摘要:

Certain geologic media are known to have small permeability; subsurface environments composed of these media and lacking well developed secondary permeability have groundwater flow sytems with many distinctive characteristics. Moreover, groundwater flow in these environments appears to influence the evolution of certain hydrologic, geologic, and geochemical systems, may affect the accumulation of pertroleum and ores, and probably has a role in the structural evolution of parts of the crust. Such environments are also important in the context of waste disposal. This review attempts to synthesize the diverse contributions of various disciplines to the problem of flow in low‐permeability environments. Problems hindering analysis are enumerated together with suggested approaches to overcoming them. A common thread running through the discussion is the significance of size‐ and time‐scale limitations of the ability to directly observe flow behavior and make measurements of parameters. These limitations have resulted in rather distinct small‐ and large‐scale approaches to the problem. The first part of the review considers experimental investigations of low‐permeability flow, including in situ testing; these are generally conducted on temporal and spatial scales which are relatively small compared with those of interest. Results from this work have provided increasingly detailed information about many aspects of the flow but leave certain questions unanswered. Recent advances in laboratory and in situ testing techniques have permitted measurements of permeability and storage properties in progressively “tighter” media and investigation of transient flow under these conditions. However, very large hydraulic gradients are still required for the tests; an observational gap exists for typical in situ gradients. The applicability of Darcy's law in this range is therefore untested, although claims of observed non‐Darcian behavior appear flawed. Two important nonhydraulic flow phenomena, osmosis and ultrafiltration, are experimentally well established in prepared clays but have been incompletely investigated, particularly in undisturbed geologic media. Small‐scale experimental results form much of the basis for analyses of flow in low‐permeability environments which occurs on scales of time and size too large to permit direct observation. Such large‐scale flow behavior is the focus of the second part of the review. Extrapolation of small‐scale experimental experience becomes an important and sometimes controversial problem in this context. In large flow systems under steady state conditions the regional permeability can sometimes be determined, but systems with transient flow are more difficult to analyze. The complexity of the problem is enhanced by the sensitivity of large‐scale flow to the effects of slow geologic processes. One‐dimensional studies have begun to elucidate how simple burial or exhumation can generate transient flow conditions by changing the state of stress and temperature and by burial metamorphism. Investigation of the more complex problem of the interaction of geologic processes and flow in two and three dimensions is just beginning. Because these transient flow analyses have largely been based on flow in experimental scale systems or in relatively permeable systems, deformation in response to effective stress changes is generally treated as linearly elastic; however, this treatment creates difficulties for the long periods of interest because viscoelastic deformation is probably significant. Also, large‐scale flow simulations in argillaceous environments generally have neglected osmosis and ultrafiltration, in part because extrapolation of laboratory experience with coupled flow to large scales under in situ conditions is controversial. Nevertheless, the effects are potentially quite important because the coupled flow might cause ultra long lived transient conditions. The difficulties associated with analysis are matched by those of characterizing hydrologic conditions in tight environments; measurements of hydraulic head and sampling of pore fluids have been done only rarely because of the practical difficulties involved. These problems are also discussed i

ISSN:0043-1397    

DOI:10.1029/WR022i008p01163

年代:1986

数据来源: WILEY

摘要:

A heuristic solution procedure is developed for capacity expansion problems encountered in water resources development. The heuristic solution procedure consists of two steps. In the first step, project sequence is generated and capacity refinement is done in the second step. A shortest path dynamic programming solution scheme is used as a “return function” for the sequence generation step, and a univariate direct search algorithm is used for capacity refinement. The proposed solution procedure is tested for both capacitated and uncapacitated example problems with a finite planning horizon. The results indicate that the proposed heuristic solution procedure is superior to other solution procedures selected for comparison in finding an optimum expansion policy consisting of a sequence of project developments, investment timing, and sizing for each proj

ISSN:0043-1397    

DOI:10.1029/WR022i008p01197

年代:1986

数据来源: WILEY

摘要:

