Antonella Sciortino

Inverse modeling for locating dense nonaqueous pools in groundwater under steady flow conditions

In this work we develop an inverse modeling procedure to identify the location and the dimensions of a single-component dense nonaqueous phase liquid (DNAPL) pool in a saturated porous medium under steady flow conditions. The inverse problem is formulated as a least squares minimization problem and solved by a search procedure based on the Levenberg-Marquardt method. Model output is calculated by an existing three-dimensional analytical model describing the transport of solute from a dissolving distributed noise upon the forward model-generated concentration field. We further test the algorithms ability to predict the location and size of a DNAPL pool placed in a controlled three-dimensional bench-scale experiment. In this case we apply the Levenberg-Marquardt algorithm to the minimization of three types of residuals: ordinary residuals, weighted residuals with weights equal to the square of the inverse of the observations, and weighted residuals with weights obtained by adding a constant term to the observed concentrations. The results are sensitive to the location of the observation wells and to the type of residuals minimized. In general, better results in terms of pool location and dimensions were obtained by the minimization of weighted residuals with weights obtained by adding a constant term to the observed concentrations. The results also indicate that the inverse problem is nonunique and nonconvex even in the absence of observation errors. Finally, the sensitivity of the inverse modeling scheme to transport parameter uncertainty was addressed. The inverse solution was found to be extremely sensitive to errors in the dispersion coefficients and relatively insensitive to errors in the mass transfer coefficient.

Modeling Transport in Dual-Permeability Media with Unequal Dispersivity and Velocity

AbstractPorous media such as fractured rock and aggregated soils consist of two pore domains with distinct transport properties. A numerical code was developed to simulate solute concentrations in the two domains using a partitioned solution procedure to efficiently model transport in dual-permeability media. Furthermore, an approximate analytical solution was obtained that allows for different advective and dispersive terms in both flow domains, for a first-type or a third-type inlet condition. Solutions were obtained for local concentrations in both domains as well as effluent concentration and the concentration per medium volume. The problem was solved by decoupling the transport equations using diagonalization. This involves an error for the dispersivity matrix that is related to the difference in dispersivity of both domains. The correctness of the solution was assessed by comparison with numerical results. For low Damkohler numbers the solution was accurate even for a dispersivity ratio of 10. The n...

Experimental and Numerical Investigation of the Effect of Diesel or Ethanol- Blended Diesel on Hydraulic Properties of Unsaturated Soils

Reformulated fuels such as diesel-ethanol blends (“diesohol”) are currently used or being considered in many countries to reduce atmospheric emissions and to reduce the dependence on foreign oil. Additional emulsifiers are usually added to enhance the stability of the mixture. A common concern is that the ethanol and additives, upon (accidental) release of diesohol in the subsurface, may adversely affect the water quality because of increased solubility and decreased degradation of other organic compounds. In this study we investigated the hydraulic properties of unsaturated soils containing an aqueous solution saturated with either diesohol or diesel or containing just plain water. The soil water retention data and the saturated hydraulic conductivity Ks for two common California soils, of the Vista and Los Osos series, were measured. The experimental data showed that the retention parameter α was significantly higher for diesohol compared to plain water and somewhat higher compared to diesel for both soils. With respect to plain water, the value for Ks was slightly lower in the presence of diesohol for the Vista soil, while it was significantly higher in the presence of either diesel or diesohol for the Los Osos soil. This effect cannot be explained by the presence of ethanol because the higher viscosity would result in a commensurate decrease of the saturated hydraulic conductivity. The higher Ks is attributed to changes in soil structure caused by the ethanol. Water flow in the vadose zone was simulated using the HYDRUS 1-D model. The model was applied to simulate a ponding scenario for the Vista and Los Osos soils. Ethanol greatly affected the water retention for the Vista soil, whereas Ks was similar for all three solutions. For the fine-textured Los Osos soil the increase in Ks for diesel and diesohol compared to the plain solution was far more important for the flow regime than the change in retention due to ethanol (diesohol).

Effect of ethanol on water flow in the vadose zone.

Ethanol is increasingly used as an automotive fuel in an effort to diminish the use of fossil fuels and to reduce the emission of contaminants into the atmosphere. This study is concerned with the potential negative impact of the (accidental) release of ethanol to the subsurface. Spillage and leakage of ethanol may occur accidentally during production of blended fuels,, transport, storage, handling, or use. It will often occur in the vadose zone, i.e., from the soil surface to the water table where the soil or rock contains both water and air. Ethanol is highly miscible with water. It will reduce the air-water surface tension and alter the viscosity of aqueous solutions. Therefore, ethanol may affect the transmission and retention of water and dissolved contaminants in the vadose zone. There is an urgent need to quantify this effect. In the present study changes in the hydraulic properties of porous media due to ethanol are examined. In particular, we incorporated the surface tension of aqueous solutions as a function of the ethanol concentration in the water retention function with the “air entry” value. Furthermore the changes in hydraulic conductivity with ethanol content have been quantified through the definition of a function that relates the solution viscosity to ethanol content. The HYDRUS-1D model developed for water flow and solute transport in unsaturated soils was modified by introducing a subroutine that performed the scaling of the water retention curve and the hydraulic conductivity function according to the ethanol concentration. The code was used to model three different scenarios. In the first scenario, ponding and redistribution of pure ethanol in an ethanol-free soil column is compared with ponding and redistribution of pure water. The results show that the amount of pure water that enters the soil is larger than the ethanol and that pure water will move faster than pure ethanol. In the second scenario we compare spillage with no spillage of ethanol at the soil surface followed, in both cases, by a dry atmospheric condition, precipitation, and again dry atmospheric conditions. The ethanol spillage will lead to a higher water content in the upper part of the soil. In the third scenario, we compare water flow in columns with a 10% initial ethanol content (contaminated soil) and without initial ethanol (pristine soil). Both columns are subjected to precipitation (infiltration) followed by dry atmospheric conditions (redistribution). The higher conductivity of the pristine soil causes a rapid infiltration of water during precipitation and advancement of the liquid front during redistribution compared to the contaminated soil.

Vertical Flow Aggregation in the Vadose Zone with Spatial- and Cross-Correlated Hydraulic Properties

Knowledge of hydrological data at or near the soil surface of “large” areas – a field, catchment or watershed – is of considerable interest for the management of surface and subsurface water resources and environmental research dealing with mass transfer at the earth-atmosphere interface. The gap between the local scale for conceptual modeling and observations, and the much larger application and validation scale is bridged with aggregation procedures. Especially for the vadose zone, aggregation is subject to uncertainty due to cross- and auto-correlation of hydraulic properties and nonlinear flow. This study compares a priori aggregation of hydraulic properties with a benchmark a posteriori aggregation of simulation results. Monte Carlo simulations of evaporation were conducted for a synthetic watershed. First, probability density functions (pdf’s) and cross-correlations of hydraulic parameters were established from a set of 126 retention curves and saturated hydraulic conductivities from the UNSODA database. Second, random fields of two hydraulic parameters were generated with imposed auto- and cross-correlation. Third, evaporation was simulated using a stream tube model. Fourth, a posteriori and a priori aggregation were employed to estimate water content and moisture fluxes for different field scales and to quantify the behavior of field-scale mean and variance. A posteriori aggregation yielded the correct mean and a realistic decrease in variance with increasing field scale. On the other hand, a priori aggregation resulted in erratic predictions of the mean and variance of the evaporation flux at the field scale. The cross-correlation between hydraulic parameters enhanced clusters with low or high water contents and evaporation fluxes.