Printer Friendly

A computer simulation and evaluation of groundwater resources in the Evangeline aquifer in the area of Kleberg County, Texas.

Abstract. -- A computer simulation of groundwater flow in the Evangeline Aquifer was conducted to determine future groundwater availability within a 5776 square mile (14,960 [km.sup.2]) area southwest of Corpus Christi, Texas. This aquifer is a major source of fresh water for the region, especially in the Kingsville and surrounding areas that rely on the aquifer for municipal, agricultural, industrial, and domestic use. A three-dimensional groundwater model simulating fluid flow within the study area shows maximum drawdown in the Kingsville area based on low to high projected pumping estimates. These results are very similar to an earlier 1985 USGS study of computed potentiometric surface for the area. Although the Kingsville Dome in situ leach uranium mine is currently in restoration phase of operation, the model shows, for the period 1988 to 2020, that groundwater withdrawals of 83.0 acre-ft/year (10.2 X [10.sup.3] [m.sup.3]/year or 3.24 X [10.sup.-3] [m.sup.3]/s) and 41.5 acre-ft/year (51.2 X [10.sup.3] [m.sup.3]/year or 1.62 X [10.sup.-3] [m.sup.3]/s) from uranium mining operations will contribute 5.1 ft (1.6 m) and 2.6 ft (0.8 m) of drawdown respectively to the potentiometric surface at the mine area. Thus, a future startup and extraction of groundwater for uranium operations at previous rates will not adversely affect the levels of the water table in the Kleberg County area. Additionally, a discrepancy with the results of a computed potentiometric surface in the 1985 USGS study for the low estimates of projected pumping may be due to errors in data input or excessive pumpage used in the computer simulation.

**********

The availability of freshwater from the Evangeline aquifer (Goliad Formation) in the Kingsville area (population 25,500), southwest of Corpus Christi, Texas is the focus of this hydrogeologic study involving computer simulation of groundwater flow (Figure 1). A previous study was conducted by Groschen (1985) for the United States Geological Survey in cooperation with the Coastal Bend Council of Governments. Groschen's study involved a computer simulation of groundwater flow and solute transport in the Evangeline area of approximately 5776 sq. miles (14,960 [km.sup.2]) and was conducted to determine the potential degradation of groundwater from cross-formational flow (leakage) during intensive pumping.

[FIGURE 1 OMITTED]

The present investigation compares the results of fluid flow simulations from Groschen's (1985) work and similar hydrogeologic data from the United States Geological Survey to data from a more recent simulation. A previous study of the geology and uranium mineralization of the Goliad Formation within the Kingsville Dome area of south Texas has been described by Arredondo (1991) and Arredondo & Thomann (1996). Investigations of the geohydrology, water chemistry, and numerical groundwater modeling of the regional Gulf Coast Aquifer (which includes the Evangeline) can be found in the 2001 Coastal Bend Regional Water Plan (CBRWP) and the 2002 Draft Report by Waterstone to the Texas Water Development Board (TWDB).

STUDY AREA

Hydrogeology of the Goliad Formation. -- The Goliad Formation is within the Evangeline Aquifer, a major fresh water source in the Texas Gulf Coast (Knape 1984; CBRWP 2001) supplying water for municipal, industrial, and agricultural use. The Evangeline Aquifer is one of four aquifer formations (Catahoula, Jasper, Evangeline, Chocot) of the Gulf Coast Aquifer System that extends from northern Mexico to Florida (CBRWP 2001). A climatic atlas published by the Texas Water Development Board shows the study area is subhumid to subarid, the average annual precipitation from 1951-1980 is 27 inches (69 cm) and the average gross lake surface evaporation rate for the period of 1950-1979 is 65 inches (165 cm) (Larkin & Bomar 1983).

The Evangeline aquifer is both a confined and an unconfined aquifer. The unconfined part of the aquifer is approximately 22 miles (35 km) west of the city of Kingsville where the Goliad Formation crops out and forms a belt of sediments which parallel the coastline (Figure 2). The confined portion of the aquifer exists below a low-relief surface to the southeast of where the Goliad crops out. Closer to the Texas coastline, the Goliad sands of the Evangeline aquifer are about 400 ft (122 m) thick, and lie at depths of approximately 500 to 2000 ft (152 to 610 m) below the ground surface.

Recharge occurs at the unconfined portion of the Evangeline aquifer in Jim Wells and the west-central area of Duval County where the Goliad Formation crops out. Recharge may also occur by vertical leakage during periods of high withdrawals as pressures in the confined portion of the Evangeline are reduced through intensive pumping to allow downward leakage from the overlying Chicot aquifer through leaky clays. Groschen's (1985) estimations of the effective recharge required to maintain the predevelopment hydraulic gradients in the Evangeline are 0.23 percent (0.06 in. per year or 0.15 cm per year) of the average annual precipitation of 26 inches (66.0 cm) for the study area.

