# 184W101

Primary Investigator: Keith Halford

## Well Data

USGS Site ID |
Local Name | Altitude | Uppermost Opening |
Lowermost Opening |
Primary Aquifer | Transmissivity (ft2/d) |
---|---|---|---|---|---|---|

383933114190501 | 184W101 | 6214 | 777 | 1729 | CARBONATE ROCKS | 11000 |

## Aquifer Test

All Aquifer Test Files (zip)

### Spring Valley

Aquifer Test (pdf) || Groundwater levels (NWIS)

A multiple-well aquifer test was conducted by Southern Nevada Water Authority (SNWA) in southeastern Spring Valley, HA184, near Great Basin National Park to estimate the hydraulic properties of the carbonate-rock aquifer (184W101; Figure 1). Well 184W101 was pumped for 72 hours at 2,520 gpm between April 9 and 12, 2007. Results from the well 184W101 aquifer test were reinterpreted to investigate the effects of induced flow in the observation well on hydraulic property estimates. These estimates will constrain calibration of regional ground-water flow models that encompass Spring Valley.

## Site and Geology

The aquifer test occurred in southeastern Spring Valley where groundwater development has been proposed (Figure 1). Fractured limestone was encountered primarily from land surface to more than 1,800 ft below land surface (Prieur and others, 2009). A few stringers of clay exist that were less than 20 ft thick. More than 1,300 ft of saturated carbonate-rock aquifer were observed because the unpumped depth to water was about 480 ft. The carbonate-rock was interpreted as a homogeneous, vertically anisotropic aquifer with a saturated thickness of 2,000 ft. A finite thickness was assigned to the carbonate-rock for interpretation because the actual thickness is unknown (Welch and others, 2007).

Figure 1. Location of wells 184W101 and 184W502M in Spring Valley, Nevada.

Observation well 184W502M was 175 ft north of the pumping well, 184W101 (Table 1, Figure 1) and was completed with 8.625 in. diameter screens between 481 and 1,780 ft below land surface (Figure 2). The screen was in a 14.75 in. diameter open hole that extended from the water table to 1,820 ft below land surface with no fill in the annular space.

Figure 2. Radial cross-section about pumping well 184W101.

Table 1. Well location and construction data for pumping and observation wells.

## Water Levels, Drawdowns, and Temperatures

Water levels were measured more than 4,300 times in each well during the three-day aquifer test (Prieur and others, 2009). Water levels in wells 184W101 and 184W502M were 486 and 480 feet below land surface, respectively, prior to pumping.

Drawdowns were estimated by subtracting static water levels from water levels after pumping began April 9, 2007 at 0900. The number of drawdown observations was reduced to less than 30 in each well by averaging in sub-periods (Figure 3). Sub-periods were of variable duration so observations were near equally spaced on a logarithmic time scale.

Figure 3. Original and averaged drawdown estimates in wells 184W101 and 184W502M.

Water temperature at the top of observation well 184W502M decreased more than 2 degrees Celsius,°C, during the 3-day aquifer test (Figure 4a), which indicated flow occurred in the observation well. This was possible because the long screen and open borehole allowed drainage from the water table to migrate directly through observation well 184W502M. Flow through an observation well is assumed to be minimal in most analytical solutions so the potential effect on hydraulic property estimates warranted further investigation.

Figure 4. Measured temperature at top of well 184W502M and simulated flow through well 184W502M while pumping well 184W101 at 2,520 gpm.

## Analysis

The carbonate-rock aquifer was conceptualized as a homogeneous, vertically anisotropic, thick unconfined aquifer which was characterized with a transmissivity, vertical-to-horizontal anisotropy, specific yield, and specific storage. These hydraulic properties were estimated with the Moench, analytical solution (Barlow and Moench, 1999) and a numerical model. The numerical model primarily differed from the analytical solution by simulating well 184W502M as a high hydraulic conductivity interval, which allowed flow to be simulated in the observation well.

Two sets of hydraulic properties were estimated to investigate the effect of flow through an observation well. Hydraulic properties of the carbonate-rock aquifer were estimated by minimizing differences between simulated and measured drawdowns. Drawdowns were simulated with both the analytical solution and a three-dimensional, MODFLOW model (Harbaugh and McDonald, 1996). Parameter estimation was performed by minimizing a weighted sum-of-squares objective function where the Moench solution was minimized with the Solver in Excel and the numerical model was minimized with MODOPTIM (Halford, 2006).

Hydraulic property estimates from the Moench analytical solution were reasonable for a carbonate-rock aquifer (Table 2). Hydraulic conductivity is 5 ft/d if a 9,800-ft^{2}/d transmissivity is divided by a 2,000-ft aquifer thickness. Specific-storage of 1.2 x 10-6 ft-1 and specific yield of 0.024 generally agree with other estimates for carbonate rocks. A vertical-to-horizontal anisotropy of 1.2 exceeds an expected ratio of less than 1 but bedding is absent.

