Stochastic analysis of solute transport in partially saturated heterogeneous soil: 1. Numerical experiments

In an endeavour to describe quantitatively the water flow and solute transport in soils and other heterogeneous porous media, various different approaches have been introduced in the past decades, including double porosity, double permeability and other multiple-continua approaches. Recently, a promising methodology to identify experimentally the pore structure of porous media has been proposed, where a discrete distribution of effective pore radii is established based on saturated flow experiments with non-Newtonian (shear-thinning) fluids, as described by Abou Najm and Atallah (2016) and in other works. In this particular concept, the porous media is idealised as a bundle of capillaries with only a reasonably small number of distinct values of their radii. This allows to identify the pore radii and the contributions of the corresponding pore groups to the total flow by performing and evaluating a reasonable number of flow experiments.In an attempt to understand better the relation of the effective discrete pore radii distribution concept (with a given number of distinct pore radii allowed) to the structure of the porous media, we perform numerical experiments with other idealised geometries of the pore space. The saturated flow experiments with shear-thinning fluids are simulated by finite element method and then, based on the resulting flow, the discrete pore radii distributions are established and compared with the original geometry. For simplicity, we stick to one-dimensional models analogous to Poiseuille or Hagen-Poiseuille flow. The idea is to examine pore size distributions that are continuous rather than discrete, while keeping the advantage of a perfectly controlled and comprehensible idealised geometry. This in-silico approach may later serve as a supporting tool for studying various aspects of the addressed experimental methodology, e.g., in taking into account realistic non-Newtonian rheology, proposing an optimal set of experiments, or contemplating links with solute transport models.

Download Full-text

Comparison of alternative models for simulating anomalous solute transport in a large heterogeneous soil column

Journal of Hydrology ◽

10.1016/j.jhydrol.2009.08.036 ◽

2009 ◽

Vol 377 (3-4) ◽

pp. 391-404 ◽

Cited By ~ 43

Author(s):

Guangyao Gao ◽

Hongbin Zhan ◽

Shaoyuan Feng ◽

Guanhua Huang ◽

Xiaomin Mao

Keyword(s):

Solute Transport ◽

Soil Column ◽

Alternative Models ◽

Heterogeneous Soil

Download Full-text

Solute Transport Through an Integrated Heterogeneous Soil-Groundwater System

Water Resources Research ◽

10.1029/95wr01330 ◽

1995 ◽

Vol 31 (8) ◽

pp. 1935-1944 ◽

Cited By ~ 66

Author(s):

Georgia Destouni ◽

Wendy Graham

Keyword(s):

Solute Transport ◽

Groundwater System ◽

Heterogeneous Soil

Download Full-text

Comparison between field measurements and numerical simulation of steady-state solute transport in a heterogeneous soil profile

Hydrology and Earth System Sciences ◽

10.5194/hess-1-853-1997 ◽

1997 ◽

Vol 1 (4) ◽

pp. 853-871 ◽

Cited By ~ 8

Author(s):

J. Vanderborght ◽

D. Jacques ◽

D. Mallants ◽

P.-H. Tseng ◽

J. Feyen

Keyword(s):

Solute Transport ◽

Water Flow ◽

Random Fields ◽

Hydraulic Properties ◽

Soil Hydraulic Properties ◽

Field Scale ◽

Flow Heterogeneity ◽

Heterogeneous Soil ◽

Soil Hydraulic ◽

Second Procedure

Abstract. Abstract: Field-scale solute dispersion is determined by water flow heterogeneity which results from spatial variability of soil hydraulic properties and soil moisture state. Measured variabilities of soil hydraulic properties are highly sensitive to the experimental method. Field-scale dispersion derived from leaching experiments in a macroporous loam soil was compared with field-scale dispersion obtained with numerical simulations in heterogeneous random fields. Four types of random fields of hydraulic properties having statistical properties derived from four different types of laboratory measurements were considered. Based on this comparison, the measurement method depicting heterogeneities of hydraulic properties most relevant to field-scale solute transport was identified. For unsaturated flow, the variability of the hydraulic conductivity characteristic measured on a small soil volume was the most relevant parameter. For saturated flow, simulated dispersion underestimated the measured dispersion and it was concluded that heterogeneity of macroscopic hydraulic properties could not represent solute flow heterogeneity under these flow conditions. Field-scale averaged solute concentrations depend both on the detection method and the averaging procedure. Flux-averaged concentrations (relevant to practical applications) differ from volume-averaged or resident concentrations (easy to measure), especially when water flow is more heterogeneous. Simulated flux and resident concentrations were subsequently used to test two simple one-dimensional transport models in predicting flux concentrations when they are calibrated on resident concentrations. In the first procedure, solute transport in a heterogeneous soil is represented by a 1-D convection dispersion process. The second procedure was based on the relation between flux and resident concentrations for a stochastic convective process. Better predictions of flux concentrations were obtained using the second procedure, especially when water flow and solute transport are very heterogeneous.

Download Full-text