Transmissivity and storage coefficient estimates from slug tests, Naval Air Warfare Center, West Trenton, New Jersey

• Home
• List of Figures and Tables
• Conversion Factors
• Abstract
• Introduction
• Methods
• Results
• Summary
• Acknowledgments
• References Cited
• Index

Methods

Mechanical slugs that raise the water level when instantaneously placed into the well (the slug-in test) and lower the water level when instantaneously removed from the well (the slug-out test) were used for testing. Four slugs were utilized: a 2-inch (in.) diameter by 4.5-foot (ft) long sand-filled polyvinyl chloride (PVC) slug, a 4-in. diameter by 4-ft long gravel-filled PVC slug, a 1-in. diameter by 6-ft long sand-filled PVC slug, and a 1-in. diameter by 5-ft long solid aluminum slug. The 2-in. x 4.5-ft slug was used most often because it fit inside most well casings and caused ample displacement. The 4-in. x 4-ft slug was used only when the well had a large casing diameter and an anticipated high T. The 1-in. x 5-ft and 1-in. x 6-ft slugs caused less water-level displacement and were used only if the well had an anticipated low T. A 10-pound per square inch (PSI) pressure transducer was used to measure water-level change during each test. The transducer transmitted the pressure in millivolts (mV) to a data logger that recorded the data at 0.25 to 60 second intervals. Shorter intervals were used for wells with faster water-level recovery, whereas larger intervals were used for wells with slower water-level recovery. A laptop computer was used to interface with the data logger and retrieve the data from the logger. Manual water-level measurements were obtained with an electric sounding tape for transducer calibration and conversion from pressure in mV to water levels in feet.

Slug test equipment at a tested well at the Naval Air Warfare Center, West Trenton, New Jersey.

Standard Operating Procedure

Standard operating procedures for the slug tests, as recommended by USGS guidelines (Cunningham and Schalk, 2011), are summarized below:

1. Open the well and take a water-level measurement. Take measurements every few minutes until several consecutive readings are identical to make sure the water level was static and not responding to stresses such as recharge or pumping.
2. Place the transducer in the well several feet deeper than the slug bottom is to be placed. Note the time and allow at least 10 minutes for the transducer to equilibrate and the water level to stabilize because installing the transducer down the well is itself a mini-slug test.
3. Insert the slug into the well, fully submerged below the water table. Insert the slug as quickly and cleanly as possible to cause a rapid rise of water level without fluctuations in the water level at the beginning of the test. Tie off the slug to prevent it from moving during the test. When the water level in the well returns to its static level or is no longer changing substantially, the slug-in test ends.
4. Remove the slug from the well, again as quickly and cleanly as possible. When the water level in the well returns to its static level or is no longer changing substantially, the slug-out test ends. Retrieve data from the data logger.

Methods of Analysis

The original (static) water level (H₀) and the water level at time from start of test (H) were recorded along with well construction information (table 1, XLSX, 24KB) (Lewis-Brown and Rice, 2002; P.J. Lacombe, USGS, written commun., 2013). The slug-out test data were used for analysis of most tests because slug removal is cleaner (faster with less splashing) than slug insertion. Slug-in test data were used for analysis in cases for which slug-out data were unavailable as a result of equipment malfunction or other circumstances. Manual water-level measurements were also used for analysis in these instances but only for slower-responding wells where manual measurements were adequate for analysis. Wells that showed no measurable change in water level within 20 minutes after slug insertion were not included in the analyses.

The analytical method used in calculating transmissivity for 33 bedrock wells was the Cooper-Bredehoeft-Papadopulos method (hereafter referred to as “Cooper”) (Cooper and others, 1967). For wells that recovered in less than 1 minute and indicated an oscillatory response, the high-K method of Butler and others (2003) (hereafter referred to as “Butler,” also known as the “Kansas Geological Survey high-K method”) was used alternatively for analysis. Both methods require a plot of normalized water-level displacement through time in which graphed test data are matched to a type curve (figs. 2–9, 10–19, 20–29, 30–39, 40–46). The shape of the Cooper type curve is governed by S and the radii of the open hole and casing. The shape of the Butler type curve represents the degree of oscillation controlled by a damping parameter value taken from a damped spring solution in classical physics. Type curves in both methods were generated automatically by the programs used in the analysis. The Cooper type curves were computed by use of the software "AQTESOLV" (Duffield, 2000). The Butler method type curves were computed by use of a USGS spreadsheet program (Halford and Kuniansky, 2002; June 2013 version from K.J. Halford, USGS, written commun., 2013). The Butler method yields values of hydraulic conductivity (K) that were subsequently converted to transmissivity by multiplying K by well open-interval thickness.

There are several assumptions incorporated into the Cooper method: (1) the aquifer is confined and has an infinite areal extent; (2) the aquifer is homogeneous, isotropic, and of uniform thickness; (3) the well completely penetrates the aquifer; (4) the potentiometric surface is initially horizontal; (5) water-level change in the well occurs instantaneously; (6) flow is unsteady; (7) water is discharged instantaneously from storage with the fall of aquifer water level (or vice versa); and (8) the well diameter is finite. These assumptions are not fully met by any real aquifers, but the approximations associated with these assumptions yield reasonable estimates of aquifer properties in many field settings (Shapiro and Hsieh, 1998). Because of the small volume of aquifer affected by the slug tests, the second assumption is considered applicable. Groundwater flow occurs in thin fractured units alternating with thin confining units (Lacombe, 2000), so it is likely a confining unit is situated near the top and bottom of each well opening. Therefore, the third assumption is considered applicable. The overlying surface layer at NAWC is a 10–25 ft thick weathered zone that includes clay-rich regolith and highly weathered bedrock (Lacombe, 2000), which act as low-K unconsolidated media (Tiedeman and others, 2010). Wells 63BR and 65BR are partially open to the weathered zone (table 1, XLSX, 24KB) (Lacombe, 2000). During slug tests of these wells, groundwater is released primarily from fractures intersecting the well, and the weathered zone functions as a confining unit. Thus, the confined assumption is adequate for these wells. Wells 57BR and 64BR are located where the weathered zone is comparatively thin (Lacombe, 2000), so the relatively shallow open intervals for these wells are completely below the weathered zone (table 1, XLSX, 24KB) and the confined assumption is applicable. The Butler method requires many of the same conditions as the Cooper method, except flow is deemed to be “steady-state” because the method neglects aquifer storage. This approximation is often adequate in fractured-rock aquifers with low bulk  porosity.