[FIGURE 2 OMITTED]

Prior to development of the groundwater resources, groundwater flowed in a southeast direction towards the Texas coastline. However, extensive pumping of groundwater has altered the direction of the groundwater flowpaths with the formation of a cone of depression at the center of the study area. Rettman (1983) measured the water levels in the Evangeline in the Kingsville area during 1982 and found a large cone of depression in the potentiometric surface which occurs below Kingsville. This cone of depression formed as a result of high water withdrawals for municipal and industrial usage. Prior to 1907, and perhaps as recently as 1928, water wells in the Goliad were flowing artesian in western and southern Kleberg County (Livingston & Bridges 1936; J. Arredondo, pers. comm.). In 1933, hydraulic pressure in the Evangeline aquifer had decreased such that only wells in west-central Kleberg County (west of Riviera and south of the city of Kingsville) were flowing artesian. Inspection of maps of the 1982 potentiometric surface prepared by Rettman (1983) shows that the capacity for water in the Evangeline aquifer to flow freely above the surface is greatest in east Kenedy and Kleberg counties along the shores of the Laguna Madre. All remaining water wells in Kleberg County west of Laguna Madre are non-flowing artesian.

Hydrochemistry. -- The aquifer sands within the Goliad contain interstitial water of the meteoric flow regime. Chemical analysis of major constituents in Goliad Formation groundwater from the Texas Water Development Board data base shows that groundwaters sampled from the period of 1959 to 1984 are fresh to brackish and average 997 milligrams per liter of total dissolved solids. Chloride concentrations in the Evangeline Aquifer increase in an easterly direction with progressively greater depth of the aquifer (Shafer & Baker 1973; CBRWP 2001). Higher concentrations of chloride are attributed to zones of fault-related discharge, and typically increase basinward where marine sediments with saline connate water are dominant.

METHODS AND MATERIALS

Hydrogeological analysis. -- The specific objectives of this study were to: (1) simulate the potentiometric surfaces computed by Groschen (1985); (2) determine the effects of groundwater withdrawals from uranium mining operations on the potentiometric surface of the Evangeline aquifer; (3) test several boundary conditions used by Groschen (1985) in a 38-row by 38-column grid model and determine the appropriate use of the boundary conditions; and (4) determine if there may be any significant dewatering of the Evangeline Aquifer due to in situ leach mining operations at the Kingsville Dome plant (currently in restoration phase).

Calibration of the fluid flow model in this study utilizes Groschen's (1985) data for steady-state and transient-state flow simulations for the years 1901 through 1982. The potentiometric surface as defined by Groschen (1985) is the elevation at which groundwater levels would stand in tightly cased wells. Hydraulic heads were mapped and potentiometric surfaces were contoured with SURFER Version 4.00 (Golden Software Inc. 1989) contouring and mapping software. This surface was mapped with the Kriging gridding technique at contour intervals of 25 ft (7.6 m) to allow comparison with Groschen's (1985) results. The 25 foot interval is a commonly used contour interval of most hydrogeologic maps in this study area.

Data collection. -- Data from the Texas Water Development Board's data bank of observation wells in Kleberg, Jim Wells, Nueces, Kenedy, and Brooks counties, the U.S. Geological Survey, Texas state reports, and King Ranch archives were utilized for the verification of aquifer heads. Historical information was gathered from discussions with area scientists and local citizens who were knowledgeable of the history of groundwater use.

Data analysis -- A quasi three-dimensional finite difference model was used to simulate recharge, hydraulic conductivity, storativity, multiple pumping wells, aquifer sand thickness, aquifer heads, and leakage from an overlying aquifer. Design of the aquifer model was similar to Groschen's (1985) U.S.G.S. study of the Evangeline aquifer. Model simulations were performed to determine the sensitivity and accuracy of head calculation with respect to boundary conditions designnated as no-flow in Groschen's (1985) model. The methods and techniques which are used to calibrate an aquifer model are not presented because the aquifer model was calibrated using Groschen's (1985) data. An analysis was undertaken to determine the sensitivity of the calibrated model to uncertainties in the data, of the aquifer properties, and the assumptions of the model. This sensitivity analysis was employed during the steady-state and transient-state simulations to determine the appropriate use of the no-flow boundary conditions by the United States Geological Survey in a groundwater simulation (Franke et al. 1987). The elevations of the potentiometric surfaces of the one-layer three-dimensional fluid flow aquifer simulations were compared to those results by Groschen (1985).

Simulation of the Evangeline Aquifer. -- The fluid flow for the Evangeline Aquifer in the study area was modeled to examine the following: (1) the transient-state effects on the potentiometric surface for the years 1983-2020; (2) the effects that potential pumping from uranium mining operations at the Kingsville Dome would have on the groundwater resources in the future; and (3) the results of a sensitivity analysis on Groschen's (1985), no-flow boundary conditions on the east and south sides of the modeled area.

