Muddy Creek Formation (Figure 9). Geraghty and Miller, Inc., (1980) report
                                                                                     N
       Figure 9. -  Configuration and altitude of the top of the Muddy Creek Formation, Henderson,
                   Nevada (elevation contours in feet above mean sea level) (adapted from U.S. Bureau
                   of Reclamation, 1984).

that these channels are often filled with moderately well-sorted deposits of sand and gravel resembling "gravel
trains," which have been buried by subsequent alluvial fan deposits. Near the study site, U.S. Bureau of
Reclamation (1982) labels these deposits as the younger Quaternary alluvial fan and valley-fill mixtures
described above.

Muddy Creek Formation

       Most, if not all, of the Las Vegas Valley  is believed to be underlain by the Tertiary Muddy Creek
Formation, described by Longwell et al. (1965) as valley-fill deposits consisting of several fades which intergrade
laterally from coarse-grained  deposits near mountain fronts to progressively finer-grained deposits in basin
lowlands. It is believed to have been deposited in an interiorly drained basin, or basins, prior to development of
the present Colorado River drainage system. The formation is characterized lithologically as clays, sandy clays,
                                                  19

-------
silty days, gypsiferous sandy days, clayey sands and conglomerates (Fordham a a/., 1984). Malmberg (1%5) and
Geraghty and Miller, Inc. (1980) describe the formation as consisting of thin layers of sand with some gravel,
interbedded with thick beds of clay. Finer-grained fades are typically light-colored, ranging from reddish-tan
to light green or white (Geraghty and Miller, Inc., 1980). Exposures of Muddy Creek Formation within the
eastern portion of the Las Vegas Valley are rare, occurring only along the southern edge of Frenchman Mountain
and near Whitney Mesa, west of Henderson, where exposed sediments consist of a sequence of reddish to pink
days and silts up to 6 m thick (Kaufman, 1978).

       lithologic logs from wells installed along the Pittman Lateral reveal a sharp delineation between the
white to brown days and silts of the Muddy Creek Formation and the overlying coarse sands and gravels of the
alluvial fan deposits (Figure 10). The sharp contrast allows for easy determination of the top of the Muddy Creek
Formation at depths ranging from approximately 3 m to 20 m below ground surface (Fordham et al., 1984).
However, U.S. Bureau of Reclamation (1984) notes that the top of the "first day formation" may be a more recent
deposit than the Muddy Creek Formation and may or may not be contiguous with it.
                    660
          West
646      640
                               0        500
                                Sato In M
63B     630
Station Number
620
618
                                    0      	200
                                      SCH6) In ITMtMS
                                                                                      I
610
East
           Figure 10. - Hydrogeologic cross-section of the Pittman Lateral site (adapted from U.S.
                      Bureau of Reclamation, 1984).
       The thickness of the Muddy Creek Formation is largely unknown, but probably varies greatly throughout
 most of the valley. Malmberg (1965) reported approximately 930 m of sediments logged as Muddy Creek
 Formation from an oil test well near Whitney Mesa. Another deep exploration well, drilled within the BMI
 complex by the Stauffer Geology Department, contained 660 m of Muddy Creek sediments.

 HYDROGEOLOCY

       The aquifer system within the Las Vegas Valley is complex, consisting of coarse-grained alluvial sands
 and gravels interbedded with finer-grained valley fill deposits. These interfingering units of the fine-grained
 Muddy Creek Formation have led to the development of artesian, semi-artesian and unconfined ground-water
 conditions throughout the valley. Maxey and Jameson (1948) identified four principal aquifer zones in the Las
 Vegas Valley. Three artesian aquifers, designated as shallow, middle and deep, are penetrated by wells at 60 m
 to 140 m, 150 m to 215 m, and over 215 m. Fault zones and semi-confining materialsprovide hydraulic connections
 which allow  for generally upward leakage between aquifer  zones. A fourth major  zone, the  so-called
 "near-surface" aquifer, is generally found overlying the Muddy Creek Formation within the Quaternary alluvial
 fan  and valley-fill deposits. Ground water in the near-surface aquifer occurs primarily under water-table
                                                 20

-------
(unconfined) conditions although small areas of artesian conditions may exist where ground water has been
confined by caliche layers or by lenses of low-permeability sediments (Kaufman, 1978; Fordham et al., 1984).

       Prior to extensive development of the ground-water resources in the valley, virtually all recharge to the
near-surface aquifer was by upward leakage from deeper, artesian aquifers. Recharge due to precipitation is
generally considered to be negligible  (Malmberg, 1965). Natural discharge was primarily to  springs, direct
evaporation (where the water table was near the ground surface), and phreatophyte evapotranspiration, with
little or no water discharging to Las Vegas Wash (Malmberg, 1965; U.S. Bureau of Reclamation, 1982).

       In recent years, however, recharge to the near-surface aquifer, particularly in the lower Las Vegas Valley
(the "lower" valley is defined here as that portion of the valley which lies to the east of Whitney Mesa and south
of Frenchman Mountain), has changed considerably. Urbanization of the valley has increased recharge to the
near-surface aquifer by increasing irrigation infiltration and by removing phreatophyte vegetation near the
washes. Primary sources of recharge to the shallow system in the Henderson area now include infiltrating
irrigation water, sewage-treatment-plant effluent, industrial effluent, and the upward leakage of ground water
from deeper aquifers (Kaufman, 1978).

       From the early  1940s to 1977, it was the policy of several of the companies within the BMI complex to
discharge industrial waste water into  unlined ditches and evaporation ponds north and east of Henderson.
Malmberg (1965) noted that the ensuing recharge caused ground-water levels in the Pittman area to rise to within
a few meters of land surface. As noted  earlier, rising ground-water levels flooded basements in Pittman as early
as 1942. To alleviate this flooding, waste-disposal ponds south of Pittman were abandoned, and over the ensuing
two years, the water table near the center of Pittman dropped up to 3.5 m (U.S. Bureau of Reclamation, 1984).


       Discharge from the near-surface aquifer in the lower Las Vegas Valley area is by seepage into Las Vegas
 Wash and by evapotranspiration. No large production wells tap the near-surface aquifer in this area, primarily
 because of  poor water quality (Geraghty and Miller, Inc., 1980).

        Geraghty and Miller, Inc. (1980), in their ground-water investigation of the nearby Stauffer Chemical
 Company property, noted three primary geologic factors which govern the occurrence and movement of ground
 water at that  site. It  may be assumed that these factors also play a major role at the  current study site,
 approximately 3 km north of the Stauffer site. These factors are:

         -  The configuration and slope of the top surface of the Muddy Creek Formation,

         -   The lithology of the Muddy Creek Formation, and
         -   The presence of deep erosional channels in the surface of the Muddy Creek Formation
             that contain relatively high-permeability gravel train deposits.
        Although the Muddy Creek Formation may control direction of groundwater movement, the primary
  water-bearing zone at the study site is the Quaternary alluvial fan and valley-fill deposits overlying the Muddy
  Creek Formation. Near the Pittman Lateral site, the ground surface, water table and the top of the Muddy Creek
  Formation all dip gently to the north and northeast toward Las Vegas Wash (Figures 7,9, and 11). Ground water
  generally flows to the north, but gradients may vary locally with changes in the surface of the clay aquitard, or
  with  heterogeneities within the Quaternary deposits. Depth to ground water along the Lateral is approximately
  3 m,  although it may vary from 2 m to 4 m. Saturated thickness of the near-surface aquifer ranges from 8 m to
  approximately 14 m across the breadth of the Pittman Lateral Transect. The similarities between the gradients
  of the water table and the top of the Muddy Creek Formation suggest that the configuration of the surface of
  the Muddy Creek Formation is controlling the direction of ground-water flow at this site, just as Geraghty and
  Miller, Inc., (1980) concluded for the Stauffer site.
        The Muddy Creek Formation, at the Pittman Lateral site and throughout much of the surrounding area,
  generally consists of thick units of clay and silt interbedded with a few thin layers of sand and gravel. The low
                                                  21

-------
        CONTOUR INTERVAL - 10 feat
                                                     5000500  1500  feet
                                                                                          N
          Figure 11. - Water table elevation map, Henderson, Nevada (elevations in feet above
                      mean sea level) (adapted from U.S. Bureau of Reclamation, 1984).
permeability of these sediments creates an effective lower barrier to ground-water flow from the overlying
aquifer. However, very little information regarding the hydraulic characteristics of the alluvial aquifer or the
Muddy Creek Formation in the lower valley region is available. While some information has been published for
sites within the BMI complex, little has been produced for the Pittman area. Geraghty and Miller, Inc., (1980)
performed  no  formal aquifer tests on wells penetrating a saturated sand layer in the upper Muddy Creek
Formation, but noted the extremely low yield indicated by the large drawdowns measured while pumping at low
rates. However, vertical permeability tests were performed on undisturbed cores taken from the upper Muddy
Creek Formation. Values ranged from 2.0 x 10~8 to 5.6 x 10~10 m/s, confirming the extremely low permeability
of the upper Muddy Creek Formation. Hall (1986) reported a transmissivity of 1.3 x 10~5 m2/s obtained during
an aquifer test on the Stauffer site.
                                                22

-------
       Aquifer tests performed by Geraghty and Miller, Inc. (1980), on two wells located within the BMI complex
and completed in the Quaternary alluvium resulted in estimated transmissivities ranging from 1.9 x 1(H to 2.3
x 1Q-4 m2/s. These values were considered to be typical of the near-surface aquifer. A third well, completed in
one of the deep channels in the surface of the Muddy Creek Formation, yielded a calculated transmissivity of
2.1 x 10-3 m2/s. Another well, located in a larger channel, yielded a transmissivity of 9.1 x 10~3 m2/s. Hall (1986)
reported transmissivities ranging from 2.0 x 10~5 to 4.5 x 10~3 m2/s within the alluvial aquifer on the Stauffer site.
The large range in values was attributed to the anisotropic conditions common in alluvial fan deposits. Calculated
hydraulic conductivities were found to be greatest in the channel-fill deposits within the paleochannels of the
Muddy Creek Formation surface. Fordham et al. (1984) note that on the TTMET, Inc., site (on BMI property),
ground-water flow is generally down the slope of the Muddy Creek Formation toward the deep channels, which
transmit the bulk of the flow to the north toward the Pittman Lateral site and Las Vegas Wash.

       Estimates of ground-water flow rates vary from site to site. Geraghty and Miller, Inc., (1980) estimated
a maximum ground-water velocity of 1.5 m/day to 6.1 m/day, based on the greatest calculated transmissivity at
the Stauffer site. Similarly, Fordham et al. (1984) estimated a maximum flow rate of 10.7 m/day based on a similar
transmissivity, but with a slightly higher gradient and a porosity of 0.3. Ecology and Environment, Inc., (1984)
report a hydraulic gradient of 0.012 and a flow velocity in excess of 0.8 m/day in the Pittman area. Hall (1986)
reported a velocity of 1.6 m/day with a maximum of 7.3 m/day at the Stauffer site.

       The barrier to downward vertical ground-water flow presented by the low-permeability sediments of the
Muddy Creek Formation is augmented by an upward vertical gradient from  the deeper artesian aquifers.
Geraghty and Miller, Inc., (1980) cite head data from wells tapping confined sand zones within the Muddy Creek
Formation. Measured heads in these wells ranged up to 10.4 m higher than in adjacent wells completed in the
near-surface aquifer. The U.S. Bureau of Reclamation (1982) noted upward gradients in the Pittman area and
below the BMI ponds near Las Vegas Wash.

HYDROCHEMISTRY

       Ground water in the near-surface aquifer in the lower Las Vegas Valley is generally of very poor quality.
The chemical composition of the water is a reflection of the type of material through which the water moves,
residence time, length of flow path, and the direct and indirect effects of urbanization and industrialization within
the valley (Malmberg, 1965; Fordham et al., 1984). The ground water in this region is characterized by high
 concentrations of total dissolved solids (TDS), sodium, chloride, and sulfate. Locally, significant levels of organic
 contaminants are found.

        Large portions of the lower valley contain shallow ground water with  greater than  5,000 mg/L TDS
 (Figure 12). The high dissolved solids content may be attributed to several factors. Where high water table
 conditions exist, evapotranspiration processes concentrate salts in the subsurface. Dissolution of evaporite
 minerals along flow paths may add considerable amounts of dissolved constituents to ground water. U.S. Bureau
 of Reclamation (1982) noted that many high-TDS areas correspond to large deposits of evaporite minerals.
 Sulfate, chloride, and sodium ion concentrations tend to increase from west to east across the valley, and these
 increases are generally attributed to dissolution of gypsum  and halite, both of which are abundant in the valley
 fill materials, particularly  in the Muddy Creek Formation (U.S. Bureau of Reclamation, 1982). Urbanization
 within the valley has contributed both directly and indirectly to the high dissolved solids content of ground water
 in the near-surface aquifer. Infiltrating water from lawn watering, agricultural development, or industrial
 waste-water disposal leaches salts from the soil profile and into the underlying aquifer. Where infiltration is
 excessive, a rising water table brings ground water into direct contact with readily dissolved salts stored in the
 vadose zone. Dissolution of these minerals increases the TDS content of the ground water (U.S. Bureau of
 Reclamation, 1984; Kaufman, 1978).

        Ground water  at the study site is contaminated by a variety of organic and inorganic compounds. Much
 of the organic contamination may be traced to the accidental storage tank leak in 1976 noted earlier. Since that


                                                  23

-------
               CONTOUB
             < iaooo -  2,000 mi/1
             > 10,000 -  10.000 mg/l
                                                    500 0 500   1500 feet
N
    Figure 12. - Contour map of total dissolved solids concentrations (mg/L) in ground water, Henderson,
               Nevada (adapted from U.S. Bureau of Reclamation, 1984).
time, a benzene plume has migrated to the north and northeast toward Las Vegas Wash (Figure 13). Ecology
and Environment, Inc., (1984) and Geraghty and Miller, Inc., (1980) cite benzene concentrations which range
from  in excess  of 500,000 mg/L near the source of the leak down to approximately 10 mg/L near Pittman.
Chlorobenzene and chloroform have also been detected in relatively high concentrations in the wells which make
up the Pittman  Lateral Transect.

       Passage of the plume under older waste disposal areas as it moved downgradient has raised additional
concerns. Geraghty and Miller, Inc., (1980) and Ecology and Environment, Inc., (1985)  suggest the possibility
that other organic compounds present in these disposal areas have been mobilized in the presence of the nigh
benzene concentrations and have moved downgradient with the contaminant plume.

-------
                                                                                      N
                                     800
                                                       ago 9 ago . iqeo
     Figure 13. - Map of benzene concentrations (ppm) in ground water, Henderson, Nevada (adapted
                from Ecology and Environment, Inc., 1984).


MONITORING WELL INSTALLATION

       Due to the design and operating procedures of the selected sampling devices, as well as the objectives
of the study, it was necessary to install the wells close to each other. Six wells were installed in late July 1987
near the mid-point of the Pittman Lateral Transect. Although it was desirable for the wells to be close to each
other, it was also necessary for the wells to be spaced far enough apart so that the effects of pumping at one well
would not significantly impact the distribution of VOCs near adjacent wells. To determine the minimum well
separation, drawdown, and capture-zone, calculations were made utilizing estimates of hydraulic properties of
the aquifer. These calculations indicated that a well-spacing of 6 m would provide sufficient hydraulic isolation
between wells. The 6-m spacing was also felt to be small enough to result in similar geologic and hydrochemical
conditions over the site. The wells are arranged approximately 6 m apart in a rectangular grid with the long axis
of the rectangle roughly paralleling the northward direction of ground-water flow (Figure 14).

       A dual-tube percussion hammer drill rig was utilized to  install the wells. Lithologic  logs from each
borehole typically showed approximately 6.4 m to 7.5 m of silty sands and gravels overlying a very hard layer of
                                                 25

-------
                                                                         ,637^

                                                                         Fence
                                       Gate
                                                        Building
                                          Well Designation
                                     Wei	SampHng Device
                                       1        Bladder Pump
                                       2        BATPTFE
                                       3        Westbay
                                       4        Bat Well Probe
                                       5        BAT HOPE
                                       6        Pump/Packer
                                    637,639     Pittman Lateral Welts
N
           Figure 14. - Monitoring well layout, Pittman Lateral site (not to scale).


cemented sands and gravels (caliche), which was found at depths ranging from 6.4 m to 7.6 m below ground
surface. This layer was penetrated in well 6 at 9.6 m, at which point the drill bit immediately entered a
greenish-white clay interpreted to be the Muddy Creek Formation. Drilling for all other wells was terminated
before or directly after contacting the top of the caliche layer, although some of the holes were driven further
into the caliche to facilitate placement of the well intakes. Depth to ground water was found to be approximately
3m.

       To further establish the similarities between the six wells, as well as to select the proper depths for well
intake placement, two boreholes at opposite corners of the rectangular grid (wells 1 and 6) were logged with
magnetic induction and natural-gamma logging tools. Prior to logging, the holes were temporarily cased with
10.2-cm PVC pipe to prevent collapse of the borehole walls. Interpretation of the geophysical and lithologic logs
suggested that each well intake should be placed within a zone extending from 6 to 7 m below ground surface.
                                                 26

-------
The logs suggested a lower clay content and perhaps better flow characteristics in this zone than in the deposits
above or below. Since the two logged boreholes were at opposite comers of a very small rectangular grid, it was
assumed that the hydrogeologic conditions would be very similar in the other wells. After running the geophysical
logs, the temporary casing was removed from the borehole.

       Five of the six wells were cased with 5.1-cm-diameter Teflon casing with well intakes placed within the
targeted zone described above. Three of the wells (1, 4, and 6) were screened over a 03 m-interval. Two more
contained the HDPE and PTFE filter tips with 5.1-cm-diameter Teflon riser pipes (wells 2 and 5). The sixth
monitoring well consisted of the stainless steel multi-port with measurement ports placed at 3 m, 4.5 m, and 6
m below ground surface (well 3). Only data from the lowermost port would be utilized in the comparison study.
The other two ports were installed to investigate vertical variations within the contaminant plume. Each well
was backfilled with a gravel pack extending approximately 0.6 m above and below the well intake point. A thin
layer of fine silica sand was installed above the gravel pack followed by a surface seal consisting of a 5 percent
cement-bentonite slurry. The multi-port well design was somewhat complicated by the multiple monitoring
zones, but the installation was completed in the same general manner as for the other wells. Well completion
diagrams and generalized lithologic logs are presented in Appendix A.

       The three conventional monitoring wells were developed by pumping at a rate of approximately 2.8 L/min
with one of the bladder pumps. Volumes of 278, 320, and 354 L were removed during development from wells
6,1, and 4, respectively. The multi-port installation was developed by opening the pumping port and airlifting
about 150 L from the casing at a rate of approximately 1.9 L/min. The two filter tip installations were developed
by collecting five 500 mL sample vials from each. At all installations, development continued until sample
temperature, pH, and electrical conductivity stabilized and a clear sample was obtained. All water was disposed
of off-site. The lower volume needed for the development of the filter tips may have resulted from the smaller
volume around the filter tip that was impacted by development and the much lower sampling rate (estimated
to be about 60 mL/min under these conditions). After development of the multi-port installation was completed,
the pumping port was closed and any water remaining inside the casing airlifted out.
                                                 27

-------
                                           SECTION 4

                                         EXPERIMENT A
EXPERIMENTAL DESIGN
      The experimental design of the initial portion of this study included the evaluation of differences between
the six wells, selection of a suitable sampling frequency, and implementation of the sampling program. Because
all field sampling techniques introduce bias into VOC determinations,  a true assessment of accuracy, and
therefore representativeness of ground-water conditions is not possible in a field study. However, because of
the physical and chemical properties of most VOCs, losses of VOCs from the sample are much more likely than
increases. Therefore, a relative approximation of accuracy can be made based on the concentrations of VOCs
recovered during the sampling process (i.e., those devices which recover the highest levels of VOCs may be
considered the most accurate). A detailed accounting of the experimental design, as well as sampling procedures
and quality assurance objectives, may be found in Desert Research Institute (1987), an EPA-approved quality
assurance plan.

      Because sampling device variability was the source of variability of most interest to this study, all other
Sources Of vanabil'ly present in any field study werp mimmrepd as much as possible. As previously described, since
the comparison of sampling devices could not be based on samples collected from a single monitoring well, six
wells were installed in a manner that resulted in similar geologic conditions at each well. However, due to the
natural variability in lithology and stratigraphy of alluvial materials, geologic conditions cannot be considered
identical at each well. The variability in geologic conditions may be an important element of the total variability
observed in this study, but could not be adequately quantified in a way that would relate to effects on VOC
concentration variability.

