&EPA
               United Stoles
               Environmental Protection
               Agency
              Environmental Monitoring
              SysternsLaboratoiy
              P O. Box 93478
              Las Vegas NV 89193-347S
  Pre4ssue Copy
  July 1989


EPA 600 -J 39 032
               Reseaich and jPevebpment
Sampling  Frequency
for  Ground-Water
Quality  Monitoring
V
\

-------

-------
   SAMPLING FREQUENCY FOR GROUND-WATER QUALITY MONITORING
                             by

Michael J. Barcelona, H. Allen Wehrmann, Michael R.  Schock,
            Mark E.  Sievers,  and Joseph R. Karny
                   Water  Survey  Division
    Illinois Department of Energy  and  Natural  Resources
               Champaign,  Illinois 61820-7495
           Cooperative Agreement  No.,  CR812165-02
                       Project  Officer

                        Jane  E.  Denne
            Advanced Monitoring Systems Division
        Environmental  Monitoring  Systems  Laboratory
                Las Vegas,  Nevada 89193-3478
      This  study was conducted in cooperation with the
                   Water  Survey Division,
    Illinois Department of Energy and  Natural  Resources
               Champaign,  Illinois 61820-7495
        ENVIRONMENTAL MONITORING  SYSTEMS  LABORATORY
             OFFICE OF RESEARCH AND DEVELOPMENT
            U.S.  ENVIRONMENTAL PROTECTION AGENCY
                LAS VEGAS,  NEVADA 89193-3478

-------

-------
                                    NOTICE


The  information  in  this document  has  been funded  wholly or in part by  the
United  States  Environmental  Protection  Agency  under cooperative  agreement
number CR812165-02  to the  University of Illinois.    It has  been subjected to
the  Agency's  peer and  administrative  review,  and  it has been  approved for
publication as an EPA  document.    Mention of trade names  or commercial
products does not constitute endorsement  or recommendation for use.
                                       11

-------

-------
                                  ABSTRACT
     The  primary goals  of  this project  were  to  collect  a benchmark
water-quality dataset  and evaluate  methods to optimize  sampling frequency  as
a  network  design  variable.    Ground water was  collected  biweekly  for
18  months from  twelve wells at two sites  in  a  shallow sand and  gravel
aquifer in  Illinois.    Sampling and analyses were conducted  for twenty-six
water  quality  and geochemical  constituents with careful quality  control
measures  to  allow  statistical analysis of variability  in  ground-water
quality  data. The  results demonstrate  that  natural  variability over time can
exceed the  variability introduced  into the data  from  sampling  and  analysis
procedures.  Natural  temporal  variability  and the  highly  autocorrelated
nature  of ground-water quality data seriously  complicate the selection of
optimal sampling  frequency  and  the  identification of  seasonal trends  in
ground-water  quality variables.   Quarterly sampling  frequency is a good
initial starting point for  ground-water  quality  monitoring network  design,
though  bimonthly   frequency   may  be  preferred  for   reactive   chemical
constituents.   Analysis of data collected  during this  project  suggests that
the  collection of a  long-term  (i.e., more than two years) dataset  is
necessary  to determine optimal sampling frequency and  to  identify  seasonal
trends in ground-water monitoring  results.    This report  was  submitted  in
fulfillment of Cooperative Agreement Number  CR812165-02 by the Water  Survey
Division of  the Illinois Department  of Energy and  Natural Resources under
the partial  sponsorship of the U.S.   Environmental Protection  Agency.   This
report  covers a period from May  1,   1985  to Sept.  30,   1988; and work was
completed  as of May  19,  1989.
                                     111

-------
IV

-------

-------
                                    CONTENTS
Abstract  	    i i i
Figures  	    vii
Tables  	     ix
Acknowledgments  	     xi

      1.   Introduction 	        1
             Variability in Ground-Water  Quality 	       3
                Hydrologic  Transience  in Ground-Water Systems  	      3
                   Infiltration  Recharge  	       3
                   Aquifer  Loading  	      4
                   Ground-Water Discharge 	       5
                Water  Quality Variability  and Network Design  	      6
      2.   Conclusions   	      8
      3.   Recommendat  ions  	     10
      4.   Experimental Design and Procedures 	      11
             Regional   Location  and Description  of  Field  Sites  	     11
                Regional  Geology  	     11
                Regional  Hydrology and Climate  	      14
                General  Site  Descriptions 	      16
                   Sand  Ridge State Forest  	      16
                   Beards town  	     17
             Field  Site Instrumentation   	      17
                Well  Construction Details 	      17
                   Monitoring Wells  	      18
                   Piezometers  	     20
                   Hydrologic  Monitoring  Systems 	      20
             Sand  Ridge  State  Forest  	      22
                Background  	     22
                Piezometer  Network  	     22
                Monitoring Wells  	      24
                Hydrologic Instrumentation  	      24
             Beardstown  	     24
                Background  	     24
                Piezometer Network  	      26
                Monitoring Wells  	      29
                Hydrologic Instrumentation  	      29
             Field  Activities  	     29
                Sampling Trip  Logistics   	      29
                Use of the Van  	     31
                Field  Parameters  and  Sampling  Protocols  	     31

-------

-------
                             CONTENTS (concluded)


            Laboratory Activities  and  Analytical Protocols 	      33
               Sample Preparation  	       33
               Sample Tracking 	       36
               Analytical  Protocols  	       38
               Reporting  and  Verification of Analytical Data  	      42
     5.   Results and Discussion 	       46
            Field  Activities 	       46
               Routine Activities   	       46
               Major  Difficulties   	       46
            Laboratory Activities   	       47
               Routine Activities   	       47
               Difficulties 	       47
            Data Quality  Evaluation  	       50
               Method Performance  	       50
               Sample Data Quality  	       62
            Characterization of Ground-Water Hydrology 	      65
               Sand Ridge  State Forest  	      65
               Beardstown  	       70
            Chemical  Data  Characteristics  	      81
            Statistical Structure and Temporal  Variability  	      88
               Estimation  of Sources of Variation 	      88
               Temporal Variations in Ground-Water Quality  	      91
               Sampling Frequency  	       94

References  	     10?
Appendices  	     117

     A.   Summary of  Analytical  Results  for Sampling Wells
             (Constituent  concentrations are expressed  in  mg.L"1
              except  as noted)   	    118
     B.   Time Series  of  Individual  Constituent Concentrations
            for Biweekly  Sampling Runs for Each Well  at the
            Sand Ridge Site and the Beardstown  Site 	     130
     C.   Ground-Water Elevations Measured  During Each Biweekly
            Sampling Run at the Sand Ridge and  Beardstown Sites
             (Elevations at the Sand Ridge  site  are in feet relative
            to  an arbitrary 1000-foot  reference  point.  Elevations at
            Beardstown are in feet relative to mean  sea  level)  	    I89
                                       vi

-------

-------
                                    FIGURES


Number                                                                     Page

   1    Location map  of  the  two field  sites at Sand Ridge  State  Forest
       and Beardstown,  Illinois  	    12

   2   Surficial geology  map of the Havana  Lowland  region  	    13

   3   Water  table  contour map of the Havana Lowland region (1960)  ....    13

   4   Monthly precipitation recorded  at Havana,  IL,  1983-1987  	    15

   5   Depth to ground  water recorded  at Snicarte,  IL,  1983-1987  	    15

   6   Typical  piezometer and  sampling well  construction  for  the
       Sand Ridge  and  Beardstown field sites 	     19

   7   Sampling well and piezometer  network  for the  Sand Ridge  field
       site  	    21

   8   General plan  of  the  Beardstown field  site  	     25

   9   Detailed view  of  the  sampling well and piezometer network  at
       the Beardstown  field  site  	    28

  10   Sample  handling  flow diagram  	     32

  11    Typical  set  of sample  bottle  labels  	     35

  12   Project  sample processing  scheme 	     37

  13   Example of  the  lead  sheet  of the sample tracking system 	    39

  14   Summary  of the  quality  assurance contribution to the  analytical
       workload 	    43

  15   Data handling flow chart  	     45

  16   Ion balance  error  summary  	    63

  17   Monthly precipitation recorded at  Havana  (a)  and depth  to  ground
       water in Well SR3 at  Sand Ridge field site (b)  for the period
        1983-1987  	    66
                                       vn

-------

-------
                              FIGURES (concluded)


Number                                                                     Page

  18   Monthly precipitation  recorded  at Sand  Ridge State Forest  (a)
       and relative ground-water elevations in Wells  D035  and DO 105  at
       Sand Ridge field  site (b)  during  the  field  sampling period
       (Elevations  in feet  relative to a 1000-foot datum)	    67

  19   Potentiometric surface  at Sand  Ridge field  site on  May 5,  1986
       (a) and August 24, 1987 (b)  (Elevations in  feet relative to a
       1000-foot  datum)   	    69

  20   Monthly precipitation  recorded  at Beardstown (a),  ground-water
       elevations  in  Wells BT23, BT30, and BT33  (b),  and ground-water
       elevations  in the  B8  piezometer nest  (c)  at Beardstown field
       site during the field  sampling  period  	    71

  21   Precipitation, barometric  pressure, and  ground-water elevation
       in  piezometer WLR2.1 recorded  at the Beardstown  field site  in
       March, 1988 	    72

  22   Upper, middle,  and  lower  potentiometric surfaces at the
       Beardstown field site on  April 21, 1987 and August 25,  1987  ....    73

  23   Equipotential  lines beneath  anaerobic impoundment  3 at  the
       Beardstown field  site  on April  21,  1987 (a) and  August 25,
       1987 (b)  	    76

  24   Calculated ground-water velocities for  four  areas  in  the
       vicinity of  anaerobic  impoundment 3 at  the  Beardstown  field
       site  	    80

  25   General  chemical  characteristics for the Sand  Ridge wells
       (a), the  upgradient  wells at  Beardstown (b), and the
       downgradient  wells at Beardstown (c)  	    82

  26   Profiles  of Eh,  dissolved  oxygen,  and  ferrous iron with depth at
       Sand Ridge field  site 	    84

  27   Average  concentrations  of  redox-active  chemical species with
       distance from contaminant  source (Concentration is on a
       logarithmic scale,  Eh is on a  linear scale and distances from
       source are schematic  and  not to scale)	    86
                                      Vlll

-------

-------
                                     TABLES


Number                                                                     Page

   1    Causes for  Temporal  Variations in  Ground-Water  Elevation  	     3

   2   Sand Ridge  Well Construction Details  	     23

   3   Beardstown  Well Construction Details  	     27

   4   Project Sampling Run Schedule  	     30

   5   Sample Handling,  Preservation  and Analysis According  to
       Parameter  	    34

   6   Analytical  Methods Used  in the Project 	     40

   7   Instrumentation Used for  Analytical Determinations  	    41

   8   Detection Limits for  Selected Inorganic Constituents Computed
       for Repetitive Analyses of Low  Calibration  Standards  	    51

   9   Summary of the Mean Accuracy and Precision (one  standard
       deviation)  of Field Standards  (Expressed in percent)  	    53

  10   Summary of the Mean Accuracy and Precision (one  standard
       deviation)  of Field Spikes (Expressed  in percent)  	    54

  11    Percent  Recoveries  and Relative Standard Deviations  of  External
       Quality  Control Standards (Concentrations  in  mg.L"1)  	    55

  12   Analytical Precision  Estimates from External Reference
       Quality  Control Standards  	     60
  13   Pooled  Analytical  Precision  Based  on Replicate Laboratory
       Analyses  of Field  Duplicate  Samples 	     61

  14   Vertical Hydraulic Gradients  at  the B8  Piezometer Nest  	    75

  15   Hydraulic Conductivities  (in  10  2  cm/s) at  Selected  Beardstown
       Wells  	    77

  16   Hydraulic Gradients  and Ground-Water Velocities at Beardstown  . .    79

  17   Mean  Saturation Indices for Selected Minerals
       in  Ground  Water from  the  Project Wells 	     87
                                        IX

-------

-------
                              TABLES (concluded)


Number                                                                    Page

  18   Percentage  of Variance  Attributable to  Laboratory Error, Field
       Error,  and Natural Variability by Chemical  and Site  	    90

  19   Observations  of Temporal Variations  in Ground-Water Quality:
       Short-Term  Variations 	    92

  20   Observations  of Temporal Variations  in Ground-Water Quality:
       Long-Term Variations  	    93

  21   Subjective  Estimate  of Strength of Seasonality  or  Trend in
       Variables by  Location 	    97

  22   Ranking of Average Lag  One  Correlation Over All Sites, From
       Smallest  to  Largest   	    98

  23   Sampling Intervals (in weeks)  for  Given Ratio  of  Effective to
       Independent Sample Size,  Based on the  Estimated  Lag One  Markov
       Model  	    99

  24   Estimated Ranges  of Sampling Frequency (in  months) to Maintain
       Information Loss  at <10%  for  Selected Types  of  Chemical
       Parameters  	    103

  25   Minimum Sampling Frequency  (in months) to  Estimate the  Mean  of
       the Base  Dataset  Within  10  Percent  	    105

-------

-------
                                ACKNOWLEDGMENTS
     The authors acknowledge the professional help and advice of Dr.  Dennis
Lettenmaier (University of Washington),  Greg George,  Carl Lonnquist,  Pamela
Beavers,  Ed Carske,  Sarah Smothers,  Osia Smith, Midge O'Brien,  Carolyn
Hohenboken and  Eleanor  Hopke.    Valuable  assistance  and  support  were
contributed to the effort  by Les McMillion, Robert  Snelling, Jane Denne,
Joseph DLugosz,  and Ann  Pitchford of the USEPA-EMSL,  Las Vegas. We are also
grateful  to  Ken Hlinka, John Helfrich,  John Brother, and Linda Riggin for
their  help  over  the course of the project.    The collaboration of the
Illinois Department of Conservation and the plant  proprietors  at  Beardstown
is very  much appreciated.
                                     XI

-------

-------
                                  SECTION  1

                                 INTRODUCTION
     This  study  was designed to  address  the  effects  of  temporal variability
on the reliability  of ground-water guality monitoring results.   it was  ini-
tiated in  May of  1985  as a  logical  extension of  previous research which
established a basis for the identification and control of  sampling  error  in
ground-water  monitoring  network  designs.

     The  primary  goals  of  the  project   were   to collect a  benchmark
water-quality dataset and to evaluate methods to  optimize sampling frequency
as a  network design  variable.     The density of sampling  points and the
frequency of  sampling are the principal  cost multipliers in  the  design and
operation of  monitoring  networks.    The  optimization of  sampling frequency
can reduce the  costs  involved in monitoring  network  operation without
affecting the information  return.

     There  are two  principal  sources  of  variability in  ground-water quality
data,   "natural"  variability  and  variability resultant from  the  network
design and operation.   The components of  "natural" variability  arise  from
temporal or  spatial variability related  to hydrologic  processes such  as
pumpage,  recharge or discharge,   as well  as  influences of these  processes  on
the release  and distribution of chemical  constituents  from a variety  of
chemical  sources.    The sources may be natural mineral assemblages,  precipi-
tation and percolation through the  unsaturated zone, in  addition to  numerous
point and non-point sources of chemical contaminants.   In general,  "natural"
sources of variability cannot be controlled  although they may be  quantified
through effective monitoring  network  design.

     Water-quality  data  variability  may also arise from the sampling and
analytical  components of  monitoring  network  design.    Sampling variability
includes  variations due to the  selection of the  locations  and  construction
of sampling points  in apace,  sampling  frequency,  well purging and  the execu-
tion  of the sampling protocol.   The sampling  protocol  consists of  the proce-
dures used to collect, handle,  preserve,  and transport water samples  to the
analytical  laboratory.     Elements of  the sampling protocol  have been  evalu-
ated   for   their   relative  contributions  to  variability  or   errors  in
water-quality  data  in previous research  (1-6).

     Analytical  variability in water-quality data  arises principally  from
the errors  involved in analytical methods and the  subsequent data  processing
steps.    These errors can be  controlled once suitable  water-quality indi-
cators or  chemical constituents  have  been  selected and  a thorough data
quality assurance/quality  control  program  has been  designed and  executed.

-------
     This study  was planned to control the sources of variability in water
quality data which  result from  network design components  such as sampling
location,  frequency,  sampling methods and analytical procedures.   The sam-
pling  frequency  was held  constant at  a  biweekly interval between  sample
collection  dates.   The benchmark  dataset  that resulted  from this  experimen-
tal design could then  be analyzed  to  determine  the optimal  sampling fre-
quency for selected water-quality  variables  at  both uncontaminated and
contaminated  study sites.   The objectives  of this analysis are enumerated
below:

     1.  Establish the degree of natural  variability  in  both
         physical and  chemical  parameters  under known hydrologic
         conditions.

     2.  Determine the  statistical reliability    of annual water
         quality  trends derived  from biweekly, monthly,  bimonthly,
         and  quarterly data  sets.

     3.  Assess  the impact  of  natural variability,  serial correla-
         tion  (persistence) and  seasonal  trends  on  the  sensitivity
         of simple  parametric statistical tests  to  detect water
         quality  changes.

     4.  Identify optimum sampling  frequencies for  annual mean and
         minimum  detection  sensitivity for  selected conservative
         and  reactive  chemical  constituents  under  known hydrologic
         conditions.

     5.  Analyze available  water quality  data  from  other studies
         which  may allow the extension of  the conclusions  to  other
         hydrogeologic   situations.

     6.  Explore the,  effect which the geochemical  variability of
         ground  water in unconfined aquifers has  on the utility of
         chemical   speciation   models   for   the   selection  of
         waste-specific indicator parameters.

     Optimization of sampling frequency in ground-water quality  monitoring
networks should provide  sufficient sensitivity  for chemical constituent
detection  and  adequate  characterization of average chemical conditions.  This
should  be  accomplished with a minimum  number of sampling dates.   It was not
within  the  scope of this work  to  provide  an integrated discussion of the
hydrologic and  chemical processes  which  give rise to natural  variability  in
the benchmark dataset.   Some introductory  and background  material has  been
presented to place  the project results  in  proper perspective.

-------
VARIABILITY IN GROUND-WATER QUALITY

Hydrologic Transience in Ground-Water Systems

     The  impact of the sources  of variability mentioned above will be influ-
enced  by the hydrology  of the  ground-water  system.   It is important  to
understand  that although aquifer  hydraulic  properties may not vary  signifi-
cantly  at a single  measurement point over time,   spatial variability may be
substantial.   This is  a very  active area of research  with application  to
monitoring network design  (7,8,9,10,11).

     Temporal and spatial  variations in ground-water elevation  may affect
ground-water  flow  rate and the direction of movement.  Such  changes may
influence the  quality of the ground water in the  vicinity  of  a  sampled well
by directing water  from a  different  upgradient  area or changing the velocity
with which dissolved  constituents move along  a  flow path. Examples abound  in
the literature detailing ground-water response (i.e.,  elevation  change) to  a
wide variety  of influences.   In addition to seasonal  fluctuations produced
in response  to  short-term (i.e.,  months to one  year)  events,  ground-water
levels  also reflect changes in long-term  (i.e.,  years to decades) condi-
tions.   Table I presents a  number of  natural and  artificial (man-induced)
influences which can cause changes in  ground-water elevation (the paren-
thetical  sign  denotes whether the water table  would  be expected to  rise  or
fall due  to the  listed  cause).
      TABLE 1.   CAUSES FOR TEMPORAL VARIATIONS IN GROUND-WATER ELEVATION


     Infiltration/recharge  (+)
          Natural --  rainfall,  snowmelt, flood (bank  storage)
          Artificial --  pipe or tank leakage, injection, irrigation
     Aquifer loading  on  confined  & semi-confined aquifer  systems  (+ or -)
          Barometricchanges -- air pressure changes
          Concentrated  loads  -- trains,  automobiles,  etc.
          Other  external loads --  earthquakes, tides
     Discharge (-)
          Natural --  evaporation,  transpiration, surface  water interaction
          Artificial  -- pumpage
Infiltration/Recharge

     Atmospheric  water Palling to the  earth's surface in its  various forms
is  the principal source of  ground-water  recharge.    Numerous studies have
investigated the  mechanisms  of  natural recharge  (12,13,14,15,16,17)  and
recharge   rates  through  various   geologic  materials  (18,191.     Factors
affecting  recharge include:   1) the  character and  thickness of  the  soil and
other deposits  above  and below the  water table,   2)  topography,  3)  vegetal
cover,  4) land use,   5) antecedent  soil moisture  content,  6)  depth to the
water  table, 7) the  intensity,  duration and seasonal distribution  of  rain-

-------
fall,  8)  the form of precipitation (i.e.,  snow,  rain, sleet,  etc.),  and
9) the air temperature, wind velocity, humidity,  and other  meteorologic and
climatologic factors (18,20).

     In  most  areas  of  the  U.S.,  particularly the  subhumid Midwest,  natural,
shallow  ground-water recharge generally occurs during limited times  of the
year, commonly in  the  spring  and late  fall.   During  these  periods, rainfall
tends to be of the  long-duration, low intensity  form.   Cool  weather and lack
of vegetative  growth are conducive to low  evapotranspiration rates  and  con-
tribute to greater  antecedent soil moisture  conditions thus enhancing the
potential  for  recharge to  occur.   Alternatively,   high evapotranspiration
rates caused by  growing  plants  and hot weather  during  the  summer months  may
eliminate ground-water recharge.   During the  winter,   when the ground  is
frozen,  recharge is  almost  nonexistent.

     Artificial  recharge  may  occur in  any location where man returns water
to the subsurface  environment.    Methods  of artificial recharge  can be
generally divided into  three  categories:   1) land  application  (e.g., spray
irrigation, land flooding, surface  impoundments),   2) subsurface  percolation
(e.g.,    septic  system   drainfields,    cesspools,    leaking  sewers),   and
3)   subsurface injection  (e.g.,   injection   wells).     Recharge   can   be
intentional  as  with the case  of  recharge pits and injection  wells  (21,22,23)
or inadvertent as  with  leaking sewers,  and waste treatment  lagoons  (24,25).
However the  recharge is  realized,   local   hydraulic gradients  will  be
affected.   Suter and  Harmeson (21)  documented  rises from 5 to 15  feet (1.5
to 4.5 m) in  wells  near recharge pits  along the  Illinois River  in Peoria.
Morton et al.  (25)  noticed  a rise of  nearly 8 feet (2.4 m)  beneath  a septic
waste impoundment  during  its  early  stages of operation.  A  number of papers
have described  mathematically the growth of ground-water  mounds beneath
areas where water has  been applied to the land  surface (26,27,28,29).

Aquifer  Loading

     Water levels   in  artesian  wells,   and  to  a   lesser   degree  in  some
water-table wells,  have been shown to respond  to  changes in  atmospheric
pressure and  other external loads  (30,31).    Such responses  are due  to
changes  in fluid pressure  and stress in the  aquifer matrix.   Roberts and
Romine (32)  observed a change of over  0.5 foot  (0.15 m) in  the piezometric
surface  of a well in  central  Illinois  as the  barometric  pressure increased
about one foot (0.3 m) for a  barometric  efficiency of approximately  0.5.
Barometric  efficiencies  generally have been found to  fall in  the  range from
0.20 to  0.75  (33).    Atmospheric effects have been observed to  a lesser
degree in water table  aquifers and have been attributed to  the presence  of
entrapped air  (34,35).   Atmospheric pressure  changes  caused water table
fluctuations from  0.6  to  2.4 inches (0.02  to 0.06  m) in  a fine-grained
aquifer in Utah  (36).

     Water  level fluctuations in  wells in confined aquifer systems  have been
attributed  to a variety  of other  causes  including concentrated loads such as
trains  (32,37,38) and  shocks caused  by  earthquakes  (39,40,41).  Again, water
level response  is related  to  changes  in fluid  pressure and stress in the
aquifer  matrix.   As might be  expected, the magnitude  of  water level response

-------
is a  function  of the  magnitude  of  the load (or shock),  the elastic proper-
ties  of  the  aquifer,  and the degree  of  confinement  (41).  Roberts  and Romine
(32) reported  short-term  fluctuations up  to  0.06  (0.02  m) foot  in response
to a  passing  train  in  central  Illinois.  Jacob  (38)  reported  a similar
response to a passing train on  Long Island.

     Scott  and Render (41)  discussed the  water level  fluctuations  in several
wells in Canada  due to the 1964 earthquake in Alaska.    The  earthquake
registered 8.5 on  the Richter  scale  and caused  water  level fluctuations
throughout the  conterminous  United States  as well  as Hawaii.  For the  wells
in confined aquifers  that  they  discussed,   water  levels  fluctuated  from
approximately  0.5 to 0.8  foot  (0.15 to 0.24  m) as the shock wave passed.
Residual  increased water levels  were on the order  of  0.1  foot  (0.03  m);
these residuals dissipated or  were  assimilated into the ground-water  trend
within  several  hours to days.  A  hydrograph  of water  levels  in  a well  in an
unconfined  aquifer  showed an  initial increase  of 0.2  foot (0.06  m) followed
by  a decline  of  1.1  feet  (0.33  m).   It  took  4  days  to recover  to  its
pre-shock  level.     This  dramatic change  was attributed  to  an increase in
ground-water  discharge due  to the initial shock followed by a  reduction of
water in  storage.    Thomas  (40)  found similar  responses  (up to  0.87  feet,
0.26 m) in  wells  in Utah, California,  and New Mexico to  two earthquakes in
Alaska and  Chile.

Ground-Water Discharge

     Diurnal variations  in  water  table  elevation have been  well documented
(31,42,43).   Most  of  these fluctuations have been  attributed to evapotrans-
piration losses, but  some  fluctuations may also  be  due to daily  atmospheric
pressure changes  (36).   Diurnal  fluctuations appear as small "waves"  upon
the regional ground-  water  trend.    The  change  in water  level  occurs as
ground-water storage is  reduced  to meet  the water  needs of plants and to
satisfy soil  evaporation requirements.     Drawdown  generally is greatest
during  the  daylight hours;  the maximum rate of drawdown occurs near midday.
The  magnitude  of  the fluctuation  is  relatively  small and depends  upon
weather conditions  and plant  water requirements.  Fluctuations are generally
greater  during the  height of the  growing  season (July)  than they  are  later
in the  season  (August).  Water level fluctuations  due to  evapotranspiration
are  essentially negligible after   killing   frosts.  Typical  daily  fluctuations
documented  in investigations  conducted in  the United  States  (31), Canada
(42), and  Sweden  (43) were on  the order of 0.1 foot  (0.03 m).

     In  areas  of ground-water  withdrawals,  ground-water  level changes caused
by pumpage  are superimposed on  seasonal  and long-term  fluctuations produced
by natural  ground-water recharge and  discharge.    When a well is pumped,
water levels decline,  forming a  characteristic cone  of depression.    The
shape and depth  of this cone depend on  the  amount, rate,  and duration of
withdrawal,   the  hydraulic characteristics  of the  well structure, and the
hydraulic  properties of the aquifer.   Water level declines are directly
proportional  to pumpage  and inversely  proportional  to the distance from the
withdrawal point.   The  hydraulic  properties  of an aquifer  remain  essentially
unchanged  over time; however,  pumpage normally exhibits  seasonal  variations
in response  to  water supply demands.

-------
Water Quality Variability  and Network Design

     Ground-water quality monitoring networks  are  designed for  a  number  of
purposes,   including   ambient  resource  studies,  contaminant detection and
assessment,  Contaminant  source  evaluation, litigation,  and research investi-
gations.   The  effective design of virtually  any such network,  regardless  of
purpose,  depends on  knowledge  of the  hydrogeologic system of interest, an
indication  of the presumed  contaminants  or preferred  water quality indica-
tors and  an assessment  of the  relative contributions  of  sources  of  varia-
bility.   These  aspects  of  monitoring network design have  been addressed
previously  in  the  literature  (44,45,46,47,48,49).    The common recommenda-
tions  in  these  works  are  that  background information must be supplemented
with  the results  of  a preliminary  sampling experiment to progressively
refine the  network  design to  account for error and variability in the
chemical results.

     Variability   in  the  analytical   results  for  particular ground-water
chemical   constituents   may  arise   due    to   "natural"   causes    such as
nonhomogenous  spatial distributions of the  constituents and  temporal  varia-
bility in   recharge.     Variability   may  also   arise   due   to   network
design-related  variables such as well  design,  sampling devices  and  sampling
protocols.   The  apparent sources of variability in water-quality data which
are often attributed  to  natural (i.e.,  temporal and spatial)  effects include
hydrologic   transience, the  fluctuations in  contaminant source  strength and
composition and  the   interactions between  reactive  chemical,  biochemical and
mineral constituents in recharge  water  and  ground  water.   Our  understanding
of the interdependence of hydrologic,  biological and chemical  processes  in
the subsurface is limited. However,  it may not be necessary to fully  under-
stand  the relationship between these processes,  contaminant  sources and the
resultant  chemical distributions  in order to monitor potential  contaminant
releases.

     The temporal and spatial  variability which is  observed in water quality
results over time at  discrete monitoring points  is the  result of the
processes  noted  in the preceding discussion  as  well as  the  sample collection
and  measurement  errors  inherent  to  network  design and  operation.    This
variability,  or  "noise",  in the  data embodies  the stochastic distribution  of
possible values  for particular  chemical  constituents  and the effects of  both
determinate  (i.e.,  systematic)  and  indeterminate  (i.e.,  random)  error.
Determinate error can be  measured  as  inaccuracy or  bias  if the "true  value"
is known.    Indeterminate error can be  estimated as imprecision or irrepro-
ducibility if a  sufficient  number of replicate determinations  can be made  to
faithfully  estimate the mean  or  the "true"  value.   In practice, determinate
errors can only be  estimated  and controlled by  careful  quality assur-
ance/quality  control  measures  exercised  over  appropriate  sampling   and
analytical  procedures  because  the  true value in environmental  distributions
is  unknown  and some disturbance of the subsurface  is  inevitable  in
ground-water quality work.  Identifying  and  controlling these design-related
errors  have been the  focus  of  much  of  our  recent  research (1,3,50,51,52).

-------
     Statistical   measures  of  short- term  temporal   variability   include
seasonal effects (e.g.,   consequences of  recharge  or  temperature  effects)
which can be  assigned to the seasons of the  year,  periodic  effects (e.g.,
consequences  of  anthropogenic contaminant sources  or pumping effects)  and
serial correlation  or dependence effects  which tend to  make data points
following maxima or  minima  in temporal  data series higher or  lower,  respec-
tively,  than one would attribute to random processes  alone. Trends  in data,
on the  other hand,   are  long-term variations  compared  to those which  may
occur  within  a  hydrologic  year (53).    This  categorization of temporal
effects  is somewhat  artificial  in that  the  combination  of seasonal,  periodic
or correlative  components  may result  in  a water-quality  time  series which
cannot  be  differentiated  quantitatively.    It  must  be recognized  then,  that
the  identification of short-  or long-  term trends in  water  quality is condi-
tional on some knowledge of the  proximity of the  sampling location to the
location and time  of chemical release as  well  as  the statistical  character-
istics of ground-water quality variables.

     Statistical   measures  of temporal   variability have  been reviewed
recently by Loftis et  al.  (54), Montgomery et  al.  (55) and  Harris et al.
(56).    They cite numerous examples of both short-  and long-  term temporal
variability which supplement  the earlier reviews  of Porter  and  Trautman  (53)
and  Colchin  et al. (57).

     Loftis  et  al.   (54)  note in  their  review that  although  there are
numerous reports of  apparent  seasonality  or  periodic  effects in  ground-water
quality  data,   very  few  long-term  datasets  exist at  sufficiently  high
sampling frequency  (i.e.,  more frequent than quarterly) to  statistically
distinguish these  effects  from those  of serial dependence or autocorrela-
tion.    These  characteristics  of the  limited  ground-water quality datasets,
together with  the fact that the quality variables  are  frequently not  nor-
mally  distributed,  constrain the use  of  simple  parametric statistical  tests
of significance to compare means or  identify trends  (55).   The development
of benchmark  ground-water quality  datasets  at  high  sampling frequency (i.e.,
monthly or biweekly)' for  time periods  in  excess of  one  year would  be  most
useful  in determining minimum  statistical  criteria  for  cost-effective  net-
work  design.   This type  of dataset would also  be very useful in the identi-
fication of applicable  statistical  methods  for  trend analysis  and  for
significance  testing of  comparisons  of  background versus contaminated water
quality conditions (56).

     Spatial and temporal  variability in ground-water  quality may affect the
sensitivity  of  contaminant detection and  the estimation  of mean chemical
concentrations.   To  some extent,  spatial chemical data  collected  at  discrete
points along a horizontal  flow path  may  be quite  similar to  data collected
over time at  a single point in the path.   This supposition  is,  of  course,
dependent on  a number of factors related  to  hydrologic conditions  as  well  as
the nature of the chemical source,  reactivity and mobility constraints.  The
substitution  of  spatially intense samples  for  use  in  temporal  variability
studies  could  be applied  to  studies  of ambient  concentrations  of conserva-
tive  chemical  species for regional  assessments  in  rather unique  hydrologic
situations.

-------
                                  SECTION 2

                                 CONCLUSIONS
     1)   Sampling  and analytical  errors  can be controlled  to  within about
+20%  of  the  annual mean  inorganic  chemical constituent concentration  in
ground water if the protocols are  properly  designed  and executed.    The use
of previously published  guides  for  ground-water  monitoring  can  provide
reproducible,  accurate  results  for  such  studies.   The  effects of sampling
and  analytical  errors  may  be far more serious  for  trace  organic contami-
nants.

     2)   The results of the  study  concentrate  mainly on inorganic chemical
constituents  in  ground water.     The  statistical  characteristics  of the
time-series data for  reactive chemical constituents (e.g.,  Fe(II),  sulfide,
H202   ®2  anc^ N02") disclose that temporal variability  is often  lower than
the 'magnitude of concentration  changes  observed  during purging  of stagnant
water  prior to sampling.  This  means that  improper  well purging can result
in gross  errors  and the introduction  of artifacts into  ground-water  quality
datasets.

     3) In agreement  with the  results of previous studies,   distributions  of
ground-water quality variables  show   little  consistent tendency towards
either  normality or nonnormality,   and  values within  a time-series  exhibit
significant  autocorrelation  or serial  dependence.    Autocorrelation  effects
detected  within datasets of  relatively  short duration  (i.e., less than one
or two years) suggest  that  it  is very  difficult  to  quantify seasonality  or
time-trends under  stable  hydrologic and  steady  contaminant source  release
conditions.    The potential  implications for the  design of source  detection
and  contamination  assessment monitoring  systems may be  serious.  Sampling
more frequently  than  quarterly  may result in significantly  reduced informa-
tion  return for the cost and effort involved in  data collection.   The actual
magnitude  of this "loss  of information"  will  be  dependent on  both the
sampling  frequency and the proposed duration of the monitoring  effort.  A
relatively long-term data collection period  may be  required  for  reasoned
decision-making  in judging plume capture or cleanup efficiencies, since high
sampling  frequencies may not yield significant increases  in information.

     4)   Temporal variability in ground-water quality  which is  seasonal  in
nature (i.e.,  duration or cycle of less than  one year) may  not be easily
identified, given the  highly  correlated statistical distribution of chemical
parameter values and the sensitivity  of results to purging and pumping.  It
may  be necessary,  therefore,  to have five- to ten-year background datasets
before  seasonal  components can be  distinguished statistically.    Quarterly
sampling  frequency represents a good initial  choice  for monitoring network
designs.    The  frequency should be  reevaluated as  site-specific  data are
collected  and with respect  to  the duration of the program.    Background

-------
water  quality datasets  collected in a single year may only  be  snapshots of
the actual conditions  prevalent  during the sampling  period.

     5)   Regional changes  in water quality (i.e.,  short-term variations or
long-term trends)  may be easily recognized in datasets from  wells upgradient
from  contaminant sources  while source  contributions  may mask any such
changes  in  downgradient  wells.    This observation complicates  simple para-
metric statistical  comparisons of upgradient  and  downgradient contamination
effects in ground  water.

     6) Major ionic  constituents (i.e.,  Na+,  Cl", NH4 + ,  HC03", etc.)  were
also major  components  of the  waste  stream at the  Beardstown  site.   Under
these  conditions,   the extent of natural  or geochemical variability  imposed
no  significant constraints on  chemical speciation model  results for identi-
fying  waste-specific  indicators.

-------
                                  SECTION 3

                               RECOMMENDATIONS
     1)    Several ground-water  research  sites  should  be established to
collect  long-term  (i.e.,  in excess  of  five years)  water  quality  datasets
under known  hydrologic conditions in order to permit the quantitative  treat-
ment  of  seasonality  and serial  dependence effects in the  data.

     2)  Given the fact that natural variability of major ionic  constituents
of ground water exceeds the errors involved in reliable  sampling and ana-
lytical  protocols,   it is  recommended that network  design  efforts  emphasize
longer  term monitoring  of natural and contaminant-related  variability  rather
than  high  frequency (i.e.,  more  frequent than biweekly or  monthly) intensive
sampling  activities.

     3)    Research  into  the  statistical  characteristics  of the  inorganic
water quality variables should  be  extended to include organic  constituents
where  few  long-term datasets  exist and the sources of variability inherent
to network design and  operations may be much  greater than natural variabil-
ity.

     4)    Reliable  methods for statistical  analysis of "trend"  data are
needed,  particularly where  gaps exist  in temporal  datasets.  The  dataset
collected in this  study provides a  sound  basis  for the design of further
research in  the detection  of water quality trends.

     5)    The sensitivity  of overall variability in ground-water quality
results to purging procedures prior  to sampling needs to be  addressed in a
series  of  experiments  which   integrates  hydrologic and  chemical   data
collection methods.

     6)   The  integration  of chemical  and hydrologic data collection  and
interpretation should be encouraged  in  future  research  and  regulatory  data
analysis efforts.   The  use  of  models  which incorporate solute transport and
reaction  components could  be quite  useful  in  determining  hydrologic versus
chemical   control   over the natural variability  in ground-water quality
constituents.
                                      10

-------
                                  SECTION 4

                      EXPERIMENTAL  DESIGN AND PROCEDURES


REGIONAL LOCATION AND DESCRIPTION OF FIELD SITES

     One of the principal objectives of  this study  was to assess  the  rela-
tive contributions of natural and network-design-related  variability  to
ground-water quality  monitoring results.    Two sites  were chosen  to enable
the  isolation of the effects of network  design  variables from those due  to
natural  or  contaminant-related  sources.    To  maximize the  potential  for
temporal  variability,   the chosen  sites  were located over water table
aquifers of  moderate  to high  yield.  One site was in  a pristine environment
far removed from  any sources of contamination.   The other site was in an
industrial environment under the  influence  of a leaking anaerobic waste
impoundment.

     The two field sites  are  located in west-central  Illinois (Figure  1)  on
a wide,   sandy outwash plain  known as the Havana Lowland region.   The "clean"
site  is in Sand  Ridge  State  Forest  at the widest part  of the Lowland several
miles south  of  Peoria.  The contaminated site is near  the  city of Beardstown
at the southern  boundary of  the Lowland  approximately  40 miles (64  km)  down
the  Illinois  River from the Sand  Ridge  site.    As shown in Figure  I, the
Lowland  quickly  broadens south of Peoria  to a width  of over  20 miles (32  km)
and  then gradually narrows to  about 10  miles  (16 km) at Beardstown.

