Lower Aquifer Research Articles

Full‐scale demonstration of in situ cometabolic biodegradation of trichloroethylene in groundwater 1. Dynamics of a recirculating well system

Recirculating well systems provide an engine for the in situ treatment of subsurface contaminants. Although numerous recirculating wells have been installed in the field, for such systems, there is a paucity of comprehensive monitoring data and models constrained to data appearing in the research literature. Here we present an extensive data set combined with detailed inverse and simulation analyses for a two‐well groundwater recirculation system used for in situ bioremediation at Edwards Air Force Base in southern California. The “conveyor belt” flow system, which was established for in situ treatment of trichloroethylene (TCE) in two bioactive zones, was created by pumping water upward in one well and downward in another well, each well being screened in both the upper and lower aquifers. A bromide tracer test was conducted and extensively monitored for 60 days. Combined inverse analysis was conducted on hydraulic heads from 38 monitoring wells, 32 bromide concentration histories, and a constraint on the degree of recirculation that was based on TCE concentration data. Four different formulations involving alternative weighting schemes used in a nonlinear weighted least squares simulation‐regression analysis were explored. The best formulation provided parameter estimates with tight bounds on estimated covariances, suggesting that the model provides a reasonable description of the hydrogeologic system. Our investigation indicates the geometry of the recirculation zone and the degree of recirculation under two different sets of operating conditions. Surprisingly, our analysis suggests that the effects of aquifer heterogeneity are not significant at this site under the conditions of forced recirculation. Furthermore, anomalous flow through an open monitoring well created significant vertical short‐circuiting between the generally insulated aquifers. Flow through this small open conduit was equivalent to as much as 33% of the flow through the pumping wells. Using the model as a guide, we treated the aquifer system and bioactive zones as an equivalent mixed reactor to develop simple expressions relating effluent concentrations to influent concentrations. We demonstrate how these expressions are useful in predicting the removal of TCE that had undergone in situ bioremediation in the recirculatory treatment well system. The finite element model developed in this work serves as the foundation for a reactive transport simulator that we developed to analyze bioremediation which occurred during a 444 day experiment [Gandhi et al., 2002].

On the accuracy of parameters determination from SNMR measurements

When performing Surface Nuclear Magnetic Resonance (SNMR) measurements in some regions of Russia where quite conductive layers are encountered, it has been observed that erroneous interpretations were sometimes made with the standard SNMR data inversion scheme that assumes a resistive ground. This has prompted an investigation of the effects of the phase shift induced by conductive layers, using a simplified approach. Quite strong effects of phase shifts occur between SNMR signals from two aquifers, when separated by an electrically conductive layer, which has an influence on the apparent amplitude and decay time of the measured signal. A specific analysis is recommended to avoid errors in the interpretation. Such an analysis is made easier if the equipment used allows larger energising moments. Another case of signal distortion has also been investigated, namely the occurrence of different Larmor frequencies for an upper and a lower aquifer, caused by inhomogeneities of the local earth magnetic field. In order to overcome the difficulty of interpreting SNMR data in those various cases, it is recommended that further developments be made to better take into account both low-resistivity environments and spatial geomagnetic disturbances.

Analytical solutions of tidal groundwater flow in coastal two-aquifer system

This paper presents a complete analytical solution to describe tidal groundwater level fluctuations in a coastal subsurface system. The system consists of two aquifers and a leaky layer between them. Previous solutions of Jacob [Flow of groundwater, in: H. Rouse (Ed.), Engineering Hydraulics, Wiley, New York, 1950, pp. 321–386], Jiao and Tang [Water Resour. Res. 35 (3) (1999) 747], Li and Jiao [Adv. Water Resour. 24 (5) (2001a) 565], Li et al. [Water Resour. Res. 37 (2001) 1095] and Jeng et al. [Adv. Water Resour. (in press)] are special cases of the new solution. The present solution differs from previous work in that both the effects of the leaky layer's elastic storage and the tidal wave interference between the two aquifers are considered. If the upper and lower aquifers have the same storativities and transimissivities, the system can be simplified into an equivalent double-layered, aquifer–aquitard system bounded by impermeable layers from up and down. It is found that the leaky layer's elastic storage behaves as a buffer to the tidal wave interference between the two aquifers. The buffer capacity increases with the leaky layer's thickness, specific storage, and decreases with the leaky layer's vertical permeability. Great buffer capacity can result in negligible tidal wave interference between the upper and lower aquifers so that the Li and Jiao (loc. cit.) solution applies.

