Georgia Destouni

Field scale mass arrival of sorptive solute into the groundwater

The effect of spatial variability in sorption parameters is investigated for field scale solute transport in the unsaturated zone. The considered sorption reaction is kinetically controlled sorption-desorption coupled with a parallel linear equilibrium isotherm. The source of solute is assumed to be distributed over a heterogeneous field where the horizontal extent of the source area is significantly larger than the depth to the ground water table. The transport model is expressed in terms of the mass flux and in dimensionless form, where the effect of equilibrium sorption is incorporated in the dimensionless parameters of the kinetic sorption model. The influence of different correlation properties between the sorption parameters and the hydraulic conductivity at saturation, Ks, on the expected field scale mass flux of solute is illustrated. The results indicate that spatially variable sorption rate coefficients with negative or no correlation to Ks may result in an earlier mass arrival and a considerably altered form of the breakthrough curve compared to the case of constant sorption parameters. In addition, a comparison between the flux-averaged and the resident concentration models indicates that spatial variability in the sorption rate coefficients has similar effects on the temporal variation of the field-scale resident concentration as on the field scale mass flux of solute.

Solute Transport Through an Integrated Heterogeneous Soil‐Groundwater System

The coupled transport process through an integrated soil-groundwater system is quantified for kinetically sorbing solute that originates from a time dependent source at the soil surface and is transported by steady random velocity. The derived expressions of ensemble mean solute breakthrough at some arbitrary control plane normal to the mean flow direction involve probability density functions (pdfs) of advective solute travel time through the unsaturated and the saturated zone of the transport domain. A nonstationary travel time pdf is derived for the saturated zone, to account for possible effects of flow nonuniformity due to recharge of water from the unsaturated zone. Nonuniform mean flow in the saturated zone decreases the relative influence of spatial variability within that zone on the ensemble mean solute breakthrough curve. Factors such as the longitudinal extent of the solute source and the unsaturated zone variability become more important for the spreading of the expected solute breakthrough as the degree of flow nonuniformity in the saturated zone increases. This implies that possible far-field simplifications based on the assumption that the transport process in an integrated soil-groundwater system is dominated by the transport conditions in the saturated zone may not be valid in cases with significant groundwater recharge from the unsaturated zone.

Solute transport through preferential pathways in municipal solid waste.

We modelled lithium transport through an undisturbed solid waste sample and a pilot-scale experimental landfill by use of the same probabilistic Lagrangian approach. The waste sample contained 3.5 m3 of old, well-degraded waste, and the experimental landfill 545 m3 of fresh municipal waste. In the waste sample, four tracer tests were performed under either constant water head or sprinkling input boundary conditions, and either steady state or transient flow. All experimental breakthrough curves (BTCs) exhibit greater spreading than can be explained by a local dispersion mechanism, even when combined with diffusive mass transfer between mobile and immobile water. The observed non-uniform flow and transport could only be represented meaningfully by a two-domain conceptual model, assuming mobile water and advective solute transport both through preferential flow paths and in zones of slow flow. In the undisturbed waste sample, the preferential flow quantification implies that 55–70% of the total infiltrated water moves through only 5–16% of the total water content. In the landfill, it is 90% of the vertically flowing water that moves preferentially through 47% of the total water content. The difference between the landfill and the waste sample appears to be related to waste material properties (which may vary both between and within different waste systems) rather than to the prevailing flow conditions (average flow rate or degree of saturation).