Regional Geology

     The Havana  Lowland  region  is a  section of  the broad Pennsylvanian
Lowland that covers  most of  Illinois and much of Indiana.   It  was formed
during  pre-glacial  times at the  convergence of the ancient Mississippi and
Teays Rivers and other smaller tributaries  (58).  In  the area of Sand  Ridge
State Forest,   the ancient river  channel  cut  through the  Pennsylvanian
shales,   sands tones,  and  limestones into Mississippian limestones,  shales,
and  dolomites.   The channel continued to the south  of  Beardstown,  eventually
eroding  into  formations of Devonian,  Silurian,   and  Ordovician age before
joining  the  present Mississippi River about 20  miles  (-32  km) north of the
Missouri River  at  St.  Louis.   The  modern Illinois  River joins  the channel  of
the ancient  Mississippi at a point just  west of Sand Ridge  State  Forest.

     During  the early Pleistocene period, the Havana  Lowland was  filled with
sand and gravel outwash which was later at least  partially  removed and  con-
tinually  reworked,  covering the  outwash plain with successional terraces  of
sand  and gravel valley train  or outwash fans  (59).    These  sand and gravel
deposits reach  a depth of about 200 feet (61 m)  in the deepest  portions  of
the now-buried  ancient valley.   The major physiographic areas of the Havana
Lowland  are  shown  in  Figure 2.
                                      11

-------
                      90*
                                 3rd PM
                                                        Q Chicago
                     £.  .-J	r	1
                                ILLINOIS
                            .^.-j^ALL
                            Sand Ridge "  " ~,~?\
                                                           -40*
        STATUTE MILES
    0  10 JO 30  40 SO 60
    0  30  40  80  80 100
         KILOMETERS
                                  3rdPM
Figure  1.    Location  map of the two  field  sites  at
             Sand Ridge  State  Forest and Beardstom, Illinois.
                               12

-------
   Figure  2.   Surficial  geology  map of the  Havana  Lowland region.
                  WATER TABLE. 1960
          EXPLANATION
        /j-1 Contour, innrvil
      / &     10 fttt

           Well location
                  * C *  I   • * *  I   •••  1  • r * ' I   •«*
Figure  3.   Water  table  contour map of the  Havana Lowland  region (1960).
                                     13

-------
Regional Hydrology and Climate

     The saturated thickness  of  the  sand and gravel aquifer in the  Havana
Lowland ranges  from 50 feet (15 m)  or leas  along the Illinois River  to more
than  200 feet  (61  m) in areas east of Sand Ridge State Forest.  Greater than
100  feet (31 m) of saturated  deposits underlie both field sites.    A water
table  contour map drawn in 1960 for the  Havana Lowland is  shown in Figure 3.
Contours were  not plotted  at the Beardstown  area,  but the 440 and 450  foot
(134-137 m) contour levels should  continue  to  follow  the Illinois  River to
the south of the Sangamon River  through Beardstown.   More limited coverage,
recent  data indicate that the  present water table topography is  similar to
the 1960 map.  Regional ground-water flow has a  stable  directional component
toward the Illinois  River,  although recent pumpage  for a  state fish hatchery
near  Sand Ridge State Forest has influenced  water levels  locally.

     The Illinois  River  is a major   shipping  route between  Chicago  and
St. Louis.   It is controlled through  levee and lock  structures  and  does  not
usually flood  large areas of the adjoining  land.   The average  annual  dis-
charge rate of the river is  approximately 15-22  cubic  feet  per  second (cfs)
[0.42-0.62  m /s] with  a record low  of about  1.5 cfs (0.04 m /s)  and record
highs  of 90 cfs  (2.5 m /s)  at  Havana and 125 cfs (3.5 m /s) at Beardstown
(60).

     Mean annual precipitation in the Havana Lowland  area is approximately
35.5 inches  (0.9 m).   Total  monthly precipitation recorded  at Havana, midway
between the two field sites,  for the period  1983 through 1987 is shown in
Figure  4.   As  shown,   seasonal variations can be  great,  with the  greatest
amounts of precipitation occurring typically  in the  spring and fall  months.
An anomalously high  monthly rainfall occurred  in  November 1985, shortly
before  the  start of the biweekly sampling  period, when  nearly 11  inches
(0.3  m) of rain fell.

     Ground-water levels  have been  continuously recorded  by  the  Illinois
State  Water  Survey  (ISWS)  at Snicarte  (located between  Havana  and
Beards town) since  1958.   A  well  hydrograph for the period  1983  through 1987
at Snicarte appears  as Figure 5.  The well is  situated in  an area remote from
pumping centers and,  therefore,  should reflect  only natural fluctuations in
shallow  ground-water  levels caused by  seasonal  and long-term responses to
precipitation  and climatologic  conditions.    Annual fluctuations of 3 to 4
feet  (0.9 to 1.2 m) are  common.   Recharge typically  occurs  during  the spring
and late fall months.   A shift  in the  timing  of peak ground-water levels  can
be seen  in the two-year period from 1985  to  1987.    Ground-water level highs
occurred in the  late fall and early winter of those years rather than in  the
spring  as  had occurred in  previous years.   This  shift is apparently  the
result of low snowfall  during  the winter  and leas rainfall during  the spring
months  of  those years.   Note the  severe decline  in ground-water levels
during  1987 when  the  total  annual  rainfall was only  29 inches (0.7  m),  about
20% below  normal.   The greatest  amounts of  rain  occurred in June,  August,
and December  which are usually low recharge  months.  Similar declines were
found  in monitoring wells  and  piezometera at the  two field  sites.
                                        14

-------
  12
  10 •
2"
o
a.
0
LU
£  6
I
I-
2  2--
           lll
I,
III
          1983          1984          1985          1988          1987

 Figure 4.  Monthly precipitation  recorded  at  Havana,  IL, 1983-1987.
   30
    40
           1983
                        1984
                                    1985
                                                 1986
                                                              1987
  Figure  5.   Depth  to ground water  recorded at  Snicarte, IL,  1983-1987.
                                   15

-------
     Studies by  O'Hearn and  Cibb (61)  and Walton (62) indicate  that  during
years  of normal precipitation,   ground-water baseflow  to  streams  in  the
Havana Lowland  area  ranges from 4 to 5 in/yr  (0.1-0.13 m/yr or approximately
11-14%  of  mean annual  precipitation).    O'Hearn and Gtf)b  (61)  estimated
ground-water baseflow to streams  in  this  area for  low  (Q )  and high (Qi0)
streamflows as  1.8  in/yr  (0.04  m/yr) and 13.6  in/yr  (0.34  m/yr), respec-
tively.   Walton (62) also estimated that  recharge  to the aquifer, based  on
the water table map  of  1960  (Figure 3),  was  10 inches (or  0.25 m,  29% of the
mean annual precipitation).

General Site Descriptions

Sand Ridge State Forest

     Sand   Ridge State  Forest is an  Illinois  Department of Conservation
(IDOC)  facility  located  5 miles  (8 km) southeast of the Illinois River in
the north-central  Havana  Lowland.  The  Illinois  State  Water  Survey's  experi-
mental field site  is  located  in the  middle of  the State Forest near  the
center  of Section 34,  T23N, R7W, 3rd PM,  Mason County, Illinois.

     The land  surface is dominated by rolling  sand dunes covered  with  oak,
hickory,  and pine.   Within the  forest, several  areas have been  cleared for
growing  grains that provide  cover  and food for  wildlife  during the  winter
months.   The absence  of  any nearby source  of ground-water contamination  and
the ability  to  work in a remote,  open  site led  to  the  choice of  one  of  the
cleared  areas as  our "pristine"  field   site.

     Privately-owned land surrounding  the  State  Forest  is used  primarily as
farm land  on which  cash grain crops  such  as  corn  and  soybeans are  grown.
Center-pivot irrigation systems  are used extensively throughout  the area to
maintain high crop  yields on  the  sandy soils.    In  1983,  a  state  fish
hatchery near  the  Illinois River  in the  northwest corner of the  State  Forest
began  operation.    Water is supplied to the hatchery by 9 wells  located  one
mile (1.6 km)  north  of the field site  used in this study. Pumpage  at  the
hatchery has  averaged over  9  million gallons per day (0.4  m3/s), with
occasional  lows of about 7 million gallons  per  day  (0.3 m3/s). The influence
of this  well field  on the local ground-water flow  domain at the  field  site
is described later in this report.

     Three  distinct horizons  comprise the unconsolidated deposits at Sand
Ridge:    at  the surface  is 30  feet  (9 m) of dune  sand (the Parkland  sand);
from 30  feet (9 m) to a depth  of 55 feet (17  m)  is the Manito Terrace  of the
Wisconsinan outwash,  consisting of a sometimes  silty,  sometimes  coarse sand
to medium  gravel;  and  from 55  feet (17 m)  down  to bedrock  below 110  feet
(34 m),  and possibly as deep  as  150 feet (46 m), is the  medium  sand  to fine
gravel   of  the Sankoty sand  (Kansan outwash).   Illinoian deposits  are
missing.     The bedrock  beneath the  site  is probably  early  Pennsylvanian
limestone,  shale,  or dolomite.

     Depth  to  the water  table  is  greater than 30  feet (9  m) below  the  ground
surf ace.   Ground-water  movement is generally  toward  the Illinois River.  The
hydraulic gradient measured  at  the  site in  1983 was  approximately  0.0016
                                      16

-------
 (63).  Aquifer tests conducted  on the water  supply wells at the  state fish
 hatchery  indicate that the hydraulic conductivity of the sand  and gravel  at
 approximately 100 feet (30 m) depth (in the Sankoty sand) is about 2000
 gpd/ft2 (0.094 cm/sec). Tracer  experiments conducted  in  1983  (63)  indicated
 smaller  hydraulic conductivities (from  250  to 1900  gpd/ft ,   0.01  to  0.09
 cm/sec) may be  experienced in  shallower materials.   Hydraulic  conductivity
 values  of  350  to  900  gpd/ft2  (0.02   to 0.04  cm/sec) were obtained  by
 empirical  methods  of analysis  based on  the grain  size  distributions  of
 shallow  aquifer  samples  (63).    In  that same study,  the  porosity of  the
 saturated  terrace materials  was found to be 25%.

 Beardstown

     The  "contaminated"  field site  is  located in the  vicinity  of  several
 liquid  waste impoundments serving a pork slaughtering facility  approximately
 I  mile (1.6 km)  southeast of  Beardstown (in the SE  corner, Sec. 23, T18N,
 R12W,   3rd PM,  Cass  County,  Illinois).   Beardstown  is  situated at  the
 southern  end of the broad  Havana Lowland  where the  Illinois  River turns
 southward on its way to the Mississippi River.

     While the field  site  lies two miles southeast of the  river,  it  is  only
 about  5  feet  (1.5 m)  higher than the floodplain.   Farmland  and wooded areas
 surround the facility.  The unconsolidated deposits lying above the bedrock
 consist of the clayey  sands of  the Beardstown Terrace  on the  Wisconsinan
 outwash  plain.  The bedrock  surface  is of Mississipian age and  lies at about
 100 feet (30 m) below  the  ground surface.

     Facility  memos  and  consultant's reports  provided preliminary informa-
 tion on ground-water conditions at the  site.   Owing  to  land  surface eleva-
 t ion  changes,  depth  to  water varies  from about 5 to  15  feet (1.5  to  4.5  m)
 below the  ground surface.    Similar  to the Sand  Ridge site,  regional
 ground-water flow is  toward  the Illinois River (hydraulic  gradient,  0.002).
 However,    the   influence of   several   waste   treatment   impoundments  on
 ground-water elevations at the  site had not  been evaluated.    Due to  the
 presence  of silt  and  clay,  the aquifer  is less permeable than  it is at  the
 Sand  Ridge site.   One falling  head  permeability  test produced a hydraulic
 conductivity  value of only  130  gpd/ft2 (6 x  10"3 cm/sec).   However, this
 value may represent the  lower  end of hydraulic  conductivities  at  the
 Beardstown facility as  a  slug test on another well on-site could not be con-
 ducted because of "an  inability  to fill  the well  with water" (64).  While it
 is unclear from the  report  exactly  how the  well was  being  filled, it is
 clear  that the aquifer could  accept  water  faster than it was  being  applied
 to the well.   This  suggests  a  hydraulic  conductivity greater than  that
 calculated for the  other well.
FIELD SITE INSTRUMENTATION

Well Construction Details

     Each field  site  contained:    1) a network of piezometers to describe
local ground-water conditions,   2)  nested  sampling wells, to define the
                                       17

-------
ground-water  quality  variability,   and  3)  real-time  hydrologic  monitoring
systems  that have provided  a continuous record of selected hydrologic  and
meteorologic  parameters since mid-1987.   The placement and construct ion  of
these structures and  instruments  within the  field site  were critical to the
study, This  was particularly  true at  the  Beardstown site where placement  of
sampling wells upgradient  and within  a contaminant plume  downgradient of  an
anaerobic waste  impoundment was  vital  to the project objectives. The moni-
toring devices  (sampling wells,  piezometers, and hydrologic  monitoring
systems)   were  constructed  in  a   similar   manner at   both  field sites.
A  description of general  construction details  is  provided here  before  dis-
cussing  each site  in detail.   Previous research investigations   (1,2,3,4)
provided the basis for the design  and construction of the monitoring wells
and  selection of pumping mechanisms.

Monitoring Wells

     Bore holes  for construction  of all monitoring  wells were drilled with a
4.25-inch (11  cm) inside  diameter  (I.D.)  hollow-stem  auger.    Prior  to
drilling  at  each location,  all auger  flights  and aquifer solid samplers were
steam cleaned  thoroughly.   Steam  cleaning was not done when moving between
successively  deeper holes  at  the  same  site but was done  when  moving to  a
different  site (e.g.,  moving from areas downgradient  to  upgradient of the
anaerobic  impoundments  at  Beardstown).    In  addition,   all well casing
materials and well  protectors  were  steam  cleaned  immediately  before place-
ment  in  the  bore hole.

     Due to  the  sandy  environment,  it was necessary to use  water  to  maintain
an  open hole  as  auger flights were  added.   Only water  taken from nearby
water wells  with similar water quality to that of the upgradient wells  was
used. No drilling mud or other method  of bore hole stabilization was used.
Casing lengths and the screen were  screwed together as they were placed down
the inside of  the auger  flights before the  flights were removed to ensure
that the  sandy materials  would  not collapse in the bore hole after  drilling.
All casings  and  screens were  2-inch  (5 cm)  I.D.    In every case, the
saturated formation  collapsed around the casing  and screen  as the auger
string was  removed,  leaving an open  bore hole only above  the water  table.  A
plug of bentonite pellets  two to  three  feet thick  was  placed  in  the  annulus
directly  above the top of the caved  material.   Drill cuttings  (principally
sand) were  backfilled above  the  bentonite seal to  within  three  feet  of land
surf ace.  A  concrete plug  approximately 3 feet  (0.9  m) deep  and 2  feet (0.6
m)  in diameter was  placed around  the well casing at the  ground surface  to
prevent surface drainage from moving down  the well casing.  A six-inch (15
cm)  diameter  steel  well  protector  with  locking cap  was  installed  over the
protruding  well casing to  complete  the well.

     The construction details  of the  sampling wells (Figure 6)  at both sites
are  identical in all ways   other than  the  length  of casing and casing
materials  in  two wells  at  Beardstown.    One     well at Beardstown  was
constructed  of  stainless steel  (SS) and  one other of  polyvinylchloride
        All  of the other  sampling  wells  at  both  sites were constructed with
polytetrafluoroethylene (PTFE-Teflon(R), DuPont).  All wells have 2-inch
                                      18

-------
                          Well Protector









—




Drive Point ,.
><
— 1













—
X
Ground Surface
\
1
1
^*



Well Casing \
s
1.25" dia. or 2" d!a-
2" dia. Teflon,
Galvanized PVC, or
Steel Stainless
Steel

^7


~
/
7
\
^
k
I,
h
Well Screen I
^,-12" or 36" 60" ^^ ,
^r long long -^\
L












—
—


I
.s Concrete Seal
*
1
t
t
I
T
^ * Dri" Cuttings
\^S Backfill

^ ^^- Bentonite Seal

\
|^ Caved Formation
N Material
s

N
H
s!
H
        V
    Piezometer
Sampling Well
figure 6.  Typical piezometer and  sampling well construction
          for the Sand Ridge and Beardstown  field sites.
                           19

-------
 (5 cm)  I.D.  flush-threaded casing.   Screens  were 5  feet (1.5  m)  long with
0.01-inch (0.02  cm) slot openings.

     Dedicated   tubing   and   positive  displacement    bladder   pumps  (Well
Wizard(R),   QED Environmental Systems) were  permanently installed  in  each
sampling  well with pump intakes  positioned 2  feet (0.6 m) below  the top of
the screen in each well.  Each bladder and all  tubing  were made  of PTFE, but
the pump housings  and  fittings  were made  of the same  material as that of the
well casing  in  which they were  installed (i.e.,  PTFE pumps in PTFE wells,
PVC pump in PVC well, and SS  pump in SS well).

     Well development  was performed within  hours  of well installation  by
evacuating the  water in the  casing with compressed  air.   Surging  to  first
evacuate  the  casing,  then allowing  complete  recovery of  the  water in the
well  was the best available method  to obtain water movement through the well
screen  to remove  fine  materials  from the permeable sands   and  gravels.
Development times  from 30 minutes to one hour  were  required  until  the  dis-
charge  was  free of sediment.   Steady  development  with a constant stream of
compressed  air  was  equivalent to a pumping rate of 20 to 30 gallons per
minute  (I to  2  1/s) at  both sites.

     Following development, hydraulic conductivities  at all  sampling  wells
were  determined by  slug  tests performed through a  casing pressurization
technique similar  to  that  detailed by  Prosser (65).   The  casing  pressuriza-
tion  technique was used because these  highly permeable aquifers possess very
rapid (less  than one minute) recovery times.    A  microcomputer-controlled
data acquisition system was used  at the two  field sites  to  allow collection
of up to  100 data points  (depth-to-water measurements) per  second  (66).
Casing  pressurization tests could  not be performed at the Sand Ridge  well
D035 (#1 in  Figure 7) because the  height  of the  water column  over  the  well
screen  was  insufficient to allow for  significant displacement of water.

Piezometers

     The piezometers  at both sites (Figure  6)  consisted of  drive points
constructed of  galvanized steel  casing.    Screens on  the  drive  points  were
either  12 inches (30.5 cm) or 36 inches (91.4  cm)  long and had slot widths
of 0.01  inches  (0.02 cm>.  Casing  diameters were  either 1.25 inches (3.2  cm)
or 2 inches  (5 cm).

Hydrologic Monitoring  Systems

     Data logging  systems for continuous recording  of selected hydrologic
and  meteorologic parameters   were installed late in  the project to  provide
more information on  the timing of recharge events at both sites.  Instrumen-
tation  included:

     1)  submerged pressure  transducers   to  measure  ground-water
         levels  in  a  piezometer nest,

     2)  a tipping-bucket rain  gauge to measure rainfall timing  and
         intensity,
                                      20

-------
                                             SRI
                                 \
                                 I
                                  \
                                  \
                                  \
                                   \
                                   I
                    Direction of Regional  \

                    Ground-Water Movement
A

 SR2
                        Data Logger/*^^~

                          Shelter ^    •/
                                                 A

                                               SR3
                                                                . 0035 (f 1)

                                                               . D050 (/2)

                                                              . 0065 (|3)

                                                              00105
                                EXPLANATION
  SCALE  OF FECToQ           .  Sampling Well
Piezometer

Sampling W

Existing Well
          Figure  7.   Sampling well and piezometer network for

                      the Sand  Ridge  field site.
                                  21

-------
     3) a temperature probe to measure  ambient air  temperature,  and

     4)  a  barometric  pressure  sensor  to  measure   atmospheric
         pressure.

Parameters sensed  by these instruments  were recorded on  a multichannel,
battery-operated,  portable  data logging  system  (Easylogger(R)  Omnidata).
This particular  data logging system  provided  a means to record  sensed data
every  15  minutes for a period in excess of thirty days  without  downloading
the  recorded  data.    Data  were stored on 128 kilobyte nonvolatile  EPROM
(erasable  programmable  read only memory) storage packs. Storage packs were
capable of being replaced in the  field,  brought  back to the  office,  and
downloaded  to transfer the  stored data  to  a  microcomputer.
SAND RIDGE STATE FOREST

Background

     The  placement of sampling wells for the purpose of evaluating natural
variations in  ground-water  quality at this "clean" site  was  not as critical
as at the  "contaminated"  Beardstown  site.   The  site  at  Sand Ridge has been
used in previous (63)  and other current (67)  State  Water Survey  experiments,
so the hydrogeology  of the site  was well known.   In particular,  the depth  to
water  and the direction and rate  of ground-water movement had  been
documented.

     One  principal   concern for this site  was  the construction  of the
shallowest sampling well.  To evaluate shallow  ground-water quality changes,
it was important  to complete a well  as  near to the water table as possible.
The  risk  of a shallow  well completion was that of the well going dry during
the  study period and the  subsequent  loss  of that  data  collection  point.
Since 1982,  the  water table had been  falling  principally  as a result  of
water use by the nearby fish  hatchery.    Consideration was given to both
natural  ground-water  fluctuations  and  the  maximum drawdown that could be
expected  from the fish hatchery's production wells.

     A  second  concern at this  site  was  to  avoid  the potential  influence  of
residual  dye tracers used in experiments conducted  in  1983 (63).  Informa-
tion from previous  and ongoing research  at the site provided excellent
information to  determine a  location upgradient  from such potential  influ-
ences.   A summary of the well construction  details for  the  Sand Ridge site
appears  in Table 2.

Piezometer Network

     Many of  the  piezometers at  this site  have  been in place since  1982 and
have  provided  a  record of water  level fluctuations  since  that  time.    Wells
SRI,  SR2,   and  SR3 (Figure  7)  are 40 feet  (12 m) deep with 20-foot  (6 m)
screens that intersect the  water table.    Ground-water elevation data from
these wells were  used to  calculate  the gradient and direction of flow  at the
                                      22

-------
                      TABLE 2.   SAND RIDGE WELL CONSTRUCTION  DETAILS


Well
name
SRI
SR2
SR3
ND2.1 .1
ND2.1 .2
ND2.1 .3
D035 (#1)
D050 (#2)
D065 (#3)
D0105 (#4)

Well
type3
5 cm PVC
5 cm PVC
5 cm PVC
3.2 cm SP
3.2 cm SP
3.2 cm SP
5 cm PTFE
5 cm PTFE
5 cm PTFE
5 cm PTFE

Date
completed
1982
1982
1982
1982
1982
1982
10/814
10/814
10/814
10/85
Measuring
point elev.b
(ft)
993.59
991 .78
991 .73
991 .29
991.31
991 .32
992. H3
992. i4H
993.33
992.09
Bottom
elev.b
(ft)
951 .6
9149.8
91*9.7
956.3
952.3
9147.3
955.3
91*0.2
925.1
885.0
Well
depth
(m)
12.0
12.1
12.1
10.0
11 .2
12.8
10.7
15.2
19.8
32.0
Well
riser0
(m)
0.73
0.61
0.6J4
0.58
0.58
0.58
0.6J4
0.67
0.67
0.614
Screen
length
(m)
6.1
6.1
6.1
0.6
0.6
0.6
1 .5
1.5
1 .5
1 .5

a PVC,  polyvinyl chloride;  SP,  sand point;  PTFE,  Teflon(R)
b Elevations in feet relative to 1000-foot  datum,  not mean  sea level
c Height of well above land surface

-------
site.   Nested wells ND2.1.1,  ND2.1.2,   and ND2.1.3 are  1.25  inch  (3.2 cm)
diameter drive  points with 12  inch  (30.5 cm)  long,  0.01-inch (0.02 cm)
slotted screens  completed  at  depths of 33, 37,  and 42  feet  (i.e.,  10, 11,
and 12.5 m),  respectively.

Monitoring Wells

     The  sampling  wells, #1, #2, #3, and #4,   are located as close  to  the
upgradient  boundary  of the field site  as  possible  (Figure 7).    The wells
were  placed  upgradient of well  sites  used in the previous tracer experiments
according to  ground-water  level information  collected from  the SR wells.
The sampling wells  were completed at depths of 35, 50 65, and  105  feet (11,
15;  20, and  32  m)  below ground.   Wells #1,  #2  and #3 were completed in
October of 1984.   Well #4  was  completed in September, 1985.   All four wells
have  PTFE casing and screen.

Hydrologic Instrumentation

     The  data logging  system  for this  site was  installed  inside  a walk-in
shelter built  around the ND nest of  wells.   The  system was designed to moni-
tor:   a) water levels  in the 3 ND wells by means  of 3  submerged pressure
transmitters,  b)  precipitation  in a  tipping-bucket  rain  gauge, c) ambient
air  temperature  from a  temperature probe,  and  d)  atmospheric pressure from a
barometric pressure sensor.   Installation  of these instruments did not  occur
until  August,  1987.   The usefulness  of  data  from the  instruments was greatly
reduced due  to the  lack of  measurable recharge during the  remainder of the
sampling period.


BEARDSTOWN

Background

     The  Beardstown site  is located on the property  of  a  commercial  pork
processing operation.    The  average plant  waste discharge for this facility
is approximately  1.1  million gallons  per day (0.05 m3/s).  The  wastewater is
composed of animal slaughtering wastes,   processing wastes,  and  a  small
volume of sanitary sewage.    Treatment of the wastewater is accomplished  by a
three-stage series of impoundments  (Figure  8).  Waste flow from the plant is
initially sent to  a  set of three small anaerobic impoundments.  These three
impoundments are arranged  in  parallel and  cover a total  of  approximately 3.2
acres   (1.3  ha).  Discharge  from these impoundments  is sent to a 30-acre (12
ha) intermediate  impoundment  capable  of  storing approximately  29 million
gallons (110,000   m3).   Five floating aerators  in  the intermediate  impound-
ment  provide  oxygen to the  water to  enhance  treatment.  Wastewater is sent
from the intermediate  impoundment to another 30-acre  (12 ha) impoundment for
contact chlorination treatment.   Discharge from  the final impoundment is
accomplished  by three  1400  gpm  (0.09  m3/s)  pumps  which  deliver the water to
four  central-pivot spray irrigation  systems.   The  irrigation systems provide
water to  400 acres (160 ha)  of cropland  situated  southeast of the liquid
waste  impoundments.   Discharge  of water through the irrigation system occurs
throughout the year whenever the  final impoundment becomes full.
                                      24

-------
                 SCALE OF FEET
                 100    200         400
     EXPLANATION
    o    Piezometer
    •    Piezometer nest
    a    Sampling  Well
    •    Sampling  Well nest
    A     Existing Well
   u—g   up—gradient
   d —g   down—gradient
           85
                                  r
ST.BS.BP d-g

*WLR2
                                                    87
                                                              W18
  Pump
  House   s~ Anaerobic Impoundments -r—^^^
                                 Bfl
                                                        B6_
                                                     B1
                Intermediate
                Impoundment
                                             BT u-g
                                                     \
                                           B2
           Direction of Regional \
           Ground-Woter Movement
                                               To Spray Rig
                                   To Spray Rigs

                                   W9
Figure 8.   General plan of  the  Beardstown  Held site.
                            25

-------
     The  process of selecting specific well locations at Beardstown  was much
more involved than  at  the Sand Ridge  site.    The  local hydrogeology  and
chemistry of the ground water  near  the  impoundments were unknown  in  any
great detail at the beginning of the project.    Very few ground-water data
for the  Beardstown area  were available.   No  data existed  to describe  the
influence  of the impoundments  on local ground-water  flow and quality.
Therefore,   the  design of  the  piezometer  and  sampling  well  networks
progressed in stages  with  the  goals of, first,  accurately  describing  the
regional gradient  and direction  of  flow,  and second, defining the location
(vertically and  horizontally)  of the contaminant plume presumed to be
emanating  from  the three anaerobic treatment impoundments.

     Based on a  consultant's  report  (64)  regarding regional gradient  and
direction  of flow,  a triangular  array  of piezometers was  placed south of the
easternmost  anaerobic  impoundment  (Figure 8) for the purpose of providing
more site-specific information.    Later,  more  piezometers  and piezometer
nests were installed  at several locations around  and between the impound-
ments to  determine the presence and  magnitude  of vertical gradients  caused
by mounding  of water beneath the impoundments  (Figure 9).  The data  obtained
from these piezometers were used  to establish the  positions  of  the
upgradient and downgradient nests  of sampling  wells on a  line  parallel to
the regional ground-water flow path passing beneath anaerobic impoundment 3.

     Before  designing and installing the sampling  well network,  it  was
necessary to find the horizontal  and vertical position  of  the  leachate plume
presumed  to be  emanating  from  the anaerobic impoundments.  In July of 1985,
small diameter sand  points   were  driven at  several locations  along  the
downgradient  side of the easternmost anaerobic  impoundment (#3).  At  each  5
foot  (1.5  m) depth interval,  the well point  was bailed  until the discharge
was  clear  of sediment.   A sample was subsequently collected and  tested for
electrical  conductance.     Higher   electrical   conductivity   indicated  the
presence  of greater dissolved solids concentrations that  would  accompany  a
contaminant  plume at this site.  The highest conductivity readings occurred
in a  15-foot (4.6 m) thick zone  between depths  of 20 and 35 feet  (6.1 to
10.6  m).   This zone was chosen for the position of the downgradient sampling
wells.   Aquifer solid  samples taken from the  formation  at the depth of the
sampling well screens  indicated that the zone of interest  consisted of a
silty medium-grained  sand.   A summary of the well construction  details at
the Beardstown  site appears in Table  3.

Piezometer Network

     Most of the piezometers at this site were installed  specifically for
the purposes  of the project,  and their locations  are shown in Figure  9. Well
W18  had been  drilled previously and  formed part  of a  larger sampling network
used by  facility  consultants.   Piezometers  WLR2.1 and  WLR2.2  are nested
2-inch (5.1 centimeter) diameter  sand point wells that were  monitored con-
tinuously by the data  logger  at  this  site.   All other piezometers with names
beginning with a  "B"  are  1.25-inch (3.2  centimeter) diameter  sand  points.
                                      26

-------
                     TABLE 3.  BEARDSTOWN WELL CONSTRUCTION DETAILS
Well
name
B1
B2.1
B2.2
B3.1
B3.2
B4
B5
B6.1
B6.2
BY
B8.1
B8.2
B8.3
WLR2.
SLR2.
BT18
BT23
BT25
BT30
BT35
BS30
BP30
BT33
W18













1
2
(#5)
(#6)
(#8)
(#9)
(#10)
(#11)
(#12)
(#13)

Well
type*
3.2 cm SP
3-2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
3.2 cm SP
5 cm SP
5 cm SP
5 cm PTFE
5 cm PTFE
5 cm PTFE
5 cm PTFE
5 cm PTFE
5 cm SS
5 cm PVC
5 cm PTFE
5 cm GS
Date
completed
6/25/85
6/25/85
10/29/85
6/25/85
6/25/85
6/26/85
6/26/85
10/29/85
10/30/85
10/30/85
10/30/85
10/30/85
4/7/87
10/30/85
10/30/85
10/10/85
10/10/85
10/9/85
10/9/85
10/9/85
10/9/85
10/9/85
10/9/85
4/19/82
Measuring
point elev.
(ft MSL)
450
450
450
451
451
458
460
453
453
458
462
462
462
458
458
450
451
457
457
457
458
457
458
457
.92
.55
.52
.71
.79
.20
.55
.67
.51
.82
.60
.62
.54
.68
.47
.99
.20
.48
.56
.72
.04
.70
.74
.52
Bottom
elev.
(ft MSL)
433
433
423
434
421
430
431
435
425
434
435
425
442
430
425
431
426
430
425
420
426
425
423
439
.9
.5
.5
.7
.8
.2
.5
.7
.5
.8
.6
.6
.6
.7
.5
.0
.2
.5
.6
.7
.0
.7
.7
.5
Well
depth
(m)
4.2
4.3
7.3
4.3
8.2
7.5
8.0
4.6
7.7
6.4
7.3
10.3
5.2
7.6
9.2
5.6
7.1
7.7
9.2
10.8
9.2
9.2
10.2
5.2
Well
riser**
(m)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
.91
.88
.88
.85
.88
.94
.75
.85
.82
.91
.88
.91
.88
.88
.82
.45
.45
.45
.45
.45
.45
.45
.45
• 30
Screen
length
(m)
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.91
0.61
0.61
0.61
0.91
0.91
1.52
1 .52
1.52
1 .52
1.52
1 .52
1 .52
1 .52
1.52

 * SP,  sand point;  PTFE,  Teflon(R);  SS,  stainless steel;  PVC,  polyvinyl  chloride;
   GS,  galvanized steel
** Height of well above land surface

-------
          BT35 (#10) BS30 (
                 • •    ^
           BT30 (#9j»B   •WLR2.2

          BP30 (#12) 8T25 (#8)

                       BT33 (#13)
   Anaerobic
Impoundment #2
88.1
B8.3
88.2
  Anaerobic
Impoundment  #3
                                    B7
                                                      W18;
                    B3.2
                    BT23 (|6)•
                     BT18 (#5)«
                                                      B1
                                                             86.1,
'B6.2
 Intermediate
Impoundment
                                   B2.1
                                     B2.2
                            Direction of Regional \
                            Ground—Water Movement
                      SCALE  OF FEET
                   0    50  100        200
                             ±
                                 EXPLANATION

                                 •  Piezometer
                                 •  Sampling Well
                                 A  Existing  Well
     Figure 9.  Detailed view  of the  sampling  well and piezometer
                 network  at the  Beardstown field  site.
                                       28

-------
     The  piezometers  were  installed  in  a pattern that facilitated  the draw-
ing  of potentiometric surface maps at  three  different  levels beneath the
impoundments.  The deepest  Well in each nest,  Bn.2,  monitors the  regional
gradient and  direction of flow.  The shallow  wells, Bn or Bn.l,  indicate the
effect of  mounding under  the impoundments.   Piezometer B8.3 was  installed  on
the  berm between the impoundments as  a means of locating the  top  of the
ground-water mound formed by water leaking beneath impoundments  2  and 3.

Monitoring Wells

     The  downgradient  nest  of sampling wells  (Figure 9) contained three  PTFE
wells, #8, #9  and #10, with screens placed at three discrete  5-foot (1.5  m)
intervals.   The wells were  completed at  25 feet (8 m),  30 feet (9 m), and  35
feet (11 m)  below  ground  surface,  respectively.   These three wells were
located  in such a way as to be parallel to  the regional ground-water flow
path, while two other 30-foot (9 m)  wells,  #11 and  #12, (one SS  and the
other PVC) were placed on either side of the  30-foot (9 m)  PTFE  well, #9,  in
a  line perpendicular to regional  ground-water  flow.

     The two  upgradient  wells, #5  and #6,  were completed at 18 feet (5.5  m)
and  23  feet (7 m).   The wells  were placed  at  the  foot  of the impoundment
berm on  a  cross  section  parallel with the  regional  ground-water,  flow and
through the downgradient wells #8, #9, and #10.   Wells #5 and  #6 also were
constructed of PTFE.   Sampling  well #13  was completed at a depth of 33 feet
(10  m)  and was placed along the  same  cross  section  as the other  sampling
wells at  a  point mid-way  between the impoundment and the downgradient nest
of wells.   Because  of land surface elevation changes,  the screened sections
of wells  #6, #9 and  #13  are approximately the same  depth below the water
table.

Hydrologic Instrumentation

     The  data logging system at Beardstown  recorded  water  levels  in the
piezometer nest WLR2.1  and WLR2.2, precipitation,  barometric pressure and
ambient  air temperature.   Installation of  this  system was completed  in April
1987.

FIELD ACTIVITIES

Sampling Trip Logistics

     Biweekly sampling was the highest  frequency  which  would permit  sample
analysis  within the  recommended  holding  times  for each  constituent and allow
preparation for the next run on  a  routine  basis.       It was  necessary  to
carefully  consider the  timing of each step of the operation to maintain the
sampling  frequency  within  a day of the  biweekly  timing.   The sampling crew
was  charged  with establishing a  sample  collection routine which would work
within the travel,  sampling  and analytical  constraints.   The schedule out-
lined in  Table 4 was  used  during this  project.

     The  travel involved in this  schedule proved to  be quite manageable,
even in inclement weather,  provided  the  sampling  itself  could be  completed.
                                      29

-------
             TABLE  4.   PROJECT SAMPLING RUN SCHEDULE
 7:00 AM    Depart  Champaign
 9:30 AM    Arrive  Manito,  IL; purchase ice for sample  coolers
 9:45 AM    Arrive  Sand Ridge State  Forest
10:00 AM    Set  up;  measure ground-water  levels  in  monitoring  wells;
               insulate    pump   output   lines   and   calibrate    field
               measurement equipment
10:30 AM    Begin  well  purging procedures and flow  measurements
11:00 AM    Measure  ground-water  levels   in  piezometers;   purge  and
               sample PTFE wells #1-4 in order.   Field determinations of
               alkalinity and  dissolved  oxygen  are  made  as samples are
               collected.
 2:30 PM    Take blank samples for  Day  1  (well #7) and  prepare to leave
               for Beardstown site
 4:30 PM    Arrive at Beardstown upgradient  well site;  repeat  setup,
               ground-water  level  measurement, calibration and purging
               procedures
 5:30 PM    Sample  upgradient  monitoring  wells #5 and  6 in order  and
               perform alkalinity  and  dissolved  oxygen determinations;
               prepare to leave  site
 8:00 AM    Purchase  ice  and refuel vehicle, if necessary
 9:00 AM    Arrive  at  Beardstown downgradient well site; repeat  setup,
              ground-water level measurement,  calibration  and purging
              procedures;  measure ground-water levels  in  piezometers
10:30 AM    Sample nested  PTFE wells  #8, 9, 10 in order, then  complete
              wells #11,  12 and 13;  perform alkalinity and dissolved
              oxygen  determinations  as samples are collected
 2:00 PM    Take blank  samples for  Day  2 (well #14);  prepare  to leave
              site
 4:30 PM    Arrive Champaign;   unload  samples from van,  sort  and select
              two  wells  for  duplicate  analyses  and  two  wells  for
              spiking
 5:30 PM    Spike   and  split  samples  appropriately  and transfer to
              refrigerators and  freezers  for storage prior  to analysis
                                  30

-------
The most  critical holding periods  for  specific  analytes (i.e.  sulfide,
ferrous iron,  methane and ammonia) could still  be met  even if road  closures
or  other problems  necessitated an  additional  overnight  stay prior  to
returning  to  Champaign.

Use of the Van

     In order to  establish control  over the conditions  of  purging,  sampling,
and field determinations,  a sampling  vehicle was designed, purchased, and
outfitted.  Built on a 20-foot  (6 m) stock van chassis,  it  was equipped with
a  6.5  kw generator, heating  and air-conditioning,  a refrigerator,  and  a  I HP
on-board air compressor  to  supply drive gas for the sampling pumps.   The van
also has  approximately  12 linear  feet  (3.6  m)  of  bench space for field
determinations.    It seats three individuals  which make up the foul-weather
sampling  crew.   The  vehicle was received in mid-October of 1985 from the
manufacturer.