Groschen's (1985) simulation strategy was to accomplish the following: (1) identify a period when the aquifer was in equilibrium, and use aquifer heads as the initial aquifer potentiometric surface; (2) perform a steady-state simulation to adjust hydrologic input and parameters until computed heads matched the field heads for the period of equilibrium; (3) use the steady-state computed heads as the initial potentiometric surface; and (4) model all stresses which had occurred on the aquifer, as well as project stresses to a future period.

Conceptual model. -- The Evangeline and the Chicot aquifers were modeled using Groschen's (1985) conceptual model which describes the actual physical boundaries of the aquifer flow system (Figure 3). The conceptual model consists of the confined and unconfined Evangeline aquifer flow system which exhibits heterogeneous and anisotropic conditions. The Evangeline is confined where the Beaumont and Lissie Formations overlie the Goliad Formation, and is unconfined aquifer in the western part of the study area where the sands of the Goliad Formation crop out at the surface. The aquifer dips to the southeast below the Gulf of Mexico where the freshwater/saltwater interface occurs several miles east of Padre Island. The Evangeline aquifer also extends beyond the southern borders of the modeled area towards the Rio Grande River, and to the north beyond the study area.

The Chicot aquifer is an overlying unit which is included within the Beaumont clays and the low permeability Lissie sands (C. Bartels pers. comm.). The Chicot aquifer has permeable sandy units that are approximately 30 to 100 ft (9.1 to 30.5 m) thick, and clay lenses of low permeability which confine the aquifer.

[FIGURE 3 OMITTED]

Mathematical model. -- The mathematical model is a numerical version of the conceptual model and usually much simpler in design than an exact model of the aquifer. The numerical groundwater model was developed with MODFLOW--A Modular Three-Dimensional Finite-Difference Groundwater Flow Model (Version 1634) by MacDonald & Harbaugh (1988). The flow model calculates hydraulic heads, and determines groundwater flow in a quasi three dimensional space using the block centered finite difference approach. An iterative solution technique using the Strongly Implicit Procedure (SIP) was used to solve the finite-difference equations. A three dimensional simulation was performed for a one-layer by 38-row by 38-column grid. The model was specified for constant head and no-flow boundary conditions, horizontal and vertical hydraulic conductivities, aquifer thicknesses, elevations of aquifer layers, pumping cells, storativities for each block cell, and source terms for discharge, recharge, and wells. Groschen's (1985) simulation grid for the discrete-modeled conceptual aquifer system was applied for the study area (Figure 4). The boundary conditions of the aquifer are those assigned by Groschen (1985) in which the west boundary was modeled with a constant head to simulate recharge from infiltration of surface water. The north model boundary was designated as a no-flow boundary because it corresponds with the regional aquifer discharge area at the Nueces River and Nueces Bay. The southern and eastern boundaries were designated as no-flow boundaries. The eastern no-flow boundary was designated as a fixed stream surface (no-flow) boundary because the freshwater/saltwater interface was considered fixed under stress. The southern boundary was designated as a no-flow boundary but, according to Groschen (1985), is not valid for the transient-state simulation.

[FIGURE 4 OMITTED]

The Lissie sand (of the Chicot aquifer) acts as a confined aquifer in the URI mine area, but very little regional hydrologic data for the Chicot is available. Therefore, a Cauchy boundary condition (mixed boundary) (Franke et al. 1987) was used to simulate a source from the Chicot aquifer and provide head-dependent leakance across the overlying Chicot aquifer into the Evangeline.

Steady-state flow simulation. -- The steady state simulation was performed to simulate the initial potentiometric surface, and the steady-state model is calibrated to Groschen's 1901-1982 data. This initial surface was computed by using Groschen's (1985) calibrated initial aquifer head and confining layer head arrays for the steady-state condition. Groschen's (1985) steady-state simulation required adjustments of input data and hydrologic parameters for the model calibration. The water balance was used to determine when the aquifer system was in a steady-state flow condition. Groschen (1985) used the computed steady-state heads as initial heads for the transient-state simulations. The amount of water which flows through Groschen's (1985) predevelopment aquifer model is 7.0 [ft.sup.3]/s (0.20 [m.sup.3]/s), whereas the amount of water which flows through Arredondo's (1991) predevelopment aquifer model using MODFLOW is 5.7 [ft.sup.3]/s (0.16 [m.sup.3]/s). The volume of water which flows into and out of Arredondo's (1991) pre-adjusted steady-state aquifer is 9.0 X [10.sup.9] [ft.sup.3] (2.74 X [10.sup.9] [m.sup.3]).