Combined effects of dissolution kinetics, secondary mineral precipitation, and preferential flow on copper leaching from mining waste rock

A probabilistic Lagrangian approach to reactive subsurface transport is used to investigate possible processes that may explain reported discrepancies between modelled and observed copper concentrations at the Aitik waste rock heaps in northern Sweden and to quantify the implications of these processes for the long-term dynamics of copper leaching. The presented Lagrangian transport system couples two different types of primary dissolution kinetics with flow heterogeneity and pH-dependent equilibrium precipitation/dissolution of secondary copper-bearing minerals. Of the investigated processes, which we chose to study based on previous geochemical studies of the site, only flow heterogeneity in the form of preferential flow paths can provide a possible explanation for the low drainage water concentrations of copper currently measured in the field. The investigated pH-dependent precipitation-dissolution of secondary copper bearing minerals provides an explanation for even lower copper concentrations observed in laboratory column experiments in fresh waste rock; these experiments correspond to the initial part of the weathering phase in the field. If both preferential flow and secondary copper precipitation/dissolution occur at the Aitik site, the assumed prevailing type of primary dissolution kinetics does not influence the modelled expected long-term copper leaching nearly as much as it does under homogeneous flow conditions without formation of secondary minerals. The present sensitivity analysis does not provide a comprehensive prediction model for the Aitik site, since not all possible hydrological and chemical processes have been included in the investigation. This study rather exemplifies the application of the probabilistic Lagrangian approach as an investigation methodology for field-scale geochemical problems by using and extending results of detailed geochemical modelling.

Prediction uncertainty in solute flux through heterogeneous soil

The Lagrangian framework presented by Dagan et al. (1992a) is used to analyze the uncertainty in predictions of the field scale mass flux of solute through the unsaturated zone. Transport of both nonreactive and degradable solutes is investigated for input sources that are located at the soil surface of fields with spatially variable hydraulic conductivity at saturation. The variances of the solute flux and accumulated mass, which quantify the corresponding prediction uncertainties, are illustrated at an arbitrary depth below the soil surface for different sizes and shapes of the input domain, and for different flow and degradation conditions. The greatest solute flux variances arise when the expected breakthrough curve has a steep slope. The coefficient of variation for the solute flux is minimum at the peak arrival time of the expected breakthrough curve; this minimum value is relatively insensitive to the assumed distribution for solute travel time and to the loss rate coefficient for degradante solute. The prediction uncertainty decreases with increasing size of the input domain and is smaller for a planar source than for a linear one. The relative uncertainty in the total leached mass of degradable solute increases with increasing loss rate coefficient.

Applicability of the Steady State Flow Assumption for Solute Advection in Field Soils

A comparison between solute travel times predicted by a transient and a steady state flow model is made. Data for five different soil profiles with detailed measurements of their hydraulic properties and their variation with depth are used. Daily measurements of meteorological data are used as input parameters in the transient simulations that include snow and frost dynamics, interception of precipitation, and evapotranspiration. The parameters of the steady state flow model are related to the measured soil properties and the hydrological characteristics of each transient simulation. Furthermore, the influence of solute injection time on the predicted travel time is analyzed, and the effect of root water uptake on the applicability of the steady state flow assumption for solute advection is investigated. The results indicate that the steady state flow model may provide estimates of the mean solute advection that are compatible with those of the transient flow model. The constant rate of recharge in the steady state flow model should then be interpreted as the average annual effective infiltration (i.e., infiltration minus actual evapotranspiration). When root water uptake is accounted for, an arithmetic depth-averaging of the soil parameters appears to yield steady state estimates of arrival time that are closest to the transient predictions. When root water uptake is neglected, a harmonic depth-averaging of the soil parameters provides the best steady state results. The discrepancy between the arrival times predicted with the two flow models decreases with the travel distance from the soil surface.

Comparative analysis of laboratory and field tracer tests for investigating preferential flow and transport in mining waste rock

Abstract A comparison of different tracer tests, conducted in mining waste rock from the Aitik site in northern Sweden, is made by use of temporal moment analysis and advection–dispersion modelling of experimental breakthrough curves. The tracer tests were carried out in both laboratory columns and field lysimeters, and the experimental conditions included both saturated and unsaturated flow, ranging from being close to steady-state to being highly transient. The objectives were to investigate the possible occurrence of preferential flow and the transport behaviour of two tracers, bromide and uranine, under geochemical conditions prevailing in mining waste rock. No indications of significant preferential flow were found in the laboratory columns. In the field, there were indications of water flowing preferentially in about 55–70% of the total water content. In relatively old waste rock, the tracers bromide and uranine behave differently, with uranine being both retained and delayed relative to bromide. These effects on uranine are most probably due to low pH conditions in parts of, or along the entire length of the tracer pathway in the mining waste rock, suggesting that uranine may be useful as an indicator of limited acidic zones within a larger transport domain; in mining waste rock, acidic zones are associated with higher pollutant generation.