      To compare between sampling devices installed in separate wells, the further assumption was made that
hydrochemical conditions were the same at each of the six well sites. To test this assumption and to determine
the magnitude of expected laboratory errors, a "survey" sampling round was conducted in which an initial set
of samples was collected from wells 1,4, and 6. These three wells were utilized for the survey sampling because
they were the only wells that could be sampled with the same device, a Teflon bladder pump, and  because their
locatkms represented a fairly complete coverage of the study site. Due to their design, the other three wells (2,
3, and 5) could not be sampled with the bladder pump and, therefore, were not included in the survey sampling.
Each of the sampled wells was purged at a rate of approximately 1 L/min until a minimum of five well volumes
had been removed. A bulk sample was then collected from each veil in a 19-L container and divided into five
sample splits. Samples were analyzed for pH, electrical conductivity, gross chemistry, total dissolved solids,
dissolved silica, iron, manganese, total organic caibon, benzene, and  chlorobenzene.  Laboratory analysis
methods are briefly summarized in Appendix B. The analytical results for the sample splits and calculated means
and standard deviations are presented in Appendix C.

      Plots of the benzene, chlorobenzene, and total organic carbon (TOC) data and their mean values and
standard deviations are shown in Figures 15 through 17. The means of the five splits from each well are all within
                                                28

-------
   400
a.  300
§
To
o>
e
at
to
    200
    100
                               Well Number
  Figure 15. - Plot of benzene concentration (ng/L) for each of five
             sample splits from wells 1,4, and 6.
    2000


 _ 1800


 f" 1600
 o

 I 1400


 o> 1200


 | 1000


 0  800


     600
                                Well Number
 Figure 16. - Plot of chlorobenzene concentration Oig/L) for each of five
             sample splits from wells 1,4, and 6.
                                   29

-------
                    11.0
                 — 10.0
                     9.0
                     8.0
                                                                         r
                                                                         I;
                                                Well Number
                  Figure 17. - Plot of TOC concentration (mg/L) for each of five
                             sample splits from wells 1,4, and 6.
approximately one standard deviation of each other, while the differences between wells for all the splits fall
well within the stated analytical error of .±.20 percent for benzene and chlorobenzene and ±_5 percent for TOC.
The ranges in concentration means between the three wells were approximately SO ug/L, ISO ng/L, and 0.4 mg/L
for benzene, chlorobenzene, and TOC  respectively. Inspection of the inorganic data showed that, with the
exception of manganese, the differences between the sample means of the three wells  were also all within
laboratory-stated analytical error.

       The low, although present, hydrochemical variation indicated by the survey sampling, close proximity of
the wells, and the similar lithologic and geophysical logs, all suggest that the geological and hydrochemical
conditions at the site are fairly uniform.  For the purposes of the sampler comparison, these conditions formed
the basis for the assumption that geologic and hydrochemical conditions were essentially the same at each of the
six well sites. However, the magnitude of the small hydrochemical variation observed during the survey sampling
was kept in mind when analyzing the data for sources of concentration variation. Unfortunately, because of the
nature of this field study, the devices used, and  the time period over which the study took place, the
hydrochemical variation could only be  approximated during the survey  sampling and could not be further
investigated as the study proceeded.  As a result, temporal and spatial hydrochemical variation remained a
possible source of variation in sampler performance.

       Because of the overall project emphasis on sampling device effect on VOC concentration, only the VOC
and TOC data were utilized in the experiment A comparisons. While inorganic chemistry samples were collected,
the resulting data were not statistically analyzed for this report. The inorganic data were used primarily for
understanding geochemical conditions.

       Sampling frequency was chosen based on the need to collect a large number of samples at closely spaced
intervals and the desire to examine temporal variations associated with varying sampling intervals. Samples were
collected weekly during the first month, biweekly during the second month, and during the  13th and 19th weeks.
The first set of samples was collected during the week of February 8,1988. Each sampling week, the wells were
                                                30

-------
sampled in a random order determined with the aid of a table of random digits. Table 1 lists the sampling devices
that were utilized during Experiment A.
                   TABLE 1. SAMPLING DEVICES UTILIZED IN EXPERIMENT A

                         Device Type                            Well Number

                        Bladder Pump                               1
                        Bladder Pump/Packer                         6
                        HOPE Filter Tip                             5
                        Multi-Port                                  3
                        PTFE Filter Tip                              2
                        Standard Bailer                              1
	Well Probe	    4	

       Wells 1 and 6 were purged prior to sampling by pumping (with the respective dedicated bladder pumps)
at a rate of 1 L/min until solution parameters (temperature, pH, electrical conductivity) stabilized (±.10 percent
over two successive well volumes) and at least five well volumes had been removed from the well. Pump discharge
was then reduced to less than 500 mL/min for sampling. Sampling from well 1 was always accomplished by purging
and sampling with the bladder pump prior to sampling with the bailer. The first two bailer volumes collected were
discarded in order for the device to be thoroughly rinsed with formation water prior to collecting a sample. Bailer
samples were first transferred to a dean glass beaker, then poured into appropriate sample bottles.

       Samples were collected from well 6 with the bladder pump/packer. The well was purged and sampled only
 after inflation of the packer. Well volume calculations accounted only for that volume of water in the well below
 the packer.

       Sampling procedures for the in situ devices and well probe were carried out as previously described.
 Because the wells were shallow and hydraulic heads low, a small vacuum was applied to the multi-port sample
 bottles to obtain a full bottle. This was accomplished through the use of a special adapter and a hand-held vacuum
 pump. This procedure was not expected to adversely affect sample integrity because the flow-through design
 of the sample probe meant that the sample collected was the last water to pass through the bottle, when pressure
 conditions were near static.

       Sample temperature, pH, electrical conductivity, dissolved oxygen, and bicarbonate concentration were
 measured in the field. Measurements taken after the final purge volume had been removed from wells 1 and
 6 were considered to be representative of in situ conditions. Measurements from bailed samples took place only
 after the two initial bailer volumes had been discarded. Measurements from the in situ installations and the well
 probe took place only after at least one sample volume had been removed and discarded.

        Each full set of samples collected during the week also included two sets of equipment blanks (one set
 each for the bailer and bladder pump), one set of standards solutions, one set of trip blanks, and one set of
 duplicate samples collected from a randomly selected well.

 RESULTS

        The ground-water chemistry data resulting from the analysis of samples collected with each sampling
 device, as well as standard solutions and blank samples, are given in Appendix C. As noted earlier, project
 emphasis was placed on  the effect of the devices on VOC and TOC concentrations. The comparison described
 below was based on the  organic chemistry data alone.


                                                 31

-------
Benzene and CWorobenzene

      The laboratory analysis results for benzene and chlorobenzene recovery revealed variation in the data
between sampling devices as well as with time. Plots of recovered concentration versus time (sampling week)
for both benzene and chlorobenzene graphically illustrate the differences that exist between sampling devices
or sites with regard to VOC recovery (Figures 18 and 19). For benzene, the HDPE filter tip recovered the highest
concentrations followed by the well probe and bladder pump, while, with the exception of sample round 6, the
multi-port recovered the lowest concentrations. For chlorobenzene, the results are more variable but the general
trends are that all devices recovered comparable concentrations although the multi-port and bailer recovered
lower concentrations.

       In a subsequent experiment, the cause  of the low concentrations recovered by the multi-port was
identified as being related to the manufacturer's recommended procedures for sample collection. Correction
of this problem resulted in collection of much higher VOC concentrations, and will be further discussed later
in this report. Owing to this change, the multi-port results from experiment A are not representative of the more
effective sampling procedures incorporated in later experiments, but are included here because they were
collected for this experiment using the originally-stated manufacturer's procedures.

       Figures 18 and 19 also reveal a definite decrease in VOC concentration with time, particularly with regard
to the benzene concentrations. Possible  explanations include natural degradation of the organic contaminants,
as the plume moves downgradient, a gradual shift of the spatial configuration of the plume, or the movement
of cleaner recharge water, possibly due to npgradient irrigation, ground-water remediation efforts, wastewater
disposal  practices, or precipitation patterns throughout the study area. Because the concentration  decreases
appear to have affected all  six wells in the same general manner, and because pre-sampling in situ conditions
were considered to  have been essentially the same at each well, it may be suggested that VOC conditions are
changing at each well location in the same manner.

       A comparison of experiment A results with the results of the survey sampling indicates that, in general,
apparent sampler variation  exceeded estimated hydrochemical variation. The experiment A benzene data from
wells 1,  4, and 6 (the only wells that could be sampled during the survey sampling) showed a range  in
concentrations between wells at least twice the range in  benzene concentration means observed during the
survey sampling. It is important to note, however, that a  significant portion of sampler variation in benzene
concentrations observed in some rounds of experiment A may be attributed to hydrochemical variation, although
the magnitude of this variation was not known during the sampling rounds themselves. This suggests that the
differences between sampling devices, at least for the bladder pumps, bailer, and well probe, which were used
in wells 1, 4, and 6, may in  some cases be small enough to be indistinguishable from hydrochemical variation.
For chlorobenzene, the sampler variation was considerably greater than the range in concentration means
observed in the survey sampling, so it appears that sampler variation may be an important source of variation
in the experiment A chlorobenzene data.

       The benzene and chlorobenzene data were statistically  analyzed by means of a two-way analysis of
variance to determine if the sampling devices and/or sampling time introduced a significant source of variation
to the data. For both compounds, the results of theanafysisslHW that both factors were significant at meSpercent
level, suggesting that the devices differ in their ability to recover volatile organic compounds in ground-water
samples. However,  the analysis does not reveal where the significant differences occur. Although beyond the
scope of this study,  additional research will evaluate the temporal trends observed in the chemical data.

       To quantify the importance of the observed concentration  differences between devices, a Tukey multiple
comparison test (Steel and Tonic, 1960) was conducted on the  data set. The Tukey test, which tests for
significance between individual pairs of sample means, can determine which means of individual devices differ
at the dCCTTd yigntfin»nre liyd. The ftevices may then he gronped accordingly by bracketing those devices whose
mean concentrations show no significant differences. To apply the Tukey test to Experiment A data, the means


                                                 32

-------
500
                                                      Bladder Pump
                                                      Bladder Pump/
                                                       Packer
                                                      Multi-Port
         HOPE Filter Tip
         PTFE Filter Tip
         Well Probe
         Bailer
    123468                   13                      19
                                 Sampling Week
    Figure 18. - Benzene concentration (p.g/L) recovered by each sampling device
                as a function of time.
  500


  250
HOPE Fitter Tip
PTFE Filter Tip
Well Probe
Bailer    	
 Bladder Pump
 Bladder Pump/
  Packer
• Multi-Port
                                              •    •
                                                             -I	L.
     123468                  13                      19
                                  Sampling Week
     Figure 19. - Chlorobenzene concentration (ng/L) recovered by each sampling
                 device as a function of time.
                                      33

-------
of the concentrations recovered by each device over the eight sample rounds were calculated and compared to
each other. This results in an overall comparison based on all the samples collected over the 19-week experiment,
but the application of the Tukey test in this way is only valid if the concentrations follow similar trends over time
for each device. As shown on Figures 18 and 19, the benzene trends for each device are fairly similar while the
chlorobenzene trends are variable.  For this reason, the results of the Tukey test may be less valid for the
chlorobenzene data than for the benzene data.

      The results, shown in Table 2, indicate that at the 5 percent confidence level the HDPE filter tip recovered
significantly higher benzene concentrations than the other devices. The well probe and bladder pump results,
which were indistinguishable from each other, were lower than the HDPE filter tip and significantly higher than
the other devices. The bladder pump/packer, PTFE filter tip, and bailer formed a third grouping lower than the
devices mentioned above but significantly higher than the multi-port, which recovered the lowest benzene
concentrations. Again, these low multi-port results were related to sample collection procedures that were later
modified to incorporate more effective methods.

                TABLE 2.   TUKEY TEST RESULTS FOR EXPERIMENT A

            Rank*                  Benzene                    Chlorobenzene
               1                  HDPE Filter Tip

               2                f Well Probe

               3                I Bladder Pump
               4
               5
               6
Bladder Pump/Packer

PTFE Filter Tip
 Well Probe

 Bladder Pump/Packer

 Bladder Pump

 PTFE Filter Tip

.HDPE Filter Tip
Standard Bailer               I   Standard Bailer
               7                 Multi-Port                      Multi-Port
 •Ranked from highest to lowest concentrations

       The lack of consistent chlorobenzene concentration trends between devices suggests that the Tukey test
 may not be strictly valid for this compound, however, some conclusions may be drawn if the results are used in
 conjunction with the trends evident on Figure 19. Although considerable variation is evident in the concentration
 trends, the HDPE filter tip, PTFE filter tip, well probe, bladder pump, and bladder pump/packer form a group
 that exhibits similar mean values over the eight sample rounds. In contrast, the bailer shows considerable
 variation  and lower concentrations than this group and the multi-port consistently recovered the lowest
 concentrations. By combining this information with the chlorobenzene Tukey test results, it appears that all of
 the devices form a general grouping with the exception of the multi-port, which recovered significantly lower
 concentrations, and the bailer, which varied in concentration between the multi-port and the group of other
 devices.

 Total Organic Carbon

       A plot of the concentrations of total organic carbon (TOC) recovered by each sampling device versus
 sampling week (Figure 20) shows an overall decrease in concentration with time similar to that seen in the
 benzene and chlorobenzene data. However, these data are highly variable, and a  clear distinction between
 sampling devices is not readily apparent. For most sample rounds, the variation between samplers in wells 1,4,
 and 6 was less than the TOC concentration variation observed during the survey sampling. In addition, continual
                                                34

-------
problems with contamination of TOC samples added a component of variation that makes most of these results
suspect. In general, however, the bladder pumps, well probe, and PIPE filter tip recovered the highest TOC
concentrations while the HOPE filter tip recovered the lowest.
            11.0
•— • HOPE Filter Tip
•— • PTFE Filter Up
*— Well Probe
«— « Bailer
•— • Bladder Pump
•— • Bladder Pump/
Packer
«— * Multi-Port
                 1234
8
Sampling Week
13
19
                Figure 20. - TOC concentration (mg/L) recovered by each sampling
                           device as a function of time.
       The TOC data were subjected to a two-way analysis of variance similar to that performed on the benzene
 and chlorobenzene data. Again, both factors, sampling device and sampling time, were found to be significant
 at the 95 percent level of significance, but due to the high variability in the TOC data, the Tukey test could not
 be applied and no estimate of device accuracy could be made.

 Week 19 Experiments

       The relatively low VOC concentrations recovered by the multi-port prompted development of an
 additional set of experiments designed to isolate the cause of this apparent problem. These experiments consisted
 of collecting multiple sample replicates from the multi-port utilizing a variety of sampling methods and sample
 handling  procedures. Detailed descriptions of these experiments are contained in  Appendix C. The results
 indicate that VOC  sample integrity was adversely affected by  perforated septa that resulted from the
 manufacturer's recommended sample collection method. Multi-port samples that utilized unpunctured septa
 exhibited benzene and chlorobenzene concentrations that were approximately three times higher than those
 samples utilizing punctured septa. In addition, the samples  with unpunctured septa showed benzene and
 chlorobenzene concentrations comparable to those collected with the filter tips and bladder pump/packer (the
 bladder pump and well probe samples were compromised prior to analysis). The sample collection procedures
 were subsequently modified by the manufacturer to include replacement of punctured septa with new,
 unpunctured ones.
                                                 35

-------
      The Week 19 data suggest that the low VOC concentrations observed in the multi-port samples during
sample weeks 1 through 13 were the result of losses of VOCs from the sample vials themselves and were probably
not related to the nmlti-port device, installation, or the presence of lower VOC concentrations near the
multi-port installation.
      Punctured septa on the 40 mL VOA vials may have contributed to VOC losses by several mechanisms.
First, it is possible that partial evacuation of the VOA vials gave rise to large pressure differences between the
aquifer and the vial during sample collection, possibly leading to volatilization of VOCs into air bubbles inside
the vial that could pass out through the punctured septa. Second, passage of air into the VOA vials through the
punctured septa during sample shipment and storage could also have led to formation of air bubbles into which
VOCs could volatilize. Finally, direct volatilization of VOCs to the atmosphere may have occurred through the
vial septa.

DISCUSSION
      The primary objectives of this study were to compare the accuracy and precision of ground-water sampling
devices and determine if either of the in situ sampling methods yields representative data. But, since this study
was conducted in the field, a true assessment of representativeness was not possible because the actual in situ
VOC concentrations were unknown. Under these conditions, as was previously discussed, it was assumed that
those devices that recovered the highest VOC concentrations were the most accurate and produced samples that
were the most representative of in situ conditions.
       During experiment A, statistical assessments of accuracy and precision were difficult due to the trend of
decreasing VOC concentrations with time and the lack of adequate numbers of sample replicates. The analysis
of variance indicated that VOC concentrations for all devices changed significantly with time causing a high
degree of temporal variability within the data set while the lack of sufficient replicates resulted in an incomplete
understanding of the variability that each sampling method produced during a single sample round. In addition,
variation observed during the survey sampling suggests that  hydrochemkal conditions at the site may vary
spatially, and that this variation may be most important for the benzene and TOC data. These possible sources
of error made it difficult to distinguish between variability due to the device and variability due to spatial and
temporal concentration changes.
       If the assumption is made that spatial variation between samplers is of lower magnitude than sampler
variation, the results of experiment A suggest that, for benzene, the HOPE filter tip was the most accurate of
the seven devices tested; it recovered significantly higher benzene concentrations than the other devices. For
cnlorobenzene no single device was consistently the most accurate, as the bladder pumps, well probe, and PTFE
filter tip all recovered the highest concentrations for at least one sample round and as a group the  means over
time of these devices and the HOPE filter tip were similar, as suggested by the plot of  concentrations and the
Tukey test results. The least accurate devices for both compounds appeared to be the bailer and multi-port using
the original sampling procedures. The findings of additional experiments showed that the multi-port was capable
of performing as accurately as the bladder pumps and filter tips.
       The relatively tow benzene recovery of the PTFE fitter tip, in comparison to the  HOPE filter tip, was of
interest  because the devices recovered comparable cnlorobenzene concentrations and because  the operating
procedures were identical for the two devices. The only difference between the two devices is the materials used in
the filter tip. The PTFE filter tip material has a permeability an order of magnitude, lower than the HDPEfflter tip
material (LTiKH cm/s versus LTklO"3 cm/s). As a result of the PTFE filter's lower permeability, higher hydraulic
gradients may form across this filter during sampling, resulting in more turbulence as water flows through the filter.
These mote turbulent flow conditions may have a greater effect on benzene concentrations than chlorooenzene
concentratioas because of benzene's higher volatility, possftfy causing tower benzene recoveries with the PTFE fute
tip than  with the HOPE filter tip. A similar effect was observed in experiment C.
       Although the well probe recovered VOC concentrations comparable to the bladder pump, difficulties
were experienced in  recovering full  sample bottles during several sample rounds and no samples could be


                                                 36

-------
recovered during week 8. This problem seemed to be related to improper connections between the sample
housing guide sleeve and the well probe cap, which prevented penetration of the hypodermic needle into the
cap septum. Inspection of the cap revealed no damage, and little sediment had built up above the cap, so the
cause of this problem was not identified. Furthermore, water was observed in the well bore above the well probe
during several sample rounds. The source of this water may have been leakage past the probe O-rings or through
the filter tip cap, although the actual cause could not be identified. The manufacturer has subsequently made
several design changes to this well probe; the tested version is no longer commercially available

      The TOC data contained a great deal of variability but, in general, the bladder pumps, well probe, and
PTFE filter tip appeared to be the  most accurate,  much like the chlorobenzene results. The HDPE filter tip
appeared to be the least accurate of the seven devices for TOC recovery. Excessive variability in the TOC data
prevented determination of device precision.

      The overall results suggest that the well probe and in situ filter tips, which require only minimal purging,
may perform as well as a bladder pump. The unexpected performance of the multi-port was related to the loss
of VOCs through punctured septa on the sample vials during transport and storage, as shown by the results of
the week 19 sampling. Multi-port samples collected during weeks 1 through 13 were probably all affected by the
perforated septa, resulting in low VOC recoveries.  This problem was corrected during subsequent experiments
by replacing the punctured septa on each sample vial with new septa prior to packaging and shipment to the
laboratory.
               i
       The design of subsequent experiments in the comparison study was based upon the pattern of responses
observed during this  experiment,  the increased familiarity with the devices, and the sampling procedure
modifications that needed to be made to obtain the most  representative samples from each device. This was
particularly important for the multi-port sampling procedures because this  study was designed in part to
investigate in situ devices. More  complete information about device accuracy and precision was obtained from
the subsequent studies (experiments B and C) which are discussed in the following sections.