The Saltwater/Freshwater Regime in the Sedimentary Cover of the Gorleben Salt Dome

Hydrogeological investigations were conducted around the Gorleben salt dome, in the northeastern part of Lower Saxony in Germany, to study the aquifer system in the cover deposits consisting of unconsolidated Quaternary and Tertiary sediments up to 250m thick. An upper freshwater body is underlain here by saline water. The salinity of the water generally increases with depth. Saturated brines are found in the lower aquifer in a subglacial erosion channel above the salt dome. Hydrochemical analyses and geophysical and isotope studies provide information on development history and behaviour of the flow system. Highly saline groundwater formed by subrosion of salt collects in the lower aquifer above the salt dome. It generally flows north through the erosion channel in the direction of the regional flow field and collects in a basin-like aquifer structure north of the dome. The post-Pleistocene groundwater circulation is progressively replacing the Pleistocene brines with Holocene groundwater. Only a small amount of salt water is being transported into the near-surface aquifer.

Discontinuous flow of fracture water: a new technique of karst collapse prevention

A unique regime of groundwater flow in large limestone aquifers in China is presented in the context of variation of anisotropy of permeability tensors with depth. The limestone aquifers are composed of an upper region where the permeability ellipsoid is upright or prolate and characterized by Kz>Kx and Kz>Ky, a transitional zone, and a lower zone whose permeability ellipsoid is horizontal or oblate and characterized by Kz<Kx and Kz<Ky. The transitional zone partitions the whole limestone into these two upper and lower parts. The principal permeability values Kx, Ky, and Kz in these permeability tensors all decrease simultaneously and logarithmically with depth. When groundwater is discharged out of both the upper part and the lower part of the limestone aquifer by permanent mine drainage at a high yield, an initial regional drawdown cone of groundwater is formed in the aquifer domain. Thus the karst collapse may occur catastrophically. However, a special discontinuous flow environment in this permeable continuum, which we refer to as "double water table flow", could appear in a large limestone aquifer if groundwater is further abstracted out of the lower part of the aquifer through mine drainage even at small yields. Under this condition – where a second water table appears beneath the upper high stage water table due to mine drainage activity at the lower level of the aquifer – the previous regional drawdown cone disappears and the initial groundwater table reappears. Consequently, the high stage of water table is re-established in the upper part of the aquifer, and the karst collapses come to a stop. During the process of tapping the lower aquifer zone, a local drawdown cone, separated from the original groundwater table, gradually takes shape in the lower part of the aquifer. This lower water table migrates downwards with depth as controlled by mine drainage. In this case, the total yield of mine drainage does not increase with depth and approaches a constant value. Although the flow rate of groundwater is uninterrupted between these two parts of the aquifer, an unsaturated region of water cuts off the static pressure in the middle portion of the aquifer. Such an unusual phenomenon can be regarded as a special discontinuous flow condition in a permeable continuum. "Double water table flow" or DWTF is proposed as a new concept in applied hydrogeology and can also be developed into a new technique for karst collapse prevention. It is completely different from some partial discontinuous appearances of groundwater pressure or separations of flow systems that result from a stratiform aquifer combination on groundwater discharge. As the mine drainage changes from the upper part to the lower part in a large limestone aquifer, the original continuous flow (a single water table flow) of fracture water can finally evolve into a discontinuous flow (a double water table flow). The occurrence mechanism of DWTF has been fully demonstrated by the hydraulic simulation of fracture water in an infiltration tank. The prerequisite to the evolution of DWTF in a large limestone aquifer is that the permeability ellipsoid of the lower part of the limestone aquifer must be horizontal or oblate. The economic and environmental benefits of the discontinuous fracture water flow are also briefly pointed out.

