Slug Tests Research Articles

Combined performance of pumping and tracer tests: A case study

A combined pumping and tracer test was conducted at a highly fractured aquifer system. The hydrogeologic units underlying the test site are reclamation soil, weathered rock layer, and fractured layer. The fractured layer is the main aquifer for this site. Prior to pumping and tracer tests, slug tests were conducted at four test wells. The test data revealed existence of a low permeability zone near well OB-1. Generally the estimated hydraulic conductivities are in the order of 10−4 cm/sec. A pumping test with a discharge rate of 57 m3/d was performed for 1,230 min. The pumping test data analysis yielded coherent hydraulic conductivity values with those of the slug tests. However, the separate analysis for each monitoring well based on conventional analytical solutions with highly strict boundary conditions and homogeneity assumption cannot efficiently show the potential existence of the low permeability zone. During the pumping test, when the water levels of the pumping and monitoring wells are stabilized, a convergent radial tracer test was conducted. From the observed tracer concentration, a longitudinal dispersivity of 0.3 m was obtained, which is well consistent with the values in the prominent literature considering the test scale. This study excellently demonstrated a method completing a combined pumping and tracer test at one time.

Cross‐coupled flow modeling of coincident streaming and electrochemical potentials and application to subglacial self‐potential data

This study aims to demonstrate that self‐potential measurements can yield unique information about subglacial drainage conditions, which has so far been doubtful. At the bed of warm‐based glaciers arborescent and distributed drainage systems interact, and streaming and electrochemical potentials coincide. A key secondary aim of this study is, therefore, the development of a mathematical framework for the analysis of self‐potentials generated by such interacting flow systems. We consider cross coupling between hydraulic, electrical, and chemical driving forces, and apply the resulting model solution to self‐potential data recorded beneath Haut Glacier d'Arolla, Switzerland. We find that (1) the streaming and electrochemical contributions to the total self‐potential signal can be separated mathematically, (2) streaming potentials directly reflect the direction of water flow at the glacier bed, (3) combination of streaming potentials with hydraulic conductivity estimates from slug tests allows estimation of in situ subglacial water flow velocities, (4) the subglacial zeta potential, calculated from estimated values of the streaming coupling coefficient, is small (∼−0.022 V) due to systematic flushing of clay minerals from the channel area during the melt season, and (5) uncertainties in exact electrode location in the borehole are unlikely to produce self‐potential noise. Self‐potential measurements may therefore yield unique information about subglacial drainage conditions, and especially the direction and velocity of in situ subglacial water flow are difficult to obtain by other methods. We further anticipate that our mathematical framework will be applicable to other areas of subsurface hydrology, especially where spatially distributed seepage feeds preferential flow pathways.

Characterization of the hydraulic properties of fractures in chalk.

The fracture systems intersecting Eocene chalk formations in the Negev desert, Israel, and their hydraulic properties were characterized using a variety of geologic and hydrologic techniques. These included identification of the prevailing directions of fracture systems in outcrops, in cores retrieved from inclined coreholes, in coreholes using video logs, and in trenches. The orientation and inclination of these fracture systems were determined, and evidence of ground water flow on the fracture surfaces was described and ranked. Their hydraulic conductivity was determined through slug and pumping tests performed at discrete intervals. Temperature, electrical conductivity, caliper, gamma and heat-pulse logs were conducted in the same coreholes. The results from the logs, tests, and core descriptions were compared to identify reliable and cost-effective tools for investigating the hydraulic characteristics of fracture systems. We concluded that in the study area: (1) fracture mapping in outcrops and coreholes (including downhole video and caliper logs) must be supplemented by hydraulic testing of the mapped fracture sets in the coreholes; (2) inclined coreholes provide information regarding the orientation of the hydraulically active fracture systems that cannot be obtained from vertical boreholes; (3) hydraulic testing of unpacked holes provides a reasonable estimate of the maximum hydraulic conductivity; and (4) the hydraulic conductivity distribution with depth is log normal and all significant ground water flow takes place within the upper 25 m.