Cost effective management of stochastic coastal water pollution

This study develops a theoretical tool for investigating the impact on cost effective coastal water management from explicit treatment of: coastal pollutant transports, stochastic pollutant transports in the catchment areas, and wetlands as a pollutant abatement option. It is applied to a relatively well investigated estuary, Himmerfjärden, south of the Swedish capital, Stockholm. The theoretical results indicate that all three factors influence cost effective allocation of measures and associated design of economic instruments. The consideration of stochastic pollutant transports will increase costs, but the direction of influence of the other two factors cannot be determined without empirical support. The application to nitrogen transport in Himmerfjärden shows that, for target nitrogen reductions given in terms of a percentage of pre-abatement loads, the inclusion of coastal transports in the cost calculations lowers the estimated total costs for targets interpreted in terms of nitrogen loads to the marine water. The alternative investigated target interpretation was in terms of nitrogen loads to coastal waters. Depending on the ability of wetlands to abate nitrogen and to change the variance in pollutant load to the coastal recipients, costs are either increased or decreased as compared to when wetlands are excluded as nitrogen abatement options.

Stochastic modelling of solute flux in the unsaturated zone at the field scale

Abstract Spatial variability in both the saturated hydraulic conductivity and the pore-size distribution is incorporated in a stochastic model of field-scale solute transport through the unsaturated zone. The transport model considered is based on a Lagrangian framework, in which the expected field-scale solute flux and the associated uncertainty in predictions are defined from distributions of solute travel time. Expressions for such travel time distributions are derived from statistical information on measurable soil hydraulic properties and are used to analyze the statistics of the field-scale solute flux in heterogeneous soil. The spatial variability in the saturated hydraulic conductivity and in the pore-size distribution has a similar effect on the prediction uncertainty, quantified by the variance of the solute flux, as on the expected breakthrough curve. The effect on the relative prediction uncertainty, quantified by the coefficient of variation for the solute flux, is therefore small. Both the solute flux variance and the coefficient of variation decrease rapidly with increasing size of the input domain and are considerably smaller for a planar source than for a linear one.

The influence of observation method on local concentration statistics in the subsurface

By introducing the concept of a finest possible time resolution in concentration observations, we extend the probabilistic Lagrangiari transport formulation to account for the statistics of locally measured concentration values and their dependence on observation method for short pulse solute inputs. The outlined methodology for quantifying the ensemble expected value and variance of locally measured concentration values is relevant for transport under either unsaturated or saturated flow conditions and for different types of observation methods. The methodology is exemplified here for direct pore water sampling in aquifers (perfectly stratified/near-field transport, or three-dimensional isotropic and far-field transport) and for solute that is non-reactive or undergoes linear (reversible equilibrium or nonequilibrium, and irreversible) sorption. For nonreactive solute and for solute that undergoes irreversible sorption or equilibrium sorption-desorption, the observation procedure greatly influences the variance of locally measured solute concentrations: both the concentration variance and the coefficient of variation decreases considerably as the sampled water volume/sampling time increases. For solute that undergoes nonequilibrium sorption-desorption, both the concentration variance and the observation effect on the variance are considerably smaller than for nonreactive solute or solute undergoing equilibrium or irreversible sorption. The reason is that the concentration variance is highly dependent on the sorption kinetics, which implies that simple rules, based on average solute behavior, for when the sorption-desorption process can be regarded as being in equilibrium may be irrelevant for describing the concentration variance in heterogeneous fields.