Field Parameters and Sampling Protocols

     Sampling and analytical  protocols   are detailed  descriptions  of the
actual  procedures used  in  sample retrieval,  handling,  storage  and  transport,
subsampling and  analysis.   Given  the results of our previous  research in-
volving ground-water sampling,  we  were  confident in  our selection  of  relia-
ble sampling  and  purging  techniques  (1,2,3,4,5,6).    The  basis  of the
sampling  protocol was to obtain  chemically  and hydrogeologically  representa-
tive samples for volatile organic compounds.   It  was expected that  a  suita-
ble protocol for  the collection  of volatile organic  compounds would be more
than adequate  for the  other,   less error-prone chemical constituents.
Therefore,  purging would be  done at flow rates  well below those used during
well development.  The actual sampling  of the  ground  water would  be  done at
flow rates of approximately 100  mL/min.   Wells  at  both sites  showed very
good yields,  with hydraulic  conductivities  in  excess  of  400  gpd/ft2 (0.02
cm/s).    From rough  computations  (3), it was clear  that  greater than 95%
aquifer water  could  be obtained from pumps  set  at the top  of the  well screen
in ten minutes or less by pumping  at  1-2 L/min.  Thus, all of the  wells could
be  pumped  at  a  relatively low rate,  ensuring hydrologic representativeness.
This was verified  for each well  at  the time  of sampling  by monitoring the
well purging parameters:  pH, temperature,  conductivity, redox  potential and
dissolved oxygen. The remainder  of the  sampling protocol  for the chemical
constituents  of interest  involved  Agency  recommended procedures  for  sample
handling  and preservation (68,69).

     The  sample  handling scheme followed in the field  is shown  in  Figure 10,
referenced to  the notes  below:

     A.  Once  the variabilities of the  well  purging parameters  had
         stabilized to  less  than  about  10% (i.e.,  ±0.05  pH, ±5
         (iS.cm"1,  ±0.1°C,   and  ±5 mv)  over two successive  well
         volumes,  unfiltered samples  were taken first for organic
         compound determinations.   The sample containers were 40 mL
          (TOG, VOC)  or  500  mL (TOX) glass bottles with PTFE-lined
         caps.  After  flushing  the  sample  bottles with  the aquifer
                                      31

-------
              DETERMINE PUMPING RATE
               MONITOR/PURGE WELL
                     RECORD
               (pH, Eh. Temp, DO, Cond)
                     COLLECT
               • UNFILTERED SAMPLES
                                             • WINKLER D.O.
                                                (Triplicate)
                      FILTER






1
COLI
FILTERED
ACIDIFY 1
0.2% v/v ZINC
H2S04 ACETATE
1 1
COLI
1 . 1 "J
F 1 f
NH3 HS-/S-
t-P04
Fe+2
1

Filtrate

.ECT ,, Al KAI INITV
SAMPLES (Duplicate)

-ECT
NKS
NOj-
N03I
ACIDIFY
0.3% v/v
HNO3

N« Ca
^02-N M9
o-P04 F«
cr Mn
so4- .F
                                                         Store 94'C
                                                        FROZEN
Figure 10.   Sample handling flow diagram.
                      32

-------
         water,    samples   were   taken  devoid  of  headspace  and
         refrigerated.  Methane samples  were obtained by  filling  25
         mL  gas tight polyethylene  syringes.    These  were also
         refrigerated.

     B.   Triplicate  field  samples  of  unfiltered  water (Wells #1-3
         only) were collected in clean,  300 mL glass  BOD bottles
         for dissolved  oxygen determinations.   The  Azide Modifica-
         tion  of  the Winkler  Method was used (70).

     C.   A  SS and  PTFE  in-line  filter holder with 0.4 (im polycarbo-
         nate  membranes was  used  for sample  filtration, with the
         dedicated sampling pumps providing  the  pressure.   At  least
         one  liter of  water was  filtered and discarded in order to
         flush the filter  holder  and wash the membrane before  any
         samples were taken.    Samples for various analyses  were
         treated as shown  in  Table 5 (e.g.,  samples for metals
         determinations   were   taken in  acid   cleaned,   500 mL
         high-density   polyethylene   bottles   and  acidified  with
         HN03).

     D.   Duplicate  field  samples  of  filtered water  were taken in
         clean polyethylene  bottles for alkalinity determinations.

     E.   At the  end of each day,   field  blanks of double deionized
         water  were  taken  as  well  numbers  7  and  14  for  each
         inorganic  parameter and preserved  accordingly. Double
         distilled-in-glass    water was   used   for   the  organic
         parameter blanks.

     Replicate samples were taken from each well  for  all analytes  except
methane.   Two samples from  two individual  wells  (one  from each  site)  were
selected for  duplicate  analyses.   Similarly,   two samples  were  chosen  for
spiking with  suitable standards  for the inorganic species.

     In order to  expedite the sample  handling procedures, all sample  bottles
were  brought into the  field thoroughly cleaned  and  labelled for  individual
analyses or similarly  preserved analytes.   An example set of labels is  shown
in Figure 11.   The labels contain  project-specific information  such as  ID#
(well or blank),  date  of collection,  means of preservation,  handling pre-
cautions  (if any),  as well as the  determinations  to be  done on  subsequent
subsamples.   Details of how this  information  was  used in  sample  tracking  are
provided in  sections to  follow.
LABORATORY ACTIVITIES AND ANALYTICAL PROTOCOLS

Sample Preparation

     Polyethylene bottles  to  be used for. field collection were initially
washed in a  5%  Contrad 70(R)  detergent  solution.   The bottles  were then
rinsed with deionized water, followed  by a  rinsing with 50%  HC1.  The final
                                       33

-------
TABLE 5.  SAMPLE HANDLING. PRESERVATION  AND ANALYSIS ACCORDING TO PARAMETER
Chemical
parameter
Alkalinity
Winkler DO
NH3
N03N02-N
S=
Si02
o-POi4
N02-N
TOC
VOC
t-POij
CHij
Cl~
Ca
Fe
Fe(II)
K
Mg
Mn
Na
SQn=
TOX
Container Means of
type1 preservation2
HP
B
HP
HP
V
HP
HP
HP
V
V
HP
S
HP
HP
HP
HP
HP
HP
HP
HP
HP
G
4°C


Maximum
holding3
24
Fix on site, store dark 8
4°C,
4°C
4°C,
4°C
4°C
0.2$

v/v H2SOi|

Zinc Acetate




Frozen
4°C,
4°C,
4°C,
4°C,
4°C
4°C,
4°C,
4°C,
4°C,
4°C,
4°C,
4°C,
4°C
4°C,
Gas
Gas
0.2$
Gas

0.3$
0.3$
0.3$
0.3$
0.3$
0.3$
0.3$

Gas
Tight
Tight
v/v H2SOi|
Tight

v/v HN03
v/v HN03
v/v HN03
v/v HN03
v/v HN03
v/v HN03
v/v HN03

Tight
28
8
7
28
48
48
24
24
28
24
28
6
6
8
6
6
6
6
28

hours
hours
days
hours
days
days
hours
hours
hours
hours
days
hours
days
months
months
hours
months
months
months
months
days

Approximate
time to
analysis^
Immediate
Immediate
<3
<3
<3
<3
<4
<4
<4
<4
<5
<1
<1
<2
<2
<2
<2
<2
<2
<2
<2
<2
days
days
days
days
days
days
days
days
days
week
week
weeks
weeks
weeks
weeks
weeks
weeks
weeks
weeks
weeks

Footnotes:
   1.   High density polyethylene  (HP),  glass  septum vial (V) ,  amber glass
         with PTFE-lined  cap  (G) ,  polyethylene  syringe (S),  BOD bottle
         (B).
   2.   Sample preservation performed  immediately  upon  collection.
   3.   Samples analyzed as soon after collection  as  possible.
          The times listed are the maximum times  that  samples  may be
          held before analysis and still  considered  valid.   Some samples
          may not be  stable  for  the  maximum time period  given  (68,69).
   4.   Approximate time to analysis includes  two  days  of field storage
          and transport.
                                      34

-------
SAMPLE TRACKING
     ID
SAMPLE ID
PROJECT 	

PRECAUTION
ANALYTE(S)
                  SF  01  1950
                  UNFILTERED
                  GAS  TIGHT
                  DET: VOLATILES
                DATE: 03/10/86 -*-
                40 mL GLASS VIAL
                       COLLECTION
                             DATE
                  SF  01  1951
                  UNFILTERED
                  GAS  TIGHT
                  DET: TOC
                DATE: 03/10/86
                40 mL GLASS VIAL
                          STORAGE
                           VESSEL
-+-SF 01 1952
  UNFILTERED
-»GAS TIGHT
  DET: TOX
DATE: 03/10/86
500 mL AMBER GLASS
                  SF  01  1953
                  FILTERED
                  FROZEN
                  DET: N02-
                DATE:  03/10/86
                125 mL POLYETHYLENE
  SF 01 1954
  FILTERED
  0.3% V/V HNO3
-»-DET: Ca Mg Na K  Fe Mn
                                 DATE:  03/10/86
                                 500 mL POLYETHYLENE
                                 DATE:  03/10/86
                                 500 mL POLYETHYLENE
  SF 01 1955
  FILTERED
  UNPRESERVED
  DET: N03 Si02  O-P04  CL- SO4-

  SF 01 1956DATE:  03/10/86
  FILTERED       500  mL POLYETHYLENE
  0.2% V/V H2SO4
  DET: NH3 Fe(II)  t-PO4
              Figure 11.   Typical  set of sample bottle  labels.
                                  35

-------
step  in bottle preparation  was  a complete rinsing with  double-deionized
water  (resistivity  > 17.8 megohms).  Glass  bottles  for  organic  analyses  were
washed as above  and then muffled  at  450°C for 4 hours.   All Teflon-lined
bottle  caps  and septa were  rinsed  in methanol following  the  detergent wash-
ing procedure.   Double deionized  water was used for all  cleaning, rinsing,
blank  and standard  preparation for  the  inorganic parameters.    Double
distilled-in-glass  water  was  used  for  the solutions  and  rinsing  procedures
for organic  parameters.

     Standard solutions for  calibration of inorganic  analyses were  prepared
by  serial dilution with class-A volumetric  glassware.   Stock  solutions  were
obtained  from commercial  vendors  if  available,  or prepared by  dilution of
pre-measured  ampules or  a  high-purity analytical  salt  of the desired  con-
stituent.   All stock  and calibration standards were  verified by  comparison
to NBS or USEPA reference  standards,   as  well as by comparison to previous
lots of standard.   Analytical standard solutions were prepared  for  sample
spiking and  the  control standards prepared for  submission  as "field  stan-
dards".

     When samples were returned from the  field,  they  were separated into
analytical  groups  for spiking.  The method  of spiking varied  with the  set of
constituents  to  be analyzed.  Initially,  for nitrite  samples,  an aliquot of
nitrite  standard  was added to the field duplicate of each  well chosen for
spiking.    This procedure  proved to be inadequate for several  reasons.
Beginning with Run 25,  the  samples were  spiked  immediately  in the  field,
prior to freezing, using a  micropipette and a class-A volumetric flask.    For
sulfide  samples,  a glass microliter syringe was  used to inject  50 (iL of the
spike solution  through the  septum into the  field duplicate  sample of the two
spike-designated  wells.   For all  other inorganic  constituents,  an  aliquot of
spike  solution was   added  directly to the  appropriate  duplicate sample
bottle.   This procedure was  made  more  rigorous after  Run  6, in which a 25 mL
aliquot  of stock spike solution was  added  to 475 mL of sample  in a 500 mL
volumetric flask.   All spiked  sample results were corrected  for the dilution
of the  sample  and spike volumes.

     All field  duplicate  samples from  wells  which were  not designated  to be
spiked or duplicated  were  put into refrigerated  storage.  Samples for atomic
absorption analysis were  not  refrigerated prior to  analysis  because  it was
noted that the samples became turbid.  The working  analytical  dilutions  done
by  air-displacement pipettes were  more  accurate when used on  samples  which
were at room temperature.

     The  sample  processing  scheme used  for the  project  is  illustrated in
Figure 12,  which  includes the  computer databases where the analytical values
were recorded.  The  data processing aspects of the  study will be covered in
more  detail  in a  later  section of the  report.

Sample Tracking

     The  first step in the  sample tracking process  involved  the pre-labeling
of all  sample  bottles prior  to each sampling  trip,  as  described  previously.
All of the information from these labels  was  then entered  into  the  sample
                                       36

-------
           'FIELD-
                                                         • LABORATORY.
i   re
r»
Q
Q
-j
Uj
                                      UNSPIKED
                                        SPIKED
                             COLD STORAGE!
                                        BLANK
                                     INTERNAL


                                f=r\ STANDARDS

v
a.
L
a.
1
a,
'
a.
a,
1
•r-i ,
n i
LJ '
n
Q,
EXTERNAL
REFERENCE
STANDARDS
J— I L

n i
LJ *
n i
LJ '
f*mmm


c
H
E
M
1
C
A
L
A
N
A
L
Y
S
1
S













l* "^""™"
FIELD
DUPS
DATA
BASE
^mmmtm
SPIKE
&
STND
DATA
BASE

1


EXTERNAL 1
STNDS 1
DATA BASE |



*

(••••^
SPIKE
&
STND
DATA
BASE

••••••i^^
1 Exc
» Exc
                                         MEANS





                                         DATA


                                         BASE
     FIELD REPLICATES1
ANALYTICAL REPLICATES2 3
                                                                                     3 Triplicates for TOC. VOC
                           Figure 12.  Project sample  processing  scheme.

-------
tracking system,  which was maintained  as a  Lotus  1-2-3    spreadsheet.  In
this way,  even before Samples were  collected,,  sample numbers had already
been  assigned.    Once the  samples arrived at the laboratory and  had been
sorted and  spiked,  determinations  of the most  sensitive  constituents were
begun.  Each  individual  sample or subsample was  also indexed by a  laboratory
identification  number.   As each  sample  was analyzed for specific  constitu-
ents,  the analyst  noted  the date,  his or  her initials, and  any comments  on
difficulties  relevant  to  that  analysis.  This information also  was noted for
those parameters  which  were measured in the field.   Figure 13 is an example
of the first sheet  of the sample tracking system used  for  this project.

Analytical  Protocols

     The analytical methods  employed in the study are summarized in Table  6.
Most  of the test  procedures  involved the use of well-established  methods,
with  necessary  refinements  or  adjustments only for  the  particular  analytical
instruments  available.

     It  was important to optimize the analytical protocols  and minimize  the
time  for sample processing  given  the variety of parameters and the number of
corresponding  instrumental  methods.   For the  most part,  standard methods
were  used for  the  basis  of each  determination  without major modification
(68,71,72). Ammonia,  methane, sulfate,  and  metals  determinations required
special  consideration.  Ammonia determinations were made after neutralization
of the H2S04-preserved sample by 10  N NaOH by the method of  Ivancic and
Degobbis  (73).     In  comparison with standard methods  this method gave
superior results  over the  wide  concentration range  observed between the
ground-water samples at  the two sites.    Sulfate was determined by  an
automated potentiometric titration  method using lead  nitrate  titrant.   This
was   developed   specifically   for   the   project   because  of   analytical
interferences  in the water samples from the downgradient Beardstown loca-
tion.   Total dissolved metals  were  determined by the  single-solution modifi-
cation  (76) of  the  standard  atomic  absorption  spectrophotometric  method
(68) .   In  this  modification,  lanthanum  (ionization  suppressant)  and  cesium
(releasing agent)  were added  at  concentrations of 5 and  2 g/L, respectively.

     A  list  of analytical instruments  used by the laboratory  for the  study
is given in Table  7.   The analytical aspects of the project  were performed
as would  be expected  in  a  routine  monitoring  program.   Research tasks,  such
as  optimization  of some  of  the  methods for detectability,  precision,  or
minimization of  bias due  to  the sample matrix, could  not  be performed within
the  scope  of  the  project.     However,   some  problems  with  established
procedures that prevented any reliable data  collection were investigated  to
improve method  performance.   These  investigations are discussed in  detail  in
Section 5.

     A  large  part  of the  analytical work load was the overall Quality
Assurance (QA) effort,   which  included the  use of calibration standards,
external  Quality Control (QC)  standards (USGS, USEPA, NBS),  field  standards
and  blanks,   field  spikes,   field  duplicates and analytical duplicates.
Analytical duplicates were performed  on those  samples  and QC  solutions for
which duplication  was possible.   The samples themselves  accounted  for less
                                      38

-------
L«b 10
        Pro] ID   UCLL NO Dttt t»k«n PRCSERV  FILTER
D.O.  *naly«t  Alk   »n»ly»t
FIELD 1 24-Aug-87 UP 0.1 C
FIELD 2 24-Aug-87 UP 0.1 C
FIELD 3 24-AUB-87 UP 0.1 C
FIELD 4 24-AUQ-87 UP 0.1 C
FIELD 5 24-AU9-S7 UP 0.1 C
FIELD 6 24-Aug-B7 UP 0.1 C
703298 SF-0109780 1 24-Aug-B7 VIAL uf
703299 SF-0109781 1 24-Aug-87 VIAL uf
703300 SF-0109782 1 24-Aug-87 AMI GL uf
703301 SF -0109783 1 24-Aug-87 FROZEN 0.1
701302 IF-01097S4 1 24-Aug-87 HN03 0.1
703J03 SF-0109785 1 24-Aug-S7 UP 0.1
703304 SF-01097B6 1 24-Aug-87 H2S04 0.1
703305 SF-0109787 1 24-Aug-87 ZnAc« 0.1
703306 SF-0209788 2 24-Aug-87 VIAL uf
703307 SF-0209789 2 24-Aug-87 VIAL uf
703308 SF- 0209790 2 24-Aug-87 AM CL uf
703309 SF-0209791 2 24-Aug-87 FROZEN 0.1
703310 SF-0209792 2 24-Aug-87 HN03 0.1
703311 SF-0209793 2 24-Aug-87 UP 0.1
703312 SF -0209794 2 24-Aug-87 H2SO4 0.1
703313 SF-0209795 2 24-Aug-87 ZnAe* 0.1
703314 SF-0309796 3 24-Aug-87 VIAL uf
703315 SF-0309797 3 24-Aug-87 VIAL uf
703316 SF- 0309798 3 24-Aug-87 AM GL uf
703317 SF-0309799 24-Aug-67 FROIEH 0.1
703318 SF -0309800 24-Aug-87 HN03 0.1
703319 SF-0309801 24-Aug-S7 UP 0.1
703320 SF-0309602 24-Aug-87 H2SO4 0.1
703321 SF -0309803 24-AU9-87 ZrVkc* 0.1
703322 SF- 0409804 4 24-AU9-87 VIAL uf
703323 SF- 0409805 4 24-Aug-87 VIAL uf
703324 SF-0409806 4 24-Aug-87 AM GL uf
703325 SF- 0409807 4 24-Aog-87 FROZEN 0.1
703326 SF- 0409808 4 24-Aug-87 HN03 0.1
(0/eon/pH/Eh/Ttiip UnfUt»r«d
0/ean/pH/Eh/T«ip Unfilttrcd
K3/eon/pN/Eh/T«np Unfftttrtd
X3/con/pH/Eh/T«it> unfilttrtd
IO/con/pH/Eh/T«rp Unfdtirtd
)0/cen/pH/Eh/Temp unff lured




























24-AU9-8?
24-Aug-87
24-Aug-8;
24-Aug-87
24-Aug-S;
24-Aug-87




























' JK
' JK
' JK
' JK
' JK
' JK




























24-Aug-87
24-Aug-87
24-AUB-85
24-Aug-87
24-Aug-B7
24-Aug-87








_



















JK
JK
' JK
' JK
' JK
' JK



























	
      Figure  13.  Example  of the lead  sheet  of the  sample tracking system.
                                             39

-------
                               TABLE 6.  ANALYTICAL METHODS USED IN THE PROJECT
-tr
O
Parameter
PH
Conductance
TOC/VOC
TOX
Total Alkalinity
Chloride
(N02~ + N03~)-N
30^ =
Si02
o-POij=
t-PO|,E
Ca
Mg
Na
K
Eh
DO
Fe(II)
Fe(total)
Mn
NH3
CHu
S=
N02~-N

Type of method* Std. Methods (71)
E
E
P 505b
APT 506
PT
PT 407C
AS
AS, PT 426D
AS
S
S
AAS
AAS
AAS
AAS
E
RT, E 421B, 421F
S
AAS
AAS
S
GC-FID
S 427
S 419
Reference Method**
USEPA (68) USGS (72) Other***
5
5
415. 1

5

353-2
(see text)
00955
365.3
365.3
76
76
76
76
5

75
76
76
73
74



      * Types of methods:  AAS = atomic absorption spectrophotometric; APT = adsorption-pyrolysis,
           titrimetric; AS = automated spectrophotometric; E = electrometric; GC-FID = gas chromatographic
           - flame ionization detection; P = persulfate, oxidation; PT = potentiometric titration;
           RT = redox titration; S = spectrophotometric
     ** Method number corresponding to, or most closely to, that used in this project.
    *** See reference or text for discussion.

-------
      TABLE 7.  INSTRUMENTATION USED FOR ANALYTICAL DETERMINATIONS
        Parameters
            Instrument
Ca, Mg, Na, K, Fe, Mn
Na, K
Si02,  (N(V  + N02")-N, N02"-N,
S04=  (Wells  1-4)
Cr,  S04= (W_ells  8-13),
Fe(II)  and S~ standardizations

N02"-N, S=,  NH3,  Fe(II),
  0-P04, t-P04
TOG, NVOC, POC
CH4
TOX

All inorganic  calibration
  curves and calculations
GBC 903  Atomic Absorption Spectro-
  photometer (with  D2 background
  correction for Fe,  Mn,  Mg)

Allied  Analytical Systems; Video
  HE  Atomic Absorption Spectro-
  photometer*

Technicon  AAII  system with  Orion
  AS-150  autosampler and chart
  recorder

Metrohm  E-636  Titroprocessor
Beckman DBG UV/VIS Spectrophotometer
  Beckman DU UV/VIS Spectrophotometer
  With Updatel(R)  electronics

Oceanographic International  Model 700
  TOG analyzer

Varian 3740 Gas Chromatograph
  Hewlett-Packard  3390 Integrator

Dohrmann DC-20  TOX Analyzer

IBM PC/XT, IBM PC/AT, or PC Designs
  Ft  286i  microcomputers using
  least-squares  first  or
  second-degree  linear  regressions
  done with RS/1(R)**  software
 * Used as a backup  for some  of the metal  analyses in other runs.
** Trademark; Bolt,  Beranek and Newman, Inc.
                                    41

-------
than  22%  of all inorganic  constituent determinations, with  the  exception  of
sulfide.   A summary of  the entire QA/QC analysis  effort  is  given  in
Figure 14.

Reporting  and Verification of Analytical Data

     Once  a set of analytical determinations was  completed  and  the  analyst
had made  the appropriate  entries  in the  sample  tracking  log,  the  results
were  reported  in  a  summary  which included:

      1.   Calibration  curve with slope,  r2,  intercept and  the  coeffi-
         cients  of the first-  or second-degree  linear  equation fitted
         to the  calibration  points  (sensitivity was included  for  atomic
         absorption and other spectrophotometric  methods),

     2.   Instrument  response  readings  (e.g.,  absorbance,  peak height),

     3.   Dilution  factors,  if any,

     4.   Computed values  for  the calibration standards and  all QC
         samples using  the  regression equation,

     5.   Predicted  concentrations for each unknown, and

     6.   True values for all calibration standards and QC samples,  with
         bias calculated as a percent recovery.

     In order to maintain  complete and accurate analytical records,  each
analyst was responsible for keeping backup  copies of their  results in a lab
notebook,  computer file,  or  both.

     The chemical  data generated from each sampling trip were  recorded  on
master data sheets  according to  sample type (primary water sample, internal
spikes and standards,  field duplicates  and  external  standards).   The compo-
sition of  these  subsets,  from collection  to  data  handling, is  shown  in
Figure 12.

     Replicate analytical subsampling (triplicate  for  some analytes) and the
resulting magnitude of each biweekly data  set required that the project  data
base  be kept  in  the most manageable form  possible.   Thus,   for the primary
water samples   (including the  two  field  blanks), only  mean values were
recorded.  The results of all  field  determinations were  also recorded  with
the "means."  Missing data were marked  with either "NA"  (not available  or
lost)  or "ERR"  (error  in value  obtained).   A complete explanation or state-
ment  of the problem was noted  in a separate data validation notebook.

     The  internal  spike and standard  data subset  contained  all reported
values from the two  field  blanks (called Wells  7A and  14A), the  two field
standards  (called Wells 7B and 14B)  and the two  wells chosen  for spiking (as
well  as their unspiked counterparts).   Mean values were used  in  computing
percent  recovery  of each spike and  standard.
                                      42

-------
 QUALITY CONTROL SUMMARY
 for Na, K,  Ca, Ma, T-Fe,  Ma
                           QUALITYCONTROLSUMMARY
                                 FORCHLORIDES
[\^ cMJVunon ito i;*

QBE ClTIftMM. OC   • «

K2 FCLfl VKCS  1 ««

^| r«LO ^104   « •*

L_J r«L£ 0*^1   i «*

QQ UMPUS     t4 »

^3 1MM.TKM. 0*^1 Ml
                                                                                   OI *M»  OU«   J t<
QUALITY CONTROL SUMMARY
  for  NH3, T-PO4, O-PO4
                            CM0unoN trtn •*«

                                 OC   i J«
QUALITY CONTROL SUMMARY
       FOR SIUCA
                           QUALITY CONTROL SUMMARY
                           FOR NO2-N  & (N03+N02)-N
                        PCI
                                  tin in
                                 oc   i ««
                        g%l noA im   i •«

                        I  I >«LO 0*1   1 ««

                        CO V>»U>    M '«

                        B53 NMLVTK41 0^1 W«
                                                                              CM«N«noM Mm it*
                                                                              IitCMMi OC  J 2
                                                    I JMM.rTCM. 01^1 US
                             QUALITY CONTROL SUMMARY
                                    FOR SULFIDE
                                                               %m m

                                                                 41II

                                                         raiA wti  r >•

                                                         MW1T1CM Oun 7 I*

                                                         r»iaouw   7 ii
     Figure  14.    Summary  of the  quality  assurance  contribution
                      to  the  analytical workload.
                                             43

-------
     Finally,  additional  data  sheets were used to record all values obtained
from the  two  wells  designated as  field  duplicates.    With replicate
subsampling,  as many as four  values could be  reported for a  given parameter.
For  those  analytes  which were  subsampled  in triplicate, only the first two
reported values  were recorded.

     The  original  analytical  results  were filed  after the mean analytical
results  were  recorded onto  the appropriate master data sheets.

     Ion balances for each  of  the  12 monitoring wells were computed using a
modified version of  the  WATEQF  (77)  chemical equilibrium  speciation model.
Data could be  entered  interactively  or  through ASCII files  derived  from the
master  means database.    Data  which  did  not agree  to within  ±  5%  were
examined more closely  to  determine the possible source(s)  of error  (i.e.,  a
particular analysis out of control).  Ion balance  results  (expressed  as a
relative percent) were  then recorded on the  master  data  sheets  along  with
the primary water  sample data.   An outline  of the  flow of  data  as  well as
the relationships among the individual databases is  shown in Figure 15.

     The next  step  in  data handling involved transcription  of  the  data  from
the  paper data sheets  to computer  files.    Again,   the  data were  kept
according to sample type.    Lotus  l-2-3(R)  was  selected as the  database
management  package because  of its  simplicity,  in addition  to its graphical
and  statistical  capabilities.    By converting  a Lotus worksheet  to an ASCII
file,  other  software  packages  could be  used  as well.

     The task  of data verification was an ongoing process, beginning  with
entry into  the  l-2-3(R)  spreadsheets.   Data outliers and unexplained  gaps  in
the  data on the master data  sheets  were  checked against the original
results.   Changes  then were noted in the data validation notebook.  Once the
data had been  entered,  the following checks  were made  to  insure  agreement
among the  various databases:

     1.   Averages  were  computed for  all replicate  values  in  the  field
         duplicates  database and  compared  with  the value recorded  in
         the  "means" database.

     2.   Averages  were  computed for  all replicate  values  of  unspiked
         samples and compared  to those values recorded  in  the "means"
         database.

     Quality control charts were also employed  in the  data  verification
effort.   Reported  values for  each parameter were  plotted as a function  of
time (run number).   Lines also  were drawn  showing  the average and control
limits of two  and three  standard  deviations.   Values which fell outside  of
the  three  standard deviation  limit  from the mean were  noted, as  were
successive values which differed  by  more than two  standard  deviations.   All
notations  regarding data  outliers  were recorded in  the  data validation
notebook. This same procedure  also was  used  for the spike  and standard data
by plotting percent recoveries  on control charts.
                                      44

-------
    FIELD
  DUPLICATES
   PRIMARY
   SAMPLES
    SPIKES
-t     &
^ STANDARDS
    DATA
   ENTRY
     &
VERIFICATION

 (LOTUS 1-2-3)
             —SFDUPS.WK-
-MASTER44.WK-
                                                                                      FORMAT
                                                                                    CONVERSION
                                                                                  (FORTRAN, 1-2-3)
              SR##SPK.WK
              BT##STD.WK"
                                                                                     SFSPKS.WK
                                                                                     SFSTDS.WK
                                                                                                             DATA
                                                                                                           ANALYSIS
                                                                                      COMBINE
                                                                                      ALL RUN
                                                                                      RESULTS
                                                        SFINQCDB.WK
               PAPER DATA
                 SHEETS
                           .COMPUTER FILES,
                             (LOTUS 12-3)
                                     Figure 15.  Data handling flow chart.

-------
                                   SECTION 5

                             RESULTS AND DISCUSSION
FIELD ACTIVITIES

Routine Activities

     Five preliminary sampling runs were completed between  November  1985 and
March  1986.   Then  thirty-nine biweekly  sampling trips were  conducted during
the period of March 10, 1986  through August 25, 1987.   These  field  activi-
ties involved purging and sampling  the monitoring  wells 526  times  and
measuring more  than 2,000  ground-water  levels.   Only  two wells were missed
out of the 528  sampling opportunities,  The 105 ft (32 m) well  (#4)  at  Sand
Ridge was not  sampled on January 26,  1987 because  of a compressor breakdown
which  was complicated by  a  cold weather failure  of an  air  line fitting at
the well-head,  The well was  again operational by the  end of that  week.   The
other instance also  occurred at  Sand Ridge on August  10, 1987.   Faulty air
seals in the pump  of well  #3  could not be corrected  in the  field,  even  after
replacing the  bladder and  all gaskets.   The pump  was returned  to the manu-
facturer  but a replacement was not obtained  until after  the  last regular
sampling trip. Samples were collected from well  #3 on  that  last trip using a
PVC pump.

     Despite the replacement of one  of our most  experienced field  and
laboratory personnel and the need to  move  our entire laboratory operation
twice,   the 14-day interval between sampling dates  was maintained,  although
Run  #18 was delayed by one day due to an engine  problem with  the  sampling
van.

Major  Difficulties

     Although  a near-perfect sampling record was established over  the course
of this  project,  minor  problems invariably  occurred in the  field.  Broken
probe/meter cables,  worn  tubing  fittings, dead  batteries,  leaking pump
bladders,  erratic monitoring probes and other such difficulties were, for
the most part,  quickly solved.   The van was stocked  with the proper  tools
and  adequate replacement  parts  and  supplies.   A backup  was kept for nearly
every piece of equipment,  from pump assemblies  to  pipettes.   Even with  regu-
lar equipment  maintenance,  however, many of  the  backups were used at one
time or  another.

     The most  significant  problem regarding  field equipment  involved the
mobile van's air compressor system.   From the  time  of the  first breakdown,
mentioned in the preceding section  (January  26, 1987),  the compressor ran
rough  and erratically,  often  tripping  its circuit  breaker.   The compressor
was down through mid-August and during this time a  portable  backup was used.
                                        46

-------
A  new pressure  switch (the second one  during the  project period) was
installed  in  March,  but that did  not  improve the  compressor performance,
The  compressor  itself was  finally  sent  out for servicing in June but  no
problems were  diagnosed.  No further problems  occurred  after the  unit was
reinstalled in June  1987.

     Well temperature data  from July  through September of 1986  were not
obtained  due to  a  faulty thermistor probe and  cable.    The  length  of new
cable required (140 ft,  43  m) had  to be  back ordered, causing the inordinate
gap  in the data.

     The  battery operated  digital buret  used in the field  for alkalinity
titrations was sent out  for  repair during  the  period of  November through
December of  1986.     During this  span,  the Sand Ridge  and  upgradient
Beardstown  samples were titrated in the  field using an Eppendorf  micro-
pipette  (100-1000  microliters).     The downgradient  Beardstown  samples,
however,  required  a titrating resolution less than  0.1 mL.  Therefore, these
samples  were  refrigerated upon collection  and titrated  after  return  to  the
laboratory.

     Problems with the  calibration  and  response of the  electrometric dis-
solved oxygen probe  resulted in  missing or partial data  on 8 of  the  39
biweekly sampling  trips.   This problem  did not  present  itself in a specific
period,  but  appeared  sporadically  over  the entire project period.  Al though
the  membrane/electrolyte modules were  replaced when necessary,  the  age of
the  probe  assembly itself was probably  the largest  factor causing  this  dif-
ficulty.


LABORATORY ACTIVITIES

Routine  Activities

     The  laboratory activities generally went  very well despite  the fact
that the  entire laboratory  operation was moved  twice  during the  sampling
phase of the project.  This achievement  was  due to the dedication and
commitment  of the  project  staff.    They put in frequent night, weekend,  and
holiday  hours to complete the  work from  one sampling run prior to arrival of
samples  from the next  run.   More  than  95% of the individual analytical
determinations were completed  within  the prescribed Agency holding and
storage  periods. Exceptions  to this  level  of performance were  documented in
a logbook.

Difficulties
     A  number of analytical  and instrument difficulties  were encountered
during  the  project,   In most cases,  alternative  instrumentation or  methods
(i.e.,   manual vs.  automated) were  used to complete the work in a timely
fashion.   Analytical  problems were  more  difficult to deal with, particularly
due  to  the  differing  background  matrices represented in water  from Sand
Ridge versus that  from Beardstown.
                                      47

-------
     Foremost  among the analytical  problems was  the inability of the auto-
mated  methyl  thymol blue procedure  for  sulfate to yield  measurable  results
for many of the downgradient Beardstown samples.    The  interference mani-
fested itself  as a  sudden  drop in  the spectrophotometer  signal  baseline
whenever downgradient samples were  introduced  into the autoanalyzer reaction
train.    The peak  heights of affected  samples  were inconsistent  with those
from  either  the calibration  standards or quality  controls.   This meant that
even  if the baseline  could be brought back  on scale,  quantitative  interpre-
tation was  impossible.   The interference  presumably  was  due  to  the  organic
contaminants in that  system.  Dilution could  not be used to  reduce the inter-
ference because  the sulfate  levels were too  low to remain  detectable.   The
interference  did not  affect  all downgradient wells  to  the  same  degree, and
it was  inconsistent  over  time.   Attempts  were made to remove  the interfer-
ence by  oxidative pretreatment with hydrogen  peroxide  and  by adding a carbon
adsorption column  in line, but  they  were unsuccessful.  Several  alternative
automated sulfate  procedures could  not be  investigated because  of cost and
time  constraints.   Tests  of  a  manual  and automated turbidimetric  method
revealed  poor precision and  accuracy  for the Beardstown downgradient sample
group,  so they were  not used.

     These  problems led to the development  of a potentiometric titration
procedure for  sulfate,  which employed the  Metrohm E-636  Titroprocessor for
analysis  and  endpoint  detection.   Calibration  was done  by analyzing stan-
dards containing various amounts  of  sulfate covering  the concentration range
of the samples.  The titration  results  were  used to compute the  sample sul-
fate values  by  linear regression calibrated  on  sulfate standards  (78).

     A method  that  combined attributes  of  several  published procedures was
developed  (78,79,80,81,82) to  avoid  serious inorganic  interferences  to
potentiometric  sulfate techniques (such as orthophosphate,  copper,  and hard-
ness ions).   The first step  in the procedure was  the  addition of lanthanum
nitrate solution to  the  aliquot  of sample  taken  for analysis,  in  order  to
precipitate  orthophosphate.   The  treated sample was mixed  and then held for
ten to fifteen  minutes until precipitation was  complete.   It was then vacuum
filtered through a  0.45  (im  pore-size cellulose  acetate membrane  filter.   The
samples  were  then  passed through a  strong acid cation exchange column  (i.e.,
Dowex  50W-X8; 20-50  mesh).   Then 25  mL of sample was mixed with an equal
volume of methanol  solution,  which  contained 3 to 4 drops per liter  of
formaldehyde as  an antioxidant.    If  the  sulfate concentration  of  the  sample
aliquot was  below about 0.5  mg  S047L,  the solution was spiked with  2 mL of
1000  mg/L sulfate  standard.   One  drop of bromcresol  green/methyl  red  indi-
cator  was then added  to the sample,   and  the  pH was adjusted to approximately
7  with 2 N  NaOH.    Approximately one drop  of  10% volume/volume  (v/v)
perchloric acid  solution then was added,  reducing  the pH to between 5.5 and
6, and the  sample was  titrated with 0.010 M lead  nitrate solution.   The
indicator electrode for the titration was a  PbS  solid-state ion-selective
electrode (ISE) and  the reference  electrode  was a double-junction  electrode
with a  10%  KN03  outer filling  solution. Frequently,  the  lead  electrode had
to be cleaned between titrations with  a mildly abrasive polishing strip.
                                       48

-------
     The  ammonia analysis method was another source of analytical  problems.
Although  samples  were preserved  with  0.2% v/v  H2S04 in the field per USEPA
recommendations  (69),  the standard manual  ammonia procedures (i.e.,  indo-
phenol blue method, etc.) were  designed  for unpreserved  samples.   Hence  the
procedure  did not  contain an adequate pH buffer for the optimum pH 01  the
color-forming reaction  (i.e.,  pH 10.5-11).    The high alkalinity and  the
presence  of the sulfuric acid  caused  erratic  results  which could not always
be  identified by  the external standards, field spikes, or field standards.
Manual neutralization was performed before reagent  addition to  overcome  this
problem,  but  the  test was very sensitive  to the performance  of the analyst.
Problems were particularly  serious  during Runs  18-22, and  in  the five
preliminary   runs.     There  were  additional   difficulties  in  obtaining
ammonia-free double deionized  water due to ambient contamination from
laboratory operations and the influence  of nearby sheep,  swine, corn,  and
soybean farming  operations  on  the outside air  ducted into  the  laboratory.

