EPA 510-K-92-812

                            Report No. KT-88-003(R)

       FUEL VAPOR BACKGROUND
CONCENTRATION MEASUREMENT
             AND  TRACER TESTING
                 IN UNDERGROUND
        STORAGE TANK BACKFILL
            SAN JOSE,  CALIFORNIA
                  Revised  March  1988
                                 Printed on Recycled Paper

-------
f

-------
                               :               Report No. KT-88-003(R)

                           REPORT
FUEL VAPOR BACKGROUND CONCENTRATION MEASUREMENT
                    AND TRACER TESTING
        IN UNDERGROUND STORAGE TANK BACKFILL
                       San Jose, California
                      Revised March 1988
                     Subcontract  216-8900-9
                          Submitted to:
                    Midwest Research Institute
                       5201 Leesburg Pike
                            Suite 209
                     Falls Church, Virginia
                             22041
                              By:

                         Kaman Tempo
                         816 State  Street
                    Santa Barbara, California
                             93101
                         (805)963-6479

-------

-------
                                EXECUTIVE SUMMARY
     This study was performed in two parts:   1)  measurement  of the fuel  vapor
concentrations  in  underground  tank backfill,  and 2)  development of  tracer
testing methods for  determing  the  efficacy of vapor  monitoring  systems.   All
sites were located in the City of San Jose, California.

Background Concentration Measurement
     Total fuel  hydrocarbon vapor  concentration  was  measured in  75  existing
vapor wells  at  11  service  stations  and  one emergency power  generation  site.
The wells were  located within  tank  and  pipeline  backfill.   Tank contents were
limited to gasoline and diesel.

     Concentration was  measured downhole,  one  foot  from the bottom of  the
well, using  a  Photovac  TIP-I connected to  a  25-foot  length of  1/16"  ID PTFE
tubing.   Readings  ranged from 0 to  2,000  ppm,  with a mean value  of  300 ppm.
Concentrations were reported by volume as isobutylene.

     The  following conclusions can be made from the background measurements:

     •    Background concentrations are commonly greater in tank backfill than
          in pipeline backfill.  Kaman  Tempo  believes that this is ascribable
          to the common occurance of tank overfills.
     t    Stratification of soil vapor within tank backfill  can occur.   If a
          well  is  perforated over  a short  interval,  leakage  in an unscreened
          stratum  may be  undetected.   Therefore,  each  vapor well  should be
          perforated over the entire backfilled depth.

Tracer Testing
     All  sites  selected for tracer  tests  had very low  fuel  vapor background
concentrations.  Measured volumes of the tracers were  injected into a "source"
well  after  which concentration was monitored for  several  hours in "observa-
tion" wells within the pipeline and  tank backfill.


-------
     Three  tracers were  used:   sulfur  hexafluoride  (SFg),  fluorocarbon-11
(F-ll) and propane.  SFg was detected by a thermal conductivity meter, whereas
F-ll and propane were  detected  using  a  dual-column gas chromatograph equipped
with a  photoionization detector.  The  most  conclusive data  obtained  was  for
propane.  Propane  travel  velocities  were observed  to  be on  the  order of  0.5
feet per hour.   SFg travelled more rapidly at  approximately  7 feet per hour.
F-ll proved to be a poor tracer due to its poor detectability.

     A  method  was  developed  for predicting  a  tracer's concentration  at  any
distance from the  source well at  any  time after its release using a mathemat-
ical model.  Such a model can aid the design of well spacings and the detector
selection for future monitoring systems.

     The following  conclusions can be made from the tracer tests:

     •    Vapor  wells  can  be effectively used  to monitor  underground product
          releases  from both tanks and pipelines.
     •    Because  tracers were  not detected on the side of the tank opposite
          the  source  well,  in  multiple  tank  installations,   shorter  well
          spacings  are recommended  in  the direction  transverse to  the  tank
          axis than in the  longitudinal direction.
     •    Additional  testing is  needed  to  relate  the diffusion  rates  (and
          concommittant  detection time)  of petroleum  products  and surrogate.
          compounds used as tracers.
                                         -n-

-------
                               TABLE OF CONTENTS
Section

   1         INTRODUCTION ...........................................      i'1

   2         INVESTIGATIVE METHODS ....................... . ..........      2'1

             21   DESCRIPTION OF VAPOR MONITORING WELLS ............      2-1
             22   BACKGROUND CONCENTRATION TESTING METHODS .........      2-2
             2.3   TRACER TESTING ............................... -----      2-4
                   2.3.1  Field Procedures, Sites #1 and #2 .........      2-9
                   2.3.2  Field Procedures, Site #3 .................      2-12

   3         RESULTS ................................................      3'1

             3 . 1   BACKGROUND MEASUREMENT RESULTS ...................      3-1
             3.2   TRACER TEST RESULTS. ......................... ....      3-6
             3.3   GAS DIFFUSION MODEL  DEVELOPMENT AND FIELD TESTING.    3-8

   4         CONCLUSIONS .............. . .............................      4~1

 Appendix

   A         SELECTED CITY  OF SAN JOSE  HAZARDOUS MATERIALS DIVISION
             SITE  INFORMATION SHEETS.. ..............................     A~l

   B         LETTER OF AUTHORIZATION ................................     B-l

    C         ADDITIONAL  SITE SKETCHES  ..............................     C-l

    D         VAPOR MONITORING EQUIPMENT SPECIFICATIONS ..............     D-l

    E         SITE  #3  TRACER TEST DATA ............................. . •     E-l

    F         CURVE MATCHING SOLUTION OF INSTANTANEOUS LINE SOURCE
             GAS DIFFUSION MODEL ....................................     F-!

-------

-------
                                 LIST.OF TABLES


Table                                                                    Page

 2-1      Physical Properties of Selected Gases 	      2-8

 2-2      Detector Sensitivity for Selected Compounds 	     2-10

 3-1      Background Fuel Vapor Concentrations at Underground
          Tank' Sites 	      3-2

-------

-------
                                LIST OF FIGURES
Figure                                                                   Page
  2-1        Sketch of Typical  Underground Storage Tank Site	      2-3
  2-2        Sketch of Sites #1 and #2	      2-5
  -2-3        Sketch of Site #3	      2-6
  3-1        Frequency Plots of Concentration Measured in Pipeline
             and Tank Backfill	      3-7
  3-2        Time Versus SFg Concentration Plots	      3-9
  3-3        Time Versus Propane Concentration Plots	      3-10
  3-4        Selected Tracer Test Chromatograms	      3-11
  3-5        Well V-2, Test A  - Solution of D1 and A by Curve
             Matching Method	      3-13
  3-6        Well V-5, Test A - Solution of D1 and A by Curve
             Matching Method	      3-14
  3-7        Well V-5, Test B  - Solution of D1 and A by Curve
             Matching Method	      3-15

-------

-------
                                    SECTION 1
                                  INTRODUCTION

     This  report  describes  the methodology  and results  of background  fuel
vapor measurements and gaseous  tracer  tests  in  Underground Storage Tank (UST)
backfill.  The twofold purpose  of  the  study was:  1) to  expand  the data  base
of fuel  vapor  concentration in tank backfill,  and  2) to  develop  methods  for
determining the efficacy  of vapor monitoring systems.   Kaman Tempo performed
this  study during November and December  1987  under subcontract  to  Midwest
Research  Institute  as a  contribution  to the  EPA  UST Federal Guidelines  and
Regulations.

     Tracer tests are commonly  used  to construct gas  transport models  and  are
easily applied to fuel vapor applications.   Such models can provide an  approx-
imation of the volume and rate of a theoretical product leak detectable by  the
lower detection threshold of the monitoring system.   Ultimately,  this informa-
tion  can  be  used  for system design and  operation optimization and for deter-
mining regulatory performance standards.

      In  a  previous  study  (Kaman,  September 1987), Kaman  Tempo measured back-
ground levels in tank backfill at known leak sites and demonstrated that these
concentrations were statistically higher than  those measured at  sites  with no
leak  history.  These  tests did not  address  vapor migration or  timeliness of
leak  detection.   Bench scale  tracer  tests using  actual  field  products  have
been  conducted by other workers to  study migration  rates; however, the appli-
cation of  their  results  must assume boundary  conditions  that may differ from
actual tank  conditions.   Therefore, Kaman  Tempo believes  that on-site tracer
testing  is the most valuable tool  available for evaluating vapor migration in
tank  backfill,  excluding  monitoring  of actual  leaks  in progress, both  for
testing of individual installations and for design purposes.

      The  field  work  for  this  study was conducted  in  the City  of San Jose,
California, where vapor monitoring  is  required in the backfill of all existing
USTs.   The San  Jose Fire  Department  Hazardous Materials  Program maintains
information sheets for each site which lists vapor well and tank construction
data.  The sites  visited  and  tested  in this  study were  selected from these
                                        1-1

-------
sheets.  Seventy-five vapor wells at twelve service stations were measured for
fuel  vapor concentration.   Tracer  tests  were  conducted  at  three of  these
sites.   All  chemical  analyses and  measurements were  performed with  field-
portable equipment.

     Kaman Tempo would like to thank Charles J. Wilhelm and Joseph C. Afong of
the City  of San Jose  Hazardous Materials  Program  for their  cooperation  and
assistance in the field portion of this study.
                                       1-2

-------
                                   SECTION.2
                             INVESTIGATIVE METHODS

     Twelve test sites were selected from 58  site information  sheets  provided
by the City of San Jose according to the following criteria:

     Sites
          •    distributed within a four mile radius
          •    minimum number of fuel  company brand names

     Wells
          •    uniform construction and design
          •    not instrumented with continuous monitoring devices.

These criteria minimized  logistics, reduced the  number  of variables  affecting
results,  and  minimized disturbance of the existing monitoring  systems.   Site
accessibility  and  operator  cooperation were  also  optimized.   Information
sheets  for  the selected sites are  presented  in Appendix A.  Site access  was
further aided by a Letter of Authorization from the City of San Jose  which was
presented to the attendant of each station (see Appendix B).

     The  site information sheets contain entry spaces for facility address and
name, number  of tanks, tank  and piping construction data,  well  construction
data, soil  analysis data,  depth to groundwater, and vapor monitoring  data.
Because the information sheet program  is still  being developed, many items on
the  sheets,  including age  of  tanks,  depth  to groundwater,  and  site  leak
history,  were not available for most of  the sites.

2.1  DESCRIPTION OF VAPOR MONITORING WELLS
     Total  fuel  hydrocarbon vapor concentration measurements were taken at 11
non-instrumented service  stations and one instrumented emergency power genera-
tion site.   In all,  75 wells were tested.  All service stations were operated
either  by Arco or  Unocal.  Vapor well  configurations at these sites typically
included  two  or more  wells within the tank backfill area, wells at the end of
pump islands  in  the pipeline  backfill,  a well  in  the  vent  line  backfill, and  a
                                        2-1

-------
background well  at some distance  from pipelines or tanks  (Figure  2-1).   The
configuration of wells  for  Site Numbers 1, 2, and 3 are  shown  in Figures 2-2
and 2-3.   Configurations  of the remaining sites  are presented  in Appendix C.
Tank  contents  were  primarily gasoline,  although some diesel  and waste oil
tanks were encountered.

