Model analysis of mechanisms controlling pneumatic soil vapor extraction

The efficiency of traditional soil venting or soil vapor extraction (SVE) highly depends on the architecture of the subsurface because imposed advective air flow tends to bypass low-permeable contaminated areas. Pneumatic SVE is a technique developed to enhance remediation efficiency of heterogeneous soils by enforcing large fluctuating pressure fronts through the contaminated area. Laboratory experiments have suggested that pneumatic SVE considerably improves the recovery rate from low-permeable units. We have analyzed the experimental results using a numerical code and quantified the physical processes controlling the functioning of the method. A sensitivity analysis for selected boundary conditions, initial conditions and parameters was carried out to examine how the method behaves under conditions different from the experimental set-up. The simulations show that at the laboratory level the pneumatic venting technology is superior to the traditional technique, and that the method is particularly efficient in cases where large permeability contrasts exist between soil units in the subsurface.     

Experimental investigation of pneumatic soil vapor extraction

Soil Vapor Extraction (SVE) is a common remediation technique for removing volatile organic compounds from unsaturated contaminated soils. Soil heterogeneities can however cause serious limitations to the applicability of SVE due to air bypassing low permeable areas of the soil, leading to diffusion limitation of the remediation. To enhance removal from areas subject to diffusion limitation a new remediation technique, pneumatic soil vapor extraction, is proposed. In contrast to traditional SVE, in which soil vapor is extracted continuously by a vacuum pump, pneumatic SVE is based on enforcing a sequence of large pressure drops on the system to enhance the recovery from the low-permeable areas. The pneumatic SVE technique was investigated in the laboratory using TCE as a model contaminant. 2D-laboratory tank experiments were performed on homogeneous and heterogeneous sand packs. The heterogeneous packs consisted of a fine sand lens surrounded by a coarser sand matrix. As expected when using traditional SVE, the removal of TCE from the low permeable lens was extremely slow and subject to diffusion limitation. In contrast when pneumatic venting was used removal rates increased by up to 77%. The enhanced removal was hypothesized to be attributed to mixing of the contaminated air inside the lens and generation of net advective transport out of the lens due to air expansion.     

Non-aqueous phase liquid spreading during soil vapor extraction

Many non-aqueous phase liquids (NAPLs) are expected to spread at the air-water interface, particularly under non-equilibrium conditions. In the vadose zone, this spreading should increase the surface area for mass transfer and the efficiency of volatile NAPL recovery by soil vapor extraction (SVE). Observations of spreading on water wet surfaces led to a conceptual model of oil spreading vertically above a NAPL pool in the vadose zone. Analysis of this model predicts that spreading can enhance the SVE contaminant recovery compared to conditions where the liquid does not spread. Experiments were conducted with spreading volatile oils hexane and heptane in wet porous media and capillary tubes, where spreading was observed at the scale of centimeters. Within porous medium columns up to a meter in height containing stagnant gas, spreading was less than ten centimeters and did not contribute significantly to hexane volatilization. Water film thinning and oil film pinning may have prevented significant oil film spreading, and thus did not enhance SVE at the scale of a meter. The experiments performed indicate that volatile oil spreading at the field scale is unlikely to contribute significantly to the efficiency of SVE.     

Scaling DNAPL migration from the laboratory to the field

A particular problem with the release of dense nonaqueous phase liquids (DNAPLs) into the environment is identifying where the DNAPL is and if it is still moving. This question is particularly important at sites where thousands of cubic meters of DNAPLs were disposed of. To date, results from laboratory models have not been scaled to predict analogous migration at the larger length and time scales appropriate for sites where large volumes of DNAPLs were released. Modified inspectional analysis is a technique for developing scaling relationships through nondimensionalizing the governing equations. It was applied in this study to scale observations of DNAPL migration in a laboratory model to four hypothetical scenarios in the field where large volumes of DNAPL were released. One scenario was compared to a large DNAPL spill site. The length and time scales of DNAPL movement predicted from our analysis are consistent with those predicted from a numerical model of this site. To our knowledge, this is the first application of modified inspectional analysis for release of DNAPLs in a laboratory model. This methodology may prove useful for scaling results from other laboratory investigations of DNAPL migration to field-scale systems.     

