Combined interpretation of pumping and tracer tests: theoretical considerations and illustration with a field test

doi:10.1016/S0022-1694(03)00090-8

Journal of Hydrology

Volume 277, Issues 1–2, 1 June 2003, Pages 134-149

https://doi.org/10.1016/S0022-1694(03)00090-8 Get rights and content

Abstract

Pumping and tracer tests are two well-established ground water reservoir tests. Hydraulic conductivity can be derived from both. Conductivities are studied in a mainly layered heterogeneous coastal aquifer in Belgium with a pumping and tracer test and results of both are compared. Characteristics and important points in performance and interpretation of the tests are discussed. The interpretation method of pumping tests must be chosen very carefully. Application of interpretation methods based on models underestimating or even neglecting the leakage towards the pumped permeable layer results in an overestimation of the horizontal conductivity of this layer. The larger the distance of the used observation well(s), the larger this overestimation becomes. This is illustrated here with the interpretation of the drawdown data obtained during the pumping test using different analytical models and a numerical model. With an inverse axi-symmetric numerical model incorporating layered heterogeneity, results of the pumping test are in good agreement with results from the tracer test. Many discrepancies between pumping and tracer test data performed in layered heterogeneous media can be explained when realistic pumping test interpretation methods are applied and when leakage is not underestimated.

Introduction

Mass that is dissolved in ground water undergoes transport and redistribution by flow. This solute transport is involved in many problems in ground water systems, for instance in contaminant flow, distribution of fresh and salt water, water quality assessment, etc. Understanding of the parameters controlling solute transport in a ground water system is therefore a major concern in hydrogeological research. Two processes, advection and dispersion, govern this solute transport. Advection is a mass transport process due to the flow of water in which mass is dissolved. The direction and rate of transport coincides with the ground water flow. This process determines therefore when a mass dissolved in ground water is observed at a certain place. Advection is completely determined by ground water velocity. Hydraulic conductivity, porosity and hydraulic gradient determine ground water velocity. Dispersion is a process of fluid mixing that causes the development of a mixing zone between the dissolved mass and the displaced fluid with a different composition. Dispersion determines the evolution of this zone or, in other words, the evolution of the concentration during the passage of a tracer plume. The solute transport parameters, longitudinal and transverse dispersivity describe this process along with the ground water flow velocities. This paper focuses on advection and how to predict reliably the time of maximum breakthrough of dissolved masses.

Hydraulic and solute transport parameters can be determined by hydraulic testing of aquifers. Two basic and well proven methods are available in hydrogeological research: pumping and tracer tests. During a pumping test, ground water is extracted from a permeable layer. Hydraulic heads are logged in the pumped and surrounding layers. Horizontal conductivities, specific elastic storages and hydraulic resistances (or vertical conductivities) can then be derived (Lebbe, 1999). During a tracer test, a tracer is injected in the ground water reservoir and its movements and distribution are observed. Many different working methods have been developed (Domenico and Schwartz, 1997). Longitudinal and transverse dispersivity and the ratio of hydraulic conductivity and effective porosity can be determined. Hence, both tests deal with advective properties of the aquifer. With a pumping test, hydraulic conductivities are derived from drawdown measurements and with a porosity estimate the velocity field around the pumping well can be calculated. With a tracer test, travel times are measured between observation wells. These allow the derivation of ground water velocities. Along with the observed gradients between the wells, the ratio of the horizontal conductivity to the effective porosity can be derived. It is expected that both tests result in approximately the same value for the horizontal conductivity when performed on the same location and in the same aquifer. With both tests, breakthrough times should be predicted and/or explained coherently. Few articles dealing with combined interpretation of pumping and tracer tests are available in literature. Results of those few published studies, however, question seriously the conformity of pumping and tracer test interpretations. Niemann and Rovey (2000) for instance find hydraulic conductivities derived from a pumping test which are 10 to 20 times larger than derived with a tracer test. Differences are attributed to scale effects and influences of heterogeneities. Thorbjarnarson et al. (1998) on the other hand point to the difference between average aquifer horizontal conductivity derived with a pumping test and the horizontal conductivity of different layers in an aquifer derived with a tracer test. Horizons where tracer movement would be six times larger than calculated with the mean horizontal conductivity are found. Use of pumping test-derived average conductivities would underestimate the travel distance of tracers. In this context Mallants et al. (2000) point to the importance of good estimates of conductivity in solute transport. In most studies, tracer tests are considered to be more accurate because they would better implement the heterogeneity. Freeze and Cherry (1979) have seriously questioned the uniqueness of a pumping test interpretation. They argue for ‘simpler’ well tests.