     Nitrite-N analyses were performed   initially  by  a  manual procedure
(Table 6), but time and  staff considerations prompted an attempt to  use  the
standard USEPA automated procedure (68).   This method was used from  Runs 23
to 32.  Inconsistent  instrument  performance  and  poor  detection  limits rela-
tive to  sample levels  prompted a  return to the manual procedure at Run  33.
Concentrations in  the samples were  uniformly  lower than 0.05 mg/L which is
very  close to the reported detection limit.    During the  course of  trouble-
shooting  the  nitrite-N problems,  it was  found that  once  the  frozen  samples
were thawed,   the   nitrite-N  levels quickly deteriorated. Therefore,  samples
could not be  refrozen if there  were analytical problems, and  once the dupli-
cate samples  were used  (which  did  not  allow  for backups  for  the field spike
and field duplicate samples),  further  reanalyses  were  not  possible.

     The  preservation technique for  the sulfide  samples  was  tested  using
standard  solutions before  the  initiation of  field  collection.   The  analyti-
cal results showed that the preservation  technique was  effective  and gave
reliable quantitative  recoveries  for at  least  two  days  after  collection.

     Titrations of  samples  from the Beardstown sites with the highest alka-
linities  showed that  more than  1.5  mL of concentrated nitric acid  might be
necessary to  lower the pH of the samples to  pH  <2  on  many occasions (69).
Therefore,  a  concentration of 3  mL/L  (0.3% v/v) nitric acid was used for  the
metals samples.   Tests also  were  made of the accuracy of  total iron determi-
nations using sulfuric acid  preserved  samples against standards  made  up in
nitric  acid.   The tests showed that  the results would be in error by a
factor  of  approximately  2 to 3.   All  atomic  absorption  standards were made
up  in this acid  solution  in order to match the  matrix of the  samples.

     Staff analytical task  assignments  were  made in order to attain  the
holding-time  goals (Table  5) and to  avoid  conflicts in instrument usage.
Analysts  had  backup  assignments to  cover absences  of primary responsible
persons and  maintain the  necessary  turn-around  time. Following Run  22,
weekly  or biweekly  Analytical Support Group Meetings were  introduced in
order  to  review   all data  as a  group  and  to  try  to resolve analytical
problems.   Problems  could not always  be  resolved immediately for the current
or  immediately following runs.    The meetings were, however,  important in
                                      49

-------
identifying the  generation of potentially bad  data from  sources such  as
spike or   standard   preparation   problems,   building   water  system   and
contamination  problems,   test interferences,  or  instrument  malfunctions.
These  sources  of error  were  resolved  routinely before tests became  out of
control for  extended periods.


DATA QUALITY EVALUATION

Method Performance

     There are numerous ways in which to assess the performance  of the ana-
lytic procedures,  using different aspects  of the  overall  QA/QC program.   The
detection limits attained  in  practice  were  computed  by pooling  replicate
analytical  results   of the  lowest  calibration  standard  used  for  each
analytical run.   An  additional low  standard was  analyzed  in duplicate  for
this procedure.   For trace  constituents (iron,  manganese,  sulfide,  etc.),
the lowest calibration standard was  close to the actual detection  limit,  and
was easily within  the factor  of 10  suggested by the American  Chemical
Society (ACS)  procedure  (83).   For  the major constituents,  whose  concentra-
tions  were higher than  the routine accepted detection  limits  by  a factor of
100 or more,  no attempt was  made to  optimize method  or instrumental sensi-
tivity.    In  these  cases an extra  low  calibration  standard  was  included to
compute  an approximate  detection limit  by the  ACS  procedure  (83).

     Detection limits  were computed for  several randomly  distributed runs
before pooling the data  from the low  calibration standards  over  all of  the
runs.   This was done  to  see if  detection limits  were  consistent for  the
duration  of the study.   That assumption was  valid.

     The  detection limits  computed for  various inorganic  constituents  by  the
above  procedure are presented  in  Table  8.    For  spectrophotometric  proce-
dures,  the sensitivities  (the concentration  producing  0.0044 absorbance
units)  were computed  for  each calibration to monitor method  consistency. In
accordance with the ASTM procedure  for  the  reporting of  low-level data (84),
all  computed  concentrations were recorded in the  database.   This was done
rather than following  common practice  of censoring the  data  distribution by
reporting low values  as "less than" a detection limit  or as  zero.  No
detection limits are reported for some  analytical  determinations, such as
Cl",  potentiometric  S04=,  and total alkalinity,  where the detection limit is
as much  a  function  of  sample  size as instrumental performance  factors or
reagent concentrations.

     Three  different  types of quality control  samples  were used in this
investigation  as a control for  methodological accuracy.     These control
samples  were   "blind"  or  field  standards,  spiked  samples,  and  external  refer-
ence standards.

     A summary of the mean percent  recoveries and  one  standard deviation  for
"blind" standards  is presented in  Table 9.   These standards  were  prepared by
field  sampling  personnel,  as described  in Section 4.   A  separate standard
                                      50

-------
 TABLE 8.   DETECTION LIMITS FOR SELECTED INORGANIC CONSTITUENTS
COMPUTED FOR REPETITIVE ANALYSES OF LOW CALIBRATION  STANDARDS
            Analyte              Detection  limit  (mg/L)
NH3
(N02~ + N03~)-N
N02~-N
Sulfide
Sulfate (autoanalyzer)
Si02
0-P04
t-POij
Ca
Mg
K
Ma
Fe(total)
Fe(II)
Mn
0.012
0.030
0.002
0.003
1 .5
0.98
0.008
0.009
0.21
0.10
0.047
0.30
0.032
0.015
0.013
                               51

-------
was  prepared for each type of preservation method (e.g.,  atomic absorption
metals,  sulfate  +  chloride  + nitrate +  orthophosphate,  etc.).    Two  field
standards   were   submitted with  samples   for each run,   one  representing
concentrations similar to wells  #1-6  and the  other similar to  wells  #8-13.
Computations   of the  mean  accuracy for the field standards  in Table  9
indicated a  hypothesis  of 100% recovery could  be  accepted  at the 95%
confidence interval for all  constituents  except calcium,  potassium,  ortho-
phosphate,  and  nitrite-N.

     The mean  percent recoveries and standard deviations for  spiked samples
are summarized  in Table 10.    Like  the field standards, spiked samples  were
prepared  by  field personnel.    A  sample  from  one  well for   each  of the
collection  days was  selected to be  spiked with a known laboratory standard
solution.   This procedure provided a total of two  spiked samples per  run.
The  preparation of the  spiked  samples was described in Section  4. The
apparently large relative  standard  deviations  resulted in part because  com-
puted recoveries were the product of the  two field standard mean values.
Each of the  means  included  a certain level of analytical  imprecision.  Also
each  spiked  control sample was  a field duplicate  of  the unspiked  one.
Standard error propagation formulas were used to monitor  the reasonableness
of the percentage errors   given the analytical data for each  run.   Then the
error values  were incorporated into the spike  and  standard spreadsheets. By
this  criterion,  all  constituents  could  be considered  to represent 100%
recovery at  the 95%  confidence  interval,  except for  total iron (atomic
absorption)  and  total sulfide  which  tended to be  systematically  low.   The
increased variation in  iron values was a problem because it was  necessary to
compute  Fe(III)  by  difference, using  the  Fe(II) analytical result and  total
iron concentration data.

     The third  type of control samples  used  to assess  the accuracy  of the
analysis methods  were  USEPA,  U.  S.  Geological Survey (USGS)  and National
Bureau of Standards (NBS) reference standards.  The USEPA standards  contain
a known mass  of the analyte of interest and were diluted or  concentrated
(made up in  less volume) to  be  more similar to project samples. The  NBS
trace  metal  standards  have "certified"  reference values.  The U.S.  Geological
Survey "Standard Reference Water  Samples   have mean  concentrations and  stan-
dard  deviations  computed from  a  critical  evaluation of laboratory  results
from  "round  robins"  of participating  analytical  laboratories.

     A summary of the percent  recoveries  and relative  standard  deviations
for  the  external  reference standards that were included  in replicate in  each
analysis  set is  presented in  Table 11.    These samples were analyzed  more
than  10 times over  the course of the sampling and analytical period of the
project.  The accuracy and precision of analysis  of these  standards  were
excellent.    This  was true  even for  iron and manganese concentrations  near
the detection limit  of the instrumental  procedure.   External  stable stan-
dards do not  exist,  or are of  questionable  validity  (due  to preservation and
stability problems),   for   several constituents of interest, such  as ferrous
iron,  sulfide and  nitrite-N.    These  standards  generally  showed  higher
accuracy  and  better  precision  than  did the  field  standards and  spikes.  Part
of this  result  may  be attributed to the consistency  of preparation  of the
control  solutions by laboratory  personnel.  Matrix  variability  in  the samples
                                      52

-------
       TABLE 9.   SUMMARY OF THE MEAN ACCURACY  AND PRECISION
           (ONE  STANDARD DEVIATION)  OF FIELD  STANDARDS
                         (Expressed in Percent)

Overall
Par am.
NH3
T-P04
Fe(II)
N02 -N
Sulfide
(N02- + N03 ) -N
Si02
o-P04
Cl
S04 =
Ca
Mg
Na
K
Fe
Mn
Ace.
95.90
99.64
96.07
82.17
NA
100.35
99.47
103.44
105.78
95.77
98.36
99.15
101 .69
97.85
99.22
101.04
Prec.
23.49
8.60
18.80
36.29
NA
10.27
5.03
15.38
32.59
21.85
3.88
8.70
12.17
5.17
5.80
6.46
Sand Ridge
Ace.
91.99
100.95
NA*
81.07
NA
98.85
100.21
106.54
112.01
94.73
98.65
99.90
103.51
99.10
100.34
101.28
Prec.
29.80
9.28
NA
35.00
NA
7.82
2.97
20.77
46.55
6.58
3.76
10.72
16.16
5.15
7.20
8.17
Beards town
Ace.
100.09
98.24
96.07
83.27
NA
101.97
98.71
100.12
100.18
97.24
98.07
98.42
99.95
96.63
98.04
100.79
Prec.
12.54
7.56
18.80
37.50
NA
12.17
6.41
2.32
1.52
33.07
3.98
6.03
5.87
4.89
3.46
3.92

No field standard  used due to low concentration in monitoring
wells.
                                53

-------
     TABLE  10.   SUMMARY OF THE MEAN ACCURACY AND PRECISION
(ONE  STANDARD DEVIATION)  OF FIELD SPIKES  (Expressed in Percent)

Overall
Par am.
NH3
T-P04
Fe(II)
NCV-N
Sulfide
(N02- + N03')-N
Si02
o-P04
Cl
s04 =
Ca
Mg
Na
K
Fe
Mn
Ace.
98.08
99.44
102.82
101.29
90.32
104.74
97.09
101.75
111.86
105.05
105.43
96.80
96.47
96.88
86.61
95.99
Prec.
28.32
25.11
27.59
26.55
13.97
28.06
14.19
30.11
59.07
43.92
41.46
17.22
31.41
13.02
17.01
12.16
Sand Ridge
Ace.
97.62
97.78
102.12
102.44
89.08
100.59
99.00
100.54
123.71
94.02
101.44
95.85
87.79
96.20
86.08
96.92
Prec.
31.52
29.48
18.18
25.83
16.20
13.04
11.50
6.12
72.06
20.63
15.39
14.66
19.76
12.57
19.85
12.80
Beardstown
Ace.
98.54
100.99
103.33
100.23
91.44
108.70
95.41
102.88
101.95
118.57
109.11
97.67
104.93
97.55
87.12
95.07
Prec.
24.80
20.08
32.75
27.15
11.48
36.66
16.01
41.38
42.97
58.66
55.30
19.25
37.71
13.41
13.67
11.42
                               54

-------
  TABLE 11.  PERCENT RECOVERIES AND RELATIVE STANDARD DEVIATIONS OF EXTERNAL QUALITY CONTROL STANDARDS.
                                        (Concentrations in mg/L)
Standard
USEPA 882-1 (+10)
USEPA 882-1 (+2)
USEPA 882-2
USGS M-76 (+25)
USGS M-86 (+10)
USGS M-86
USGS M-2 (+10)
USGS M-2
USGS M-84 (+10)

-------
                                             TABLE 11.  (CONTINUED)
un
cr>

Fe (tot)
Standard
USEPA 281-1
USEPA 281-2
USEPA 281-2 (x2)
USEPA 281-2 (xl)
USEPA 281-2 (+2)
USEPA 386 (xl)
TV
0.021
0.797
1 .580
3.188
0.399
0.100
% R
112.1
99.5
99.1
99.0
99.1
99.6
RSD
51 .28
2.10
3.09
3.28
1.66
1.00
TV
0.0129
0.318
0.696
1.392
0.171
0.100
Mn
* R
102.6
100.6
102.1
100.9
100.5
102.9

RSD
15.6
2.35
2.61
3.12
10.76
3-38
                                                                      continued  on next page

-------
    Standard
                                             TABLE  11.   (CONTINUED)
                             cr
                     S0i4~ (autoanalyzer)     SO^-  (potentiometric)
TV    $ R   RSD
% R   RSD
                                                        % R  RSD
                                                                              Si02
TV     % R   RSD
USEPA 882-1 (+10)
USEPA 882-1 (+2)
USEPA 882-2
USGS M-76
USGS M-76  (+25)
USGS M-86  (+10)
USGS M-86
USGS M-2 (+10)
USGS M-2
USGS M-84  (+10)
USGS M-84
USGS M-88  (+20)
USGS M-88
USGS M-100 (+25)
USGS M-100
USGS M-82
USGS M-54
USGS M-6
USGS M-2 (+2)
USEPA 384-1
17.8  105.0  it.37



1414.14   qq.H  14.52

32.0  100.n  2.77



37.1   99.8  2.33

79.0  101.9  0.63




fin.fi   99.6  1.29
                       23.5  93.9   14.0
                      3*4.7  98.5     3.53

                      27.3  95.60    5.52
                                               74.5
                                              105.4
                                               95.1
                         99.0  3-33
                         94.6  1.29
                         95.5  1.80
                                                                         q.76   108.8  5.03
                                          13-3
                                                                               98.6  2.27
                                                                        11.2    102.4  3.31
                                                                                           concluded  on  next  page

-------
Co
TABLE 11 . (CONCLUDED)
NH3 (N02~ + N03~)-N
Standard
USEPA 284-1
USEPA 284-2
USEPA 284-2 (+2)
USEPA 284-3
USEPA 284-4
USEPA 284-5
USEPA 284-6
USEPA 284-7
USEPA 486-1
USEPA 486-3
TV
1.84
—
0.340
2.310
—
—
—
2.43
—
% R
101 .0
—
96.4
103.4
—
—
—
109.3
—
RSD
12.70
—
3.65
11 .24
—
—
--
14.02
--
TV $ R
0.18 94.8
1.60 98.5
0.14 103.5
1.43 99.2
—
—
—
—
—
RSD
5.90
3.83
5.24
4.35
--
—
—
—
—
TV
--
—
--
—
0.399
3.159
0.303
--
—
t-POij
VR
--
--
--
--
92.9
89.5
89.4
--
—

RSD
—
--
--
—
4.68
7.24
2.99
--
—

TV
0.83
--
0.15
1 .07
--
—
—
1.53
	
o-P04
% R
103.4
--
99.2
99.3
--
—
—
98.7
	

RSD
7.93
—
7.24
1 .05
--
—
—
2.3
	

-------
contributed  to  the  poorer precision of the spikes.   Sources of the  slight
discrepancies between the variability  in the external reference standards
and  the  field standards include differing reagent sources  and  techniques  for
preparation of  the  control   solutions  from  outside  agencies  and  our
laboratories.

     The external QC standards covered a broad  concentration  range.   Plots
of observed  standard deviation versus  true  or reference  concentration were
evaluated by unweighted first-degree  linear regression  to  see  if there  was a
trend in precision with concentration.   Significance tests for the slope of
the  regression line were performed  with the  RS/1(R) statistical  and
graphical  software  package.   A summary of these results and  the  concentra-
tion  ranges covered  for each analyte are shown  in Table  12.   Note that in
most  cases,   the equations of  the  lines only  give  an approximation  of the
standard deviations  of a given concentration because  of inherent  regression
uncertainty.   A statistical summary of the chemical  constituent concentra-
tions over the  study  period for  each  well is provided in Appendix A.
Detailed time-series  graphs of the actual mean data for  each well are  pro-
vided in Appendix B.

     Another way used  to  estimate analytical method precision  was to  pool
the duplicate laboratory analysis  data from  the database of field  duplicate
samples  (84,85).   Outliers were removed by application  of  the  Grubbs  test
(86).   The results  are presented in Table  13.   This is a reasonable approach
when  samples are grouped by  similarity  of matrix and  approximate  concentra-
tion.    It could be  used  further  to  estimate  the  variance attributable to
sampling when combined with  the  overall sample  variance.   For Eh and total
alkalinity,  the  precision was  calculated  using all  analytical data.

     In general,  small variations  in accuracy and precision were  randomly
distributed throughout the  sampling and  analytical  period.  The  major  excep-
tion  was ammonia,  which suffered  from a  particular  period  of  probable
inaccuracy.   Some of the  quality  controls  also tended  to  support  a low  bias
in the  atomic  absorption  iron determinations  and in the  determination of
total  phosphate.   The latter problem  may be due in part to the difficulty in
properly controlling  the acidity of digested samples and standards  that  had
been  acid-preserved  because  the  method  is  extremely  sensitive to sample
acidity.

     Analytical  performance was verified  to be  within acceptable  limits  (or
samples  were identified for reanalysis) by several  techniques:

     1.   Comparison  to previous data  from the same location

     2.   Ion balance  computations

     3.   Computation of alkalinity  from  the  major ion  composition

     4.   Examination of QC  and/or calibration  data  for  each ana-
         lytical  run
                                      59

-------
          TABLE 12.  ANALYTICAL PRECISION ESTIMATES FROM EXTERNAL
         	REFERENCE QUALITY CONTROL STANDARDS	
    Analyte
 Calibration
range (mg/L)a
Equation
A       B
P(slope)c   pooled sd
NH3
T-P04
(N02~ + N03~)-N
Si02
o-POjj
Cl~
S0i|= (AA)
S0i4 = (AA)
S0i4= (P)
Ca
Mg
Na
K
Fe
Mn
Mn
0-3
0-3-5
0-2
5-15
0-1 .6
0-100
10-40
50-550
10-110
0-7.5
0-7.5
0-15
0-5
0-5
0-0.5
0.3-1 .4
0.1464
0.0768
0.0197
—
—
--
--
0.0587
—
0.0845
0.0115
0.0239
0.0363
0.0309
--
0.0350
-0.0381
-0.0125
+0.0067
—
--
--
--
-4.0845
—
-0.2695
0.0396
+0.0342
-0.0071
+0.0046
--
-0.0023
0.020
0.004
0.013
0.59
0.74
0.71
0.48
0.02
0.57
0.078
0.099
0.051
0.001
0.001
--
0.005
	
—
—
0.41
0.046
1 .09
1.23
--
1 .89
--
--
--
--
--
0.012
"

a All  samples were  analyzed  in  this range, after  appropriate  dilution.
     For example, 70 mg/L calcium  sample would be analyzed after a
     tenfold dilution, so the actual analysis precision would  be 10
     times that computed  with  the  equation  or  pooled  s  given  here.
b Precision  (Sest)  =  AC  +  B,  where C is the analyzed concentration.
c Significance of the slope (0.05  or smaller is  highly  significant)
d For analytes showing no significant dependence  of analytical precision
     on  concentration, this is  the precision estimate using all
     standards.
                                    60

-------
     TABLE 13.  POOLED ANALYTICAL PRECISION BASED ON REPLICATE
         LABORATORY ANALYSES OF FIELD DUPLICATE SAMPLES.*
              (STANDARD DEVIATION VALUES IN MG/L)

Parameter
NH
T-k>4
Fe(II)
N02~-N
(N(V + N03 ) -N
Si02
o-P04
cl"
S04 =
Ca
Mg
Na
K
Fe
Mn
TOG
TOX (micrograms/L)
VOC
Total alkalinity
Eh (volts)

(1-4)
0.004
0.006
0.02
0.00
0.006
0.59
0.005
0.11
0.92
0.70
0.22
0.05
0.01
0.01
0.003
0.03
1.2
0.002
3.59
0.020**
Wells
(5-6)
0.006
0.004
0.11
0.00
0.00
—
0.009
0.16
- -
0.67
0.33
0.86
0.05
0.04
0.080
0.02
1.9
0.00
1.08
0.011**

(8-13)
4.9
0.36
0.02
0.00
0.00
0.08
0.35
0.62
1.2
0.95
0.33
1.87
0.50
0.04
0.008
0.12
3.2
0.18
6.30
0.019**

 * Pooled on the basis of matrix  similarity
** Wells  1-3,  4-6,  8-13,  respectively
                                61

-------
     5.  Computation of pH   with   the   geochemical    computer
         equilibrium model PHREEQE,  using  analytical concentrations
         and  historical  saturation  index  information as   inputs
         (87).

Questionable  data  were  noted in a  log of  sample verification notes when
reanalyses provided data that still seemed  to be  unusual, or  when  reanalyses
could  not  be  done  (e.g.,  pH,  Eh,  N02"-N,  sulfide).   For the purpose of this
project,  the remaining  questionable  values  were not deleted,  although they
might  be in future studies of the  geochemistry of the test sites.

Sample Data Quality

     During the course of the study,  more than 55,000 analytical  determina-
tions were  made on blanks, standards  and samples.  The final dataset was  96%
complete,  that is,  96% of the  maximum possible number of samples and  sub-
sequent  analytical  determinations were successfully  completed.   Outliers
were  screened  successively at ±3  and ±2  standard  deviations  from the mean
levels.  In  moat cases,  this  screening  revealed apparent  errors  in  calcula-
tions,  calibration,  or data  entry which were corrected prior to data  analy-
sis.   For  all  wells and  constituents,  the maximum number of samples which
were identified as possible outliers and  for  which no  documented  error  was
identified  was  four  percent  of the total.   No adjustment was made to  appar-
ent outliers for which no  documented error  could  be identified.   Some other
screening methods are given  in the immediately preceding section.

     Mean  ion balance errors for  the  complete  dataset  are  reported in  Figure
16.  The  error bars  represent  one standard deviation from  the mean.   The
mean  error  for all  wells  was  leas than  ±2%.    There  tended to be  a  alight
positive  bias in  the  Beardstown  downgradient  samples.    This may reflect
complexation of divalent metals by  unidentified organic  ligands or  the in-
fluence of  ammonia  determinations that  appeared to  be erroneously high for
several runs,   It is,  important to note  that the project analytes  appear to
cover  virtually  all  charged  species present  in significant concentrations
and that the  average  performance on Beardstown downgradient samples  was
comparable  to that  for the wells  in  the  uncontaminated environments.  The
larger  standard deviations for  the downgradient wells (#8-#13) most  likely
reflect  the  imprecision in the ammonia analyses  (Tables 9  and  10) from  run
to run.  This  effect is also  evident  in the  larger standard deviations  from
wells #5 and #6, at Beardstown relative to  the Sand  Ridge  results.   Since
Na+, Ca++,  S04~ and Cl" made up  a  large  proportion of the total ionic solids
in  samples  from  wells #5 and #6  the  analytical precision  alone  translates
into a higher  relative error in the ion balance.

     Analyses  were  performed on  the  quality assurance/quality  control
(QA/QC) data  to  determine if there was  evidence  of temporal  variability in
the blanks and  standards or  field standard recoveries and to determine  if
there  was  a concentration-related  dependence  in  the  accuracy or precision  of
these  datasets.   A time series was  constructed  for the field and  laboratory
QA/QC samples  (i.e.,  field  standards  and  blanks).    Missing values were
removed to  create  a  complete  series of  lower total  sample size.   A  linear
regression  was  then fit to the  series  with time  as the  independent variable.
                                       62

-------
1U.U:
8.0-
6.0-
o
t 4.0-
LJ
o) 2.0-
0
J 0.0-
m -2.0-
c :
o -4.0-
^ -6.0-
-8.0-
1 n n .

-



| Y
• i i i i I
;'TTiI;





4
1





' <




> ^

•



| 1




1 1





1 «


"

1

:



i 	 1 	


1 	 1 1

1   2   3   4  5   6   7   8   9  10  11  12  13  14
                      Well
  Figure 16.  Ion balance error aumnary.
                63

-------
The significance of the  slope (time  coefficient) was tested  at the  0.05
significance  (95% confidence  level).   No  cases were found in which the  slope
was significant.   This result was somewhat  surprising since  the expected
number of rejections of the null hypothesis  (that the  slope is zero)  is  0.05
times the number  of analyses, or approximately  seven.  This result  may be
explained  by the  lack of correlation  in the  data (especially  blanks) and
perhaps by  nonnormality or outliers,  all  of which may  make the  true signifi-
cance level less  than the nominal value of 0.05 and hence,  reduce the  number
of  rejections.    The  test  should be  powerful enough,  however,  to  detect
trends  over  long periods  of  time in  the  laboratory  and  field  procedures.
The  absence  of  significant  trends may be  taken  as evidence that  there  was  no
significant  temporal trend  in the blanks and  standards  which  might bias the
sample  dataset  over time.

     The  procedure outlined above  also  was  applied to the  sample  percent
recoveries for  three  groupings of the wells:     1) the uncontaminated  Sand
Ridge site  (wells #l-#4), 2)  the  upgradient wells  at the Beardstown contami-
nated site  (wells #5  & #6)  and 3)  the  downgradient wells at this site  (wells
#8-#13). Time series  were  formed  for  each of two  sets of determinations  for
each chemical for each group  of wells.   The  first reflects laboratory error
only (i.e.,  two replicate  determinations  were performed  on  samples spiked
within  the  laboratory),  while the second incorporates both  laboratory  and
field error (two replicate  determinations  were performed  on actual samples).
Separate time series  also were constructed from the mean, median,  and stan-
dard deviation  of the  percent recovery data.

     No significant trends  were found in field  or laboratory recoveries  for
any chemicals  in  the Sand  Ridge wells.  Significant increasing  recovery
trends  with  time in the mean, median,  and standard deviation of the  percent
recoveries were found in  NH3  (laboratory  and field), S04~  (field only),
Fe(II) (laboratory Only) and MnT  (laboratory  only)  for  the Beardstown  wells.
The  apparent recovery  trends  were due largely to difficulties  encountered in
the five preliminary  sampling  runs of the study  between  November 1985  and
March  of  1986  when  the biweekly  sampling began,   When these  early  samples
were  eliminated, the  time  trends were  no  longer  statistically  significant.
As  in the blank and  standards analysis, the  number of significant trends  was
less  than  would be expected baaed on chance.

     The  average overall analytical accuracy  and precision  values  for the
chemical constituents were  given in Tables 9,  10,  12, and  13.   The levels of
overall accuracy and  precision achieved  during  the project  were routinely
within  acceptable  limits  for  the  laboratory  analytical methods (Table 6).
Exceptions to  this level of performance  were noted for constituents  which
were consistently present at or  near analytical detection  limits  (e.g.,
precision of Fe(II),  sulfide and  C1-).   The QA/QC data analyses demonstrate
that the  sampling and analytical  protocols  employed  in the  study were in
control.
                                       64

-------
CHARACTERIZATION OF GROUND-WATER HYDROLOGY

     In  addition  to chemical determinations,   measurement   of  hydrologic
parameters  was performed  to  provide a  basis for the interpretation of varia-
bility  in the ground-water   chemistry due  to  hydrologic processes.   The
hydrologic data collection  was not  an  original  objective  of  the study,  but
it  is a necessary  part of the benchmark  dataset.   It  will provide a basis
for future interpretations  of  coupled hydrologic  and geochemical  processes.
Ground-water elevation  measurements  in  piezometers  and sampling  wells were
the principal  data collected at the two sites.   Ground-water level  hydro-
graphs and  contour maps were prepared  from these data  to  provide  information
on ground-water response  to  recharge,   the  direction of ground-water move-
ment,   and the  rate  of ground-water movement.    A summary discussion of the
ground-water hydrology  of  each site  follows.

Sand Ridge State Forest

     A five-year hydrograph  of ground-water levels  recorded at Sand Ridge
Well SR3 and  precipitation recorded  at  Havana  appears in Figure 17.   The
five-year  hydrograph depicts   natural,  seasonal  variations  in  the water table
as well as  the  effects  produced by ground-water withdrawals at the nearby
State   Fish  Hatchery.    Annual maxima in the  hydrograph are  indicative  of
seasonal recharge which typically occurs in  the late fall and spring. Super-
imposed on this  yearly cycle is  a  ground-water level recession beginning
with the start of pumping  for   the fish hatchery  in  September of 1983.
Ground-water  levels have dropped  at   a rate  of approximately  1.2 feet
(0.37   m)  per year since  the  hatchery started operations.

     Hydrographs  of sampling wells #1 (D035) and #4 (DO 105), prepared from
water  level  measurements made every two weeks  during  sampling runs, along
with precipitation  recorded  at Sand Ridge  for the study period,  are shown  in
Figure   18.  Elevations  of water levels  in sampling  wells  #2  (D050)  and  #3
(D065) at  the  Sand Ridge  site usually fell between wells #1 and #4, denoting
a  slight upward gradient from the deeper well  #4.    Downward  gradients from
well #1 to wells #2 and  #3 were observed occasionally.   However, the  head  in
well #4 was always above  the  other three wells.   [Note:   Because no nearby
benchmark was  available from which  to  survey  a level  circuit, the  elevations
shown  in Figures  18 and 19  were  calculated  from a benchmark created  on-site
and given  an arbitrary datum of 1000 feet.   Therefore,   the  elevations shown
on the  figures  should  not be  considered  as feet above mean sea  level.  Actual
ground-water elevations  are known to be  approximately 460 feet above mean
sea level in this  area (see   Figure  3).]  Examination of Figure   18  shows
little   obvious correlation  between rainfall  and  ground-water  levels.  From
March  through  October  1986,  ground-water levels declined  approximately one
foot  (0.3  m), even  though 4.7 and 8.1  inches  (12 and 20  cm) of rain  fell  in
June and July, respectively.   These rains  were probably responsible for  a
slight  decrease  in  the  rate of ground-water  level  declines, but  an extremely
low August rainfall caused the recession to proceed.   Fall  and winter pre-
cipitation caused  water levels to rise  approximately  0.6  foot (0.2  m)  by
March  1987.   This rise,  however, left water levels  nearly  0.5 foot (0.15 m)
below  the  previous year's  high.    Fortunately,  even though  ground-water
levels  fell  through the  remainder of  the study period,  sampling well #1,
                                     65

-------
  11

8 10
-5
.£ 9'

1 8'
o 6
01
cc
I  3
"
          1983
                     1984
 1985
YEAR
                                            1986
                                                       1987
    Figure 17.   Monthly  precipitation  recorded at  Havana (a)
               and  depth to ground water in  Well SR3 at
               Sand Ridge  field site  (b)  for the period 1983-1987.

                              66

-------
       0
       Feb 1986
May 1986   Aug 1986   Nov 1986    Feb 1987    May 1987   Aug 1987
    962.6
    961.0
      02/18/86
          08/25/86
02/23/87
                                                                        08/24/87
Figure 18.   Monthly precipitation recorded  at Sand Ridge State Forest  (a)
             and  relative  ground-water elevations  in  Wells D035  and DO 105 at
             Sand Ridge field  site  (b)  during  the field  sampling period.
             (Elevations in  feet  relative  to  a  1000-foot  datum)
                                        67

-------
completed  at  a  depth of 35.1  feet  (11 m)  below ground, did not have  water
levels  drop below the top of the 5-foot  (1.5 m)  well  screen.

     Only  slight vertical variations in head  were observed  at  this  site,  As
shown in Figure 18(b),  the head  difference between the  shallowest (35  ft,
11  m) and  deepest (105 ft,  32  m) nested sampling wells was fairly  constant.
The  gradient  was always upward throughout the  study  period.   The largest
head  difference  observed between the two wells was  0.1  foot (0.03  m) for  an
upward  vertical  gradient of 0.0014  ft/ft.    While  gradient reversals  were
observed between wells #1, #2  and  #3,   the timing  and  duration of the
gradient changes  could  not be  resolved with the  biweekly  frequency of
observations.     A  data logging  system  was  installed-  at the site to
continuously record water  levels  in three  nested piezometers located within
100  feet of the  sampling  wells. However,  the  installation of the  recording
instruments occurred  late   in the project  (April  1987))  and transient
vertical  ground-water  gradient reversals  were  not observed before  the end of
the sampling period  in August.

     Although ground-water  levels generally declined,  the  direction of
ground-water  flow did  not  change significantly.   Potentiometric  surface  maps
prepared from water level measurements at  the  SR wells  on each sampling run
show that  the  direction  of flow  remained  in a northwesterly direction.
Potentiometric  surfaces prepared  from data  collected  on May  5,  1986 and
August 24,   1987 are shown  in Figure 19.   Even  though ground-water  elevations
dropped approximately  0.8 feet  (0.24 m) during  that  time period,   the
direction of flow remained  the  same.   The  horizontal hydraulic gradient was
computed to be  0.002  for both  surfaces.

     Results of casing pressurization  tests performed  on sampling well  #2
revealed  that  the hydraulic  conductivity  of the  sand  at  50 feet  (15  m)  depth
at Sand Ridge  is approximately 600  to 750  gpd/ft2 (0.028 to 0.035 cm/s).
These  results  were lower than  might be  expected for  the  clean,  medium sand
found  at this site.  "Underdamped"   responses in water level recovery, indica-
tive  of a very rapid  ground-water  level response to  induced changes in head,
were  experienced in the deeper sampling wells 83 (65 ft,  20 m) and #4 (105
ft,  32  m).    Such  water  level  responses cannot  be  directly analyzed  by
typical Hvorslev methods  (88)  but  are indicative of hydraulic conductivities
greater than the 600  to 750 gpd/ft2 calculated.

     Substitution of  a hydraulic  conductivity  of  750 gpd/ft2  (0.035  cm/s), a
hydraulic gradient of 0.002,  and  an  effective  porosity  of 0.23 into Darcy's
equation produces a  ground-water velocity  of  only 0.9 ft/d  (0.3 m/d). Tracer
experiments conducted at  this  site in 1983 (63)  indicated the  ground-water
velocity  approaches  2  ft/d  (0.6 m/d).   Given  the same  hydraulic gradient and
porosity,   a hydraulic conductivity  of  1700  gpd/ft2 (0.08  cm/s) would  be
necessary  to  yield this velocity.   Such a  hydraulic  conductivity is possible
in these materials  and  may  account  for  the underdamped water  level
recoveries  experienced  in  the  two  deeper  wells.     Estimated rates of
ground-water  movement  fell  within  a  range of 1 to  2 ft/d  (0.3-0.6 m/d).
                                      68

-------
       Direction of Regional  \
       Ground-Water Movement
       • Piezometer
       • Sampling Well
       * Existing Well
                                                            .•70
                                 a. 5/5/86
       Direction of Regional ',
       Ground-Water Movement
       SCALE OF FEET
     0  28  80     100
            _C
                                 b.  8/25/87
Figure 19.   Potentiometric  surface  at Sand  Ridge field  site  on
              May  5,  1986  (a)  and  August 24,  1987  (b).
              (Elevations  in  feet  relative  to  a  1000-foot datum)

                                   69

-------
Beardstown

     Ground-water  hydrographs of all  piezometers  and sampling wells at
Beardstown  were prepared using water level  measurements  collected  every two
weeks  during  sampling  runs.   Hydrographs of sampling wells #6,  #13, #8, and
#10 generally plotted  parallel with each  other.    The hydrographs  appear
along with the monthly  precipitation record at Beardstown in Figure  20a & b.
The hydrographs  of  the nested B8 piezometers  located between  anaerobic
impoundments 2 and  3 appear in Figure  20c (note  that B8.3, the  most shallow
piezometer,  was installed in April of  1987).   Strikingly similar patterns
are evident in the  hydrographs for  wells up- and down-gradient from, as  well
as  between, the anaerobic impoundments.

     Figure  20 shows  clearly that water levels rose  and fell during the
course  of the study in response  to rainfall  events.   Ground-water level
fluctuations were  more  frequent and  of  greater  amplitude  at  the Beardstown
site than at  Sand Ridge,  possibly  because the  water table is closer  to the
land surface  at the  Beardstown  site  (5  to  15  feet deep, 1.5-4.5   m, as  com-
pared  to  30  feet,   9   m,   at  Sand  Ridge).     An extremely wet  July was
experienced in  1986  with over   10   inches   (25   cm)   of   rainfall,  and
ground-water  levels  rose more than one foot  (0.3  m)  shortly  thereafter.  Low
August  1986  rainfall resulted in a decline  of  over one foot (0.3 m).  Rains
in  September and October  1986  contributed to a rise  of approximately  2 feet
(0.6 m). Generally  low  rainfall throughout the  remainder of the  study  period
resulted in  a  decline   in  ground-water  levels,   with   slight   recovery
experienced during the  spring and early  summer  of 1987.

     Information collected from  the hydrologic monitoring system  revealed
that ground-water  levels  in the  silty water-table aquifer  beneath the
Beardstown  site responded to changes in atmospheric  pressure.  An example  of
this response  is shown in  Figure  21.    The figure  displays data recorded
every 15  minutes  for the barometric  pressure,  the  water level in piezometer
WLR2.1,   and daily  precipitation for the month  of March  1988.   The  figure
shows  that increases  in atmospheric pressure  often  were accompanied by cor-
responding drops   in water level.    Similarly,   decreases  in   atmospheric
pressure  were accompanied by increases in water  level.   Changes  in  water
level are  not always  easily attributed  to  atmospheric pressure  changes
because of the interference  of  rainfall recharge.  Barometric  efficiencies
calculated for  the data  shown in Figure  21 ranged from  13  to  83  percent  with
a mean of 36 percent,

     Because  of   the   presence  of   the   wastewater   impoundments,  the
ground-water  flow pattern at the  Beardstown site  was much more  complex than
at  the  Sand  Ridge  site.   Potentiometric surface maps were prepared from
water level  measurements collected  at all  sampling wells and piezometers to
describe  three  surf aces (upper,  middle,  and lower)  according  to  the  depth of
the well used.   Figure 22  shows the three  potentiometric surfaces for two
dates  following  the installation  of piezometer  B8.3,  which  was used to
define the top  of the mound created beneath  anaerobic  impoundment  3.   April
21,   1987 (sampling run #30) was chosen to  represent  a period  where
ground-water  levels  were high and  August  25,   1987  (sampling run #39) was
chosen  to represent a period when  ground-water  levels  were  low.
                                      70

-------
                  F«b 1986  May 1986 Aug 1986 Nov 1986  Feb 1987  May 1987  Aug 1987
                 02/19/86
                   08/26/86
                                                  02/24/87
                                                    08/25/87
                 446'
              < 445
              UJ
              to
              z
UJ
>
                444
                 443
              IT 442
              2 441
              UJ
                 440
                      C.
                                     \jf
                                              '88.3
               ~ 02/19/86           08/26/86          02/24/87          08/25/87
Figure  20.   Monthly precipitation recorded at  Beardstown (a),  ground-water
              elevations  in Wells BT23, BT30,  and  BT33 (b),  and  ground-water
              elevations  in  the  B8  piezometer nest (c)  at  Beardstown field
              site  during the field sampling period.
                                         71

-------
          Daily amounts
             0.09       0.11
441.2-^
       1
10
                                             15
                                          MARCH. 1988
20
30
                                                        1-L28.6
         Figure 21.  Precipitation,  barometric pressure, and ground-water elevation
                     in  piezometer WLR2.1  recorded at the Beardstown  field site in
                     March.  1988.