Modeling radium and radon transport through soil and vegetation

A one-dimensional flow and transport model was developed to describe the movement of two fluid phases, gas and water, within a porous medium and the transport of 226Ra and 222Rn within and between these two phases. Included in this model is the vegetative uptake of water and aqueous 226Ra and 222Rn that can be extracted from the soil via the transpiration stream. The mathematical model is formulated through a set of phase balance equations and a set of species balance equations. Mass exchange, sink terms and the dependence of physical properties upon phase composition couple the two sets of equations. Numerical solution of each set, with iteration between the sets, is carried out leading to a set-iterative compositional model. The Petrov-Galerkin finite element approach is used to allow for upstream weighting if required for a given simulation. Mass lumping improves solution convergence and stability behavior. The resulting numerical model was applied to four problems and was found to produce accurate, mass conservative solutions when compared to published experimental and numerical results and theoretical column experiments. Preliminary results suggest that the model can be used as an investigative tool to determine the feasibility of phytoremediating radium and radon-contaminated soil.     

Effect of water content on transient nonequilibrium NAPL-gas mass transfer during soil vapor extraction

The effect of water content on the volatilization of nonaqueous phase liquid (NAPL) in unsaturated soils was characterized by one-dimensional venting experiments conducted to evaluate the lumped mass transfer coefficient. An empirical correlation based upon the modified Sherwood number, Peclet number, and normalized mean grain size was used to estimate initial lumped mass transfer coefficients over a range of water content. The effects of water content on the soil vapor extraction SVE process have been investigated through experimentation and mathematical modeling. The experimental results indicated that a rate-limited NAPL-gas mass transfer occurred in water-wet soils. A severe mass transfer limitation was observed at 61.0% water saturation where the normalized effluent gas concentrations fell below 1.0 almost immediately, declined exponentially from the initiation of venting, and showed long tailing. This result was attributed to the reduction of interfacial area between the NAPL and mobile gas phases due to the increased water content. A transient mathematical model describing the change of the lumped mass transfer coefficient was used. Simulations showed that the nonequilibrium mass transfer process could be characterized by the exponent beta, a parameter which described the reduction of the specific area available for NAPL volatilization. The nonequilibrium mass transfer limitations were controlled by the soil mean grain size and pore gas velocity, were well described by beta values below 1.0 at low water saturation, and were well predicted with beta values greater than 1.0 at high water saturation.     

A reaction-based paradigm to model reactive chemical transport in groundwater with general kinetic and equilibrium reactions

This paper presents a reaction-based water quality transport model in subsurface flow systems. Transport of chemical species with a variety of chemical and physical processes is mathematically described by M partial differential equations (PDEs). Decomposition via Gauss-Jordan column reduction of the reaction network transforms M species reactive transport equations into two sets of equations: a set of thermodynamic equilibrium equations representing N(E) equilibrium reactions and a set of reactive transport equations of M-N(E) kinetic-variables involving no equilibrium reactions (a kinetic-variable is a linear combination of species). The elimination of equilibrium reactions from reactive transport equations allows robust and efficient numerical integration. The model solves the PDEs of kinetic-variables rather than individual chemical species, which reduces the number of reactive transport equations and simplifies the reaction terms in the equations. A variety of numerical methods are investigated for solving the coupled transport and reaction equations. Simulation comparisons with exact solutions were performed to verify numerical accuracy and assess the effectiveness of various numerical strategies to deal with different application circumstances. Two validation examples involving simulations of uranium transport in soil columns are presented to evaluate the ability of the model to simulate reactive transport with complex reaction networks involving both kinetic and equilibrium reactions.     

Gel barrier formation in unsaturated porous media

The gel barrier formation by a gelling liquid (Colloidal Silica) injection in an unsaturated porous medium is investigated by developing a mathematical model and conducting numerical simulations. Gelation process is initiated by adding electrolytes such as NaCl, and the gel phase consisting of cross-linked colloidal silica particles grows as the gelation process proceeds. The mathematical model describing the transport and gelation of Colloidal Silica (CS) is based on coupled mass balance equations for the gel mixture (the sol phase plus the gel phase), gel phase (cross-linked colloidal silica particles plus water captured between cross-linked particles), and colloidal silica particles (discrete and cross-linked) and NaCl in the sol (suspension of discrete colloidal silica particles in water) and gel phases. The solutions in terms of volumetric fraction of the gel phase yield the gel mixture viscosity via the dependency on the volumetric fraction of gel phase. This dependency is determined from a kinetic gelation model with time-normalized viscosity curves. The proposed model is verified by comparing experimentally and numerically determined hydraulic conductivities of gel-treated soil columns at different CS injection volumes. The numerical experiments indicate that an impermeable gel layer is formed within the time period twice the gel-point in a one-dimensional flow system. At the same normalized time corresponding to twice the gel-point, the CS solutions with lower NaCl concentrations result in further migration and poor performance in plugging the pore space. The viscosity computation proposed in this study is compared with another method available in the literature. It is observed that the other method estimates the viscosity at the mixing zone higher than the one proposed by the authors. The proposed model can simulate realistic injection scenarios with various combinations of operating parameters such as NaCl concentration and NaCl mixing time, and thus providing guidelines in performing this technology on site.     