Here we want to discus the performance of pumping and tracer test executions and the validity of pumping test analyses. These theoretical considerations are demonstrated with a field test in a layered heterogeneous aquifer in the Belgian coastal plain. A pumping test and a tracer test using a conservative tracer (salt water) were performed. It is argued that in predominantly layered heterogeneous media, pumping test data and tracer test data can brought into good agreement if the pumping test is analysed with a realistic aquifer model, incorporating geological data on the layering. The remainder of this paper is organised as following. First, pumping test interpretation with different analytical models and an inverse numerical model is discussed. Then, some major points of attention in pumping and tracer test performance are given. Finally, interpretation of a pumping and tracer test in the Belgian coastal plain is presented.

Section snippets

Pumping tests analyses

Fig. 1 shows schematically the flow in permeable and semi-permeable layers of a ground water reservoir during pumping. Layers A and C are semi-permeable and layer B is a permeable layer. When water is extracted from layer B, ground water flow occurs in this layer towards the screen. This ground water flow is characterised by the horizontal conductivity K_h and specific elastic storage S_s from layer B and the effect of pumping can be observed by the lowering of the hydraulic head in observation

Interpretation methods derived from analytical models

The model of Theis (1935) describes the unsteady state ground water flow in a confined aquifer. Specific boundary conditions and assumptions are made for the solution of the ground water flow equation. The permeable layer is considered bounded above and below by an impermeable layer. Further, all layers are considered of infinite lateral extend, homogeneous and of constant thickness. Discharge rate is constant and the pumping well is screened over the entire thickness of the permeable layer.

Inverse numerical model

HYPARIDEN (HYdraulic PARameter IDENtification) (Lebbe, 1999), is a set of computer codes developed as a generalised interpretation method for single and multiple pumping tests. It is based on an axial-symmetrical model. The ground water reservoir can be subdivided in a large number of layers, each with a horizontal conductivity, a specific elastic storage and a value for the hydraulic resistance. The hydraulic resistance of a layer is its thickness divided by its vertical conductivity. In this

Combined pumping and tracer test

Observation wells with long screens are often used during tracer tests. This provides the opportunity to take water samples on different levels without the need to install well nests. Although this method is very practical for tracer tests, using these as observation wells during a pumping test is not advisable. In most cases, layers with different hydraulic parameters are intersected. Drawdowns observed in such wells are in most cases not representative for one particular layer but depend on

Practicle example: the Houtave test site

The theoretical considerations are illustrated with a pumping and tracer test conducted in the phreatic aquifer of the Belgian coastal plain, near the village of Houtave. Fig. 2 shows a schematical cross-section of the site. The aquifer is bounded below by clay. This clay is considered here as impermeable. On top of the clay layer, a semi-permeable layer of sandy clay to clayey sand with horizons of sandstones is found. Above this, a permeable layer of fine siltous sand is present. The top of

Discussion and conclusions

A pumping and tracer test performed at the Houtave test site illustrates that proper analysis of both leads to coherent results. This is demonstrated by the successful prediction of the maximum breakthrough times in two observation points using hydraulic parameters derived from the pumping test. The important step herein is the careful interpretation of the pumping test in which the selection of the proper model is very important. The importance of this selection is illustrated by the

Acknowledgements

The authors would like to thank Prof. Dr K. Thorbjarnarson from San Diego State University (Department of Geological Sciences) and one anonymous reviewer for their critical suggestions to improve the paper.