Transient-state flow simulation. -- Transient-state flow simulations were conducted for the periods of 1901 to 2020. The steady-state head configuration was used as the initial conditions for the transient flow simulations. Calibrated hydrologic data used for the computer model is from Groschen (1985). During the calibration phase, Groschen (1985) matched the computed heads to Rettman's (1983) observed field heads to an error [+ or -] 40 ft (12.2 m). The rates of withdrawal from pumpage, as well as low and high pumping rates, simulated in this report are those from Groschen (1985). Withdrawals of 90.4 [ft.sup.3]/s (2.6 [m.sup.3]/s) for the entire model were used during the calibration period. The low estimate of withdrawal for the projected pumping period of 1983-2020 was 127.9 [ft.sup.3]/s (3.6 [m.sup.3]/s), and the high estimate of withdrawal for the projected pumping period of 1983-2020 was 223.7 [ft.sup.3]/s (6.3 [m.sup.3]/s). Computer simulations were also conducted to determine the effects from the mining uranium on the potentiometric surface for high withdrawals at 9625.0 [ft.sup.3]/day (272.5 [m.sup.3]/day), and for low withdrawals at 4812.5 [ft.sup.3]/day (136.3 [m.sup.3]/day).

Sensitivity analysis of the aquifer model. -- A sensitivity analysis was conducted to determine the response of the simulation model to the use of no-flow boundary conditions in the eastern and southern borders. Both the eastern and southern boundaries were selected because as Groschen (1985) stated, there were no corresponding physical boundaries. The no-flow boundaries assigned by Groschen (1985) in his 38-row by 38-column grid at the eastern and southern edges were tested by expanding the model grid to 48 rows by 48 columns for steady-state and transient-state simulations. The modeled area was expanded to include 10 cells (20 miles) to the east and 10 cells to the south. Input data for the aquifer cells in rows 39 to 48 and cells in columns 39 to 48 were identical to that of row 38 and column 38 respectively. Simulations were conducted for Groschen's (1985) calibration period of 1901-1982, and the period of 1983-2020 with the high estimate of projected pumping. Contours of the potentiometric surfaces from simulations of the 38-row by 38-column grid were compared to those of the 48-row by 48-column grid. The differences in the potentiometric surfaces of the 38-row by 38-column grids and the 48-row by 48-column grids were calculated and mapped as the difference in drawdown, and as the percent difference of the total drawdown.

RESULTS OF SIMULATIONS

Results of the simulation with the data for the calibration period (1901 to 1982) in Figure 4 using 48 by 48 simulation cells are similar to Groschen's (1985). Table 1 summarizes the modeled withdraw from the Evangeline Aquifer, Table 2 summarizes the modeled inflow into the Evangeline Aquifer, and Table 3 lists the summary and results of the computer simulations and sensitivity analysis. Maximum drawdown occurs in the area surrounding the City of Kingsville, where the elevation of heads range from -184 to -150 ft (-56.0 to -45.7 m). Comparison of the computed potentiometric surface to Groschen's (1985) results, simulated with the use of Konikow & Bredehoeft's (1987) Method of Characteristics (MOC) simulation model, reveals no significant differences with the exception of increased drawdown at the eastern no-flow boundary (the zero elevation contour is closer to the eastern no-flow boundary). Differences in elevations between these results and Groschen's (1985) are approximately 4 ft (1.2 m) at the eastern boundary.

Comparison of the computed potentiometric surface with the low estimates of projected (1983 to 2020) pumping for Groschen's (1985) computed potentiometric surface reveals that aquifer simulations using MODFLOW are locally more sensitive to reductions of the pumping rates than Groschen's (1985) computed surface with MOC. The aquifer model using MODFLOW responded to the reduced rates of withdrawal by computing a potentiometric surface which has an elevation of approximately -47.8 ft (-14.6 m) for the Kingsville area. The elevation of Groschen's (1985) computed potentiometric surface presented for the Kingsville area is approximately -250 ft (-76.2 m). Differences in heads of the potentiometric surface with the low estimates of projected pumping in this study are approximately 202 ft (61.6 m) higher than Groschen's (1985) computed potentiometric surface with MOC.

Results of the computer simulation of pumping of the aquifer for the calibration period (1901 to 1982), and for the 1983 to 2020 simulation using high estimates of projected pumping are shown in Figure 5. Maximum drawdown occurs in the City of Kingsville area where the elevation of heads range from -469 to -400 ft (-142.9 to -121.9 m). These results are very similar to Groschen's (1985) computed potentiometric surface.

The Kingsville Dome uranium in situ leach mine is currently in a restoration phase but the source of the low and high pumping rates used in this paper are projected from historical pumping rates. The simulation of pumping of groundwater from the production within the Kingsville Dome uranium mine projected for the period 1988-2020 decreases the elevation of the potentiometric surface by 2.6 ft (0.8 m). Comparisons of the computed potentiometric surface from the high estimate that does not simulate mining with the computed potentiometric surface which does simulate mining do not reveal significant differences in the elevations of those potentiometric surfaces respectively.