Field study of hydraulic conductivity in a heterogeneous aquifer: Comparison of single‐borehole measurements using different instruments

This field study compares three techniques for estimating the vertical distribution of horizontal hydraulic conductivity Kr in a heterogeneous aquifer and evaluates possible support volume effects. The dipole flow test (DFT), multilevel slug test (MLST), and borehole flowmeter test (BFT) are based on different kinematic flow structures and the shape and the size of the support volumes. The experiment design employed an identical characteristic linear scale for all tests. Vertical profiles of Kr ranging up to 260 m/day from tested wells in an alluvial aquifer exhibit a strong correlation in spite of the differences between test hydraulics. Results suggest that tested screen length is an important indicator of the averaging mechanism for hydraulic tests. Correlation between the DFT and MLST is especially strong. Correlation between data from the BFT and other tests is not as strong due to the absence of a distinct physical vertical scale, among other factors. The differences between the tests are discussed using the concept of a weighting function associated with the magnitude of instantaneous local velocity.

Characterizing submarine groundwater discharge: A seepage meter study in Waquoit Bay, Massachusetts

A seepage meter study was performed in Waquoit Bay on Cape Cod, Massachusetts to characterize the amount, pattern, and origin of submarine groundwater discharge. Measurements from grids of 40 seepage meters provide a detailed representation of groundwater flux in both space and time. At the head of the bay, a distinct band of high, saline discharge was observed between 25 and 45 m from the shoreline. Slug tests indicated no pattern of permeability to explain the band of discharge, and the band was not observed offshore of an island where freshwater discharge is negligible. Experiments using clusters of seepage meters showed large variability in discharge at the meter scale and similar temporal variation throughout the domain, reflecting tidal influence primarily near shore. The small‐scale variability challenges the assumption of locally homogeneous flow used in many models, and the band of discharge contradicts predictions that total outflow is largely fresh and decreases monotonically from shore.

Slug-test derived differences in bed hydraulic properties between a surge-type and a non-surge-type Svalbard glacier

AbstractWe investigate the differences in subglacial hydraulic properties between Bakaninbreen, a surge-type glacier in southern Svalbard, and midre Lovénbreen, a non-surge-type glacier in northwest Svalbard, using slug tests. At Bakaninbreen, underlain by fine-grained glacial till and marine sediments, slug-test responses were underdamped and are analyzed with the Van der Kamp method using a fully penetrating screen. At midre Lovénbreen, underlain by metres-thick permafrost consisting of coarse clasts, ice and water, slug-test responses were overdamped and are analyzed with the Butler–Garnett method using a partially penetrating screen. We calculate typical hydraulic conductivities of 8.2 ± 7.8 x 10–3 ms–1 for Bakaninbreen, and 1.9 ± 0.5 × 10–5 m s–1 for midre Lovénbreen, after correction for a high-conductivity skin. At Bakaninbreen, late surge-induced subglacial sediment dilation probably caused marked hydraulic conductivity enhancement, which could be widespread during times of peak ice flow. We argue that the flow pathways in the permafrost beneath midre Lovénbreen are present, though limited in terms of their discharge capacity, which in combination with drilling-based observations and independent evidence suggests that midre Lovénbreen is not capable of surging.

What does a slug test measure: An investigation of instrument response and the effects of heterogeneity

We consider parameters determined by the inversion of slug‐test head recovery data with the homogeneous‐parameter model of Cooper et al. [1967] to be weighted spatial averages of transmissivity and storage defined at a smaller scale. We quantify the spatial averaging using a power‐average spatial filter expression. We determine the form of the filter function and the power exponent using numerically simulated slug‐test data. The filter function that describes how smaller‐scale transmissivity is weighted by slug tests displays an approximate 1/r2 behavior, with r the radial distance from the well. The radius of the cylinderical volume that is averaged by the slug test is inversely proportional to the square root of the storage coefficient (larger averaging volume with smaller storage). The power exponent grows from −0.19 to 0.345 as the ratio of the characteristic scale of the heterogeneity to the characteristic scale of the averaging volume grows, although a power exponent of zero, corresponding to geometric averaging, provides good results for most simulations. Our results show that while slug tests are useful to estimate transmissivity, they have dubious value for estimating storage coefficients. We find that the transmissivity estimate is unbiased and does not appear to be strongly influenced by storage properties. The storage coefficient estimate is, however, strongly influenced by the transmissivity and is biased. We investigate the interaction between storage coefficient and transmissivity by examining an analytical slug‐test model that contains two annular zones, each with distinct transmissivity and storage coefficient.