References (25)

G.L. Moltyaner
Mixing cup and through-the-wall measurements in field-scale tracer tests and their related scales of averaging
J. Hydrol.
(1987)
Boulton N.S., 1955. Unsteady radial flow to a pumped well allowing for delayed yield from storage. IASH Assemblée...
R.L. Cooley
A finite difference method for unsteady flow in variably saturated process media, application to a single pumping well
Water Resour. Res.
(1971)
R.L. Cooley
Numerical simulation of flow in an aquifer overlain by a water table aquitard
Water Resour. Res.
(1972)
R.L. Cooley et al.
Effect of a watertable aquitard on drawdown in an underlying pumping aquifer
Water Resour. Res.
(1973)
H.H. Cooper et al.
A generalized graphical method for evaluating formation constants and summarising well field history
Am. Geophys. Union Trans.
(1946)
P.A. Domenico et al.
Physical and Chemical Hydrogeology
(1997)
C.E. Feehley et al.
A dual-domain mass transfer approach for modeling solute transport in heterogeneous aquifers: application to the Macrodispersion Experiment (MADE) site
Water Resour. Res.
(2000)
Franceschi, G., 1975 Geological study of the superficial layers in the area of Houtave. MSc Thesis, Ghent University....
R.A. Freeze et al.
Groundwater
(1979)

M.S. Hantush

Modification of the theory of leaky aquifers

J. Geophys. Res.

(1960)

M.S. Hantush et al.

Non-steady radial flow in an infinite leaky aquifer

Trans. Am. Geophys. Union

(1955)

Cited by (26)