     The wells  were constructed of  either one-  or two-inch  nominal  diameter
PVC pipe,  slotted  to within two to five feet  below the  surface and fitted at
the top with a threaded coupling and threaded plug.  The majority of the wells
had a one-inch  diameter.  An  uncapped  i-inch hose fitting was threaded into a
hole drilled into either the side of the threaded coupling or top of the plug.
Well  head  assemblies were enclosed with  steel  traffic-rated  monuments having
12-inch  threaded  steel tops.   Some  sites where measurements were  not taken
included a lock box with a  keyed  padlock within  the  monument.   Wells inter-
cepted  the entire  backfill  depth interval.   Wells  within pipeline  backfill
tended  to  be less  than five feet deep,  whereas tank backfill wells  were as
deep  as  18 feet.

2.2   BACKGROUND CONCENTRATION TESTING  METHODS
      Fuel  vapor concentrations  were  measured downhole using  a  Photovac Model
TIP-I connected to a 20-foot, probe constructed of 1/16-inch I.D. PTFE tubing.
The  intake end of the probe was  weighted with  a i-inch  brass  bell  reducer
which acted  as  a centralizer to avoid aspirating water condensation from the
walls of the well  casing  into the instrument.  TIP-I specifications are  given
in Appendix D.

      Prior to well  measurement at  each  site,  the  instrument was  zeroed to
ambient  air  and calibrated to  100 ppmV isobutylene.   The same probe was  used
for  downhole  measurement and  instrument  calibration  in  order  to correct for
possible progressive contamination of  the  probe.

      Each  well  was opened by removing  the threaded PVC  plug  and then  sounded
for water  with  an  electric water  level sounder.  No downhole measurement was
made  if the  well  was flooded.  The  total  depth  of the well was  also  recorded
and the  well immediately  recapped  until ready for concentration measurement.
                                        2-2

-------
    Product Lines
    i
/
Background Vapor Well
                                                 Vapor Well
                                              Vapor Recovery Line
                                                    Backfill
                         Vapor Well In
                      Product Line Backfill
                                                                      .Vapor Recovery
                                                                           Lines
                          Vapor Wei!
                        In Tank Backfill
                                                                   •Vent Lines
                                                                         VENTS
         Figure 2-1.   Sketch  of typical  underground  storage  tank site.
                                          2-3

-------
     The concentration was first  measured  with the TIP probe  held  two inches
below the top of  the  casing.   The probe was then  slowly  lowered  at a rate of
about two feet per second to within one foot of the bottom of the well and the
concentration recorded at that depth.  Concentration versus depth was recorded
in  rare  cases  where  unusual  stratification  was  observed.   In most cases,
concentration generally  increased  with  depth as anticipated and concentration
versus depth was not recorded.

     No attempt was made to  aspirate the sample from  the hose fitting on the
well head assembly due to the long pumping times anticipated to reach a stable
concentration level.  Previous  field experience has shown  that more than ten
well casing  volumes of  soil  vapor must  be removed to  achieve a stable pump-
discharge  reading  that   suitably approximates  a  downhole reading  obtained
without pumping (Kaman Tempo, March  1988).

2.3  TRACER TESTING
     The  two main criteria used  to  select sites  for  tracer  testing were low
background  concentration level  and  the  presence of at least  two vapor wells
within either the pipeline or tank  backfill.   Six of  the twelve sites tested
for  background  concentration qualified  as tracer  candidate  sites, and three
sites  (Sites #1,  #2, and  #3) were actually  tested.   Sites  #1  and  #2 are
non-instrumented  sites,   whereas  Site #3  is instrumented  with an aspirating
system  with a  central  control  panel manufactured by Soil  Sentry.  The  well.
configurations  at the sites are shown in Figures  2-2 and  2-3.

     Site #3 well head  assemblies differed somewhat from the  non-instrumented
sites.   The instrumented wells were constructed of one inch PVC  fitted with  a
slip-cap connected to a  filter and restricter valve,  which was  in  turn  fitted
with i-inch tubing to the aspirator pump and  detector in the  control  panel.
Well monument boxes at  Site  #3 were sealed with a silicone sealant.   In  gen-
eral,  these  well monuments  were  drier than  those  at  the  non-instrumented
sites.

      An  ideal  tracer compound  for  fuel  leak studies  should  have  transport
 properties identical  to  one of the  principal  fuel components  of environmental
concern (such as  benzene, toluene, or xylene), be non-reactive, non-hazardous,..

                                        2-4

-------
                                                   9-2
                                                                     Bfanham Lane
                                                                       Pump Island  ]
                                                                       Pump Island  ]
c
-s-
n>

ro
 i
ro
CO
rt-
n
o
-h

CO



CD



=«=
=tfc
ro
           O

           CO


           3=>


-------
ro

 7500 gallon     |
Underground  \/
 Diesel Tank
                                             2000 gallon
                                               Above-
                                               Ground
                                             Water Tank
                                                 Fuel Lines
                                                                                                       V-6
Concrete Pad
Covering Two
10,000 gallon
Diesel Tanks
                                                                                      V-5
                                                                        Hospital Boiler Room,
                                                                        Shops.Offices, etc.
                                   Fuel Lines
                                                                                              Doors  ', !     .
                                I
                                                                                                          Soil  Sentry
                                                                                                          Control Panel
                                                                                                          and Sensor
                                                                10     20    30     40     50 Feet
                                                 Figure 2-3.   Sketch of Site #3.

-------
readily available  and  inexpensive, and detectable with  field  portable equip-
ment.  For this study, the following compounds were used:  sulfur hexafluoride
(SF6), trichlorof1uoromethane  (F-ll),  and propane, each  of  which  satisfies a
unique set  of  the ideal criteria.   The transport properties  in soils of all
three tested tracers  were unknown at  the onset of this  study.  All  are-non-
reactive with  fuel components, comrnercially available  and reasonably afford-
able  in the needed quantities.  In  its  highest commercially available purity
(99.99 percent)  SFg  is the most  expensive of  the tracers  used.   Purities of
the  other  tracers used  were  greater  than 99.0  percent.   Although detection
limits varied  between  compounds,  all were detectable  with field instruments.
Propane was selected largely because it is already in common use in evaluating
soil  gas monitoring systems.

      Because of  significant disadvantages of propane  as a tracer,  the current
study attempts to evaluate the  two other compounds  as possible substitutes.
One  of propane's  drawbacks is  that  it is detectable  by most monitoring sys-
tems.  A tested  system's alarm threshold  level  must be  raised  after propane is
used  unless all  residual propane  is  evacuated from the  backfill after  testing.
It  is more  desirable to use a tracer  that does not degrade the effectiveness
of  the monitoring  system because  threshold adjustments  of this type can result
in  false  negative alarms.  In  addition,  propane is  a  hazardous  substance on
the basis of its flammability.

      A comparison of physical properties of the selected tracers  and compon-
ents  of  gasoline vapor is presented in Table 2-1.  All  three  tracers  would be
expected  to  diffuse  more rapidly  through  tank backfill  than  actual   fuel
components  on  the  basis  of  their higher  vapor  pressures  and  Henry's  Law
Constants   and  lower  boiling  points.   The  molecular  weight and  density  of
propane  approaches that of benzene,  whereas  F-ll and  SFg more closely approach
the density and molecular weight of xylenes.   The vapor pressures  and Henry's
Law Constants  for all  three  tracers are higher  than those  of fuel components
 by a  factor of about  100.   The  boiling  points  of  the  tracers are  generally
much lower than those of the fuel components, F-ll having the closest approxi-
mation.   Because  both  gasoline  and  F-ll are  liquids  at  the expected  soil
 temperatures between 20°C and 24°C, F-ll's transport is of special interest.
                                        2-7

-------
                                                Table 2-1.
                             Physical Properties of Selected Gases*



Compound
TRACERS
Propane(2)

F-11®(2)
(Trichlorofluor
Photo-
ionization
Potential
MW fEW1} bp.-C

44.11 11.07 -42.1

137.37 11.77 23.8
•omethane)
Henry's Law Constant
Vapor
Pressure
p. kPa

941

88.5
(20°)
Solu-
bility
S. 9/m3

62.4

1100
(20°)
Density

a/I

1.969
(calc. 20°)
6.133
(calc. 20°)
kPa m3/mol
Experi-
Calculated mental

71.6

11.05 81.20
(20°) (20°)

Recom-
mended

71.6±2.4

..

SF6(3)        146.05     >12.0     -63.8*     753(3)
(Sulfurhexafluoride)                'Sublimates  (-63.5°)

SELECTED COMPONENTS OF GASOLINE VAPOR
Sparingly    6.602
 in water      (3)
Benzene (2) 78.11 9.25
Toluene (2) 92.13 8.82
m-Xylene 106.2 8.56
o-Xylene 106.2 8.56
(2)
Afr.ffl 	 = 	 = 	
80.1 12.7 1780 3.487 0.557 0.562
(calc. 20°)
110.6 3.8 515 4.113 0.68 0.67
(calc. 20°)
139 1.1 162 4.741 0.721
(calc. 20°)
144.4 0.882 175 4.741 0.535
(calc. 20°)
1.200
0.550
±.025
0.670
±.035
0.700
±.1
0.500
±0.06
      * Properties at 25°C except where noted.

      1) From Photovac Technical Bulletin No. 11.

      2) From: Dev5tt,D.A., Evans.R.B, Jury,W.A., and Starks.T.H.. 1986, Soil Gas Sensing for
             Detection and Mapping  of Volatile Organics, Draft, EPA Cooprative Agreement
             CR 812189-01.

      3) From: Handbook of Chemistry and Physics, 1973-1974, 54th edition.Chemical Rubber
             Company Press, p B-143.

      4) ibid, p D-187.

      5) m-Xylene is the predominant xylene isomer in fuel products, whereas o-Xylene is of greater
             concern to health.

      6) From: Cummins, A.B.,and Given.LA., 1973, SME Mining Engineering Handbook, Society of
             Mining  Engineers, p 16-13.
                                                       2-8

-------
All three tracers and gasoline  vapor :have Densities greater than air.   SFg is
one of the heaviest known gases.

     Tracer  concentrations  were  measured  with a  Photovac  Model  10S70  gas
chromatograph with photoionization  detection (PID)  and a  Gowmac Model  20-600
gas analyzer with thermal conductivity  detection (TCD).   Both instruments are
described in  detail  in  Appendix D.   Table 2-2  shows the  relative  detector
sensitivity to  these compounds.  The  PID  detector is sensitive  to  propane,
F-ll  and  fuel  components,  whereas  the TCD detector  is  sensitive to  all  of
these compounds.  Therefore,  the tracer tests  with SFg  had to  be  run  before
the  propane  and  F-ll  tests,  in  order to  avoid shifts  in  background level
during the duration of the test.