       The analysis of duplicate samples, standard solutions, and blank samples (Appendix C) yielded results
which fell well within the quality assurance objectives outlined in Desert Research Institute (1987). However,
these quality assurance procedures did reveal a few problems, most notably the possible contamination of the
TOC samples, which may nave contributed to significant variability in the TOC results.

       Also of some concern were the duplicate VOC samples and VOC standards. Duplicate samples collected
during weeks 2 through 4 varied greatly in terms of benzene and chlorobenzene recovery. The lack of analytical
reference standards during these weeks made it difficult to determine whether most of the variation lies within
the laboratory analysis or in the sampling methodology. In either case, the high variation among these samples
pointed out the need to collect multiple replicate  samples from each device during each sampling period.

       The laboratory analysis of the VOC standard solutions resulted in concentrations which were consistently
lower than the theoretical value.  While this deviation is most likely due to poor standard preparation, or
analytical variability, it may also be indicative of a limited sample shelf life. Although all of the samples were
analyzed within the maximum allowed holding time of 14 days (U.S. Environmental Protection Agency, 1986),
most analyses were completed near the end of this time period. If it can be assumed that the solutions were
prepared and handled properly, then sample degradation during storage should be considered as a possible
source of sample bias.
                                                  37

-------
                                          SECTIONS

                                        EXPERIMENT B
EXPERIMENTAL DESIGN
      The results of Experiment A provided a great deal of information regarding site and sampling device
characteristics and resulted in a preliminary comparison of the seven sampling devices. However, the original
experimental design needed modification to obtain a more complete evaluation of each type of sampling device.
Experiment B was conducted to resolve some of the problems encountered during Experiment A.

      Many of the original assumptions and procedures  outlined in the design  of  Experiment  A were
incorporated into the new experimental design. Most importantly, the assumption that there were no significant
differences between wells, as well as most sampling procedures, remained unchanged. However, the high natural
variability observed in the Experiment A data required that changes be made with regard to sampling frequency
and the number and type of samples collected. To avoid the high temporal variation found within the VOC data,
sampling was restricted to a single occasion.  Only VOC samples were collected for this  experiment. Multiple
replicate samples were collected in an effort to make the ensuing statistical analysis less susceptible to outliers
and high variability within the data (including analytical variability),  and to reduce the effect of lost sample data.
Six replicate VOC samples were collected with each sampling device and full  VOC  analysis results were
requested (as opposed to only the benzene, chlorobenzene, and chloroform concentrations requested during
Experiment A). Quality assurance samples (two standards, one trip  blank, and two equipment blanks) were also
prepared or collected.

      Sampling procedures for the second experiment remained virtually identical to those used during the first
phase of the study. However, after experiment A, known or suspected problems with the HOPE filter tip and well
probe precluded their use during this phase of the  comparison study. The HDPE filter tip installation was
damaged by on-site construction activity, and problems experienced with the well probe in experiment A were
discussed previously. Because well probe design was significantly modified by the manufacturer and the tested
probe was no longer commercially available, it was decided to remove it from further comparisons. In addition,
although the bladder pump/packer appeared to perform well in experiment A, this device was removed from
subsequent study because the project focus was primarily on in situ devices. VOC sampling with the multi-port
continued as in Experiment A, except that the  septum in each sample vial was carefully replaced with an
unperforated septum after sample recovery. Samples were also collected with the PTFE filter tip, bladder pump,
and bailer, as shown in Table 3. Sample replicates were collected with six separate trips of the sampling devices in
the PTFE filter tip and multi-port, in sequence from the bladder pump discharge, and with two volumes with the
standard bailer. All sampling for Experiment B took place on November 8,1988.

RESULTS

      Ground-water chemistry data  resulting  from analyses of samples collected for Experiment B are
presented in Table C.17 of Appendix C Nine VOCs were detected in each ground-water sample. A 10th,


                                               38

-------
                 TABLE 3.    SAMPLING DEVICES UTILIZED IN EXPERIMENT B
                          Device Type                    Well Number
                         Standard Bailer                       1
                         Bladder Pump                        1
                         Multi-Port                           3
                         PTFE Filter Tip                      2
acetone, was also detected in a few of the samples. Results for the quality assurance samples (standards, blanks)
are presented in Table C.18 of Appendix C.

      The VOC data are presented in Figures D.I through D.9 in Appendix D and in Table 4. In general, the
results of these analyses reveal little of the variation that was so notable during the first experiment. This is to be
expected since this experiment was designed to minimize the potential for temporal variations produced by
sampling over long time periods. A preliminary review of the data also suggests very little difference between the
four sampling devices with regard to recoveries of VOCs. However, there are a few notable exceptions (benzene
recoveries, bailer recovery of chlorobenzene).

      The tabulated data shown in Table 4 allow for a generalized comparison of the devices based on recovered
VOC concentrations. All four of the sampling devices recovered comparable levels  of all the VOCs except
benzene (and bailer recovery of chlorobenzene). The bladder pump recovered the highest mean concentrations
of benzene, followed by the bailer,  the multi-port and the PTFE filter tip. The PTFE filter tip recovered the
highest concentrations of six compounds, while the multi-port recovered the highest concentrations of the
remaining two. The raw data suggest little overall difference between the performances of these devices and that
of the bladder pump (with the exception of benzene, the multi-port and PTFE filter tip sample means were
generally greater than, or within one standard deviation of,  the respective bladder pump mean recovered
concentrations). In addition, overall variation between devices was slightly less (for benzene) and of roughly the
same magnitude (for chlorobenzene) than the variation observed in the concentrations of these compounds
during the survey  sampling. Because  benzene  concentrations had declined significantly since  the survey
        TABLE 4.   VOC CONCENTRATION MEANS Oig/D AND STANDARD DEVIATIONS
                    OF THE FOUR SAMPLING DEVICES UTILIZED IN EXPERIMENT B
PTFE
Filter Tip
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
1,1-Dichloroethane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
42
1290
219
20
358
33
10
29
10
±.
±.
±_
±_
±_
±_
+.
±_
±_
1.4
25.3
5.72
0.98
15.4
0.75
0.41
1.1
0.41
Standard
Bailer
53
1060
195
18
306
31
9
22
11
±.
^^^
±.
"*"
±_
±_
±.
±_
+.
1.2
25.1
7.12
1-7
11.8
0.82
0.41
0.52
0.84
Bladder
Pump
65 ±_
1250 +_
210 ±_
19 +.
338 ±_
33 ±_
10 +.
25 ±_
11 +.
1.1
23.5
6.25
0.82
6.81
1.7
0.75
0.52
0.85
Multi-
Port
50
1240
203
18
335
34
9
28
11
+
+_
+_
±_
+.
+.
+.
+
+_
1.0
20.4
7.31
0.52
12.8
1.0
0.75
1.2
0.00
                                                39

-------
sampling, is was difficult to relate the survey sampling results to the benzene concentration variation observed
during experiment B.

      With the exception of benzene and trichloroethene recoveries, the bailer exhibited the lowest mean
concentrations  (only the  mean recovered concentrations  of  trichloroethene,  1,3-dichlorobenzene,  and
1,1-dkhloroethane were within one standard deviation of those recovered by the bladder pump). The standard
deviations of the means differed little between individual sampling devices. This suggests that the precision of the
PTFE filter tip and multi-port is similar to that of the bladder pump. It is also interesting to note that the bailer
samples produced a standard deviation very similar to that of the other devices for each organic compound.

      The VOC data were statistically analyzed by a multivariate analysis of variance, followed by one-way
analyses of variance and multiple comparison tests. The multivariate analysis, run on software produced by
BMDP Statistical Software, Inc. (1988), revealed significant differences between sampling devices when all nine
detected VOCs were included.

       The one-way analyses of variance were designed to determine if significant differences existed between
the sampling devices based on recoveries of individual VOCs, With the exception of trichloroethene, this analysis
of variance showed that the mean values of the replicate samples for each device were significantly different from
each other. This suggests that, at the 5 percent significance level, the devices differed significantly in their ability
to recover VOCs. For the recovery of trichloroethene, the devices did not differ significantly.

      The liikey multiple comparison test was utilized to test for significant differences between individual pairs
of sample means. The results, shown in Table 5, confirm much of what may be interpreted from an inspection of
the  data shown in Table 4.  In all  cases (except benzene recovery), the in situ devices recovered VOCs at
concentrations at least as great as, or not significantly different from, those of the bladder pump. The PTFE filter
tip was part of the groups recovering the significantly highest concentrations for eight of the nine compounds,
while the bladder pump was included in this group for six compounds. The multi-port was part of the highest
group for three compounds, while the bailer was part of the highest group for one compound.

       The bailer yielded the lowest mean concentrations for seven of the nine detectable VOCs, and for three of
these compounds  (chlorobenzene, ethylbenzene, and 1,4-ctichlorobenzene), the means  differed significantly
from those  of all other sampling devices. For six of the nine VOCs, the bailer produced  mean concentrations
which were significantly lower than those of the bladder pump.

 DISCUSSION

       The objectives of this phase of the study were to compare the accuracy and precision of each of the
 sampling devices. Again, because of the nature of a field study, the assessment of sampling device accuracy must
 be based on the relative concentrations of VOCs recovered (higher mean concentrations imply greater accuracy).
 Precision is estimated from a comparison of the standard deviations about the mean values.

       Based on inspection of the data and the results of the statistical analyses, it may be stated that Experiment
 B has shown that the PTFE filter tip produces samples with an accuracy as great  as or greater than those collected
 with the bladder pump, multi-port, or the bailer. The PTFE filter tip and bladder pump results were fairly
 comparable, while the multi-port  and bailer appeared to be less accurate. The bailer appears to be the least
 accurate of the four devices tested. Considering the likely spatial variation in  hydrochemistry at the site, these
 results also suggest that under field conditions the bladder pump, PTFE filter tip, and multi-port may produce
 comparable results. Because the bailer sampled the same well as the bladder pump, spatial variation should not
 have been a factor in the bailer's response, further indicating its lower accuracy. Based on a comparison of
 standard deviations, all four of the devices exhibited comparable precisions.

        The PTFE filter tip proved to be capable of providing  ground-water samples  with an accuracy and
 precision rivaling that of the bladder pump. It is unfortunate that the HDPE filter tip installation was damaged


                                                 40

-------
                  TABLES.   TUKEY TEST RESULTS FOR EXPERIMENT B
Rank* Benzene
1 Bladder Pump
2 Standard Bailer
3 Multi-Port
4 PTFE Filter Tip
Rank* 1,3-Dichlorobenzene
1 1 PTFE Filter Tip
2 I Bladder Pump
3 Multi-Port
4 L Standard Bailer
Rank* 1,2-Dichloroethane
1 [ PTFE Filter Tip
2 [ L Bladder Pump
3 [ f Multi-Port
4 L Standard Bailer
Chlorobenzene
PTFE Filter Tip
[Bladder Pump
Multi-Port
Standard Bailer
1,4-Dichlorobenzene
PTFE Filter Tip
[Bladder Pump
Multi-Port
Standard Bailer
Ethylbenzene
IPTFE Filter Tip
Multi-Port
Bladder Pump
Standard Bailer
1,2-Dichlorobenzene
IPTFE Filter Tip
Bladder Pump
[Multi-Port
Standard Bailer
1, 1-Dichloroethane
Multi-Port
PTFE Filter Tip
Bladder Pump
. Standard Bailer
Trichloroethane
Multi-Port
Bladder Pump
Standard Bailer
. PTFE Filter Tip
    'Ranked from highest to lowest concentration
and could not be included in this phase of the study. A continued evaluation of the differences between the two
materials may have provided valuable information that might explain some of the discrepancies between the two
types of filter tips noted during Experiment A.

      Experiment B also confirmed the need to replace the perforated sample vial septum on those samples
collected with the multi-port device. The six replicate samples showed no sign of the apparent high losses of
volatiles noted during the Experiment A sampling phase. Precision of the samples produced by the multi-port
was comparable to those of the bladder pump and PTFE filter tip, although accuracy appeared to be somewhat
less.

      The bailer again produced relatively low concentrations of most of the VOCs and a precision comparable
to the other devices. However, these results are somewhat misleading: samples collected with the bailer were
transferred from the bailer to a glass beaker, from which they were then transferred to individual sample vials. As
a result, the six samples collected with the bailer were not obtained from six successively collected bailer volumes.
Instead, these samples are actually splits from two successive bailer volumes (the beaker was filled and the
contents used to fill three sample vials), and, therefore, are not the kind of results that might have been obtained
with individual bailer samplings. This would explain, in part, why the precision of the bailed samples appears
                                                41

-------
higher than might be expected based on the results of other studies. Overall accuracy and precision may have
been different if the replicates had been collected from six successive bailer volumes. In light of the procedures
used to collect the bailer samples, these results are of little use in evaluating the performance of the bailer.

      As was found during Experiment A, the VOC standard solution concentrations were much lower than the
stated true values (Table C.18). Again, the deviations can probably be attributed to volatilization during sample
preparation rather than sample degradation. The two standards are duplicates produced from the same batch
sample, and both analysis results are in close agreement.

      The source of detectable levels of acetone in the standards, trip blank, bailer equipment blank, and two
bladder pump field samples was not determined during this experiment. However, the bailer decontamination
process can probably be eliminated as the source in the bailer eqmproent blank because acetone was not used for
equipment decontamination during this experiment. It is also unlikely that the source of acetone was the ground
water at the site because acetone has not been detected during any other sampling at this site and because acetone
was only detected in two out of the 24 field samples collected during this sample round. The presence of acetone
in several types of samples (field samples, trip blanks, equipment blanks, and standards) suggests that the
contamination occurred at a time when many of the samples could be affected simultaneously; for example,
during the VOA bottle-cleaning process. In an effort to eliminate the problem, the remaining bottles in that
batch were discarded and new, pre-cleaned VOA bottles were utilized during subsequent experiments. In any
case, the acetone contamination did not appear to affect the results of the experiment.
                                                42

-------
                                           SECTION 6

                                        EXPERIMENT C
EXPERIMENTAL DESIGN

     While experiment B resulted in a valid comparison of the four devices for a single sampling event, that
experiment did not include all the devices of interest, nor did it provide information about long-term VOC
concentration trends and device performance. The final phase of the comparative study, experiment C, was
designed to monitor long-term VOC recovery patterns of the included sampling devices. Collecting more
samples over several sample rounds would allow comparisons to be made over a longer time period so that
patterns between devices could be better defined. This information would be useful in the determinations of
individual device accuracy  and precision as well as understanding how the devices as  a group responded to
changing conditions. In addition, the ability of the devices to collect samples over long time periods without
mechanical failure was of additional interest to the study. It was hoped that this information would provide an
indication of device reliability.

     The six devices utilized during experiment C included the four from experiment B plus the HDPE filter
tip and the BED bailer (bailer with a bottom emptying device) as shown in Table 6. The HDPE filter tip
installation, which had been damaged by construction activity prior to experiment B, was repaired and returned
to use for this experiment. As the filter tip itself was not affected by the damage, results of subsequent sampling
were not expected to be adversely impacted. The  BED bailer, which was described in an earlier section of this
report, had not been used in any of the previous experiments. It was tested here to determine if its design would
reduce the variability observed with the standard  bailer in experiments A and B.

                 TABLE  6.   SAMPLING DEVICES UTILIZED IN EXPERIMENT C

                           Device Type	          Well Number	

                         BED Bailer                                  1
                         Bladder Pump                                1
                         HDPE Filter Tip                             5
                         Multi-Port                                   3
                         PTFE Filter Tip                              2
                         Standard Bailer                              1
      The experiment involved four sampling rounds conducted at intervals of approximately 12 weeks for the
 duration of one year. Three sample replicates were collected with each device, as opposed to the six replicates
 collected during experiment B. Samples were collected for complete VOC analysis (EPA Method 624), trace
 metals Fe and Mg, and gross chemistry. As in the previous experiments and with the exception of well 1, the order


                                                 43

-------
in which each device was sampled during each round was randomized to prevent bias. The order in which well
1 was sampled was chosen randomly from the group of four wells but the device sampling order within well 1
was generally not random. During the previous experiments and during the early rounds of this experiment the
bladder pump samples were collected fust, directly after the purging procedures were complete. This meant that
the pump did not have to be removed from the well, stored during sampling with the other devices, and then
decontaminated prior to reinstallation in the well for sampling. As this procedure might have introduced bias
into the experiment, the order with which the well 1 devices were used was randomized during sample rounds
3 and 4.

      Quality assurance samples (trip blanks and equipment blanks) were collected during each round. Although
standards were not prepared during this experiment because of the difficulties encountered in their preparation
(as previously discussed), laboratory replicate analyses were performed on approximately 20 percent of the
samples submitted

      This  experiment  utilized the same assumptions  as the previous experiments,  namely that VOC
concentrations were uniformly distributed across the study site and that the highest VOC concentrations
recovered provided the most accurate representation of aquifer VOC concentrations.

RESULTS

      The VOC analytical data resulting from experiment C are tabulated in Tables  C.19 through C.22 and
graphically presented in Figures D.10 through D.18. The  ame nine VOCs that were detected in experiment B
were also detected in  each of the  samples in this experiment. In contrast to that experiment, however,
considerable variation is apparent in these data, both between devices and over time for each device. The
compounds continued the generally declining VOC concentration trends observed in the earlier experiments.
In fact, concentrations of several of the compounds fell well below 50 jig/L, making their usefulness for further
analysis limited.

      Appendix C presents the results of the analyses of the quality assurance blanks and laboratory analytical
precision evaluations. Contamination was not detected in any of the quality assurance blanks, so the change in
bottles had a positive impact on bottle cleanliness. Furthermore, the duplicate analyses of random field samples
show that all of the VOC concentration determinations were within a ±.10 percent concentration range. In fact,
many analyses fell within a ±.5 percent range. These results demonstrate that analytical precision was well within
laboratory-stated bounds and that analytical variability may be of lower magnitude than other potential sources
of variability in the study.

      Although there was considerable variation in device ranking (based on VOC concentration recovered)
during each sample round, the bladder pump and filter tips generally recovered the highest concentrations of
the nine compounds. The only exception to this pattern was sample round 1 where the standard bailer also
recovered the highest concentrations of some compounds. However, the bailers, particularly the BED bailer,
recovered the lowest concentrations of most compounds during sample rounds 2 through 4, while the multi-port
and standard bailer each recovered the lowest concentrations for several compounds during sample round 1.
Most devices recovered the lowest concentrations of at least one compound at some time during the experiment,
though the  bailers usually recovered  the lowest  concentrations. The  lack of dear device-dependent
concentration trends suggests that site conditions may have been changing non-uniformly during the study and
that a certain amount of variation might be expected from any sampling device under these conditions.

      Because three sample replicates were collected with each device in experiment C, unlike the six collected
in experiment B, the estimates of standard deviation made in this experiment are likely to be less representative
of the true standard deviations. In addition, loss of one of the three replicates for two devices in round 2 resulted
in only two replicates for those devices. As a result, for individual sample rounds in experiment C, the variation
 exhibited may be more strongly influenced by  random fluctuations than device-dependent  fluctuations.


                                                 44

-------
          TABLE 7.    TEST OF EFFECTIVENESS OF BOTTOM-EMPTYING DEVICE,
                       SAMPLE ROUND 4
Mean Concentrations and Standard Deviations of Three Replicates
Compound
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
1, 1-Dichloroethane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
Bottom-Emptying Device
11.0
684
125
13.0
172
26.3
7.4
19.4
8.3
JL
±.
JL
±.
±.
±.
±.
+.
±.
1.9
134
17.6
0.7
47.0
1.1
0.4
3.0
03
Decanted from Bailer Top
9.0
523
98.4
11.9
112
26.1
7.6
15.8
7.9
±.
±.
±.
±.
±.
JL
±.
±.
±.
0.0
5.6
2.3
0.1
2.5
1.4
1.1
0.1
0.3
However, variations that exhibit consistent patterns over the four sample rounds are more likely to result from
device-dependent fluctuations than chance fluctuations.

     The standard deviations of the replicate sample means for each device generally showed less variation
between sample rounds than the sample means themselves. Overall, the multi-port exhibited the lowest
standard deviations for most compounds and sample rounds. As a group, the HOPE filter tip, bladder pump,
and multi-port consistently had the lowest standard deviations. The standard deviations of the PTFE filter tip
were variable over the four rounds, while the bailers generally had the highest standard deviations. The BED
bailer produced the highest standard deviation for almost every compound and sample round.

