supplemental modeling and analysis report
TRANSCRIPT
SupplementalModeling and AnalysisReport
Atlas Corporation Moab Mill, Moab, Utah
February $1998
Prepared for the U.S. Fish and Wildlife ServiceLincoln Plaza, Suite 404Salt Lake City, Utah 84115
Prepared by Oak Ridge National LaboratoryEnvironmental Technology Section2597 B 314 RoadGrand Junction, Colorado 81503
Table of Contents, .
1.0 Introduction . . a . . . . . . a 1 . a 1 I . . e a . e ~ . . I . 1 . D . . . . . J.
I.1 Project Scope . . . . 1 1 ~ . a a . a . a ~ s D . 1 . ~ 1 . j ~ ~ . ~ 1 . ~ . D 1
2.0 Groundwater and Contaminant Transport Modeling . D . . 3
3* -2.1 Modeling Constraints 1 . a . . e . a a e * * 0 . . *
2.2 Drainage Modeling ~ . a . . . . ~ . . ~ , . . . . . ~ . . 4. -
2.3 Contaminant Transport Simulations . . . . . * . . . .
2.3.1 Plume Simulations . . . . . . . . . . . . . . . . * . . . * . I
2.3.2 Source Removal Simulations . . . . . . . . . . . . . . .
3.0 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . fi
4.0 Retardation of Contaminants . . . . . . . . . . . . . , . . . . . . . 15
6* -
4D -
11
5.0 Trend Analysis of Groundwater Contamination Data . . . 16
6.0 River Fluctuations on Contaminant Discharge Rates . . . 20
ir?7.0 Review of Historical Water Level Data . . _ . . , . . . . . . . 2,. ,jl, ,. __.i_ ,^ ,_:
1: 1 8.0 Impact of Moving Tailings . . . . . . . . .“. . . . . . . . . . . . . 23
9.0 Ccnclusions ~ . . , D . ~ . ~ . . ~ . . . . . . . . . . . . . . . . 1 . . . . 3
10.0 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95. . 5
Appendix A: Statistical Analysis of Historical Water Quality Data
Appendix B: Historical Water Level Data
-- ,_ _i ‘. .- .,a. . 7 a,,:, . ,,Y.,>. ., 1
1 .O Introdutitibn
Rm
1:; 1 The purpose of this report is to provide additional numerical modeling ano data evaluation for the
r,Atlas tailings pile near Moab, Utah. A previous report (Tailings Pile Seepage Model: The Atlas
1 Corporation Moab Mill, Moab, Utah, January 9, 1998) prepared for the Nuclear Regulatory
Commission (NRC) by Oak Ridge National Laboratory/Grand Junction (ORNL/GJ) presented the
results of steady-state modeling of water flow and subsequent discharge to the underlying
groundwater system. At the request of the Fish and Wildlife Service (FWS), this model was.)
expanded to evaluate the impact of drainage from the tailings pile in addition to recharge from
precipitation in a transient mode simulation. In addition, the FWS requested transient simulations
of contaminant transport in the alluvial aquifer. Subsequently, NRC requested an evaluation of
additional hydrologic issues related to the results presented in the Tailings Pile Seepage Model
(ORNL/GJ 1998a) and the Limited Groundwater Investigation (ORNL/GJ 1998b). Funding for
the report was provided by the U.S. Department of Energy. The following section lists the
individual tasks with subsequent sections providing the results. A map for the Atlas Moab Millr?i ! site is presented in Fig. 1.1.
ri / I.% Project Scope
The scope of this report was based on requests by‘the FWS and’NRC during a January 13
conference call with ORNL/GJ. Listed below are the individual tasks that are addressed in this.‘ ;, _-,,, ,^ ._. .., I. L 3 ‘. : “.,l~ _ ,., 1, _report.
FWS REOUESTED TASKS
l Task FWS-1: Transient simulations of pile drainage.
l Task FWS-2: Transient simulations of the contaminant concentrations discharging from
the pile.
0 Task FWS-3: Impact of tailings pile removal on contaminant flux discharging to the
alluvial aquifer and the Colorado River.
Supplemental Modeling and Analysis Report 1
._
3’ I ,.’ ;,
., ,,,_ -’ ‘. ‘. / ‘” “, ._
a’+ai&&;i: ~e~;iti.t~-~;al;sis of& hydrol~gi:gic pa;g;net;rs;used i& ahe model.
f7 .Task NRC3;%Impacts of retardation on contaminant migration in the alluvial
1 groundwater.
a Task NRC-3: Transient effects of river stage fluctuation.“I ,. _. I ;. I _.0 Task NRC-4: Review of historical water level and water quality data and the subsequent
I-i, ;i :
impact on seepage rates from the tailings pile.
‘a Task NRC-5: Impact of construction activities during tailings pile removal on
contaminant discharge to the groundwater.
Each of these tasks are addressed but at varying levels of detail. The bulk of this report discusses
the results of transient numerical modeling for the drainage of the tailings pore water and
contaminant transport simulations. Because’of time and budget limitations, several of the tasks
are addressed only in limited detail. In particular, ORNUGJ acknowledges that the sensitivity
analysis task (Task NRC- 1) has not been adequately addressed.
rt:
2.0 Groundwater and Contaminant Transport Modeling
Ail”
2.1 Modeling Constraints
The,information presented in this and the previous reports (ORNL,/GJ 1998a, b) is preliminary‘ ,, ‘%“‘. -’and is intended- to provide an order of magnitude estimate of the geochemical and hydrologic
processes at the site. Further work, particularly regarding sensitivity analysis of the model and
impacts of contaminant retardation as well as heterogeneities in the tailings pile, variable
saturation of the pile, and transient river effects is needed to improve the reliability of the model
predictions.
Supplemental Modeling and Analysis Report 3
c
. . ’
2.2 Drainage Modeling
The existing two-dimensional model (Fig. 2.1) used in’ the previous simulations (OR.NL/GJ
(”I+
1998a) was modified for the pile drainage simulations (Task FWS-1). Based on recent and-.“?._*_ Jo ,._ ,_historic soil borings drilled through the pile, very fine-grain slimes located at the base of the pile
were included in the simulation. It was assumed that the slimes cover the entire bottom of the
tailings pile. Boring logs indicate that there are areas where the slimes are absent but there is
Fii .,
insufficient data to construct a reliable map of their extent and thickness. The same permeabilityI,and unsaturated hydraulic characteristics (0.0003 ft/d [IO-’ cm/s]) used for the clay cover were
used to represent the slimes. It should be noted that the tailings material above the slimes is
mostly fine to very fine-grained sand.p*
ti
t;
Boundary conditions consisted of mixed conditions of constant head or flux boundaries. Thet?I; lateral boundaries consisted of constant head values for the alluvial acluifer immediately“_ I. ,. ,,I,, ,. _
upgradient of the tailings pile and the river elevation downgradient of the pile. The lateral
Fi boundaries above the water table were set as “no flux” boundaries. Head values were based onF- , . /water-level measurements from monitoring wells or surveyed river elevations taken during
1c:1 J December 1997. The lower boundary, which was’also set as a “no flux’ boundary, may not be% , .