A numerical modelling approach to investigate the fate of brine reject of farm scale desalination plants on groundwater aquifers in arid environments
2024, Science of the Total Environment
Farm-scale desalination units are gaining popularity for agricultural irrigation in arid countries, such as Kuwait to meet freshwater demands. However, less attention has been given to the management of environmentally hazardous brine reject water they produce. In this study we investigated the fate of brine water produced by the inland desalination units on the underlying aquifers using numerical modelling and field investigations. The methodology involved developing groundwater flow and solute transport models using Flex VMF-SEAWAT to simulate the movement of reject brine. The field investigations included collecting 150 water samples and conducting pumping tests on newly drilled wells. This numerical simulation considered advection, dispersion, and adsorption processes with variable groundwater density following rigorous validation and calibration of the developed numerical models. The results show that the RO reject brine will significantly increase groundwater salinity, exceeding 10,000 mg/L when accounting for advection, dispersion, and adsorption processes. The sustainable yield of the aquifer, with a salinity of <10,000 mg/L, averages 500 Mm³ but is expected to be depleted within 16 years with the current extraction rate. The resulting hydraulic properties are favourable with K about 100 m/d, T > 1000 m²/day, and Sy just >0.1. The adopted values for dispersivity and adsorption coefficients for chloride and sulphate salts in the aquifer were 10 m and 1 × 10⁻⁷ [mg/L]⁻¹ respectively. Chemical and numerical analyses indicate a mixing ratio between the reject brine and groundwater in the study area of approximately 10 %. Uncontrolled groundwater extraction, combined with the surface disposal of RO reject brine, has led to a significant decline in groundwater levels and an increase in the salinity. The adsorption ratio of simulated brine plume was 13 %. The authors recommend to dispose the RO reject water in a safe location or transfer it to the nearest wastewater treatment plant for proper treatment and reuse.
Characterization of flow and transport in a fracture network at the EGS Collab field experiment through stochastic modeling of tracer recovery
2021, Journal of Hydrology
Citation Excerpt :
The successful exploitation of a subsurface reservoir requires a comprehensive understanding of flow and transport characteristics in the reservoir, particularly in the context of unconventional oil/gas extraction (Middleton et al., 2015), geothermal heat recovery (Brown et al., 2012; Fu et al., 2016; McClure and Horne, 2014; U.S. Department of Energy, 2019), CO2 storage (Fu et al., 2017; Sun and Tong, 2017), as well as radioactive and toxic industrial wastes containment (Cuss et al., 2015; Sudicky and Frind, 1984; Sun and Buscheck, 2003; Tang et al., 1981; Tsang et al., 2015). Quantitative characterization of flow and transport processes in subsurface reservoirs is commonly based on flow (or pressure) and tracer tests in conjunction with various geological and geophysical investigations such as core logging, outcrop analysis, and seismic and electrical imaging (Berkowitz, 2002; Goovaerts, 1997; Juliusson and Horne, 2013; Karmakar et al., 2016; Neuman, 2005; Vandenbohede and Lebbe, 2003). The inference of spatially variable hydraulic and transport properties can be achieved by matching the measured flow and tracer data with results from either analytical or numerical models constrained by geophysical investigations (Bullivant and O'sullivan, 1989; Cacas et al., 1990b; Castagna and Bellin, 2009; Hawkins et al., 2017a, 2017b, 2018; Radilla et al., 2012).
Energy extraction from subsurface reservoirs is important for addressing the increasing energy demand and environmental concerns such as global warming. However, the characterization of subsurface reservoirs, particularly reservoirs dominated by fracture networks remains a challenge due to the lack of means to directly observe subsurface processes. This study explores the feasibility and efficacy of characterizing fracture flow and transport processes in an enhanced geothermal system (EGS) testbed through stochastic tracer modeling. There are two enabling factors that allow application of stochastic modeling to characterize a subsurface reservoir. First, an abundance of geological and geophysical measurements enables the development of a high-fidelity and well-constrained fracture network model. Second, high-performance computing (HPC) allows running massive realizations efficiently. Six conservative tracer tests were stochastically modeled and produced satisfactory realizations that successfully reproduce field tracer recovery data from each tracer test. The evolution of flow and transport processes in the fracture network was then analyzed from these satisfactory realizations. The present study demonstrates that stochastic tracer modeling on a high-fidelity fracture network model is feasible and can provide important insights regarding flow and transport characteristics in subsurface fractured reservoirs.
An empirical specific storage-depth model for the Earth's crust
2021, Journal of Hydrology
Specific storage, a basic hydraulic property of an aquifer, is required for transient groundwater studies. Previous investigations have shown that specific storage decreases with depth. Here, we collected field-specific storage-depth data from the literature and proposed an empirical model that describes the decreasing trend of specific storage with depth in the Earth's crust. The results show that the model agrees well with the field-specific storage-depth data. There are three parameters in the proposed model, which is flexible in fitting specific storage-depth data. The model also yields a finite value at the ground surface and a reasonable decay rate with depth. This is the first model describing the decreasing trend of specific storage with depth. The model is simple and can readily be applied to transient numerical models of groundwater flow and solute transport in porous media.
A review of specific storage in aquifers
2020, Journal of Hydrology
Specific storage, which represents the aquifer’s capacity to release water from or take into storage when water level changes, is an important aquifer property required for transient groundwater modeling and water resources management. This review synthesizes the published literature on using different methods to estimate field specific storage of aquifers. The field methods include pumping tests, slug tests, and analyses of sea tides, atmospheric loading and earth tides, and seismic waves. The present state of knowledge concerning the analytical and numerical models and parameter estimating procedures are presented and applications of these models are discussed. Specific storage values from 182 different field sites are collected, covering a wide range of aquifer materials, including unconsolidated deposits and rocks. The geometric mean of specific storage for unconsolidated deposits mainly range from 10⁻⁵ to 10⁻⁴ m⁻¹. The geometric mean of specific storage for different types of rocks mainly range from 10⁻⁷ to 10⁻⁶ m⁻¹. The corresponding hydraulic conductivity shows a wider range of variation than the specific storage. There is not a clear relationship between specific storage and hydraulic conductivity for all the data but a positive correlation can be found for some types of aquifer materials. Specific storage is positively correlated with porosity of the aquifer. A comparison of the different field methods is presented and future research directions are recommended.