     Because the replicate samples collected with the bailers during this experiment were from separate bailer
volumes (unlike experiment B), these high deviations suggest a true lack of repeatability or precision for sampling
with bailers. In fact, the BED bailer standard deviations were unexpectedly high, especially during the sample
rounds 3 and 4, considering the use of the  bottom-empty device which was expected to provide  more
representative samples. In addition, these high standard deviations usually correlated with mean concentrations
that were lower than the other devices. The effectiveness of the bottom emptying device was tested during
sample round 4 by decanting three replicate samples from the top of the bailer (samples could not be obtained
by the method used with the standard bailer because the bottom ball valve could not be accessed to release the
sample). The results show (Table 7) that decanting the sample generally led to lower mean concentrations and
lower  standard deviations about those means than using the bottom emptying device. It  appears that, in
comparison to decanting from the top of the bailer, the bottom emptying device may lead to more variation in
the concentrations of recovered samples  although the overall resultant concentrations are higher.

      The VOC data were analyzed using a multivariate analysis of variance (BMDP Statistical Software, 1988)
to determine if there existed statistically significant differences between  device sample means and sample
rounds. The analysis was conducted for all nine detected VOCs with the six sampling devices and four sampling
rounds as the classification variables. The hypothesis tested was that the observed VOC concentrations were
independent of sampling device, with the effect of time removed. The multivariate analysis indicated that there
were significant differences between device sample means and sample rounds at the 5 percent confidence level.

      Univariate analyses of variance were then conducted to determine if significant differences existed
between device sample means based on recovery of each individual compound. Nine one-way analyses were
 conducted for each sample round, one for each compound, thereby removing the effect of interactions between


                                                45

-------
devices over time that was observed in the VOC data plots. These analyses revealed that for most compounds
significant differences existed between sample means at the 5 percent level, although there were a few
exceptions. Trichloroethene and 1,2-dichloroethane sample means of the six devices did not differ during sample
rounds 1, 2, and 3;  1,4-dichlorobenzene  sample means did not  differ during sample  round 1; and
1,1-dichloroethane  sample  means did not  differ during  sample  round  3.  With  the  exception  of
1,4-dJchlorobenzene, the lack of significant differences between the devices for these compounds is due to their
recovered concentrations near the detection levels for the analytical methods used (1 jig/L). Because distinctions
could not be made below 1 fig/L, resolution at these low concentrations was reduced by the loss of a significant
figure in the reported values and led to grouping of the concentrations at lower levels of resolution (higher
concentration). This resulted in the appearance of more similar means than actually existed in the data set.

      Tukey multiple comparison tests were then conducted to determine which individual pairs of sample means
during each sample round were significantly different at the 5 percent level. The results of these analyses are
shown in Tables  D.I through D.4. The devices are divided into groups of devices with sample means that are
similar. Devices in each individual group have sample means that are significantly different from those in other,
separate groups, although devices can be part of more than one group if their sample means are similar to the
sample means in both groups. There is considerable overlap between groups but generally the devices with the
highest sample means are significantly different from those with the lowest.

      The results of all the Tukey tests are summarized in Tables 8 and 9, which show the number of times each
device had the highest or lowest sample means of each compound or the sample mean for that device was not
significantly different from the highest or lowest sample mean for that compound. Individual Tukey test results
show that during sample round 1, the standard bailer and BED bailer most often recovered concentrations that
were significantly higher than the other devices. The bladder pump and in situ devices recovered slightly lower
concentrations than the bailers, with the multi-port in the lowest grouping most often. During sample round
2, the filter tips and bladder pump were most often in the top grouping while the bailers were most often in the
lowest groupings. It is interesting to note, however, that although the filter tips were most often in the top
grouping, they also occasionally appear in the lowest group. The bladder pump, which was also often in the top
grouping, in round 2 never appeared in the lowest group. The bladder pump and all three in situ devices were
all in the top grouping during sample round 3, with all six devices occasionally appearing in the lowest grouping,
although the bladder pump and filter tips were in the lowest group least often. During sample round 4, the
bladder pump and filter tips were again part of the highest grouping most of ten. As in sample round 2, the bladder
 pump was never in the lowest group of round 4, although the filter tips occasionally were. The BED bailer was
 in the lowest group for  eight of the nine compounds and was never in the top group for round 4.

       Overall, the bladder pump and PTFE filter tip were both most often in the top group (recovering the
 highest concentrations), followed closely by the HOPE filter tip. The multi-port, standard bailer, and BED bailer
 then followed in order of frequency of their appearance in the top group. The bladder pump was least often in
 the lowest group, followed by the PTFE filter tip, HOPE filter tip, standard bailer, multi-port, and BED bailer.
 The bailers were least often in the top group and, with the multi-port, most often in the lowest group.

 DISCUSSION

       The results of experiment C provide an estimation of device accuracy and precision based on the ranking
 of individual device sample means and standard deviations. Much like experiment B, the most effective devices
 were the bladder pump and in situ devices while the least effective were the bailers.

       Based on the univariate analysis and Tukey tests, the bladder pump appears to be the most accurate of the
 six devices tested. It most often collected the highest VOC concentrations or concentrations that were not
 significantly different from the highest recovered concentrations. The bladder pump also recovered the lowest
 concentrations of the nine compounds considerably fewer times than the other devices which suggests that there


                                                 46

-------
     TABLE 8.    RANKINGS OF DEVICES BASED ON TUKEY TEST RESULTS - HIGHEST
                 VOC RECOVERY

                                        Number of times in highest concentration group*
                                              Sample Round
          Device
Overall
Bladder Pump
PTFE Filter Tip
HDPE Filter Tip
Multi-Port
Bailer
BED Bailer
5
5
5
5
7
6
7
6
7
4
2
2
8
8
7
8
4
3
6
7
6
3
3
0
26
26
25
20
16
11
 •Number of times the device recovered the highest mean concentration of a compound, or the
 concentration was not significantly different from the highest concentration recovered by any device.
     TABLE 9.   RANKINGS OF DEVICES BASED ON TUKEY TEST RESULTS - LOWEST
                 VOC RECOVERY

                                         Number of times in lowest concentration group*
                                              Sample Round
          Device
Overall
BED Bailer
Multi-Port
Bailer
HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
2
4
1
2
3
3
5
3
4
3
2
0
5
4
4
2
2
2
8
5
5
5
1
0
20
16
14
12
8
5
  'Number of times the device recovered the lowest mean concentration of a compound, or, the
  concentration was not significantly different from the lowest concentration recovered by any device.
  Results only include those compounds where there were significant differences between devices.
might be less variability inherent to the bladder pump sampling procedures between sample rounds when
compared to the other devices. The precision of the bladder pump was on the same order as the HDPE filter
tip and multi-port, which, as a group, produced the lowest standard deviations about their replicate sample
means. It is not surprising that the bladder pump performed well during these experiments as it has been proven
during several previous investigations to be an accurate and precise device for the collection of VOCs.

     The PTFE filter tip collected significantly higher VOC concentrations than the other devices as often as
the bladder pump, but, on the other hand, collected samples that had significantly lower concentrations than
the other devices more often than the bladder pump. This difference was slight, however, making the PTFE filter
tip just as accurate as the bladder pump under field conditions. The PTFE filter tip replicate sample standard
                                              47

-------
deviations were more variable and usually higher over the four rounds than those of the bladder pump, HDPE
filter tip, and multi-port suggesting that the PTFE filter tip is less precise than these devices.

     The HDPE filter tip was roughly as accurate as the bladder pump and PTFE filter tip, based on its overall
high VOC sample means, and was as precise as the bladder pump and multi-port. It is interesting to note that
although overall they recovered about the same mean concentrations, the HDPE filter tip was considerably more
precise than the FIVE filter tip. This disparity is difficult to explain because their design is similar and their
operating procedures are identical. As discussed in experiment A, the source of imprecision may lie within the
characteristics of the individual filter tips which are composed of different materials and have different pore sizes
and apparent permeabilities. The FIFE filter has an average pore size twice that of the HDPE (35 >im versus
15 jun) but the permeability is an order of magnitude lower (1.7xl(H cm/s versus l.TxlO"3 cm/s> The lower
permeability of the PTFE filter may result in slightly more turbulence and possible volatilization of VOCs as
ground water passes through the filter material and could lead to more variation in the recovered concentrations.
The production of the PTFE filter tip has been discontinued by the manufacturer, although for different reasons.

      The multi-port recovered somewhat lower mean concentrations than the filter tips and bladder pump and
so might be considered slightly less accurate. It did, however,  generally result in lower standard deviations than
most devices during most sample rounds. A possible explanation for this combination of lower apparent accuracy
and high precision is that the device f unctions well but that ambient VOC concentrations may be somewhat lower
at the multi-prat installation than at the  others. Another possible  explanation might  be loss  of volatiles
associated with the sampling device design, such as potential orifice effects around the measurement port and
sampling probe valves, excessive vacuum applied to the sample vial holder, escape of volatiles through the vent
needle and into the interior of the VOA bottle holder, or well installation and development procedures. The
replacement of punctured sample vial septa with unpunctured septa did apparently lead to consistently higher
concentrations than had been observed during experiment A.

      The standard bailer and BED bailer results confirmed some of the conclusions reached during other
investigations regarding the relative inconsistencies associated with sampling VOCs with bailers. Of the six
devices tested in this study, the bailers exhibited the lowest overall accuracy and precision, with the BED bailer's
performance the lower of the two. The inaccuracy of a standard bailer is generally attributed  to problems
associated with how sampling personnel handle the bailer and the resultant sample. In this study, there is little
doubt that the practice of discharging the sample from the standard bailer through the check valve and into a
beaker contributed to the loss of VOCs from the sample. Pouring the sample from the bailer vent hole without
 causing agitation was also found to be difficult and was not an effective way of transferring the sample from the
bailer to a sample bottle. It was thought  that this problem would be reduced by using the BED bailer because
 it was designed to gently release the sample through the bottom check valve by means of the bottom emptying
 device. In practice, however, the bottom emptying device seemed to increase the accuracy of the device slightly
 but reduced the precision when compared to decanting the sample from the top of the bailer (again, not a very
 effective way of retrieving the sample, but the only alternative available considering the design of the BED
 bailer). The problem may be related to the formation of air bubbles at the bottom of the bailer when the bottom
 emptying device  is installed into the bottom check valve  assembly. These bubbles, which appeared to be
 introduced into the bailer body from  the bottom emptying device, traveled up through  the bailer body (and
 sample)to the top of thedevice. It is possible that as these bubbles traveled through the sample VOCs in solution,
 they volatilized into the available head space and were carried out of the sample as the bubbles rose to the surface.

       As in experiment B, a portion of the variation observed between devices in experiment C may be attributed
 to spatial variation in hydrochemkal conditions. If the assumption of identical conditions between devices is
 made, the above described relationships apply. However, if spatial variation in these parameters exists at the site,
 it may be an important part of the total variation observed In this case, as in experiment B, the overall variation
 observed in the bladder pump and in situ device results did not greatly exceed the spatial variation to be expected
 under these field conditions.


                                                 48

-------
     With the exception of the well probe, few mechanical problems were experienced with the sampling devices
during the more than two year extent of this experiment. The bailers caused no problems at all, due to the
simplicity of their construction and operation. Handling, decontamination, and storage were easily accomplished
under all conditions. The bladder pump experienced only one minor problem: the nitrogen supply tube fitting
on the pump loosened during purging in sample round 4, allowing nitrogen to enter the well casing and causing
considerable agitation of the well water. This problem was easily repaired in the  field and, after additional
purging, did not appear to affect sampling results for that round. Otherwise, the bladder pump performed
smoothly.

      Although none of the problems were major, the in situ devices tended to have more difficulties associated
with them due to their specialized and more complicated design. Both filter tips  occasionally had problems
delivering samples when foreign materials inadvertently entered their well bores and covered the filter tip septa.
This problem was rectified by rinsing the interior of the filter tip casing with distilled water and pumping out the
materials with a peristaltic pump. This treatment was required only once during the experiment. In addition,
although not a mechanical problem, decontamination of the specially designed sample bottles in the laboratory
was a time-consuming procedure whose consistency and effectiveness were difficult to  evaluate.

      The multi-port sampling device performed well throughout the experiment with  only two problems. An
air leak in the pneumatic control box required an interruption in sampling and repair in the shop. Although not
generally recommended  by the manufacturer, the multi-port sampling probe, and all stainless-steel sample
bottles,  were  disassembled  after each  sample  round and   thoroughly cleaned.  This  procedure was
time-consuming and considerable care was required when reassembling all the components. A second problem
involved the inability to collect a sample from the measurement port at the 4.5-m depth. Although samples from
this port were not designed to be included in the comparison study, several attempts were made to make the port
operational. All methods attempted were unsuccessful, and the problem was attributed to a small leak in the
 sampling probe-measurement port connection.
                                                  49

-------
                                            SECTION?

                                 SUMMARY AND CONCLUSIONS


     The collection of representative ground-water samples is commonly a very difficult and often hazardous
undertaking. Many of the problems relate to the process of purging a well of stagnant water prior to sample
collection, a necessary procedure when using most commercially available sampling devices in conventional
monitoring wells. During this investigation, field experiments were designed to evaluate the effectiveness of four
relatively new sampling systems (two filter tips, a well probe, and a multi-port system) which require little or
no pre-sample purging. The evaluation involved a comparison of these newer devices with two relatively common
sampling devices, a bladder pump and a bailer.

     The devices were field-tested at a site where a shallow alluvial aquifer was contaminated by a VOC plume.
Six wells were installed on 6-m centers in a rectangular grid pattern, with the sampling zones of each device
placed at a depth of approximately 6 m. Heterogeneous geologic and hydrochemical conditions represented
possible sources of variability in the results, although they were assumed to be uniform over the site in order
to make comparisons between the samplers. The comparisons were based on the ability of each device to
accurately and  precisely recover samples of ground water contaminated with VOCs. Those devices that
recovered the highest concentrations were considered the most accurate.

     The results of Experiment A suggest both of the filter tip devices and the well probe recovered benzene
and chlorobenzene with an accuracy greater than that of the bailer, and at levels rivaling those obtained with
the bladder pump. Before determining the source of the anomalously low concentrations, the multi-port
produced VOC  samples which were much less accurate than those collected with the bladder pumps, filter tip,
and well probe devices. Additional experiments with the multi-port have shown that replacement of the
perforated VGA bottle septum with a new septum can prevent sample degradation and allow this system to
extract accurate samples which can be  preserved until the  time of analysis. The design of this system is also
flexible enough  to allow for other methods of sample collection which do not involve a punctured septum, but
still have the potential to recover accurate samples. This system and the various methods of sample collection
could be evaluated farther.

     Experiment B, based on many of the results and problems noted during the first experiment, confirmed
many of the original findings. Multiple replicate samples were collected with four of the original sampling devices
during a single sampling interval. The statistical analysis indicated that for eight of the nine detected VOCs,
concentrations recovered by the PTFE filter tip were significantly higher than, or not significantly different from,
concentrations recovered by the bladder pump.  Although the multi-port was slightly less accurate than the
bladder pump and filter tip, precision was comparable for all three. Samples collected with the bailer appeared
to be less accurate than those collected with all other devices. However, the bailer sampling procedures used
did not allow for the collection of true replicate samples, and, therefore, complete assessments of the accuracy
and precision of the bailer could not be made. Experiment B also confirmed the ability of the multi-port to collect
more accurate samples when the perforated sample vial septum is replaced.
                                                50

-------
     The results of experiment C, which included four sample rounds, six devices, and multiple replicate
samples from each device, followed several of the trends established during the previous two experiments, but
provided more information. This experiment suggested that the bladder pump and the filter tips were the most
accurate of the devices tested and that the bailers were the least accurate. The multi-port was found to be
somewhat less accurate than the bladder pumps and filter tips. The multi-port, bladder pump, and HDPE filter
tip provided the most precise samples while the bailers, the BED bailer in particular, provided the least precise
samples.

     The variability observed in the survey sampling, although less than the stated analytical error, indicates
that some spatial variability may exist at the site and that the assumption of identical conditions between devices
may not have  been strictly valid. Because of the nature of this study,  a complete understanding of the
hydrochemical variation between the well installations could not be determined independently of the tested
devices. As a result, the possibility of spatial variability contributing to the overall observed variability cannot
be discounted. If spatial variability typical of many field sites is included, these experiments indicate that the
variability associated with the bladder pump and in situ samplers is of similar magnitude and that there may be
little difference in the accuracy and precision of these devices. The bailers, on the other hand, which sampled
the same well as the bladder pump, were not subjected to the uncertainty of varying spatial conditions, so the
samplers were the primary source of variability.

      The results show that, under the field conditions  of this study, the tested in situ devices provide samples
with essentially the same precision and accuracy as bladder pumps and greater precision and accuracy than
bailers. It appears that the designs of these in situ devices significantly reduce the volume of stagnant water
normally associated with conventional monitoring wells. As a result, the in situ devices can provide consistently
representative VOC samples in sand and gravel aquifers, while virtually eliminating the need for well purging
prior to sample collection.

      In addition to collection of representative samples and minimizing purging volumes, the in situ devices also
 allow samples to be collected quickly, while reducing both exposure of sampling personnel to potentially
 hazardous materials and the volume of purge  water to be  disposed. Furthermore, these sampling systems are
 relatively easy to operate and maintain, the standardized sampling methodology reduces variability potentially
 introduced by sample handling, and  fewer sampling personnel are required to  obtain samples. The few
 operational difficulties experienced with the two in situ devices are described fully in this report.

      To obtain a more accurate evaluation of these sampling devices, it is suggested that additional studies be
 developed and implemented, such as application of these systems in various hydrogeochemical conditions,
 including low-permeability environments, or areas with different types of contaminants. In addition, the effects
 of these devices on the concentrations of naturally-occurring ground-water chemical constituents should be
 studied. Other installation methods and devices at varied depths also should be investigated. Further studies,
 performed at a variety of sites and involving these and other commercially available in situ sampling devices, are
 needed to improve understanding of the applicability of these devices to a variety of monitoring situations.
                                                   51

-------
                                         REFERENCES

Barcelona, MJ., J.A. Helfrich, E.E. Garske, and J.P. Gibb, 1984. A laboratory evaluation of ground water
    sampling mechanisms: Ground Water Monitoring Review. Vol. 4. No. 2. pp. 32-41.

BAT Envitech, Inc., 1987, BAT groundwater monitoring system, special equipment notice for Desert Research
    Institute (DRI> 10 p.

Black, W.H., H.R. Smith, and ED. Patton, 1986. Multiple-level ground water monitoring with the MP system. In:
    Proceedings of the Surface and Borehole Geophysical Methods and Ground Water Instrumentation Conference and
    Exposition. Denver, Colorado. National Water Well Association. Water Well Journal Publishing Company,
    Dublin, Ohio. pp. 41-61.

Blegen, R.P., K.F. Pohlmann, and J.W. Hess, 1987. Bibliography of ground-water sampling methods. U.S.
    Environmental Protection Agency. EPA-600/X-87/325. 43 p.

BMDP Statistical Software, Inc-, 1988. BMDP Statistical Software Manual. Volume 2, Method PV4. ed. by WJ.
    Dixon, a al.. University of California Press.

Desert Research Institute, 1987. Quality assurance plan for subsurface geophysical methods for hazardous waste
    site investigations and monitoring. Prepared for U.S. Environmental Protection Agency, Environmental
    Monitoring Systems Laboratory, Las'Vegas, Nevada. Cooperative Agreement CR812713-01 (unpublished).

Ecology and Environment, Inc., 1984. Technical summary of soil and in situ gas sampling study. Bask
    Management, Incx, Henderson, Nevada. Report Number C(84)C345.28 p.

Ecology and Environment, Inc., 1985. Phase HB sampling plan. Stauffer Chemical  Company, BMI Complex,
    Henderson, Nevada. 52 p.

Fordham, J.WL,  H. Kolterman, and J.W. Hess, 1984. Groundwater monitoring plan, part 1 - review of site
    hydrogeology and existing monitoring network. Prepared for T1MET, Inc., Henderson, Nevada. Water
    Resources Center, Desert Research Institute. 37 p.

Geraghty and Miller, Inc., 1980. Ground-water investigation. Stauffer Chemical Company, Henderson, Nevada.
    42 p.

Gillham, R.W., MJ.L. Robin, J.E Barker, and J.A. Cherry, 1983. Groundwater monitoring and sample bias.
    American Petroleum Institute, Environmental Affairs Department. API Publication No. 4367. 206 p.

Gillham, R.W., MJ.L. Robin, J.E Barker, and J. A Cherry,  1985. Field evaluation of well flushing procedures.
    American Petroleum Institute, Health and Environmental Sciences Department. API Publication No. 4405.
    110 p.
                                               52

-------
Hall, L.E., 1986. Control of ground water contamination in an alluvial fan aquifer by a dual hydraulic barrier. In:
    Proceeding? of the First National Outdoor Action Conference on Aquifer Restoration, Ground Water Monitoring
    and Geophysical Methods. Las Vegas, Nevada. National Water Well Association, Dublin, Ohio. Water Well
    Journal Publishing Company, Worthington, Ohio. pp. 125-142.