Ecompletely accurate considering the evidence of vertical flow from the underlying salt formations
,_,.4 ,
,; as i;ldic;t;d by the eibv;ted .~~oride’~oncentrations y-wng;;i;;& &ie pile ((L&.&
1998b). Nevertheless, any vertical flow should have little impact on transport calculations through
the cover and pile. The upper boundary was set as a ‘fixed f&x value that corresponds to the
estimated yearly infiltration rate which was estimated to be 0.0002 ft/yr based on a fixed
percentage of the average yearly rainfall of 8 in/yr (Blanchard, 1990). This recharge rate,
cc: Iresulting from precipitation, is an estimate and is subject to a range of interpretations.
Initial conditions consisted of saturated moisture contents for the tailings material based on the
assumption that the slimes and tailingswere saturated when placed in the impoundment.
Simulations were run in the transient mode with small incremental time steps that increase based
on the convergence criteria. Moisture content and vertical flux values for water discharging
through the pile were recorded during the simulations.
m Supplemental Modeling and Analysis Report 4
I‘
Figures 2.2 and 2.3 show the vertical flux and saturation as a function of time for the center of the
fine-grain sand tailings and at the base of the slimes immediately above the alluvial aquifer. The
graphs show that the bulk of the pile drainage occurs within the first 100 years. Steady-state
conditions, defined as the point in time where the flux at the base of the pile matches the recharge
rate, occur after 238 years. The relatively flat nature of the curves (Figs. 2.2 and 2.3) between
100 and 238 years represents continued drainage of bore water in the tailings and slimes but at a
rate of approximately 5 percent of the recharge flux fi-om precipitation.
6” .
s
.
“./ :t,
F?
SDll
eC”
PT”
2.3 Contaminant Transport Simulations
Two separate contaminant transport simulations were conducted for this investigation. The first
simulation (Section 2.3.1) consisted of the transport of contaminants from the pile into the
alluvial aquifer for a period of 41 years (Task FWS-2). Forty-one years represents the available
time for transport since the initial placement of tailings in the impoundment in 1956. The second
simulation (Section 2.3.2) assumed that the sources of contamination, the tailings pile and all
other potential sources, were removed (Task FWS-3). This simulation predicts the time required
for contaminant levels in the aquifer to return to pre-1956 levels.
For all simulations performed, contaminant concentrations were normalized to the average
oontaminant concentrations in the tailings pile pore water. It was assumed that the contan&ants
were conservative, therefore, Kd values were set to zero: Based on published values (Freeze and
Cherry 1979) of the coefficients of dispersivity for a sand aquifer, the longitudinal dispersivity was
set at 0.5 ft”/d and the transverse dispersivity was set at 0.05 fi2/d. Future sensitivity analyses
should be conducted to evaluate the importance of these parameters on the contaminant transport
simuiations.
2.3.1 Plume Simulations.; :. ;_.
Initial groundwater flow conditions consisted of a saturated tailings pile at a normalized
contaminant concentration of 1 .O. Contaminant concentrations in the underlying alluvium were
Supphnental Modeling and Analysis Report 6
_ ,
-L :
i : DISCHARGE vs. TIMEv i‘0.0020 -l--
l-9t :
b”nb 0 . 0 0 1 6 -t !
0 . 0 0 0 8 -
0 . 0 0 0 4 -
- l - l * Center of tailings- - - - Base of slimes
i 1i :i 1\i 1.
0.0000F i 0 2 5 5 0 7 5 1 0 0 1 2 5 1 5 0 1 7 5 2 0 0 2 2 5 2 5 0*L # ~\“t.lkl,L\~l~*lg~\mrc145.dwg
Time, years
Figure 2.2. Flux rates in the tailings pile during desaturation.. _.,
\/
.:
L-Y: I SA TURA T/ON VS. TIMEf > 1.0. , \
I
iiiiii
- - - - Base of slimes-.-.- Center of tailings
‘\‘\
‘\‘\ -w --.a. -.-.-.-.l.-.-.-.-.~.~.~.
0 . 5 -j- I0 2 5 50 7 5 1 0 0 1 2 5 1 5 0 1 7 5 2&O 2 2 5 2 5 C
hlusr\klrk\mscdwgf\msc2~.dwg
p”, T I M E , Y E A R S ,i !a,
Figure 2.3. Moisture contents (6s) changes with timein tkiiings pile ri&ed.“bn ntini~ri~al tktidelirig simulation.
”
F”
1‘ ;
b 1
f””
i..
r
”L j
m
i
” :
Fa
: ’?
set to zero.‘. ‘,,
Figure 2.4 illustrates the contaminant plume in the alluvial aquifer after 41 years. The simulations
predict a mature plume that reached the river several years prior to the end of the 41-year
simulation period. This is consistent with contaminant plume maps based on water sample results
presented in the 0RNIJG.I groundwater report (ORNIJGJ 1998b).
The model predicts that the contaminants emanating from the tailings pile are diluted by 60
percent by the groundwater. Although the 60 percent dilution rate is consistent with mixing
simulations for ammonia, the extent of retardation by aquifer sediments and the extent of
oxidation of ammonia to nitrate is unknown, therefore, reducing the reliability of the ammonia
simulation. In a similar fashion, sulfate simulations are questionable because the solubility limits of
several of its salts are probably exceeded. The mixing simulations are, perhaps, best served by the
uranium data due to the conservative nature of the uranyl carbonate ion in this geochemical
environment. For uranium mixing calculations, however, a 60 percent dilution results in a
reduction of the tailings pore water concentration of 23.5 mg/l to 9.4 mg/l in the alluvial
groundwater downgradient of the pile. The actual average uranium concentration in the alluvial
groundwater is 4.62 mg/l--approximately one-half the amount predicted by the model. Thus,
these results suggest that the initial input of contaminated water is high by a factor of two during
the first 30 years of the simulation. Most likely, the initial condition of total saturation is not
correct and a lower saturation value should be used for future simulations. Using a lower
saturation value will permit a more accurate simulation of the concentrations that are now being
measured. Relative concentrations, as described in the next two paragraphs, are also affected by
the choice of initial saturation conditions. The relative concentrations predicted later
are probably too high but still within a factor of 2 or 3 of the actual values._- - ,./
A simulation was conducted to evaluate the long-term contaminant concentrations near the river
based on the assumption that the tailings pile is not removed. Figure 2.5 shows the normalized
contaminant concentrations at a node in the center of the groundwater contaminant plume/. , ,. ,v.- _ .,a, ./. ..,‘ /.A,^ ,.r_ ,I :. ,L%%.-r . j.l”., ,i (_ ‘._ ,, >- ,,.. I . . . . . . . ., .(“.&1 _.,,I) .,~ri,,r,<l--‘ .,ii”,*“r;.adjacent to the Colorado River. This node is only one foot thick. Actual contaminant
concentrations as measured in a monitoring well would be lower because the well is screened over
Supplemental Modeling and Analysis Report 9
51laA!y
-0pelo~o~
rk ,
NORMALIZED CONTAMINANT CONCENTRATION
P P P P 00
P P Pd N W P In 0, ;iPm
000”0
‘I0Itt
‘:I
several nodes that exhibit Power contaminant concentrations.
The model simulation predicts that after 50 years, the decline in the contribution of contaminants
due to the drainage of tailings pore water begins to impact concentrations at the river (see Figs.P,_
2.2 and 2.3). Because 41 years have passed’ since the tailings were first placed in the
impoundment, these results suggest that contaminant concentrations at the river will continue to
increase for 9 more years. The simulation further suggests that the peak of the contamination is
located near the downgradient edge of the tailings pile. This result is consistent with the trend
analysis conducted in Section 5.0 that indicates that there is no discernable trend in contaminant
concentrations.