-------
             4/21/87
8/25/87
                        Upper Potentlometric Surface
                         Middle Potentlometric Surface
                         Lower Potentlometric Surface
Figure  22.  Upper,  middle,  and  lower potentiometic  surfaces  at the
             Beard'stown field  site on  April  21,  1987  and August 25,  1987.
                                    73

-------
      The  upper and  middle potentiometric  surfaces in Figure  22,  show that a
 ground-water  mound formed beneath  the  anaerobic impoundments.     The B8
 piezometers placed between the anaerobic  impoundments  2  and 3  gave  a  clear
 indication that a  vertical  gradient existed   beneath  the impoundments.
 A hydrograph of the  B8  piezometer  nest (Figure  20c) shows the vertical
 gradient which was  present between  anaerobic impoundments 2 and 3.    For
 mapping purposes,   the elevation  of  the  ground-water  surface beneath  the
 impoundments was  estimated to be slightly higher than the  elevation of the
 surface at B8.3  because  no piezometers  were  placed in the impoundments.
 Further,  vertical gradients beneath the center of impoundment 3 were assumed
 to be the  same  as  those  at the  B8  piezometers  located  between  anaerobic
 impoundments 2 and  3  (Figure 9).

      The  height  of  the mound  above  regional water  levels and the change in
 height in  comparison  to  regional ground-water  levels  were of considerable
 interest.   Figure  22  shows that the  height  of  the mound (on the upper
 potentiometric surface) above regional ground-water  elevations (on the lower
 potentiometric  surf  ace) changed significantly between April and August.  Over
 this period, ground-water levels  fell approximately 2.7 feet (0.8  m) on  the
 lower  surf ace.  However,   ground-water elevations on  the upper surface
 directly beneath the impoundment fell only  1  foot (0.3 m).

      Regional ground-water elevations  appear to be predominantly affected  by
 rainfall  recharge.      Regional  ground-water   elevation declines   can   be
 attributed  largely  to  a lack of rainfall  (almost  40%  below  normal  for  the
 April to August  period).   The  height of the mound,  however, appeared  to be
 affected not only by the rate of impoundment leakage but  also  by the  magni-
 tude of the  leakage in comparison  to the regional  ground-water discharge
 rate.    This  hypothesis  is supported by  surface and    ground-water  level
 measurements made  through the course of  the project.

      Staff gage readings of the  water surface in the  anaerobic impoundments,
 made during every sampling run,  indicated  that the volume of water  stored in
 the impoundments  was  fairly  stable  throughout the study  period.   Water
 levels  in  each anaerobic  impoundment were controlled by a fixed-elevation
 circular weir.   In  controlling the  water  levels  within the impoundments,
 heads on  the  impoundment bottoms were  held within a range of  1  to  2  feet,
 (0.3 to  0.6 m).    It  is  likely that  leakage  through the  bottom of  each
 impoundment was reasonably constant.

      If  the rate  of  leakage from the  impoundment was constant,  it  would  be
 expected that the leakage rate  would  be  proportionately smaller in compari-
 son to  the  regional  discharge  when  regional ground-water levels were high.
 The result  might then be observed as  the assimilation of the ground-water
 mound into the  regional  flow  field.   This effect is shown  by the  April 21
 potentiometric  surfaces in Figure 22.   When regional ground-water levels
were low,  regional  ground-water  discharge is  less and  the  constant  leakage
 rate from the impoundment would be expected  to  be  greater  in proportion to
 the regional discharge.   The ground-water mound then would  be expected  to be
 more pronounced as shown  by the August 25 potentiometric  surfaces in Figure 22.
                                        74

-------
     TABLE 14.   VERTICAL HYDRAULIC GRADIENTS AT THE B8 PIEZOMETER NEST

Between Vertical
piezometers April 21, 1987
B8.3 & B8.1 0.03
B8.1 & B8.2 0.05
gradient Relative
August 25, 1987 change
0.09 3x
0.26 5x

     The  mounding  effect was  exemplified further by  the vertical  gradients
induced beneath the  impoundment.   Vertical  gradients beneath  the  impoundment
are seen  most easily in the hydrographs of the B8 piezometers (Figure 20c).
The head  difference  between the shallowest and  deepest  piezometers, B8.3  and
B8.2,  was less than 1 foot  for  most of the period in  which water levels  were
recorded  in B8.3.   However,  as water  levels fell during  the summer, the  head
difference  between piezometers   gradually   increased  to  over   2.5  feet
(0.76 m).   The  changes  in the vertical  gradients between  piezometers  B8.3
and B8.1  and between piezometers B8.1 and B8.2 for  the two dates are shown
in Table  14.   Vertical gradients increased  by a factor  of  3  to  5  over this
four-month  period.

     The  mound had an effect  on the  wells  located do w n - gr adie nt, and
potentially "upgradient,"  from anerobic impoundment 3.   Equipotential  lines
drawn on a vertical cross  section parallel to the regional ground-water flow
path  beneath anaerobic impoundment 3  and  through  the  sampling  wells are
shown in  Figure 23.   The  equipotential  lines  shown  are interpretations of
the head  data collected from the piezometers and  sampling  wells at the  site
on April  21,  1987  (Figure  23a) and August 25,  1987  (Figure  23b).   These are
the same dates  for which  potentiometric  surfaces  are shown in Figure 22.
Ground-water elevation data  from the B8 piezometers were  used to draw the
contours beneath  the,  impoundment  (see  Appendix C). The equipotential  lines
are considered conceptual because  head  values  at  depth  in the  aquifer  were
not available  and  no flow modeling  was conducted  to  provide a mass balance
of the regional and impoundment  fluxes.

     The  conceptual cross sections  provide a  reasonable interpretation of
ground-water movement  beneath  the  impoundment. The cross sections  show  that
water moving  out from the  bottom of the  impoundment forces  "regional"  ground
water (ground  water from an  upgradient  direction) downward  as it  approaches
the impoundment.   Water from the impoundment would  move through the screens
of the downgradient sampling wells as it was diverted  into  the regional flow
system.    At certain  times,  particularly when  the regional  water table was
low (or  when the regional flow was  low compared to the  impoundment leakage)
as in  Figure 23b, water leaking from the impoundment  also may be expected to
move  toward the upgradient sampling  wells #5  and #6.  Therefore,  while the
downgradient wells  always received water which  originated  in the  impound-
ment,   the  upgradient wells may  have  been  affected only periodically
depending on  the  strength  of the regional  flow  system.   This may have caused
ground-water  quality  changes  in wells  #5 and #6.
                                        75

-------
   a. 4/21/87
340
   b.  8/25/87
460r
340
      Figure 23.   Equipotential lines beneath  anaerobic impoundment 3
                   at the Beardstown field site  on April  21,  1987  (a)
                   and August  25,  1987  (b).
                                        76

-------
     Results of casing pressurization tests  performed on  selected sampling
wells  at  Beardstown are shown  in Table  15.   Two  variations  of the Hvorslev
(88) method of analysis  were used.   The  first  method  is an approximation
based  on a well point extended  in  uniform soil  (Hvorslev's  Case  8)  and is
also the method commonly outlined in ground-water texts (e.g., Freeze  and
Cherry,  33).    The  second method  is  an approximation for a well point
extended through  a permeable  layer  between  impervious  strata (Hvorslev's
Case 9).   This  method  was used for  comparison with the  first  method  because
it was  felt that the  siltiness  of layers above and  below  screened sections
may  cause   the  horizontal hydraulic  conductivity  to govern ground-water
movement.  As shown in Table  15,  the hydraulic conductivity ranged from 2.4
to 3.0 x  10"2  cm/s for Hvorslev's Case  8  method and  from 3.1 to  4.2 x  10"2
cm/s  for Hvorslev's  Case 9 method.     Basically,   results from the  second
method were 1.3 times greater than  those  from the first.    It  is not possible
at this  time to  determine which values are more  accurate.   These  values are
within  a range reasonable  for  a  silty to  clean sand  (33)  and  are quite
similar to  those observed at the Sand Ridge  site.   The differences  between
the  results for  the  two methods at  the  Beardstown site may  reflect
differences  in  hydraulic  conductivity  between the developed  bore hole  and
that of the natural  geologic  material.

     'The potentiometric surfaces  in  Figure  22 indicate that the slope of the
potentiometric  surface varied greatly with  location and with time.  This is
in marked  contrast to  the  constant slope of  the potentiometric  surface
observed at the Sand  Ridge site.

     Because of the  large  vertical  head differences near the  impoundment,
hydraulic gradients  and  ground-water  velocities were  calculated in three
space dimensions according to  a method  outlined by  Abriola and Finder  (89).
The  method is  based  on the differentiation  of  a three-dimensional linear
interpolation of hydraulic heads  at  four irregularly spaced points to derive
hydraulic  gradients  in  three  dimensions.   Substitution of  the calculated x-,
Y",  or z-gradient  into  Darcy's law provides  the  ground-water  velocity in the
respective   direction.


             TABLE 15.   HYDRAULIC CONDUCTIVITIES (IN 10'2 CM/S)
                        AT SELECTED BEARDSTOWN WELLS

Method
of
calculation*
Hvorslev
#8
Hvorslev
#9

BT18
(#5)

3.0

3.9

BT23
(#6)

3.2

4.2

BT25
(#8)

3.0

3.8

BT35
(#10)

3.0

3.8

BS30
(#11)

2.5

3.3

BP30
(#12)

2.5

3.2

BT33
(#13)

2.4

3.2

  * Hvorslev's Case 8 and Case 9,  see reference
                                      77

-------
     For  this  site,  the magnitude  Of the three-dimensional  gradients (and
velocities) calculated by this method  depended  on the choice of the four
wells used in the computation.   This  was particularly true in areas  where
infiltration from the  impoundment  had a  profound effect on vertical
gradients.    Four areas were  chosen to more closely examine the nature of
ground-water velocity changes across the  site.    These  were:   I ) upgradient
from the anaerobic impoundment,  utilizing head differences   and  distances
between piezometers  Bl, B2.1,  B3.1,  and  B3.2,  2) downgradient  from the
anaerobic  impoundment,  utilizing head differences and distances between
wells WLR2.1,  BT33 (Well  #13),  BT30  (Well #9), and  BT35  (Well #10), 3)
beneath the impoundment on the  "upgradient" side, utilizing  head differences
and  distances between piezometers B8.1, B8.2, B3.1,  and  sampling well BT18
(well  #5),   and 4)  beneath the impoundment on the  downgradient side,
utilizing head differences and distances between piezometers B8.1,  B8.2  and
sampling wells BT33 (well  #13),  and BT25 (well #8). Water level data from
B8.3 were not used in these calculations because a  full  length  of  record
over the entire sampling period  was not available.

     A  summary  of the  ranges  and means of  the  hydraulic  gradients  and
ground-water velocities  calculated  for  the  four  areas around anaerobic
impoundment 3 is presented in Table 16.  The  values  in the table  represent
the range  and means calculated  from  observed  heads  measured at the  site
during the 39 biweekly sampling runs conducted between  March  11,  1986  and
August 25,  1987.   A horizontal hydraulic  conductivity  ten times the vertical
value was  used in the calculation of ground-water  velocities  because  of the
siltiness and  layering of the  aquifer materials.

     Resultant ground-water  velocities calculated  over the  length of the
project  for these four areas  are  shown in  Figure 24.   Shallow ground-water
velocities upgradient of the  impoundment (mean, 0.22 m/d) were slightly  less
than downgradient of the  impoundment (mean, 0.30 m/d).   This  was due to  the
presence of the  ground-water mound beneath the  impoundment.   Velocities
decreased  as the  mound  was approached and increased again after the  mound
was  passed.   Ground-water velocities beneath  the  impoundment were much
greater than the upgradient  and downgradient  velocities.    This  was  due
principally  to the large vertical gradients  created  within  the  mound.
Vertical gradients beneath the  impoundment  were  10  to  25 times greater than
vertical gradients  outside  the boundary of the  impoundment.    Resultant
velocities  beneath the impoundment  on the downgradient side  of the mound
(mean, 1.15 m/d)  were greater than those on the upgradient side of the  mound
(mean  0.65 m/d).

     Velocities computed for  upgradient and downgradient  locations beneath
the impoundment paralleled each other because of  the  use  of head data at
piezometers B8.1 and B8.2  for both  locations.   The strong  vertical  gradient
between these  two piezometers had  a  large influence on the  calculated ve-
locities for  these two  areas.   The steep increase  in velocities  over the
last five  sampling runs was  further indication  of the  influence of the
vertical gradient  on ground-water velocities.   As mentioned previously, when
regional water levels fell over the course  of the  summer of  1987, the  mound
beneath the impoundment  gradually became  more  pronounced.   This caused both
                                      78

-------
                                   TABLE  16.   HYDRAULIC  GRADIENTS  AND GROUND-WATER VELOCITIES AT BEARDSTOWN*

Locat ion
Upgradient of
Impoundment 3
Beneath Impoundment 3
on "Upgradient" aide
heneath Impoundment 3
on "Uowngradient" Side
Downgradient of
Impoundment 3


Parameter
Cjrad lent
Velocity
Grad lent
Velocity
Gradient
Velocity
Gradient
Velocity
(m/m)
(m/d)
(m/m)
(m/d)
(m/m)
(m/d)
(m/m)
(m/d)
X-Dlrection
Range
0.00083 to 0.0017
-0.23 to -0.1 1
0.001 1 to 0.0020
-0.27 to -0.15
0.0020 to 0.01 10
-1 .51 to -0.19
-0.000096 to 0.001 3
-0.17 to 0.009
K-Dlrection
Mean
0.001 1
-0.15
0.0015
-0.20
0.0056
-0.78
0.00057
-0.077
Range
-0.0025 to
0.09 to 0.
-0.00093 to
-0.63 to 0.
-0.0069 to
0.12 to 0.
-0.0033 to
0. 16 to 0.
-0.00065
31
0.0015
13
-0.0031
95
-0.0012
15
Mean
-0.0012
0.16
0.0013
-0.18
-0.0012
0.57
-0.0021
0.28
Z-Directlon
Range
0.00051 to
-0.01 to -0
0.026 to 0
-1.17 to -0
0.028 to 0
-1 .27 to -0
0.00035 to
-0. 12 to -0
0.0029
.007
.085
.36
.092
.38
0.0089
.005
Mean
0.0017
-0.025
0.012
-0.58
0.016
-O.b3
0.0039
-0.051
Resultant
velocities
Range Mean

0.16 to O.M2 0.22
0.12 to 1 .35 0.65

0.7b to 2.20 1.15

0.20 to 0.16 0. 30
"  Velocities calculated using horizontal  hydraulic conductivities,  Kx & K¥ - 27 m/d;  vertical hydraulic conductivity.  K^ -  O.I  Kx - 2.7  m/d;  and
  effective porosity,  n =  0.2.   Ranges given are the extremes (most negative to most  positive) from 39 biweekly sampling runs conducted  between
  3/11/86 and 8/25/87.   Kor velocity vectors, positive X-direction  Is east,  positive  Y-directlon Is north, and positive Z-direction is upward.

-------
  2.6


  2.4



  2.2



£ 2.0
£  1.8
  1.6-

o

3  1.4
UJ

E  1.2
UJ

ILOH

o
§ 0.8

O

S0.6
   0.4-



   0.2-



    0-
        0 Upgradient of impoundment 3


        0 Beneath impoundment 3 on upgradient side


        + Beneath impoundment 3 on downgradient side


        A Downgradient of impoundment 3
                         10
                                   15        20         25

                                    SAMPLING RUN NUMBER
                                                                 30
35
40
   Figure  24.   Calculated ground-water velocities  for four  areas  in  the  vicinity

                of  anaerobic  impoundment  3 at  the Beardstown field site.
                                            80

-------
the vertical  hydraulic  gradients  and  ground-water velocities to increase as
a  result.

     The  ground-water velocities and hydraulic gradients  observed at the
Beardstown site were indicative of a much more  complex flow system than at
Sand  Ridge.   Water level response  to  rainfall  at  Beardstown was much more
rapid  and  of greater amplitude than  at Sand Ridge.    The response  to changes
in atmospheric pressure, even under water table  conditions,  was apparent at
the Beardstown site.    The leaking  impoundments coupled with  a  shallower,
more  responsive   water table and more  variable  geologic materials at
Beardstown comprised a  more transient  hydrologic system than at Sand Ridge.

     According to  individual well  hydrographs,   ground-water levels  fell  a
maximum of over  three  feet (0.9  m) at Beardstown  during the field  sampling
period but less than two  feet (0.6 m) at the Sand  Ridge site.    This was
despite the  fact  that  the Sand  Ridge site was within the influence  of  a
continuously  pumping  well  field.   In spite  of the pumpage, the direction and
rate of movement  of ground water at Sand  Ridge remained remarkably constant
throughout the study  period as   shown in the  potentiometric surfaces  in
Figure 19.   Conditions  at  Beardstown,  however,  changed markedly and are best
exemplified by the potentiometric  surfaces shown  in Figure  22.    Vertical
hydraulic  gradients,  created  by the  leaking impoundments at  Beardstown,  were
particularly  large and  the principal  cause for  velocity  changes.    However,
at locations distant  from the  immediate  vicinity  of  the impoundments,
ground-water   velocities   (and  hydraulic gradients)   remained  relatively
constant (Figure  24).    This  information is consistent  with the  ground-water
chemistry  of each site  as  will be related in the following sections.
CHEMICALDATACHARACTERISTICS

     The completed  sampling at the uncontaminated  site extended over   more
than  2.5 years  (1020,  days)  with field and laboratory  determinations of 26
chemical and physical parameters   at  established intervals over  the time
period.   The sampling and  analysis at the contaminated  site were  conducted
over 1.8  years (600  days)  for the same suite  of major parameters.   Figure 25
shows graphically the  relative  fractions of anions  or cations present, on
the average, for each well.   For these summaries, the mean inorganic concen-
trations for  each well  (Appendix A)  were processed by the WATEQ4F  computer-
ized aqueous  chemical speciation model (90).   Ion pairs and  complexes (e.g.,
CaHC03+, MgHC03+, CaS046,  MgC03°  etc.) were counted with  their anionic and
cationic  components  for  simplicity  in  these  figures.   The vertical  order of
the ligand  represents  the  order of the components in  each of the bars.

     Ground water at the  Sand Ridge  site contained  calcium, magnesium, and
bicarbonate  components introduced during recharge  through the  soil zone.
Small amounts of sodium,   sulfate  and  silica  were  also  present.  Calcium was
consistently higher  (i.e.,  approximately 1.5 times)  than magnesium    on  a
molar basis.  Calcium, magnesium,  and bicarbonate decreased distinctly  with
depth.   The decrease  corresponded  approximately to  the relative ages  of the
ground  water (i.e.,  oldest = deepest)  determined by  tritium (3H)  dating.
                                      81

-------
                                                       C/A/U
=d
f si 5
- o
< E
-E4
z _-
0 D
<23
-* ^
O :5
§ < 2
z u
0 Z
u
Q
a.
- |
Qj
1
v
$
• ^
\
\
. \
\
\
\
N
__
-








•I
\
r
o
|
8
5c
"T
\
\]
\
N
^

y
X
5
X
$
\
s
s
s
\
\
-i

^





I
3
V
5<
\
\
\
S
^




-

.
^
\/X Cu/HP04
g^ NH4/CI

KX1 Mn/P04

| | Fe/HC03

•1 K/C03

ffl% Na/S04/SiO2
[gg Mg/NO3/H2S
[\^ Ca/NO2/NH3
CAU CAU CAU CAU
Well =1 2 3 4
C/A/U
10

I_J g
„ "5
< |
§56
z Q
< LU
.- a
o a: 4
? o
o z
P=2
u

b.
.


-
m
y
/
y
\

\
s
	
vaj-
P
SS
®
P
1
1


// rm-*

'///
w
lp
^
1

\\

^
NNN
i
1
1


////
%,
-







7773

\/A Cu/HP04
t^xj NH4/O

I^SJ Mr PO4
| | Fe/HCO3
^1 K/C03
^ N./S04/Si02


^g Mg/N03/H2S
£\3 Ca/N02/NH3
CAU CAU
Well =5 6
r/A/u
30

?_j 25
i 1 20
1 =
w (3

jj «E
y< 10
Z 0
o z
< D 5
u

c.
-
.


•
.


_ i
!
• ^
I |
! 5

^





$
1






f(
^
i 5


>
'
; /
; «
• :
w \
> ;
;

!
t
' -



i

/



^ Cu/HP04
{^3 |NH4/Q
|^X,j |Mn/P04
| | |Fe/HC03
•1 K/C03

&% |Na/S04/Si02

^g |Mg/N03/H2S

^^ Ca/N02/NH3
                 Well
13    8    9    11    12   10
Figure 25.   General  chemical characteristics  for  the  Sand Ridge wells,
             (a),  the upgradient wells  at  Beardstown (b),  and the
             downgradient  wells  at Beardstown  (c) .
                                     82

-------
     Comparison of the results for water samples from wells #1  and #4  (i.e.,
the  shallowest and deepest wells, respectively,  at  Sand  Ridge)  provides  a
measure  of the difference  between fully  oxidized and suboxic subsurface
redox  conditions.   Average  profiles  of 02,  Eh  and Fe(II)  with  depth at the
Sand Ridge  site are shown  in  Figure 26 (a  two  standard  deviation range  is
included  for  each parameter).    Very strong oxygen concentration  and redox
potential gradients were  present  and were  reflected  by average  concentra-
tions of ferrous  iron.   These  chemical  gradients were very stable  during the
study.  It is likely that the dissolved oxygen levels measured in water from
well #4 at  Sand  Ridge represented  an upper  limit  of concentration  since
oxygen  diffusion through the PTFE tubing  alone could increase  the  apparent
02 levels by  0.2 to 0.3  mg • L"1 (91).

     Samples of the uncontaminated ground water from Beardstown upgradient
wells were  similar to  each  other in relative proportions of components, but
were considerably different in ionic strength.   The  water from well #5 had
            strength   similar to   the  Sand Ridge  wells   (0.007  versus
0.005-0.008),    but sodium  accounted  for approximately one-third of the
equivalents  of  cations  present.  Sodium was present at about twice the molar
concentration of calcium.    Water from well #6 had  the  same  general ionic
character  as well #5,  but contained proportionally more  of all  components.
The  average ionic  strength of water  from well #6  was 0.011  compared to 0.007
for well #5, and the ferrous iron in well #6  was 3.5  times the  average level
in well #5.   For  both wells,  chloride  and sulfate made  up  the bulk  of the
anionic composition  on a  charge basis with  chloride  1.5 to  3 times the
bicarbonate  and sulfate  concentrations on a  molar basis.

     The  effect of the lagoon  seepage on downgradient ground-water composi-
tion was striking.    The average  ionic strengths of water from the  down-
gradient wells  were 2 to 4  times as high as wells #5  and  #6.   Ammonium ion
and  sodium were  the  dominant cations,  with  potassium  approximately  seven
times the  upgradient levels on a molar  basis.    The  sodium  concentration was
consistently 9  to  10 times the potassium concentration.   The molar ratio  of
Ca2+ to Mg2+  was  consistently  1.5.  Ion  pairing  and  complexation of Ca + and
Mg2+ by HCO3", HPO4=,  SO4= and  CO3=  accounted  for  approximately  10-20%  of
the  dissolved  metal species in samples  from these wells.   The ferrous  iron
concentrations  were twice the  upgradient  concentrations,  but they  remained
insignificant  in  the  overall  constituent balance,  relative to the other
cations.  Bicarbonate  was by far the  dominant  anionic  constituent,  making up
approximately  three-quarters  of  all anions on a charge basis.  Chloride  was
consistently the next most important  anion in  concentration (about  0.25  that
of bicarbonate), with SO4=, HPO4= and PO4= the  remaining significant  anionic
forms.

     The results  for the ground-water  samples  from  Beardstown showed the
marked differences one would expect  between locations upgradient  and down-
gradient from a leaking anaerobic treatment impoundment.  Dissolved oxygen
levels at both  upgradient and downgradient locations  (Tables A-5 through
A-12) were consistently at  or  near the  detection  limit  for oxygen  electrode
measurements  and below  the  corresponding  limit  for  iodometric titration  by
the Winkler method. Ground-water quality conditions were changed markedly  at
downgradient  locations  by the leakage of leachate  from  the anaerobic
                                        83

-------
                     100     200	300      400      500    (mv)   • Eh
    10
=   20
t
LU
Q
   30
                                       8      10   (mg-L'1)  •PROBE   o WlNKlER. 02
             0.1      0.2       0.3      0.4      0.5    (mg-L-1)
Fe2
    Figure  26.  Profiles of Eh,  dissolved oxygen,  and ferrous iron

                 with  depth at  the Sand  Ridge field site.

-------
impoundment.   The  downgradient  ground-water  samples were  generally  much
higher  in dissolved  solids, temperature, total organic  carbon, methane,
alkalinity, ammonia,  sulfide, iron,  chloride,  sodium and potassium than were
samples  from upgradient positions.

     The  average results  of  water  analyses  for  redox-related  species  and Eh
in samples from wells at the Beardstown site are shown in  Figure 27. The
results  from the  upgradient nested wells  (i.e.,  #5 and #6)  showed  some
increase  in ferrous  iron  and  sulfide  concentrations  with depth but no
significant change in  Eh or  concentrations  of  dissolved oxygen and methane,
The  downgradient wells  along a flow  path from  the  impoundment  accessed
progressively more  reducing  ground water,  judging  from average methane,
dissolved  oxygen, sulfide and Eh measurements.   Ferrous  iron values  did not
increase  or  decrease significantly downgradient.   This  observation may be
explained in part  by iron  mineral  controls  on ferrous iron solubility.
Saturation indices  for  common  ferrous-carbonate,  -phosphate  and  -sulfide
minerals  were  all at  or  above saturation for  ground-water samples  from the
downgradient wells, using the  equilibrium  speciation model (90).

     The  analytical  results from  samples from the three wells finished  at 30
ft  (10  m)  depth,   only  I  to 1.5 m  apart  and  constructed  of dissimilar
materials  (i.e.,  #9 -PTFE;  #11 -SS; and #12 -  PVO are shown in Tables A-8,
A-10,  and A-ll.   The  results  reflect  the sensitivity of the iron and sulfide
systems to well casing  material  effects and presumed  differences in the
geochemical microenvironment near  the wells.   A  more detailed  discussion of
these results appears in  Barcelona et al.  (92).

     Table 17 gives a summary  of  the  saturation indices computed for several
minerals  of  interest using  mean values of the chemical analyses over the
course  of the project.   The  indices  suggest  an  approach to calcium carbonate
equilibrium from initial  oversaturation at the  Sand Ridge site.   The deepest
well (#4) represents water  closest to  that of the regional base flow,  which
would have an "age"  since recharge  takes place on the order of decades.  The
shallower  wells  would be expected to  have younger water (on  the  order of
days to years in age) containing C02  picked up from the overlying soil  and
calcium,   magnesium,  and  carbonate  leached from  surrounding materials. As
the C02  gas is depleted with  depth,   the water approaches equilibrium with
calcite.   The  ground water  at Sand  Ridge tended to be near saturation with
respect to chalcedony,  a cryptocrystalline  form of  quartz. This  phenomenon
may represent  the product of reactions of  the recharge  water with clay
minerals  in  the soil  zone and  weathering reactions  of silicate  minerals in
the aquifer materials.

     In the  Beardstown  downgradient wells,  there appears to be a  trend
towards  equilibrium with  calcium carbonate  from upgradient  to  downgradient
positions.  Insufficient  calcium is  introduced  from the waste plume to  reach
saturation,  so  the driving force  towards equilibrium  appears to be the
generation of bicarbonate from organic  material decomposition.    Increases in
temperature  due to microbial activity  also favor  this change.    Iron concen-
trations may be controlled  by ferrous carbonate  (siderite).   The uncertainty
in the  equilibrium  constants for many sulfide  and phosphate  minerals  and
their slow precipitation rates or uncertain  stoichiometries make their  role
                                       85

-------
    10
     ,-3
    10-'
_
o
z
LU
O
z
o
(J
    10-'
    10"
        : UPGRADIENT
             A Fe2*

/
         £ -- o Eh
             .D a
J-*
               A
               N
               A
               E
               R
               0
               B
               I
               C

               I
               M
               P
               0
               U
               N
               D
               M
               E
               N
               T
                DOWNGRADIENT
          -17m  -16m
                    •SOURCE-
                       D ---- 0
                                                          400
                                                          300
                                                                     200-B
                                                                        j=
                                                                        01
                                                          100
                         13
                         ->
                                           8 9,11.12 10
                                              WELL No.
                  +23m   +40m  +41m +42m
                                        -> DISTANCE FROM SOURCE
  Figure 27.   Average concentrations of redox-active chemical species with
              distance from contaminant source.   (Concentration is on
              a logarithmic scale,  Eh is on a linear scale and distances
              from source are schematic and not  to scale)
                                           86

-------
do
                           TABLE  17.   MEAN  SATURATION INDICES* FOR SELECTED MINERALS
                                    IN GROUND WATER FROM THE PROJECT WELLS
Well
1
2
3
H
5
6
8
13
9
11
12
10
Calcite
0.32
0.33
0.36
-0.02
-1.68
-1 .78
-0.28
-0.18
-0.22
-0.20
-0.18
-0.11
* Saturation In


Dolomite
0.27
0.32
0.39
-0.42
-3.65
-3.89
-0.81
-0.58
-0.68
-0.63
-0.59
-0.48


lx<
Siderite
-1 .78
-1 .85
-0.7-3
-0.14
-1.43
-1 .21
0.25
0.49
0.27
0.46
-0.03
-0.16


»P
Chalcedony FeS
0.09
0.09
0.06
n.DQ
0.02
0.07
0.14
0.14
0.19
0.15
0.20
0.20
(ppt) Vivianite
—
--
--
—
1 .18
1 .09
0.28
0.92
0.94
0.95
0.77
0.92
IAP = ion activity product
KSp = solubility
constant
-8.62
-8.28
-4.28
-1 .61
-4.93
-4.39
1 .74
2.51
2.16
2.50
1.35
1 .92
and

Hydroxyapatite
-2.85
-1 .82
-0.70
-1.23
-9.16
-10.03
0.30
0.97
1 .12
0.83
1 .46
1.75


for reactions as written below.
              CaC03(3)  t  Ca2+ + CO
              CaMg(C03)2(s)  t  Ca2+
              FeC03(s)  £  Fe2+ + C03
              HijSiOn(s)  £  H^SiOij0
              FeS (ppt) + H+  %  Fe2+
              Fe3(POij)2.8 H20(s)  J  3
              Ca5(POij)3OH(s) + 4 H+  X
            calcite:
            dolomite:
            siderite:
            chalcedony:
            FeS  (ppt):
            vivianite:
            hydroxyapatite
+ Mg
2-
    2 +
 HS"
Fe2+ +
 5 Ca2+
         co32~
           P0ij3
            3
 8  H20
-~  +  H20

-------
(if any)  difficult to  assess.   Sulfide  may  be controlled  by precipitation
with  ferrous  iron,  which is  suggested  by the relative consistency of the
saturation index  for  ferrous  sulfide.

     The  principal dissolved contaminants  at the Beardstown site  were major
ionic  constituents  of the  anaerobic  leachate which seemed to  be at or near
chemical  equilibrium  with the aquifer materials.   The  usefulness  of any  of
these  constituents as waste-specific indicators  under similar conditions
would  not appear to be  constrained significantly by assumptions  of chemical
equilibrium and the use  of  speciation  models.
STATISTICAL STRUCTURE AND TEMPORAL VARIABILITY

Estimation of Sources of Variation

     In order  to  insure that monitoring  resources  are optimally allocated,
it is important  to  identify the  sources  of natural  (i.e.,  spatial and
temporal)  and sampling and  analytical variance.   The  general  statistical
approach  to  reducing the effect  of any source of variation is to  randomize
and  collect  replicates.   Therefore,  the  effects  of natural  variability can
only be reduced  by increasing the  sample size (that is, either the sampling
frequency or  the length of the  data collection  series).    For instance,
increasing the sample  size  has the effect of reducing  the  component  of the
variance of the long-term mean attributable to field and lab  errors.   In  ad-
dition,  if the component of the  total variance due  to  laboratory  and/or
field errors  is  large,  it can be reduced by  taking more  than one field  or
laboratory replicate  at  each  sampling occasion.   Whether or not collect  ion
of laboratory  or  field replicates  is  cost  effective  depends  on the fraction
of the  total  variance attributable to each source.   The  conceptual model
used to estimate sources of variation was:

        at2  =  On2 *• °12 * °f2

where:   o^  =  total variance

        an2  =  natural variance

        0]_2  =  laboratory analytical variance

        Of2  =  field sampling variance

     Generally, the natural  variations  in water  quality time  series  are  of
interest.   For instance,  the  difference between  the time series  of a given
contaminant at a  downgradient  and  an up gradient  well  may  give  an indication
of whether  contaminant release has  occurred.  However,  the difference  series
is inevitably corrupted by  errors  in the field data collection and labora-
tory analysis procedures,  both of which  introduce  what may be  considered
"noise"  into  the  time  series.   Each  of these noise processes  has a variance,
and  the total variance is the sum  of the  three variance  terms.   This  model
assumes that  the three sources of  variation  are  statistically independent,
which   is   a  reasonable  assumption  because the sources are  physically

-------
independent.    A possible exception is that the magnitude  of the  field and
laboratory   errors   may  depend  on   the  true   value  of  the   chemical
concentration.  This  consideration is addressed  below.

     The  sources of variation were estimated as follows.   First,  the labora-
tory analytical variance was   estimated  by taking  the difference of the
laboratory calibration standards  series,  since  each standard was subjected
to two  replicate analytical  determinations.  Each difference was  normalized
by the  (known) true  value  of the standard.    Then,  a normalized  standard
deviation was   computed using the  inner-quartile  differences (i.e.,   75th
minus  25th  percentiles)  multiplied by an adjustment  factor appropriate  to
the  normal  distribution.   The  use of the adjustment  factor for the normal
distribution  does  not imply  that the  distributions  are  in fact normal (most
are  not);   it is only a  convenience.   Next, a  similar procedure was applied
to the  field replicate  series.    Because the  field replicates  include the
effect  of laboratory analytical variability as well,  the  analysis of the
field replicates provided estimates of the  sum  of the normalized  laboratory
and  field  sampling variance.   The normalized field sampling  variance was
then estimated  by  subtraction.  This  subtraction  occasionally gave  negative
values,  which  were reported  as  "NA".   An alternative procedure would  be  to
simply set  these values (i.e.,  the normalized  field  variance)  to  zero.
Finally,   the normalized total variance  was  estimated from  the entire time
series for each  well,  and for each  chemical constituent where the  normaliza-
tion  was  by  the median  for the  given well.   The normalized  natural variance
was  then  estimated by  subtraction.

     The  results are summarized in Table  18  for  the  three  groups of wells.
For  almost  all  of the groups, and for almost  all  of the  chemical constitu-
ents,  a high fraction of the total  variation was natural.   In fact,  with the
exceptions of  calcium,   magnesium and potassium  which showed little or  no
natural variability and  manganese which  was usually near  detection limits,
the combined lab and field variances were  generally below ten percent  of the
total variance.   This  is  consistent  with the  QA/QC  data  analyses,  which
showed that the  data  collection  errors were  generally quite  small.   The
entries  in the  table  have been  separated  into  water quality  parameters and
chemical  parameters of  geochemical  interest.    The results  confirm that  if
careful sampling  and  analytical  protocols are used, the  analytical and
sampling  errors can be held  to less than  about  20%.   Therefore,   the natural
variability in the  major  ion  chemistry of  the system can be  identified.  For
TOG and  TOX  it is clear that "natural"1  sources of variability  are greater
than the combined  lab  and  field variance. However,  the  level  of overall
variability in TOX  results was quite large  in comparison  to  the  mean values
for  each well.  The  significance of  these  determinations  at the microgram per
liter concentration level is  doubtful.

     A  similar attempt  to  estimate sources of variations  in ground-water
quality  data  was performed  by Summers et al.  (93).  These  workers analyzed
data from two sources  on related ground-water  samples  collected in the
vicinity of power plant waste disposal  impoundments.   They  reported that,  in
general,   combined  sampling and  analytical variability was less  than the
natural  variability.    They noted that combined sampling and  analytical
variability was usually  less than  about 15%  of the total  variability which
                                       83

-------
   TABLE 18.   PERCENTAGE OF VARIANCE ATTRIBUTABLE TO LABORATORYERROR,
        FIELD ERROR, AND NATURAL VARIABILITY BY CHEMICAL AND SITE

Type of
parameter
Water Quality
N(V
so4-
Si02
o-P04"
T-P04-
cr
Ca
Mg
Na
K
Geochemical

Nff2
s~
Fe2+
FeT
MnT
Contaminant
Indicator
TOC**
TOX**
Sand Ridge
lab

0.0
0.0
0.0
1.2
0.0
7.2
0.0
0.0
0.0
0.0

0.0
NA
NA
NA
0.0
0.0

lab
15.4
0.0
field

0.0
0.0
NA
1.2
NA
NA
45.7
20.0
NA
NA

0.0
NA
NA
NA
NA
NA

+ field


nat

100.0
100.0
100.0
97.6
100.0
92.8
54.3
80.0
100.0
100.0

100.0
NA
NA
NA
100.0
100.0


84.6
100.0
Beardstown
(upgradient)
lab

0.1
0.2
0.0
0.0
2.8
0.0
0.0
0.0
0.0
33.9

0.0
0.1
NA
0.0
0.0
0.0

lab
29.9
12.5
field

NA*
NA
20.0
0.0
NA
3.3
2.3
2.2
0.3
NA

0.0
NA
NA
0.1
0.0
40.1

+ field


nat

99.
99.
80.
100.
97.
96.
97.
97.
99.
66.

100.
99.
Beardstown
(downgradient)
lab

9
8
0
0
8
7
7
8
7
1

0
9
NA
99.
100.
59.

9
0
9


0
1
0
0
0
0
0
0
0
87,

0
0

.2
.4
.0
.0
.9
.0
.0
.0
.0
,1

.0
.3
NA
0
0
0

.0
.0
.0

lab +
70.
87.
1
5
40
24,
.6
,6
field

NA
0.1
6.8
0.0
NA
17.2
3.6
2.8
7.1
NA

0.0
NA
NA
5.9
NA
73.6

field


nat

99.8
98.6
93.2
100.0
99.1
82.8
96.4
97.2
92.9
12.9

100.0
99.7
NA
94.1
100.0
26.4


59.5
75.4

 * NA indicates  that the number of observations  on which the estimated
      variance was based was less than 5,  or  the estimated variance was
      negative.
** True  field  spiked standards not available  for these constituents,
      demanding  combined estimates of laboratory and field variability.
                                   90

-------
is consistent with the results  of  the  present  study.    Summers et al. did
report  exceptions in the  cases  of N03",  silicate,  and Zn where the  combined
sampling and analytical variance exceeded 30% of the  total  variance.   The
potential  sensitivity of  these  constituents  to  well   construct  ion  and
sampling errors was not discussed  in their report.