Houghton, R.L. and M.E. Berger, 1984. Effects of well-casing composition and sampling method on apparent
    quality of ground water. In: Proceedings of the Fourth National Symposium and Exposition on Aquifer Restoration
    and Ground Water Monitoring. Columbus, Ohio, National Water Well Association. Water Well Journal
    Publishing Company, Worthington, Ohio. pp.  203-213.

Imbrigiotta, T.E., J. Gibs, T.V. Fusillo, G.R.  Kish, and JJ. Hochreiter, 1986. Field evaluation of seven sampling
    devices for purgeable organic compounds in ground water. Prepared for ASTM Symposium on Field
    Methods for Ground Water Contamination Studies and Their Standardization. February 2-7,1986. Cocoa
    Beach, Florida. 23 p.

JRB Associates, Inc., 1981. Henderson Industrial Complex hazardous waste investigation. U.S. Environmental
    Protection Agency Contract No. 68-01-5052. Directive of work #23.

Kaufman,  R.F., 1978. Land and water use  effects  on ground-water  quality in  Las Vegas Valley.  U.S.
    Environmental Protection Agency. EPA-600/2-78-179. 216 p.

Longwell, C.R., E.H. Pampeyan, B. Bowyer, and RJ. Roberts, 1965. Geology and mineral resources of Clark
    County, Nevada. Nevada Bureau of Mines and Geology. Bulletin 62. 218 p.

Malmberg, G.T., 1965. Available water supply of the Las Vegas ground-water basin, Nevada. U.S. Geological
    Survey Water-Supply Paper 1780.116 p.

Maxey, G.B. and C.H. Jameson 1948. Geology and water resources of Las Vegas, Pahrump, and Indian Springs
    Valleys: Clark and Nye Counties, Nevada. State of Nevada, Office of the State Engineer. Water Resources
    Bulletin 5. 121 p.

Muska, C.F., W.P. Colven, V.D. Jones, J.T. Scogin, B.B. Looney, and V. Price, 1986. Field evaluation of ground
    water sampling devices for volatile organic compounds. In: Proceedings of the Sixth National Symposium and
    Exposition on Aquifer Restoration  and  Ground Water Monitoring. Columbus, Ohio, National Water Well
    Association. Water Well Journal Publishing Company, Worthington,  Ohio. pp. 235-246.

Nielsen, D.M. and G.L,Yeates, 1985. A comparison of sampling methods available for small-diameter ground
    water monitoring wells. Ground Water Monitoring Review. Vol. 5. No. 2. pp. 83-98.

Pearsall, K.A. and D.A.V. Eckhardt, 1987. Effects of selected sampling equipment and procedures on the
    concentrations of trichloroethylene and related compounds in ground water samples.  Ground Water
    Monitoring Review. Vol. 7. No. 2. pp.  64-73.

 Plumb, Jr., R.H. and A.M. Pitchf ord, 1985. Volatile organic scans: Implications for ground-water monitoring. In:
    Proceedings of the Petroleum Hydrocarbons and Organic Chemicals in Ground Water - Prevention, Detection, and
     Restoration - A Conference and Exposition. Houston, Texas. National Water Well Association and American
     Petroleum Institute. Water Well Journal Publishing Company, Dublin, Ohio. pp. 207-221.
                                                 53

-------
Pohlmann, K.F., W.L. Halligan, and J.W. Hess, 1990. Laboratory evaluation of purging techniques for low-yield
    wells. U.S. Environmental Protection Agency Internal Report. In preparation.

Pohlmann, K.F. and J.W. Hess,  1988. Generalized  ground-water sampling device matrix.  Ground Water
    Monitoring Review. Vol. 8. No. 4. pp. 82-84.

Scalf, MIL, J.F. McNabb, W.J. Dunlap, and R.L. Crosby,  1981. Manual of ground water quality sampling
    procedures. U.S. Environmental Protection Agency. Robert S. Kerr Environmental Research Laboratory.
    Ada, Oklahoma. 105 p.

Skougstad, M.W., JJ. Fishman, L.C. Friedman, D.E.  Erdmann, and S.S. Duncan, eds. 1979. Methods for the
    determination of inorganic substances in water and fluvial sediments. U.S. Geological Survey Techniques of
    Wrter Resources Investigations. Book 5. Chapter Al. 626 p.

Steel, R.G.D. and J.H. Torrie, 1960. Principles and Procedures of Statistics. McGraw-Hill. New York. 481 p.

Stolzenburg, TR. and D.G. Nichols,  1985. Preliminary results on chemical changes in groundwater samples due
    to sampling devices. Electric Power Research Institute, Inc.. EPRIEA-4118. Madison, Wisconsin, 84 p.

Torstensson, B., 1984. A new system for ground water monitoring. Ground Water Monitoring Review. Vol. 4. No.
    4. pp. 131-138.

Torstensson, B, and  A.M.  Petsonk, 1986.  A hermetically  isolated  sampling method for ground  water
    investigation. Prepared for ASTM Symposium on Field Methods for Ground Water Contamination Studies
    and Their Standardization. February 2-7,1986. Cocoa Beach, Florida. 23 p.

U.S. Bureau of Reclamation,  1982.  Hydrogeologic analysis of lower Las Vegas Valley. Colorado River Basin
    Salinity Control Project, Title n - Las Vfegas \%ish Unit 106 p.

U.S. Bureau of Reclamation, 1984.Verification plan report Pittman verification program. 43 p.

U.S.  Environmental Protection Agency,  1979.  Methods for chemical analysis of  water and wastes.
    EPA-#»/4-79-020.

U.S. Environmental Protection Agency, 1986. RCRA ground-water monitoring technical enforcement guidance
    document. U.S. Environmental Protection Agency Office of Solid Waste and Emergency Response.
    OSWER-9950.1.208p.

Unwin, J.P., 1984. Sampling ground water for volatile organic compounds: the effect of sampling method,
    compound volatility and concentration. In: Proceedings of the Fourth National Symposium and Exposition on
    Aquifer Restoration and Ground Water Monitoring. Columbus, Ohio, National Water Well Association. Water
    Well Journal Publishing Company, Worthington, Ohio. pp. 257-261

Whitbeck, M.R. and R. Williams, 1990. A simulated well environment for testing ground-water samplers. U.S.
    Environmental Protection Agency Internal Report. Submitted to U.S. EPA for review.

Yeskis, D., K. Chiu, S. Meyers, J. Weiss, and T Bloom, 1988. A field study of various sampling devices and then-
    effects on volatile organic contaminants. In: Proceedings of the Second National Outdoor Action Conference on
    Aquifer Restoration, Ground Water Monitoring and  Geophysical Methods. Las Vegas, Nevada, National Water
    Well Association. Water Well Journal Publishing Company, Dublin, Ohio. pp. 471-479.
                                                54

-------
                          APPENDIX A

                WELL CONSTRUCTION DIAGRAMS
S.
o
U

1"
2-
4-

5-
6-

7—



B

~|
















•* ^x»
•v. -
•x^ "•«,
^-.
:«•:
m
»!•!»!»
&»
1*1*
I***
I*!*
»*£'
:j§
»%%!
==
^^
pj^
•








•••^





a—
_



^
S
S
s&
:>::»:
:>:»
'^*t
.*•*.*
==:


^^^=
==
LITHOLOGY
5.1 cm Teflon casing Fine silts, sand.
and screen Boulders to 0.6 m.
0.3 m screened interval
(0.254 mm slots)
Silts, sand, and gravel to 0.6 cm.
Brown to dark brown.
Sands and clays.
_ . 	 t .. Gravels to 0.6 cm.
5% cement-bentonite Laf boulders at 3-1 m.
slurry surface seal
Sand, clay (20%).
Large cobbles (to 15 cm)
Brown to dark brown.

Fine silica sand
(90-grit)
Gravel pack
(1 6-grit silica sand) Sand, silt, and gravel.

Fine silica sand


0 64 cm bentonite pellets
U.DH m K Haf(j jjgp^gnjejj sands and
gravels at 7.6 m.
Figure A.1 - Well 1 construction diagram and lithologic log. Sampling devices: Bladder
           Pump, Bailer, BED Bailer.
                                 55

-------
o
u
1-
2-
3~
4~



6—


7 —














•»- -v.
•v. -
N^. •>,
V -N»
-<^.
^^^ ^
w -V
^ **s
^w ^
•*- *V
s^. -x.

# * *
• » »
* * *

•
'









ITJ
yji
*»"
+1*
»l*




">» •>
•>^ •*
»« *v
•^- -V
>^ •>
^ ••>>
^. ^,
>s. •>
s. **s
"-C"
* * *
» + * *
* * *





LITHOLOGY
5.1 cm Teflon riser pipe Fine sand, silt, and clay.
Some gravel.
Brown to dark brown.
Sand, silt, and clay.
Some gravel to 5.1 cm.
5% cement-berrtonite Brown to dark brown.
slurry surface seal
Clay, sand, and gravel.
Gravel, cobbles to 10.2 cm.
Fine silica sand (90-grit)
Gravel (to 2.5 cm), sand and clay.
PIPE finer tip
Gravel pack (16 grit Gravels' some §and- "* clay'
silica sand) Hard- cemented sand and
gravel at 6.4 m.
Figure A.2 - Well 2 construction diagram and lithologk log. Sampling
            device: PTFE Filter Tip.
                                 56

-------
Q.
L>
         »LM
       >I>
 4-
 5-
       1*1*
       •:ga
                        4.83 cm stainless
                        steel casing
                        0.9 m screened section
                        (0.254 mm slots)
5% cement-bentonite
slurry surface seal

Bentonite flour

Fine silica sand
Measurement port
Gravel pack


0.6 cm bentonite pellets


Measurement port
Gravel pack

0.6 cm bentonite pellets
                         Measurement port

                         Gravel pack (16-grit
                         silica sand)
                         Pumping port
                                                            LITHOLOGY
                         0.6 cm bentonite pellets
                                 Sands and silts.
                                 Gravels (to 0.6 cm).
                                 Brown to dark brown.
                                                          Clays, sand, and
                                                          gravel (to 0.6 cm).
Clays, sand, and
gravel (to 5.1 cm).
                                                           Clays, sand, and
                                                           gravels (to 0.6 cm).
                                                           Brown to dark brown.
                                  Sand and gravel.
                                                           Hard, cemented sand
                                                           and gravels at 7.5 m.
         Figure A.3 - Well 3 construction diagram and lithologic log. Sampling
                    device: MP System.
                                          57

-------
a
a
                                                            LITHOLOGY
 1-
 2-
 3-
 4-
 5-
  8-
  7-
  8-1

g
*^-
• -x.
^^^
* * * 1
*»*£
» *
:t
•







•5S
v.

s

*-^"
^^ "^
^W ^^^
>• "S
*x. •>
«- -s.
' "^^
* * *•?'
*t»£»
r » »
> ^ »
* * •
» » »
» » •
: » i
====
^

5.1 cm Teflon casing
andsaeen
0.3 m screened interval
(0.254 mm slots)
5% cement-bentonite
slurry surface seal


Fine silica sand
(90-grit)
Gravel pack ( 1 6-grrt
silica sand)

Fine silica sand
0.64 cm bentonrie pellets
Sand, clay, and gravel.
Brown. Possibly fill.
Sand and clay.
Gravel (to 2.5 cm).
Brown.
Gravels (to 2.5 cm) with
sand and clay.


Gravels with sand and clay.
Cobbles to 7 6 cm

Hard, cemented gravels
at 6.4 m.

            Figure A.4 - Well 4 construction diagram and lithologic log. Sampling

                       device: Well Probe.
                                           58

-------
Q.
. -s.
'SSS
i
»
+


5.1 cm Teflon riser pipe Fine sands and silts.
Some gravel (< 10%).
Brown to dark brown.
Silts, sand, and gravels to 0.6 cm.
5% cement-bentonite
slurry surface seal
Large cobbles (to 15 cm).
Clays, silts, and sand.
Gravel (to 2.5 cm, 50%).
Silts and sand.
Fine silica sand

HOPE filter tip
Gravel pack (16-grit Rounded gravels (to 7.6 cm).
silica santfl Sand and Sllt-


          Figure A.5 - Well 5 construction diagram and lithologic log. Sampling

                      device: HDPE Filter Tip.
                                           59

-------
 Q.
 2
 1-
 2-
 3-
 4-
 5-
 7-
 8-
 9-
10-
                                                           LITHOLOGY
5.1 cm Teflon casing
and screen
0.3 m screened interval
(0.254 mm slots)
                        5% cement-bentonite
                        slurry surface seal
                        Fine silica sand
                        (90-grit)
Gravel pack (16-grit
silica sand)
                        Fine silica sand
                        0.64 cm bentonite pellets
                        Collapsed backfill
Fine silts and sand.
Large cobbles and
boulders (to 0.6 m).
Light brown.
                          Dark brown silt/sand.
                          Gravel (0.6 cm to 5.1 cm).
                          Rounded gravels (80%) (to 0.6 cm).
                          Few cobbles (to 10.2 cm).
                          Clay (20%), silty/sandy.
                          Dark brown.
Rounded gravels and
cobbles (to 15.2 cm).
Sandy clay (30% clay).
Dark brown.
                          Coarse sand to cobbles (tc>12.7 cm).
                          80% sand, 20% gravels.
                          Brown.
                          Hard, cemented gravels and
                          coarse sands.
                          Rounded volcanic clasts.
                          Light gray to white (7.6 to 9.6 m).
                          Green clay, gypsiferous.
        Figure A.6 - Well 6 construction diagram and lithologic log. Sampling
                    device: Bladder Pump/Packer.
                                         60

-------
                                         APPENDIX B

                           LABORATORY ANALYSIS METHODOLOGY

     Samples for inorganic constituents and total organic carbon were analyzed by the Desert Research Institute
Water Chemistry Laboratory in Reno, Nevada. Samples were analyzed for pH, electrical conductivity (EC),
bicarbonate (HCO3), chloride (C1-), sulfate (SO2.' ), nitrate (NO?), sodium (Na+), potassium (K+),
magnesium (Mg2+), calcium (Ca2+), manganese (Mn2+), iron (Fe2+), dissolved silica (SiO2), total dissolved
solids (IDS), and total organic carbon (TOC).

     Analyses for volatile organic compounds were conducted by Alpha Analytical, Inc., in Sparks, Nevada.
Analysis requests for Experiment A samples were for benzene,  chlorobenzene, and chloroform only. For
Experiments B and C, a full EPA Method 624 analysis was conducted for all samples.

     Table B.I lists each chemical analysis, the appropriate laboratory method reference,  and laboratory
equipment used in the analysis.


REFERENCES

Skougstad, M.W., MJ. Fishman, L.C. Friedman, D.E. Erdmann, and S.S. Duncan, eds., 1979. Methods for
    determination of inorganic substances in water and fluvial sediments. U.S. Geological Survey Techniques of
    Water Resources Investigations. Book 5. Chapter Al. 626 p.

U.S.  Environmental  Protection Agency,  1979.  Methods  for  chemical  analysis  of water and  wastes.
    EPA-600/4-79-020.
                                                61

-------
   TABLE B.1 - CHEMICAL SPECIES, REFERENCED METHOD ANALYSIS,
              AND LABORATORY ANALYSIS EQUIPMENT USED
 Measurement
  Parameter
   Reference
    Method
           Equipment
pH



Electrical Conductivity



Bicarbonate (Alkalinity)



Chloride



Sulla te



Nitrate



Sodium



Potassium



Magnesium



Calcium



Manganese



 Iron


 Silica



 IDS



 TOC
 Volatile Organic
 Compounds
EPA-600/4-79-020
      (1979)
  Method 150.1

EPA-600/4-79-020
      (1979)
  Method 120.1

EPA-600/4-79-020
      (1979)
  Method 305.1

EPA-600/4-79-020
      (1979)
  Method 325.1

EPA-600/4-79-020
      (1979)
  Method 375.4

EPA-600/4-79-020
      (1979)
   Method 353.2

EPA-600/4-79-020
      (1979)
   Method 273.1

EPA-600/4-79-020
      (1979)
   Method 258.1

EPA-600/4-79-020
      (1979)
   Method 242.1

EPA-600/4-79-020
      (1979)
   Method 215.1

EPA-600/4-79-020
      (1979)
   Method 243.1

 EPA-600/4-79-020
      (1979)
   Method 236.1

  Skougstad a at.,
      (1979)

 EPA-600/4-79-020
      (1979)
   Method 160.1

 EPA-600/4-79-020
      (1979)
   Method 415.1

 EPA-600/4-82-057
      (1982)
    Method 624
                                                                   Beckman 4500
                                                                 Automated Titrator
     Beckman Model RC-19
       Conductivity Bridge
       Brinkman Metrohm
       Automated Titrator
      Model E636/series 02

        Coulter Industrial
            KemoUb
        Hack Model 2100
          lUrbidimeter
     Two Channel Technicon
          Autoanatyzer
 Instrumentation Laboratory AA/AE
      Spectrophotoineter 952
 Instrumentation Laboratory AA/AE
      Spectrophotometer 952
 Instrumentation Laboratory AA/AE
      Spectrophotometer 952
 Instrumentation Laboratory AA/AE
      Spectrophotometer 952
 Instrumentation Laboratory AA/AE
      Spectrophotometer 952
 Instrumentation Laboratory AA/AE
      Spectrophotometer 952
    Coulter Automated Analyzer
Barnstead T-100A Vfcter Bath/Mettler
        Analytical Balance
        Astro Model 2001
      Hewlett-Packard 5970
      Mass Selective Detector
       Hewlett-Packard 5890
        Gas Cbromatograph
                                          62

-------
                                   APPENDIX C

                             ANALYTICAL RESULTS OF
                        SAMPLING METHODS EXPERIMENTS

RESULTS OF SURVEY SAMPLING
     Wells 1,4, and 6 were sampled as a part of a survey sampling plan to determine expected laboratory
errors and random errors between wells. Each well was sampled with the same all-Teflon bladder pump,
which was decontaminated prior to use in each well. A19 Lbulk sample was then collected from each
well and split into five sets of duplicate samples. The resulting data and calculated means and standard
deviations are presented in Table C.I, C.2 and C-3.
                     TABLE C.1 - SURVEY SAMPLING RESULTS, WELL 1
  Well No. 1
  Sampling Date: December 18,1987
  Sampling Device: Bladder Pump
Replicate
Temp. (°C)
Field EC
Field pH
Field HCOs
Field DO
LabpH
Lab EC
TDS (mg/L)
SiO2 (mg/L)
HCOa (mg/L)
Cl (mg/L)
SCX, (mg/L)
NO3 (mg/L)
Na (mg/L)
K(mg/L)
Ca (mg/L)
Mg (mg/L)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (jig^)
Chlorobenzene (jig/L)
1
22.4
27460
run
run
0.60
7.42
22000
16700
97
414
7350
2830
<0.04
4740
68.7
546
251
0.04
1.55
9.7
303
1390
2





7.33
22000
16700
97
415
7390
2830
<0.04
4690
64.4
556
249
0.04
1.48
9.7
331
1480
3





7.37
22000
16700
97
414
7230
2830
<0.04
4780
64.9
551
250
0.03
1.50
9.8
300
1390
4





7.41
22000
16700
97
414
7380
2830
<0.04
4720
64.0
556
250
0.04
1.48
9.8
122
636
5





7.47
22100
16700
97
414
7410
2820
<0.04
4740
63.5
551
249
0.04
1.51
9.7
344
1590
X





7.40
22040
16700
97
414
7352
2828
-
4734
65.1
552
250
0.04
1.50
9.7
280
1297
s





0.05
55
-
-
1
72
5
-
33
2.1
4
1
0.01
0.03
0.1
90
379
  nm = no measurement
                                          63

-------
                    TABLE C2 - SURVEY SAMPLING RESULTS, WELL 4
Well No. 4
Sampling Date: December 17,1987
Sampling Device: Bladder Pump
Replicate
Temp. (°C)
Field EC
Field pH
Field HCO3
Field DO
LabpH
Lab EC
TDS (mg/L)
SiO2(mg/L)
HCOa (mg/L)
Cl (rag/L)
SO4 (mg/L)
NOaOng/L)
Na(mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
Fe(mg/L)
MnOngTL)
TOC(mg/L)
Benzene (jig/L)
Chlorabenzene (n-g/L)
1
21.6
28920
7.23
nm
0.40
7.23
22500
16800
97
415
7480
2830
<0.04
4800
663
567
262
0.10
1.29
9.6
350
1600
2





730
22500
16900
98
415
7460
2830
<0.04
4800
66.3
562
260
0.09
130
9.6
267
1350
3





7.21
22500
16900
97
415
7550
2820
<0.04
4690
65.4
562
258
0.09
130
9.6
266
1320
4