Improving the analysis of slug tests

This paper examines several techniques that have the potential to improve the quality of slug test analysis. These techniques are applicable in the range from low hydraulic conductivities with overdamped responses to high hydraulic conductivities with nonlinear oscillatory responses. Four techniques for improving slug test analysis will be discussed: use of an extended capability nonlinear model, sensitivity analysis, correction for acceleration and velocity effects, and use of multiple slug tests. The four-parameter nonlinear slug test model used in this work is shown to allow accurate analysis of slug tests with widely differing character. The parameter β represents a correction to the water column length caused primarily by radius variations in the wellbore and is most useful in matching the oscillation frequency and amplitude. The water column velocity at slug initiation ( V o) is an additional model parameter, which would ideally be zero but may not be due to the initiation mechanism. The remaining two model parameters are A (parameter for nonlinear effects) and K (hydraulic conductivity). Sensitivity analysis shows that in general β and V o have the lowest sensitivity and K usually has the highest. However, for very high K values the sensitivity to A may surpass the sensitivity to K. Oscillatory slug tests involve higher accelerations and velocities of the water column; thus, the pressure transducer responses are affected by these factors and the model response must be corrected to allow maximum accuracy for the analysis. The performance of multiple slug tests will allow some statistical measure of the experimental accuracy and of the reliability of the resulting aquifer parameters.

Test methods for characterizing contaminant transport in a glaciated carbonate aquifer

It is widely recognized that karstification can substantially influence flow and transport characteristics in carbonate aquifers. Surface features such as sinkholes are widely used to diagnose the presence of a karst aquifer, but specific borehole tests for karst have not been well defined. Such tests are especially important in glaciated areas where karst features have been eroded or buried by till. One such area is Smithville, Ontario, where more than 60 boreholes at a PCB-contaminated dolostone site provided an opportunity for a wide range of downhole tests and monitoring to be carried out. It was found that there were a number of useful tests for indicating karstification. These included (1) the order of magnitude differences between pump, slug and packer test results, (2) the presence of water table troughs, (3) rapid water level response following recharge events, (4) the rapid changes in water quality following recharge events, (5) water undersaturated with respect to calcite following recharge events, and (6) a wide range in fracture apertures along major bedding planes. Most parameters vary over a similar range at Smithville as they do at Mammoth Cave, Kentucky, indicating that the Smithville aquifer behaves as a typical karst aquifer.

Effects of Electroosmosis on Soil Temperature and Hydraulic Head. II: Numerical Simulation

A numerical model to simulate the distributions of the voltage, soil temperature, and hydraulic head during a field test of electroosmosis was developed. The two-dimensional governing equations for the distributions of the voltage, soil temperature, and hydraulic head within a cylindrical domain are derived based on the principles of charge, energy, and mass conservations, Darcy’s law, Ohm’s law, and Fourier’s law of heat conduction. We assumed that the voltage distribution was at steady state, whereas the soil temperature and hydraulic head were at transient states during the test. The simulated domain was segmented with a block-centered finite-difference scheme and the resulting equations were solved numerically with the successive overrelaxation method. The parameters (such as electrical, thermal, hydraulic, and electroosmotic properties of the soil, graphite, and sand) that were required by the model were measured either using core samples or slug tests. The model is able to predict the pattern as wel...

A simple correction for slug tests in small-diameter wells.