Influence of rate-limited sorption on the cleanup of layered soils by vapor extraction

The conditions under which rate-limited sorption is important for cleanup of layered soils by vapor extraction are investigated. The investigation includes two steps: (a) the cleanup time is estimated for a number of scenario cases by means of a numerical model and (b) the numerical results are approximated using analytical solutions derived for simplified models. In this way, equations are derived, which give insight into the influence of different parameters characterizing the properties of the soil, the geometry of the formation, the mass transfer mechanisms in it, and the distribution of the contaminant mass in the different phases (gas phase, water phase and solid phase). The numerical model used is based on the advection-dispersion differential equations for Darcian isothermal airflow, local equilibrium contaminant mass transfer between gas phase and soil water and first-order kinetics for mass transfer between soil water and solid phase. The numerical results are approximated combining an analytical solution to estimate cleanup time in layered formations for local equilibrium sorption, which has been presented in a previous work (J. Contam. Hydrol., 36 (1999) 105). with an analytical solution based on the well-mixed reservoir model under consideration of rate-limited sorption. The analytical approximation of the cleanup time is in reasonable agreement with the numerical results and allows its estimation with small computational effort.     

Axi-symmetric simulation of soil vapor extraction influenced by soil fracturing

Fracturing, either pneumatic or hydraulic, is a method to improve the performance of soil vapor extraction (SVE) in relatively low permeability soils (< 10(-5) cm/s). A two-dimensional model is presented to simulate trichloroethylene (TCE) soil vapor extraction modified by fracturing. Flow and transport is modeled using mobile macropore and micropore networks, which also have been identified in the literature as dual porosity, dual permeability, or heterogeneous flow models. In this model, fluids can flow in both the macropore and micropore networks. This represents a more general model compared to immobile micropore, mobile macropore models presented thus far in the literature for vapor flow and transport in two dimensions. The model considers pressure- and concentration-driven exchange between the macropore and micropore networks, concentration-driven exchange between the gas and sorbed phases within each network, and equilibrium exchange between the gas and water and a sorbed phase within each network. The parameters employed in an example simulation are based on field measurements made at a fractured site. Considered in the simulations were the influence of the volume percentage of fractures, the length of fractures, the relative location of the water table, and the influence of pulsed pumping. For these simulations, internetwork concentration-driven exchange most significantly affected mass removal. The volume percentage of fractures more significantly influence flow and mass removal than the length of fractures. The depth of the water table below the contamination plume only significantly influenced flow and mass removal when the water table was within 60 cm of the bottom of the contaminated soil in the vadose zone for the parameters considered in this study. Pulsed pumping was not found to increase the amount of mass removed in this study.     

A multi-flowpath model for the interpretation of immiscible displacement experiments in heterogeneous soil columns

This work focuses on the phenomenon of the immiscible two-phase flow of water and oil in saturated heterogeneous soil columns. The goal is to develop a fast and reliable method for quantifying soil heterogeneities for incorporation into the relevant capillary pressure and relative permeability functions. Such data are commonly used as input data in simulators of contaminant transport in the subsurface. Rate-controlled drainage experiments are performed on undisturbed soil columns and the transient response of the axial distribution of water saturation is determined from electrical measurements. The transient responses of the axial distribution of water saturation and total pressure drop are fitted with the multi-flowpath model (MFPM) where the pore space is regarded as a system of parallel paths of different permeability. The MFPM enables us to quantify soil heterogeneity at two scales: the micro-scale parameters describe on average the effects of pore network heterogeneities on the two-phase flow pattern; the macro-scale parameters indicate the variability of permeability at the scale of interconnected pore networks. The capillary pressure curve is consistent with that measured with mercury intrusion porosimetry over the low pressure range. The oil relative permeability increases sharply at a very low oil saturation (< 10^− 3) and tends to a high end value. The water relative permeability decreases abruptly at a low oil saturation (~ 0.1), whereas the irreducible wetting phase saturation is quite high. The foregoing characteristics of the two-phase flow properties are associated with critical (preferential) flowpaths that comprise a very small percentage of the total pore volume, control the overall hydraulic conductivity, and are consistent with the very broad range of pore-length scales usually probed in soil porous matrix.     

