&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
-------
------- |