A simple procedure is presented for correcting hydraulic conductivity (K) estimates obtained from slug tests performed in small-diameter installations screened in highly permeable aquifers. Previously reported discrepancies between results from slug tests in small-diameter installations and those from tests in nearby larger-diameter wells are primarily a product of frictional losses within the small-diameter pipe. These frictional losses are readily incorporated into existing models for slug tests in high-K aquifers, which then serve as the basis of a straightforward procedure for correcting previously obtained K estimates. A demonstration of the proposed procedure using data from a series of slug tests performed in a controlled field setting confirms the validity of the approach. The results of this demonstration also reveal the detailed view of spatial variations in K that can be obtained using slug tests in small-diameter installations.

A dual-tube direct-push method for vertical profiling of hydraulic conductivity in unconsolidated formations

Abstract A new direct-push procedure has been developed for the purpose of conducting discrete-interval slug tests to define vertical variations in hydraulic conductivity (K.) This approach is an extension of existing dual-tube methods developed for soil sampling. In this procedure, nested rods (tubes) are simultaneously advanced to predetermined test intervals. The inner rods are then removed and a screen is inserted into the formation for slug testing and possible water sampling. Once testing and sampling are completed, the screen is retrieved, the inner rods reinserted, and the system is advanced to the next test interval. A series of field tests were performed in a highly permeable sand and gravel aquifer to assess the effectiveness of this new approach. Dual-tube profiling results were compared to multilevel slug tests conducted in conventional monitoring wells for intervals in which hydraulic conductivity ranged from 175 ft/day to over 800 ft/day. An initial evaluation found that the dual-tube profiling results were in good agreement (< or =12 percent difference) with K values obtained from multilevel slug tests in the closest monitoring well. Two more-detailed profiles demonstrate that the dual-tube method can effectively delineate small-scale vertical and horizontal variations in hydraulic conductivity. This field assessment shows that the dual-tube method is an accurate and efficient procedure for obtaining information about spatial variations in hydraulic conductivity. This information can be useful for selecting intervals for well installations, for assessment of various remediation alternatives, and for identifying preferential flow paths and other features that can control contaminant movement in the subsurface. The information is obtained without the need for permanent wells. Because this is a direct-push procedure, drill cuttings are eliminated and the volume of development water generated is significantly reduced.

Analysis of slug tests in bypassed wells

In Berlin many wells are constructed with an additional small diameter bypass tube flanged on to the large diameter casing and running parallel to the later upward to the surface. The response to a slug test conducted in such a well is usually characterised by significant water level oscillations developing inside the branched pipe system. In the present paper, a general non-linear model is developed for the analysis of slug tests performed in bypassed wells. The model includes skin effects, non-linear head losses due to internal well bore fluid friction, minor losses originating at radius changes along the flow path inside the well, and inertial effects of the water columns contained within the primary casing and the bypass. Pipe flow inside the branched tubing system is described by the one-dimensional conservation law of the mechanical energy of the well bore fluid. Aquifer flow is quantified by a convolution-type approach. The coupled well-aquifer system is governed by an integro-differential equation representing the mean mechanical energy of the total well bore fluid and, coupled to it, by a non-linear ordinary differential equation which accounts for mechanical energy fluctuations between the water columns inside the main casing and the bypass. Both equations are solved by finite difference approximations coupled to point iterative numerical techniques. An application of the model to field data suggests that non-linear hydraulic head losses originating inside the well bore may cause the measured head responses to depend on the initial displacement. A packer used to initiate a slug test can induce these non-linear distortions through the combined effect of minor losses originating at the inlet and the outlet of a small diameter packer flow-through tube and internal fluid friction potentially developing inside that packer passage. As classical linear theories do not allow for the analysis of such phenomena, a model accounting for these non-linear processes should be used if field data show significant dependencies on the initial slug height.