[FIGURE 5 OMITTED]

Results of the 48-row by 48-column grid sensitivity analysis for the calibration period are different simulated heads than results with the 38-row by 38-column grid. Generally the elevations of the potentiometric surfaces for the 38-row by 38-column grid are lower than those of the 48-row by 48-column grid respectively. The difference in the elevations of the potentiometric surfaces of the 38-row by 38-column grid and the 48-row by 48-column grid is 13 ft (4.0 m) of drawdown. This difference is attributed to boundary effects from the southern no-flow boundary and the additional volume of water in storage. The contour line representing 1 foot (0.3 m) elevation is approximately 7 to 10 miles (11.3 km to 16.1 km) south of Kingsville. The eastern no-flow boundary does not significantly influence the elevation of the potentiometric surface throughout the eastern portion of the study area.

Results from the sensitivity analysis with the 48-row by 48-column grid for the period of 1983-2020 with the high estimates of projected pumping also show differences in the simulated heads in comparison to results from the 38-row by 38-column grid (Figure 6). The greatest difference in elevation of the potentiometric surfaces of the 38-row by 38-column grid and the 48-row by 48-column grid is approximately 30 ft (9.1 m) of drawdown which occurs at the southern no-flow boundary. The 1 ft (0.3 m) contour line remains approximately 7 to 10 miles (11.3 km to 16.1 km) south of Kingsville but lies closer to the high pumping center west of Kingsville. The eastern no-flow boundary does not greatly influence the elevation of the potentiometric surface in the eastern portion of the study area. Boundary effects from the eastern no-flow boundary range from 1 ft (0.3 m) to approximately 3.7 ft (1.1 m) of drawdown at the eastern edge. The percent of the differences in drawdown of the 38-row by 38-column the 48-row by 48-column grid potentiometric surfaces to the total drawdown of the 48-row by 48-column grid potentiometric surface for the calibration period is presented in Figure 7. The sensitivity analysis supports Groschen's (1985) report which states the eastern boundary can be designated as a no-flow boundary and will not contribute significant boundary effects. The analysis also supports Groschen's (1985) conclusion that the southern boundary is not valid under transient-state conditions. As Groschen (1985) had determined in his sensitivity analysis for two different boundary conditions, the computed potentiometric surface is not significantly different in the area by the designation of two different grid sizes.

[FIGURE 6 OMITTED]

The simulations were not an attempt to match the exact levels of the potentiometric surface of the Evangeline aquifer, but rather to obtain the range of effects that pumping would have on the potentiometric surface of the aquifer. Based on the elevation of the top of the aquifer sands of the Goliad Formation, model results indicate that de-watering of the aquifer will not occur in the Kingsville area until the elevation of the potentiometric surface is less than approximately 580 ft (176.8 m) below sea level (Arredondo 1991).

[FIGURE 7 OMITTED]

CONCLUSIONS

Groundwater steady-state and transient-state computer simulations of the Evangeline aquifer using MODFLOW, a 3-dimensional groundwater model yields several results. A fluid flow simulation for the calibration period of 1901-1982 shows that elevations of heads for the Kingsville range from approximately -184 to -150 ft (-56.0 to -45.7 m) below sea level. A simulation with the low estimates of projected pumping for the period of 1983-2020 yields heads with elevations of approximately -47.8 to -25 ft (-14.6 to -7.6 m) for the Kingsville area, and is 202 ft (61.6 m) higher in elevation than the value computed by Groschen (1985). A simulation with the high estimates of projected pumping for the period 1983-2020 yields heads with elevations which range from -469 to -400 ft (-142.9 to -121.9 m) for the Kingsville area.

The Kingsville Dome in situ leach uranium mine has been operating in a restoration phase since 1999 (URI company news release, November 16, 1998), but the following conclusions based on computer simulations, structural geology, and water-table elevations (Arredondo 1991) can be drawn on the hydrogeology of the Kingsville area should mine operations start up again in the near future. Withdrawals of 41.5 acre-ft/year (51.2 by [10.sup.3] [m.sup.3]/year or 1.62 by [10.sup.-3] [m.sup.3]/s) from mining for the period of 1983-2020 with the high estimates of projected pumping will contribute to 2.6 ft (0.8 m) of additional drawdown at the mine area. Withdrawals of 83 acre-ft/year (102.4 by [10.sup.3] [m.sup.3]/year or 3.24 by [10.sup.-3] [m.sup.3]/s) from uranium mining operations for the period of 1988-2020 will contribute to 5.1 ft (1.6 m) of additional drawdown of the computed potentiometric surface in the mine area. The high pumping estimates of mining operations will contribute 0.12 percent to the total simulated withdrawals of groundwater. Thus, groundwater withdrawals for uranium operations, if continued at simulated rates, would not adversely affect the levels of the water table in the Kleberg County area.