     The implication  of the results discussed above  and those  of this study
is that  network  design  optimization efforts  should focus  primarily on the
natural  variability.   The  use of field and laboratory  replication  for pur-
poses other than  QA/QC will be difficult  to justify as long as  the sampling
and  analytical  protocols are  in  control.   This  conclusion  must be qualified,
however.    The  chemical constituents present  at appreciable concentrations
(i.e.,  mg-L"1)     at either site  were the major  cations and anions and  general
water  quality indicators.  The analytical  and  sampling variances for trace
organic  contaminants would be  expected  to be higher,  and their analytical
recoveries  are frequently  found  to  be a function of  concentration.   For  such
contaminants,  the field and laboratory variations may  not be  independent,
which would  violate a basic assumption  in  this model.

     The field and  analytical  data were collected  with  very careful QA/QC  in
the  course of  this research project.   It  is unlikely that combinations  of
different laboratories  and  field sampling crews would be able to achieve  and
maintain such  low levels  of  error.   Further,  the  sites used in this study
provided fairly stable conditions of ground-water  flow  rates  and   direction
as well  as a steadily  leaking  source of  contamination.  In  addition,  the
contaminant  source itself was of fairly uniform composition.    The   effect  of
these conditions  would be  to minimize natural variability  making  the   degree
of sampling  and  analytical control  all  the more  critical.

Temporal Variations in Ground-Water Quality

     There are  numerous examples of both  short-  and long-term  variability  in
ground-water quality, in the  literature.    Recent  reviews by  Loftis  et al.
(54)  and Montgomery  et  al. (55)  pointed out the need for  very  careful selec-
tion of  statistical methods  and for qualified  interpretations of  existing
datasets.  The  well-documented  cases  of both short- and long-term   temporal
variability  in  ground-water quality  have been  tabulated in  Tables  19  and 20,
respectively.   These  observations cover temporal variability caused  by agri-
cultural  and  nonagricultural sources in high-volume  water  supply production
wells and  in  low-volume observation, monitoring,  and shallow  private wells
all in a variety  of hydrogeologic settings.   The  concentration variations
are noted mainly  as  multiples above and below an  arbitrary baseline  or back-
ground  concentration.  In a  few instances,   where the trends were clearly  very
long-term or cyclic  (i.e., due  to  alternate pumping and  nonpumping condi-
tions),   the  variations  have  been  entered  in  concentration units. Although
the  details of purging, sampling,  filtration/preservation, and  analysis were
frequently lacking in the  reports,  quite  substantial  variability has been
documented over  time-frames  ranging from minutes to  decades. Significant
short-term temporal concentration variability has  been  observed  in low yield
wells  (i.e.,  monitoring  and  observation  wells)  largely resulting from
purging  effects (6,101).  Similar variations from  one  to ten times  the  ini-
tial  or   background   concentrations   have   been   noted  in   samples   from

-------
      TABLE  19.  OBSERVATIONS  OF  TEMPORAL  VARIATIONS  IN  GROUND-MATER  QUALITY;  SHORT-TERM  VARIATIONS
Constituents
(concentration variation)
Agricultural Se (±2 mg-L~1)
Sources
304= (3-7X)
N03~ (1-UX)
N03~ (1-10X)
S04= (1-1 .5X)
N03~ (0.5-2X)
Atrazine (1-5X)
v£>
™ Non-Agricultural H2S (1-5X)
or mixed sources S04= (1-1. 2X)
NH? (1-3X)
J
S04= (1-2X)
Fe (1-3X)
Mn (1-1 .5X)
PCE, TCE, 1,2-t-DCE (1-10X)*
TCE (2-10X)
Fe2+ (1-110X)
3= (1-15X)
Volatile halocarbons (1-8X)
Nature
Period
Monthly
Minutes
Minutes
Monthly
Hours to weeks
Minutes to
hours
Minutes to
hours
Minutes
Minutes
Monthly to
weekly
Minutes
of variability
Probable cause
Irrigation/return/
indeterminate
Pumpage/head changes
and leaching from
unsaturated zone
i umpage/vertical
stratification
Irrigation/ fertilizer
applications/
leaching; locational
differences apparent
Surface runoff
recharge
Pumping rate and well
drilling
Pumping rate and
purging
Purging
Pumping rate and
purging
Pumping rate and
development of
cone of depression
Purging
Reference
9*4
95
96
97
98
57
99
100
101
102
6

* PCE = perchloroethylene,  TCE = trichloroethylene,  1,2-t-DCE = 1,2 trans-dichloroethylene

-------
       TABLE 20.   OBSERVATIONS OF TEMPORAL VARIATIONS IN GROUND-WATER QUALITY: LONG-TERM VARIATIONS
Constituents
(concentration variation)
Agricultural
Sources





soir
N03~
SOi,"
N03~
N03~
N03~
N03~
(+1.5X)
(2-1X)
(3-6X)
(3-7X)
(±18 mg-L~1/yr)
(1-12X)
(1-1. 5X)
(1-5X)
(1-1 .5X)
Nature
Period
Decades
Seasonal
Seasonal
Seasonal
Seasonal
Years-seasonal
of variability
Probable cause
Irrigation/ recharge
Irrigation/ precipita-
tion
Leaching/ recharge
Irrigation/fertilizer
applications
Recharge/ fertilizer
applications
Infiltr at ion/ recharge
Reference
103
101
105
97
106
98
                    Pesticides (1-1.5X)
Non-Agricultural Conductance (2~3X) Seasonal
or Mixed Sources S0ij= (1-3.5 X)
Hardness (2-6X)
Conductance ( + 2,000 yS-cnT1) Decades

N03~ (±55 mg-L~1/yr) Seasonal


Cl~ (1-3X) Seasonal

PCE (±1-20X)* Seasonal


TCE (±1-3X)* Seasonal


H20 level fluctuations 107
freezing/ thawing
recharge
Irrigation/upconing 108
of saline water
Sewage/fertilizer 109
recharge and
applications
Oil field brine/ 110, 111
recharge
Infiltrated surface 112
water quality
variations
Pumping rate and 102
patterns in well
field

* PCE = perchloroethylene,  TCE = trichloroethylene

-------
high-volume production wells due  to  pumping rate,  initial pumping  after
periods of inactivity and  cone of  depression  development (57,95,96,102,103).

     In general,  the major  ionic  chemical constituents  determined  in  this
study showed differences between their  overall maximum and minimum values
from the mean for  each  well on the  order of one or two times the mean  value
(see Appendices A and B).   One or two  times  the mean value places the varia-
bility  noted in  this study in the  same range as  the  long-term, seasonal
variability  noted in Table 20.    The magnitude of overall  long-term varia-
tions observed  in  this  study  and the literature  (Table 20) is often much
lower than  those noted for short-term variations  due to pumping and  local
recharge  effects.    The  magnitude of short-term  concentration  variations
noted  in  the literature  strongly suggests  that the  analysis of ambient
resource,  water  quality datasets  must be undertaken with careful attention
to the pumping  procedures used  in purging and sample collection.     This
observation is  particularly  Critical in relatively sparse datasets where
annual   "mean"  concentrations may be determined  from  programs with low
sampling  frequency (i.e.,  annually, biannually,  etc.).  Similar cautions  in
interpretations   of  long-term  datasets apply in the  analysis  of  trends at
varying or unequal  sampling  frequencies  due  to the relatively short  duration
of the  records  in comparison  to the  length of apparent annual  to  multi-year
variations.

     It was  expected that the high sampling frequency  (i.e.,  biweekly)  and
consistent purging  and  sampling  procedures   employed in  this  study would
permit the  identification of optimal frequencies for monitoring water
quality variations  under stable hydrologic  and contaminant source condi-
tions.   For  this reason,  field sampling  and  laboratory analytical protocols
were  carefully  controlled.

Sampling Frequency

     The  primary  purpose of the project was to  investigate  the  optimal
sampling  frequency  for  ground-water quality  monitoring.    Strictly  speaking,
there is no  minimum sampling frequency.   However,  there  is a  relationship
between the  information  content  of the  data and the sampling frequency.  The
term "information"  is sometimes  used  loosely, but  in  a  statistical  context,
it  can  be given  a more precise definition,  depending on the use of the  data.
The  most  common definition  of information (e._g.,.  in the  Fisher sense)  is  in
terms of  the variance  of  the mean, Var(x)  =  o2/n,  where x is  the sample
mean,   n  is  the  sample size,  and  a2  is  the  variance  of the data.   The
reciprocal  of the variance of  the mean  is a measure of the information  con-
tent of the  data. If the a2 is large, or the  sample size  small,  the informa-
tion content is low.  While this  definition of  information  applies  to
estimation of the mean,  the  power  of  trend  detection (in  space or time)  is
related to the variance  of the mean as well.

     As noted in the preceding discussion,  the  total variance is made up  of
the natural  variance and the variance attributable  to the  sample collection
process (field sampling  and  laboratory error).   Most monitoring  programs are
intended  to  discriminate  some effect (e.g., the  long-term  mean in the  case
of baseline  sampling,  or  the  difference  in  the  mean between upgradient and
                                      94

-------
downgradient wells  in the case  of  RCRA sampling)  from the total variation  in
the time  series.   The  effect of sample  collection variance might  be reduced
by  replicate sampling.    Although,  as  shown in  the previous section, the
sample  collection variance  made up such a  small  fraction of the  total  vari-
ance  that it  probably  would  not  be  worthwhile  for data  similar to  that
described here. The effect  of  natural variation can only be quantified by
increasing  the  number  of samples  (increased sampling  frequency or  length  of
sample  collection).  Increasing  the  number of  samples  also reduces the  effect
of the sampling  variance.   Seemingly,  the  information content of the data
could be increased  arbitrarily,   since it  depends  linearly  on the  sample
size.   In practice,  though,   ground-water  quality data  are  correlated in time
(autocorrelated),   and  the autocorrelation  increases   with the  sampling
frequency.   When  the  data are autocorrelated,  the variance  of the  mean can
be  reexpreaaed as  Var(x)  = 
-------
correlation was retained  and averaged with  the  other  estimates  for the same
chemical in the given well group.

     Seasonality and long-term trends in the data  presented  a major compli-
cation  in  the  analysis.    There are  well-structured  methods  for  handling
seasonality in  time  series,  but  none are  applicable  to  the  relatively short
(i.e.,  in terms of total duration) chemical time series  that were available
for analysis.    The problem is  that,  to properly  estimate  a seasonal  model,  a
relatively large number  of seasonal  cycles  (e.g.,  at  least  10) are required;
this corresponds to,  say, ten  years  of  data,  which  greatly  exceeds  the
length of the sampling  horizon.   Ignoring  the  seasonality tends to  inflate
the  estimate  of  the autocorrelation coefficient,  as  does  the existence of
trends in the  data.   There is no completely satisfactory  solution  to this
problem.   Our  approach was  to identify  series with apparent  strong seasonal-
ity or  long-term trends  subjectively  (seasonality in some variables, such as
temperature,  is apparent, and  can be argued  from first principles). Table 21
identifies those series for  which there  was  apparent  strong  seasonality,  as
well as the number  of violations of  the diagnostic checks for each variable
and  well group.  The  maximum possible  number  of  violations for each  variable
was  twice the  number of wells in the group, since two  tests were applied.
Subsequent results for series  showing  a high  number of rejections,  or for
which  there was strong  apparent seasonality  or  long-term trends, should be
interpreted  with caution.   However,  these  problems were not an  issue for a
large number  of series.   By  summarizing the results  over  well  groups,  and to
a  more limited  extent,  over  chemical constituents,  it  is  possible to  give a
general  picture of  the  sampling frequency  dependence  of the effective
independent  sample size,  which is relatively unaffected by  the peculiarities
of individual variables  or  sites.

     Table 22  gives  the average lag  one correlation for each variable  and
well  group,  ordered by the sum of the ranks over all well groups.  Variables
at the  top  of  the list tended to have the  lowest autocorrelation,  while
variables at  the bottom  were  most highly  autocorrelated.   Also given is  the
average  autocorrelation  over all  three  well  groups.  Autocorrelations  tended
to be stronger  at the Beardstown  wells than at Sand Ridge  and were higher at
the Beardstown upgradient  wells  than at the downgradient wells.   The  latter
effect  may be  due to randomness introduced by the  release, migration  and
transformation  of the contaminants.   Autocorrelations  for  almost all  varia-
bles, even  those with no  apparent trends  or  seasonality, were quite  high,
suggesting that there  was considerable redundancy  in  the  data at a biweekly
sampling frequency.

     To  illustrate  the  effect of  the  autocorrelation on sampling frequency,
we solved for  the  sampling  interval,  in weeks,  that  would  result in ratios
nef/n =  0.5, 0.8,  and 0.9  using  equation 13  of  Lettenmaier  (113).  Alterna-
tiVely,  these  can be interpreted as relative  losses  of  information due to
autocorrelation  in  the data  of  50, 20, and  10 percent.  The results are given
in Table 23.   At Sand Ridge,  the implied  loss  of  information was about 50
percent for  many  variables at  a  weekly sampling  frequency,  20 percent for
many variables  at  sampling intervals  in the range  of 4-8  weeks,  and  10 per-
cent for the  majority of variables  at  a sampling interval of 8 weeks or
more.  At  the  Beardstown wells,  the loss  of information at high sampling
                                       96

-------
        TABLE 21.  SUBJECTIVE ESTIMATE OF STRENGTH OF SEASONALITY
                   OR TREND IN VARIABLES BY LOCATION

Sand Ridge Beardstown
(1-4) (up gradient)
pH
Cond +
Temp C + +
Temp W + +
Eh
Probe 02
Wink 02
Alkalinity +

ND33"N
N03'N(VN
HS
S04=
Si02 *
o-P04 =
T-P04
cr
Fe2+
Ca * *
* K.
Mg *
Na *
K *
FeT
MnT
TOX
voc
NVOC
TOG
Beardstown Number of
(downgradient) violations
0
+ 2
+ 6
+ 4
1
0
0
1
3
1
0
*
0
0
0
* 1
* 1
+ 2
* 3
+ 1
2
* 3
* 3
0
+ 0
2
6
* n
3

     +  Indicates strongly seasonal
     * Indicates apparent trend  or possible seasonality
TEMP C = Flow cell temperature,  TEMP W = temperature  reading  in  well
   TOG = VOC + NVOC;  Total Organic Carbon = Volatile  Organic Carbon +
         Nonvolatile Organic Carbon
  Cond = Conductance

-------
TABLE 22.  RANKING OF AVERAGE LAG ONE CORRELATION OVER
         ALL SITES, FROM SMALLEST TO LARGEST

Sand Ridge
(1-4)
N02~N
Fe2+
pH
S"
NTJ
i\ n Q
Si02
MnT
Probe 02
T-P04~
o-P04=
Eh
N03N02~N
TOC
S04 =
FeT
K
Ca
Mg
Cl"
Na
Alkalinity
Ion
balance
Temp C
VOC
Cond
TOX
Temp W
NVOC
.27
.01
.51
.16
.29
.37
.51
.41
.06
.10
.46
.75
.46
.59
.21
.31
.45
.49
.19
.47
.73
.73

.54
.54
.80
.80
.66
.66
Beardstown
(5-6)
.42
.86
.47
.36
.82
.76
.47
.66
.20
.19
.60
.35
.60
.53
.90
.89
.92
.91
.96
.95
.69
.69

.92
.92
.94
.94
.97
.97
Beardstown
(8-13)
.37
.56
.20
.67
.26
.24
.20
.44
.86
.91
.60
.42
.60
.52
.66
.71
.66
.65
.75
.65
.76
.76

.79
.79
.75
.75
.78
.78
Summed
rank
17
18
25
26
28
28
28
30
32
33
34
36
37
39
40
46
50
50
54
56
62
62

69
70
73
74
76
77
Average
(over all
three well
groups)
(rho)
.35
.48
.39
.40
.46
.46
.39
.51
.37
.40
.55
.51
.55
.55
.59
.64
.68
.68
.63
.69
.73
.73

.75
.75
.83
.83
.80
.80
                       98

-------
TABLE 23.  SAMPLING INTERVALS (IN WEEKS) FOR GIVEN RATIO OF EFFECTIVE TO
  INDEPENDENT SAMPLE SIZE, BASED ON THE ESTIMATED LAG ONE MARKOV MODEL


Sand Ridge
N02" N
Fe2+
pH
S=
Ha,
MnT
Probe-02
T-P04=
0-P04=
Eh
N03N02'N

S04=
FeT
K
Ca
Mg
cr
Na
Alkalinity
Ion
balance
Temp C
voc
Cond
TOX
Temp W
NVOC
Beardstown Upgradient
N02" N
T-I 2+
Fe
pH
S~
NH2
Si02
MnT
Probe 02
T-P04=
o-P04=
0.5

2
1
4
2
2
3
4
3
1
1
3
8
3
5
2
2
3
4
2
3
7
7

4
4
10
10
6
6

3
15
3
3
1 1
8
3
6
2
2
orf'"

4
1
7
3
4
5
7
5
2
2
6
16
6
9
3
4
6
7
3
6
14
14

8
8
20
20
11
1 1

6
29
6
5
22
16
6
11
3
3
0.9

5
2
9
4
5
6
9
7
3
3
8
21
8
12
4
5
8
9
4
8
19
19

10
10
27
27
15
15

7
39
8
6
30
22
8
15
4
4
                                     continued  on  next page


                                99

-------
TABLE 23.   (continued)


Eh
N03N02~N
TOC
S04 =
FeT
K
Ca
Mg
Cl"
Na
Alkalinity
Ion
balance
Temp C
VOC
Cond
TOX
Temp W
NVOC
Beardstown Downgradient
N02~N
Fe2+
pH
S"
NH,
Si02
MnT
Probe 02
T-P04=
o-P04=
Eh
N03N(VN

S04 =
FeT
K
Ca
Mg
Cl"
Na
Alkalinity
0.5
5
3
5
4
21
19
26
23
53
42
6
6

26
26
35
35
71
71

3
4
2
6
2
2
2
3
15
23
5
3
5
4
6
7
6
5
8
5
8
nef/n
0?8
9
5
9
7
42
38
53
47
107
85
12
12

53
53
71
71
143
143

5
8
3
1 1
4
4
3
6
29
47
9
6
9
7
11
13
1 1
1 1
16
1 1
16
0.9
12
6
12
10
56
51
71
62
144
114
16
16

71
71
95
95
192
192

6
11
4
15
5
5
4
8
39
62
12
7
12
9
15
18
15
14
21
14
22
            concluded on next page
       100

-------
                TABLE 23.   (concluded)

                                      nef/n
                             0.5      0.8
Ion                            8       16         22
  balance
Temp C                         10       19         25
VOC                            10       19         25
Cond                           8       16         21
TOX                            8       16         21
Temp W                         9       18         24
NVOC                           9       18         24
                        101

-------
 frequencies  was  much greater.    At  the  upgradient  wells,  which  had the
 highest  autocorrelation,   the inferred  loss  of information of 50  percent
 occurred  for Several variables  at a  sampling  interval  of  over 26  weeks.
 Information  loss  of between 20 and  10 percent was  inferred for some varia-
 bles at sampling intervals  exceeding one year. This  effect was particularly
 evident for Na + ,  Cl"  and well-head  temperature  (TEMPW) which  showed an
 increasing trend  over the study  period.

      The  results  of the study indicate that,  for the major chemical  constit-
 uents (i.e.,  water  quality  or  contaminant indicator),  quarterly  sampling
 represents  a good  starting point for  a preliminary network design.    Some
 estimated ranges of  sampling  frequency to  maintain  information losses below
 ten percent are  shown in  Table  24.    This frequency,  of course, must be
evaluated with respect to  the  purpose  and time-frame over which the  network
 will  be conducted.   Under  the conditions of this  study, sampling four to  six
 times per year would provide an  estimated information loss below 20% and
 minimize  redundancy.  The  results  for  reactive,  geochemical  constituents
 suggest that bimonthly sampling frequency  would be a  good  starting point if
 chemical  reactivity  and  transformation are  of concern.

      It is clear that for  common chemical constituents,  a suggested sampling
 frequency of bimonthly  or  quarterly would represent  a  reasonably  efficient
 monitoring  design for evaluation of general water quality.  The wide ranges
 of sampling frequencies  shown in Table  24  for geochemical and trace-level
 reactive  constituents are in some  measure  a reflection of random analytical
 and  natural variability.    This  is particularly  true  for species  like  dis-
 solved  oxygen,  NH3, N02",  sulfide  and  ferrous iron under  oxidizing or
 suboxic conditions where  their concentrations  were expected to  be at  or near
 detection limits.   Determinations of these species can be diagnostic of oxi-
 dation-reduction  intensity situations in ground water and have major impli-
 cations for  the  design of remedial action  activities predicated  on microbial
 transformations  or   chemical reactions  to encourage  contaminant   removal
 (67,115,116,117).      Determinations  of  total organic halogen (TOX) are
 probably  not relevant  to uncontaminated  situations because the variability
 represented  in  these values in this  study  was very  nearly all analytical.

      A  subsampling  experiment was performed  to  evaluate the  usefulness of
 the  estimated  sampling  frequencies   for ambient  average or unusual event
 detection (i.e.,  90th percentile  excursions)  of selected types of chemical
 constituents (Table  24).    The procedure has  been fully  described elsewhere
 (119).    Reduced sampling frequency  subsets  were derived from  the  39 run
 biweekly  base  dataset.   In this  experiment  it was  assumed  that the base
 dataset represented  the  "true"  existing ground-water  quality  conditions at
 the two sites.    This assumption was  made for  the  sake of  practicality, and
 one  should  recognize that ambient  ground-water quality is  the result of
 stochastic processes.  Conclusions  drawn  from this analysis are diagnostic
 and  should  be used  only  as preliminary design parameters.

      The  base dataset was  broken  down by  F sequential sampling intervals (in
 weeks) and  N subsets, as follows:
                                       102

-------
    TABLE 24.  ESTIMATED RANGES OF SAMPLING FREQUENCY (IN MONTHS) TO
         MAINTAIN INFORMATION LOSS AT <10% FOR SELECTED TYPES OF
	CHEMICALPARAMETERS	
                        Pristine background   	Contaminated	
Type of parameter           conditions        Upgradient   Downgradient
Water  Quality

  Trace constituents
    (<1.0  mg-I/1
  Major  constituents

Geochemical
  Trace constituents
    (<1.0  mg-*^1
  Major  constituents

Contaminant  Indicator

  TOG
  TOX
  Conductivity
  pH
2 to 7

2 to 7



I to  2

1 to 2
2
6 to 7
6 to 7
2
1 to 2

2 to 38
7 to  14
3
24
24
2
2 to 10

2 to 10



1 to 5

1 to 5
3
7
7
1
                                   103

-------
          F      N              Description

          2      1      base dataset  of  39  runs
          421   subset   of runs,  1 o f 19 runs
          6      3      subsets of 13 runs
          843   subsets   of runs,  1 o f 9 runs
         12      5      4   subsets   of runs,  1 of 7 runs
         24      3      subsets of 12 runs
         78      1      subset of 39  runs

These  consecutive sampling intervals corresponded to 2-week,  1-month, 1.5-
month,  2-month, quarterly,  semiannual  and  1.5-year sampling  frequency.   The
base dataset and  the  subsets  were then  each rank ordered.   The mean,  median
and  the 90th percentile statistics were  calculated  then for the base dataset
as well as the  subsets.    At  a given sampling  frequency a  measure  of  the
"error"   of  the  subset  statistics  relative  to the  "true"  statistic  from  the
base dataset  was  defined as the  average percent  deviation from  the  "true"
value.    The acceptable  sampling frequency  for a particular parameter was then
selected as the  lowest  frequency  for which the average  percent  deviation  fell
below  an arbitrary cutoff of 10 percent. This  type of approach (i.e.,  distri-
bution-free,  nonparametric  method for percentile estimation) has been used in
previous studies  of  water  quality variability  (118,  119,  120) when assumptions
of normality and independence  in the  data  are inappropriate.

     Under the  conditions of this study there were  several  apparent trends in
the  results of the  subsampling experiment  which  are  of note.   The minimum
sampling frequencies  to  estimate the  means of selected  types of chemical
parameters within  ±10 percent are  provided in  Table  25.    Similar  to  the
results  in  Table 24,   it  is clear that  a  quarterly  frequency  represents  a
conservative  estimate  for an  initial sampling interval for determining  the
major ionic constituents as well  as pH,  TOG  and  conductivity.   These results,
therefore,   support the  regulatory  minimum  sampling frequency  for  those  param-
eters.    Reactive and  trace level constituents  seem to require a  somewhat more
frequent sampling interval for  adequate estimation of the "true"  mean.  Within
the  limits  of  this subsampling experiment,  the minimum  sampling  frequencies to
estimate the  median (i.e.,   50th percentile)  base dataset  values within ±10
percent were  not much  different  from that  of  the  mean.   The minimum sampling
frequency  necessary  to estimate  extreme values  (i.e.,  90th percentile) was
usually more  frequent  than  quarterly  under the  range of hydrogeochemical
conditions  of  the  study.

     Caution  must  be exercised in interpretation of these  results due  to  the
effects  of seasonality  and  long-term trends.   However,  it should be clear
that there  is considerable redundancy in the data at the  two-week sampling
interval,  and that,  at similar  sites and  for most of the  variables studied,
operational  sampling programs  would be  inefficient at sampling intervals more
frequent than bimonthly.  The practical  implication of this  is  that,  for  many
operational  monitoring programs,  a relatively long  time horizon (e.g.,  on  the
order of ten years)  may be required  to  obtain  adequate information for deci-
sion-making purposes,  given that high frequency sampling will  not yield much
increase in information.
                                     104

-------
     TABLE 25.  MINIMUM SAMPLING FREQUENCY (IN MONTHS) TO ESTIMATE THE
               MEAN OF THE BASE DATASET WITHIN 10 PERCENT
Type of parameter
Pristine  background
    conditions
   (Wells #1-4)
 Contaminated  conditions
Upgradient   Downgradient
(Wells #5,6) (Wells #8-13)
Water  Quality

  Trace  constituents
    (<1.0 rng-L'1)
  Major  constituents

Geochemical

  Trace  constituents
    (<1.0 mg-L"1
  Major  constituents

Contaminant Indicator

  TOC
  TOX
  Conductivity
  pH
                                 >6
                                 >6
                                                    6

                                                   26
                           6
                           1
                           6
                           3
                                         3

                                        26



                                        <1
                  >6
                   6
                                      105

-------
     It  is  important  to  emphasize that the information from sampling  depends
on  the effective  independent  sample size,  not just  the  ratio nef/n.
Therefore,  if the autocorrelation is large so that a  relatively low sampling
frequency  is necessary  to avoid  sampling redundancy, the  total length of the
sampling  period must be increased  to  achieve  sufficient information return.
These results  cannot  simply be  interpreted  to mean, for instance,  that
quarterly sampling is  adequate, unless that  interpretation is couched  in  terms
of the time horizon  of the sampling  program.   This conclusion of the study has
been  supported recently  by the work of Bell and  DeLong  (121) who reported the
results  of  an eight-year long,  monthly  data  evaluation  for  tetrachloroethylene
in ground  water.  They  noted that  the tetrachloroethylene  concentration data
showed no evidence of seasonality,  normality  or  serial  correlation in the
first  three years  of  monthly sampling,  but some definable  trends appeared  in
the fourth through eighth years  of sampling.   Their work supports  our general
recommendation of quarterly  sampling frequency  as an initial design parameter
in network design and  underscores the  need to collect long-term  datasets  in
order to define  temporal trends  in chemical constituent  concentrations.
                                     106

-------
                                 SECTION 6

                                 REFERENCES
 1.   Barcelona,  M. J.,  J.  P.  Gibb,  and R.  A.  Miller.  A guide to the selec-
     tion  of materials for monitoring well construction  and ground-water
     sampling.   ISWS Contract Report 327,  prepared  for USEPA-RSKERL, Ada,
     OK, and USEPA-EMSL, Las Vegas, NV,  EPA  600/S2-84-024, 78 pp.,  1983.

 2.   Barcelona, M. J.,  J. A.  Helfrich,  E. E. Garske, and J.  P.  Gibb.  A
     laboratory  evaluation of ground-water sampling mechanisms.    Ground
     Water Monitoring Review 4(2) :32-41,  1984.

 3.   Barcelona, M.  J.,  J.  P. Gibb, J.  A. Helfrich,  and E.  E. Garske.
     Practical guide  for  ground-water sampling.    ISWS Contract  Report 374
     prepared for USEPA-RSKERL, Ada, OK,  and USEPA-EMSL, Las Vegas,  NV,
     EPA  600/S2-85-104, 94  pp.,  1985.

 4.   Barcelona,  M. J., J.  A.  Helfrich, and E. E. Garske.  Sampling tubing
     effects  on  ground water  samples.   Analytical Chemistry 57(2):  460-464,
     1985.

 5.   Garske,  E.  E.,  and M. R.  Schock.  An  inexpensive  flow- through cell and
     measurement  system for  monitoring  selected chemical  parameters   in
     ground  water.   Ground  Water Monitoring  Review 6(3):79-84,  1986.

 6.   Barcelona, M. J.,  and J.  A.  Helfrich.  Effects of  well construction
     materials on  ground  water  samples.   Environmental  Science  and Tech-
     nology  20(11):1179-1184,  1986.

 7.   Delhomme,  J.  P,.   Spatial variability  and uncertainty in  groundwater
     flow  parameters:  a  geostatistical approach.   Water  Resources  Research
     15(2):269-280, 1979.

 8.   Russo,  D.,  and  E.  Bresler.    Soil hydraulic  properties as  stochastic
     processes: I. An  analysis  of field  spatial  variability.    Soil Sci._
     Soc.  Am. 
-------
12.   Horton,  R.  E.    The  role  of infiltration in  the hydrologic cycle.
     Trans. Amer.  Geophys. Union  14:  446-460,  1933.

13.   Rubin, J.,  and R. Steinhardt.   Soil water  relations during rain infil-
     tration:  I.  Theory.   Soil Sci, Soc. Amer.  Proc.. 27: 247-251,   1963.

14.   Rubin,  J.,  R. Steinhardt,  and  P. Reiniger.    Soil  water  relations
     during  rain  infiltration:   II. Moisture  content profiles  during rains
     of low  intensities.   Soil Sci.  Soc.. Amer. Proc.. 28: 1-5,  1964.

15.   Sophocleous,  M.,  and  C.  A. Perry.   Experimental studies in natural
     groundwater-recharge dynamics:     The analysis  of observed  recharge
     events.   Journal of Hydrology 81:  297-332,  1985.

16.   Freeze,  R.  A.  The  mechanism of natural groundwater  recharge  and  dis-
     charge:     1.  One-dimensional,  vertical, unsteady,  unsaturated flow
     above  a recharging  or discharging groundwater flow  system.   Water
     Resources Research  5(1):153-171,  1969.

17.   Freeze,  R.  A., and J. Banner.   The  mechanism of  natural  ground-water
     recharge and discharge:    2.  Laboratory column experiments  and field
     measurements.  Water Resources Research 6(1): 138-155,  1970.

18.   Walton,  W.  C.  Groundwater resource  evaluation. McGraw-Hill Inc., New
     York, 664 pp., 1970.

19.   Johansson,  P.   Estimation  of groundwater recharge in sandy till with
     two  different methods using  groundwater  level  fluctuations.    Journal
     of Hydrology  90: 183-198, 1987.

20.   Chow, V. T.   Handbook of applied hydrology. McGraw-Hill  Book  Company,
     New  York.  1418  pp.,  1964.

21.   Suter,  M.,  and R. H.  Harmeson.    Artificial ground-water  recharge at
     Peoria,   Illinois.     Bulletin  48,   Illinois   State Water   Survey,
     Champaign,  Illinois,  48 pp., 1960.

22.   Winquist, G., and K.  Marcelius.   The  design of artificial  recharge
     schemes.     In:    Proceedings for Artificial  Ground  Water Recharge
     Conference,  Water Research  Association,  England, 65 pp.,  1970.

23.   Naymik,  T.  A.,  T.  E. Larson,  and  M.  R. Schock.   Water  reuse  through
     groundwater  recharge  in northeastern  Illinois.   Unpublished coopera-
     tive ground  water  report.    Illinois  State Water  and  Geological
     Surveys,  Champaign,  IL,  183 pp.,  1984.

24.   Kimmel,  G. E.   Nitrogen  content of ground water  in  Kings  County,  Long
     Island, New  York.   Professional  Paper 800-D,  U.S. Geological Survey,
     pp. D199-D203, 1972.
                                     108

-------
25.   Morton, B.  L., J. J.  Kolega,  and  H.  D. Luce.   Ground water  monitoring
     at a  septage lagoon  facility  in  Connecticut.   Ground Water  Monitoring
     Review  4(2):45-50,  1984.

26.   Hantush, M. S.  Growth and decay of  ground water mounds  in  response to
     uniform  percolation.    Water  Resources Research 3(l):227-234,  1967.

27.   Marino,  M. A.  Water-table fluctuation in  response to recharge. Paper
     No. 10582,  J.  Irrigation and  Drainage Div., ASCE,  100  (IR2): 117-125,
     1974.

28.   Singh,  R.    Prediction  of Mound Geometry  Under Recharge  Basins.   Water
     Resources Research 12(4):775-780, 1976.

29   Rao,  N.  H.,  and  P.   B.  S.  Sarma.    Growth of Ground-Water Mound in
     Response  to Recharge.   Ground Water  18(6):587-595,  1980.

30.   Parker,  G. G., and V. T. Stringfield.  Effects  of  earthquakes, trains,
     tides,   winds,   and  atmospheric pressure   changes on  water  in  the
     geologic  formations  of southern Florida.    Econ. Geol._45: 441-460,
     1950.

31.   Russell,  R.  R.   Ground-water levels in  Illinois through 1961. Report
     of Investigation  45,   Illinois State Water  Survey,  Champaign, Illinois,
     51 pp.,  1963.

32.   Roberts, W.  J.,  and  H.  E.  Romine.    Effect of train loading on  the
     water  level in a  deep  glacial-drift  well  in  central  Illinois.   Trans.
     Amer. Geophys. Union, 28:  912-917, 1947.

33.   Freeze,  R. A., and J. A. Cherry.    Groundwater.   Prentice-Hall,  Inc.,
     Englewood Cliffs,  New Jersey, 604 pp., 1979.

34.   Peck, A.  J.   The water  table  as  affected by atmospheric  pressure.
     Journal  of  Geophysical Research  65(8):2383-2388,  1960.

35.   Norum,  D.  I., and J.  N. Luthin.    The effects of entrapped air  and
     barometric  fluctuations on the  drainage  of porous mediums.    Water
     Resources Research 4:  417-423, 1968.

36.   Turk, L.  J.   Diurnal fluctuations of  water tables induced  by atmos-
     pheric pressure changes.  Journal of Hydrology  26(1):1-16, July  1975.

37.   King,  F. H.   Seismic oscillations of  the  ground-water surface. U.S.
     Weather Bureau  Bulletin 5:  67-68,  1892.

38.   Jacob,  C. E.   Fluctuations  in  artesian  pressure produced by  passing
     railroad trains as  shown in a well on Long  Island, New York.   Trans.
     Amer. Geophys. Union  20: 666-674,  1939.
                                      109

-------
39.   Piper,  A.  M.   Fluctuations of water-surface  in  observation wells and
     at stream-gaging stations in the  Mokelumne  area,  California,  during
     the earthquake of December,  1932.   Trans. Amer.  Geophys. Union  14:
     471-475, 1933.

40.   Thomas, H.  E.   Fluctuations of ground-water levels during the  earth-
     quakes  of  November  10,  1938  and  January  24,  1939.   Seis. Soc. Amer.
     Bull.,   30:   93-97,  1940.

41.   Scott,  J.  S.,  and F. W. Render.    Effect of an  Alaskan earthquake  on
     water  levels  in wells  at  Winnipeg and  Ottawa, Canada.   Journal  of
     Hydrology  2(3):262-268, 1964.

42.   Meyboom, P.   Unsteady groundwater flow near a willow ring  in hummocky
     moraine.  Journal of Hydrology 4(l):38-62,  1966.

43.   Johansson,   P.  Diurnal groundwater level  fluctuations in  sandy till -a
     model  analysis.  Journal  of Hydrology 87:125-134,  1986.

44.   Todd,  D. K.,  R. M.  Tinlin, K.  D.  Schmidt, and  L.  G. Everett.  Monitor-
     ing ground water quality:   monitoring methodology. EPA  600/4-76-026,
     U.S.   Environmental    Protection  Agency,   EMSL,   Las   Vegas,   Nevada,
     172 pp., 1976.

45.   Sanders, T.  G.,  R.  C. Ward, J.  C. Loftis,  T.  D. Steel,  D. D. Adrian,
     and V.  Yevjevich.    Design of networks for monitoring  water quality.
     Water  Resources Publications,  Littleton, Colorado, 328  pp., 1983.

46.   Moss,  M.  E.,  D. P.  Lettenmaier,  and  E.  F.  Wood.    On the  design of
     hydrologic    data   networks.      EOS-Transactions of  the   American
     Geophysical  Union   59(8):772-775,   1978.

47.   Liggett,  W.  S.,  Detecting elevated contamination  by comparison with
     background. In: G.  E. Schweitzer  and  J.  A. Santoluci to  (Eds.),
     Environmental Sampling  for  Hazardous  Wastes,  Chapter  13.    American
     Society  Symposium  Series  #267,  Washington, D.C.,  pp. 119-128,  1984.

48.   Liggett, W. S.  Statistical aspects of designs for studying sources  of
     contamination in quality assurance for  environmental measurements.  In:
     J.  K.  Taylor and T.   W. Stanley  (Eds.),  ASTM STP 867. American  Society
     for Testing and  Materials, Philadelphia, Pennsylvania, pp.  22-40,
     1985.

49.   Gillham,  R.  W.,  J.   L.   Robin,   J.  F.  Barker,   and  J.   A.   Cherry.
     Ground-water  monitoring and  sample  bias.    API  Publication    #4367,
     American Petroleum  Institute, 206  pp., June,  1983.

50.   Barcelona,  M.  J.  Overview  of  the sampling  process.  In:   Lawrence H.
     Keith (ed.), Principles  of Environmental  Sampling,  Chaser  1.  American
     Chemical  Society  Professional Reference  Book.    American  Chemical
     Society,  Washington,  D.C., pp.  4-23,  January, 1988.
                                      110

-------
51.   Barcelona, M.  J., J. A. Helfrich,  and  E.  E.  Garske.   Field verifica-
     tion of  sampling methods  and materials selection for  ground-water con-
     tamination  studies. In: ASTM-STP 963,  Proceedings  of  Symposium  on
     Field Methods  for  Ground Water  Contamination Studies  and  Their
     Standardization  for the American Society for  Testing and  Materials,
     Cocoa Beach, Florida,  pp.  221-231,  February 2-7,  1986.

52.   Barcelona,  M.   J.,   and   J.P.   Gibb.    The  development of effective
     ground-water sampling protocols. In:   ASTM STP-963,  Proceedings  of
     the Symposium on Field Methods for Ground Water Contamination  Studies
     and Their Standardization for  the American Society for Testing and
     Materials, Cocoa Beach, Florida, pp.  17-26, February 2-7,  1986.

53.   Porter, K.  S.,   and  N.  M.  Trautman.     Seasonality in groundwater
     quality.   Draft  report by  Cornell  University Center  for  Environment
     Research for the U.S.   Environmental Protection  Agency, EMSL, Las
     Vegas,  Nevada, 26 pp., 1984.

54.   Loftis,  J. C.,  R. H.  Montgomery, J.  Harris, D.  Nettles,  P. S.  Porter,
     R. C. Ward,  and  T.  G.  Sanders.   Monitoring strategies  for  ground-water
     quality  management.    Final  report,   USGS Grant  14-08-0001-G-1060,
     Colorado  Water  Resources Research  Institute,  Fort Collins,  Colorado,
     125  pp., 1986.

55.   Montgomery,   R.  H.,  J.   C.  Loftis,   and J.   Harris.     Statistical
     characteristics  of  ground water  quality  variables.    Ground Water
     25(2):176-184,  1987.

56.   Harris,  J.,  J. C. Loftis,  and R. H. Montgomery.   Statistical methods
     for characterizing ground-water quality.   Ground  Water 25(2): 185-193,
     1987.

57.   Colchin,  M.  P.,  L. J.  Turk, and M. J. Humenick.   Sampling of ground
     water  baseline and monitoring data  for in-situ  processes.    Report
     #EHE-78-01, CRWR157,  Texas  Water Resources Research Center,  Austin,
     Texas,  171 pp.,  1978.

58.   Horberg, L.    Bedrock topography of Illinois.    Bulletin 73,  Illinois
     State Geological  Survey,  Champaign,  Illinois, 111 pp.,  1950.

59.   Walker,  W. H., R. E.  Bergstrom, and W. C. Walton.   Preliminary  report
     on  the  ground-water  resources of the Havana  region in west-central
     Illinois.  Cooperative  Ground-Water Report  3, Illinois  State Water
     Survey  and  Illinois  State Geological  Survey,  Champaign,  Illinois,
     61  pp.,  1965.

60.   United  States Geological  Survey.    Water  resources data,  Illinois,
     water  year  1986: Volume  2. Illinois  River Basin.    Water Data report
     IL-86-2,  U.S. Geological Survey, 415 pp.,  1987.
                                    ill

-------
61.   O'Hearn, M.,  and  J.  P.  Gibb.    Ground-water  discharge  to  Illinois
     streams.     Contract Report CR246,   Illinois  State  Water  Survey,
     Champaign, Illinois, 112  pp.,  1980.

62.   Walton, W. C.    Ground-water  recharge and runoff  in  Illinois.   Report
     of   Investigation,  RI48,   Illinois  State  Water Survey,  Champaign,
     Illinois,  55 pp.,  1965.

63.   Naymik,  T.  C., and M. E.  Sievers.   Groundwater tracer  experiment (ll)
     at  Sand  Ridge  State Forest,  Illinois.   Contract  Report  334,  Illinois
     State  Water Survey, Champaign,  Illinois,  105  pp.,  1983.

64.   CH2M Hill.    Beardstown,   Illinois Ground-Water  Study.    Milwaukee,
     Wisconsin, 45 pp., 1982.

65.   Prosser,  D. W.    A method of  performing response tests  on highly
     permeable  aquifers.  Ground Water 19(6):588-592,  1981.

66.   Collins,  M. A.,  and D.  J.  Kelly.    A  microcomputer  controlled  data
     acquisition system  for  ground-water applications.     Illinois  State
     Water Survey  publication, in preparation, Champaign, Illinois,  1989.

67.   Holm, T.  R., G. K.  George, and M. J.  Barcelona.  Dissolved oxygen and
     oxidation-reduction potentials in ground  water,  EPA 6OO/S2-86/042.
     U.S.  Environmental  Protection  Agency,  56  pp.,  June,  1986.

68.   United States  Environmental  Protection Agency.   Methods for chemical
     analysis  of water and  wastes.   EPA  600/4-79-020.    U.S. Environmental
     Protection Agency,  EMSL, Cincinnati,  Ohio, 460  pp., March, 1980.

69.   United States  Environmental Protection Agency.   Technical additions  to
     methods  for chemical analysis of water and wastes.   EPA 600/4-82-055.
     U.S.  Environmental Protection  Agency, EMSL,  Cincinnati,  Ohio, 6  pp.,
     December, 1982.

70.   Carpenter,  J.  H.     The  Chesapeake  Bay  Institute technique  for the
     Winkler dissolved  oxygen  method.   Limnol.  and Oceanogr. 10:  141-143,
     1965.

71.   APHA-AWWA-WPCF,  Standard Methods  for the Examination of Water and
     Wastewater,  16th  Edition,  1268  pp.,  1985.

72.   Fishman,  M. J.,  and L.  C. Friedman (Eds.).  Methods  for  the  Determina-
     tion  of Inorganic  Substances  in  Water  and  Fluvial Sediments.  U.S.
     Geological  Survey.    Open File  Report 85-495,  Denver, CO,  709  pp.,
     1985.

73.   Ivancic,   I., and  D, Degobbis.      An  Optimal Manual Procedure for
     Ammonia  Analysis in Natural Waters  by  the  Indophenol Blue Method.
     Water Research  18(9): 1143-1147,  1985.
                                       112

-------
74.   McAuliffe,  C.    Gas Chromatographic Determination of  Solutes by
     Multiple Phase  Equilibration.  Chem. Technol.  I:  46-51, 1971.

75.   Brown,  E.,  M. W.  Skougstad,  and M.  J. Fishman.  Methods for Collection
     and  Analysis  of Water Samples for Dissolved  Minerals  and Gases.  USGS
     Techniques   of Water-Resources  Investigations  of the  United  States
     Geological  Survey,  Book 5, Chapter  Al, pp. 103-105, 1970.

76.   Smith,  R.,  E. M.  Bezuidenhout, and A.  M. van Heerden.    The use of
     interference  suppressants  in the direct  flame atomic  absorption  deter-
     mination of metals in water.  Water Research  17(11): 1483-1489,   1983.

77.   Plummer,  L. N.,  B. F. Jones, and A. H. Truesdell.   WATEQF - A  Fortran
     IV Version  of WATEQ,  A Computer Program for  Calculating Chemical
     Equilibrium of Natural Waters.  U.S. Geological  Survey Water Resources
     Investigations 76-13,.  Revised and Reprinted,  Dec.  1978.

78.   Midgley, D.,  and  K.  Torrence.   Potentiometric Water Analysis.    John
     Wiley & Sons, 409 pp., 1978.

79.   Siskos, P.  A.,  E. P.  Diamandis, E.  Gillieron,  and J. Colbert.
     Potentiometric titration  of sulphate,  sulphite  and dithionate mix-
     tures,    with  use   of   a  lead   ion-selective   electrode.   Talanta
     30(12):980-982,  1983.

80.   Moscini,  M.   Titration of  sulphate  in mineral waters  and  sea water by
     using the  solid-state lead electrode.    Analyst 98: 325-328,  May,
     1973.

81.   Hulanicki,  A., R. Lewandowski, and  A. Lewenstam.   Elimination  of ionic
     interferences in  the  determination of sulphates  in water  using the
     lead-sensitive   ion-selective   electrode.     Analyst   101:   939-942,
     December, 1976..

82.   Selig,  P.   Micro  and  semimicro determination of sulfur in organic com-
     pounds by  potentiometric titration  with  lead  perchlorate. Mikrochim.
     Acta  [Wien]  1: 168-175, 1970.

83.   American Chemical  Society.  Guidelines for Data  Acquisition and Data
     Quality Evaluation in Environmental Chemistry.    Anal.  Chem. 52:
     2242-2249,   1980.

84.   American  Society  for  Testing  and  Materials.    Standard  Practice  for
     Interlaboratory   Quality   Control  Procedures .and  a Discussion on
     Reporting  Low-Level  Data, D4210-83,  Philadelphia,  PA, 9  pp.,  1983.

85.   Skoog,  D.  A., and D.  M.  West.   Fundamentals of Analytical  Chemistry,
     Third  Edition.  Holt,  Rinehart  and Winston, New York, 804 pp.,  1976.

86.   Taylor,  J.  K.    Quality  Assurance  o_f  Chemical Measurements.    Lewis
     Publishers,  328 pp.,  1987.
                                      113

-------
87.   Parkhurst, D. L.,  D.  C. Thorstenson, and  L.  N. Plummer.   PHREEQE - A
     Computer    Program    for   Geochemical    Calculations.        U.S.G.S.
     Water-Resources  Investigations 80-96,  210 pp., 1980.

88.   Hvorslev,  M. J.  Time  lag and  soil permeability in ground-water obser-
     vations.    Bulletin  No.  36, U.S. Army  Corps of Engineers,  Waterways
     Experiment Station, Vicksburg, Mississippi,  50 pp.,  1951.

89.   Abriola,  L. M.,  and G.  F.  Finder.   Calculation  of velocity  in three
     space  dimensions from hydraulic head measurements.    Ground Water
     20(2):205-213,  1982.

90.   Ball,  J.  W.,  D.  K.  Nordstrom,  and D.  W.  Zachmann.    WATEQ4F  - A
     Personal  Computer Fortran Translation of the  Geochemical Model  WATEQ2
     with Revised Data  Base. U.S.  Geological  Survey Open-File  Report 87-50,
     108  pp.,  1987.

91.   Holm,  T.  R.,  G. K.  George,  and  M.  J.  Barcelona.   Oxygen Transfer
     Through  Flexible  Tubing  and Its   Effects  on Ground-Water  Sampling
     Results.  Ground  Water  Monitoring  Review  8(3):83-89,  1988.

92.   Barcelona,  M. J., G. K. George, and  M. R. Schock.  Comparison of water
     samples from PTFE,   PVC and SS monitoring wells.    EPA  Internal Report
     EPA 600/X-88/091 prepared for USEPA-EMSL, Las Vegas,  37  pp., February,
     1988.

93   Summers, K. V., G. L.  Rapp,  G. F.  Davis, and S.  A. Gherini.   Ground
     water  data analyses  at utility waste  disposal sites.    Project report
     #EA-4165 prepared  by Tetra Tech Incorporated for the  Electric Power
     Research  Institute,  Palo Alto,  CA,  420 pp.,  July,  1985.

94.   Crist,  M. A.    Selenium in  waters in and adjacent to Kendrick Project,
     Natrona County,, WY.   U.S.  Geological Survey Water Supply  Paper #2023,
     39 pp., 1974.

95.   Schmidt,  K. D.   Water  quality variations for pumping wells.   Ground
     Water 15(2): 130-137, 1977.

96.   Eccles, L. A.,  J.  M. Klein, and W.  F.  Hardt.   USGS scientists bring
     California water supply into  compliance  with  federal  regulations.
     Water  Well  Journal 13(2):42-45,  1977.

97.   Spalding,  R.  F.,  and M. E.  Exner.   Areal,  vertical and  temporal  dif-
     ferences   in  ground  water  chemistry,  I.   Inorganic   constituents.
     J. Env._ Quality  9(3): 466-479,  1980.

98.   Libra, R. D., G. R. Hallberg,  B. E. Hoyer,  and L. G.  Johnson.   Agri-
     cultural  impacts on ground water  quality:  the Big Spring Basin study,
     Iowa.   Proceedings of Agric.  Impacts  on  Ground  Water, August 11-13,
     Omaha, NE, National  Water Well Association,  Dublin, OH, pp.  253-273,
     1986.
                                       114

-------
 99.   Humenick,  M. J.,  L. J.  Turk,  and M. P. Colchin.    Methodology  for
      monitoring  ground water  at  uranium   solution  mines.    Ground Water
      18(3):262-273,  1980.

100.   Wilson, L.  C., and J.  V.  Rouse.   Variations in water quality during
      initial pumping  of monitoring wells.    Ground Water  Monitoring Review
      3:   103-109,  1983.

101.   Keely,  J.  F., and F.  Wolf.  Field applications of chemical  time-series
      sampling.   Ground  Water  Monitoring Review  3(4):26-33,  1983.

102.   McReynolds,  L.   Monitoring  organic contaminants  in  Los Angeles  San
      Fernando  Valley  ground water basin.   Proceedings of the 15th  Biennial
      Conference   on   Ground   Water,   p.   53-60.   In:   J.   Devries,    Ed.,
      California Water  Resources  Center,   State  of  California,  September
      23-25,  1985, San Diego, CA,  169  pp.,  1986.

103.   Evenson,  R. E.    Suitability  of irrigation water  and  changes in
      ground-water  quality in the  Lompoc  Subarea  of  the Santa Ynez River
      Basin,  Santa  Barbara,  CA, U.S.  Geological Survey  Water Supply Paper
      1809-S, 20 pp.,  1965.

104.   Tenorio, P.  A., R.  H. F. Young, and  H. C. Whitehead.   Identification
      of  return  irrigation  water in the subsurface.   Tech. Report #30,  Water
      Resources Research  Center,  University of Hawaii, Honolulu,  90 pp.,
      1969.

105.   Tyron,  C.   P.     Ground-water  quality variation  in  Phelps  County,
      Missouri.   Ground  Water  14(4):214-223,  1976.

106.   Rajagopal, R., and R.  L.  Talcott.  Patterns  in  ground-water  quality:
      selected observations in Iowa.  Environmental Management 1(5) :  465-474,
      1983.

107.   Feulner,  A.  J.,  and  R. G. Schupp.   Seasonal changes in the chemical
      quality  of  shallow ground  water in  northwestern Alaska. U.S.
      Geological  Survey  Prof.  Paper  #475-B, pp.  B189-B191,  1963.

108.   Handy, A.  H.,  R.  W.  Mower, and G.  W.  Sandberg.    Changes in the
      chemical  quality of  ground water  in three  areas in the Great Basin,
      UT.   U.S.  Geological  Survey  Prof. Paper #650-D, pp. D228-234,  1969.

109.   Perlmutter,  N. M., and E. Koch.   Preliminary hydrogeologic appraisal
      of  nitrate in ground  water and streams, Southern Nassau County, Long
      Island, NY.   U.S. Geological  Survey Prof.  Paper #800-B, pp. B225-B235,
      1972.

110.   Pettyjohn, W.  A.    Monitoring  cyclic fluctuations in  ground  water
      quality.   Ground Water 14(6):472-480,   1976.

111.   Pettyjohn, W.  A.   Cause   and effect of cyclic  changes in ground  water
      quality.   Ground Water Monitoring Review 2(l):43-49,  1982.
                                      115

-------
112.   Schwarzenbach, R.  P.,  W.  Giger, E.  Hoehn,  and  J.  K. Schneider.
      Behavior of organic compounds during infiltration  of river water  to
      ground  water:  field studies.    Environmental  Science  and  Technology
      17(8):472-479,  1983.

113.   Lettenmaier, D.  P.    Detection of trends  in  water  quality  data from
      records   with   dependent   observations.     Water Resources Research
      12(5):1037-1046,  1976.

114.   Box,  C.   E.  P.,   and  G. M.  Jenkins.   Time  series analysis:  forecasting
      and control,  Holden Day, 416 pp., 1970.

115.   Wilson,  J.   T.,   and  J.  F.   McNabb.    Biological transformations  of
      organic  pollutants  in groundwater.   EOS,  Transactions.  Amer. Geophys.
      Union 64:  505-506,  1983.

116.   Vogel, T. M.,  C. S. Griddle,  and  P.  L. McCarty.    Transformations  of
      halogenated aliphatic compounds.   Environmental Science and  Technology
      21(8):722-736  (errata   Env.  S.ci.  and  Techn.  21(12): 1140), 1987.

117.   Macalady, D. L., P.  C. Tratnyek, and  T. J.  Crundl.    Abiotic  reactions
      of anthropogenic organic chemicals in  anaerobic systems:  A  critical
      review.   J.  Contaminant  Hydrology  1:1-28,  1986.

118.   Ellis,  J.  C.,   and   R.   F.   Lacey.    Sampling:   defining  the  task and
      planning the scheme.  In:   River Pollution  Control,   Chapter  17, M.  J.
      Stiff, Ed.,  Ellis Horwood,  Ltd. Publ.,  Chichester,  U.K.,  pp. 247-274,
      1980.

119.   Harmeson,  R. H., and  M. J. Barcelona.    Sampling frequency for water
      quality  monitoring.   Illinois  State  Water Survey  Contract Report #279
      prepared for USEPA-EMSL,  Las  Vegas,  NV,  under contract  CR804337,
      88 PP.,  November, 1981.

120.   Crabtree, R.  W., I.  D. Cluckie,  C. F.  Forster.    Percentile  estimation
      for water quality data.  Water  Research  21(5):583-590,  1987.

121.   Bell,  H. F.,  and H.  P. DeLong.   Data Characteristics:  Ground Water
      Monitoring's  "Catch 22."   Proceedings  of the American Chemical  Society
      Division  of Environmental Chemistry  196th  National  Meeting,  September
      25-30, Los  Angeles,  CA,  American  Chemical Society,  Washington,  D.C.,
      pp.  20-24,  1988.
                                       116

-------
                                  SECTION 7

                                  APPENDICES


A.   Summary  of Analytical Results for  Sampling Wells.    (Constituent concen-
    trations are expressed in mg-L"1 except  as  noted.)

5.   Time  series of  individual  constituent   concentrations for  biweekly
    sampling runs  for each well at the Sand Ridge Site  (Wells #1,  #2, #3  and
    #4) and the Beardstown  site  (Wells #8,  #9,  #10,  #11, #12, and  #13).

C.   Ground  Water Elevations  Measured  During Each Biweekly  Sampling Run  at
    the Sand Ridge  and Beardstown Sites.  (Elevations  at the  Sand  Ridge  Site
    are in  feet  relative  to  an  arbitrary  1000  foot reference point.
    Elevations  at Beardstown are in feet relative to mean sea level.)
                                      117

-------
                            APPENDIX A

          SUMMARY OF ANALYTICAL RESULTS FOR SAMPLING WELLS
(CONSTITUENT CONCENTRATIONS ARE EXPRESSED IN mg-L1 EXCEPT AS NOTED.)
        TABLE A-l.   SUMMARY OF ANALYTICAL RESULTS FOR WELL 1

Parameter
Conductivity
(US-cm'1)
pH (pH units)
TOX (^g-I/1)
TOG
Temperature (°C)
Alkalinity
Chloride
(NO3 + NO2)-N
Sulfate
Ortho-PO4
Total PO4
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
41

44
44
44
37
44
44
44
44
44
44
43
44
44
43
42
44
38
42
43
22
44
44
44
43
39
44
44
44
Mean
359.27

7.75
3.65
0.85
1.98
216.26
2.19
0.95
36.18
0.02
0.04
15.54
65.85
22.58
3.17
0.70
0.47
9.00
8.83
0.01
0.01
0.00
-0.01
0.00
0.00
-0. 12
0.20
0.74
0.93
Std dev
10.569

0.532
5.011
0.260
0.924
11.582
0.707
0.200
5.761
0.015
0.048
0.315
3.344
1.226
0.265
0.037
0.090
0.503
0.875
0.037
0.031
0.009
0.014
0.005
0.002
0.788
0.566
0.183
2.280
Maximum
381.00

9.70
26.70
1.77
14.60
245.80
5.74
1.25
49.90
0.08
0.31
16.15
78.40
26.20
3.72
0.77
0.68
9.93
11.90
0.12
0.12
0.03
0.02
0.01
0.01
0.16
3.66
1.16
6.62
Minimum
340.00

6.80
0.00
0.00
10.50
184.10
1.36
0.62
20.60
-0.01
-0.00
14.85
60.55
21.07
1.80
0.61
0.30
7.85
7.17
-0.05
-0.03
-0.01
-0.06
-0.01
-0.00
-4.91
0.01
0.18
-2.50
Median
358.00

7.62
2.40
0.83
12.00
218.95
2.09
1.02
36.95
0.02
0.02
15.60
65.33
22.15
3.16
0.71
0.49
8.93
8 . 9
0.01
0.01
0.00
-0.01
0.00
0.00
0.00
0.07
0.75
0.99
Missing
8

5
5
5
12
5
5
5
5
5
5
6
5
5
6
7
5
11
7
6
27
5
5
5
6
10
5
5
5
                                118

-------
TABLE A-2.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 2

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX Oig-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
40

43
41
43
36
43
43
43
43
43
42
42
43
43
42
41
43
38
41
40
20
43
43
43
42
38
43
43
43
Mean
280.40

7.95
3.17
0.53
11.75
175.15
1.54
0.74
24.08
0.03
0.06
15.37
49.38
18.03
2.43
0.61
0.46
7.61
7.36
0.02
0.00
0.00
-0.01
-0.00
0.00
-0.06
0.19
0.42
-0.41
Std dev Maximum
19.002

0.351
3.689
0.202
0.189
13.894
0.712
0.067
4.013
0.015
0.061
0.421
4.258
1.491
0.225
0.038
0.083
1.421
1.576
0.040
0.020
0.009
0.016
0.008
0.002
0.376
0.54'
0.119
2.251
300.00

8.95
19.80
1.37
12.00
195.20
4.47
0.87
31.75
0.09
0.38
16.05
55.76
20.75
3.01
0.67
0.66
14.20
13.80
0.18
0.05
0.03
0.01
0.01
0.01
0.09
3.46
0.59
6.13
Minimum
245.00

7.42
0.00
0.08
11.10
151.30
0.91
0.62
16.20
0.00
0.02
14.50
39.80
15.20
1.98
0.53
0.30
5.88
5.35
-0.04
-0.02
-0.01
-0.06
-0.04
-0.00
-2.31
0.01
0.00
-6.83
Median Missing
288.00

7.90
2.30
0.51
1.80
179.10
1.36
0.74
25.50
0.03
0.04
15.45
50.53
18.35
2.48
0.61
0.48
7.55
7.63
0.01
0.00
0.00
-0.01
0.00
0.00
0.00
0.06
0.45
-0.29
9

6
8
6
13
6
6
6
6
6
7
7
6
6
7
8
6
11
8
9
29
6
6
6
7
11
6
6
6
                       119

-------
TABLE A-3.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 3

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX Oig-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
46

49
43
46
42
49
46
48
46
49
49
44
46
46
44
43
49
42
46
47
48
49
49
46
43
39
46
46
46
Mean
276.22

8.01
2.77
0.47
11.79
166.86
1.68
0.32
28.20
0.07
0.08
14.64
47.65
17.87
2.75
0.60
0.45
2.98
3.05
0.02
0.06
0.00
-0.01
0.00
0.00
-0.01
0.17
0.37
-0.35
Std dev Maximum
15.463

0.269
3.496
0.192
0.370
8.986
0.655
0.067
3.159
0.015
0.053
0.372
2.437
0.922
0.234
0.030
0.085
0.630
0.702
0.042
0.287
0.009
0.016
0.006
0.002
0.056
0.496
0.117
2.078
299.00

8.82
18.55
1.17
12.80
196.40
3.79
0.41
35.45
0.12
0.42
15.20
53.50
21.00
3.45
0.69
0.64
4.07
4.05
0.20
1.99
0.03
0.01
0.02
0.01
0.15
3.26
0.59
5.98
Minimum
205.00

7.48
0.00
0.00
11.00
147.00
0.88
0.11
19.10
0.03
0.04
13.80
43.20
15.80
2.20
0.50
0.26
1.73
1.21
-0.04
-0.03
-0.01
-0.06
-0.01
-0.00
-0.30
0.01
0.00
-3.77
Median Missing
279.00

7.98
1.85
0.44
11.90
166.60
1.46
0.34
29.41
0.07
0.07
14.72
47.36
17.83
2.79
0.61
0.45
3.12
3.27
0.01
0.01
0.00
-0.01
0.00
0.00
0.00
0.05
0.39
-0.95
3

0
6
3
7
0
3
1
3
0
0
5
3
3
5
6
0
7
3
2
1
0
0
3
6
0
3
3
3
                        120

-------
TABLE A-4.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 4

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX Oig-L-')
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
45

48
41
45
39
48
45
48
45
48
48
43
45
45
43
42
48
41
1
47
46
48
48
45
42
38
45
45
45
Mean
225.44

7.81
3.07
0.62
12.13
132.32
1.66
-0.02
22.34
0.11
0.13
15.78
38.42
12.27
3.62
0.72
0.13
0.62
0.42
0.44
0.50
0.15
0.06
0.00
0.00
0.01
0.13
0.56
-0.70
Std dev Maximum
11 .587

0.364
3.948
0.398
0.336
5.379
0.715
0.026
4.652
0.023
0.048
0.334
2.373
0.613
0.317
0.045
0.071
0.264
—
0.096
0.076
0.026
0.039
0.004
0.001
0.084
0.323
0.394
2.613
248.00

8.80
22.38
2.89
12.50
146.70
4.94
0.04
31.85
0.20
0.40
16.45
45.50
14.20
4.32
0.83
0.34
1 .81
0.42
0.53
0.73
0.18
0.20
0.01
0.00
0.36
2.03
2.86
5.85
Minimum
207.00

6.93
0.00
0.00
11.10
121 .10
0.75
-0.07
15.10
0.06
0.09
15.05
34.70
11 .43
3.00
0.59
0.08
0.05
0.42
0.05
0.33
-0.00
-0.05
-0.01
-0.00
-0.31
0.01
0.00
-5.18
Median Missing
220.00

7.83
2.10
0.58
12.30
132.40
1.48
-0.01
20.80
0.11
0.12
15.80
38.00
12.10
3.60
0.72
0.10
0.55
0.42
0.46
0.50
0.16
0.06
0.00
0.00
0.00
0.05
0.52
-0.96
4

1
8
4
10
1
4
1
4
1
1
6
4
4
6
7
1
8
48
2
3
1
1
4
7
11
4
4
4
                        121

-------
TABLE A-5.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 5

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX (ng-L'1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
41

44
43
44
37
44
44
44
44
44
44
43
44
44
43
42
44
38
0
43
43
44
44
44
43
39
44
44
43
Mean
374.95

6.48
6.26
3-08
12.49
65.45
66.60
-0.02
76.76
0.06
0.10
13.31
38.50
14.72
33.99
2.88
0.23
0.36
—
1 .02
1.04
0.09
0.26
0.04
0.00
-0.00
0.13
3.03
-0.47
Std dev Maximum
96.759

0.352
4.593
0.745
2.690
5.662
38.943
0.023
13.132
0.029
0.047
1 .090
8.592
3.V34
11.986
0.826
0.033
0.134
--
0.226
0.176
0.017
0.151
0.011
0.002
0.036
0.347
0.740
3.199
578.00

7.45
22.90
5.75
17.20
78.50
140.20
0.01
127.00
0.16
0.35
15.60
56.65
20.35
54.40
4.04
0.32
0.95
--
2.00
1 .42
0.14
0.62
0.06
0.01
0.1 1
2.19
5.53
8.17
Minimum
252.00

5.42
0.10
0.52
8.70
51 .60
19.37
-0.07
49.20
0.02
0.05
11 .80
27.20
10.80
16.40
1.38
0.10
0.08
--
0.72
0.70
0.07
-0.05
0.01
-0.00
-0.15
0.01
0.51
-8.08
Median Missing
400.00

6.45
4.60
3.06
11 .60
66.20
68.81
-0.01
77.87
0.05
0.09
13.20
35.21
13.50
30.60
2.79
0.22
0.34
--
1 .01
1 .02
0.09
0.28
0.03
0.00
0.00
0.04
3.02
-1 .04
8

5
6
5
12
5
5
5
5
5
5
6
5
5
6
7
5
11
49
6
6
5
5
5
6
10
5
5
6
                       122

-------
TABLE A-6.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 6

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX (ng-L4)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
43

46
43
46
38
46
46
46
46
46
46
45
46
46
44
43
46
39
0
45
46
46
46
46
43
39
46
46
45
Mean
647.88

6.19
9.19
2.32
12.57
73.83
149.50
-0.02
109.36
0.07
0.16
15.11
60.84
21 .44
66.98
1 .94
0.23
0.36
—
3-39
3.47
0.31
0.40
0.06
0.01
-0.09
0.16
2.23
-1.18
Std dev
114.828

0.389
4.925
0.487
1 .898
1 1 .491
40.893
0.024
24.647
0.020
0.074
0.708
12.017
4 . 0'05
30.788
0.671
0.024
0.145
—
0.748
0.913
0.108
0.190
0.009
0.005
0.660
0.390
0.485
6.327
Maximum
850.00

7.26
22.25
3.37
15.80
100.00
231 .80
0.01
161 .95
0.16
0.59
16.60
92.80
30.30
111 .50
4.47
0.29
0.94
--
5.13
5.20
0.50
0.80
0.08
0.01
0.21
2.43
3.21
4.13
Minimum
475.00

5.02
1 .20
0.34
10.00
56.40
91 .80
-0.07
59.20
0.04
0.08
13.25
41.45
14.23
22.50
0.77
0.14
0.09
--
1 .15
0.00
0.10
-0.05
0.04
-0.01
-4.10
0.01
0.00
-40.58
Median
661 .00

6.15
8.50
2.35
11.90
71 .45
154.60
-0.01
104.45
0.07
0.14
15.20
57.85
21 .25
74.50
2.14
0.24
0.36
--
3.27
3-35
0.29
0.47
0.06
0.01
0.01
0.06
2.25
-0.72
Missing
6

3
6
3
1 1
3
3
3
3
3
3
4
3
3
5
6
3
10
49
4
3
3
3
3
6
10
3
3
4
                        123

-------
TABLE A-7.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 8

Parameter
Conductivity
(\iS • cm4)
pH (pH units)
TOX Oig-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
N i t r i t e - N
Methane
voc
NVOC
Ion Balance (%)
Number
42

45
43
44
37
45
45
45
36
45
45
44
45
45
44
43
45
36
0
43
45
45
45
45
42
39
44
44
45
Mean
1605.21

6.88
10.90
6.78
15.61
692.44
141.35
1.85
35.48
14.84
14.91
18.99
44.13
17.49
117.49
22.80
0.10
0.36
" "
2.23
2.27
0.63
173.60
0.14
0.01
1.33
0.39
6.42
1.53
Std dev
171.503

0.208
7.645
1.156
1.612
80.972
9.997
2.062
5.534
7.760
7.608
4.913
7.060
2.155
14.442
2.442
0.027
0.178

0.861
0.912
0.105
49.179
0.072
0.003
0.709
0.796
0.900
6.086
Maximum
2000.00

7.53
49.30
11.85
18.30
926.80
163.60
8.76
46.60
40.63
40.72
27.35
75.75
25.20
138.00
27.10
0.14
0.70
- -
4.69
4.99
0.90
333.50
0.33
0.02
3.34
4.21
9.14
21.14
Minimum
1300.00

6.54
0.85
4.65
11.20
534.50
125.32
0.00
20.97
6.93
7.80
6.90
34.05
14.20
87.25
18.05
0.03
-0.51
- -
0.71
0.92
0.48
120.00
0.05
0.00
0.02
0.03
4.52
-19.88
Median
1600.00

6.85
9.60
6.63
15.20
694.20
142.67
1.26
36.78
11.80
12.20
20.48
43.50
16.90
119.50
23.20
0.10
0.34
- -
1.97
1.94
0.63
166.00
0.12
0.01
1.28
0.14
6.44
1.29
Missing
7

4
6
5
12
4
4
4
13
4
4
5
4
4
5
6
4
13
49
6
4
4
4
4
7
10
5
5
4
                        124

-------
TABLE A-8.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 9

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX (ng-L4)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
40

43
43
43
35
43
43
42
34
43
43
41
43
42
43
42
43
35
0
41
43
43
43
43
42
39
43
43
42
Mean
1706.37

6.86
11.09
8.12
15.83
786.64
146.01
-0.01
22.14
27.14
27.03
21.47
46.59
18.91
126.48
22.69
0.06
0.31
- -
2.22
2.40
0.89
185.09
0.75
0.01
3.82
0.32
7.81
0.74
Std dev
167.305

0.223
6.191
1.132
1.353
84.607
10.223
0.035
5.404
6.531
6.163
1.303
7.506
2.943
15.516
2.204
0.037
0.196
- -
0.380
0.478
0.142
50.105
0.170
0.003
1.957
0.570
0.953
4.937
Maximum
2125.00

7.45
39.80
11.22
17.70
1006.90
170.50
0.14
36.34
43.17
39.85
23.30
62.20
24.75
164.00
29.00
0.13
0.68
" "
2.92
3.66
1.12
336.00
1.16
0.01
9.51
3.54
9.94
16.67
Minimum Median Missing
1400.00

6.48
1.05
5.71
10.50
616.20
128.50
-0.07
11.20
15.25
16.00
17.30
27.90
13.60
96.25
18.93
-0.02
-0.65
- -
1.41
1.53
0.66
121.88
0.44
0.00
1.29
0.05
5.58
-8.03
1700.00

6.87
10.50
7.90
15.50
782.70
143.70
0.00
20.90
28.40
27.50
21.80
46.40
18.27
126.00
22.55
0.07
0.30
" "
2.31
2.40
0.86
176.00
0.71
0.01
3.52
0.17
7.69
0.95
9

6
6
6
14
6
6
7
15
6
6
8
6
7
6
7
6
14
49
8
6
6
6
6
7
10
6
6
7
                        125

-------
TABLE A-9.   SUMMARY OF ANALYTICAL RESULTS FOR WELL 10

Parameter
Conductivity
(|iS • cm" )
pH (pH units)
TOX (^g-I/1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-PO4
Total PO4
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide'
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
40

43
43
42
35
43
43
43
34
43
43
41
43
43
43
42
43
35
0
41
4.3
43
43
43
42
39
42
42
43
Mean
1925.87

6.85
13.07
9.19
15.76
930.17
161.56
-0.00
15.53
37.52
37.44
21.82
55.72
21.19
143.87
26.67
0.05
0.27
• '
1.66
1.80
0.56
213.29
1.08
0.01
4.62
0.40
8.65
0.07
Std dev
127.198

0.171
6.917
0.980
0.960
69.568
9.072
0.055
9.560
5.788
5.939
1.731
5.864
2.021
14.041
2.047
0.040
0.192
• '
0.266
0.350
0.119
46.908
0.163
0.016
3.492
0.737
1.516
4.954
Maximum
2200.00

7.26
36.50
12.30
16.80
1103.00
179.80
0.26
50.07
51.55
49.80
24.60
68.10
26.00
172.00
30.80
0.11
0.46
• '
2.52
3.05
0.94
386.00
1.44
0.10
18.10
4.22
10.59
18.79
Minimum
1650.00

6.50
5.30
7.92
11.30
785.00
137.20
-0.07
6.02
28.30
27.10
14.70
41.60
17.22
118.00
22.60
-0.03
-0.78
• '
1.33
1.33
0.33
151.00
0.64
-0.01
0.65
0.05
0.57
-9.02
Median
1972.50

6.85
11.20
9.00
15.90
941.60
163.99
-0.01
12.90
37.80
36.90
22.25
55.30
21.10
146.50
26.98
0.05
0.29
• "
1.59
1.66
0.55
206.20
1.09
0.01
3.50
0.17
8.71
-0.28
Missing
9

6
6
7
14
6
6
6
15
6
6
8
6
6
6
7
6
14
49
8
6
6
6
6
7
10
7
7
6
                          126

-------
TABLE A-10.
SUMMARY OF ANALYTICAL RESULTS FOR WELL 11

Parameter
Conductivity
(\iS • cm4)
pH (pH units)
TOX Oig-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
40

43
43
43
35
43
43
42
34
43
43
41
43
43
43
42
43
35
0
41
41
43
43
43
42
39
43
43
43
Mean
1614.87

6.84
11.34
7.21
15.80
741.19
138.62
0.39
28.29
19.20
19.32
19.85
53.09
21.91
119.25
21.27
0.05
0.27
- -
3.64
3.79
1.06
160.02
0.47
0.01
2.54
0.34
6.92
0.40
Std dev
169.280

0.196
6.080
1.292
1.236
86.080
16.702
0.273
6.630
7.213
7.263
2.285
11.672
5.109
13.378
2.569
0.034
0.185

0.547
0.478
0.191
47.652
0.121
0.008
1.765
0.573
1.130
4.746
Maximum
2140.00

7.33
31.00
11.49
17.70
970.90
179.90
1.03
40.42
33.83
32.96
22.30
80.90
36.50
152.00
27.30
0.12
0.48
- -
4.97
4.73
1.54
340.50
0.68
0.05
7.13
3.13
10.17
12.47
Minimum
1400.00

6.47
1.20
5.42
12.00
645.00
119.50
-0.00
15.54
7.08
6.64
10.65
30.80
14.00
96.50
18.00
-0.01
-0.72
- -
2.65
2.99
0.81
95.90
0.17
0.00
0.03
0.04
5.23
-11.18
Median Missing
1582.