Groundwater Hydraulics and Pollutant Transport

3R55. Groundwater Hydraulics and Pollutant Transport. - RJ Charbeneau (Dept of Civil Eng, Env and Water Resources Eng Program, Univ of Texas, Austin TX). Prentice Hall, Upper Saddle River NJ. 2000. 593 pp. ISBN 0-13-975616-7. $105.00.Reviewed by SA Sherif (Dept of Mech Eng, Univ of Florida, 228 MEB, PO Box 116300, Gainesville FL 32611-6300).This book is intended for use as a textbook for upper level undergraduate and graduate courses in groundwater hydrology, groundwater hydraulics, and mass transport of subsurface contaminants. It is also intended to serve as a reference for practicing hydrologists, hydrogeologists, and environmental engineers. The book contains nine chapters, nine appendices, 460 references grouped at the end of the book, and an index. Chapter 1 is an introduction to groundwater hydrology with topics dealing with porous media, distribution of subsurface water, porosity and related properties of soil, subsurface hydrologic cycle, and hydrogeologic formations. Chapter 2 presents a discussion of Darcy’s Law, continuity relations for flow in porous media, and groundwater management models. Chapter 3, entitled Groundwater and Well Hydraulics, addresses topics such as steady and transient flow to a well in an ideal confined aquifer, pumping tests, slug tests, well tests, multiple well problems, potential flow for stratified aquifers, the interface in coastal aquifers, and other transient flow problems. Chapter 4, entitled The Vadose Zone and Groundwater Recharge, presents detailed discussions of soil water in the vadose zone, soil water characteristic curve, Darcy’s Law and Richard’s Equation, measurement of soil properties, infiltration models, redistribution of soil water, evaporation and desorption models, evaporation from a shallow water table, and water balance and groundwater recharge. Chapter 5 deals with sources of subsurface contamination, mass transport processes, the general continuity equation, solute partitioning, degradation losses of soil and groundwater contaminants, and simplified forms of the continuity equation. Chapter 6 entitled Solute Transport by Advection, includes topics such as advection transport, potential theory, potential and stream functions, some applications of potential theory, residence time distribution theory, standard flow patterns, and evaluating the environmental consequences of groundwater contamination. Chapter 7 is entitled Solute Transport by Diffusion and deals with Fick’s laws, molecular diffusion coefficients, diffusion in porous media, diffusion in multiphase systems, some applications of the diffusion equation, and volatilization losses of soil contaminants. Chapter 8 presents detailed discussions of one-dimensional flow and column experiments, radial flow from a well, transverse dispersion, the mechanical dispersion tensor, moments of the transport equation, analytical models of chemical spills and contaminant plumes, numerical simulation of solute transport, nonideal flow in porous media, and subsurface mass transport through the vadose zone. Chapter 9 deals with multiphase flow and hydrocarbon recovery. This includes topics such as capillary trapping and residual saturation, NAPL behavior in fractured media, monitoring of free-product petroleum hydrocarbon, NAPL infiltration in the vadose zone, screening models for fate/transport of organic chemicals in soil and groundwater, soil-vapor extraction systems, and free product recovery of petroleum hydrocarbon liquids. Many of the appendices have spreadsheet modules for topics discussed in the book. Examples include modules for calculating well functions, slug test well function, LNAPL distribution, and LNAPL recovery analysis. Groundwater Hydraulics and Pollutant Transport is very well written and organized, contains topics appropriate for its stated title and objectives, has numerous examples and chapter problems, and has very clear illustrations and diagrams. The author is obviously very knowledgeable in the subject matter and is able to use his skills in teaching the subject over many years to convey his ideas clearly to the reader. This reviewer recommends that the book be acquired by academic and research libraries as well as practicing hydrologists, hydrogeologists, and environmental engineers.