      It should be noted that although the TIP  and 10S70  have 10.6 eV  detector
lamps,  propane,  whose  photoionization potential  is  11.0  eV,  is  reasonably
detectable  (see  Table  2-1).   F-ll,  which has  an  11.7 eV  photoionization
potential is much less detectable.  The  12.0 eV ionization  potential  of SFg
renders it undetectable by photoionization.

      The most sensitive detector  of choice  for the  fluorinated compounds F-ll
and  SF,  would  be an electron  capture  detector (ECD); however  this  detector
requires  a  degree of  temperature  control  and  a power  supply  not currently
available in portable equipment.  The  metal oxide semiconductor (MOS)  type of
detector  commonly used in  most continuously  monitoring  equipment, including
that  at Site #3,  is  sensitive to fuel components, propane and  F-ll, but not to
SF6.

      The  gas chromatograph  has  the  added  feature  of  reporting  tracer and
background  fuel  vapor  concentrations  as  separate  peaks.   Therefore,  any
possible  diurnal  changes  in fuel  vapor  background  levels do  not influence
tracer test results.

2.3.1 Field Procedures, Sites  #1 and #2
      Preliminary  tracer tests were  conducted simultaneously at noninstrumented
Sites #1 and #2 in order to gauge the volume of tracer and test  duration  that
                                        2-9

-------
             TABLE 2-2.  DETECTOR SENSITIVITY FOR SELECTED COMPOUNDS
       Criteria
MOS detector sensitivity
PID detector sensitivity
TCD detector sensitivity
ECD detector sensitivity
Flamiability Hazard
Propane
Fll
          BTX
   0

  Yes
+
+
u
0
0
+
+
No
No
 0

Yes
 Benzene, Toluene, Xylenes.

Explanation:

  0 = Not detectable
  + = Detectable
 -H- * Highly detectable
                                       2-10

-------
would be required for a full  scale  test.   Full  scale tests were not conducted
at these sites  because  of automobile traffic through  work areas during busi-
ness hours.  SF-6 and F-ll were used, however,  propane was not used so as not
to disturb the City's weekly monitoring of the sites.

     In general, tests  were  performed by placing measured volumes of tracers
into a designated source well while monitoring concentration for several hours
in observation  wells within  the  line or  tank  backfill.   In  order  to avoid
convection effects,  injection volumes were minimized  and  in no test was more
than one tracer  injected  at  a time.  Background  concentrations were measured
in the  observation  wells prior  to tracer  injection in order  to  establish a
concentration base line.

     At Site  #1, a  single  test was  performed  using  100  ml  of  liquid F-ll.
Well V-4 was  the source  well and  V-2 the  observation well.   Four readings,
including the background measurement were taken over a 115 hour interval.  The
tracer was poured from the supply cylinder  into a beaker and drawn into a 100
ml syringe.  These  preparations were  made  at a distance of 100 feet from the
test wells, in order to avoid surface contamination of the observation well.
The  syringe  was then connected  to a  10 foot  length  of  teflon  tubing.   The
tubing was lowered  into  the  source well  and the tracer injected opposite the
perforated section  of well  casing.   The  source well  was  then  quickly capped
and the steel monument covered tightly.

     In the observation well, measurements were made  with the Photovac 10S70
probe at a depth of  14 feet.   Using the internal pump  on the 10S70, two tubing
volumes (30 ml)  were removed to insure that the  sample was representative of
downhole  conditions rather  than  residual  ambient  air in  the  tubing.   The
sample was then  injected by  the  automated  probe method (Appendix D)  with an
injection volume of 1.0 ml  and gain setting of 50  to  500.  Immediately after
injection, the probe was  removed  from the well and  the PVC plug and monument
cover  replaced.   The observation  well  was uncapped  for  a  maximum  of four
minutes at a time.

     A  pipeline  and  a  tank  backfill  test  were  performed at Site .#2.   The
pipeline backfill  was tested  by  injecting 500 ml  of  liquid  F-ll, using the
                                      2-11

-------
same method as described for Site #1, into source well V-7.  Observations were
made  in  Well  V-8 for a  period  of approximately 24 hours.  Wells  V-7  and V-8
are  located in  the  vapor  recovery  line backfill.   Only two  measurements,
including the background measurements, were made with the 10S70 for each well.
Measurements were taken in the same way described for Site #1.

     The tank backfill at Site #2 was tested by injecting 10 liters of SFg gas
into well V-3 with  three measurements taken in observation wells  V-2  and V-4
over a 92 hour period.   The  SFg measurements required more sample preparation
in  comparison  to the F-ll  measurements, because  the  Gowmac gas  analyzer's
internal  pump was not strong  enough  to  draw a sample through  the total length
of  the sampling  probe.   Therefore,  samples  were collected in  5-layer bonded
sample bags using  an oilless vacuum  pump.   Between  each sample,  bags  were
flushed five times  with  ambient air and  filled to  capacity with  ambient air.
The ambient air sample was then used to zero the Gowmac instrument.  The probe
was then lowered  to within  one  foot of the  bottom  of each well  and connected
to the vacuum pump.   The sample bag was then evacuated, the vacuum pump intake
connected to the  probe,  and the  bag  filled  with sample gas.  The  sample bag
was  then  immediately  connected  to the instrument and  the  concentration
recorded.

2.3.2  Field Procedures, Site #3
     Before tracer  tests were  conducted  at Site  #3, all  wells were  suction
tested for  plugged  perforations.   An in-line vacuum  gauge was  placed  between
the well  head  fitting and  a vacuum  pump.   No  response  was  observed  on the
gauge during pumping, indicating good communication with the backfill.   Strong
vacuum readings  would  indicate  saturated  or  clayey soils  having very low
permeabilities to gas flow.   Had  strong vacuums been observed,  tracer testing
would not have been conducted at site #3.

     Two complete sets  of tracer tests  were  run,  Test A during  November 18
through November  20,  1987,  and  Test B  during November 30  through  December 2,
1987.  In both tests, Wells  V-l and V-4  were used  as source  wells, and Wells
V-2 and  V-5 used as  observation  wells.   Wells  V-3 and V-6 were  monitored as
                                      2-12

-------
observation wells only during  the  first test set of tests.  The  major  diffe-
rence  in  field conditions  between sets was  that the  site's  aspirating-type
automated monitoring  system, manufactured  by Soil  Sentry,  was  in  operation
during the  first  set.   Each well  was  aspirated  three times daily during  the
first  set,  for  7  minutes  at 1.8L/min,  with cycles starting  at  midnight,  8:00
a.m. and 4:00 p.m.  Therefore,  13  liters  of soil  gas were aspirated per cycle
at each well, or  5 to 20 well  casing volumes,  depending on depth.  The  second
set  of tests was  conducted to determine  if the site's  automated monitoring
system significantly affected the tracer test results.

     During Test  A, 500 ml  of liquid F-ll  and 7.5 liters  of propane gas  were
injected into Well V-4, whereas 5.6 liters  of  SFg and 3.75 liters of propane
were injected  into Well V-l.  During Test  B, 7.5 liters  of  SFg and 15  liters
of  propane  were injected  into Well V-4,  whereas  3.75 liters of  SFg  and  3.75
liters of propane were injected into Well V-l.

     As with  Sites  #1 and  #2,  tracers  were not  mixed in  any test due  to  the
large  injection volumes  required and positive interference  by  propane  to  SFg
detection.

     Test  A measurements  were  taken according to the methodologies described
for Site  #1.   During  Test  B dedicated  probes were placed in  each well for the
duration  of testing in order to minimize dilution by ambient surface air.  The
dedicated  probes  also avoided  cross contamination  between observation wells.
During both sets  of tests,  the  annular  space  between the well casing and probe
was sealed  with a one-hole stopper.   It was decided  not to use F-ll in Test  B
due to F-ll's  poor detectability as  shown  by  Test A.
                                       2-13

-------

-------
                                    SECTION 3
                                     RESULTS

3.1  BACKGROUND MEASUREMENT RESULTS
     Background  fuel  vapor  concentrations  were  measured  in  75  wells at  12
sites.  The  measured concentrations  are  reported with  well and  site obser-
vations in Table 3-1.

     Downhole concentrations, regarded  as  representative of the concentration
in the  backfill, range from  0  to 2,000  ppm,  with a  mean  value of  300  ppm.
Concentrations are reported by volume as isobutylene.

     Top of  casing  readings  proved to be poor indicators of true  soil condi-
tions  as  they  were greatly  influenced by  dilution  with  surface air.   For
example, concentrations  less than 10 ppm  were measured at  Site #2  in wells
with downhole concentrations  as  high as 1,600 ppm.   Conversely,  the  greatest
top  of casing  readings  were  obtained  in  wells giving  the  lowest  downhole
readings at that same site.

     Concentration  in  each  .well  generally increased  with  depth,  with  the
greatest concentration being measured  in the  bottom  one-quarter of  the  per-
forated casing.  The inverse  condition  was  encountered  only  in Wells  V-3, V-4
and V-5 at Site V-10.  These three wells were among the deepest encountered in
the  study.   The inverse  concentration gradient may  be  ascribable to  the
partitioning and stratification of individual fuel components having different
TIP detection sensitivities  within the  well casing.   Because the  well casing
is screened  throughout almost  its entire  length,  the  inverse concentration
gradient in  the  well  may also  represent  actual  stratification  of  backfill
contamination.

     Measurements were unobtainable or questionable at six wells due to either
inaccessibility, broken casings  or flooded  conditions.   Bentonite was encoun-
tered in several wells, indicating that the  sandpack  did not cover the entire
length of perforations.
                                       3-1

-------
                                      Tab!? 3-1.   Background  fuel vapor  concentrations at
                                                    underground tank sites.
CO

ro


Site Well
# Address No.
1 858 W. Branham 1
2
(Tracer test site) 3
4
5
6
7
2 4610 Pearl Ave 1
2
(Tracer test site) 3
4
5
6
7
8
3 2105 Forest Ave 1
2
(Tracer test site) 3
4
5
6
7
4 1501 Parkmoor 1
2
3
4
5


6

Weil
Depth
fm
8
15
8
15
5
5
5
8
18
18
15
5
18
5
5
3
13
4
14
14
14
14
8
14.5
5
15
15


5
TIP
(Top of
Casing)
(ppm) ,
1
1
10
3
4.5
2
1
10
2
118
8
160
0
3
6
0
0
0
0
0
0
0
1
1
3
5
9


1
TIP
at
Depth
fpprrrt
30
15
9
15
6
3
1.5
9
33
160
15
-
-
22
16
0
0
0
0
0
0
0
1
3
5
9
300
240
350
8
Depth of
Meas-
urement
ffn
7
12
7
12
4
4
4
7
17
17
14
4
17
4
4
2
12
3
13
13
13
13
7
13
4
14
13
12
10
4

Well
Diameter
(Inches)
1
2
1
2
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
2
2


1


Product
Type
G
G
WO
G
G
G
G
R
R
R
SU
-
-
-
.
D
D
D
D
D
D
D
G
G
G
G
G


G



Remarks Ml





Well on supply line.
Well on supoly line.