Inverse modeling to estimate LNAPL plume release timing

Methodologies are presented for dating releases of light nonaqueous phase liquids (LNAPLs) using an inverse modeling approach with simple analytical models. Models for LNAPL plume migration are presented to predict LNAPL plume velocity in the unsaturated and saturated zones as a function of basic soil and fluid properties. A relative mobility factor is introduced for LNAPL movement at the water table that depends primarily on the van Genuchten n parameter (related to the breadth of the soil pore size distribution) and the magnitude of water table fluctuations. Estimated LNAPL plume velocities compare reasonably with more rigorous numerical models, which may be used in cases where data availability warrant the greater effort entailed.Two methods of estimating release timing and its uncertainty are investigated. A direct estimation method is described that determines travel time for a single observed travel distance based on estimated soil and fluid properties. Release date uncertainty may be determined using the first order (FO) or Monte Carlo (MC) methods. The second method for estimating release date involves nonlinear parameter estimation utilizing distance vs. time measurements and other data.A case study is presented for a field site where independent estimates of release timing were obtained from a numerical modeling analysis. Release timing estimates based on direct inversion of the analytical timing model agree well with the numerical analysis. Results for a second field site indicate that release date confidence limits estimated by the FO method, assuming log-normally distributed travel times, are close to values determined by the MC method, which makes no assumption regarding the form of the travel time probability distribution.Results for a hypothetical problem indicate that LNAPL velocity and travel time may be accurately estimated if sufficient data on travel distance vs. time are available. Incorporating prior information on relevant soil and fluid properties into the objective function reduces the uncertainty in release date if prior estimates are accurate. However, biased prior estimates may lead to over- or underestimation of release date uncertainty. Simultaneous estimation of soil and fluid properties and release date is possible if prior information is available to condition the parameter estimates.     

土壤、地下水中有机污染物的就地处置

有机化合物对土壤、地下水的污染已引起世界各国的普遍关注.地层介质中的有机物主要以自由态、挥发态、溶解态和固态4种形态存在.有机污染物的自然降解能力较差,如不进行人工清除,在自然环境中它们可能存留长达几十年之久,对土壤、地下水资源构成长期的威胁.传统的开挖处理技术不仅费用昂贵,而且当贮油设施的地表被利用时则无法进行开挖处理(如有建筑物等).近年来,以地下冲洗法、土壤抽水法和地下水曝气法为代表的有机污染物就地处置技术得到了迅速的发展.本文对这3种技术进行概要的介绍,总结指出决定这些技术可能性的主要因素是地层介质的通透能力,有机物的挥发、溶解能力及其可生物降解能力,并列出目前的主要有机污染物挥发、溶解及生物降解能力的相对强弱作为制定具体处置技术的参考指标.     

Electrokinetic ion transport through unsaturated soil: 1. Theory, model development, and testing

An electromigration transport model for non-reactive ion transport in unsaturated soil was developed and tested against laboratory experiments. This model assumed the electric potential field was constant with respect to time, an assumption valid for highly buffered soil, or when the electrode electrolysis reactions are neutralized. The model also assumed constant moisture contents and temperature with respect to time, and that electroosmotic and hydraulic transport of water through the soil was negligible. A functional relationship between ionic mobility and the electrolyte concentration was estimated using the chemical activity coefficient. Tortuosity was calculated from a mathematical relationship fitted to the electrical conductivity of the bulk pore water and soil moisture data. The functional relationship between ionic mobility, pore-water concentration, and tortuosity as a function of moisture content allowed the model to predict ion transport in heterogeneous unsaturated soils. The model was tested against laboratory measurements assessing anionic electromigration as a function of moisture content. In the test cell, a strip of soil was spiked with red dye No 40 and monitored for a 24-h period while a 10-mA current was maintained between the electrodes. Electromigration velocities predicted by the electromigration transport model were in agreement with laboratory experimental results. Both laboratory-measured and model-predicted dye migration results indicated a maximum transport velocity at moisture contents less than saturation due to competing effects between current density and tortuosity as moisture content decreases.     

Laboratory analog and numerical study of groundwater flow and solute transport in a karst aquifer with conduit and matrix domains

New mathematical and laboratory methods have been developed for simulating groundwater flow and solute transport in karst aquifers having conduits imbedded in a porous medium, such as limestone. The Stokes equations are used to model the flow in the conduits and the Darcy equation is used for the flow in the matrix. The Beavers–Joseph interface boundary conditions are adopted to describe the flow exchange at the interface boundary between the two domains. A laboratory analog is used to simulate the conduit and matrix domains of a karst aquifer. The conduit domain is located at the bottom of the transparent plexiglas laboratory analog and glass beads occupy the remaining space to represent the matrix domain. Water flows into and out of the two domains separately and each has its own supply and outflow reservoirs. Water and solute are exchanged through an interface between the two domains. Pressure transducers located within the matrix and conduit domains of the analog provide data that is processed and stored in digital format. Dye tracing experiments are recorded using time-lapse imaging. The data and images produced are analyzed by a spatial analysis program. The experiments provide not only hydraulic head distribution but also capture solute front images and mass exchange measurements between the conduit and matrix domains. In the experiment, we measure and record pressures, and quantify flow rates and solute transport. The results present a plausible argument that laboratory analogs can characterize groundwater water flow, solute transport, and mass exchange between the conduit and matrix domains in a karst aquifer. The analog validates the predictions of a numerical model and demonstrates the need of laboratory analogs to provide verification of proposed theories and the calibration of mathematical models.     

The effect of multicomponent diffusion on NAPL dissolution from spherical ternary mixtures

This paper investigates the dissolution characteristics of ternary nonaqueous phase liquid (NAPL) mixtures with the goal of comparing the relative contributions of multicomponent (intra-NAPL) diffusion, film transfer and thermodynamic nonideality. These contributions are compared at the pore scale and intermediate scale (several centimeters downstream from the source). Trichloroethene (TCE), tetrachloroethene (PCE) and 1,1,1-trichloroethane (TCA) were selected to model a reasonably ideal mixture; TCE, PCE and octanol were selected as a relevant nonideal mixture. A multicomponent diffusion-based dissolution model incorporating hydrodynamic theory was formulated to estimate intra-NAPL concentration gradients and associated aqueous interfacial concentrations for ideally shaped (spherical) NAPL blobs. Pore scale dissolution times for this model were compared to those generated using the conventional well-mixed NAPL dissolution model, applying the same film transfer boundary condition in both cases. Activity coefficients (spatially and temporally variable for the diffusion model, temporally variable for the well-mixed model) were estimated using UNIFAC. NAPL interfacial concentration histories generated using the pore scale models were used as input in a three-dimensional groundwater transport model (MT3DMS) to compare downstream concentration distributions. For the relatively large NAPL bodies simulated (r=0.6 cm), intra-NAPL diffusion effects were found to be significant at the pore scale and were strongly impacted by the mixture's thermodynamic ideality. At the intermediate scale, and for the conditions tested, modest differences in the simulations suggested that intra-NAPL diffusion effects would be negligible compared to those associated with mixture composition uncertainty, dissolution rate processes related to NAPL-induced permeability effects and hydrodynamic issues associated with flow field heterogeneity.     

Volatile chlorinated hydrocarbons in Antarctic superficial snow sampled during Italian ITASE expeditions

In order to detect the presence of some volatile chlorinated hydrocarbons (VCHCs) and to understand their transport and deposition mechanism, superficial snow was sampled during two Italian ITASE (International Trans Antarctic Scientific Expedition) expeditions: the first traverse was carried out in 1998/1999 from Terra Nova Bay to Dome Concordia; the second traverse was carried out in 2001/2002 through Adélie, George V, Oates and Northern Victoria Lands. Some VCHCs (chloroform; 1,1,1-trichloroethane; tetrachloromethane; 1,1,2-trichloroethylene; tetrachloroethylene) were analysed using a highly sensitive and selective hyphenated technique composed of a purge-and-trap injector coupled to a gas chromatograph with a mass spectrometric detector (PTI-GC-MS) operating in SIM mode. Investigated VCHCs were present in all analysed snow samples with concentration levels of several units, tens, or sometimes hundreds of ng kg(-1). VCHC snow concentration levels remained approximately constant with changing distance from the coast and the comparison between fresh and aged snow did not show any substantial differences; on the basis of this evidence marine aerosol and dry deposition may be rejected as principal VCHC transport and deposition mechanism hypotheses. VCHC concentration levels in Antarctic snow samples were comparable to or greater than those found in snow from temperate zones.     