Dynamic interpretation of slug tests in highly permeable aquifers

Considerable progress has been made in developing a theoretical framework for modeling slug test responses in formations with high hydraulic conductivity K. However, several questions of practical significance remain unresolved. Given the rapid and often oscillatory nature of test responses, the traditional hydrostatic relationship between the water level and the transducer‐measured head in the water column may not be appropriate. A general dynamic interpretation is proposed that describes the relationship between water level response and transducer‐measured head. This theory is utilized to develop a procedure for transforming model‐generated water level responses to transducer readings. The magnitude of the difference between the actual water level position and the apparent position based on the transducer measurement is a function of the acceleration and velocity of the water column, test geometry, and depth of the transducer. The dynamic approach explains the entire slug test response, including the often‐noted discrepancy between the actual initial water level displacement and that measured by a transducer in the water column. Failure to use this approach can lead to a significant underestimation of K when the transducer is a considerable distance below the static water level. Previous investigators have noted a dependence of test responses on the magnitude of the initial water level displacement and have developed various approximate methods for analyzing such data. These methods are re‐examined and their limitations clarified. Practical field guidelines are proposed on the basis of findings of this work. The soundness of the dynamic approach is demonstrated through a comparison of K profiles from a series of multilevel slug tests with those from dipole‐flow tests performed in the same wells.

Characteristics of preferential flow and groundwater discharge to Shingobee Lake, Minnesota, USA

AbstractSmall‐scale heterogeneities and large changes in hydraulic gradient over short distances can create preferential groundwater flow paths that discharge to lakes. A 170 m2 grid within an area of springs and seeps along the shore of Shingobee Lake, Minnesota, was intensively instrumented to characterize groundwater‐lake interaction within underlying organic‐rich soil and sandy glacial sediments. Seepage meters in the lake and piezometer nests, installed at depths of 0·5 and 1·0 m below the ground surface and lakebed, were used to estimate groundwater flow. Statistical analysis of hydraulic conductivity estimated from slug tests indicated a range from 21 to 4·8 × 10−3 m day−1 and small spatial correlation. Although hydraulic gradients are overall upward and toward the lake, surface water that flows onto an area about 2 m onshore results in downward flow and localized recharge. Most flow occurred within 3 m of the shore through more permeable pathways. Seepage meter and Darcy law estimates of groundwater discharge agreed well within error limits. In the small area examined, discharge decreases irregularly with distance into the lake, indicating that sediment heterogeneity plays an important role in the distribution of groundwater discharge. Temperature gradients showed some relationship to discharge, but neither temperature profiles nor specific electrical conductance could provide a more convenient method to map groundwater–lake interaction. These results suggest that site‐specific data may be needed to evaluate local water budget and to protect the water quality and quantity of discharge‐dominated lakes. Copyright © 2002 John Wiley & Sons, Ltd.

Slug tests in a confined aquifer: experimental results in a large soil tank and numerical modeling

Variable-head (slug) tests in a confined aquifer can be interpreted using different methods that either consider or neglect the influence of the instantaneous deformation of an elastic solid matrix. This paper defines a unified interpretation for slug tests: it is based on the velocity graph describing the conservation equation underlying all methods. If the storativity S has no influence, the velocity graph is a straight line. If S has an influence, the theory considering this influence predicts the graph should be a curve. Numerous slug tests were performed in a large tank containing a confined aquifer. Other tests were used to determine independently the transmissivity T and S values of the confined aquifer which are compared with those obtained from slug tests. The velocity graphs of the slug tests provided straight lines instead of the smooth curves as predicted by the theory. A numerical analysis of these tests in the sand tank was performed using a finite element method. The analysis gave straight lines instead of curves for any S value and therefore confirmed the experimental observation (in velocity graphs) that slug test results do not depend of S and thus cannot be used to determine the S value.Key words: slug test, hydraulic conductivity, storativity, numerical modeling.