Well on vent line.
Well on vent line.
Well on supply line.
Well on supply line.
Well on supply line.
Well on vent line.
Well on supply line.

Well on supply line.




Background well.






Well on supplv line.


Well
Condition









Vent line leak history.

Wet bentonite in hole.
Wet bentonite in hole. Flooded.


All wells fitted with Soil-
Sentry/ Genelco aspirated
monitoring system instru-
mentation.











        (1)  Wells  are within tank backfill  unless otherwise stated.

-------
                                                 Table 3-1 (Continued).
co


Site Well
# Address No.
5 1708Tully 1
2
3
4
5
6 2690 Union Ave 1
2
3

4


5
6
7 3010 Almaden Expr. 1
2
3
4
5
6
8 1152Tully 1
2
3
4
5
6
7
8
9 1405 Branham Ln. 1
2

Well
Depth
fftt
14
14
14
-
.
8
4.5
4.5

14.5


14.5
8
14
15
15
8
5
5
6
15
5
5
5
-
15
15
8
8
TIP
(Top of
Casing)
(ppm)
23
400
40
-
30
0
0
20

4


0
0
2
0
0
0
0
0
425
10
10
4
160
-
1
0
107
9
TIP
at
Depth
(ppm)
441
1808
1835
-
.
0
0
0
6
87
33
18
0
0
0
0
0
0
0
0
2000
250
10
6
800
-
1370
1050
270
30
Depth of
Meas-
urement
rm
12
12
12
-
.
7
3
3
2
13
10
5
13
7
10
14
14
7
4
4
5
12
4
4
4
-
12
12
7
7

Well
Diameter
finches)
2
2
2
1
1
1
1
1

2


2
1
2
2
2
1
1
1
1
2
1
1
1
-
2
2
1
1


Product
Type
G
G
G
G
G
G
G
G

G


G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
WO
WO


Well
Remarks Condition



Well on supply line. PVC bushing stuck.
Well on supply line. Well flooded.
Background well.
Well on supply line. Pipe broken.
Well on supply line.






Concentrations at this site
verified w/ 40 ppm isobutyl-
ene standard.
Background well.
Well on supply line.
Well on supply line.


Pipe broken.
Well on supply line.
Well on supply line.
Steel lid stuck.





-------
                                                   Table 3-1 (Continued).
to


Site Well
# Address No.
10 4995 Almaden Expr. 1
2
3




4



5






6
7

11 1722 Meridian 1
2
3




4
5
6

Well
Depth
ifrt
8
15
15




15



15






5
8

8
15
14.5




15.5
5
5
TIP
(Top of
Casing)
(ppm)
82
110
950




116



225






23
10

2
4
0.6




Q
Q
56
TIP
at
Depth
(ppm)
176
400
600
700
750
900
500
780
800
800
240
48
62
85
105
120
158
130
15
0

4
6.5
60
85
105
60
6
0
0
110
Depth of
Meas-
urement
rm
7
14
13
12
9
6
3
12
10
9
5
13
12
11
10
9
6
3
4
7

7
14
13
12
10
7.5
5
14
4
4

Well
Diameter
(Inches)
1
1
1




1



2






1
1

1
1
2




1
1
1


Product
Type
G
G
G




G



G






G
G

G
G
G




G
G
G


Well
Remarks Condition
Background well.

Note inverse concentration
stratification. (500 ppm in
monument)


(40 ppm in monument)



Strong odor.
Note inverse concentration
stratification.




Well on supply line.
Well on supply line.
Ambient air =11.0 ppm.
Background Well.



Greatest reading 4 feet
from bottom of well.

No Stratification observed.
Well on supply line.
Well on supply line. 6" bentonite on bottom.

-------
CO
I
en
                                                Table 3-1  (Cont'd)


Site
# 	 Address
12 3147Senter








Well
No.
1
2
3
4
5
6
	 7_

Well
Depth
(ft)
15
15
5
15
5
14
7
TIP
(Top of
Casing)
(ppm)
1
0
4
0
10
1
2
TIP
at
Depth
Ippm)
1600
1300
8
1000
50
1500
16
Depth of
Meas-
urement
(ftt
12
12
4
12
4
12
6

Well
Diameter
(inches)
1
1
1
2
1
1
1


Product
Type Remarks
G
G
G Well on supply line.
G
G Well on supply line.
G
G


Well
Conditbn








-------
     Fuel throughput  data for each  site are not  available at this  writing,
however,  those  sites  having  the  greatest  background  concentrations  were
self-service stations with heavy business traffic during our site  visits.   The
lowest concentrations measured in  the  study were at Site #3, whose  tanks  are
only rarely used and certainly have the lowest throughput of sites visited.

     Leak history  information  is  also  incomplete at this time.   According to
the Site  #2 station operator, the highest background  concentration  was  mea-
sured in  Well  V-3 adjacent to  a  tank  having a  vent line  leak history.   The
lowest concentrations measured in the study were at Site #3, the newest of the
tank installations where a leak was highly unlikely.

     A comparison  of  concentrations measured  in  pipeline and tank backfill is
presented in  Figure 3-1.  Product, vent and  vapor  recovery pipeline backfill
were considered equally  in the comparison.  Most concentrations were less than
20  ppm  in both pipeline  and  tank backfill.  The  tank  backfill  concentration
data  exhibited  the  highest  overall  values   and most   uneven  frequency
distribution.

3.2 TRACER TEST  RESULTS
     Of  the three compounds  tested, propane yielded the most conclusive data.
Propane  was used  extensively at Site #3.   The greater effectiveness of propane
was largely  due  to  its  detectability  with  the  Photovac 10S70.   With the
methods  used  in this  study,  the  detection limits  for  propane,  SFg, and  F-ll'
were  0.01  ppmV,  0.5 percent  and 1 to  10 ppmV  respectively.   The  detection
limit for F-ll increases greatly with lowering temperature  below  65°F  (18°C).
Ambient  temperatures  varied  between  62°   and  72°F during the  tests.   As a
result,  F-ll vapor was never detected in  the backfill, including  measurements
in  source wells.

Sites  n and  12
      In the  preliminary tracer test at  Site #2,  SFg  was  observed  to  travel
 from Well  V-3 to  V-4 in less than  17 hours,  exhibiting  a 4%  concentration
 increase.  The distance between wells  was 15 feet.  There were not enough data
 to define tracer breakthrough in the other wells at Site #2.
                                        3-6

-------
    500-1000
                 PIPELINE BACKFILL DATA
     100-200
E
f
.o
1    50-100

1
o     20-50
O '
        0-20
            02468  1012141618202224262830
                           Frequency
                    TANK BACKFILL DATA
 E
 a
_Q.

I

1

I
5
            02468 1012141618202224262830
                           Frequency
Figure 3-1.  Frequency plots of concentration measured in pipeline and
           tank backfill.
                                3-7

-------
Site #3
     Conclusive tracer test data were obtained for SFg and propane in tests at
Site #3 (see Appendix E).  Time versus concentration plots for SFg and propane
for these tests are presented  in  Figures  3-2 and 3-3.   Selected chromatograms
from Site #3 tracer tests  are  presented in Figure 3-4.   A rise in SFg concen-
tration was observed in approximately two hours in both Well V-2 during Test A
and Well  V-5  during Test  B.   During  Test A, SFg was not  used at observation
well V-5/source well V-4.  No distinct concentration rise could be interpreted
from the Test B SFg well V-2 data.

     Over a distance of  15 feet, propane exhibited travel times between 24 and
30  hours  in all  tests  except  in  Test B  at  Well V-2.   This  propane data is
generally more  conclusive due  to  the increased  sampling  frequency  made pos-
sible by using the 10S70 in an  automatic cyclic sampling mode.

     The  right-hand  sides  of the field data  plots  have steep declining limbs
because the sources  were instantaneous.  Had a constant source been  used, the
declining  limb would  be  missing  and  concentration  would  remain relatively
constant  after breakthrough.   The  other ramifications of the use of instantan-
eous  versus constant  sources  upon data  interpretation are  discussed in the
following section.

     There  was no breakthrough in  Wells V-3 or V-6  from either SFg or propane.
Apparently  the density  of the tracer  gas  prevented  movement into V-3,  whose
total  depth is  12 feet above source well  V-4.  Lack of breakthrough  in  V-6  is
likely due  to  the tanks themselves, which may have  provided a  barrier to flow.
The tanks are installed parallel to a line drawn between  V-4  and  V-5.  There-
fore,  the flow path to V-6 from source well V-4 is  more tortuous  than to V-5.

3.3  GAS  DIFFUSION MODEL DEVELOPMENT AND FIELD TESTING
      Several  continuous point  source models,  such  as that of Weeks  and others
 (1982), have been tested.   However,  these models do not  satisfy  the boundary
 conditions imposed by  the subject field tests.  It  is  therefore  necessary  to
 develop a  new practical mathematical  model to predict the  concentration  of a
 gas in backfill  at any time and radius from an instantaneous  line  source  of
                                        3-8

-------
CO
 I
10
                             CQ
                             tz

                             n>

                             to

                             ro
 to
 TI
en

 o
 o
 3
 O
                              -J
                              OJ
                              O
                              rt
                              CO
                CO
                I

                O)
                                            ro
                                            o .

                                            ro .
d°"
3 co

3 o
                CO

                00
                IO
                o
                10
                to
                ro
                *.
                ro
                o>
                to
                oo
                09
                o
                                                      ro
                                                       i
                                  SF6 (%)
                                 *>     OJ
                                                                                                                SF6 Concentration (%)
                                              00

                                              I
                                                       a
                                                       DO
3!

I


Ul
O
«  s
   s
                 o
                  I

                 00
                  I

                 en
                  I

                 *.


                 10

                 O


                 lO


                 *>


                 01
                                                                                                     at

                                                                                                     CO
                                                                                                     ro
                                                                                                     o
                                                                                                     10
                                                                                                     10
                                                                                                     10
                                                                                                     CD
                                                                                                     10
                                                                                                     00
                                                                                                     CO
                                                                                                     o
           Geneico System aspirated

            7 min every 8 hrs.

           T=5.42 hrs
a
>

Tl
O)

2.

I
a
w

-------
                    Test A, Propane Well V-2 Data
  0.4'
  0.3-
ss
£
e
n.
  0.2-
  0.1 -
                      T=23.25 hrs
      Genelco System Aspirated
      7 min every 8 hrs.
      T=3.00
                               T	1	1	1	1	1	1	1	1	1	1	1	1	T
     -202468  10121416182022242628303234363840424446
                                Tlme(Hrs)
                     Test A, Propane Well V-5 Data
     •20246 810121416182022242628303234363840424446
                                 Tlme(hrs)
                       Test B, Propane Well V-5 Data	
>
2.3.0
§ 2.0
I
-•- Propane Concentration
* Time Probe Disconnected
x Time Probe Removed From Well
     -20  24  6 810121416182022242628303234363840424446
                               Time (hrs)

     Figure 3-3.  Time vs.  Propane concentration Plots.
                                   3-10

-------
   j» •  :•"
   -.  , r
                                                                      v • v -
                                                                        » 2
    I  a  j
       ' :n
       i 5*
WML-Sti *    .   rs-I 5C.-PL£a
rE""ESl»T'-FE   21   IBi'EC
G*If*       1000   "J'RSCP'VIK
     *  *
                                       ig
                                     i
  .* «  JJd.J
WV1Pt.E RUT*      OEC  1 1387  8:
IWW.TSI3 •    7  23-5 SWPLES
tE"PERl»rUfi£  30  iBSEC
(*:N       iaaa  OWBOPKK
                                                                       1 u
SIWLt RUN      DEC  1  138;   3- 6
KKOLTSIS *    3   23-S SWPLE3
TBrESuriJKE   31   IBSEC
3«1M      leee   CIWBOPKCK
                                                                  •JNKNCUM
                                                                  'JWNOUM

                                                                  •JNKNCi»<
                                               i   IS.' 392.= «oS
                                               3   3«.8 58 = «JS
                                               5   12.3 31.0 -^9


                                               3  163.3  7.« ..•:
 '••:.' <   '•SB.?
        <       OEC
    "SIS •     «   23-5
        IE   35   13SE
          iaaa
fl;t 1
t
j \ '-'=•
\ 5**
!h
i CTP
• •» ^ •••^a 31 _ .
u
: «
;5
> 16
« tf
a 21

V 9 ^?8.8 ^''C
!i
/P
5 il
« 23

~~
                  -J.i i3^.3 -JS
 VHTI.E RUN       PEC  I 1387  77:32   Sdnr-E RUN      PEC  1 138: ~n V
 «NM.TS;S <    3   2!-=. Sf
-------
known length and mass.  The model's development and application procedures are
described in Appendix F.  The  intent  of the model is to aid  in  the  design of
adequate well  spacings  and  detector sensitivities in tank  and pipeline  moni-
toring systems, and to test the efficacy of existing systems.

     A  line  source model  was required  for this  study  because of  the  great
thickness over which  the source wells  are  perforated.   The  main  drawback of
point source tests is the very low emission rate and concomitant low detection
limits  (high detector sensitivity) required.

     For additional  simplicity, it  was required  that  the source  be instan-
taneous.  In continuous  source tests,  controlling the source  release rate can
be a complex task.  However, an instantaneous source is easily approximated by
injecting a volume of tracer into a vapor well.

     By applying  the  instantaneous  line source model to the tracer concentra-
tion  versus  time data,  the  effective diffusion coefficient,  D', and  the
sorbtion corrected porosity,  A, are estimated.  These parameters characterize
both the tracer and particular backfill material.

     Average  values  of D1 and A for  propane calculated from  the test results
in  the pea-gravel backfill at Site #3  are 0.02  cm2/s  and 30% respectively.
The  field  data plots  for Site #3, Wells V-2 and V-5, showing  type curve match
points, are  presented in Figures 3-5,  3-6  and  3-7.   The two  curve fits  shown
in each field  data plot represent  the  range of  possible  solutions.

     The Test B well  V-5 test (Figure  3-7) yields the most  reliable solution
because of the plentiful data on the  rising  limb of the curve.  The model's
type curve only describes the rising  portion of field data.

     The  identification of rising-limb data during curve-fitting was greatly
aided  by referring  to the arithmetic  plots (Figure 3-3).   Because only  two
rising points  existed  for both wells  in  Test  A, off-plot  coordinates  using
zero concentrations  had  to  be used to  achieve  a solution.  The  V-2 solution
required  off-plot coordinates  (2.87,  0).   The V-5  solution  required off-plot
coordinates  (5.69, 0).
                                       3-12

-------
CO
I
            10
         0
O)
E
          10
          IU
        '2
     *
     O        0
          10  3
    10
             -4
               10
             0
                           E

                                   Cuve

                                                                 SOLLfTION
                                                         u«1Q2, V(u)«10-6,  Q= 77.5mg/cm
                                                   EiLl: r2/t=i8 cm2/sec
                                                        C*t«=0.01  mg sec/cm3
                                                        D' « 0.014 cm2/sec
                                                        A  * 0.043
                                                   Fit 2: r2/t = 20 cm2/sec
                                                        C*t • 0.03 mg sec/cm^
                                                        D' = 0.016 cm2/s*c
                                                        A « 0.013


                                                          10'
10'
                                        rz/t (cm2/sec)
            Figure 3-5.  Well V-2, Test A - Solution of D1 and A by the curve matching method.

-------
                        C*t (mg sec/cm  )
         _L  CO
         o
          o
              o
                I
               ro
                                                      o
                                                       ro
GO
I
cn
i
un
-H
n>
in
00
O

C
<-+

o
3

O
-h

O
01
=5
D.
o  —k
3  °
 io ro
          o
           CO
o

-5



S
ft-
o_


IQ

i

•
1

•
c

•

•

•
,
•
i
_—- -~


B

j








--"



\
\
-







••"V












i
\.
•o
o
o



> q c
••
„ H H
Pb'-
0 g I>
S <" =
o *2
1.?
« c
a i
o (


Ef B






E -
' To ?
: S <
> *• T K
* jk. " *i
i _ -«. —
, i ° i
'1!? -* i
r «
L w
* CO
3
o"
3
o
o.

-------
CO
i
en
O

O

 0.023 cm2/«ec
A a 0.191
Fit 2: r2/t « 0.32 cm2/sec
CM » 0.058 mg sec/cm3
D' * 0.026 cm2/««c
A x 0.359
um m
m


                 10
                 10
                                  0
                                   .10
                                  1
                                    r2/t  (cm2/sec)
10'
10'
             Figure 3-7. Well V-5, Test B - Solution of D1 and A by the curve matching method.

-------
     Once obtained, D1 and A can be used to predict the concentration of vapor
in backfill at any time in distance from a point source typical of a fuel leak
using a  point  source  model.   The relationship between  fuel  and tracer diffu-
sion parameters must first be known in order to apply the model to fuel leaks.
This relationship  would  be obtainable by  relatively  simply benchscale column
tests.  However, such tests were not within the scope of this project and were
not performed.
                                       3-16

-------
                                    SECTION 4
                                   CONCLUSIONS

1.   Vapor  wells  can  be  effectively  used  to  monitor  underground  product
     releases from both tanks and pipelines.

2.   Background concentrations are  commonly greater in tank backfill  than  in
     pipeline backfill.  Kaman Tempo believes that  this  is  ascribable  to tank
     overfill.  Such overfills are  common  and would affect  only the immediate
     vicinity of the tanks in moderate overfills.

3.   Because  stratification  of  soil vapor  within  tank backfill  is  possible,
     perforating the entire backfill thickness is desirable.  Shorter perfora-
     ted  intervals   might  cause  leakage   in  an  unscreened  stratum  to  be
     undetected.

4.   Propane is the best suited compound of those tested for use with portable
     gas detectors;  however, its flammability and interference with continuous
     site monitoring devices  are undesirable properties.   Therefore a surro-
     gate compound with the detectability of propane and a low flammability  is
     desired.

5.   The convective  flow  induced by the cyclic  aspiration  of  sample wells  in
     the normal operation  of the Genelco type system  tested was  not signifi-
     cant  enough  to  marketly  effect  the solution  of  diffusion  parameters
     calculated  from  the  time  versus  concentration  data.   The  arithmetic
     time/concentration plots  reveal  a  certain  amount of  intermittent noise
     that may be  associated with aspiration  of  the wells, however, where  an
     adequate  number  of data are  available to  filter out such  noise,  it  is
     revealed that cyclic  aspiration does  not affect the general  trend of the
     curve.

6.   Vapor wells in  shallow  pipeline  backfill may  not be effective in detect-
     ing tank leaks  even  when in close  proximity  to the tank backfill.  This
                                       4-1

-------
     conclusion is supported by the observations made in pipeline Well  V-3 and
     tank Well V-5  at Site #3.  Although Well  V-3 was closer to  source  well
     V-4 within the  tank backfill, no increase  in  concentration was observed
     in Well  V-3.   This  observation  may be  ascribed  to gas  density  effects
     where  denser-than-air  tracers  remain  within the  tank  backfill,  and
     therefore did not enter the pipeline backfill.  Conversely, when the same
     tracers were injected  in  source  well V-l in pipeline backfill, they were
     detected in tank backfill (well V-2) at a lower elevation.

7.   In multiple  tank installations,  shorter well spacings  are suggested in
     the  direction   transverse to  the  tank axis  than  in  the longitudinal
     direction due  to the  no-flow  boundary effects caused by  the tank walls
     themselves.

8.   The need  for additional  tracer testing  is  indicated whereby a relation-
     ship between the diffusion parameters of tracers and fuel products can be
     determined.  These  tests  can be performed in relatively simple benchscale
     soil column  tests.   Once this relationship has been defined, the diffu-
     sion parameters, D1  and A, for a  particular underground storage tank site
     can be  evaluated using surrogate compounds as tracers.   These diffusion
     parameters can  be used to aid the selection of well  spacings and detector
     sensitivities in tank  monitoring  systems.

9.   In order to  simplify  the procedures for field tracer  tests, it would be
     advantageous to evaluate a  surrogate  compound with sorptive  properties
     similar  to  fuel products. This would  minimize the need for the  relation
     described in Item 8.
                                        4-2

-------
                                    SECTION 5
                                   REFERENCES
1.   Freeze, R.  Allen,  and  J.A.  Cherry,  Groundwater,  pages 317-345,  1979,
     Prentice-Hall, Inc., Englewood Cliffs, New Jersey.

2.   "Evaluation of U-Tube Underground Tank Monitoring  Systems for Soil  Vapor
     Testing, Suffolk  County,  New York,"  Kaman  Tempo, March 1988,  Report  to
     Midwest Research Institute for EPA Prime Contract Number 68-01-7383.

3.   Kreamer, D.K, Personal Communication,  January 1988.

4.   Weeks, E.P. D.E.  Earp, and  G.M.  Thompson,  "Use  of Atmospheric Fluorocar-
     bons F-ll  and  F-12  to Determine the  Diffusion  Parameters of the Unsatu-
     rated Zone in  the Southern  High Plains of Texas,"  Water Resources  Res.,
     Volume 18, pages 1365-1378, 1982.