Geostatistical mapping of leakance in a regional aquitard, Oak Ridges Moraine area, Ontario, Canada

The Newmarket Till forms a regionally extensive aquitard separating two major aquifer systems in the Greater Toronto area, Canada. The till is incised, and sometimes eroded entirely, by a network of sand- and gravel-filled channels forming productive aquifers and, locally, high-conductivity windows between aquifer systems. Leakage through the till may also be substantial in places. This study investigates the spatial variability of aquitard leakance in order to assess the relative importance of recharge processes to the lower aquifers. With a large database derived from water-well records and containing both hard and soft information, the Sequential Indicator Simulation method is used to generate maps of aquitard thickness and window probability. These can be used for targeting channel aquifers and for identifying potential areas of recharge to the lower aquifers. Conductivities are modeled from sparse data assuming that their correlation range is much smaller than the grid spacing. Block-scale leakances are obtained by upscaling nodal values based on simulated conductivity and thickness fields. Under the "aquifer-flow'' assumption, upscaling is performed by arithmetic spatial averaging. Histograms and maps of upscaled leakances show that heterogeneities associated with aquitard windows have the largest effect on regional groundwater flow patterns.

The significance of buried landscape in subsurface migration of dense non-aqueous phase liquids; the case of perchloroethylene in the Sussex Aquifer, New Brunswick

Abstract At Sussex, New Brunswick, dissolved perchloroethylene (PCE) as high as 28.0 mu g/liter, was detected within the upper aquifer and in concentrations to 1.6 mu g/liter within the lower aquifer that serves as the main source of municipal water supply. Predominantly formed as an esker ridge, the lower aquifer consists of well-sorted, glaciofluvial/ice-contact, sand and gravel layers mostly confined by underlying till and/or bedrock and an overlying unit of laminated silt-clay glaciolacustrine sediments (middle aquitard). The middle aquitard is in turn overlain by an upper aquifer of variable thickness, extending from a depth of about 18 m to near surface, consisting of coarsening-upward sand and gravel, with occasional lenses of clay, silt and diamicton. A dispersal plume was identified by contouring PCE concentrations from 72 samples from the upper aquifer; PCE was not detected in 16 surface soil samples. The highest concentrations imply that the PCE originated from a surface spill near the urban business center and that subsurface migration was influenced significantly by a buried relict river channel cut into the middle aquitard. Scouring along the channel bottom cut through the middle aquitard and intersected the esker ridge of the underlying water supply aquifer, forming windows through the aquitard. Contamination of the lower aquifer has likely originated from the PCE being drawn through these windows during pumping of a water-supply well that is now out of service. Results indicate that withdrawal rates and placement of original municipal water wells, as well as the stratigraphy, conductivity and topography of subsurface units are controlling factors in Dense Non-Aqueous Phase Liquid migration at this site.

Noble gas constraints on hydrocarbon accumulation and groundwater flow in the central area of Western Sichuan Basin