A model for simulating microbial growth‐degradation processes in porous media is developed. It is assumed that the bulk of microorganisms in an aquifer grow in microcolonies attached to matrix surfaces. As developed, the model applies to the growth and decay of aerobic, heterotrophic microorganisms whose growth is limited by lack of a carbon and energy source (substrate), an oxygen source or both simultaneously as described by modified Monod kinetics. Transport of substrate and oxygen in the porous medium is assumed to be governed by advection‐dispersion equations with surface adsorption. A total of five coupled equations result describing substrate and oxygen concentrations in the pore fluid, substrate and oxygen concentrations in the microcolonies and colony density, which is assumed sufficiently small so that aquifer hydraulic conductivity is not diminished. An iterative process involving an Eulerian‐Lagrangian numerical procedure that is highly resistant to numerical dispersion in the presence of small dispersivities is used to solve the overall model, with parameter values selected from the literature or estimated. Results indicate that biodgradation would be expected to have a major effect on contaminant transport when proper conditions for growth exist. For one‐dimensional transport in a column, the most rapid microbial growth always occurred at the influent boundary where oxygen and substrate concentrations were held constant independent of colony density. Anaerobic conditions develop rapidly and aerobic biodegradation ceases if large amounts of substrate are added to the

ISSN:0043-1397    

DOI:10.1029/WR022i008p01207

年代:1986

数据来源: WILEY

摘要:

Detailed vertical profiles of the horizontal component of the specific discharge at the water table region of a phreatic aquifer were obtained, from two research wells, employing both a modified point dilution technique under natural flow conditions and a mathematical diffusion model. H218O‐depleted water was used as a tracer. A 2.5‐m multilayer sampler with dialysis cells at 3‐cm intervals was utilized both to introduce the tracer into groundwater and determine its temporal variations. Large variations in the profiles of the tracer concentration were observed in a relatively thin layer just below the water table. The horizontal specific discharge was found to increase with depth from 0.5 m/year in the first 50 cm below the water table to 4.5 m/year at 240 cm depth. Two theories are suggested to explain the presence of an almost stagnant zone at the water table region: (1) the existence of entrapped‐migrating air bubbles which reduce the hydraulic conductivity and (2) drag forces exerted by the capillary

ISSN:0043-1397    

DOI:10.1029/WR022i008p01217

年代:1986

数据来源: WILEY

摘要:

The solution for a general case of flow in a bounded two‐aquifer system with an aquitard and only one aquifer being pumped has been developed. The exact solution for the typical case of a homogeneous circular system with only one well at the center has been found. These results were obtained without imposing the usual restriction of vertical flow in the aquitard. These results provide a theoretical and practical basis for evaluating ground water resources and problems of oil production and the effects of water injection in multilayered reservoirs. They can also be used in solving problems of land subsidenc

ISSN:0043-1397    

DOI:10.1029/WR022i008p01225

年代:1986

数据来源: WILEY

摘要:

A new integral formulation, based on Green's second identity, is used to solve the unsteady transport (diffusion‐advection) equation which governs the storage and movement of pollutants in porous media. It uses the fundamental solution for the terms of the differential equation with the highest derivatives (the order of the equation) and, by applying Green's second identity, those terms are cast into a boundary integral while the remaining terms with lower derivatives are weighted with that fundamental solution and integrated over the solution domain. The resulting integral representation is no longer a boundary integral but comprises both boundary and domain integral portions. Because the terms with the highest derivatives constitute the Laplacian, the fundamental solution employed in the formulation is the logarithmic function. The boundary integral portion of the formulation is similar to that encountered when the boundary element method is applied to elliptic equations but the domain portion, which contains the temporal derivative and advection terms, is evaluated by discretizing the solution domain into elements, as is frequently done in finite element formulations. The classic one‐dimensional semi‐infinite and a two‐dimensional semi‐infinite diffusion‐advection problems are solved for a wide range of local Peclet numbers in order to verify and demonstrate the usefulness of the present formulation. The results are s

ISSN:0043-1397    

DOI:10.1029/WR022i008p01237

年代:1986

数据来源: WILEY

摘要:

Recent field investigations in crystalline rock give strong indications that water flows in largely isolated channels in fissured rock. The present paper sets out to investigate radionuclide transport in such channels including diffusion into the rock matrix. It is shown that the uptake from a cylindrical hole in a matrix is much more effective, per unit contact area, than a flat surface (semi‐infinite solid). Comparison of diffusion from cylindrical and slit‐formed channels shows that the differences in interfacial flux are minor. A slit‐formed channel may therefore be approximated by a cylindrical one. Diffusional transport in the matrix is then treated as one dimensional instead of two dimensional, which gives a considerable numerical simplification. A simple estimate of the “penetration depth” into the cylindrical region, i.e., the location of the diffusional front, is proposed. By using the integrated finite difference method a number of radionuclide transport calculations are performed for the case of flow and dispersion in a cylindrical channel coupled to diffusion and sorption in t

ISSN:0043-1397    

DOI:10.1029/WR022i008p01247

年代:1986

数据来源: WILEY

摘要:

In this paper we describe a simple combination of variances technique for incorporating the spatial variability of infiltration rates into the irrigation uniformity predictions of border irrigation models. The mean and variance of the depth of infiltrated water across a border can be calculated directly from the infiltration opportunity time results of a single border irrigation simulation and the distributions of the infiltration parameters for the border. Infiltration depth results are not needed. The single irrigation simulation is run using the expected values for the infiltration parameters to describe uniform infiltration across the border. We also show how these results can be used to calculate the irrigation distribution uniformity coefficientDU, where care must be used if the depths of infiltration are assumed normally distributed. Although only normal and lognormal distributions are discussed, the techniques described can be applied to any infiltration parameter distribution.

ISSN:0043-1397    

DOI:10.1029/WR022i008p01257

年代:1986

数据来源: WILEY

摘要:

Production of storm runoff in highly responsive catchments is not well understood. We report in these papers a comprehensive set of hydrometric and natural tracer data for rainfall, soil water, and streamflow for catchments in the Tawhai State Forest, Westland, New Zealand, which reveal some of the important runoff processes. The catchments are small (<4 ha), with short (<300 m) steep (average 34°) slopes and thin (<1 m) permeable soils. Long‐term (1977–1980) weekly observations of oxygen 18, electrical conductivity, and chloride in the stream, groundwater, and rain in the main study catchment indicate that catchment outflow reflects a well‐mixed reservoir with a mean residence time of approximately 4 months. A preliminary storm hydrograph separation using oxygen 18 (for a storm hydrograph exceeded by only 22% of events since 1979) indicates that only 3% of storm runoff could be considered “new” (i.e., current storm) water. Rapid subsurface flow, such as macropore flow, of new water therefore cannot explain streamflow response in the study area. More detailed hydrograph separation studies on throughflow as well as streamflow are described in parts 2 (M. G. Sklash et. al., this issue) and 3 (M. G. Sklash et. al., unpublished manuscr

ISSN:0043-1397    

DOI:10.1029/WR022i008p01263

年代:1986

数据来源: WILEY

摘要:

Previous hydrometric and dye tracer studies in Maimai 8, a highly responsive catchment in the Tawhai State Forest, Westland, New Zealand, suggest that storm runoff generation is by rapid transmission of “new” (current storm rain) water to the stream via macropores. We used hydrometric and natural tracer (oxygen 18, deuterium, electrical conductivity, chloride) observations in two first‐ and one second‐order stream and in six throughflow pits, to evaluate the roles of “old” (stored) and new water during three storm events (return periods from 4 weeks to 3 months) in September 1983. New water contributions were small (<25% of the hydrograph volume) and could be accounted for by saturation overland flow. Hillslope response varied areally but all sites issued old water‐dominated throughflow. Ridge top sites had larger new water contributions (∼ 30–40%) than valley sites (<10%). Macropore flow of new water therefore cannot explain streamflow or throughflow response in the

ISSN:0043-1397    

DOI:10.1029/WR022i008p01273

年代:1986

数据来源: WILEY