-------

-------
                  APPENDIX A

    SELECTED CITY OF SAN JOSE HAZARDOUS
MATERIALS DIVISION SITE INFORMATION SHEETS
                 A-l

-------

-------
                                                                                    ' t
Facility Address:
Facility Name:
File Number:
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                           SITE
                   rOv\
                                              K
Uv\Qgig.£L
                                                            T. 3
CONSTRUCTION INFORMATION
Number of Underground Tanks:
Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping
Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
	        Vadose or Groundwater Well

Number of Wells: 	[	

Well Number:                 1      2

Vadose or Water Well         \S
                                  _     -	
Depth of Well               fS""   \ ^~    %    $    *5~     £"
Depth of Groundwater	:	
Is Well in backfill?	
Soil Anal. (Installation)	
Soil Anal. Result (ppm)    	  	
Frequency of Monitoring:  Daily, Weekly, JtonthjVor Other (specify)
Brand Name of Monitoring Device:      L^>(^A p	
Principle of Detection:	
Person performing sampling: Owner or Consultant:
Periodic Calibrations?	
Alarm Hook-Up?	:	

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint  (ppm): 	
      Initial  Calibration Date:
      Calibration Chemical: 	
	        Interstitial  Space Monitoring

Frequency of Monitoring:   Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:	
Principle of Detection:	

-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                            SITE
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation

Piping
Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:
Facility Address: _
Facility Name:        fKLc 0	 Ht\? .2.3 N
-------
                                                                Page 2
MONITORING INFORMATION
             Vadose or Groundwater Well

Number of Wells:  	

Well Number;                 1      2345      67

Vadose or Water Well          ^     v    ^    ^   ^     ^    V
Depth of Well                (8   _&_ JJL J*	^-  -£- _®L
Depth of Groundwater	
Is Well in backfill?	
Soil Anal. (Installation)   T     Y	Y   Y    Y
Soil Anal. Result (ppm)    ^O   r\ o  ^Q  ^\ o  . T>  ^ \     	
Principle of Detection:	
Person performing sampling: Owner or Consultant:
Periodic Calibrations? 	
Alarm Hook-Up?	;	

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint (ppm): 	
      Initial Calibration Date:
      Calibration Chemical: 	
	        Interstitial Space Monitoring

Frequency of Monitoring:  Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:	
Principle of Detection:  	;	'.

-------
                           UNDERGROUND TANK MONITORING SYSTEM
                                   INFORMATION SHEET
Facility Address:
Facility Name:    _ OCTov^v>G.r  VW-&r\v ^~a.J»L     t*fr'?   it ( 3 •/£>£> "
File Number:      _ __

CONSTRUCTION INFORMATION
Number of Underground Tanks:
Tank                           1            2          - 3

Size (Gallons)
Material Stored              t>           T>          O
Construction Material      _     _
Single/Double Wall         _ _     _ _
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
_        Vadose or Groundwater Well

Number of Wells:       \  \ >r  w

Well' Number:        36_     1      2345      67

Vadose or Water Well  ^    ^   ___!_ _ ¥ _zL _^ __ ^L __±1
Depth of Well         ^  \Q    T£~ _M_ _U_ _V3 __ 1B_
Depth of Groundwater       _  ____
Is Well in backfill?       _  __ ; __  _
Soil Anal. (Installation)  _  ____
Soil Anal. Result  (ppm)    _  ____
 Frequency of Monitoring:  Daily, Weekly, Monthly or Other (specify)
 Brand Name of Monitoring Device:
 Principle of Detection:  _ _
 Person performing  sampling: Owner or Consultant:
 Periodic  Calibrations? 	
 Alarm Hook-Up? 	,	

 Vadose Well  Only
      Aspirated  or  Static:  	
      Alarm Setpoint  (ppm): 	
      Initial  Calibration Date:
      Calibt-ution Chemical: 	
.	        Interstitial  Space  Monitoring

 Frequency of Monitoring:   Daily, Weekly, Monthly or Other (specify)
 Brand Name of Monitoring  Device:		
 Principle of Detection:	

-------
                          UNDERGROUND TANK MONITORING  SYSTEM
                                  INFORMATION SHEET
                          svre

Facility Address:      ^O I
Facility Name:         UKO
File Number:      	
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
Number of Wells:
             Vadose or Groundwater Well

                        ^-^
Well Number;                 1      2345      67      8

Vadose or Water Well        ^    ^    ^    ^    ^   _A/	
Depth of Well               ^ ST    (C    \T  ( *T    S"  _§		
Depth of Groundwater		
Is Well in backfill?        ^    ^\      ^    \A     CA   	u_	  	
Soil Anal.  (Installation)     y	;	1_  	1	
Soil Anal. Result (ppm)	
Frequency of Monitoring:  Daily, Weekly^onthjpor Other  (specify)
Brand Name of Monitoring Device:       ^           ^
Principle of Detection:  _ _
Person performing sampling: Owner or Consultant:
Periodic Calibrations? _
Alarm Hook-Up? _

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint  (ppm): 	
      Initial Calibration Date:
      Calibration Chemical:
.	         Interstitial Space Monitoring

Frequency of  Monitoring:  Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:		
Principle of  Detection:	

-------
                           UNDERGROUND TANK MONITORING SYSTEM
                                   INFORMATION  SHEET
 Facility Address:
 Facility Name:
 File Number:
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
 \/         Vadose or Groundwater Well

Number of Wells: 	^	

Well Number;                 1      2345      67      8

Vadose or Water Well       	              	_ ___X   ^	
Depth of Well               \  C"    f !T   ^   C"   ^~  _§		
Depth of Groundwater	
Is Well in backfill?         O\    ^)     (/)   U   	La -^	
Soil Anal. (Installation)	^	
Soil Anal. Result (ppm)	  	
Frequency of Monitoring:  Daily, Weekly.(Konthiy or Other (specify)
Brand Name of Monitoring Device:     V— ^ f^
Principle of Detection:	
Person performing sampling: Owner or Consultant:
Periodic Calibrations? 	
Alarm Hook-Up? 	

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint  (ppm): 	
      Initial Calibration Date:
      Calibration Chemical:
	         Interstitial Space Monitoring

Frequency of  Monitoring:  Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:	
Principle of  Detection:	

-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                           SITE:
Facility Address:
Facility Name:
File Number:
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:.
Secondary Contained Piping? 	
Method of Secondary Containment:
Year of Installation:

-------
                                                                 Page 2
MONITORING INFORMATION
	        Vadose or Groundwater Well

Number of Wells: 	
      ^)o&L  -
Well Number:
Vadose or Water Well        V/.  ^    V	
Depth of Well               \5>    < !T    ft    *>
Depth of Groundwater	,	
Is Well in backfill?        VA    ^)    
-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
Facility Address:    3O \ O
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation

Piping
Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:
Facility Name:       OrxOCaJ?   l£>   _    MA?  23
File Number:      _ _   T*.f>c&L

CONSTRUCTION INFORMATION
Number of Underground Tanks:

Tank

-------
                                                                Page 2
MONITORING INFORMATION
	        Vadose or Groundwater Well

Number of Wells: 	
                •»     t
Well Number:                 1      2345

Vadose or Water Well        \/    V     V    w     V/
Depth of Well
Depth of Groundwater
Is Well in backfill?
Soil Anal.  (Installation)
Soil Anal. Result  (ppm)

Frequency of Monitoring:  Daily, Weekly, (M^ntMyjor Other (specify)
Brand Name of Monitoring Device:
Principle of Detection:  	
 Person  performing  sampling: Owner or Consultant:
 Periodic  Calibrations?	
 Alarm Hook-Up?	

 Vadose  Well  Only
      Aspirated  or Static:  	
      Alarm Setpoint  (ppm): 	
       Initial  Calibration Date:
       Calibration Chemical:
 	        Interstitial  Space Monitoring

 Frequency of Monitoring:   Daily, Weekly,  Monthly  or Other  (specify)
 Brand Name of Monitoring  Device:	
 Principle of Detection:	

-------
                           UNDERGROUND TANK MONITORING SYSTEM
                                   INFORMATION SHEET
 Facility Address:        \ t
 Facility Name:           U A * <:_ gg    T
 File Number:
 CONSTRUCTION  INFORMATION

 Number  of  Underground  Tanks:

 Tank
Size  (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping?
Method of Secondary Containment:
Year of Installation:

-------
                                                                                        I
                                                               Page 2
MONITORING INFORMATION
             Vadose or Groundwater Well

Number of Wells:  	
 Person  performing sampling: Owner or Consultant:
 Periodic  Calibrations?	
 Alarm Hook-Up?		

 Vadose  Well  Only
       Aspirated  or Static:  	
       Alarm Setpoint  (ppm): 	
       Initial  Calibration Date:
       Calibration Chemical: 	
.	        Interstitial  Space Monitoring

 Frequency of Monitoring:   Daily, Weekly, Monthly or Other (specify)
 Brand Name of Monitoring  Device:		_	_	
 Principle of Detection:   	.	_	—
Well Number:                 12345678

                             /                                     /     /      ^
Vadose or Water Well        _Y_  _J£_ __^_   ^   v    —*-	^-  -^—   '	'
Depth of Well               \ ^~   \ ^ —2—
Depth of Groundwater	
Is Well in backfill?       _^	  _S£	^,
Soil Anal.  (Installation)  	  	
Soil Anal. Result (ppm)    	  	

Frequency of Monitoring:  Daily,  Weekly^MontMyJr Other (specify)
Brand Name  of Monitoring Device:
Principle of Detection:

-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                            SITE

Facility Address:     f  A OST
Facility Name:         A~^C O
File Number:
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
  ^         Vadose or Groundwater Well

Number of Wells:	

Well Number:                 1      2
Vadose or Water Well        \/
Depth of Well               ^
Depth of Groundwater       _
Is Well in backfill?       _
Soil Anal. (Installation)
Soil Anal. Result (ppm)    <^(<->   *M C
Frequency of Monitoring:  Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:
Principle of Detection:	
Person performing sampling: Owner or Consultant:
Periodic Calibrations?	
Alarm Hook-Up? 	;	

Vadose Well Only
      Aspirated  or Static:  	
      Alarm Setpoint  (ppm): 	
      Initial  Calibration Date:
      Calibration Chemical:
              Interstitial  Space  Monitoring
 Frequency of Monitoring:   Daily, Weekly, Monthly or Other (specify)
 Brand  Name of Monitoring  Device:	
 Principle of Detection:		

-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                           	.  10
                    x7^a /o t—    l/\  \
Facility Address:
Facility Nair
File Number:
Facility Name:          Afl^ O
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page 2
MONITORING INFORMATION
  \J         Vadose or Groundwater Well
                         7
Number of Wells: _ / _

Hell Number;                 1      2

Vadose or Water Well         ^   _^_   _ __ _         _ _-
Depth of Well              _l£  _^_ _l£ j£_ j£l  _£L .!_
Depth of Groundwater       _  ____  ___
Is Well in backfill?       _  ____  __
Soil Anal. (Installation)    ^    ^  VA __ ^- J^ _  M
Soil Anal. Result (ppm)     ^a   <\ ^  c ( o 4 Q  
-------
                          UNDERGROUND TANK MONITORING SYSTEM
                                  INFORMATION SHEET
                          S\TE:
                 .. •»
I T 2- ^-  MS '~ i tfu^
Facility Address:
Facility Name:         frvlC Q
File Number:
CONSTRUCTION INFORMATION

Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation
Piping

Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                 Page 2
MONITORING INFORMATION
	        Vadose or Groundwater Well

Number of Wells: 	

Well Number:
Vadose or Water Well       __  __     _ _
Depth of Well               '3     ^   \ S"7   IS"    ST
Depth of Groundwater	
Is Well in backfill?	,	
Soil Anal. (Installation)  	  	  	
Soil Anal. Result (ppm)    	  	  	

Frequency of Monitoring:  Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:      *— T> f^V y	
Principle of Detection:	
Person performing sampling: Owner or Consultant:
Periodic Calibrations?	
Alarm Hook-Up?	;	

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint  (ppm): 	
      Initial Calibration Date:
      Calibration Chemical: 	
	        Interstitial  Space Monitoring

Frequency  of Monitoring:   Daily, Weekly, Monthly or Other (specify)
Brand Name of Monitoring Device:	
Principle  of Detection:	

-------
                           UNDERGROUND TANK MONITORING SYSTEM
                                   INFORMATION SHEET
Facility Address: 	^_             	
Facility Name:        fenj^C. Q	 nAv9  2 "3
File Number:      	

CONSTRUCTION INFORMATION
Number of Underground Tanks:

Tank
Size (Gallons)
Material Stored
Construction Material
Single/Double Wall
Year of Installation

Piping
Piping Construction Material:
Secondary Contained Piping? _
Method of Secondary Containment:
Year of Installation:

-------
                                                                Page  2
MONITORING INFORMATION
  \/         Vadose or Groundwater Well

                        -7
Number of Wells:  	'	
Well Number:
Vadose or Water Well                      _     _     _
Depth of Well                ST    b"    (^  1 \    f ^S"
Depth of Groundwater       _  ____
Is Well in backfill?       _  ____
Soil Anal. (Installation)   V     ^    Y    Y    Y
Soil Anal. Result (ppm)     ^Q   ^(c   ^O  ^o^ ^U   <[O_ _£U:
Frequency of Monitoring:  Daily,  Weekly,  Monthly or Other (specify)
Brand Name of Monitoring Device:   _ ^- S> t^\ V
Principle of Detection:  ___^
Person performing sampling: Owner or Consultant:
Periodic Calibrations? 	
Alarm Hook-Up?	;	

Vadose Well Only
      Aspirated or Static:  	
      Alarm Setpoint (ppm): 	
      Initial Calibration Date:
      Calibration Chemical: 	
	        Interstitial Space Monitoring

Frequency of Monitoring:  Daily, Weekly, Monthly or Other  (specify)
Brand Name of Monitoring Device:  	_____
Principle of Detection:	

-------
        APPENDIX B
LETTER OF AUTHORIZATION
           B-l

-------

-------
ROBERT E. OSBY
  Fire Chief
             CITY OF SAN JOSE, CALIFORNIA


             FOUR NORTH SECOND STREET, SUITE 1100                          SAN JOSE FIRE DEPARTMENT
             SAN JOSE, CALIFORNIA 95113-1305
             (408) 277-4444
                               November 10, 1987
     TO WHOM IT MAY CONCERN:


     The city of San Jose Fire  Department,  in cooperation with the
     United States Environmental  Protection Agency ("EPA"), is
     conducting a research study  underground storage tank vapor
     monitoring wells in the San  Jose area.  Mr. Bill Bergmann of
     Kaman Tempo, a consultant  to the EPA,  is performing this study at
     service stations and other underground storage tank locations
     such as yours which are equipped with  vapor monitoring wells in
     tank backfill.

     Please give your cooperation to  this study.  The information
     being gathered is for research purposes.  Please do not consider
     this an enforcement action.   Every effort will be made to
     complete the work in a rapid manner so as not to significantly
     interfere with your normal flow  of business.

     Should a question or problem arise related to this project,
     contact Joseph Afong of the  Hazardous  Materials Program at
     277-4659.
                                         sincerely,
                                        CHARLES J.  WILHELM, Manager
                                        Hazardous Materials Program
                                        San Jose Fire Department
    CJW:jwt
    CJW/SURVYLTR/1

-------

-------
       APPENDIX C
ADDITIONAL SITE SKETCHES
          C-l

-------

-------

\U7WA

                         £©'
©
                   -5 ©
                             A
                              "
                                   31IS
                              P3
                              U




                                                        m
                               31IS

-------
                              311S


\-
                     so
                               31IS

-------
     £
     o

     P>

                          T®
              0

                            C
->
                             6#  31IS
I
        r

        c

                        S © C ^wtrv )

                         H(7) (-
                             8# 31IS

-------
                     T©
                            Wcf/Wtf
                     c
r
%
             ©
                                                  I
                                   0 !-#  31IS

-------
0
2
-I
0

r

              A,  f\
              \J


              ^  (&
             or
                      X
                                                m

-------

-------
               APPENDIX D
VAPOR MONITORING EQUIPMENT SPECIFICATIONS
                  D-l

-------

-------
                                  APPENDIX.D
                   VAPOR MONITORING EQUIPMENT SPECIFICATIONS
TIP-I:
     The  TIP-I  is a device  that measures total  hydrocarbon  concentration in
air by  photoionization.   In operation, sample gas  is  pumped  at approximately
275 ml/min into a  10.6 eV  detection  cell.   Concentrations between 0 and 2,000
ppm by  volume are read from  a  digital LCD display.  The  intake  port is ter-
minated with  a swagelok fitting which simplifies attachment of sample probes,
and a 15 micron sintered dust filter.

Photovac 10S70 Gas Chromatograph;
     The 10S70 is a dual-column portable gas chromatograph with internal power
and carrier  gas  supplies   using  the  same detection  cell  as in the  TIP-I  de-
scribed above.  The  pre-column  is  backflushed  during analysis, allowing short
cycle times between samples.  A carbopack column was selected for this project
to adequately separate  propane  and F-ll.  Carrier  flow  rates  were maintained
at 10 ml/min.

     Sample  injections  are made  either by direct  syringe injections  into  a
septum  or  by an  internal  sampling loop that  is  loaded  by an  internal  pump.
The pump intake port is terminated with a 1/8" swagelok connector identical to
that  of the  TIP-I.   Loop  injection  volumes  are  controlled  by  varying  the
duration that the  sample loop is  in  series  with  the carrier gas flow into the
column.  Loop injections can  be made  automatically  at  any time cycle interval
enabling the  10S70 to be operated while unattended.

     Individual  sample components  are reported as  separate peaks  on a  strip-
chart recorder.    Concentrations are  proportional  to the  integrated area under
each peak, printed in volt-seconds (VS) by the internal  software  at the base
of each chromatogram.  By  calibration with  known standards, concentrations in
ppm can be  calculated  by  hand or  by the  instrument's  internal  circuitry.
Kaman has detected benzene at levels as low as 1 ppb with the 10S70.

-------
Gowmac Model 20-600 Gas Analyzer:
     The Model  20-600  is a device which  measures total concentration  of  any
gas that  has  a thermal  conductivity  difference from air.  The  4-inch  analog
readout dial has 0.2% divisions  on the 0  to  10% scale and  2%  divisions  on  the
0 to  100% scale.   Span  potentiometers  for the  two  scales  were  independently
calibrated  for use with  SFg in this project.   Readings  are susceptible  to
interferences  from  other gases  and do  not necessarily  represent  SFg concen-
tration.  However, for SFg tracer tests one may assume that changes in concen-
tration in a particular vapor well are due to SFg.

     The  Model  20-600 has an  internal  power supply,  200 ml/min  pump,  and a
1/8" swagelok  intake-port  termination like the  TIP-I and 10S70.   However,  the
pump is not strong enough  to maintain an  adequate flow rate through more than
5 feet of tubing.  Therefore, sample  bags must  be used.

-------
       APPENDIX E
SITE #3 TRACER TEST DATA
          E-l

-------

-------
Test A   Tracer Tests: Wells  V-2 and  V-5,  Site  #3
Hours
after source
initialized
Well #2 0.00
0.02
0.67
1.70
2.15
3.68
5.02
6.23
8.68
8.87
9.62
10.37
11.12
11.87
12.62
13.37
14.12
14.87
15.62
16.37
23.25
23.47
24.17
24.67
25.55
Hours
after source
initialized
Well #5 0.33
0.93
1.28
2.70
3.33
4.63
6.17
8.27
23.25
24.07
28.00
Concentration
(VS)
0
0
0.01
0.006
0
0.023
0.01
0
0.009
0
0
0.007
0
0
0
0.009
0.006
0
0.007
0
0.357
0.123
0.256
0.117
0.157
Concentration
(VS)
0.398
0.118
0.129
0.154
0.023
0.008
0
0
6.6
13.7
2.9
Propane
Concentration
trna/rr\\}
OE+00 PROPANE
OE+00
2.29E-07
1.38E-07
OE+00
5.27E-07
2.29E-07
OE+00
2.06E-07
OE+00
OE+00
1.60E-07
OE+00
OE+00
OE+00
2.06E-07
1.38E-07
OE+00
1.60E-07
OE+00
8.18E-06
2.82E-06
5.87E-06
2.68E-06
3.60E-06
Propane
Concentration
(mo/rnn
9.12E-06 PROPANE
2.70E-06
2.96E-06
3.53E-06
5.27E-07
1.83E-07
OE+00
OE+00
1.51E-04
3.14E-04
6.65E-05

-------
                Test  A   Tracer Tests: Wells  V-2 and  V-5,  Site #3
Well #2
Hours
after source
initialized
-0.28
0.74
1.89
6.17
6.25
26.00
SF6
Concentration
(%}
3.0
2.5
5.0
0.8
0.7
2.0
SF6

-------
Test  B   Tracer  Tests:   Wells V-2 and V-5, Site #3
11/30 Well#2 SF6
           Hours
         after source
 Date     initialized
                                     Injection Volume = 2  Bags = 3.6L
    Concentration
11/30
-0.1
1.1
1.6
1.9
4.2
11/30 Well #5 SF6

Hours
after source
Date initialized
11/30 -0.9
-0.1
0.8
1.0
1.6
2.0
2.4
4.5

12/1 Well #2 Propane
Hours
after source
Date initialized
12/1 -0.4
-0.3
3.6
12/1 4.6
12/2 16.2
16.6
17.9
18.4
19.4
20.9
21.9
22.6
23.7
24.4
25.0
25.8
26.3

7.4
1.6
5.2
4.5
65



Concentration
(%}
0.5
0.8
3.4
1.2
1.5
8.5
8.0
7.3


Propane
Concentration
(VS)
0.008
0.020
0.000
0.000
0.230
0.006
0.010
0.005
0.008
0.009
0.000
0.008
0.008
0.006
0.008
0.012
0.005
SF6 INJECTED INTO WELL 1 AT T=0.
70°F Temperature In hole T=0.08.



End

Injection Volume=4 BAGS=7.2 L



Temp 63°F T=0.






End
Injection Volume=2 bags = 3.6L Injected Into well 1.

Propane
Concentration
(ma/mh
1.83E-07 73°F at T= -0.67.
4.58E-07
OE+00
OE+00
5.27E-06 ?: Residual from W#5 rdg: See
1.38E-07 W#5 T=37.95 rdg.
2.29E-07
1.15E-07
1.83E-07
2.06E-07
OE+00
1.83E-07
1.83E-07
1.38E-07
1.83E-07
2.75E-07
1.15E-07

-------
Test  B   Tracer Tests:   Wells  V-2 and V-5, Site #3
Hours
after source
Date initialized
1 1/30 0.0
0.8
1.6
1.6
11/30 3.1
12/1 3.1
3.1
15.5
16.3
17.1
17.9
18.5
19.0
19.5
20.0
20.4
21.0
21.4
22.4
22.9
23.2
23.4
23.7
23.9
24.2
24.4
24.7
25.4
25.9
26.2
27.0
27.2
27.5
27.7
28.0
28.2
28.5
28.7
29.0
Injection Volume :
Source stre
Propane Propane
Concentration Concentration
(VS) (ma/ml) ...
1.300
1.300
1.200
0.790
0.550
0.010
0.000
0.030
0.120
0.150
0.120
0.290
0.220
0.270
0.290
0.321
0.370
0.310
0.600
0.770
0.880
0.990
1.200
1.000
1.000
0.910
2.900
0.820
1.400
1.400
0.570
1.000
1.200
1.600
1.800
1.600
1.600
2.000
2.200
2.98E-05
2.98E-05
2.75E-05
1.81E-05
1.26E-05
2.29E-07
OE+00
6.88E-07
2.75E-06
3.44E-06
2.75E-06
6.65E-06
5.04E-06
6.19E-06
6.65E-06
7.36E-06
8.48E-06
7.10E-06
1.38E-05
1.76E-05
2.02E-05
2.27E-05
2.75E-05
2.29E-05
2.29E-05
2.09E-05
6.65E-05
1.88E-05
3.21 E-05
3.21 E-05
1.31 E-05
2.29E-05
2.75E-05
3.67E-05
4.13E-05
3.67E-05
3.67E-05
4.58E-05
5.04E-05
                                                                Page 1 of 2

                                 8 bags*(1.8L/bag)*1969mg/L  = 28400mg
                    Source strength =  (28400mg)/14ft = 2029 mg/ft = 66.6mg/cm
                               Pumped 2 L from W5 to enhance flow
                                  atT=14.59.
                               Temp =64°F @ T=16.59
                                Hiatus in rdgs to check instr. flow
                                rates. Adjusted for longer loop
                                loading time after 1=24.7 .

-------
                  Test  B   Tracer  Tests:   Wells  V-2 and V-5,  Site  #3
11/30 Well 5 Pronane (continued)
Hours
after source
Date initialized
12/1 29.2
29.5
29.7
30.0
12/1 30.2
12/2 30.5
30.7
31.0
31.2
31.5
31.7
32.0
32.2
9 32.5
32.7
33.0
33.2
33.5
33.7
34.0
34.2
34.5
34.7
35.0
35.2
35.5
35.7
36.0
36.2
36.5
36.7
37.0
37.2
37.5
37.7
38.0
38.4
39.3
39.7
42.6
44.9
45.1
Propane
Concentration
/VS)
2.100
2.400
2.400
2.500
2.700
2.600
2.600
2.400
2.700
2.700
2.600
2.700
2.500
2.100
2.100
2.200
2.200
2.200
2.500
2.700
2.700
2.700
2.600
2.600
2.600
2.700
3.000
3.000
2.600
2.200
2.500
2.600
2.700
2.800
2.600
2.500
0.140
0.150
0.320
1.500
1.000
0.690
Propane
Concentration
(mg/ml)
4.81 E-05
5.50E-05
5.50E-05
5.73E-05
6.19E-05
5.96E-05
5.96E-05
5.50E-05
6.19E-05
6.19E-05
5.96E-05
6.19E-05
5.73E-05
4.81 E-05
4.81 E-05
5.04E-05
5.04E-05
5.04E-05
5.73E-05
6.19E-05
6.19E-05
6.19E-05
5.96E-05
5.96E-05
5.96E-05
6.19E-05
6.88E-05
6.88E-05
5.96E-05
5.04E-05
5.73E-05
5.96E-05
6.19E-05
6.42E-05
5.96E-05
5.73E-05
3.21 E-06
3.44E-06
7.33E-06
3.44E-05
2.29E-05
1.58E-05
                                                                                  Page 2 of 2
                                                 See W#2 T=16.18.
                                                 After sampling W 2. Dilution effect.
                                                END

-------
                  Test B   Tracer Tests:   Wells V-2 and V-5,  Site  #3
WELL #2 DISCONNECT TIMES
           Hours     Disconnected
        after source      Duration
         initialized        (hrsL
Remarks









WELL


DATE
12/1






12/2




0.0
3.7
12.6
13.0
14.1
14.6
15.6
16.7
18.1
#5 DISCONNECT
Hours
after source
initialized
14.6
16.6
17.9
21.5
24.7
25.4
26.2
38.0
38.5
39.8
40.9
42.8
3.67
1.00
0.45
1.17
0.10
0.50
1.00
1.33
1.48
TIMES
Disconnected
Duration
fhrs>
0.17
0.17
0.17
1
0.67
0.75
0.75
0.25
0.75
1.25
1.85
2.23
Probe not removed
Probe not removed
Probe not removed
Probe not removed
Probe removed
Probe not removed
Probe not removed
Probe not removed
Probe not removed
















WELL #5 Probe  REMOVAL  TIMES
           Hours       REMOVED
         after source     Duration
 DATE    initialized        fhrsl
12/1




14.6
16.6
17.9
24.7
40.9
0.17
0.17
0.17
0.67
1.85

-------
               APPENDIX F
CURVE MATCHING SOLUTION OF INSTANTANEOUS
     LINE SOURCE GAS DIFFUSION MODEL
                   F-l

-------

-------
                         Curve  Matching  Solution
           Of Instantaneous  Line-Source Gas  Flow  Model

Development of Model

A general  mathematical  diffusion model for  a sorbing gas  in 3-phase  porous
media has  been developed from Pick's 2nd Law  by Weeks  et al.(1982).  Pick's
2nd Law is stated mathematically  as
              rj2^«   -jx>
            i-> o O   oLf
           o^-TT                            (D

      where
           D  = the general coefficient of diffusion  of gas  into air  (L2/T)
           C  = the concentration of the gas (M/L3)
           x= the distance from the source (L)
           f= time (T).

In order to account for properties of  3-phase media, Eq(1) is modified to
      where
                  T= tortuosity  of  the medium
               fliAi,= gas-filled,  liquid-filled  and  total  porosities
               p, s = densities of the  liquid and solid phases  of medium
               Kw „ = partition ing coefficients for liquid and  solid phases.
Let,
which  is defined as  the sorption-corrected porosity (dimensionless).
Dividing both sides  of Eq  (2) by A  , the following diffusion  model is obtained:

-------
Solving Eq(4) for instantaneous, infinite line-source boundary conditions, we
have the solution
                              -
             c__   Q     Um                     (5) (Kreamer,1988)
                 4nD'At

      where
                   Q  = mass per unit length of instantaneous source (M/L)
                   D'  = effective diffusion coefficient  in porous  medium
                       (L2/T), defined by,

                                                   (6)
The solution in Eq(5) is  analogous to the Theis solution for groundwater flow
into a well,  where  groundwater parameters of drawdown,  transmissivity and
storage coefficient are analogous  to gas tracer parameters of concentration,
effective diffusion  coefficient  and sorption-corrected  porosity.  The vapor
well function e-u  , which we will abbreviate V(u) , is  analogous to the Theis
well function W(u).

By  further  simplification, the solution  is

            C = —Q7—V(u)                       (7)
                4nDAt                            *  '

      where

                       -                            (8)
Solution Procedure

By injecting a source of mass Q  into the source well and measuring the
resulting increase  in tracer concentration  with time in an  observation well,
D' and A  can be  evaluated by a curve-matching technique  similar to the
Theis solution method (Freeze &  Cherry, 1979).  A type curve is generated  by
plotting u  versusV(u)  on log-log paper (Figure F-l).  The field data are
plotted  on  a separate,identical sheet  of log-log paper.   The field plot  is then
superimposed upon the type curve, keeping the coordinate axes parallel, and
moved  until most of the data points fall on  the type curve (Figure F-2).  Using

-------
    10
    ID
       '2
 §10-3
    10
      '4
                         Type Curve
    •j Q~5 I    i—I I i ii ill—»_i—. .... ..II  	-I
       10
          "1
10
10
 u
              1
                                 10
10'
10
                            Field Data
                           1
                   10
10^

r2/t
                       10
                                                        10
                                                        ID
                                                           °
                                                           '1
                                                        10
                                                           ~3
10'
Figure F-l.  Type Curve and example  field data.

-------
    Type Curve .
,-1              *     -
                                      Field Data
1U '
10'2
10'3
io-4
10"5
10
-
-1
\




.10? 	

\
\
\ Mat
J
.1rf 	


ch Point: u=lC
r2/t

.10? 	


°, V(u)
=32,Ct=

/
•

=10-4 :
30 :
"
.10?.....'
10°
1C'1
10-25
ID'3
         10
102    u    103

      »       9
             rz/t
10'
10'
Figure F-2e Curve-matching procedure.  Maintaining parallel axes,
field data  overlay is shifted to  make the best visual  fit with the
type curve. A match point is selected and the coordinates  (u, V(u))
and (r /t»  C t) read from both sets of axes.

-------
f
 C
         an arbitrary type curve match-point with coordinates (u, V(u) ), the
         corresponding  coordinates  (r2/t, C*t) are read  from  the field data plot
         overlay.    For simplicity, type curve match point coordinates  with unit
         multiples of integral powers of ten (eg. 1*10-3,  1*10+6.  etc.) are preferred.

              Solving Eq(8) for D',

             ;      D'=4^r)                           w
              and solving Eq(7) for A ,
                    „_     0
         The match-point coordinates from both curves  are  subsituted into  Eq's(9) and
         (10), from which D'  and A are caculated.

         Once  parameters D' ,A , and source mass O are known, the concentration, C
         can be predicted in  tank or pipeline  backfill at any distance r from a line
         source at any time  /  after the injection.   It is  simply necessary to  calculate
         u  from  Eq(8), take  its  natural exponential, and calculate concentration, C ,
         from Eq(7).
        Freeze, R. Allen, & J.A. Cherry.  1979. Groundwater.  Prentice-Hall,Inc.,
        Englewood  Cliffs, N.J.,  pp317-345.

        Kreamer,  D.K.,January 1988, Personal Communication.

        Weeks.E.P., D.E.Earp, & G.M.Thompson.  1982. Use of atmospheric
        fluorocarbons F-11 and F-12 to determine  the  diffusion  parameters of the
        unsaturated zone in the southern high plains of Texas. Water Resources
        Res,  v.18, pp. 1365-1378.

-------

-------