The noble gas concentrations and isotope ratios of seven natural gas samples from the central area of the Western Sichuan Basin were measured. The samples all have40Ar/36Ar ratios greater than the atmospheric values, and the3He/4He ratios (RIR a ) are entirely consistent with the crustal radiogenic He values. The vertical variation of the calculated Ch4/36Ar ratios with depth clearly indicates that the CH4 and36Ar are intimately associated, indicating a common reservoir intermediate to the sampled reservoirs, where they are well mixed and stored together prior to entrapment into gas reservoirs. Meanwhile, the calculated CH4/36Ar ratios range between 8×106 and 64×106 very much greater than the CH4/36Ar values for pure water and 5 mol/L NaCl brine at low temperature and hydrostatic conditions, reflecting the presence of “excess” thermogenic CH4 over that supplied by a CH4-saturated groundwater at low temperature, and the excess CH4 saturation and dissolution to be at depth greater than the sampled reservoirs. This conclusion is consistent with the δ13C(CH4) and δ13C(C2H6) values. In addition, the4He/36Ar ratio is correlated with depth, showing that the crustal radiogenic4He are well mixed with the atomosphere-derived36Ar before introducing into gas reservoirs. The4He/36Ar ratio vertical variation with depth can be attributed to the preferential transport of4He relative to36Ar in fluxes from lower aquifers through water-filled pores into the upper one. The increasing4He/40Ar ratios with decreasing depth, from 1.3 times to 29 times greater than the crustal production ratio, are also assumed to be the results of preferential transport4He relative to40Ar from the production site into groundwater.

Isotope constraints on the hydraulic relationship of ground-waters between Quaternary and Tertiary aquifer in Xi'an area; Shaanxi province

Isotope techniques are increasingly used in evaluating the hydraulic connections between groundwaters in different aquifers, and can improve the geological and hydrogeological database. In the Xi'an area, the upper Quaternary aquifer may have hydraulic connection with the lower aquifer, the Tertiary aquifer, based on geological analyses. δ2H, δ18O,3H, and14C data of groundwater samples provide further evidence for the existence of the hydraulic connection between both aquifers.

Stabilisation de la berge rive nord de l'aménagement hydroélectrique La Grande 1

The La Grande 1 (LG-1) hydroelectrical project, part of the La Grande Complex in Northern Quebec, required the construction of a 2444 m long dyke on the north bank of the river. The presence of sensitive marine clay, covered with deltaic sand and silt deposit, and river sand deposit, called for special design features such as downstream bank and upstream bank stabilization berms to avoid the occurrence of potentially disastrous retrogressive slides. This paper describes the geotechnical and hydrogeological conditions of the northern terrace and presents the different construction phases of the riverbank stabilization, with emphasis on the control of groundwater pressures in the lower aquifer by the use of relief wells.Key words: sensitive clay, river bank, dyke, rockfill, relief well, slope stability.

Numerical simulation of multiphase flow and phase partitioning in discretely fractured geologic media

The three-dimensional compositional model CompFlow has been extended to allow the simulation of the multiphase advective, dispersive and diffusive flux of non-aqueous phase liquid (NAPL) contaminants in a discrete-fracture network, allowing for phase partitioning and dynamic interactions between the fracture network and the surrounding low-permeability rock matrix. The approach used to couple fluxes between the fractures and matrix allows representation of capillary pressure differences within the fractures and matrix and makes no assumption of equilibrium hydraulic conditions between the two. The model is verified for the case of aqueous-phase solute transport by comparison with an analytical solution. An example problem is presented involving the migration of a dense non-aqueous phase liquid (DNAPL) consisting of trichlorethylene (TCE) in a single vertical fracture within a low-permeability material with significant matrix porosity. The simulation results demonstrate that matrix diffusion acts to transfer significant amounts of contaminant to the matrix in the aqueous phase. After the DNAPL source is removed, the NAPL ultimately disappears from the fracture due to partitioning of contaminant into the aqueous phase with concomitant matrix diffusion. It is shown that as the porosity of the matrix increases, the rate of migration of the TCE DNAPL front within fractures is retarded, due to dissolution and matrix diffusion. The sensitivity of DNAPL migration within the fracture to the form of the relative permeability relationship is also discussed. The model is then used to highlight the potential for deep DNAPL penetration through a vertical cross-section consisting of a shallow unconfined sand aquifer and a deeper sand aquifer separated by a layer of fractured clay. Vertical fractures through the clay that hydraulically connects the shallow and deep aquifers are shown to be capable of transmitting both dissolved contaminant and DNAPL to the underlying aquifer, with DNAPL travel times through the 5-m thick clay unit being on the order of days. For the scenarios examined, the DNAPL entering the lower aquifer via the fractured clay unit may completely dissolve at the aquifer–aquitard interface due to lateral groundwater flow in the lower aquifer, provided the aperture of the vertical fractures in the clay layer are less than about 30 μm; however, some DNAPLs may penetrate into the lower aquifer and exist as a non-aqueous phase if the fracture apertures in the clay layer are 50 μm in size or larger. In this latter case, the presence of the DNAPL in the lower aquifer acts as a persistent source of groundwater contamination and can produce an extensive plume in the direction of groundwater flow.

Lakes, groundwater and palaeohydrology in the Sahel of NE Nigeria: evidence from hydrogeochemistry

The geochemistry of natural waters from rain, soil, the unsaturated zone, shallow aquifers, lakes and the stratified aquifer system of the Chad Basin sediments has been investigated to interpret modern hydrological processes and to reconstruct the palaeohydrology of NE Nigeria, a type region of the southern Sahel. Recharge to the confined Middle and Lower aquifers, recorded in NE Nigeria, occurred between 24 and 18.6 ka BP, prior to the last glacial maximum. The mean annual temperature at this time derived from dissolved noble gas ratios was at least 6°C cooler than at the present day. This groundwater is not coupled to the active modern recharge cycle and was not reactivated during the Holocene wet phases as elsewhere in the Saharan region, a reflection of changing lake levels and/or of changing climatic regimes. The absence of groundwater recharge at the time of the last glacial maximum supports other evidence for aridity at this time. Present day direct recharge rates in the Manga Grasslands are high (mean 44 mm a −1 ). However, regional recharge in NE Nigeria at the present day is even higher (60 mm a −1 ) emphasizing the importance of infiltration from surface runoff as input to groundwater. The present study confirms that the shallow aquifer in the region contains significant renewable groundwater resources. However the confined aquifer of the Chad Basin clearly contains palaeowater and this declining artesian basin will need careful conservation prior to a return to traditional water use methods via improved management of the renewable waters in the shallow phreatic aquifers.

Environmental Geology and Petroleum Geology: Problem Solving at the Lexington-Bluegrass Army Depot, Kentucky

Abstract Use of gas chromatography (GC) as a tool for characterizing source rocks and understanding hydrocarbon migration and emplacement is well established in petroleum geology. Petroleum geologists experienced in GC and source-rock characterization, and in other concepts such as structural and stratigraphic trapping, reservoir compartmentalization, emplacement theory, sequence stratigraphy, and diagenesis, can establish themselves as key team players in environmental investigations, especially those involving hydrocarbons. Environmental professionals lacking petroleum geology backgrounds can possibly bias field sampling and analysis of hydrocarbons because they may not consider natural hydrocarbon charging at sites. This problem was recognized recently at the Lexington-Bluegrass Army Depot in Kentucky, where the occurrence of hydrocarbons upgradient from underground storage tanks initially was not adequately deciphered. The nearly 3.25-km2 (800-acre) site, situated near the crest of the Cincinnati Arch, is located on the Middle to Late Ordovician Lexington Limestone. This shaly limestone displays epikarst features and contains two contaminated aquifers traditionally subdivided as the Tanglewood Member (upper aquifer) and the Grier Member (lower aquifer). Original remediation plans included pump and treat, based on the belief that hydrocarbons were anthropogenic. Expensive remedial efforts, however, are being avoided because it has been suggested that a significant portion of the site may be “contaminated” by natural sources. Site re-evaluation should consider (1) retention times required to detect natural hydrocarbons versus refined product; (2) regional structural position of the site; (3) sequence stratigraphic concepts; (4) chemical variations in formation waters versus compartmentalization of aquifers; and (5) expected characteristic GC signatures associated with source rocks in adjacent basins.