A sensitivity analysis of the simulation model examining the effects of the southern and eastern no-flow boundaries on the 38-row by 38-column grid shows that 13 ft (4.0 m) of drawdown is expected from the southern no-flow boundary for the calibration period of 1901-1982. Less than 1 ft (0.3 m) of drawdown is attributed to the eastern no-flow boundary on the 38-row by 38-column grid for the calibration period. Approximately 30 ft (9.1 m) of drawdown is expected at the southern no-flow boundary for the period 1983 to 2020 with the high estimates of projected pumping. Projected drawdown at the eastern no-flow boundary on the 38-row by 38-column grid for the same period is approximately 3.7 ft (1.1 m).

De-watering of the aquifer would occur in the Kingsville area when the elevation of the potentiometric surface attains an elevation of less than -580 feet (-176.8 m). However, historical declines in the Kingsville area have ceased with water levels rising due to the city's increasing use of surface water from nearby reservoirs owned by the City of Corpus Christi (Groundwater Conservation District Operations Manual 1999). The discrepancy with the results of a computed potentiometric surface in the U.S.G.S. study (Groschen, 1985) for the low estimates of projected pumping could be due to Groschen's use of excessive pumping rates that exceeded his published low pumping rates. This discrepancy might also be due to incorrect data from original sources or incorrect input into the computer simulation.

The authors make the following recommendations on a future study of groundwater resources in the Kingsville area: (1) audit the pumping rates for several municipalities as well as all other entities with major water use; (2) record the measurements of potentiometric surfaces, and (3) run another computer simulation for comparison with current and past models.

ACKNOWLEDGMENTS

The authors thank Dr. Jon A. Baskin and Dr. Michael A. Jordan, Texas A & M University in Kingsville, Texas, and Dr. David Turner, Southwest Research Institute, San Antonio, Texas for their critical reviews and constructive criticism of this paper. Thanks also to Jose Arredondo, an employee of many years at the King Ranch in South Texas who is knowledgeable on the history of water wells in the region, and Craig Bartels who was the Plant Manager for Uranium Resources Inc. at the Kingsville Dome Project.

LITERATURE CITED

Arredondo, A. G. 1991. Geology and hydrogeology of the Kingsville dome in situ leach uranium mine, Kleberg County, Texas. Unpublished M.S. thesis, Texas A & I Univ., Kingsville, Texas, 131 pp.

Arredondo, A. G. & W. F. Thomann. 1996. Uranium mineralization in the Goliad Formation of the Kingsville dome in situ leach uranium mine in Kleberg County, Texas. Texas J. of Sci., 48(4):283-296.

Coastal Bend Regional Water Plan. 2001. Section 1 -- Description of the region, 20 pp.

Coastal Bend Regional Water Plan. 2001. Section 3 -- Evaluation of current water supplies in the region, 28 pp.

Franke, O. L., T. E. Reilly & G. D. Bennett. 1987. Definition of boundary and initial conditions in the analysis of saturated ground-water flow systems, an introduction. U.S. Geol. Survey TWI Book 3, Chap. B5, 15 pp.

Golden Software Inc., 1989, SURFER, Ver. 4.00.

Groundwater Conservation District Operations Manual. 1999. Edited by the Texas Alliance of Groundwater Districts, Chairman--Richard S. Bowers, 592 pp.

Groschen, G. E. 1985. Simulated effects of projected pumping on the availability of freshwater in the Evangeline aquifer in an area southwest of Corpus Christi, Texas. U.S. Geol. Survey Water Res. Inv. Rep., 85-4182, 103 pp.

Konikow, L. F. & J. D. Bredohoeft. 1978. Computer model of two-dimensional solute transport and dispersion in ground water: U.S. Geol. Survey Techniques of Water--Resources Inv., Book 7, ch. C2, 90 pp.

Knape, B. K. 1984. Underground injection operations in Texas: Texas Dept. of Water Res., Rep. 291, (2)7.

Larkin, T. J. & G.W. Bomar. 1983. Climatic Atlas of Texas: Texas Dept. of Water Res. LP-192, 151 pp.

Livingston, P. & T. W. Bridges. 1936. Ground-water resources of Kleberg County, Texas: U.S. Geol. Survey Water-Supply Paper 773-D, 197-232.

MacDonald, M. G. & A. W. Harbaugh. 1988. A modular three-dimensional finite-difference ground-water flow model. U.S. Geol. Survey TWI 6-A1, 586 pp.

Rettman, P. L. 1983. Water levels and salinities of water within the Evangeline aquifer in an area southwest of Corpus Christi, Texas: U. S. Geol. Survey Open-File Rep., 82-174, 26 pp.