2.3.2 Source Removal Simulations
The source removal simulation assumed that after 5 1 years of contaminant release from the
tailings pile and subsequent transport by the alluvial aquifer, the source of contamination was
I!7B ;
completely removed by remedial actions. after 41 years of contaminant transport, an additional
10 years of continuous input of contaminants was simulated based on the assumption that tailings
removal will require approximately 10 years to complete. During these 10 years, contaminant flux
rates were consistent tith existing rates under the no source removal scenario.
-Figures 2.6 through 2.8 show the extent of the plume at 10, 16 and 27 years afier source removal.
After 35 years, the model predicts that contaminant concentrations near the river will be less than
10 percent of what was present in the initial tailings pile pore water. Results of the simulations
are consistent with average linear groundwater flow velocities for the alluvial aquifer as reported
previously (ORNL/GJ 1998a). Finally, using an average linear velocity of 107 ft/yr (ORNUGJ
l998a) and approximately 3800 feet of travel distance, 3 5.5 years would be required to clean the
aquifer. This value, however, is unrealistic because this simulation does not consider the effect of
E molecular diffusion of contaminants from permeable zones into adjacent low permeable zones,_.and subsequent diffusion back to permeable zones as the concentration gradient reverses. The
net result of these factors is that there will be an increased amount of time before the aquifer
?--I
i /
Supplemental Modeling and Analysis Report 12
years after source is removed
e- --s.- J
--._ L 3 ““‘3
Inactive Zone
Vertical Exaggeration = 10xGrid spacing: X=100 ft, Y=lft 0.20 Relative Concentration
m 0.60 Relative Contamination
Fig. 2.6. Extent of groundwater contaminant plume 10 yearsafter source removal.
,I..
, , _ i * , (I _,Ck ,I_
Xg:0,r IIII *
27 years after source is removed
-w-I. 1
-or?91. . .‘3
v---w“3
Inactive Zone
Vertical Exaggeration = 10xGrid spacing: X=100 ft, Y=lft
0.20 Relative Concentration
Fig. 2.8. Extent of groundwater contaminant plume 27 yearsafter source removal.
rI :. /
completely cleans itself Results from other tailings pile removal projects should be evaluated to
improve the estimate for cleaning the aquifer.
3.0 Sensitivity Analysis
The purpose of a sensitivity analysis is to quantify the uncertainty in the calibrated model caused
by uncertainties in the estimate of aquifer parameters, initial conditions, boundary conditions, and
contaminant concentrations (Task NRC-l). A sensitivity analysis is an essential step in all
modeling applications. During a sensitivity analysis, values for hydraulic conductivity, storage,
recharge, and boundary conditions are systematically changed within previously established
plausible ranges. The magnitude of changes in head and contaminant concentrations from the
calibrated solution is a measure of the sensitivity to that particular parameter.
Because of the limited time frame available to conduct this modeling of the Atlas site, it was not
Fi , possible to conduct a sensitivity analysis. ORNL/Gi recommends that future Work be performed
to complete this task. If this recommendation is followed, the most important hydraulic
parameters will be identified. Additional field Work Can then be pro@sed to better define the most
important parameters.
L 4.0 Retardation of Contaminants
,Pl“ r ’i
No specific studies have been performed to address retardation of contaminants (Task NRC-2).
However, in general, the oxidized, sandy, gravelly, highly alkaline nature of the alluvial aquifer- ”
will promote migration of the contaminants, particularly those found as anions: uranium,
molybdenum, sulfate, and nitrate. Ammonium is a special case because it is apparently being
F:
converted to nitrate due to the oxidizing nature of the alluvial system. In addition, sulfate species
may be near their solubility limits. The discussion below, therefore, focuses on uranium because
its geochemistry in this type of environment is well understood. Because of their similar behavior,
this discussion can also be applied, in a general way, to molybdenum..(, : .(_ ” ,
Supplemental Modeling and Analysis Report 16
gaI
The distribution of uranium species is highly dependent on the carbonate concentration. The third
graph in Fig. 4.1 (Waite and Payne 1993) is probably not too dissimilar from what might be
expected for the groundwater at the Atlas site. Thus, ‘[UOi(CO 3) J2- and UO,CO, may be
important species. However, the speciation is so dependent on the partial pressure of carbon_“” ‘. ,._,’ ‘, j ‘;_ .’ “l.i., I ““.; ~.‘,,_. :;. _. ,‘I, ,” ‘.,‘,. :dioxide that it is difficult to generalize. For example, Duff and Amrhein (11996) state that “in the
m/
I /presence of dissolved carbonates, U(W) forms several strong carbonate complexes: @JO,),
C03(OH)3-, UO2CO3, UO2(CO3)2 2-, UO2(CO3),4-.” Nevertheless, it is concluded that uranium in
F”:I 1
the alluvial aquifer will be in the form of one of several negatively-charged carbonate species.
As can be inferred by the previous discussion, adsorption decreases with decreasing pH and
increasing alkalinity. The reason for this effect is that formation of the dicarbonate species is
increasingly favored and bicarbonate ion competes for available adsorption sites (van Geen et al.
1994). Soil and mineral surfaces are general&negatively-charged. Thus, the negatively-charged
or neutral carbonate species have little propensity to sorb on the soil minerals,?-i
This conclusion is supported by Duff and Amrhein (1996) who concluded that “with increasing
carbonate alkalinity, U(VI) most likely formed negatively charged carbonate complexes which did
not strongly adsorb to the soil or goethite in the study. Therefore, U(VI) adsorption to soils
dominated by permanently charged clays is not a likely factor controlling U(VI) solubility . . . ”
”L if :
To summarize, conditions at the Atlas site appear to be favorable to the formation of carbonate
complexes of U(VI). Such complexes are not strongly adsorbed by soil materials and are,
therefore, relatively mobile in the environment. Extensive sampling and analysis regarding
uranium speciation and’ the effects of other ions would be needed to apply a geochemical model
that would more accurately simulate uranium mobility.
p5.0 Trend Analysis of Groundwater Contamination Data
NRC requested that historical water quality data be reviewed and that a trend analysis be provided
(Task NRC-4). Several wells at the site have been sampled and analyzed on a regular basis since
1987. The objective of the statistidal analysis presented i‘n this se&ion is to detect changes or
Supplemental Modeling and Analysis Report 17
i
gm
i,b. ,
00
80
t
70
60
w so
g40
20
10
PH h:\atlag\corelm~~ndp.cdr
Figure 4,f. Speciatim of lU%d V(W) as a Rmction of pIi. a) in theabsence of carbonate, b) in equilibrium with the atmosphere(fC0~ = 10-‘.’ atm), c) in soIutions containing 2 mM totalcarbonate. porn waite CNW m, ~~93)
trends in contaminant levels. For the groundwater chemistry data collected, the nonparametric
ati ,
Mann-Kendall test for trend analysis was conducted. This procedure is particularly use&l because
missing values are allowed and the data need not conform to any particular distribution. Also,
Fi1,
data reported as trace or less than the detection limit can be used by assigning them a common
value that is smaller than the smallest measured value in the data sets (Gilbert, I987). Tests were
!-- conducted with the null hypothesis being no trend in the data. The null hypothesis is rejected ift;, ..( ;.. I, , I~ I , :
the probabrhty value (p) &respondingto the &mputed ‘ha&Kendall statistic (s) is less than a
specified significance level. For the Atlas site, both 90 and 95 percent confidence levels were
selected to evaluate the data. The use of a 90 percent confidence interval reflects the high
variability inherent in environmental data and is intended to provide an indication of trends but at
a lower confidence interval.
If the p value is less than 0.05 or 0. I, then the evidence is insufficient to reject the null hypothesis
that there is no trend in, the data. Mann-Kendall- tests were conducted on selected analytes from a
total of 4 wells located downgradient of the tailing pile (Fig. 5.1). Results are presented in Table
5.1. The statistical and analytical data are presented in Appendix A. Of the eight analytes tested,
four yielded a trend that was within the 90 percent confidence interval. Of these four, two
showed downward or decreasing concentrations while two showed an upward trend. The
remaining four analytes showed no trend within a 90 percent confidence interval. For
concentrations that exhibited a trend within the 95 percent confidence interval, two wells showed
”rb
a downward trend and one well showed an upward trend.
pL
The Mann-Kendall test on uranium data from well ATP-2-S showed no trend within the 90
percent confidence interval. However, the characteristics of the data for this well show a repeated
increase and decrease in concentration values.which results in an uncertainty in the statistical
analysis. Time was not available to conduct further statistical tests, but it is recommended that
the Seasonal Kendall test be conducted on this data set. Using this analysis, the cyclic trend can
be removed and it is expected that a statistically significant downward trend in uranium
concentrations for ATP-2-S will emerge.
Although there is some indication of a downward trend for a few analytes, the data taken as a
Supplemental Modeling and Analysis Report 19
., .,
.*, <.. .,‘whole inaicat‘ed’no‘reliable trend &I concenti-aiions. ‘Thus, there is not c&&tent evidence for a
ptrend, in the groundwater contaminant concentrations downgradient of the pile.
bi ,“. ., / ;:- ,
Table 5.1 Statistical analysis of water quality trends based on groundwater samples from
downgradient wells
Well Analyte P-Value Trend
ATP-2-S 1
v- L
Consequently, there is no strong evidence that contaminant lev& in ihe groundwater arer”:i :
decreasing due to a decrease in the contaminant concentrations in the tailings pile or because of
reduced discharge rates. Additional evaluation of geochemical factors is needed to provide ar ,.‘ _.1 comprehensive understanding of ihe obs&%d co;ltaminant-conceati~~s downgradi&nt bf the
tailings pile. Such an evaluation would include calculation of the solubility limits of selectedri analytes within the tailings pore water and measurement of geochemical factors that affect the
i”ik i
migration of the contaminants.
p6.0 River Fluctuations on Contaminant Discharge Rates,. . .,,_ ^.
FThe,N$C requested an evaluation of how fluctuating river stages affect groundwater discharge of
contatina&‘i&o the ‘Colorado ‘five; (Task &C-j). Specifically, as the Colorado &ver stage
I? rises during spring runoff, how are contaminant flux rates fi-om the alluvial groundwater to theh ; river affected? Although’ there was in$uffici&t time to &lly anal$ze ttie isstie;.ther&” ares reaso’tisrrr(r AEl
why the contaminant flux values presented in the OIXNUGJ groundwater report (OlWL/GJ
P1, !i i
Supplemental Modeling and Analysis Report 21 ’
1998b) are a reasonable estimate of a yearly average. H’irst, the hydraulic gradient across the
Atlas site reflects average boundary conditions for the aquifer. For example, Canonie (1994)
reported a hydraulic gradient across the Atlas site of 0.004. A potentiometric map based on water
level measurements taken by ORNL/GJ in December I997 also yielded a hydraulic gradient of
0.004 (ORNLIGJ 1998b). Second, a review of stream gage records for the nearest W.S.G.S.
station at Cisco, Utah shows the daily discharge values since 1922 (Fig. 6-la). Figure 6-lb shows
recentwater”levels for January of this’year. At the time of the ORNL/GJ field effort(December
1997), the Colorado river was discharging approximately 5000 cfs. If this discharge value is-, _ .I-_ ., ,. \, .,
compared with the historical flow, there are brief periods (two or three months) of peak now that
exceed that discharge rate, but the bulk of historical river discharge values are below 5000 cfs.,. “. ., ,, _ ,_ ._‘. ._For peak discharge rates associated with spring runoff, groundwater discharge rates will decrease
in response to the higher river levels. This decrease in groundwater discharge rates during late
spring and early fall is offset by higher groundwater discharges in the late fall and winter in
response to lower river levels. Consequently, the contaminant discharge values based on
December1997 hydrologic data are a reasonable estimate of an average yearly value.
A more definitive approach to address this question would be to vary the downstream constant
head boundary of the contaminant transport model to represent on-site river stage data. Then the
model could be used to calculate the resulting time dependent contaminant flux rates. Time and
budget were not available to perform this task!Nevertheless, as de&bed‘ al&$ there would
probably not be a substantial change regarding the present description of the contaminants in the
alluvial aquifer.
Pp.;
7.0 Review of Hist.cyical Water ,@I@ Data.
$7 ,,c 1 NRC requested that the historical water level data from the Atlas monitoring wells be reviewed0 with respect to the tailings pile dewatering program that began in mid-1990. In particular, the
&“1P ’ impact on water levels in the tailings pile and the implications to seepage rate estimates was to be1
addressed (Task NRC-4). Water-level data reviewed for this evaluations were obtained from the
Canonie report (1994) and are presented in Appendix B. Interim water level data collected by
Supplemental Modeling and Analysis Report 22.
Colorado River Nmr Cisco, UtStrtlm Nuwbcr: llY1811500
199a
PROVISIONAL DATA SUBJECT TO REVlSiON
15 16 l? l(1 1s 20 21 22
updatd: 91 l 2rn1998 99:r5 bawd on ?Oyeareofrndrd.
h:\dasiskam.cdr
Figure 6.1. (a) Flow records since 1922 for the Colorado River at Cisco, UT and’ (b) recent flow records.
Atlas since the Canon& report were not available to be in&&d in the following discussion..
Graphs B- 1 through B-4 in Appendix B compare individual alluvial wells located downgradient of
the tailings pile with river stage elevations from 1989 to 1994. Fluctuations in monitoring well
water levels correspond to river stage.fluctuations’indicating that there is hydraulic connection
between the river and the alluvial aquifer downgradient of the tailings pile.
!-ic
rI
Graph B-5 in Appendix B shows a comparison of the water elevation in the pond at the top of the
tailings pile to the river stage from 1989 to 1994. As expected, there is no discernable correlation
between the two, indicating no hydraulic connection.
r Graph B-6 in Appendix B compares water levels in wells drilled into the tailings pile with riverI
stage levels. There is a large variation in water levels among these wells. A source of the
k”: variation be that the wells are completed in different lithologic units. Review of the well
completion information presented in the Canonie report (1994) and the Dames & Moore reports
F (1973 and 1981) provided the data presented in Table 2. As shown in Table 2, the wells
represented in Graph B-6, are completed within the alluvium, the sand tailings, and the slimes./!7k! Water level data in Table 2 is from January 1989, June 1990, and February 1994 and were used to, . /.
I”calculate the change is head values before and after the dewatering program began. The largest
k I! / decline in water level is observed in well B-4 and is representative of conditions before the
mc >b :
dewatering program began. Although this‘well shows no influence from the river (Graph B-5 in
Appendix B), it yields the greatest decline in water levels. However, there is no obvious
i-e ic (I
explanation for this decline based on the available data. Comparing water levels in wells A-9 and
BAA-4 shows a higher water level in well Bti4, which is underlain by slimes, compared with
A-9 that is screened in both the sand and alluvium. This difference in water levels may be due to,_” .,’
the low permeable slimes acting as an aquitard resulting in perched water in the vicinity of well
BAA-4.
,.As shown in Table 2, there is a larger average decline in water levels for the 17 months prior to
the initiation of the dewatering program than for the 42 months after pumping began. This
evidence indicates that the pumping is having little or no effect on the dewatering of the tailings
Supplemental Modeling and Analysis Report 24
Table 2. Well completion, lithology, and water level data for wells located on the Atlas Tailings Pile
Well Surfacenumber elevation
B-4 4,040
BAA-4 4,052.3
Screen TotalI I
Lithology Waterinterval, depth? elevation,
R I f-t 1 l/14/89I I
70 to 80 110I I
Alluvial well, no 4,035indication of slimes
39.2 to 41.2 ? Based on B-15 logSand tailings well,no slimes
4,020
29 to 49 50 Sand tailings, 3,978alluvial well, noindication of slimes
80 to 82 119I I
Completed well in 3,973slime tailings
51.7 to 53.7 N/A Based on B-l 4,015Screened in sandtailings with 5 A ofslime below I
Water Elevation Water Elevationelevation, difference, elevation, difference,6/l l/90 l/89 t o 6/90 2/25/94 6/90 to 2/91
4,024 1 -11 1 4,022 1 -2
3,976 -2 3,973 -3
All data compiled from Canonie report (1994) and Dames & Moore reports (1973 and 198 1).
I1
,
‘, 1
:
‘_
Supplemental Modeling and Analysis Repot-t 25
jq .
b 4
mb !
t
.__. .<.
pile. It appears that ‘natural drainage @ior to pumping sad&d the larger water-level decline.
Canonie (1994) reported that the remedial wells pump at a combined total of approximately 3
gpm when averaged over the period from July 1990 to’1Febiuar-y 1994. Assuming that the pores
drained to the residual moisture content.of 0.63 (ORNL/GJ 1998a) and that all of the decline in._ I . ._ _ ;. _.
water levels is due solely to pumping, the resulting drop in water levels in the tailings pile in .-
response to pumping would be 0.73 ft over this 42-month period or 0.21 fi per year.
If it is assumed that water level declines in monitoring wells completed in the sand tailings are due
to the discharge of water into the underlying aquifer, then it is possible to estimate the recent
P discharge of tailings pore water. This estimate of discharge would be in addition to dischargeE:
p5P‘
occurring in response to infiltration due to precipitation. For this evaluation, recharge rates in
response to water level declines were estimated before and afler dewatering began. Wells EE-4
BAA-4, both completed in the sand tailings, exhibited water level declines. From 1989 to 1990,mi: p water levels declined three and five feet, respectively. Multiplying these water level declines byi ,
the surface area of the pile (3868103 R2), the porosity of the sand tailings (0.66)[ORNL/GJ
1998a], the drainable portion of the porosity as predicted by the numerical simulation
(O.l6)[ORNL/GJ 1998a], and dividing by the time required for the water level decline (17 mo.), af”It:I- I discharge rate of 12 to 20 gpm results. If water level declines from 1990 are used, and the
ri._ r
drainable portion of the porosity is adjusted to 0.23 to reflect the increased drainage time, then the
recharge rate due to drainage of the tailings pore water ranges from 2.5 to 5.0 gpm. This
f-7i
recharge rate is comparable to the value (6.7 gpm) predicted using the uranium mixing calculation
f-Yft I
8.0 Impact of Moving Tailings
NRC expressed the concern that excavation of the tailings pile could result in a “pulse” of
contaminated water entering the groundwater system (Task NRC-5). The additional source of
the water, according to the NRC, that could cause this pulse was attributed to dust suppression/
control measures used during tailings pile excavation activities. To address this concern,
ORNLIGJ discussed the dust suppression/control measures used by the DOE with Don Metzler, a
hydrologist with the DOE in Grand Junction. According to Mr. Metzler, the volume of water
typically used for dust suppression during tailings pile excavation above the water table would
F; i
Supplemental Modeling and Analysis Report 26
I
rt- :
not result in a “pulse” of’contamination to the groundwater system if proper management of
construction/excavation activities is provided. Water added for dust suppression during”excavation would only penetrate a few centimeters into the tailings and evaporation losses would
be significant. Further, low velocities associated with unsaturated transport would be insufficient:
to move the moisture to any significant depth before the tailings were excavated. however, for
excavation activities below the water table, DOE has found that dissolution of contaminants may
be increased by the remedial action (D. Metzler, U. S. DOE Grand Junction Office - personal
communication with F. Gardner, 2/3/98).
9.0 Conclusions
n Results of the modeling simulations have resulted in estimates of the time for the pore water in the
r!t (: ;
tailings pile to drain and for the amount of time needed for the groundwater system to clean up
after total source removal. In addition, several issues raised by the FWS and NRC were
addressed. Listed below are the individual conclusions from the modeling and data analysis.
a Model simulations indicate that the bulk of the pore water in the tailings drains after 100
years with 238 years required to reach steady state conditions
l The model predicts that the contaminant plume entering the river is mature-- a finding
consistent with site characterization data.
0 Simulations predict that the peak contaminant concentration reaches the river 50 years”after emplacement of the tailings (9 years from the present) and then declines to a steady
rate after approximately 100 years.
0 Source removal simulations indicate that it would require a minimum of 35 years for the
aquifer to clean up to pre-1956 levels.:
a Retardation of uranium and molybdenum in the alluvial aquifer is not believed to be
significant.
a Statistical trend analysis.of the downgradient water’quality indicates that there is no
consistent evidence for a trend in, the data. The contaminant transport modeling is, , . _..,consistent with this finding.
a River fluctuations are not believed to have a large impact on average contaminant* ., . . i . . .’
Supplemental Modeling and Analysis Report 27
discharge estimates presented in the initial ORNL’GJ groundwater investigation report
(ORNL/GJ 1998b).
Historical water levels indicate that the present dewatering system is having minimal
impact on the water balance in the tailings pile.
a Excavation and removal of tailings is not expected to adversely impact groundwater
I”
i 1( :quality.
Pp 4c 1
Based on the analysis presented in this and the previous ORNL/GJ reports, there is evidence
supporting a total discharge rate from 6.7 to 20 gpm (ORNL/GJ 1998 a, b). Uranium mixing
simulations and post dewatering water-level declines in the tailings pile support the lower
discharge rate. Ammonia mixing simulations and pre-dewatering water level declines in the
tailings pile support the higher discharge rate. Additional data and sensitivity analysis are needed
to better define the actual discharge rate ofwater from the tailings pile to the underlying
i*: groundwater system.!
F
g+
10.0 References
Blanchard, Paul %. 1990. Ground-Water Conditions in the Grand County Area, Utah, With Emphasis on
the Mill Creek-Spanish Valley Area. Technical Publication No. 100, S&e of Utah, Department of
Natural Resources. Prepared by the United States Geological Survey in cooperation with the Utah
r Department of Natural Resources Division of Water Rights.
i-; ’i , Canonie Environmental 1994. NRC Technical Information Request, Atlas Corporation Ground
Water Corrective Action Plan Uranium Mill Tailings Disposal Area. Moab, Utah.
Project 88-067. July 1994. Canonie’Environmental Services Corporation, Englewood,
pF/
Colorado.
Dames & Moore 1973. Supplement to Environmental Report, Moab, Utah Facility for Atlas
Aknerals. Job No. 5467-063-06. ’
Supplemental Modeling and Analysis Report 28
.,
rr Dames & Moore 198 1. Report of Engineer@ Design Stu& Additions to Tailings Pond -
cF. ,
embankment System. Moab, Utah for Atlas Minerals. Job No. 5467-O 1 S-06. February 15,
1978. New printing May 26, 198 1.
Duff, M. C., and C. Amrhein. 1996. Uranium(VI) adsorption on goethite and soil in carbonate
FI
bsolutions. Soil Science Society American Journal9 60: 13 93 - 1400.
ci
Freeze, R.A., and J. A. Cherry. 1979. Groundwater. Prentice-Hall, Englewood Cliffs,
New Jersey.P: Gilbert, R. 0. Statistical Methods for Environmental Pollution Monitoring. Van Nostrand
Reinhold, ISBN o-442-23050-8, New York.
ORNL/GJ 1998a. Tailings Pile Seepage Model of the Atlas Corporation Moab Mill, Moab, Utah.
Prepared for the U.S. Nuclear Regulatory Commission by Oak Ridge National
Laboratory, Environmental Technology Section, Grand Junction, Colorado. January 9,P
[. 1998.
cr ORNL/GJ 1998b. Limited Groundwater Investigation of the Atlas Corporation Moab Mill, Moab,
Utah. Prepared for the U.S. Fish and Wildlife Service by Oak Ridge National Laboratory,
Environmental Technology Section, Grand Junction, Colorado. January 9, 1998.
van Geen, A., A.P. Robertson, and J. 0. Leckie. 1994. Complexation of carbonate species at the
goethite surface: Implications for adsorption of metal ions in natural waters. Geochim.
C’osmochim. Acta, 58:2073-2086.
Waite, T. D., and T. E. Payne. 1993. Uranium transport in the sub-surface environment -
F-ii
Koongarra - A case study. InMetals in Groundwater, Allen, H. E., E. M. Perdue, and D.
S. Brown, eds., Lewis Publishers, Chelsea, Michigan.
p””
F
Supplemental Modeling and Analysis Report 29
APPENDIX A _
STATISTICAL ANALYSIS OF HISTORICAL WATER QUALITY DATA
_’
,;: ‘.
I, ,.. . , ,- ,, , . . 2 , I-
File: J:\FOX\GENSTAT\AMM2-NIT.CSV------------------------------------------------------------------
."^ ",_. :. ,., 5, .,I_ _, ."I, r ', ,.-A al. :, _' 'i ,I .<. : ,: _ ,, .c,l,J PROB. OF EXCEEDINGTHE ABSO~iJJTE VALUE
KENDALL OF THE Z STATISTICS Z (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTXC N IF N > 10
fy 1 1 97.00 1.4308i 34 . 153.
I7tr I
SEN SLOPECONFiDENi3E itiERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT
1 1 :050 010 -.667 -.180 :500 500 2.353 1.667100:200
000:ooo
. 500 1.429
. 500 1.136
m
: ii ,
30(
26C
2oc
p
8
'1 ,150.s80
1.00
60
0
Nitrate ConcentrationsMonitoring Well AMM-2
P-.--_-_.-i
L
-
S/1.1/87 12/23/88 s/7/90 "* 9/19/91 l/31/93
Dates Sampled
6115194 10128l95 .3
.
b-u
File: &\FOX\GEN‘kTAT~AMMiJA6.CSV
p------------------------------------------------------------------.,.
@ROB. O F EXCI&DING. d THE ABSOLUTE VALUE
KENDALL OF THE 2 STATISTIC-E s Z
STATION' SEASON STATISTIC STATfsTiC ‘N "(TWO-TAILED TEST)IF, .N," ; ,l‘o ._
F 1 1 1.35320 20 . 176L !v
F!I
SEN SLOPECONFIDENCE“INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT1 1 010
:050-.013 013
:013. 057
,000 050.lOO .ooo .013 :050.200 .ooo . 013 . 031
06/L&
L8/1118
sr*i ~I 1 .-i
File:. J:\kO%\btiNS~AT\hM2;kA8.C& ..km -------------------------------------------------- ----------m.-____t 'k ; PROB. OF EXCEEDING,. ., MANN- THE ABSOLUTE VALUE
KENDALL10 OF THE Z STATISTICIr.-; S Z (TWO-TAILED TEST)P "STATION ‘StiA&N's ,kkSTIC STATISTIC N IF N > 10
P 1 1 32.00 1.00825 20 -313f ; ."b :
SEN SLOPECONFItiENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT1 1 010
:050-.040 038
:038150
-.027 :111.lOO -.019 .038 100. 200 -.009 . 038 :083
File:J:\FOX\GENSTAT\AMM2-PA8.CSVP&nted:12/31/97 Page: 1
Linear Regression
Minimum Maxinnlm Mean Median Std Dev Slope R-Sqr Num
0.00 3.75 2.52 2.65 0.82 0.13522 0.13 20
i
*.. . 1
-z
File: J:\FOX\GENSTAT\AMMZ-URA.CSV---------_----------------------------------------.-------------------
PROB. OF EXCEEDING
89* 8 KENDALL, s *
STATION SEASON STATISTICZ
STATISTIC N
THE ABSOLUTE VALUEOF THE Z STATISTIC(TWO-TAILED TEST)
IF N > 10*I!
1 1I 1 -4"o.oo -1.36919, .,, 19 / 1.71
SEN SLOPE'CONFIDENCE INTERVALS
m STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT)i 1 1 :050 010 -113.758 -42.308 60.468_
-95.044 -42.308 17.642P- 100 -89.691 -42.308' 7.143t, . :200 -74.644 -42.308 . 000
t-7, r
<I” 1 y-j ,-J ,” ‘“J : -yj ::: .“f - 11-3 1: :f *,I .) :I:1 ;‘“-‘j “-IT3 .?I- 3 Y”1
Printed:12/31/97File:J:\FOX\GENSTAT\AMkf2-DR?L.CSV
Page: 1
I Linear Regression IMinimum Maximum Mean Median Std Dev Slope R-Sqr Num
0.00 3900.00 2650.00 2700.00 813.86 0.58730 0.00 19
S6i8ZlO 1
L8/ L 118
tm.,’ i
.._ *
File: J:\FOX\GENSTAT\AMM3-MOL.CSV_--_-_-_-__-------------------------------------------------------
PROB. OF EXCEEDING-MANN THE ABSOLUTE VALUEKENDALL OF THE Z STATISTIC
mI S Z (TWO-TAILED TEST)f STATION SEASON STATISTIC STATISTIC N IF N > 10
f?t
1 1 15.00 .49040 19 . 624
T:SEN SLOPE
CONFIDENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMITF4 . 1 1 010
:050-.015 003 028-.OlO :003 . 024
100 -.007 003 :018F :200 -.005 :003 . 014c1 ‘"4
e-5f 1k 3 . . , ,.’ _,_ ,. _. ,.
c;. ‘i&
‘.)_ ,I, ..,<* ,, ,” . . . . _ .‘
rt
n
.;.; _ .,, _” , ,, _. ,. _,,,,. ‘ ..” .,i. r. I
v-l
r
I”L. /
i i
PI
k~!,
-0-lrl-Wd.0-m&2.0
-WrJ.0
-
2.t-l
f:.T-l
-
.l-l
-z.0
i ’
‘
I”) !f
?i
sll
3
S6/8ZIOl
P6lS LIQ
88/82/Z 1
L8/ 1118“! 0
0
mi iI 2
File: J:\FOX\GENSTAT\AMM3-RA8.CSVw _---____------^---------------------------------------------------
t:PROB. OF EXCEEDINGTHE ABSOLUTE VALUE
KENDALL OF THE Z STATISTICPb S Z (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTIC N IF N > 10
P 1 1 29.00 . 96605 19 . 324I /
,
SEN SLOPECONFIDENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT1 1 010 -.035
:050 -.019039:039
.llO086
100 -.014 039 :082:200 -.002 :039 . 078
I-i i
8-4
i
m
i? !
r
0.5
0
RA228 ConcentrationsMonitoring Well AMM3
8/l 1187 12123188 517190 9119191 l/3 l/93 6/15/94 10128195 3/l l/97t
Dates Sampled
cci i
File: J:\FOX\GENSTAT\AMM3-URA.CSV:
_________________-_------------------------------------~------~~~-.--
F”I
PROB. OF EXCEEDING/ , THE ABSOLUTE'-VALUB
KENDALL OF THE Z STATISTICF S Zb (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTIC .N ". -'IF"‘N"> 10" "
E” 1 1 82.00 2.83902 19
i
SEN SLOPECONFIDENCE INTERVALS
,.
STATION SEASON ALPHA LOWER LIMIT SLOPEp 1 1 010
:050000
L : 16:66765.38565.3-85I
ioo
:20025.380 65.385
cr 33.769 65.385
. 004
UPPER LIMIT141.152112.500104.54596.913
1
m
: ”
88iEZfZ 1
L8/ 11180
m: It /
File: J:\FOX\GENSTAT\ATPlS_RG.CSV------_-----------------------------------------------------------
PROB. OF EXCEEDINGTHE ABSOLUTE VALUE
KENDALL OF THE Z STATISTICS Z (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTIC N IF N > 10
r? 1 1 -44.00 -.8d743 29 -420
SEN SLOPECONFIDENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMIT1 1
.010 -.088 -.019 043
:loo' 050 -.065 - YO53~ -.019 -.019:02502-o
P . 200 -.047 -.019 :on:
__. ,.
”
File:J:\FOX\GENSTAT\ATPlSwR6.CSVPrinted:12/31/97 Page: 1
: ~" ', ,' Linear RegressionMinimum Maximum Mean Median Std Dev Slope R-Sqr Num
0.00 7.30 1.52 1.20 1.44 -0.09141 0.03 29.
p .
F
fI 1
p (‘.:_.
”
f-
i a
c ‘
F””
t
File: J:\FbX\GENSTA‘f\ATPlS;UR.CSV-------------------------------------------~-------,---------------~--
fy
,,,IPROB. OF EXCEEDING
. _I MANN - THE P;BSOLU?E VALUEKENDALL OF THE Z STATISTIC
FI I ') s Z (TWO-TAILED TEST)k. j, ., STATION SEASON STATISTIC STATI$TIC Ia IF N > 10
1 1 -141.00," -2.77477 28 .006
SEN SLOPEr 7 'CONtiIDENCE“INTERVALSL ,"
STATION SEASON ALPHA LOWER LIMIT SLiJPE UPPER LIMITkl. 1 1 010 -.300
:050 -.278-.170 000-.170 -:045
. 100 -.265 -.170 -.065F .2po -.250 -.170 -.082
y
_1 .*
n \ ,. >_ , :. ‘ (. . .: ./
m: 1i j
16
T
Uranium ConcentrationMonitoring Well ATP-1-S
2118182 11114l84 8/11187, t. sl7190
Dates Sampled
1131193 10128195 7124198
File: J:\FOx\GENSTAT\ATPZS-MO.CSV_______-----------------------------------------------------------'
_. _. , PROB. OF EXCEEDING_. ITHE ABSOLUT'E -VAbl%
f-KENDALL OF THE Z STATISTIC
b S Z (TWO-TAILED TEST)c ‘. STATION SEASON S+ATI'STIC STATISTIC N IF N > 10
em
;1 1 46.00 1.8'5999 17 . 063
k;
SEN SLOPECONFIDENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPEr
UPPER LIMIT1 1 .OlO -.012 020 061
. 050 .ooo :020 :054
. 100 . 020Fk * .200 . Ol<O . 020 . 049i '
‘ ...
<
t
”L ;
mi :: ,i ;
z-l
: !
L6lL LIE
P6lS 119
li
1616 116
File: &\FOk\tiENSTAT\ATPZS-NI.CSV_i_____--_--------------------------------------------------------
PROB. OF EXCEEDINGTHE ABSOLUTE VALUE
KENDALL OF THE Z STATISTICS Z (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTIC N IF N > 10
lrrl 1 1 -610.00 -3.88537 60 . 000
dSEN SLOPE
CONFIDENCE INTERVALS
STATION SEASON ALPHA LOWER LIMIT SLoPE UPPER LIMIT1 1 010 '
:050-4.513 -2.739 -1.059-4.032 -2.739 -1.500
. 100 -3.765 -2.739 -1.719- .200 -3.500 -2.739 -2.055L
l-
c
File: J:\FOX\GENSTAT\ATP2S_RG.CSV____-_------------------------------------------------------------
PROB. OF EXCEEDINGTHti ABSbLtiE VALUE
STATION
STATION1
KENDALL OF THE Z STATISTICS Z (TWO-TAILED TEST)
SEASON STATISTIC STATISTIC N IF N > 10
1 -194.00 -2.09965 42 . 036
SEN SLOPECONFIDENCE INTERVALS
SEASON ALPHA LOWER LIMIT1 .OlO -.050
,050 -.040
:200 100 -.033 -.030
SLOPE-.020-.0207.020-.020
UPPER LIMIT002:ooo. 000
-.007
Printed:12/31/97 Page: 1File:J:\FO~\GEXSTAT\ATP2S_R6.CSV
Linear Regression
Minimum Maxinnlm Mean Median Std Dev Slope R-Sqr Num
0.00 lg.80 1.23 0.60 1.94 -0.05548 0.12 42
li
10
8
6
4
2
0
RA226 ConcentrationsMonitoring Well ATP-2-S
-- _. 1 -- -” ‘1
,
.---..- ---.--____-_
0- -_- -._
t -
-.---._____ _._______..._._____I___
2/18/82 lll14l84 8/11/87 WI90 1131193 1 O/28/95 7124198I
Dates Sampled
. ̂ ,.( I __., j ./.l,r”,
“1
File: J:\FOX\GENSTAT\ATPZS_R8.CSVn -------_----_-_---------------------------------------------------" 1L ,
MANN-PROB. OF-EXCEEDINGTHE ABSOLUTE VALUE
KENDALL OF THE Z STATISTICp S Z (TWO-TAILED TEST)
STATION SEASON STATISTIC STATISTIC N IF N > 10
631 1 1 26.00 1.03392 17 . 301I iI
SEN SLOPECtitiFI'D"titiC!ti' ItiTERVALS
STATION SEASON ALPHA LOWER LIMIT SLOPE1 1 010 -.071 037
:050 -.048 :037. 100 -.024 037. 200 -.009 :037
UPPER LIMIT158:121. 106.083
6-7j /
: :
Printed:12/31/97 Page: 1
Linear Regression
Minimum Maximum Mean Median Std Dev Slope R-Sqr Num0.00 3.00 1.97 2.00 0.73 0.05612 0.12 17
-3
L6ILLIB
P6l5 LIS
1616 116
06lLl9
1 88/EZIZ L0
File: J:\FOX\GENSTAT\ATPZS-UR.CSV----------------------------------------- -------------------------
I-i 1
MANi'PROB. OF EXCEEDINGTHE ABSOLUTE VALUE
KENDALL OF THE Z STATISTICm S Z (TWO-TAILED TEST)I ; STATION SEASON STATIQTIC SY!&TfST~IC N IF N > 10
P-4 1 1 -143.00 -1.38991 .165:
L7SEN SLOPE.CoN;F‘TBgHcE" .‘INTGRvAL's
STATION SEASON ALPHA LOWER LIMIT SLOPE UPPER LIMITrclll
,
1 1 . 010 -163.169 -56.696 57.143
:100 050 -141.301 -125.989 -56.696 -56.696 29.763 x2.158r* .200 -115.074 -56.696 -2.769:k;
“I:
h !. d
Printed:12/31/97 Page: 1File:J:\FOX\GENSTAT\ATP2S_UR.CSV
Linear Regression
Minimum Maximum Mean Median Std Dev Slope R-Sqr Num
0.00 9800.00 4999.33 5400.00 2717.74 -'32.59795 0.02 45b
86/m L
.-
96l8210 L,
*Ict
I. E6lLlll
I
i f
f-88
a
LSI 1 118
c”iI
P8lt II 11--
j
--Yh
8 Ii2 ,
*7
I :
APPENDIX B
h.