Assessment of Hydraulic Parameters of the Aquifers Around the Sea of Galilee

AbstractLack of storativity and transmissivity data from pumping tests in the multilayered, karstified aquifer system discharging into the Sea of Galilee resulted in the use of other techniques for parameter estimation. Water level fluctuations were used to calculate specific‐storage values, which ranged from 1.6 × 10−6 to 6.8 × 10−7 m−1 and from 4 × 10−6 to 2.7 × 10−7 m−1 in the lower and upper aquifer, respectively. Depletion curves were used to calculate transmissivity, which ranged from 0.3 to 24 m2/d and from 1 to 89 m2/d in the lower and upper aquifer, respectively. Age indicators were used to calculate hydraulic conductivity values: those obtained using radiocarbon data (0.54 to 7.5 m/d and 0.4 to 4 m/d for the upper and lower aquifer, respectively) were one order of magnitude higher than those obtained by depletion curve interpretation (10−3 ‐ 0.45 m/d), reflecting either the prevailing lower porosity (0.01) typical of the aquifers, or the impact of regional‐ vs. local‐scale measurements. The higher values obtained using radiocarbon were still a sixth of that calculated from a single pumping test in the area (40 m/d). The pumping test, however, is suggested to provide values which are too high because of leakage from adjacent formations during the test. The alternative techniques for parameter estimation presented here are recommended when pumping test data are lacking (while discharge and water level data do exist), or when regional rather than point transmissivity values are required.

Significance of Defects in Annular Well Seals

The hydraulic significance of defects encountered in annular seals around well casings has been evaluated with numerical ground-water flow and transport models. Contaminant transport was modeled through intact and defective seals surrounding a monitoring well penetrating two permeable formations (aquifers) separated by a less permeable formation (aquitard). The lower aquifer was assumed to be clean initially, whereas the upper aquifer was assumed to be contaminated. The findings indicate that seals should be no more permeable than the least permeable formation through which the well penetrates. Also, seals need to be no longer than the less permeable formations through which a well penetrates, provided the hydraulic conductivity of the seal is equal to or less than that of the less permeable formaton. Occluded defects in the seal within a less permeable formation are hydraulically insignificant provided the less permeable formation is sealed at its upper and lower boundaries.

Combined effect of topography and hydrogeology on subsurface temperature — implications for aquifer permeability and heat flow. A study from the Bohemian Cretaceous basin

The subsurface temperature field of the northern Bohemian Basin (Cretaceous) was studied by using a group of 8 boreholes (depth 75–220 m) drilled in a hydrogeologically active area (2 horizontally layered aquifers separated by an aquiclude) with a pronounced relief (400 m within 2 km from the borehole sites). Reliable temperature logs from closely spaced holes allowed evaluation of both vertical and horizontal temperature trends. The vertical temperature gradient varied from 20 to 60 mK m −1 depending on lithology. In spite of relief the effect of topography was not identified: temperature at one altitude did not vary between the individual boreholes — the isotherms were nearly horizontal. The temperature field was simulated numerically along a 2-D profile to assess the influence of horizontal water flow in the aquifers on temperature. The permeability of aquifers was varied in a broad range. The switch to the convective thermal regime occurred after an increase of permeability from 2 × 10 −12 m 2 to 2 × 10 −11 m 2. The water flow velocity was then sufficient to reduce lateral temperature variations induced by topography so that the model and measured temperatures were in agreement. The temperature field was sensitive especially to permeability changes in the upper aquifer which allowed to constrain its permeability to more than 5–10 × 10 −12 m 2. The vertical temperature gradient in the aquiclude and the lower aquifer was not substantially affected by water flow in any of the models and the gradients determined from the measured temperatures could be therefore used for the heat flow density determination. The estimate (90–100 mW m −2) corresponds well to the increased heat flow in other parts of the Bohemian Cretaceous Basin.