Field Experiment and Data Analysis of a Constant-Head Injection Test with Skin Effects in a Low-Transmissivity Aquifer

In a confined, silty sand aquifer of low transmissivity, constant-rate pumping can rapidly dewater the pumping well without creating measurable drawdown at nearby observation wells. To overcome this difficulty, a constant-head test is performed by overflowing the test well by water injection. The difference between the injection rate and the overflow rate directly gives the cumulative volume Vw(t) entering the aquifer, which is characteristic of the aquifer hydrogeologic properties and can be used for parameter estimation. The field experiments include two constant-head injection tests, CHIT1 and CHIT2, performed at two different wells, respectively. Semilog plot of t/Vw(t) of CHIT2 indicates little or no skin, and the two unknown parameters of the storage coefficient and transmissivity can be uniquely determined in a similar way by the well-known Jacob semilog method. Semilog plot of t/Vw(t) of CHITl, however, shows the ex istence of well skin, and a curve matching method is developed for estimatiog the four pertinent parameters. Among them, the aquifer transmissivity can be uniquely determined while the skin transmissivity, the skin thickness, and the storage coefficient are obtained in appropriate ranges. To avoid non-unique estimates of aquifer parameters, therefore, the test well should be carefully constructed with complete and thorough well development to prevent skin region from happening. Due to the limitation of the site hydrogeology, only slug tests can be performed for the sake of validating transmissivity. The transmissivity values determined from the CHITs and from the slug tests are in good agreement, supporting the usefulness of the CHIT. The aquifer transmissivities estimated from CHITl and CHIT2 are within 14% difference, indicating that the aquifer is relatively homogeneous in the neighborhood of the two test wells.

Hydraulic tests with direct-push equipment.

The potential of direct-push technology for hydraulic characterization of saturated flow systems was investigated at a field site with a considerable degree of subsurface control. Direct-push installations were emplaced by attaching short lengths of screen (shielded and unshielded) to the bottom end of a tool string that was then advanced into the unconsolidated sediments. A series of constant-rate pumping tests were performed in a coarse sand and gravel aquifer using direct-push tool strings as observation wells. Very good agreement (within 4%) was found between hydraulic conductivity (K) estimates from direct-push installations and those from conventional wells. A program of slug tests was performed in direct-push installations using small-diameter adaptations of solid-slug and pneumatic methods. In a sandy silt interval of moderate hydraulic conductivity, K values from tests in a shielded screen tool were in excellent agreement (within 2%) with those from tests in a nearby well. In the coarse sand and gravel aquifer, K values were within 12% of those from multilevel slug tests at a nearby well. However, in the more permeable portions of the aquifer (K > 70 m/day), the smaller-diameter direct-push rods (0.016 m inner diameter [I.D.]) attenuated test responses, leading to an underprediction of K. In those conditions, use of larger-diameter rods (e.g., 0.038 m I.D.) is necessary to obtain kappa values representative of the formation. This investigation demonstrates that much valuable information can be obtained from hydraulic tests in direct-push installations. As with any type of hydraulic test, K estimates are critically dependent on use of appropriate emplacement and development procedures. In particular, driving an unshielded screen through a heterogeneous sequence will often lead to a buildup of low-K material that can be difficult to remove with standard development procedures.

Application of a nonlinear slug test model.

Knowledge of the hydraulic conductivity distribution is of utmost importance in understanding the dynamics of an aquifer and in planning the consequences of any action taken upon that aquifer. Slug tests have been used extensively to measure hydraulic conductivity in the last 50 years since Hvorslev's (1951) work. A general nonlinear model based on the Navier-Stokes equation, nonlinear frictional loss, non-Darcian flow, acceleration effects, radius changes in the wellbore, and a Hvorslev model for the aquifer has been implemented in this work. The nonlinear model has three parameters: beta, which is related primarily to radius changes in the water column; A, which is related to the nonlinear head losses; and K, the hydraulic conductivity. An additional parameter has been added representing the initial velocity of the water column at slug initiation and is incorporated into an analytical solution to generate the first time step before a sequential numerical solution generates the remainder of the time solution. Corrections are made to the model output for acceleration before it is compared to the experimental data. Sensitivity analysis and least squares fitting are used to estimate the aquifer parameters and produce some diagnostic results, which indicate the accuracy of the fit. Finally, an example of field data has been presented to illustrate the application of the model to data sets that exhibit nonlinear behavior. Multiple slug tests should be taken at a given location to test for nonlinear effects and to determine repeatability.