Modelling of atrazine transport in the presence of surfactants

Laboratory experiments were conducted to examine the effect of detergents on transport of atrazine [2-chloro-4-(ethylamino)-6-(isopropylamino)-s-triazine] through loam and sandy loam soils under saturation conditions. The Convection Dispersion Equation (CDE) was used to model and quantify the effects of detergents on atrazine model parameters: the retardation factor (R), pore velocity (v) and dispersion coefficient (D). The transport parameters were estimated using moment technique and partition coefficient obtained from batch experiments and compared with best-fitted parameters, R and D, keeping pore velocity constant. Results indicated the CDE model was not successful in predicting atrazine transport in the presence of surfactants at high concentrations. In the case of anionic surfactant with Elora loam, the average predicted R and D from moment technique of 3.4 and 11.1 cm2/h, respectively were significantly different than fitted parameters (R = 39 and D = 227 cm2/h). The poor performance of CDE in the presence of surfactants results from physiochemical changes in herbicide solubility and retention to the soil matrix rather than changes in soil hydraulic properties since the predicted pore water velocities from moment technique were similar to those measured during leaching experiments. Nevertheless, BTC analysis with CDE showed that land application of anionic surfactant (sulphonic) significantly increased R and D and decrease v for both soils. Addition of sulphonic increased R of atrazine by 12 and 26 folds for loam and sandy loam soils, respectively. On the other hand non-ionic surfactants seemed to decrease R, especially in sandy loam soil, thus facilitating atrazine leaching through soil. Non-equilibrium conditions seemed to govern atrazine transport in the presence of surfactants; double peaks in breakthrough curves were observed, indicating a need for mathematical models to account for such phenomena. Atrazine dispersion and tailing seemed to be higher through Elora loam compared to Caledon sandy loam due to higher aggregation of the Elora soil.     

Effects of pore volume-transmissivity correlation on transport phenomena

The relevant velocity that describes transport phenomena in a porous medium is the pore velocity. For this reason, one needs not only to describe the variability of transmissivity, which fully determines the Darcy velocity field for given source terms and boundary conditions, but also any variability of the pore volume. We demonstrate that hydraulically equivalent media with exactly the same transmissivity field can produce dramatic differences in the displacement of a solute if they have different pore volume distributions. In particular, we demonstrate that correlation between pore volume and transmissivity leads to a much smoother and more homogeneous solute distribution. This was observed in a laboratory experiment performed in artificial fractures made of two plexiglass plates into which a space-dependent aperture distribution was milled. Using visualization by a light transmission technique, we observe that the solute behaviour is much smoother and more regular after the fractures are filled with glass powder, which plays the role of a homogeneous fault gouge material. This is due to a perfect correlation between pore volume and transmissivity that causes pore velocity to be not directly dependent on the transmissivity, but only indirectly through the hydraulic gradient, which is a much smoother function due to the diffusive behaviour of the flow equation acting as a filter. This smoothing property of the pore volume-transmissivity correlation is also supported by numerical simulations of tracer tests in a dipole flow field. Three different conceptual models are used: an empty fracture, a rough-walled fracture filled with a homogeneous material and a parallel-plate fracture with a heterogeneous fault gouge. All three models are hydraulically equivalent, yet they have a different pore volume distribution. Even if piezometric heads and specific flow rates are exactly the same at any point of the domain, the transport process differs dramatically. These differences make it important to discriminate in situ among different conceptual models in order to simulate correctly the transport phenomena. For this reason, we study the solute breakthrough and recovery curves at the extraction wells. Our numerical case studies show that discrimination on the basis of such data might be impossible except under very favourable conditions, i.e. the integral scale of the transmissivity field has to be known and small compared to the dipole size. If the latter conditions are satisfied, discrimination between the rough-walled fracture filled with a homogeneous material and the other two models becomes possible, whereas the parallel-plate fracture with a heterogeneous fault gouge and the empty fracture still show identifiability problems. The latter may be solved by inspection of aperture and pressure testing.