Table 2. Hydraulic properties estimated with the numerical model.

## Numerical model: MODFLOW

Results from the aquifer test also were analyzed with a numerical model to test the effect of flow in the observation well on hydraulic property estimates. Only half of the area was simulated because drawdowns and flow were assumed to be symmetrical about a line that passes through wells 184W101 and 184W502M. Model discretization conformed to the diameters of the observation and pumping wells. Each well was simulated as a zone of virtually infinite hydraulic conductivity, 500 million ft/d. Hydraulic conductivity was assumed to be homogeneous and vertically anisotropic in the undisturbed aquifer as in the analytical model.

The model domain was discretized into 29 layers of 158 rows and 57 columns (Figure 5). The numerical model extended laterally 200,000 ft away from the pumping well 184W101. The vertical extent was from 479 to 2,479 ft below land surface, which conformed to the assumed saturated thickness of 2,000 ft. Column 1 intersected both wells with a width of 0.83 ft which is the radius of well 184W101. Column 2 was 0.2 ft wide and each successive column was 1.25 times wider to the furthest column. Rows 58 and 102 were 1.23 and 1.67 ft wide, respectively, which are the diameters of wells 184W101 and 184W502M. Rows adjacent to the wells were 0.2 ft wide and successive rows are 1.25 times wider away from the wells. The maximum row width between the wells is 18 ft. Layer thicknesses ranged from 1 ft at the water table to 360 ft at the base of the aquifer and were less than 10 ft thick near tops and bottoms of the open intervals in both wells 184W101 and (Figure 5). All external boundaries were no-flow. Changes in the wetted thickness of the aquifer were not simulated because the maximum drawdown near the water table was small relative to the total thickness. The aquifer test was simulated with a 3-day stress period.

Figure 5. Discretization of the numerical model and simulated drawdown surface after 3 days of pumping well 184W101 at 2,520 gpm.

Simulated and measured drawdowns matched within 0.5 ft in observation well 184W502M (Figure 6). The root-mean-square error of 0.5 ft was less than 3 percent of the 21-ft drawdown range analyzed. Simulated point observations were sampled 175 ft south of the pumping well where the simulated aquifer was undisturbed at depths of 0, 320, and 1,300 ft below land surface. Simulated drawdowns are noticeably different than point observations where flow is not simulated in the observation well (Figure 6 ). None of these time series duplicated the simulated drawdown in observation well 184W502M, but the deepest time series was most similar.

igure 6. Measured and simulated drawdowns in observation well 184W502M and simulated drawdowns at points 0, 320, and 1,300 ft below the water table during 3-day aquifer test.

Simulated and measured drawdowns matched within 13 ft in the pumping well 184W101 which is within 6 percent of the 220-ft drawdown range (Figure 7). Simulated and measured drawdowns departed more after the first day of pumping. This likely was caused by increased losses in the pumping well. A nearby impermeable boundary might similarly affect drawdowns. This is unlikely because drawdowns in the observation well also would be affected by a nearby boundary.

Figure 7. Measured and simulated drawdowns in pumping well 184W101 during 3-day aquifer test.

Drawdown surfaces were predominantly ellipsoidal shells near pumping well 184W101 that were perturbed by flow through observation well 184W502M (Figure 5). Greater drawdowns occurred near the water table as flow entered the observation well. Some flow also was induced at the bottom of observation well 184W502M. The maximum flow rate through observation well 184W502M was 50 gpm and flow rates averaged 40 gpm during the 3-day aquifer test (Figure 4b). Flow became mostly radial more than 2,000 ft from well pumping well 184W101.

## Hydraulic Property Estimates

Hydraulic property estimates for the alluvial aquifer from the analytical and numerical models differed little (Table 2). Hydraulic conductivity is 5.5 ft/d if an 11,000-ft^{2}/d transmissivity is divided by a 2,000-ft aquifer thickness. Specific-storage was about 0.5 x 10-6 ft-1 and was 40 percent of the analytical estimate. This was the greatest difference between hydraulic property estimates, which was insignificant. A specific yield estimate of 0.02 from the numerical model agrees with the analytical estimate of 0.024. A vertical-to-horizontal anisotropy of 1.0 from the numerical model is not appreciably different than an estimate of 1.2 from the analytical solution.Â Borehole flow in the observation well did not significantly affect hydraulic property estimates from the analytical solution despite violating the assumption of no borehole flow.

The transmissivity of the carbonate-rock aquifer around well 184W101 was 10,000 ft^{2}/d after rounding to 1 significant figure. Vertical-to-horizontal anisotropy of 1 was estimated which is reasonable given the lack of bedding in the carbonate. A specific yield of 0.02 agrees with other aquifer test results and effective porosity estimates in carbonate rocks.