Temperatures and thermal gradients in the Senegalese Maastrichtian aquifer: simulated test on their effect on flow discharge

The Maastrichtian aquifer in the Senegalese sedimentary basin appears completely confined and is characterised by a relatively high geothermal gradient (0.03–0.06°C m −1). In the upper part of the aquifer, temperatures are strongly related to the depth of the roof. At the basin scale, this could mask possible thermal transfer. However, in the SE, thermal anomalies have been observed in some boreholes; moreover, radiocarbon investigations have indicated abnormally long groundwater residence times. Recharge occurs mainly in this zone, near the basement margin, by leakage from the upper aquifer through semi-permeable beds. The natural vertical discharge of the Maastrichtian aquifer occurs beyond the line of equal piezometric level between the upper and lower aquifer. By comparing the aquifer to a porous U-tube and imposing the same input and output, the flow patterns have been studied numerically using the finite element method. Different numerical simulations have involved fluctuations of three quantities: the geothermal gradient, the intrinsic permeability and the input rate. They indicate that, for the mean values observed in the aquifer, natural convection could occur. When the input rate increases, the convective cells have a tendency to break up but significant mixing remains. In such cases, interpretation of residence times with natural isotopes is more difficult.

Hydrogeochemical evidence for surface water recharge to a shallow regional aquifer in northern Victoria, Australia

The Third, Middle and Reedy Lakes system in northern Victoria is a series of interconnected lakes in which groundwater discharge and recharge occur under seasonal climatic and artificial influences. At high lake levels there is a pronounced mounding effect which tends to cause localised perturbation of flow systems and recharge into the uppermost aquifer. Three hydrostratigraphic units (referred to as ‘Lower’, ‘Middle’ and ‘Upper’) are identified beneath the lakes. An upward hydraulic gradient exists in the region under the present hydrologic regime. Despite this, hydrochemical and isotopic evidence has demonstrated a fresh water lens beneath Reedy Lake in the Middle and Lower aquifers. These waters have similar characteristics to the present-day water table which is being recharged from the lakes, suggesting that the lens represents lake-water recharge under a prior hydrologic regime when regional water tables were lower. The recognition of this lens has important implications for future management strategies in that it may be possible to implement a more natural wetting-drying cycle without the intrusion of saline water into this ecologically sensitive area.

The impact of urban development on coastal aquifers near Cotonou, Benin

Mineralization processes and water quality of Quaternary to Mio-Pliocene sanstone aquifers near Cotonou, Benin have been studied using wells sampled during May 1991, August 1991 and April 1992. Anthropogenic pollution, indicated especially by high concentrations of nitrate, P and K, has been detected in the upper aquifer. In contrast, the lower aquifer has acceptable concentrations of solutes, with the exception of the Godomey pumping area, where incipient saline intrusion is detected. Some suggestions are made for aquifer protection based on the careful understanding of groundwater flow directions.

Groundwater–Surface Water Management With Stochastic Surface Water Supplies: A Simulation Optimization Approach

A new simulation optimization model for groundwater–surface water management identifies efficient strategies for meeting water demand and controlling a regional water quality problem, while accounting for varying availability of surface water supplies. Optimal values of model variables are computed for ranges of surface water availability. Expected groundwater levels are computed using a response function approach, which incorporates a probability for each hydraulic stress. Three objective functions are considered: (1) minimize need for supplemental water, (2) minimize imposed water use reductions, and (3) minimize changes from current pumping patterns. The model imposes constraints on expected groundwater levels, expected hydraulic gradients (surrogates for controlling contaminant transport), capacities of pipeline and recharge facilities, and water demand requirements. Capacity and demand constraints must be met every year for all possible surface water conditions. Application to the Santa Clara–Calleguas Basin in southern California indicates that a large quantity of supplemental water or significant pumpage reductions, particularly in the lower aquifer, may be required to control seawater intrusion with current facilities. Supplemental water would be delivered directly to users through pipelines and artificially recharged. Results also indicate that the current artificial‐recharge program has been valuable and that construction of new artificial‐recharge facilities might be beneficial. It should be noted that local agencies are currently developing several potential sources of supplemental water. Monte Carlo simulations yield estimates of the reliability of gradient constraints in controlling advective transport and the likelihood that water level constraints will be violated.