6.84
11.15
6.96
15.40
715.80
134.20
0.31
28.42
18.40
18.20
20.65
54.10
21.80
118.00
20.25
0.06
0.29
- -
3.60
3.80
1.00
150.88
0.49
0.01
2.50
0.16
6.73
0.34
50 9

6
6
6
14
6
6
7
15
6
6
8
6
6
6
7
6
14
49
8
8
6
6
6
7
10
6
6
6
127

-------
TABLE A-ll.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 12

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX Oig-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
39

42
42
42
34
42
42
41
32
41
42
40
42
42
42
41
42
34
0
40
40
42
42
42
42
39
42
42
42
Mean
1836.03

6.87
12.77
8.57
.15.91
866.05
155.76
0.27
19.07
34.18
33.81
22.07
47.31
18.99
132.53
24.84
0.04
0.26
—
1 .08
1.24
0.80
210.60
1 .21
0.01
4.22
0.36
8.25
0.30
Std dev
129.940

0.169
5.677
1 .749
1 .160
71.367
8.965
0.233
7.101
6.960
7.142
1 .422
6.061
2.816
12.063
2.221
0.041
0.213
--
0.183
0.480
0.100
47.789
0.281
0.014
2.276
0.691
1 .628
5.218
Maximum
2100.00

7.30
28.40
11 .80
17.50
1032.90
173.13
0.94
51 .14
54.71
51 .35
23.60
60.30
25.60
163-50
30.60
0.11
0.51
—
1 .50
3.78
1 .05
373-50
1 .76
0.09
12.10
4.27
11 .62
19.67
Minimum
1600.00

6.56
4.10
0.00
12.70
726.10
138.40
0.00
8.45
20.60
19.15
16.10
37.21
14.15
1 11 .50
21 .58
-0.04
-0.86
--
0.82
0.83
0.64
131 .50
0.40
0.00
0.51
0.05
0.23
-9.50
Median
1820.00

6.87
12.20
8.83
15.55
860.70
156.90
0.20
18.68
33.73
33.60
22.55
46.20
18.49
131.50
24.35
0.05
0.29
—
1.06
1 .10
0.77
202.62
1.23
0.01
3-55
0.16
8.39
-0.56
Missing
10

7
7
7
15
7
7
8
17
8
7
9
7
7
7
8
7
15
49
9
9
7
7
7
7
10
7
7
7
                        128

-------
TABLE A-12.  SUMMARY OF ANALYTICAL RESULTS FOR WELL 13

Parameter
Conductivity
(\iS ' cm4)
pH (pH units)
TOX (ng-L-1)
TOG
Temperature (°C)
Alkalinity
Chloride
(N03 + N02)-N
Sulfate
Ortho-P04
Total P04
Silica
Calcium
Magnesium
Sodium
Potassium
Eh (volts)
Electrode DO
Winkler DO
Iron
Iron(II)
Manganese
Ammonia
Sulfide
Nitrite-N
Methane
voc
NVOC
Ion Balance (%)
Number
40

43
43
43
34
43
43
43
34
43
43
42
43
43
43
42
43
35
0
41
41
43
43
43
41
39
43
43
43
Mean
1546.90

6.81
10.54
6.87
16.61
690.37
133.49
0.03
30.47
17.75
17.90
19.63
61.80
25.22
114.34
19.18
0.06
0.43
" "
4.09
4.16
0.57
134.86
0.43
0.01
1.31
0.29
6.63
0.86
Std dev
140.846

0.159
7.059
0.904
1.121
75.376
12.141
0.070
9.767
5.500
5.472
1.172
12.778
5.557
10.788
2.486
0.032
0.259
" "
0.778
0.801
0.119
42.313
0.116
0.006
0.947
0.486
0.930
4.855
Maximum
2050.00

7.15
33.00
10.19
18.00
881.30
161.65
0.234
46.91
36.10
34.88
22.40
101.50
43.00
141.00
26.10
0.11
1.34
" "
6.68
6.96
1.00
272.00
0.64
0.02
3.83
2.98
10.38
12.27
Minimum
1350.00

6.35
0.00
5.77
10.90
555.10
116.72
-0.07
9.37
11.00
10.50
17.00
41.90
18.70
94.20
15.50
-0.01
-0.53
- -
2.88
2.27
0.44
83.80
0.04
-0.00
0.04
0.04
4.92
-11.58
Median
1510.00

6.81
8.95
6.68
16.80
680.40
130.90
0.00
30.86
17.50
16.70
19.95
59.90
24.20
114.00
18.80
0.07
0.41
" "
3 . 9
4.06
0.54
124.00
0.44
0.01
1.06
0.14
6.50
1.36
Missing
9

6
6
6
15
6
6
6
15
6
6
7
6
6
6
7
6
14
49
1 8
8
6
6
6
8
10
6
6
6
                         129

-------
                       APPENDIX B

    TIME SERIES OF INDIVIDUAL CONSTITUENT CONCENTRATIONS
       FOR BIWEEKLY SAMPLING RUNS FOR EACH WELL AT THE
          SAND RIDGE SITE (WELLS #1, #2, #3 AND #4)
AND THE BEARDSTOWN SITE (WELLS #8, #9, #10, #11, #12, AND #13)
    15
    14
  D
  6
  g13
  r
  e
  e
  8

  r 12
    11
    10
                 Well Temperature
10
15
20
25
30
                              35   40
                                                  45
                           Run
WELL &e-a     1
     •-~     4
                                  2
                              BLANK
                           130

-------
              Well Temperature
  19


  181


  17


  16

D
e 15
9
r 14
e
e
s 13


C12


  11


  10


  9


  8
        WELL
10    15    20    25    30

            Run

        j-  ^ ^ ^    £

O 0 O     v7  4~~H r    IV/  X"
>i< >H *    12  a « tt    13  e-
                                      35    40   45
    8
   11
BLANK
                         131

-------
  15
               Cell  Temperature
  14
D
e
g
r
e
e
s
13
  12-
  11-
        WELL
             i |—i  i—i i | i i i i [—i i i—i | i i i i p-1 i i i |     |

             10    15   20   25   30   35   40   45
                         Run
                   1
                   4
    2
BLANK
                        132

-------
                Cell  Temperature
  24

  23

  22

  21

  20

D 19
e
g 18
r
  17 ^
  14-

  13

  12-1

  11

  10
              10
15
20
25
30
35
40
45
        WELL
              ~d( !itc i4c
       Run


   9  i i  i
  12  n n n
         6
        10
        13
              8
              11
          BLANK
                         133

-------
  10
                           PH
u
n
   8
5    10
         WELL
                    15
  20
25
30
35
40
45
                            Run
1
4
    2  A A A

BLANK
                           134

-------
                  PH
      10    15    20    25    30

                   Run
WELL
* * *
 9
12
                          10
                          13
                    35    40    45
    8
    11
BLANK
                  135

-------
                                Eh
   0.7
  0.6-
  0.5
o
I
t
s
  0.4
  0.3
  0.2
  0.1-
  0.0
       -ii—iI—|  i I r—i—| 1

     0     5     10
          WELL
1 i  ' ' ' '  i p '  ' ' i '  ' ' '  i ' ' '  r i ' '  ' • i  r ' ' '  i
15    20    25    30    35    40   45
                                 Run
    1
    4
     2
BLANK
                                136

-------
                                 Eh
V
0
1
I
s
 0.34
 0.32
 0.30
 0.28
 0.26
 0.24
 0.22
 0.20
 0.18
 0.16
 0.14
 0.12
 0.10
 0.08
 0.06
 0.04
 0.02
 0.00
-0.02
-0.04
         0
                  10
15
20    25     30    35     40    45
                                     Run
           WELL
                          ^

                              9.  ,  .
                              I  I  I
                          12   nnn
                6
               10
               13
                        8
                       11
                   BLANK
                                  137

-------
  380
  370
  360
  350
  340
M330
  320
  300
 i
 c
 r 310
 o
 s
 ' 290
m 280
 e
 n
 s 260
  270
  250
  240
  230
  220
  210:
  200-
                    Conductivity
                                 a a DD
         WELL
                10    15   20    25   30
                             Run

               ^^     4  ^_+-+ BLANK
                                          35
-i—| r
 40
45
                           138

-------
  2000
  1900
  1800
  1700
  1600
M 1500
c 140°
r 1300
s 120°
'  1100
m 1000
e  900
   800
   700
   600
   500
   400
   300
   200
n
s
      0
            T
            5
         WELL
                    Conductivity
10
15
20
25
30
35
                          40
45
                              Run
 C
9
12
                   C
                   -in
                   Iw
                   13
                                                Q
                                                1-1
                                                I *
                                            BLANK
                            139

-------
        Total Organic Halogen
18
17
16
15
m 14
c 13
r 12
0
g 11-
>
m 9.
s
/ 8
L 7-
t 6
e
r 5-
4












1
ill
11
2

1

0
5   10
     WELL
              ' I ' ' ' ' I ' ' ' ' I ^ ^r T -[ T i

              15   20  25   30   35   40
                       45
                    Run
1
4
                    2 A A A
                 BLANK
                   140

-------
  4CH
           Total Organic  Halogen
m
i  30
c
r
o
9
r
a
m 20
s
/
L
i
t
e
r 10-
   0
0    5    10    15    20   25   30

                     Run

    WELL  B-B-e
5
9
12
                              6
                              10
                              13
                                     35   40   45
   8
   11
BLANK
                       141

-------
m
  1.6

  1.5-

  1.4

  1.3

  1.2

  1.1

  1.0
y
/ 0.9;

'  0.8
t
e 0.7-
r
  0.6

  0.5-

  0.4-

  0.3

  0.2

  0.1 -

  0.0
             Total Organic Carbon
               10
                    15
20
25
30
35
40
45
                            Run
         WELL
                       1
                       4
         2  A A A
     BLANK
                          142

-------
  13

  12

  11-

  10

  9
m
g 8
/
L 7
e
r
  5

  4

  3

  2

  1

  0
i *
0
           Total  Organic Carbon
 5    10    15   20   25    30

                Run

WELL &e-a     5  <^M>    6

            12  *MM*    13
35
                                         40
                                        8
                                        11
                                        I I

                                    BLANK
45
                       143

-------
      Non-Volatile Organic Carbon
m
  1.2

  1.1

  1.0

  0.9

  0.8
'  0.6


  0.51


  0.4


  0.3


  0.2


  0.1-


  0.0
         5    10   15   20    25   30   35   40   45
                        Run
       WELL
1
4
   2
BLANK
                       144

-------
      Non- Volatile Organic Carbon
  11


  10-

  9


  8

m 7
g
/
L 6
i
'  5
e
r
  4

  3


  2

  1
   i ^
   0
 5    10   15   20   25   30

               Rur

WELL  B-B-B    5  o o o    6  &•
            y  i I r    \\J  ^x
           12  a a «    13  e
35   40   45
                                     8
                                     11
                                  BLANK
                      145

-------
  1.0
  0.9
  0.8
           Volatile  Organic Carbon
  0.7

m
9 0.6

L
'  0.5
t
e
r  0.4
  0.3


  0.2


  0.1


  0.0
p i i i i i | i i i i | i , i i | i i i i | i
10   15   20   25   30
                                       35    40    45
                          Run
        WELL
              Q O O
                  2
               BLANK
                         146

-------
         Volatile Organic Carbon
m
L
I
t
e
r  2
  1
  o
   0
10
       WELL
15   20   25    30   35    40   45
                       Run
       5
       9
       12
            6
            10
            13
   8
   11
BLANK

-------
                     Sodium
m
g
i
t
e
r
   3
2
   1
   0
  -1
     f r T-I T ]- r T r T~[—r

     0     5    10
        WELL
                15   20   25   30   35    40    45
                         Run
                  1
                  4
    2
BLANK
                        148

-------
   170
   160
   150
   140^
   130
   120
   110
m
g  100
L  90
'   80
   70 -I
   60
   50 \
   40
   30
   20
   10
    0
   -10
t
e
r
                      Sodium
                                                   ~0
                10
                     15
20
25
30
35
40
45
                            Run
         WELL
5 ooo
*-k
12 nnn
6 A-
10 *•
13 e-
*-& 8
** 11
^^ BLANK
                           149

-------
                      Potassium
    0.9


    0.8


    0.7


    0.6

m
9   0.5

L
'    0.4
t
e
r    0.3


    0.2


    0.1


    0.0
  -0.1-
      0    5     10    15    20   25    30   35    40    45

                             Run
                        1
                        4
    2
BLANK
                           150

-------
                    Potassium
   30
   20
m
g
/
L
i
t
e
r
10
  -10
      0-&OOOOOOQ
  0    5   10
        WELL
                                   I ^' ' I
                    15    20    25   30   35   40   45
                           Run

                      9  i i i
                     12  a^M*
                             6
                            10
                            13
    8
   11
BLANK
                         151

-------
                     Calcium
m
g
/
L
i
t
e
r
   80
   70 J
   60
   50
40
30
   20
    10
   -10
           5    10    15    20    25    30    35    40    45

                            Run


               o o o     4  ^+^ BLANK
                          152

-------
                    Calcium
-10
   0    5    10
      WELL  B-B-B
15
20   25   30   35   40   45
                         Run
                    9  |  | |
                    12  nn n
             6
            10
            13
                   8
                  11
               BLANK
                       153

-------
   30
   20
m
g
/
L
i
t
e
r
10
    0
  -10-
   i i i M i
                 Magnesium
     | i T T T y T T—i—i—|—

     0     5    10
       WELL
               15    20   25   30   35   40   45
                         Run
I

                    4
                           O  A A A
                           £-  L=± L^ j^

                       BLANK
                       154

-------
               Magnesium
    OOOOOGOO
-10-
   0    5   10   15   20   25   30   35  40   45
                      Run
      « VCLL
5
9
   8
   11
BLANK
                    155

-------
    0.6
                            Iron
    0.5
    0.4

m
g
/   0.3-
L
i       :
t
e   0.2
r
    0.1-
    0.0-
   -O.U
       0     5    10
          WELL
15    20    25    30    35    40    45


        Run


2                  A  A A     T
                  i_l f_in     W
1
4
                                  BLANK
                              156

-------
                           Iron
m
L
'    3
t
e
r
  -1
                                                    e—©
     i i  i i i | , i , ,  | i ,

     0     5     10
         WELL
                                       I ^' "I
15    20    25    30    35    40    45
                             Run
9
                       12
              6
              10
              13
    8
    11
BLANK
                           157

-------
                Ferrous Iron
   1
m
g
/
L
i
t
e
r
  -1
I ' '
0
, 1 1 1 ,
5
10
, , i ,
15
20
i
25
30
i i | i .
35
i i i i
40
1 ' i
45
                       Run
       WELL
1
4
   2  A A A
BLANK
                      158

-------
                   Ferrous  Iron
m
g
/
L
i
t
e
r
3
   2
                                               e-€>
  -1
     0
           10   15   20   25   30   35    40    45
        WELL
                           Run
              >K >K *
9
                  12
 6
10
13
    8
   11
BLANK
                         159

-------
                      Manganese
m
g
/
L
i
t
e
r
 0.19
 0.18
 0.17
 0.16
 0.15
 0.14
 0.13
 0.12
 0.11
 0.10
 0.09
 0.08
 0.07
 0.06
 0.05
 0.04
 0.03
 0.02 \
 0.01
 0.00
-0.01
-0.02
             ~r
              5
          WELL
               10
15   20    25
        Run
  1  A A A      P
  4  ^+_+ BLANK
30
35    40
45
                             160

-------
                     Manganese
m
g
/
L
i
t
e
r
 1.6
 1.5;
 1.4-
 1-3:
 1.2-
 1-1:
 1.0
 0.9
 0.8-
 0.7
 0.6
 0.5
 0.4
 0.3
 0.2
 0.1
 0.0
-0.1
                 10
         WELL
                                                       e
I I . 1 I I  I     I      I     I     I
15   20    25    30    35    40

        Run
              5/J           Q
     \j (j ^     ^j  f\£3
-------
                       Ammonia
m
g
/
L
i
t
e
r
  0.20
  0.19
  0.18
  0.17
  0.16
  0.15
  0.14
  0.13
  0.12
  0.11
  0.10
  0.09
  0.08
  0.071
  0.06
  0.051
  0.04
  0.031
  0.02
  0.01-1
  0.00
-0.01
-0.02-
-0.03-
-0.04
-0.051
                  10    15    20   25    30    35   40    45
                               Run
         WELL
                       1
                       4
    2
BLANK
                            162

-------
                    Ammonia
   400
   300
m
g
/
L
i
t
e
r
200
100
     0
   -100
, . I I I I I I I I I I
0    5    10



  WELL  &e-e
                                    i ' ' ' • i
                     15   20   25   30    35   40   45
                           Run
                      5
                      9
                      12
                             6
                             10
                             13
    8
   11
BLANK
                         163

-------
   300
                        Alkalinity
m
g
/
L
j 200
t
e
r

a
s
C
a
C
O
3
100
     0
              10    15    20    25    30    35    40    45
                              Run
       WELL  EHEHB     1
                      4
                                BLANK
                            164

-------
c
a
C
O  300
3

   200

   100

     0
      0
                       Alkalinity
10
15    20    25    30    35    40    45
                               Run
          WELL
                             91 i  ,
                             i i  i
                         12  »-*-*
                    6
                   10
                   13
                          8
                          11
                      BLANK
                             165

-------
                    Chloride
   5
m
g
/  ;
L
i
t
e  2
r
   0
   -1
         WELL
               10    15   20   25   30   35   40   45

                          Run
1
4
    2
BLANK
                         166

-------
                      Chloride
    300
   200
m
g
/
L
i
t
e
r
                           I     I
      0    5    10   15   20   25   30    35    40    45
                            Run
        WELL  B-&B
 5
 9
12
                          n n n
 6
10
13
    8
    11
BLANK
                          167

-------
               Sulfate
      10    15    20    25    30   35    40    45
                   Run
WELL
1
4
    2  A A A
BL^NK
                 168

-------
  170 \
  160
  150-1
  140
  130
  120
m 110
/ 100
L  90]
t  801
e
r
   601
70
   50
   40
   30
   20
   10
    0
                           Sulfate
                                                        €>
           1 i '  ' ' ' i  '
            5    10
          WELL
                     15
20
25    30    35    40
45
                                Run
                 ^L, ^i^ ^Lf
                 •*!*• *l"> ^n
                        5
                        9
                       12
          6
         10
         13
                 8
                11
            BLANK
                              169

-------
                          Silica
   17
   16
   15
   14
   13
   12
   11
m  10
g
/   9
    8
    7-
    6-
    5
    4
    3
    2
    1
    0
   -1
   -2
L
i
t
e
r
     0
               10
         WELL
15    20   25    30    35    40   45
         Run
    1   XX ^ ^     O     .      O
    4   ^_^ BLANK
                            170

-------
    30
                            Silica
    20
m
g
/
L
i
t
e
r
10
                                                     •Q-
  -10
      0     5     10
         WELL
                *T* *n T^
                    15     20     25     30

                            Run

                      5   o o o      6  &-f
                      9   -h+n-     10  **
                     12   »-*Mt     13  e^
 i  ' ' ' • i  '
35   40
     8
    11
BU\NK
                                                            45
                              171

-------
        Total   Phosphate
       10   15   20   25   30   35   40   45
                  Run
WELL
1
4
    2
BLANK
                172

-------
                Total   Phosphate
   50 4
   40
m
g
   30
L
'   20
t
e
r

   10 1
    0
  -10
     0    5    10
        WELL
                                             | I I i i |
15   20   25   30   35   40   45
                           Run
9

                     12
             6
             10
             13
    8
   11
BLANK
                         173

-------
                 ortho -  Phosphate
m
g
/
L
i
t
e
r
  0.21
  0.20
  0.19
  0.18
  0.17
  0.16
  0.15
  0.14
  0.13
  0.12
  0.11
  0.10
  0.09
  0.08
  0.07
  0.06
  0.05
  0.04
  0.03
  0.02
  0.01
  0.00
-0.01
-0.02
-Q,Q3
        0
                 I,      I,     I.      I,     I,      I,     I      I
               10    15   20    25    30    35    40    45
                                Run
         WELL
                       1
                       4
    2   A A A
BLANK
                             174

-------
               ortho  -  Phosphate
   50
   40
m
g
/
L
'   20
t
e
r
  -10-
I ' ' ' ' r T r I r I I I

0    5   10
        WELL  B-B-B
1 i     i     i     i     i     i

15   20   25   30   35   40


       Run

  C  XN /^ ^    R   A      Q
9     i  i i     m  v y y     -i-i
     t  i i     i^^  /^\ /~\ ^^\     i i
  12  *wft    13  €HE«E> BLANK
                                                   i

                                                  45
                         175

-------
    1.31
    1.2
'   0.6
t

e   0.5
r

   0.4


   0.31


   0.2


   0,1.1


   0.0


  -0.1
            Nitrate + Nitrite Nitrogen
      0    5    10    15    20    25    30    35    40    45
                             Run
         WELL  EHEH3
1
4
    2

BLANK
                           176

-------
            Nitrate + Nitrite  Nitrogen
   9


   8


   1]
m
t
e
r
   2
   1
   0
  	H
     I ' ' ' ' I ' ' ' ' I '

     0     5    10
        WELL
15    20   25   30   35    40    45
       Run


   9   | i i
  12  «-»-*
 6
10
13
    8
    11
BLANK
                          177

-------
                    Nitrite  Nitrogen
m
    0.0071
    0.006
    0.005
    0.004
1    0.003
t
e
r
    0.002 ]
    0.001
    0.000
  -0.001
        0     5    10   15
         WELL
 20    25   30    35   40    45

    Run

yy ^ y-v     O  A A A     O

+++ BLANK
                             178

-------
                Nitrite Nitrogen
-0.01-
      WELL
              10    15   20    25    30    35    40    45
                           Run
                     9
                    12
 6  A A A     8
in  v v y    11
IV  ™7% 7>     | |
13  €M=M9 BLANK
                        179

-------
                         Sulfide
    0.03
    0.02-
    0.01
m
g
/
L
i
t
e
r
 O.OC
-0.01
  -0.02-
  -0.03
  -0.04
       0    5    10   15    20    25   30   35
        WELL  see      1
                        »
                           Run

                           «      2
                           + BLANK
                          180

-------
                            Sulfide
m
9
/
L
i
t
e
r
  1.8
  1.7
  1.6-
  1.5-
  1.4
  1.3
  1.2
  1.1
  1.0
 0.9 ]
 0.8
 0.7 ]
 0.6
 0.5
 0.4 ]
 0.3
 0.2
  0.1-
 0.0
-0.1
     0
                    10
7—I  I—7—7 7  [ 7 7 7 7  [ r 7  7 7 J I  7 7 7 ]  I 7 7 I |  7 I 1 I  |
 15    20    25    30    35   40    45
           WELL
                                Rur
                         C   £^^^^±      ^%
                         VJ   " " ™      V
                         S   H-+-+     10
                                      13
                              8
                             11
                         BLANK
                                181

-------
   7

   6
p
e
r
c
e
n
t
4

3

2

1

0
   2


   3


   4


   5


   6

   7
0
      |


      5
       WELL
              Ion Balance Error
             10
15
20
25
30
35
40
45
                        Run
                 1
                 4
                         2
                     BLANK
                      182

-------
              Ion Balance  Error
p
e
r
c
e
n
t
   30
   20
   10
  0
-10
  -20
  -30
  -40
  -50
      i • ' ^ i ' ' • ' i ' '

     0    5   10
        WELL B-H-B
                           i ' ' ' ' i
                 15   20    25   30   35   40    45
                          Run
                   5
                   9
             * * *    12  n n n
 6
10
13
    8
   11
BLANK
                        183

-------
                    Methane
m
g
/ -1
L
i
t
e _2
r
  -3
  -4
   o
0    5    10
        WELL
15    20    25    30   35   40


       Run

   I  \j^" v? v     ^*  /\ /A ZA     v
  4  ^+_+ BLANK
                                                45
                        184

-------
               Methane
m
g
t
e
r
1 1 1
0 5 10

WELL
e=^=*
i\r i\t tit
J|t J|4 J|l
1 '
15

5
9
12
i . i i i , , , i , ,
20 25 30
Run
 6 **-
-4=F=r 10 x^-
nn n 13 e-e-
< 1 | 1 . . 1 | ! 1 , 1 |
35 40 4>

A 8
* 11
e BLANK
                  185

-------
      Probe  Dissolved  Oxygen
5

4

3-1

2

1

01
         10
15
20   25

  Run
30
35
40
45
     WELL
          O (^™^^
           2
        BLANK
                   186

-------
        Probe  Dissolved Oxygen
m
g
/
L
i
t
e
r
 -1-
   0    5    10
      WELL
15   20  25   30   35   40   45

      Run
5

  9

 12
                           10
                           13
   8
   11
BLANK
                     187

-------
  14


  13


  12:


  11


  10


m 9

9
/ 8
L

'  7
t

e 6
r

  5\
  3


  2


  1


  0
        Winkler Dissolved Oxygen
        \
        5
10
15
20
25
30
35
40
45
                      Run
       WELL  &€H3
       1
       4
            2
        BLANK
                     188

-------
                               APPENDIX C

    GROUND WATER ELEVATIONS MEASURED DURING EACH BIWEEKLY SAMPLING RUN
AT THE SAND RIDGE AND BEARDSTOWN SITES.   (ELEVATIONS AT THE SAND RIDGE SITE
      ARE IN FEET RELATIVE TO AN ARBITRARY 1000 FOOT REFERENCE POINT.
     ELEVATIONS AT BEARDSTOWN ARE IN FEET RELATIVE TO MEAN SEA LEVEL.)
        TABLE C-l. GROUND-WATER ELEVATIONS IN WELLS AND PIEZOMETERS AT SAND RIDGE

SAMPLING
DATE

03/10/86
03/24/86
04/07/86
04/21/86
05/05/86
05/19/86
06/02/86
06/16/86
06/30/86
07/14/86
07/28/86
08/11/86
08/23/86
09/08/86
09/22/86
10/06/86
10/20/86
11/04/86
11/17/86
12/01/86
12/15/86
12/29/86
01/12/87
01/26/87
02/09/87
02/23/87
03/09/87
03/23/87
04/06/87
04/20/17
OS/04/87
05/11/17
06/01/87
06/15/87
06/29/87
07/13/87
07/27/87
01/10/87
01/24/17
RUN
NUMBER
l
2
3
4
5
6
7
8
9
10
11
12
13
14
IS
16
17
IS
19
1 20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
33
36
37
38
39
D035
(#1)
962.03
962.28
962.16
962.05
962.01
961.97
961.89
961.88
961.86
961.81
961.76
961.72
961.70
961.64
961.57
961.46
963.43
981.50
961.58
961.65
961.68
961.75
961.87
961.93
961.91
961.96
961.98
961.99
961.86
961.79
961.75
961.80
961.60
961.75
961.69
961.56
961.11
961.27
961.17
WELL NAME AND NUMBER
D050
(#2)
962.41
962.26
962.18
962.06
962.01
961.99
961.92
961.89
961.61
961.62
961.75
961.74
961.71
961.65
961.58
963.69
961.43
961.51
961.59
961.66
961.69
961.76
961.87
961.93
961.91
961.98
961.99
962.01
961.66
961.81
961.76
961.82
961.60
961.77
961.69
961.59
961.16
961.36
961.24
D065
(#3)
962.63
962.29
962.17
962.06
962.02
961.98
961.92
961.89
961.84
961.62
961.75
961.74
961.71
961.65
961.59
961.18
961.43
961.51
961.59
961.66
961.70
961.79
961.87
961.93
961.94
961.97
961.99
962.00
961.86
961.62
961.76
961.60
961.60
961.75
961.69
961.56
961.47
961.37
961.24
D0105
(#6)
962.66
962.32
962.22
962.11
962.05
962.01
961.95
961.93
961.87
961.64
963.77
961.76
961.74
961.68
961.62
961.56
961.46
961.55
961.62
961.66
961.73
961.80
961.89
961.99
961.96

962.03
962.02
961.90
961.65
961.79
961.83
961.64
961.60
961.73
961.60
961.49
961.39
961.26
SRI

961.80
961.64
961.53
961.41
961.37
961.34
961.20
961.21
961.20
961.16
961.11
961.08
961.04
960.99
960.91
960.62
960.75
960.65
960.95
961.03
961.07
961.14
961.24
961.30
961.29
961.34
961.37
961.39
961.20
961.21
961.12
961.14
961.15
961.11
961.05
960.93
960.81
960.71
960.59
SR2

961.95
961.60
961.70
961.57
961.53
961.49
961.38
961.41
961.36
961.31
961.26
961.26
961.20
961.16
961.08
960.99
960.93
961.01
961.10
961.17
961.21
961.28
961.38
961.15
961.45
961.09
961.52
961.53
961.39
961.36
961.26
961.30
961.32
961.26
961.22
961.09
960.96
960.68
960.75
SR3

962.39
962.23
962.13
962.03
961.97
961.93
961.66
961.84
961.80
961.77
961.73
961.70
961.67
961.61
961.54
961.44
961.39
961.66
961.56
961.61
961.66
961.72
961.62
961.89
961.90
961.92
961.91
961.96
961.62
961.76
961.72
961.75
961.76
961.72
961.66
961.54
961.61
961.32
961.20
                                   189

-------
TABLE C-2. GROUND-WATER ELEVATIONS IN BEARDSTOWN SAMPLING WELLS

SAMPLING
RUN
DATE NUMBER
03/11/86
03/25/86
04/08/86
04/22/86
05/06/86
05/20/86
06/03/86
06/17/86
07/01/86
07/15/86
07/29/86
08/12/86
08/26/86
09/09/86
09/23/86
10/07/86
10/21/86
11/05/86
11/18/86
12/02/86
12/16/86
12/30/86
01/13/87
01/27/87
02/10/87
02/24/87
03/10/87
03/24/87
04/07/87
04/21/87
05/05/87
05/19/87
06/02/87
06/16/87
06/30/87
07/14/87
07/28/87
08/11/87
08/25/87
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
BT18
(#5)
444.
444.
443.
443
443.

443.
443.
443.
444.
445.
444
443.
443.
443.
445.
445.
445.
445.
445.
444.
444.
444.
444.
444.
444.
443.
444.
444.
444.
444.
444.
443.
443.
443.
442.
442.
442.
441.
.45
.16
.93
.84
.66

.55
.94
51
.71
.16
.42
74
.29
.68
.45
.24
74
08
34
.86
69
41
36
72
27
97
07
.19
.43
.27
14
.95
70
.29
.94
.41
03
76
SAMPLING WELL NAME AND NUMBER
BT23 BT25 BT30 BT35 BS30 BP30
(#6)
444.41
444.12
443.90
443.80
443.62

443.49
443.87
443.46
444.66
445.10
444.38
443.69
443.25
443.61
445.41
445.20
445.67
445.03
445.28
444.81
444.65
444.28
444.31
444.67
444.20
443.94
444.04
444.17
444.40
444.23
444.06
443.90
443.66
443.24
442.90
442.37
442.00
441.73
(#8)
443.39
443.26
443.06
442.97
442.80
442.59
442.54
442.84
442.59
443.53
443.80
443.47
442.94
442.48
442.65
444.37
444.27
444.65
444.18
444.28
444.03
443.91
443.54
443.57
443.78
443.40
443.16
443.19
443.37
443.57
443.40
443.23
443.04
442.81
442.41
442.05
441.48
441.05
440.78
(#9)
443.38
443.23
443.04
442.95
442.78
442.56
442.58
442.82
442.58
443.50
443.78
443.45
442.91
442.45
442.62
444.34
444.26
444.63
444.16
444.24
444.01
443.88
443.62
443.54
443.75
443.39
443.12
443.16
443.34
443.54
443.37
443.20
443.02
442.78
442.39
442.04
441.46
441.03
440.76
(#10)
443.35
443.20
443.00
442.91
442.75
442.54
442.54
442.79
442.57
443.47
443.77
443.41
442.88
442.44
442.61
444.30
444.23
444.58
444.11
444.23
443.98
443.87
443.60
443.53
443.73
443.36
443.10
443.15
443.32
443.52
443.34
443.17
442.98
442.76
442.36
442.01
441.44
441.00
440.72
' (#H)
443.
443.
443.
442.
442.
442.
442.
442.
442.
443.
443
443.
442.
442.
442.
444.
444.
444.
444.
444.
444.
443.
443.
443.
443.
443.
443.
443.
443.
443.
443.
443.
443.
442.
442.
442.
441.
441.
440.
37
24
05
95
78
58
60
82
57
50
.79
46
91
46
64
35
26
62
16
26
02
89
63
56
76
40
12
18
35
54
38
21
01
80
39
03
46
03
75
(#12)
443.39
443.26
443.06
442.97
442.79
442.59
442.59
442.83
442.59
443.50
443.80
443.48
442.93
442.48
442.66
444.36
444.27
444.64
444.17
444.28
444.04
443.91
443.64
443.58
443.77
443.41
443.14
443.19
443.37
443.56
443.39
443.22
443.03
442.81
442.41
442.04
441.48
441.05
440.78
BT33
(#13)
443.50
443.34
443.15
443.05
442.90
442.69
442.71
442.93
442.70
443.62
443.92
443.59
443.00
442.55
442.74
444.45
444.37
444.72
444.25
444.37
444.11
443.97
443.71
443.66
443.86
443.49
443.23
443.27
443.45
443.65
443.49
443.31
443.14
442.90
442.56
442.14
441.57
441.18
440.89
                              190

-------
TABLE C-3. GROUNO-UATER ELEVA   fiS  IN BEARDSIOUN  PIEZOMETERS

SAMPLING
RUN
DATE NUMBER
03/11/86
03/25/86
04/08/86
04/22/86
05/06/86
05/20/86
06/03/86
06/17/86
07/01/86
07/15/86
07/29/86
08/12/86
08/26/86
09/09/86
09/23/86
10/07/86
10/21/86
11/05/86
11/18/86
12/02/86
12/16/86
12/30/86
01/13/87
01/27/87
02/10/87
02/24/87
03/10/87
03/24/87
04/07/87
04/21/87
05/05/87
05/19/87
06/02/87
06/16/87
06/30/87
07/14/87
07/28/87
08/11/87
08/25/87
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
PIEZOMETER NAME
B1 B2.1 B2.2 B3.1 B3.2 B4 B5 B6.1 B6.2 B7
444.57 444.72 444.72 444.24 444.20 443.23 442.96 444.33 444.33 443.87
444.23 444.28 444.29 443.96 443.93 443.09 442.84 444.09 444.09 443.70
443.99 444.05 444.06 443.74 443.71 442.88 442.65 443.86 443.85 443.47
443.92 443.95 443.95 443.64 443.62 442.79 442.52 443.81 443.81 443.44
443.75 443.82 443.82 443.49 443.45 442.62 442.37 443.62 443.62 443.23
443.54 443.62 443.63 443.31 443.29 442.45 442.21 443.41 443.41 443.01
443.65 443.73 443.73 443.38 443.34 442.45 442.20 443.48 443.50 443.07
444.02 444.14 444.15 443.71 443.67 442.65 442.41 443.82 443.82 443.33
443.57 443.65 443.68 443.32 443.30 442.42 442.17 443.45 443.45 443.05
444.80 444.95 444.96 444.45 444.41 443.35 443.08 444.56 444.56 444.04
445.29 445.53 445.53 444.85 444.81 443.59 443.32 444.93 444.92 444.34
444.50 444.53 444.54 444.19 444.18 443.30 443.04 444.34 444.35 443.94
443.81 443.80 443.83 443.56 443.53 442.75 442.49 443.70 443.70 443.34
443.35 443.38 443.40 443.13 443.11 442.34 442.09 443.22 443.22 442.86
443.77 443.90 443.89 443.42 443.40 442.48 442.21 443.55 443.55 443.10
445.54 445.68 445.68 445.23 445.21 444.24 443.99 445.30 445.30 444.82
445.31 445.36 445.35 444.97 444.94 444.08 443.76 445.17 445.18 444.78
445.83 445.93 445.92 445.41 445.39 444.42 444.08 445.64 445.66 445.22
445.18 445.21 445.20 444.85 444.82 443.97 443.70 445.03 445.04 444.62
445.43 445.60 445.60 445.08 445.06 444.10 443.83 445.17 445.19 444.73
444.92 444.96 444.96 444.66 444.64 443.86 443.61 444.81 444.84 444.48
444.75 444.78 444.78 444.49 444.46 443.75 443.49 444.66 444.68 444.33
444.50 444.53 444.53 444.24 444.22 443.47 443.22 444.39 444.41 444.06
444.42 444.49 444.49 444.20 444.17 443.44 443.20 444.27 444.30 443.95
444.61 444.84 444.84 444.46 444.43 443.61 443.36 444.64 444.67 444.24
444.33 444.46 444.46 444.10 444.08 443.27 443.01 444.15 444.18 443.82
444.04 444.14 444.14 443.82 443.81 443.03 442.79 443.87 443.90 443.53
444.15 444.29 444.29 443.91 443.89 443.07 442.85 443.96 443.97 443.56
444.28 444.35 444.36 444.04 444.02 443.21 442.95 444.17 444.18 443.79
444.51 444.55 444.54 444.22 444.19 443.36 443.08 444.43 444.45 444.07
444.34 444.34 444.34 444.03 444.02 443.20 442.87 444.27 444.29 443.93
444.17 444.18 444.19 443.88 443.86 443.01 442.69 444.10 444.12 443.77
444.03 444.05 444.06 443.71 443.69 442.84 442. SI 443.95 443.97 443.60
443.80 443.80 443.80 443.51 443.49 442.62 442.29 443.72 443.74 443.37
443.40 443.43 443.43 443.12 443.11 442.24 441.95 443.23 443.25 442.87
443.01 443.08 443.07 442.77 442.75 441.86 441.57 442.89 442.91 442.52
442.47 442.56 442.57 442.25 442.22 441.30 441.02 442.30 442.33 441.94
442.10 442.19 442.19 441.84 441.82 440.86 440.55 441.93 441.95 441.54
441.63 441.94 441.95 441.58 441.56 440.58 440.29 441.64 441.66 441.23


B8.1
444
444
443
443
443
443
443
443
443
444
444
444
443
443
443
445
445
445
444
444
444
444
444
444
444
444
443
443
444
444
444
443
443
443
443
443
442
442
442
.17
.06
.87
.77
.64
.42
.41
.59
.41
.29
.55
.27
.76
.32
.38
.04
.02
.31
.90
.87
.68
.55
.28
.27
.40
.08
.86
.88
.01
.13
.00
.87
.70
.51
.29
.03
.50
.24
.05


B8.2 B8.3
443
443
443
.78
.58
.38
443.28
443
442
442
443
442
443
444
443
443
442
443
444
.12
.93
.98
.22
.93
.91
.25
.81
.22
.80
.01
.73
444.57
444
444
444
444
444
443
443
444
443
443
443
443
443
443
443
443
443
442
442
441
441
441
.95
.48
.61
.32
.18
.91
.87
.07
.74
.47
.54
.68 444.45
.84 444.50
.67 444.44
.50 444.36
.34 444.23
.18 444.12
.74 444.19
.50 444.11
.85 443.74
.46 444.00
.17 443.87



ULR2.1 ULR2.2 U2
443.42
443.28
443.07
442.99
442.82
442.62
442.63
442.86
442.61
443.53
442.84
443.50
442.95
442.51
442.68
444.38
444.31
444.68
444.20
444.30
444.06
443.93
443.66
443.61
443.80
443.43
443.18
443.22
443.41
443.58
443.43
443.24
443.07
442.84
442.44
442.10
441.52
441.11
440.82
443.39
443.25
443.04
442.96
442.78
442.59
442.56
442.82
442.60
443.50
443.81
443.47
442.92
442.48
442.64
444.35
444.28
444.65
444.16
444.27
444.02
443.90
443.63
443.57
443.77
443.40
443.14
443.18
443.38
443.54
443.40
443.21
443.03
442.81
442.40
442.05
441.47
441.05
440.77
441.7
441.7
441.4
441.3
441.1
440.9
440.8
441.0
441.0
441.8
441.9
441.8
441.4
441.0
441.0
442.8
442.7
442.9
442.5
442.7
442.6
442.4
442.2
442.1
442.2
441.8
441.6
441.6
441.8
441.8
441.6
441.3
441.2
440.8
440.5
440.1
439.5
439.1
438.9

U9
445.4
444.7
444.6
444.5
444.3
444.2
444.3
444.7
444.3
445.4
446.2
446.1
444.3
443.9
444.5
446.0
445.5
446.0
445.5
445.9
445.4
445.2
444.9
444.9
445.1
445.1
444.7
444.9
444.9
445.0
444.8
444.6
444.5
444.3
444.1


442.8
442.7

U18
444.04
443.86
443.63
443.61
443.38
443.15
443.26
443.50
443.21
444.20
444.50
444.09
443.47
443.00
443.24
444.94
444.97
445.43
444.77
444.88
444.63
444.49
444.21
444.08
444.42
443.96
443.65
443.70
443.96
444.28
444.15
443.99
443.81
443.58
443.03
442.70
442.09
441.70
441.38

-------

-------