7.28
22500
16900
97
415
7430
2820
<0.04
4730
65.4
562
257
0.09
133
10.2
323
1560
5





7.36
22500
16900
97
415
7470
2820
<0.04
4800
66.3
556
259
0.10
132
10.5
276
1380
X





7.28
22500
16880
97
415
7478
2824
<0.04
4764
65.9
562
259
0.09
131
9.9
296
1442
s





0.06

45
1

44
6

51
0.5
4
2
0.01
0.02
0.4
38
129
nm = no measurement
                                    64

-------
                     TABLE C.3 - SURVEY SAMPLING RESULTS, WELL 6
Well No. 6
Sampling Date: December 16, 1987
Sampling Device: Bladder Pump
Replicate
Temp. (°C)
Field EC
Field pH
Field HCOs
Field DO
LabpH
Lab EC
TDS (mg/L)
SiO2 (mg/L)
HCO3 (mg/L)
Cl (mg/L)
SCX, (mg/L)
NO3 (mg/L)
Na (mg/L)
K(mg/L)
Ca (mg/L)
Mg (mg/L)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (jig/L)
Chlorobenzene 6ig/L)
1
22.0
22000
6.15
nm
0.38
7.14
22400
16700
99
410
7500
2800
<0.04
4770
64.4
556
253
<0.01
1.56
10.0
219
1370
2





7.08
22400
16700
99
410
7550
2800
<0.04
4810
63.5
546
248
<0.01
1.54
9.9
183
1100
3





7.18
22400
16700
99
410
7520
2800
<0.04
4930
64.9
546
247
<0.01
1.58
10.1
285
1560
4





7.19
22400
16700
100
409
7550
2800
<0.04
4870
64.4
551
250
<0.01
1.57
10.1
311
1670
5





7.36
22500
16900
97
415
7470
2820
<0.04
4800
66.3
556
259
0.10
1.32
10.5
276
1380
X





7.16
22400
16700
100
410
7538
2800
<0.04
4838
64.2
549
250
<0.01
1.56
10.0
251
1428
s





0.05
-
-
1
1
28
-
63
0.5
5
2
0.02
0.2
51
216
 nm = no measurement
                                       65

-------
                 TABLE C.4 - EXPERIMENT A - BLADDER PUMP SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (umhos/cm)
Field pH
Field HCO3 (mg/l)
Field DO (mg/l)
LabpH
Lab EC (umhos/cm)
TDS (mg/l)
SiO2 (mg/l)
HCO3 (mg/l)
Cl (mg/l)
S0« (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
EPM Bal. (AN/CAT)
Fe (mg/l)
Mn (mg/l)
TOC (mg/l)
Benzene (jtg/1)
Chlorobenzene Gig/1)
Chloroform (ng/l)
1
HFeb88
23.7
24681
6.94
41S
1.05
7.33
22300
16600
98
416
7480
2790
<0.04
<0.01
4680
67.3
562
249
1.087
0.02
1.41
10.3
304
1480
<2
2
17Feb88
23.9
28511
6.87
415
0.60
7.42
22200
16600
98
419
7350
2880
<0.04
<0.01
4750
67.7
575
248
1.065
0.02
1.50
9.5
310
1530
<2
3
24Feb88
22.3
29787
6.91
416
0.90
7.53
22200
16600
98
421
7280
2880
<0.04
<0.01
4800
68.0
552
249
1.053
<0.01
1.50
9.5
259
1360
<2
4
2Mar88
22.7
29326
6.90
410
0.70
7.59
22200
16600
97
423
7330
2830
<0.04
<0.01
4780
67.8
557
247
1.057
0.01
1.40
9.5
266
1400
<2
6
16Mar88
22.8
31158
6.87
420
nm
7.52
22300
16600
97
426
7310
2830
<0.04
<0.01
4860
64.7
540
246
1.045
0.02
1.40
9.2
287
1700
<2
8
29Mar88
22.5
29362
6.85
422
0.90
7.16
22100
16600
97
426
7250
2840
<0.04
<0.01
4670
64.2
550
248
1.055
<0.01
1.40
9.7
245
1310
<2
13
4May88
22.6
28731
6.88
409
1.00
7.56
23300
16500
95
422
7180
2830
<0.04
<0.01
4910
64.7
560
247
1.018
0.01
1.50
9.7
236
1590
<2
19
14Jun88
25.4
25800
6.87
nm
2.80
7.66
21700
16400
95
414
7160
2740
<0.04
<0.01
4890
69.2
526
242
1.020
0.01
1.40
9.6
34'
<2*
<2
* = samples may have been compromised prior to analysis
nm = no measurement

-------
                          TABLE C.5 - EXPERIMENT A - BAILER SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (nmhos/cm)
Field pH
Field HC03 (mg/1)
Field DO (mg/1)
LabpH
Lab EC (^mhos/cm)
TDS (mg/1)
Si02 (mg/1)
HCO3 (mg/1)
Cl (mg/1)
SCX, (mg/1)
N03 (mg/1)
NO3 (mg/1 as N)
Na (mg/1)
K (mg/1)
Ca (mg/1)
Mg (mg/1)
EPM Bal. (AN/CAT)
Fe (mg/1)
Mn (mg/1)
TOC (mg/1)
Benzene (ng/1)
Chlorobenzene ((ig/1)
Chloroform (ng/1)
1
HFeb88
24.2
25617
6.93
384
0.90
7.44
22200
16500
98
410
7370
2800
<0.04
<0.01
4740
67.7
S66
247
1.064
0.01
0.95
10.3
251
1330
<2
2
17Feb88
23.2
29128
6.86
394
0.80
7.46
22200
16600
98
419
7410
2860
<0.04
<0.01
4680
67.7
566
251
1.083
<0.01
1.50
9.5
243
1220
<2
3
24Feb88
22.3
29149
6.92
414
1.60
7.47
22200
16600
98
420
7310
2890
<0.04
<0.01
4930
68.3
547
251
1.034
0.01
1.50
9.4
<2*
3*
<2
4
2Mar88
22.6
29579
6.92
417
1.00
7.47
22200
16600
98
422
7370
2810
<0.04
<0.01
4760
67.5
552
248
1.064
<0.01
1.40
9.3
124
822
<2
6
16M:>i18
22.4
26020
6.89
415
1.30
7.49
22300
16600
97
426
7310
2890
<0.04
<0.01
4810
64.7
555
247
1.056
0.01
1.40
9.0
72
639
<2
8
29Mar88
22.3
30298
6.86
417
1.00
7.38
22100
16600
97
424
7300
2860
0.04
0.01
4739
64.2
541
246
1.068
<0.01
1.40
9.7
203
1160
<2
13
4May88
23.8
27880
6.89
409
0.80
7.63
23300
16500
96
419
7210
2870
<0.04
<0.01
4880
63.6
560
245
1.030
<0.01
1.50
10.0
172
1290
<2
19
14Jun88
24.9
26702
6.87
nm
nm
7.66
21700
16400
96
413
7190
2670
<0.04
<0.01
4810
69.2
522
243
1.031
0.01
1.30
10.5
8t
<2t
<2
* = anomalous results apparently the result of mislabeled sample bottle
t = samples may have been compromised prior to analysis
nm = no measurement

-------
                  TABLE C.6 - EXPERIMENT A - PTFE FILTER TIP SAMPLING RESULTS
Week
Sampling Date
Temp, (°C)
Field EC (^mhos/cm)
Field pH
Field HCO3 (mg/L)
Field DO (mg/L)
Lab pH
Lab EC Gimhos/cm)
TDS (mg/L)
Si02 (mg/L)
HCO3 (mg/L)
Cl (mg/L)
SO, (mg/L)
NO3 (mg/L)
NO3 (mg/L as N)
Na (mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
EPM Bal. (AN/CAT)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (ng/L)
Chlorobenzene (ng/L)
Chloroform (ng/L)
1
9Feb88
nm
nm
nm
nm
nm
No Data













No Data

No Data
No Data


2
16Feb88
21.9
30319
6.84
408
3.0
7.32
22600
16800
98
412
7480
2750
0.09
0.02
4740
68.6
571
251
1.069
0.02
1.40
10.1
234
1520
<2
3
23Feb88
23.6
30298
6.97
399
1.5
7.52
22500
16800
98
413
7460
2870
<0.04
<0.01
4830
69.4
547
250
1.066
<0.01
1.30
10.2
226
1550
<2
4
29Feb88
24.3
29116
6.89
405
1.7
7.40
22500
16800
98
413
7520
2820
0.04
0.01
4950
69.0
547
251
1.047
0.02
1.30
9.8
172
1240
<2
6
14Mar88
23.8
29104
6.89
412
2.3
7.64
22600
16800
99
413
7480
2880
<0.04
<0.01
4910
65.7
555
250
1.054
0.01
1.40
9.3
180
1440
<2
8
30Mar88
23.9
28737
6.91
405
1.3
7.51
22500
16800
98
412
7460
2770
<0.04
<0.01
4790
65.2
5-11
250
1.067
0.02
1.30
9.6
185
1490
<2
8
Dups.





7.46
22400
16800
98
411
7470
2810
<0.04
<0.01
4990
65.4
550
250
1.034
0.02
1.40
9.8
190
1490
<2
13
3May88
24.6
27312
6.88
400
1.9
7.77
24200
16600
96
410
7390
2890
<0.04
<0.01
4880
65.0
551
249
1.051
0.02
1.40
9.9
151
1510
<2
19
13Jun88
24.6
29674
6.90
411
2.6
7.72
22000
16600
95
408
7330
2690
<0.04
<0.01
4770
69.6
554
248
1.047
<0.01
1.40
9.4
136
1420
<2
Dups = duplicate samples     nm = no measurement
No Data = unable to collect sample,  or sample lost

-------
                   TABLE C.7 - EXPERIMENT A - MULTI-PORT SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (nmhos/cm)
Field pH
Field HC03 (mg/L)
Field DO (mg/L)
Lab pH
Lab EC (nmhos/cm)
TDS (mg/L)
SiO2 (mg/L)
HC03 (mg/L)
Cl (mg/L)
S04 (mg/L)
N03 (mg/L)
NO3 (mg/L as N)
Na (mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
EPM Bal. (AN/CAT)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (ng/L)
Chlorobenzene (ng/L)
Chloroform (ng/L)
1
9Feb88
25.5
31438
6.91
405
1.30
7.55
22200
16500
99
407
7320
2800
<0.04
<0.01
4720
65.4
539
248
1.067
0.01
1.38
9.7
162
858
<2
2
16Feb88
23.5
29936
6.98
409
0.90
7.48
22200
16500
99
408
7300
2790
<0.04
<0.01
4660
64.5
566
245
1.071
0.01
1.40
9.2
131
790
<2
2
Dups.





7.48
22200
16500
99
408
7310
2790
<0.04
<0.01
4660
65.4
557
246
1.073
<0.01
1.40
9.6
198
1240
<2
3
23Feb88
26.3
27979
6.90
405
0.90
7.58
22100
16500
99
409
7210
2850
<0.04
<0.01
4720
66.0
532
245
1.062
<0.01
1.40
9.1
109
560
<2
4
29Feb88
22.5
28989
6.92
395
0.70
7.72
22100
16500
99
409
7260
2810
<0.04
<0.01
4770
66.0
527
244
1.056
<0.01
1.40
9.8
91
504
2
6
14Mar88
20.5
30265
6.88
406
1.40
7.62
22200
16500
99
410
7270
2830
<0.04
<0.01
4800
62.4
531
244
1.053
<0.01
1.40
9.2
102
661
<2
8
30Mar88
23.0
28702
6.94
409
1.10
7.35
22100
16500
99
407
7260
2810
<0.04
<0.01
4860
62.2
536
240
1.040
<0.01
1.40
9.8
86
531
<2
13
2May88
22.6
28889
ti.;2
3>9
0.90
7.68
23800
16400
96
408
7200
2890
<0.04
<0.01
4850
62.3
542
238
1.041
<0.01
1.40
9.0
78
585
<2
19
13Jun88
27.1
25440
6.92
nm
2.50
7.78
21800
16300
98
405
7160
2680
<0.04
<0.01
4750
66.9
544
242
1.035
<0.01
1.30
9.7
53
435
<2
Dups = duplicate samples     nm = no measurement

-------
                          TABLE C.8 - EXPERIMENT A - WELL PROBE SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (umhos/cm)
Field pH
Field HC03 (mg/l)
Field DO (mg/l)
LabpH
Lab EC (nmhos/cm)
TDS (mg/l)
Si02 (mg/l)
HCO3 (mg/l)
Cl (rng/1)
S0< (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
EPM Bal. (AN/CAT)
Fe (mg/l)
Mn (mg/l)
TOG (mg/l)
Benzene (ng/1)
Chlorobenzene (n.g/1)
Chloroform (ng/1)
1
9Feb88
23.5
31257
6.95
356
0.5
7.22
22600
16900
94
417
7470
2880
0.04
0.01
4660
71.0
593
273
1.081
0.16
1.38
10.5
342
1400
<2
2
18Feb88
20.8
29354
6.86
411
3.2
7.18
22500
16900
93
420
7510
2920
<0,04
<0.01
4720
71.4
606
274
1.075
0.14
1,50
9.8
349
1770
<2
3
22Feb88
23.1
29271
6.91
419
2.0
7.48
22500
16900
94
420
7530
2910
<0.04
<0.01
4850
72.0
627
274
1.049
0.13
1.30
9.2
386
2060
<2
3
Dups.





7.67
22500
16900
94
419
7490
2850
<0.04
<0.01
4800
71.5
582
274
1.058
0.13
1.40
10.8
280
1610
<2
4
!Mar88
24.3
28884
7.08
nm
2.7
7.62
22500
16900
94
420
7490
2860
0.09
0.02
4850
70.9
582
274
1.050
0.12
1.30
9.4
286
1560
<2
6
17Mar88
23.0
30191
6.88
410
1.8
No Data













No Data

9.3
280
1730
<2
8
28Mar88
nm
nm
nm
nm
nm
No Data













No Data

No Data
No Data


13
3May88
23.5
27957
6.86
397
2.4
7.72
24200
16700
92
413
7320
2970
<0.04
<0.01
4910
66.7
591
268
1.031
0.11
1.40
9.7
192"
1620
<2
19
13Jun88
23.8
28630
6.88
nm
0.92
7.78
22100
16700
90
411
7310
2780
0.09
0.02
4770
71.3
589
271
1.037
0.25
1.40
9.6
152
1010
<2
Dups = duplicate samples     nm = no measurement
No Data = unable to collect sample, or sample lost

-------
                            TABLE C.9 - EXPERIMENT A - HOPE FILTER TIP SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (nmhos/cm)
Field pH
Field HCO3 (mg/l)
Field DO (mg/l)
LabpH
Lab EC (fimhos/cm)
TDS (mg/l)
SiO2 (mg/l)
HC03 (mg/l)
Cl (mg/l)
SO, (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
EPM Bal. (AN/CAT)
Fe (mg/l)
Mn (mg/l)
TOC (mg/l)
Benzene (ng/1)
Chlorobenzene (ng/l)
Chloroform (fig/I)
1
10Feb88
25.4
29230
6.90
398
2.20
7.28
22000
16300
98
404
7200
2900
2.92
0.66
4640
66.8
566
267
1.064
0.02
1.38
9.8
427
1390
<2
2
18Feb88
22.6
28375
6.95
400
2.70
7.23
21900
16500
98
408
7200
2890
0.04
0.01
4700
66.3
584
264
1.050
<0.01
1.50
9.1
458
1550
<2
3
22Feb88
24.7
28625
6.89
400
2.30
7.64
21800
16500
98
410
7190
2930
<0.04
<0.01
4610
66.5
562
262
1.074
<0.01
1.40
9.3
386
1500
<2
4
IMarSS
23.9
29053
6.95
406
1.30
7.66
21900
16500
98
409
7190
2860
<0.04
<0.01
4690
66.5
562
259
1.055
0.01
1.50
8.8
293
1150
<2
4
Dups.





7.61
21900
16500
98
409
7190
2870
<0.04
<0.01
4720
66.5
552
259
1.052
0.01
1,40
9.3
417
1770
<2
6
14Mar88
24.1
29833
6.91
420
2.10
7.54
22000
16300
97
409
7230
2880
<0.04
<0.01
4590
63.4
564
262
1.078
<0.01
1.50
9.0
322
1350
<2
8
29Mar88
24.3
28681
6.91
409
1.70
7.53
21900
16400
97
407
7150
2900
0.04
0.01
4690
62.9
564
262
1.052
<0.01
1.40
9.3
324
1360
<2
13
3May88
22.9
28258
6.88
403
1.30
7.39
23500
16300
95
407
7050
2890
<0.04
<0.01
4740
63.1
578
261
1.029
<0.01
1.50
9.1
263
1330
<2
13
Dups.





7.67
23500
16300
96
406
7100
2950
<0.04
<0.01
4770
62.3
582
260
1.034
<0.01
1.50
8.9
298
1550
<2
19
13Jun88
24.0
28043
6.90
412
3.00
7.66
21500
16300
94
402
7030
2730
<0.04
<0.01
4600
67.6
574
267
1.037
0.01
1.50
8.8
276
1420
<2
Dups = duplicate samples

-------
                              TABLE C.10 - EXPERIMENT A - BLADDER PUMP/PACKER SAMPLING RESULTS
Week
Sampling Date
Temp. (°C)
Field EC (nmhos/cm)
Field pH
Field HC03 (mg/l)
Field DO (mg/l)
Lab pH
Lab EC (^mhos/cm)
TDS (mg/l)
Si02 (mg/l)
HC03 (mg/l)
Cl (mg/l)
S04 (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
EPM Bal. (AN/CAT)
Fe (mg/l)
Mn (mg/l)
TOG (mg/l)
Benzene (fig/1)
Chlorobenzene (ng/l)
Chloroform (ng/1)
1
HFebSS
22.5
26511
7.00
415
0.60
7.58
22300
16600
99
408
7490
2820
<0.04
<0.01
4700
67.4
557
250
1.087
<0.01
1.45
10.5
221
1360
<2
1
Dups.





7.33
22300
16600
99
408
7480
2810
<0.04
<0.01
4700
67.3
566
250
1.083
<0.01
1.46
10.0
223
1420
<2
2
17Feb88
22.2
29681
6.95
394
0.70
7.41
22200
16600
99
409
7370
2810
<0.04
<0.01
4700
67.3
566
246
1.072
<0.01
1.50
9.5
218
1490
<2
3
22Feb88
23.2
27708
6.89
392
0.75
7.70
22200
16600
99
409
7370
2790
<0.04
<0.01
4770
67.5
532
247
1.065
<0.01
1.50
9.8
290
2020
<2
4
IMarSS
22.2
30000
6.87
397
0.80
7.56
22200
16600
99
410
7370
2820
<0.04
<0.01
4880
67.5
547
247
1.045
No Data
No Data
No Data
163
1160
<2
6
17Mar88
23.0
29840
6.87
409
1.00
7.42
22300
16600
98
409
7340
2820
<0.04
<0.01
4860
63.9
545
248
1.045
<0.01
1.40
9.1
170
1360
<2
6
Dups.





7.44
22300
16600
99
409
7340
2790
<0.04
<0.01
4840
63.9
531
244
1.050
<0.01
1.40
9.0
177
1390
<2
8
28Mar88
21.9
28632
6.83
407
0.80
7.38
22200
16600
98
408
7290
2780
<0.04
<0.01
4890
63.9
545
245
1.032
<0.01
1.40
9.6
166
1290
<2
13
2May88
22.3
29333
6,90
394
1.10
7.66
23900
16400
96
408
7220
2850
<0.04
<0.01
4880
64.5
564
247
1.028
<0.01
1.40
10.0
191
1800
<2
19
16Jun88
23.7
28280
6.89
nm
nm
7.70
21700
16400
98
405
7140
2750
<0.04
<0.01
4750
68.2
544
242
1.038
<0.01
1.30
9.5
148
1600
<2
19
Dups.