Integrated assessment of variable density-viscosity groundwater flow for a high temperature mono-well aquifer thermal energy storage (HT-ATES) system in a geothermal reservoir
2015, Geothermics
Citation Excerpt :
The associated energy loss can be related to the macro-dispersion coefficient which describes groundwater flow due to local variations in the velocity caused by spatial heterogeneity (Nick et al., 2008). Many laboratory studies and tracer measurements (Anderson, 2005; Vandenbohede and Lebbe, 2003, 2006; Vandenbohede et al., 2008, 2009) have illustrated the convection-conduction equation to provide a satisfactory description of the mixing process in groundwater applications. The Peclet number can be applied to assess the most dominant process.
The use of groundwater systems for heat storage increasingly gains interest among water managers, policy makers and researchers as a way to increase the efficiency of energy production and to allow the re-use of waste heat. Typically, mono-well storage systems are thought to require the use of separate aquifers. This study assessed the suitability of using heat and cold storage in a single deep geothermal aquifer for district heating and cooling. An integrated modelling approach was used for evaluating the controls on the energy efficiency of high temperature aquifer thermal energy storage (HT-ATES). The temperature difference (ΔT) of 40 °C between the injection temperatures for the cold and warm storages 20 °C and 60 °C was significant, which required accounting for transient variation of density and viscosity due to temperature and pressure within the modelling code SEAWAT. The developed model was applied for a geothermal reservoir from the Moesian platform, in the Bucharest area, Romania. The sensitivity of the system efficiency was analyzed with respect to the main physical (density, viscosity, longitudinal dispersivity) and operational design parameters (distance between warm and cold storage volumes, flow rates). Uncertainties in geological heterogeneity and the associated range in longitudinal dispersivity values (5–50 m) resulted in significant efficiency differences (80–55%). While reducing the lateral distance between multiple mono-well systems increased their overall efficiency due to positive thermal interference, a minimum vertical distance of (160 m) was required between the injection/extraction filters to prevent interaction between the cold and warm storage volumes. Overall, this study highlights the potential of using a cost-effective mono-well system for HT-ATES in single deep geothermal groundwater systems.
Seawater intrusion processes, investigation and management: Recent advances and future challenges
2013, Advances in Water Resources
Citation Excerpt :
Downhole measurements can sometimes be used to monitor temporal changes in groundwater salinity, although the presence of a metal casing (which acts as an electrical conductor) can be prohibitive. Vandenbohede and Lebbe [324] used a frequency-domain probe to monitor the passage of a plume during a tracer test. Nienhuis et al. [224] compared different methods for detecting the fresh–salt water interface in existing boreholes and found that fixed electrode pairs installed at fixed depths within the borehole provide the most accurate results.
Seawater intrusion (SI) is a global issue, exacerbated by increasing demands for freshwater in coastal zones and predisposed to the influences of rising sea levels and changing climates. This review presents the state of knowledge in SI research, compares classes of methods for assessing and managing SI, and suggests areas for future research. We subdivide SI research into categories relating to processes, measurement, prediction and management. Considerable research effort spanning more than 50 years has provided an extensive array of field, laboratory and computer-based techniques for SI investigation. Despite this, knowledge gaps exist in SI process understanding, in particular associated with transient SI processes and timeframes, and the characterization and prediction of freshwater–saltwater interfaces over regional scales and in highly heterogeneous and dynamic settings. Multidisciplinary research is warranted to evaluate interactions between SI and submarine groundwater discharge, ecosystem health and unsaturated zone processes. Recent advances in numerical simulation, calibration and optimization techniques require rigorous field-scale application to contemporary issues of climate change, sea-level rise, and socioeconomic and ecological factors that are inseparable elements of SI management. The number of well-characterized examples of SI is small, and this has impeded understanding of field-scale processes, such as those controlling mixing zones, saltwater upconing, heterogeneity effects and other factors. Current SI process understanding is based mainly on numerical simulation and laboratory sand-tank experimentation to unravel the combined effects of tides, surface water–groundwater interaction, heterogeneity, pumping and density contrasts. The research effort would benefit from intensive measurement campaigns to delineate accurately interfaces and their movement in response to real-world coastal aquifer stresses, encompassing a range of geological and hydrological settings.

View all citing articles on Scopus

View full text

Combined interpretation of pumping and tracer tests: theoretical considerations and illustration with a field test

Abstract

Introduction

Section snippets

Pumping tests analyses

Interpretation methods derived from analytical models

Inverse numerical model

Combined pumping and tracer test

Practicle example: the Houtave test site

Discussion and conclusions

Acknowledgements

J. Hydrol.

A finite difference method for unsteady flow in variably saturated process media, application to a single pumping well

Water Resour. Res.

Numerical simulation of flow in an aquifer overlain by a water table aquitard

Water Resour. Res.

Effect of a watertable aquitard on drawdown in an underlying pumping aquifer

Water Resour. Res.

A generalized graphical method for evaluating formation constants and summarising well field history

Am. Geophys. Union Trans.

Physical and Chemical Hydrogeology

A dual-domain mass transfer approach for modeling solute transport in heterogeneous aquifers: application to the Macrodispersion Experiment (MADE) site

Water Resour. Res.

Groundwater

Modification of the theory of leaky aquifers

J. Geophys. Res.

Non-steady radial flow in an infinite leaky aquifer

Trans. Am. Geophys. Union