Shafer, G. H. & E. T. Baker. 1973. Ground-water resources of Kleberg, Kenedy, and southern Jim Wells counties, Texas: Texas Water Development Board, Rep. 173, 166 pp.

Uranium Resources Incorporated (URI) company news release, Dallas, Texas, November 16, 1998.

Waterstone Environmental Hydrology and Engineering, Inc. 2002. Groundwater availability of the central Gulf Coast Aquifer: numerical simulations to 2050, central Gulf Coast, Texas draft report to the Texas Water Development Board, 63 pp. and appendices.

AGA at: alonzo@austintx.com

Alonzo Galvan Arredondo and William F. Thomann

Texas Commission on Environmental Quality, Austin, Texas 78754 and Department of Environmental Science, University of the Incarnate Word, C.B. 311, 4301 Broadway, San Antonio, Texas 78209
Table 1. Summary of modeled water withdraws from the Evangeline Aquifer.

 Mine
 Production Agricultural,
 Municipal Uses Industrial, and Leakage
 [ft.sup.3]/s [ft.sup.3]/s [ft.sup.3]/s
Period ([m.sup.3]/s) ([m.sup.3]/s) ([m.sup.3]/s)

1901-1982 0.0 13.24 (0.37) 1.43 (0.04)

Low Pumping Estimates of Projected Mining
1901-1990 <0.01 15.18 (0.43) 1.31 (0.04)
1901-2000 <0.01 18.06 (0.51) 1.18 (0.03)
1901-2010 0.01 (0.0003) 21.12 (0.60) 1.07 (0.03)
1901-2020 0.02 (0.0004) 24.36 (0.69) 0.98 (0.03)

High Pumping Estimates of Projected Mining
1901-1990 <0.01 15.18 (0.43) 1.31 (0.04)
1901-2000 0.01 (0.0004) 18.06 (0.51) 1.18 (0.03)
1901-2010 0.02 (0.0007) 21.12 (0.60) 1.07 (0.03)
1901-2020 0.03 (0.0009) 24.36 (0.69) 0.98 (0.03)

 Storage Total Withdraw
 [ft.sup.3]/s [ft.sup.3]/s
Period ([m.sup.3]/s) ([m.sup.3]/s)

1901-1982 2.08 (0.06) 16.76 (0.47)

Low Pumping Estimates of Projected Mining
1901-1990 1.93 (0.05) 18.42 (0.52)
1901-2000 1.75 (0.05) 21.00 (0.60)
1901-2010 1.60 (0.04) 23.81 (0.67)
1901-2020 1.49 (0.04) 26.85 (0.76)

High Pumping Estimates of Projected Mining
1901-1990 1.93 (0.05) 18.42 (0.52)
1901-2000 1.75 (0.05) 21.01 (0.60)
1901-2010 1.60 (0.04) 23.82 (0.68)
1901-2020 1.49 (0.04) 26.87 (0.76)

The rates for the agricultural, industrial, and municipal uses are from
Groschen (1985). The pumping projections for the agricultural,
industrial, and municipal uses are the high estimates of the Texas Water
Development Board and the Bureau of Reclamation as reported by Groschen
(1985). Low mine projection estimate is 5.57 X [10.sup.-2] [ft.sup.3]/s
(1.58 X [10.sup.-3] [m.sup.3]/s) and high estimate is
11.14 X [10.sup.-2] [ft.sup.3]/s (3.16 X [10.sup.-3] [m.sup.3]/s).

Table 2. Summary of modeled water inflow into the Evangeline Aquifer.

 Mine Flow From
 Production Leakage Infiltration Storage
 [ft.sup.3]/s [ft.sup.3]/s [ft.sup.3]/s [ft.sup.3]/s
Period ([m.sup.3]/s) ([m.sup.3]/s) ([m.sup.3]/s) ([m.sup.3]/s)

Calibration Period
1901-1982 0.0 7.26 (0.21) 6.80 (0.19) 2.70 (0.08)

Low Pumping Estimates of Projected Mining
1901-2000 <0.01 10.30 (0.29) 6.89 (0.20) 3.82 (0.11)
1901-2010 0.01 (0.0003) 12.31 (0.35) 6.97 (0.20) 4.53 (0.13)
1901-2020 0.02 (0.0004) 14.56 (0.41) 7.07 (0.20) 5.24 (0.15)

High Pumping Estimates of Projected Mining
1901-2000 0.01 (0.0004) 10.30 (0.29) 6.89 (0.20) 3.82 (0.11)
1901-2010 0.02 (0.0007) 12.32 (0.35) 6.97 (0.20) 3.82 (0.11)
1901-2020 0.03 (0.0009) 14.56 (0.41) 7.07 (0.20) 5.24 (0.15)