7.71
21700
16400
98
405
7160
2790
<0.04
<0.01
4770
67.6
549
241
1.040
<0.01
1.40
9.7
159
1710
<2
Dups = duplicate samples  nm = no measurement
No Data = unable to collect sample, or sample lost

-------
TABLE C.ll - EXPERIMENT A - LWWD STANDARD SOLUTIONS
Week
Sampling Date
Lab pH
Lab EC (fimhos/cm)
TDS (mg/l)
Si02 (mg/l)
HC03 (mg/l)
Cl (mg/l)
SO4 (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
Fe (mg/l)
Mn (mg/l)
TOC (mg/l)
Benzene (ng/l)
Chlorobenzene (ng/1)
Chloroform (jig/1)
1
HFeb88













<0.01
<0.01
3.2
<2
<2
48
2
17Feb88
8.03
823
532
9
155
60.4
218
1.42
0.32
71.9
4.20
68.8
25.6
0.01
1.50
3.2
<2
<2
39
3
23Feb88
8.10
825
533
9
153
60.0
215
1.37
0.31
75.5
3.97
63.7
25.7
0.02
<0.01
3.8
<2
<2
43
4
29Feb88
8.10
828
534
10
154
59.5
219
1.46
0.33
73.1
4.27
66.2
25.6
<0.01
<0.01
3.3
<2
<2
35
13
2May88
8.16
859
545
9
159
60.1
212
1.68
0.38
77.0
4.16
69.5
26.2







-------
                     TABLE C.12 - EXPERIMENT A - EPA STANDARD SOLUTIONS
Week
Sampling Date
LabpH
Lab EC Oimhos/cm)
TDS (mg/l)
SiO2 (mg/l)
HCO3 (mg/l)
Cl (mg/l)
S04 (mg/l)
N03 (mg/l)
NO3 (mg/l as N)
Na (mg/l)
K (mg/l)
Ca (mg/l)
Mg (mg/l)
Fe (mg/l)
Mn (mg/l)
TOC (mg/l)
Benzene (fig/1)
Chlorobenzene (ng/l)
True Value

6.0
267
145


52.1
20.0


20.0
5.0
20.0
5.0
0.050
0.050
61.4
19.6
19.8
95% C.I.

5.91 - 6.11
247 - 288
114 - 177


48.2 - 55.4
16.3 - 23.1


17.8 - 22.3
4.17 - 5.71
17.5 - 22.2
4.18 - 5.62
0.0475 - 0.0545
0.0423 - 0.0560
51.7 - 69.3
na
na
6
14Mar88
7.58
272
130
<1
33
50.3
19.6
<0.04
<0.01
19.9
5.22
20.1
5.1
0.04
0.05
68.7
15

8
28Mar88













0.10
0.11
66.9
15

13
2May88













0.04
0.05
69.5
14

19
13Jun88

















16
Standards are subject to error during preparation (inaccurate dilution)
na = not available

-------
            TABLE C.13 - EXPERIMENT A - EQUIPMENT BLANKS - BLADDER  PUMP
Week
Sampling Date
Lab pH
Lab EC (nmhos/cm)
TDS (mg/L)
SiO2 (mg/L)
HCO3 (mg/L)
Cl (mg/L)
S04 (mg/L)
N03 (mg/L)
NO3 (mg/L as N)
Na (mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (ng/L)
Chlorobenzene (ng/L)
Chloroform (ng/L)
1
HFeb88
4.68
1.76
<5
2
<5
<0.2
1.0
<0.04
<0.01
<0.1
<0.1
<0.1
<0.1
<0.01
<0.01
0.74
<2
<2
<2
2
17Feb88
5.95
2.75
<5
2
<5
<0.2
<1
<0.04
<0.01
<0.1
<0.1
0.25
<0.1
<0.01
<0.01
1.62
<2
2
<2
3
24Feb88
5.42
1.34
<5
<1
<5
0.2
1.4
0.04
0.01
<0.1
<0.1
<0.1
<0.1
<0.01
<0.01
5.8
<2
<2
<2
4
2Mar88
5.16
1.44
<5
<1
<5
<0.2
1.4
<0.04
<0.01

-------
               TABLE C.14 - EXPERIMENT A - EQUIPMENT BLANKS - BAILER
Week
Sampling Date
Lab pH
Lab EC (nmhos/cm)
TDS (mg/L)
SiO2 (mg/L)
HC03 (mg/L)
Cl (mg/L)
S04 (mg/L)
N03 (mg/L)
NO3 (mg/L as N)
Na (mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (ng/L)
Chlorobenzene (ng/L)
Chloroform (ng/L)
1
HFebBB
4.90
2.83
<5
2
<5
0.2
<1
<0.04
<0.01
<0.1
<0. 1
<0. 1
<0. 1
No Data
No Data
2.5
<2
<2
<2
2
17Feb88
5.26
1.42
<5
<1
<5
<0.2
<1
0.04
0.01
<0. 1
<0. 1
<0. 1
<0. 1
<0.01
<0.01
2.1
<2
<2
<2
3
24Feb88
5.36
2.39
<5
<1
<5
0.4
1.8
0.09
0.02
<0-1
<0. 1
<0.1
<0. 1
<0.01
<0.01
2.4
<.1
<2
<2
4
2Mar88
5.01
1.83
<5
<1
<5
<0.2
1.3
<0.04
<0.01
0.1
<0. 1
<0.1
<0. 1
<0.01
<0.01
1.3
<2
<2
<2
6
16Mar88
5.28
1.72
<5
<1
<5
<0.02
<1
<0.04
<0.01
<0. 1
<0. 1
<0. 1
<0. 1
<0.01
<0.01
0.5
<2
<2
<2
8
29Mar88
5.47
1.93
<5
<1
<5
<0.2
<1
0.13
0.03
<0.1
<0. 1
<0. 1
<0. 1
<0.01
<0.01
0.3
Lost
Lost
Lost
13
4May88
4.84
3.00
<5
<1
<5
<0.2
1.0
<0.04
<0.01
0.1
<0. 1
<0. 1
<0. 1
<0.01
<0.01
1.0
<2
<2
<2
19
13Jun88
5.37
2.17
<5
<1
<5
<0.2
<1
<0.04
<0.01
<0.1
<0. 1
<0. 1
<0. 1
<0.01
<0.01
1.4
<2
<2
<2
No Data = no blanks collected
Lost = blank sample lost

-------
                                 TABLE C. IS - EXPERIMENT A  - TRIP BLANKS
Week
Sampling Date
Lab pH
Lab EC (nmhos/cm)
TDS (mg/L)
SiO2 (mg/L)
HCO3 (mg/L)
Cl (mg/L)
SO4 (mg/L)
NO3 (mg/L)
NO3 (mg/L as N)
Na (mg/L)
K (mg/L)
Ca (mg/L)
Mg (mg/L)
Fe (mg/L)
Mn (mg/L)
TOC (mg/L)
Benzene (ng/L)
Chlorobenzene (ug/L)
Chloroform (ng/L)
1
HFeb88
5.33
1.63
<5
2
<5
<0.2
1.4
<0.04
<0.01
<0. 1
<0.1
<0. 1
<0. 1
<0.01
<0.01
1.2
<2
<2
<2
2
16Feb88
5.29
1.18
<5
<1
<5
<0.2
<1
<0.0-l
<0.01
<0.1
<0.1
<0. 1
<0.01
<0.01
1.6
<2
<2
<2
3
22Feb88
5.21
1.83
<5
<1
<5
<0.2
1.0
<0.04
<0.01
<0. 1
<0.1
<0. 1
<0. 1
<0.01
<0.01
1.4
<2
<2
<2
4
IMarSS
5.10
1.59
<5
<1
<5
<0.2
1.1
<0.04
<0.01
<0. 1
<0.1
<0. 1
<0. 1
<0.01
<0.01
1.1
<2
<2
<2
6
14Mar88
5.35
1.44
<5
<1
<5
<0.2
<1
<0.04
<0.01
<0.1
<0.1
<0.1
<0. 1
<0.01
<0.01
0.8
<2
<2
<2
8
28Mar88
5.43
1.50
<5
<1
<5
<0.2
<1
<0.04
<0.01
<0.1
<0. 1
<0. 1
<0.01
<0.01
1.2
<2
<2
<2
13
2May88
5.52
2.00
<5
<1
<5
<0.2
<1
<0.04
<0.01
0.2
<0.1
<0. 1
<0.01
<0.01
0.7
<2
<2
2
19
13Jun88
5.28
1.35
<5
<1
<5
<0.2
<1
0.04
0.01
<0. 1
<0.1
<0.1
<0.01
<0.01
0.6
<2
<2
<2
19*
13Jun88
8.74
10.20
10
7
4.7
<0.2
<1
<0.04
<0.01
1.31
<0.1
0.86
No Data
No Data
0.7
<2
<2
<2
* - "Organic-free" distilled water used for blanks
No Data = no samples collected

-------
                           ADDITIONAL PART I EXPERIMENTAL WORK
                                            WEEK 19

     The relatively low VOC concentrations recovered by the multi-port prompted the development of an
additional set of field experiments designed to isolate the  source of this apparent  anomaly. The basic
experimental design consisted of collecting multiple samples from the multi-port, utilizing a variety of sampling
methods and sample handling procedures. Most of the samples were collected to investigate the effect of a
perforated septum in a standard VOA bottle on sample integrity and the claim that inverting the VOA bottles
during storage would have eliminated the loss of volatiles from the sample bottles.

EXPERIMENTAL DESIGN

     During the week of June 13,1988, full sets of samples from each well were collected to duplicate previous
sampling conditions. In addition to sampling with the bladder pump and the bailer, well 1 was also sampled with
the multi-port sampling probe and a small-diameter (1.91 cm) bailer. These samples were to be used to compare
conditions between well  1 and well 3 (the multi-port) to determine if the low benzene and chlorobenzene
concentrations were the result of different chemical conditions surrounding well 3.

     A total of 14 VOA samples were collected from the multi-port. Of these, two samples (PL-531 and 536)
were stored in an upright position, while the remaining samples were stored in an inverted position. Three bottles
(PL-538,540 and 562) were sealed with Parafilm* over the cap in an attempt to "seal" the needle punctures in
the septum. Sample PL-S39 was sealed with aluminum foil for the same reason. The septum on sample bottle
PL-542 was replaced with a new, unperforated septum. Sample PL-543 was collected using a newer model of
the MF VOA sample bottle holder which is designed to give the operator better  needle control and allows a
greater volume of water to pass through the VOA bottle before sample collection. Samples PL-544 and 545 were
collected in  a vented stainless steel container, carefully poured into a VOA bottle, and capped (no septum
perforations). Sample PL-S47 was collected in a stainless steel "high pressure vial." This vial is sealed by dosing
a valve and,  therefore, does not rely upon a flexible, perforated septum.

     When  all of the standard sampling procedures had been completed, the pumping port on the multi-port
was opened.  After purging the well with a peristaltic pump, a sample (PL-561) was collected from inside the well
casing with the small-diameter bailer. As noted earlier, this sample was to be compared to a sample collected
from well 1 with the same bailer. Sample PL-S62 was collected from the lowermost measurement port to check
for variations in VOC concentration after the installation had been purged.

RESULTS

      A summary of the sampling results is presented in Table  C.16. Based on the results of the experiment, it
appears that all of the multi-port samples which were collected and stored with a perforated septum experienced
high losses of both benzene and chlorobenzene relative to those samples stored in bottles with an unperforated
septum. Inverting the bottles apparently had little or no effect on VOC losses. Sealing the bottle with Parafilm
or foil apparently diminished the losses somewhat, but did not eliminate them.

      Unfortunately, samples PL-514, 519, 520,  and 521 were all apparently compromised in some unknown
fashion. Benzene and chlorobenzene  concentrations in all of these samples were anomalously low or below
detection limits despite normal results from  the analyses of the internal and surrogate standards. Therefore,
these data are not shown in Figures 18 through 20. Samples  collected from well 1 during later experiments
contained VOC concentrations which were  comparable to those from other wells. As a result, comparison
between wells 1 and 3 based on the small-diameter bailer samples was not possible. However, the benzene and
chlorobenzene concentrations found in sample PL-S61 suggest that the conditions near well 3 are, at the very
least, similar to conditions surrounding the other wells. Therefore, it is unlikely that the low VOC concentrations
                                                78

-------
associated with the multi-port throughout the seven previous sampling weeks could be attributed to anomalously
low concentrations of benzene and chlorobenzene surrounding well 3.

     The effect of a perforated septum on sample integrity can also be seen in the standard samples submitted
to the laboratory (PL-548, 549, and 550). The septa in two of the bottles were punctured with the same type of
needles used for multi-port sampling,  while a  third bottle was  simply inverted. The chlorobenzene
concentrations in these standards should have been 19.8 ppb. The low concentrations found in samples PL-549
and 550 suggest a greater than 50 percent loss of chlorobenzene due to the perforations in the sample bottle
septum. The 16 ppb chlorobenzene concentration found in sample PL-548 may indicate possible losses through
even an unperforated septum.  However, these losses are  probably due  to volatilization during sample
preparation.

     The mechanism for the loss of VOCs from the sample is not clear, but two possibilities appear likely.

         •    Passage of air through the septum and into the sample bottle, forming bubbles into
             which the VOCs may volatilize; and

         •    Direct volatilization to the atmosphere through the perforated septum.

      Air bubbles have been known to form in the samples during storage. However, the losses may be due to
a combination of the possibilities noted above.

      Based on the results of the additional field experiments, it is apparent that  a perforated septum in a
standard VGA bottle can have a considerable effect on sample integrity and should be replaced. The questions
raised concerning perforated septa should also be applied to the filter tip system in light of the fact that this
sampling device also utilizes flexible septum in sample vials. While it appears that the filter tip sample vials have
not experienced the high VOC losses characteristic of the multi-port, the potential for such losses, particularly
after the septum has been perforated several times, should be investigated.
                                                  79

-------
            TABLE C.16 - EXPERIMENT A - WEEK 19 SAMPLING RESULTS
Sample
Number
PL-486
PL-491
PL-496
PL-501
PL-506
PL-507
PL-514
PL-519
PL-520

PL-521
PL-526
PL-531
PL-536
PL-S37
PL-538
PL-539
PL-540
PL-541
PL-542
PL-543
PL-544
PL-545
PL-546
PL-547
PL-548
PL-549
PL-550
PL-551
PL-556

PL-561

PL-562
Benzene
(M*/L)
276
136
<2
152
<2
<2
34
<2
<2

8
<2
53
44
48
81
67
80
63
144
70
129
130
<2
134
<2
<2
<2
148
159

160

99
Chlorobenzene
(Mg/L)
1420
1510
<2
1010
<2
<2
<2
<2
<2

<2
<2
435
312
399
609
542
741
616
1500
650
1380
1370
<2
mo
16
6
8
1600
1710

1450

954
Well
5
2

4


1
1
1

1

3
3
3
3
3
3
3
3
3
3
3

3



6
6
3
3

3
Device
HOPE Filter Tip
PTFE Filter Tip

Well Probe


Bladder Pump
multi-port
Small-Diameter
Bailer
Bailer

multi-port
multi-port
multi-port
multi-port
multi-port
multi-port
multi-port
multi-port
multi-port
multi-port
multi-port

multi-port
Standard0
Standard0
Standard0
Pump/packer
Pump/packer

Small-Diameter
Bailer
multi-port
Treatment


Equip. Blank - Bailer

Trip Blank
Trip Blank - Organic-free H2O

Invert
Invert


Equip. Blank - Bladder Pump
Upright
Duplicate PL-531
Invert
Invert, paraseal
Invert, foil seal
Duplicate PL-538
Duplicate PL-537
Invert, new septum
New holder, invert
Vented bottle", invert
Vented bottleb, invert
Trip Blank - Organic-Free HjO
High pressure vial
No puncture, invert
2 punctures, invert
2 punctures, upright

Duplicate PL-551
(Open Pumping Port)
Invert

Invert, paraseal
Upright = VOA bottle stored in upright position
Invert = VOA bottle stored in inverted position
* - Vented bottle, short delivery tube
b - Vented bottle, long delivery tube
0 = Chlorobenzene standard; Concentration should be 19.8 ng/L
                                         80

-------
         TABLE C.17 - EXPERIMENT B - VOC ANALYTICAL RESULTS FOR SIX
        REPLICATE SAMPLES COLLECTED WITH EACH SAMPLING DEVICE -
                            NOVEMBER 8,1988
VOC (ng/L)
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-Dichlorobenzene
1,4-Dichlorobenzene
1, l-Dichloroethane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
Acetone
PTFE Filter Tip
44
42
40
42
41
41
1290
1270
1310
1310
1310
1250
217
213
225
222
223
211
21
19
21
20
21
19
369
355
381
352
360
336
33
32
34
33
34
33
11
10
10
10
10
10
28
28
30
29
29
27
11
10
10
10
10
10
ND
ND
ND
ND
ND
ND
Bailer
54
52
53
52
55
54
1070
1050
1070
1010
1050
1080
197
1%
189
185
202
203
17
16
16
19
20
19
314
300
315
285
308
314
32
31
32
30
31
32
9
9
9
9
9
8
22
22
22
22
23
23
11
10
10
10
12
10
ND
ND
ND
ND
ND
ND
Bladder Pump
65
66
63
65
66
65
1280
1220
1250
1220
1260
1240
214
214
216
202
210
202
18
20
19
18
18
19
345
332
337
330
337
347
36
31
32
33
32
33
11
9
10
9
10
10
25
25
26
25
25
26
11
11
11
11
10
9
ND
44
ND
58
ND
ND
Multi-Port
49
51
51
49
49
50
1230
1230
1250
1210
1270
1240
1%
1%
207
202
200
215
18
18
19
18
19
18
324
315
340
345
348
™
34
34
34
33
35
32
10
8
10
9
9
9
27
29
29
27
30
28
11
11
11
11
11
11
ND
ND
ND
ND
ND
ND
ND = Not Detected
                                    81

-------
         TABLE C.18 - EXPERIMENT B - QUALITY ASSURANCE SAMPLES


                                                 Equipment Blank  Equipment Blank
VOCQig/L)     Standard 1*  Standard 2*   Try Blank      Bailer
Benzene
Chlorobenzene
1,2-Dichlorobenzene
13-Dichlorobenzene
1,4-Diehlorobenzene
1, 1-Dichloroethane
1,2-Diehloroethane
Ethylbenzene
Tnchloroethene
Acetcme
Chloroform
ND
ND
ND
15
ND
ND
ND
ND
13
92
12
ND
ND
ND
14
ND
ND
ND
ND
14
104
14
ND
ND
ND
ND
ND
ND
ND
ND
ND
97
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
79
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND - Not detected
* - Standard True Valuer. Chloroform 202 n.g/L
                       m-Dichlorobenzene 20.0
                                     82

-------
TABLE C.19 - EXPERIMENT C - VOC ANALYTICAL RESULTS FOR SAMPLE ROUND 1
HOPE
VOC(ng/L) Filter Tip
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-DichIorobenzene
1 ,4-Dichlorobenzene
1,1-Dichloroethane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
44
40
39
1120
1130
1130
213
222
227
18
18
19
354
374
379
30
30
28
8
8
7
15
17
15
9
9
8
PTFE
Filter Tip
15
14
16
1260
1190
1240
166
189
204
13
17
18
386
348
377
27
26
25
7
7
7
30
27
30
8
7
8
Standard
Bailer
28
30
30
1030
1060
1060
234
237
246
22
23
23
329
338
350
32
30
30
8
7
8
25
35
24
9
0
8
BED
Bailer
31
30
30
1100
1050
1070
280
266
250
25
25
23
369
366
340
27
27
26
7
7
7
25
23
24
9
9
7
Bladder
Pump
34
35
33
1250
1260
1230
218
211
203
18
19
18
365
361
355
26
27
26
7
7
7
28
27
27
9
9
9
Multi-
Port
20
20
20
1210
1250
1260
206
214
210
17
18
18
339
348
354
25
26
25
7
7
7
29
32
31
8
8
8
                                83

-------
 TABLE C20 - EXPERIMENT C - VOC ANALYTICAL RESULTS FOR SAMPLE ROUND 2

VOC(»ig/L)

Benzene


Chlorobenzene

1,2-Diehlorobenzene

13-Dichlorobenzene

1,4-Dichlorobenzene

1,1-Dichloroethane


1,2-Dichloroe thane

Ethylbenzene

Trichloroethene

HDPE
Filter Tip
44
45
48
869
843
884
188
181
194
16
16
17
283
269
293
34
34
34
9
9
9
14
13
15
9
8
9
PTFE
Filter Tip
14
13
ND
1040
1010
ND
179
175
ND
17
17
ND
285
273
ND
33
31
ND
8
9
ND
31
30
ND
9
8
ND
Standard
Bailer
23
23
22
894
885
867
165
163
163
IS
15
15
261
253
248
32
32
31
8
8
8
25
25
24
10
9
9
BED
Bailer
20
21
23
818
827
839
158
155
157
15
16
15
239
234
243
32
32
33
8
7
9
24
24
24
8
9
8
Bladder
Pump
ND
25
24
ND
1020
1010
ND
184
183
ND
16
17
ND
284
279
ND
37
34
ND
10
9
ND
27
27
ND
10
10
Multi-
Port
15
14
14
985
971
968
173
174
172
15
16
16
263
261
260
30
30
31
8
8
8
30
29
29
10
8
9
ND = No Data
                                 84

-------
TABLE C21 - EXPERIMENT C - VOC ANALYTICAL RESULTS FOR SAMPLE ROUND 3
HOPE
VOC(ng/L) FaterTip
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-DichIorobenzene
1,4-Dichlorobenzene
1, 1-Dichloroe thane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
28
27
27
787
769
759
174
171
175
18
17
18
286
274
285
31
31
32
9
8
8
7
6
6
8
8
8
PTFE
Filter Tip
12
13
13
977
1050
1080
167
163
181
16
18
18
258
267
286
30
31
33
8
9
10
32
33
35
9
9
9
Standard
Bailer
12
17
16
673
901
840
133
171
140
14
17
14
202
255
226
27
30
29
7
8
8
21
28
24
9
9
8
BED
Bailer
14
10
8
740
557
474
150
132
124
15
15
14
223
190
179
32
30
27
8
8
7
22
18
16
9
9
8
Bladder
Pump
18
19
20
917
959
969
153
159
158
15
16
16
243
253
250
29
30
31
8
8
8
28
29
29
8
9
10
Multi-
Port
12
12
12
910
947
940
146
154
146
16
16
15
234
241
241
30
30
32
8
8
9
30
30
31
9
8
9
                              85

-------
     TABLE C.22 - EXPERIMENT C - VOC ANALYTICAL RESULTS FOR SAMPLE ROUND 4
VOC(ng/L)
Benzene
Chlorobenzene
1,2-Dichlorobenzene
1,3-Dkhlorobenzene
1,4-Dichlorobenzene
1,1-Dichloroe thane
1,2-Dichloroethane
Ethylbenzene
Trichloroethene
HOPE
Filter Tip
20.1
18.2
17.6
666
645
625
175
169
172
17.0
16.9
16.9
294
293
292
27.8
27.4
28.6
7.7
7.5
8.4
14
1.7
1.9
8.3
8.3
7.6
PTFE
Filter Tip
12.8
12.7
1Z7
1100
1090
1090
176
172
166
16.0
15.9
15.5
289
291
275
30.4
29.8
30.1
9.1
9.2
9.3
33.2
313
32.8
9.5
9.1
9.6
Standard
Bailer
13.3
13.0
14.4
845
813
860
144
142
140
13.0
13.8
13.1
225
225
223
28.4
29.1
29.1
8.8
8.6
8.6
214
21.6
22.6
8.6
8.6
9.0
BED
Bailer
8.8
12.3
12.0
529
772
751
105
135
136
113
13.2
13.5
118
197
202
25.0
27.0
26.9
7.1
7.9
7.3
15.9
21.3
20.9
8.0
8.6
8.4
BED
Bailer*
9.0
9.0
9.0
517
524
528
95.8
99.5
100
11.8
12.0
11.8
112
115
110
24.9
25.8
27.6
6.8
7.2
8.8
15.8
15.7
15.9
8.0
7.5
8.1
Bladder
Pump
15.6
15.0
15.1
940
933
938
156
159
158
13.8
14.6
14.5
245
264
256
28.9
28.2
313
8.7
8.0
9.9
24.8
25.8
25.0
9.4
8.8
9.3
Multi-
Port
10.1
10.7
10.5
929
958
940
154
149
14S
13.8
13.5
13.6
250
240
242
26.3
27.4
26.9
7.3
7.5
7.1
26.9
27.6
26.6
8.3
9.4
8.7
'Samples collected by decanting from top of bailer, as described in text.
                                        86

-------
           TABLE C.23 - EXPERIMENT C - QUALITY ASSURANCE BLANKS

Date
Round 1
17 Feb 89
21 Feb 89
23 Feb 89

23 Feb 89
Round 2
3 Jun 89
3 Jun 89
3 Jun 89
5 Jun 89
5 Jun 89
6 Jun 89
7 Jun 89
7 Jun 89
Round 3
19 Aug 89
19 Aug 89
21 Aug 89
21 Aug 89
21 Aug 89
22 Aug 89
22 Aug 89
Round 4
5 Dec 89
5 Dec 89
6 Dec 89
6 Dec 89
6 Dec 89
6 Dec 89
7 Dec 89

Blank Type

Rinsate
Trip
Rinsate

Trip

Rinsate
Rinsate
Rinsate
Rinsate
Trip
Trip
Rinsate
Trip

Rinsate
Rinsate
Rinsate
Trip
Rinsate
Trip
Rinsate

Rinsate
Rinsate
Trip
Trip
Rinsate
Rinsate
Trip

Device

Bailer
	
Bladder Pump

—

Bailer
BED Bailer
VGA Bottle
Bladder Pump
—
—
BAT Bottle
—

Bailer
BED Bailer
VGA Bottle
—
Bladder Pump
—
BAT Bottle

BED Bailer
Bailer
—
—
BAT Bottle
VGA Bottle
—
Method 624 Compounds Detected
and Concentrations

ND
ND
chlorobenzene, 2 u.g/L,
1,4-dichlorobenzene, 4^g/L
ND

ND
ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND

ND
ND
ND
ND
ND
ND
ND
ND = Not Detected
                                    87

-------
TABLE C.24 - DUPLICATE ANALYSES OF RANDOM SAMPLES
Dale
Collected
23 Fcb 89
23 Fcb 89
5 Jim 89
5 Jim 89
6 Jun 89
21 Aug 89
21 Aug 89
22 Aug 89
7 Dec 89
7 Dec 89
7 Dec 89
7 Dec 89
ND - Not
Order of
Analysis
a
b
a
b
a
b
a
b
a
b
a
b
a
b
a
b
a
b
a
b
a
b
n
b
Detcclcil
Concrmralion (flg/l.)
Dcnzcnc
30
30
ND
ND
20
21
ND
ND
U
13
12
12
ND
ND
13
13
13
13
15
15
9
9
13
13

Chlorobenzcnc 1
1047
1095
2
2
818
818
ND
ND
1037
1028
673
685
ND
ND
1080
1060
1090
1070
938
939
529
515
845
821

i,2-Dichlorobcnzvncl,3-l)iclitorobcnzencl,4-Dichlar«bcnzciic
266
2)8
ND
ND
158
155
ND
ND
179
182
133
134
ND
ND
181
180
166
171
158
157
105
98
144
139

25
25
ND
ND
15
15
ND
ND
17
17
14
14
ND
ND
18
1R
16
16
15
14
12
12
13
13

366
360
4
3
239
237
ND
ND
285
282
202
210
ND
ND
280
279
275
288
256
256
118
112
225
217

1,1-Dicliloroclhnnc
27
26
ND
ND
32
31
ND
ND
33
31
27
28
ND
ND
33
31
30
3(1
32
30
25
24
28
28

1 ,2-Dichlornctliiiiic iiihylbvnzcne
7
7
ND
ND
8
8
ND
ND
8
9
7
g
ND
ND
10
9
9
9
10
9
7
7
9
9

23
25
ND
ND
24
24
ND
ND
31
31
21
22
ND
ND
35
35
33
33
25
25
U>
15
22
22

'IViclllOHKlllCIIC
9
9
ND
ND
8
9
ND
ND
9
9
9
8
ND
ND
9
9
10
to
9
9
8
7
9
8


-------
                        APPENDIX D


       DATA PLOTS AND STATISTICAL ANALYSES RESULTS
_o
u

o
U
0)

4>
N
    70.0
    60.0
    50.0
    40.0
    30.0
                          "\
                          mm]
                 PTFE
               Filter Tip
Standard
 Bailer
Bladder
 Pump
Multi-
 Port
                          Sampling Device

          FIGURE D.I - EXPERIMENT B - BENZENE DATA
14OU.U
1*
3-1300.0
0)
§
i
| 1200.0
o
U
1 1100.0
s
M
.a
o
o 1000.0
900.0
•
• •• |
T •
)* 1
••T
*


.
- I
"i ..
• •f l-t-Sx
1 1
|_; •

PTFE Standard Bladder Multi-
Fitter Tip Baiter Pump Port
                           Sampling Device


       FIGURE D.2 - EXPERIMENT B - CHLOROBENZENE DATA

-------
§
1
240.0
220.0
200.0
o
s
o

|
N
C
.Q

|   180.0
o
S
 160.0
              PTFE
             Filter Tip
                         Standard
                          Bailer
Bladder
 Pump
Multi-
 Port
                       Sampling Device
FIGURE DJ - EXPERIMENT B - 1,2-DICHLOROBENZENE DATA
SO
i-
i
hlorobanzen* Cor
« 8
o
5
*t
1O


""1
* 1 " i "
•• ••! •• 1 ••
•1 " ""4

I;

PTFE Standard Bladder Multi-
FilterTip Bailer Purrp Port
                      SampBng Device


FIGURE D.4 - EXPERIMENT B - 1,3-DICHLOROBENZENE DATA
                          90

-------
   400.0
o
•&
   360.0
O
o>

s
o
s
_o
o
   320.0
   280.0
                 PTFE
               Filter Tip
Standard
 Bailer
Bladder
 Pump
Multi-
 Port
                          Sampling Device
   FIGURE D.5 - EXPERIMENT B - 1,4-DICHLOROBENZENE DATA
4U
§
o 35
I
«>
o
o
0 30
1
«
o
1 25
o
0
i
20

• *
" i ""I
•••* "' " 1
• ••• 1 •• •
.. {
i
•



i+s*
1
t*
'-sr
.
PTFE Standard Bladder Multi-
Filter Tip Bailer Pump Port
                         Sampling Device


    FIGURE D.6 - EXPERIMENT B - 1,1-DICHLOROETHANE DATA
                              91

-------
*M
1
§15
•s
•E
o
Q
0 10
i
^B

o
5
1 5
a
5
™-
Q

. .



• •
•••••T •••! •• i
•••••1 •• 1 • ••?



I+ST
1
I'
'-%
i • i i
PTFE Standard Bladder Multi-
Filter Tip Bailer Pump Port
                      Sampling Device


 FIGURE D.7 - EXPERIMENT B - U-DICHLOROETHANE DATA
3-
ok
  30
_o


•E
S25
•5.20

§
  15
             PTFE
            Filter Tip
Standard
 Bailer
Bladder
 Pump
Multi-
 Port
                      Sampling Device

    FIGURE D.8 - EXPERIMENT B - ETHYLBENZENE DATA
                          92

-------
O>
t«
O
4)
O
•C
                          )     ....]
             PTFE    Standard    Bladder     Multi-
            Filter Tip    Bailer      Pump     Port
                      Sampling Device
  FIGURE D.9 - EXPERIMENT B - TRICHLOROETHENE DATA
                           93

-------
  100
-j.80

t


i"
u

I  40
£ 20
              JT

              -ST
 HOPE   PTFE  Standard
Filter Tip Filter Tip  Baiter

            Sampling Device
BED
Bailer
                                      Bladder  Multi-
                                      Pump    Port
       FIGURE D. 10 - EXPERIMENT C - BENZENE DATA
Chlorobtnzen* Concentration (ug/L)
§ § 8 i i 1
i < *
\ \ \ •
. \ \ \
\ ( \f i ^ i
* \ V ^ •
, . , >, v
• \ ^ \
\ \
\ )/ i+^-
1; '

HOPE PTFE Standard BED Bladder Multi-
Fitter Tip Filter Tip Baiter Bailer Pump Port
                      Sampling Device
    FIGURE D.11 - EXPERIMENT C - CHLOROBENZENE DATA

-------
«JVW
l>
t 400
o
•&
$ 300
0
O
s
N 200
c
0>
n
0
o
•g 100
Q
I
0

I+ST

„
*
-*

•
\ \ * ,
\ t \ \ *• *

^-* I*A^ * * fe
' ^ 1 ^ W
I » 4 ^"*
1 'W
'


           HOPE   PTFE  Standard   BED  Bladder  Multi-
          Fitter Tip Filter Tip  Bailer   Bailer   Pump    Port

                       Sampling Device


FIGURE D.12 - EXPERIMENT C - 1,2-DICHLOROBENZENE DATA
wv
C
5 25
1
c
o
" 20
lorobenzen
»A
Ol
€
0
1
m

n*
t-, •
»
i
i
i i
I
** J ; i * t
\A ri ' , 1 \
* ' ^ U \. H H
K \ * V
T

• !••!•
          HDPE   PTFE  Standard  BED   Bladder  Multi-
         Fitter Tip Filter Tip Bailer   Bailer  Pump   Port

                        Sampling Device


FIGURE D.13 - EXPERIMENT C - 1,3-DICHLOROBENZENE DATA
                            95

-------
ouu
3-
f 40°
i
•
g 300
O
c
g 200
|
i 100
5
n
i b • • • •

.
\ \ • \ \ '
\ \ \ V
\ ' \ 1 I \
' w' ^ Vi vw V '
t- V " .
l+?r .
U
,
         Filter Tip Filter Tip  Bailer   Bailer   Pump    Port

                     Sampling Device


FIGURE D.14 - EXPERIMENT C - 1,4-DICHLOROBENZENE DATA
  we
  35
o
O
• JW




1
o

125

O
   20
|?
'-%•
          HOPE   PTFE  Standard  BED  Bladder  Multi-
         Filter Tip Filter Tip  Bailer   Bailer   Pump    Port

                      Sampling Device
 FIGURE D.1S - EXPERIMENT C - 1,1-DICHLOROETHANE DATA

-------
0)
o
o in
O  w

c

i
o
o  5
o
5
                       tY
A
                                           + ST
         HOPE   PTFE Standard  BED  Bladder  Multi-
        Filter Tip Filter Tip Bailer   Bailer   Pump   Port

                     Sampling Device


FIGURE D.16 - EXPERIMENT C - 1,2-DICHLOROETHANE DATA
^Vl
3"
ok
3.30
c
0
H
CO

0
c 20
o
C
«
c
«
N
C
£ 10
S.
£.
UJ

Q
• » i P » ,
h
L- \V\
T i ^ \
v 4* 4 (k
v V
v^ h
\i i
i \
Tl
I 1
1 1
P^l
Y

\
\
\ |+S*
i I -
t t X
V l-fe
V ^

HOPE PTFE Standard BED Bladder Multi-
Filter Tip Filter Tip Bailer Bailer Pump Port
                      Sampling Device

   FIGURE D.17 - EXPERIMENT C - ETHYLBENZENE DATA
                          97

-------
t»
o
§10
o
u
                        ,H«
          HOPE  FIFE  Standard  BED   Bladder  Multi-
         Filter Tip Filter Tip  Bailer   Bailer  Pump    Port

                     Sampling Device



  FIGURE D.18 - EXPERIMENT C - TRICHLOROETHENE DATA
                           98

-------
        TABLE D.1 - EXPERIMENT C - TUKEY TEST RESULTS FOR ROUND 1
Rank* Benzene Chlorobenzene 1,2-Dichlorobenzene
1 HDPE Filter Tip
2
3 f
Bladder Pump
BED Bailer
Bladder Pump
Multi-Port
BED Bailer
Bailer
PTFE Filter Tip HDPE Filter Tip
4 [ Bailer F HDPE Filter Tip
5 PTFE Filter Tip [
6 Multi-Port
. BED Bailer I
Bailer
Bladder Pump
Multi-Port
PTFE Filter Tip
Rank* 1,3-Dichlorobenzene 1,4-Dichlorobenzene 1,1-Dichloroethane
1
2
3
4
5
6
BED Bailer
. Bailer
Bladder Pump
HDPE Filter Tip
Multi-Port
. PTFE Filter Tip
PTFE Filter Tip
HDPE Filter Tip
Bladder
BED Bailer
Multi-Port
. Bailer
Bailer
.HDPE Filter Tip
BED Bailer
Bladder Pump
PTFE Filter Tip
. Multi-Port
Rank* 1.2-Dichloroethane Ethylbenzene Trichloroethene
1
2
3
4
5
6
HDPE Filter Tip
Bailer
PTFE Filter Tip
BED Bailer
Bladder Pump
Multi-Port
PTFE Filter Tip
Bailer
Bladder Pump
BED Bailer
Multi-Port HDPE FUter Tip
Bladder Pump
HDPE Filter Tip
BED Bailer
Multi-Port
PTFE Filter Tip
Bailer
•Ranked from highest to lowest concentration
                                     99

-------
        TABLE D.2 - EXPERIMENT C - TUKEY TEST RESULTS FOR ROUND 2
Rank* Benzene
1
2
3
4
HDPE Filter Tip f
Bladder Pump L
Bailer
BED Bailer 1
5 [ Multi-Port f 1
6 [PTFE Filter Tip [
Rank* 1,3-Dichlorobenzene
1
2
3
4
5
6
PTFE Filter Tip
Bladder Pomp
HDPE Filter Tip f
Multi-Port I
BED Bailer [
Bailer [
Rank* 1^-Dichloroethane
1
2
3
4
5
6
Bladder Pump
HDPE Filter Tip
PTFE Filter Tip
Bailer
BED Bailer
Multi-Port
Chlorobenzene
PTFE Filter Tip
Bladder Pump
Multi-Port
Bailer
HDPE Filter Tip
BED Bailer
l,4-Dichlor6benzene
Bladder Pump
HDPE Filter Tip
PTFE Filter Tip
Multi-Port
Bailer
BED Bailer
Ethylbenzene
IPTFE Filter Tip
Multi-Port
Bladder Pump
[Bailer
BED Bailer
HDPE Filter Tip
1,2-Dichlorobenzene
[HDPE Filter Tip
I Bladder Pump


PTFE Filter Tip
Multi-Port
[Bailer
BED Bailer
1, 1-Dichloroethane



•


Bladder Pump
HDPE Filter Tip
BED Bailer
PTFE Filter Tip
Bailer
Multi-Port
Trichloroethene






Bladder Pump
Bailer
Multi-Port
HDPE Filter
PTFE Filter Tip
BED Bailer
•Ranked from highest to lowest concentration
                                     100

-------
       TABLE D3 - EXPERIMENT C - TUKEY TEST RESULTS FOR ROUND 3
Rank* Benzene Chlorobenzene 1,2-Dichlorobenzene
1 HOPE Filter Tip
2
3
4
5
6

Bladder Pump
Bailer
PTFE Filter Tip
Multi-Port
PTFE Filter Tip
Bladder Pump
Multi-Port
Bailer
F HDPE Filter Tip
BED Bailer [ BED Bailer
HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
Multi-Port
Bailer
BED Bailer
Rank* 13-Dichlorobenzene 1,4-Dichlorobenzene 1,1-Dichloroethane
1
2
3
4
5
6



HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
. Multi-Port [
Bailer
HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
, Multi-Port
Bailer
BED Bailer BED Bailer
HDPE Filter Tip
PTFE Filter Tip
Multi-Port
Bladder Pump
BED Bailer
Bailer
Rank* 1,2-Dichloroethane Ethylbenzene Trichloroethene
1
2
3
4
5
6
PTFE Filter Tip
HDPE Filter Tip
Multi-Port
Bladder Pump
Bailer
PTFE Filter Tip
Multi-Port
Bladder Pump
Bailer
BED Bailer
BED Bailer HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
Multi-Port
Bailer
BED Bailer
HDPE Filter Tip
*Ranked from highest to lowest concentration
                                    101

-------
        TABLE D.4 - EXPERIMENT C - TUKEY TEST RESULTS FOR ROUND 4
Rank*
1
2
3
4
5
6
Rank*
1
2
3
4
5
6
Rank*
1
2
3
4
5
6
Benzene Chlorobenzene 1,2-Dichlorobenzene
HOPE Filter Tip






Bladder Pump
Bailer
. PIPE Filter Tip
PTFE Fflter Tip
Multi-Port
Bladder Pump
f Bailer
BED Bailer f I BED Bailer
Multi-Port I HDPE Filter Tip
HDPE Filter Tip
FIFE Filter Tip
. Bladder Pump
. Multi-Port
Bailer
• BED Bailer
1,3-Dichlorobenzene 1,4-Dichlorobenzene 1,1-Dichloroethane
HDPE Filter Tip
FTFE Filter Tip






Bladder Pump
Multi-Port
Bailer
BED Bailer
HDPE Filter Tip
PTFEFflterTip
Bladder Pump
, Multi-Port
Bailer
BED Bailer
PTFE Filter Tip
Bladder Pump
Bailer
I HDPE Filter Tip
Multi-Port
BED Bailer
1^2-Dichloroethane Ethyftenzene Trichloroethene



'



PTFEFflterTip FIFE Filter Tg>
Bladder Pump \ Multi-Port
Bailer
HDPE Filter Tq>
L Bladder Pump [
f Bailer
BED Bailer [ BED Bailer
Multi-Port HDPE Filter Tip
PTFE Filter Tip
Bladder Pump
Multi-Port
I Bailer
BED Bailer
HDPE Filter Tip
•Ranked from highest to lowest concentration
                                    102

-------

-------