r-4I
: /
-
*6/Q 118P6/ 1Ell&6/Ll/LL&6/0Z/86612219C6/9Z/P86/9ZlZBG/ZZ/ 1ZG/BZ/ 11
zz ZG/ZZ/O 1Z6/6 l/8Z6fL l/L2611 LZ/QZ6lt l/PZGILZIZZ6/ I/ 116/9Z/ll16/8 L/O 1
0
0
1
O9
0
0
0’
0
D
e-0
a
2a88
I’ t8
0
-.CL,
8
c8
873
0
83
08
80
Ia I
dI
8op
8a
0
aC
0
r”4
8
a
‘L
;8
0
a8 0
C
P-
aC
8 00”
ry
0 1-
8>Ba53t
16lPZi816/QZ/L1 G/Et/Q16/OZ/P16ILZIZ16/Ql/l06/6Z/ 1061 IZIO06/L l/806/ZZ/Q06/9Zrt7
0
L1
06/L l/C06/oz/ 168/lZ/ll68/ 1 L/O 168/ZZ/868/8/f6816 L/Q68/Q LIP.68/ZZ/Z68/P I/ 1+
&Jtr6/QZfZtr6/5Zf 1
$M.
C6f LZIO 1 f=
86/OC/L %
B6ILZlQ66fZZfBE6fZZfZZ6fZZf Z 1Z6fL I/ 1126/61/6Z6fL L/8z6/QZ/9Z6/6 115Z6ILZIEZ6f LZIZ16/L t/Z 116/0Z/ 1116/zz/6 E1612218 2.16/Q l/Q d
16/81/Q g16/EZ/&
.-
;,W
Q)ILe
P
B
C
!
r16fCZfZ0616 LIZ 106/L If 11061 lZf6
.-
8’0
06/P l/806/ZZ/j306fQZft06/f LIE06fOW 1681 lZ/ 1168/l L/O 168/zz/868/8/L6816 If Q68/Q l/P
GRAPH B-3WATER LEVEL COMPARISON BETWEEN WELL AMM-3 AND COLORADO RIVER
3 9 6 3 . 0 0 7 0
n
xsECTA.XLS NM’tiB3.XLC 16/2@/941
: c.
I
%
3
0
I
3
A
0
J
‘9
I
0
0
0
e"
0
‘0
I
0
0
a
-. .+ . .
n:
lrwszfztrW5Zll8WlZlOL&6/OC/Lg&&&y&
E6fzZf EE6fzZfzZwzZ/ zZ6fL If 1ZWEW6Z6fL l/8Z6f QZ/Q
c8
J
-
t.
8
#
-0
I c
I
II
03
Is
c II
.cl
I0
1
o-I
m88 l 86l c
kI
I8
�La
8
tL-
00
l~/ZZ/6 .g16/ZZ/8 z16/Q L/Q 5
9’LW8WQ’ i 216/&Z/C 2
3*
coc! Io I
o--r!
8 Y!
16/&Z!Z0616 l/Z 106fLLf 11061 lZf6OWV 118owzw~OS/&Ytr
OWL LIE06fOW 168flWt 168/L l/O1
68/zZf8
68/8/L68f6LlQ
6819 LIP
68/ZZ/Z
mc f3. >
5fF”r
c
t ,
.3tr6fQZf Z -@
E6f6Zf6P-e
E6fOCflL p
4 BGILZI 9
66fZZfC
E6fZZf 1
~Z6/L1/11
Z6/ lZf6
Z6fL l/L
Z6/6 l/Q
Z6ILZl E
i5WQ,Irn
B.e
8
n*
m
: ! Iaxss00
sa
iii5mJ E>
z
1,0H
t
Z6ltr l/l
l6/9Zfll
16/Zz/6 5
16ILZIL 2
16/81lQ 8
iS/SZ/S
16&Z/ 1
06/f If 11
0616 l/6p?f. -
B$ 06/Z l/L.-,06/l I&
06/L 11 c:
06/0Z/ 1
68/lzfll
68/Z/6
68/8/L
68/V l/Q
6818 lf&
68/P If 1
t6/4Z/ZE6/6Z/6
BGIOEIL
EGILZIB2P ”.
- - EG/ZZ/l
- - 26/f l/l 1
- - Z6fL 1/L
- - ZGILZIB
- - Z6/Pl/l
- - 16/9Z/ 11
- - LGILZIL
- - 16/91/9
- - 16/81/E
- - I6/9Z/l
- - 06/Ll/ll
- - 06/6 l/6
06/z l/L
06/ 1 l/9
06/L l/C
061OZI 1
69f lZ/ 1
69/Z/6
69/9/f
1pn
:
69/t l/9
6919 l/E
69/t I/ 1
mI i
aP6/9Z/ZO
C6/62/60
86lOCllLO
EGILZI90
EGIZZIEO
CSIZZ! 10
Z6/Ll/ll
261 lZl60
Z6lL l/LO
Z616 l&O
. Z6lLZlCO
Z6ltrlIlO
16/9Z/ll
16/~z/60 3
l6ILZlfO 5
16/91/90 4a
16lCZlCO
16&Z/ 10
06/f l/l 1
0616 l/60
oyz l/f 0
06/l l/90
06/f l/80
06/OZ/ 10
68/lZ/ll
68lZO160
68/80/f 0
68/P l/90
6818 l/E0
68/t l/10
m..* ?:
nL ,.mib;.