 Total Inflow
 [ft.sup.3]/s
Period ([m.sup.3]/s)

Calibration Period
1901-1982 16.76 (0.48)

Low Pumping Estimates of Projected Mining
1901-2000 21.00 (0.60)
1901-2010 23.81 (0.67)
1901-2020 26.85 (0.76)

High Pumping Estimates of Projected Mining
1901-2000 21.01 (0.60)
1901-2010 23.82 (0.68)
1901-2020 26.87 (0.76)

Constant head boundaries are used for flow from infiltration. Low mine
projection estimate is 5.57 X [10.sup.-2] [ft.sup.3]/s
(1.58 X [10.sup.-3] [m.sup.3]/s) and high estimate is
11.14 X [10.sup.-2] [ft.sup.3]/s.

Table 3. Summary and results of computer simulations and sensitivity
analysis for the study area.

1. Simulation of calibration period 1901-1982.
 Simulation for this period shows similar results as Groschen's,
 (1985) computed potentiometric surface. The contour of zero
 altitude differs with that of Groschen, (1985). Drawdown could be
 the effect of the eastern no-flow boundary. Heads differ in the
 City of Kingsville area from approximately -184 to -150 feet
 (-56.0 to -45.7 meters) below sea level.
2. Simulation of low estimates of projected pumping 1983-2020.
 Simulation for this period did not replicate Groschen's, (1985)
 computed potentiometric surface. Differences in altitude of the
 potentiometric surface at the City of Kingsville are 202 feet.
 Altitudes in this simulation are approximately -47.8 to -25 feet
 (-14.6 to -7.6 meters) below sea level at the City of Kingsville.
 This simulation shows that Groschen's published withdrawals for
 this period are not sufficient to replicate his reported results.
3. Simulation of high estimates of projected pumping 1983-2020.
 Simulation for this period yields similar results as Groschen's
 (1985) computed potentiometric surface. Potentiometric surface at
 the City of Kingsville ranges from -469 to -400 feet (-142.9 to
 -121.9 meters) below sea level.
4. Simulation of high estimates of projected pumping with low mine
 production 1983-2020.
 Simulation for this period which includes a low estimate
 (4812 [ft.sup.3]/day or 0.0557 [ft.sup.3]/s) of mining production
 did not produce significant differences from the simulation of high
 estimates of projected pumping without mining production
 (simulation #3). The calculated head at the mine area is 2.6 feet
 (0.8 meter) lower in altitude.
5. Simulation of high estimates of projected pumping with high mine
 production 1983-2020.
 Simulation for this period with a high estimate (9624 [ft.sup.3]/
 day or 0.1114 [ft.sup.3]/s) of mining production did not produce
 significant differences from the simulation of high estimates of
 projected pumping without mining production (simulation #3). The
 calculated head at the mine area is 5.1 feet (1.6 meter) lower in
 altitude.
6. Simulation of calibration period 1901-1982 with simulation grid
 expanded to 48 by 48 grid.
 Simulation for this period produced differences in altitude of
 the potentiometric surfaces of the 38 by 38 and 48 by 48 grids of
 approximately 13 feet (4.0 meters), at the southern boundary, to 1
 foot (0.3 meter) 7 to 10 miles (11.3 to 16.1 km) south of the City
 of Kingsville. Maximum differences in altitude at the eastern no-
 flow boundary are approximately 0.9 feet (0.3 meter).
7. Simulation of high estimates of projected pumping 1983-2020 with
 simulation grid expanded to 48 by 48 grid.
 Simulation for this period produced a potentiometric surfaces for
 the 48 by 48 grid at the southern no-flow boundary with a
 difference of 30 feet (9.1 meter), 1 foot (0.3 m) at the southern
 edge approximately 7 to 10 miles (11.3 to 16.1 km) south of the
 City of Kingsville. Difference in altitude at the eastern no-flow
 boundary is approximately 3.7 feet (1.1 meters).
COPYRIGHT 2005 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2005 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Arredondo, Alonzo Galvan; Thomann, William F.
Publication:The Texas Journal of Science
Geographic Code:1U7TX
Date:May 1, 2005
Words:6281
Previous Article:Coelomic helminths in the speckled racer, Drymobius margaritiferus (serpentes: colubridae) from Central America.
Next Article:Structure of historic vegetation on Kerr Wildlife Management Area, Kerr County, Texas.
Topics:


Related Articles
Relict seawater as source of stratified groundwater in glaciated estuarine valleys: an example from Fredericton Junction, New Brunswick.
More resources pumped into groundwater research.
Secret to northern India's vanishing water unlocked by scientists.
Chemical peculiarities of the Silurian-Ordovician and Cambrian-Vendian aquifer systems in Estonia: an overview of hydrochemical studies/...
Securing food supply for future generations through groundwater management: a policy analysis approach.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters