Dual-porosity modelling of the pore structure and transport properties of a contaminated soil

We have developed a new method to characterize the pore structure of mineral soils. We combined data from the analysis of back-scattered scanning electron microscope (BSEM) images of resin-impregnated pore-casts, and mercury intrusion porosimetry (MIP) data, with analytical percolation models and inverse modeling algorithms. The pore space is regarded as a dual-pore network consisting of a primary Euclidean pore-and-throat network and a secondary, fractal, pore system that is accessed through primary pores. The digitized 2-D BSEM images of resin-impregnated soil samples are employed to determine the autocorrelation function. The Fourier transform of this function provides the small-angle neutron scattering (SANS) intensity function, which is extended by using the surface fractal dimension obtained from high-pressure MIP data. Inversion of the extended scattering intensity function produces the volume-based radius distribution function of spherical pore bodies (PBRD). The complete volume-based PBRD is fitted with a composite number-based PBRD composed of a lognormal primary PBRD and a power (fractal) secondary PBRD with upper and lower cut-offs. Based on the concepts of invasion percolation, an analytic mathematical model that describes Hg intrusion into dual pore networks is developed. The complete PBRD and pore-throat radius distribution (PTRD) functions of the primary network along with the drainage accessibility functions (DAFs) of the primary and secondary pore networks are estimated with inverse modelling of the Hg intrusion curve. Based on critical path analysis of percolation theory, approximate analytical relationships are developed to calculate explicitly the absolute permeability and electrical formation factor from the geometrical and topological parameters of the primary pore network. The method is demonstrated with application to four soil samples.     

Effect of leaching of DOC on pore characteristic of a sandy soil

Changes of microstructural properties of a sandy soil from a former sewage farm due to the leaching of dissolved organic carbon (DOC) were studied. The leaching of DOC at various pH values was induced using sodium hydroxide or hydrochloric acid solutions. The removal of DOC altered the pore properties of the remaining solid. The average pore radius measured from water vapor desorption isotherms (micropores range) increased with the increase of the removed DOC under alkaline conditions and did practically not alter under acid treatment. The average pore radius measured by the mercury intrusion porosimetry (mesopores range), decreases, however, for low amounts of the DOC extracted and increases on extracting higher DOC amounts. The mesopores appeared to be fractal. At moderate amounts of the DOC leached the finer mesopores occurred having the higher fractal dimensions than the coarser mesopores.     

STABILITY OF SOIL PORES DURING MERCURY INTRUSION POROSIMETRY

Mercury intrusion porosimetry is one of the techniques used to measure pore size distributions of dry soil samples. It is known that incorrect pore size distributions can be obtained for some porous materials due to sample damage occurring during mercury intrusion, but little evidence relating to soil samples has been published. The pore size distributions of aggregates from three soils, differing in texture, strength and porosity, were measured after heating to 300 °C by mercury intrusion porosimetry. After the first intrusion, all of the intruded mercury was removed from each sample by heating under vacuum. Samples were then re-intruded. The pore size distribution curves for both intrusions were very similar, indicating an absence of sample damage during the first intrusion run.     

Comparison of fractal dimensions of soils estimated from adsorption isotherms,mercury intrusion,and particle size distribution

The values of the surface fractal dimensions were determined for several samples of Cambisols and Luvisols from analysis of nitrogen and water vapor adsorption isotherms and from mercury intrusion data. Moreover, the values of fractal dimension characterizing the particle size distributions of soil samples were calculated by using a number‐based method. For almost all investigated soils the values of the surface fractal dimension, obtained from water vapor isotherms were lower than those obtained from nitrogen isotherms. Largest were the surface fractal dimensions evaluated form mercury intrusion data. No significant correlations between different kinds of surface fractal dimensions were found, and the reasons of this finding are discussed. However, the values of the surface fractal dimensions calculated from mercury intrusion data correlate with those characterizing the texture of soils. The paper also reports on correlations between of the values of surface area, fractal dimensions and some selected physico‐chemical characteristics of soils.     

Structure and self-similarity in silty and sandy soils: the fractal approach

Soil structure was studied using the concept of fractals and related to soil texture and aggregate properties such as surface charges and aggregate stability. The mass and porosity fractal dimensions (D_m and D_p) of silty and sandy soils were determined on in situ soils using a variety of soil sections (thin, very-thin and ultra-thin), by image analysis on a continuous scale from m to 10^?9 to 10^?1m. Surface fractal dimensions (D_s) of these soils were determined on < 2 mm air-dried samples using mercury porosimetry and the fractal cube generator model. The results suggest that soils are not pore fractals but mass and surface fractals with D_m= 1.1 D_s when the dimension of the embedding Euclidean space d is 3. The soil structures could possibly be described by fractal diffusion-limited aggregation with complex interconnected aggregates or by fractal cluster–cluster aggregation models. As a preliminary conclusion, the fractal approach appears to be a potentially useful tool for understanding the underlying mechanisms in the creation or destruction of soil structure.     

The relation between the moisture-release curve and the structure of soil

The limit to the precision with which moisture-release data may be related to soil structure is examined. Thin sections in two orthogonal planes were prepared from cores for which moisture release data and saturated conductivity measurements were also collected. Structural parameters are inferred indirectly through estimation of the mass fractal dimension from the moisture-release curve, and directly through measurement of the scaling properties of the pore space and solid matrix from the thin sections using image analysis. Theoretical results are presented which show that, in the absence of additional information, the interpretation of the moisture-release curve is ambiguous for several reasons. A power-law exponent is a consequence of either a fractal pore volume; a fractal solid volume; a fractal pore wall; or a non-fractal, self-similar pore wall, and one cannot infer from the measurement which is the case. Experimental results are presented which confirm that direct measurement of the fractal dimension of the solid matrix is a good predictor of the Brooks–Corey exponent for the soils studied here. Therefore, although a specific structural parameter characterizes the moisture-release curve, the latter cannot be used as a detailed measure of soil structure.     

MEASUREMENT OF PORE SIZES IN FINE-TEXTURED SOILS: A REVIEW OF EXISTING TECHNIQUES

Pore size distributions obtained from the relationship between moisture content and suction are not dependable in fine-textured soils because of shrinkage. To overcome this problem, methods such as nitrogen sorption, mercury intrusion porosimetry, non-polar liquid desorption and thin sectioning have been used. In order to pre-dry samples without changes in the pore system, freeze-drying, organic liquid replacement of soil water, and critical point drying techniques have been employed. These methods of soil drying and pore size measurement are described and compared, and the validity of their use in soil studies is examined. The measurement of pore sizes by water desorption is also discussed.     

Effect of humic acids,sesquioxides and silica on the pore system of silt aggregates measured by water vapour desorption,mercury intrusion and microtomography

Most of the information on soil aggregation and porosity comes from studies of natural soil in which the effects of the different constituents that form the structure overlap. The aim of this research was to study the effects of these constituents separately on well‐characterized artificial aggregates in order to understand them better. To do this, the pore system of model silt aggregates, amended with different amounts of humic acids, iron and aluminium hydroxides or colloidal silica, was investigated at three levels of magnification with water vapour desorption (nanometre sizes), mercury intrusion (micrometre sizes) and microtomography (tens of micrometres). Humic acid and aluminium hydroxide increased aggregate porosities measured by all methods. An increase in porosity with increasing additions of each constituent was indicated only by water desorption. We did not observe any well‐defined trends in the dynamics of average pore radii. The pore surface fractal dimension determined by mercury intrusion was negatively correlated with that measured by water desorption. The pore system in granular media comprises larger voids joined by narrower necks; therefore, we attempted to relate their sizes with a novel approach that combined microtomography with mercury intrusion and extrusion data. We observed a decrease in the size of pore necks that give access to voids of the same sizes with increasing additions of all constituents. With additions of humic acid this effect was the smallest. The mercury intrusion data showed the formation of separate concretions of iron hydroxides and silica in silt aggregates.     

Three-dimensional analysis of a loamy-clay soil using pore and solid chord distributions

Examples of pore and solid chord distributions obtained for ideal porous media are presented, and the distributions of the porous and solid phases of a soil have been studied by pore and solid chord distributions. Serial sections, 100 μm apart, were cut in a soil core, impregnated with resin, and images were obtained of them. The 2D images from 160 sections were used to build a 3D reconstruction of the core. The initial 2D images, the 3D reconstructed medium and 2D computed images from the latter were studied. We found that the solid matrix of the particular soil is homogeneous and isotropic at the scale studied, and it could be characterized with a single 2D image. For pores ranging from 500 μm to 2 mm we also found similar pore chord distributions for the 2D images in the three orthogonal directions and the 3D medium. A single 2D image can be used to study these pores. For larger pores more than one 2D image is required, and we showed that eight 2D images are sufficient to describe these pores.     

From fractal analysis along a line to fractals on the plane

Self-similar fractals are useful models for soil solid and pore sets. The scaling properties of these fractals along a line and across an area can be described by the fractal dimensions. One method, for estimating the soil areal fractal dimension from the solid and pore set distributions along the lines, was proposed and tested with real macro- and micromorphological data on three soils of Mexico. The soil areal fractal dimension (D_a) was compared with the soil mass fractal dimension (D_m) estimated by two-dimensional binary image analysis, separately for solids and pores. Both methods are based on the box-counting technique and are suitable for determining the soil ‘box' or ‘capacity' fractal dimension, that seems to be apt to estimate the alternative filling of an area by a fractal set of solids and pores. This paper examines the relations between the fractal dimensions obtained along a line, across an area and directly from the image. Analysis of D_a and D_m data seems to suggest that both soil genesis and management practice can contribute to areal fractal dimension dynamics. It was shown that fractal dimensions are useful parameters able to monitor tillage influence on soil properties and to estimate the degree of soil compaction.     

Multiscale percolation properties of a fractal pore network

Network modelling and percolation theory now occupy a firm niche in soils research, aiding in the simulation and prediction of flow and transport processes. In this study we examine the connectivity properties of a simple parameterized 2D fractal network model by direct numerical simulation and theoretically, drawing on upscaling functions from renormalization theory. We demonstrate good agreement between the two approaches and extrapolate to systems inaccessible to direct numerical simulation. We examine the relationship between porosity, the pore-size range and the connectivity of the network. We show that for fixed porosity, the connectivity is a nonlinear function of the range of scales accommodated in the fractal model. We investigate the implications of distributing a fixed porosity over different ranges of pore size and demonstrate a general impairment of pore connectivity and improvement of solid phase continuity as this range is extended.     

Shrinkage and drainage in aggregates of volcanic soils: a new approach combining mercury porosimetry and vacuum drying kinetics

The topsoils of Andisols in the wet high altitude grasslands (páramos) of Ecuador have shrunk after recent changes from permanent grassland to agriculture. The question arises as to whether all Andisols in the region behave in the same way. We have therefore studied the shrinkage on drying of topsoils from four representative types of Andisol in the páramos using a combination of drying kinetics and mercury porosimetry. We aimed to identify and quantify the pore volumes that participate in either drainage or shrinkage or both. The key concept of hierarchical dual porosity of soils used to interpret the shrinkage curves was partly validated, but we also identified more complex relations between drainage, shrinkage and structure. We found that the older was the soil, the more weathered it was, and, consequently, the greater its aggregation and porosity. In turn, the total volumetric shrinkage was controlled by the initial void ratio of the wet soil samples. Our results also show that the solid–pore interfaces of the volcanic soil aggregates were less accessible after shrinkage than before.     

Conservation Agriculture Had a Poor Impact on the Soil Porosity of Veneto Low‐lying Plain Silty Soils after a 5‐year Transition Period

Conservation agriculture practices have been proposed as a set of techniques for improving soil structure properties and related ecosystem services. This study compared conservation agriculture (CA) practices (no‐tillage, cover crop and residue retention) and conventional intensive tillage system in order to evaluate their effects on total porosity, pore size distribution, pore architecture and morphology. The experiment was set up in 2010 on four farms of the low‐lying Veneto Region plain characterized by silty soils. Almost hundred soil samples were collected in 2015 at four depths down to 50‐cm layer and investigated for porosity from micrometre (0·0074 μm) to macrometre (2·5 mm) by coupling mercury intrusion porosimetry and X‐ray computed microtomography (μCT). Indices of soil morphology and architecture were derived by analysing 3D images and mercury intrusion porosimetry pore size curves. Results suggested that silty soils of Veneto plain are microstructured because much (82%) of the porosity ranged between 0·0074 and 30 μm. CA practices positively influenced the ultramicroporosity class (0·1–5 μm) (1·86E‐01 vs 1·67E‐01 μm³ μm⁻³) that is strictly linked to soil organic carbon stabilization while no effects were observed in X‐ray μCT porosity domain (> 26 μm). Silty soils of Veneto plain showed a slow reaction to CA because of the poor aggregate stability and low soil organic carbon. However, the positive response of the ultramicropore fraction indicates that a virtuous cycle was initiated between soil organic carbon and porosity, hopefully leading to well‐developed macropore systems and, in turn, enhanced soil functions and ecosystem services. Copyright © 2017 John Wiley & Sons, Ltd.     

Probability density function: A tool for simultaneous monitoring of pore/solid roughness and moisture content

We present a novel approach to multiscale fractal image analysis for monitoring the dynamics of the soil pore/solid network roughness due to moisture content changes. Roughness of the gray-level probability density function (PDF) of subsequent images of the drying soil was expressed in terms of its Hurst exponent (H_PDF), which correlated significantly with the soil dielectric permittivity and gravimetric water content, as well as with micro-horizon depth and the temporal progress of drying. We documented an intermittent character of the water content dynamics, correlating with oscillations of the pore/solid interface roughness. Our technique of PDF roughness analysis of digital images, exemplified here for soil moisture monitoring, can be applied for monitoring the other complex systems, for instance plant growth in greenhouse.     

Investigation into the fractal scaling of the structure and strength of soil aggregates

The hierarchical nature of soil structure is examined by measuring the physical properties of a range of aggregate sizes obtained using repetitive fracture. Fractals are used to assess the change with aggregate size of the specific volume, the proportion of pre-existing cracks which link to form the aggregate failure surface, and the aggregate failure stress. The pore size distribution, evaluated using mercury porosimetry and the application of the box counting algorithm to thin sections and thick sections, is also used to obtain a fractal dimension, D. Our results show that D depends upon the measurement approach for mass fractal scaling. This finding may limit the application of fractals to predict the scaling behaviour of soil physical properties.     

基于Menger海绵模型的煤矸石粉改良膨胀土微结构特征

为解决膨胀土对工程结构以及农业生态环境的危害，进行煤矸石粉改良膨胀土的试验研究。对煤矸石粉掺量为0、3%、6%、9%的膨胀土土样进行压汞试验，测得微观孔隙特征值；选取Menger海绵模型建立孔隙分分形模型，计算土体孔隙分形维数，探究土体孔隙分形维数与孔隙特征参数以及煤矸石粉掺量变化的关系。结果表明：随着煤矸石粉掺量增加，土中大孔隙所占的含量较素膨胀减少61.5%，孔隙类型从团粒间孔隙转化为颗粒间孔隙；煤矸石粉的掺入改变了土体的孔隙结构特征，煤矸石粉与膨胀土发生胶结反应，孔隙连通性降低，使得总孔隙体积、孔隙率、孔隙平均孔径、孔隙临界孔径等孔隙特征参数呈减小趋势；基于分形理论分析孔隙分形维数，分形维数随煤矸石粉掺量的增加而增加，且与孔隙特征参数呈显著相关性。孔隙分形维数反应了孔隙特征参数以及孔隙发育程度，为土的孔隙表征提供方法借鉴。     

Soil structure and the effect of management practices

To evaluate the impact of management practices on the soil environment, it is necessary to quantify the modifications to the soil structure. Soil structure conditions were evaluated by characterizing porosity using a combination of mercury intrusion porosimetry, image analysis and micromorphological observations. Saturated hydraulic conductivity and aggregate stability were also analysed.

In soils tilled by alternative tillage systems, like ripper subsoiling, the macroporosity was generally higher and homogeneously distributed through the profile while the conventional tillage systems, like the mouldboard ploughing, showed a significant reduction of porosity both in the surface layer (0–100 mm) and at the lower cultivation depth (400–500 mm). The higher macroporosity in soils under alternative tillage systems was due to a larger number of elongated transmission pores. Also, the microporosity within the aggregates, measured by mercury intrusion porosimetry, increased in the soil tilled by ripper subsoiling and disc harrow (minimum tillage). The resulting soil structure was more open and more homogeneous, thus allowing better water movement, as confirmed by the higher hydraulic conductivity in the soil tilled by ripper subsoiling. Aggregates were less stable in ploughed soils and this resulted in a more pronounced tendency to form surface crust compared with soils under minimum tillage and ripper subsoiling.

The application of compost and manure improved the soil porosity and the soil aggregation. A better aggregation indicated that the addition of organic materials plays an important role in preventing soil crust formation.

These results confirm that it is possible to adopt alternative tillage systems to prevent soil physical degradation and that the application of organic materials is essential to improve the soil structure quality.     

Applicability of cavity-throat connecting model for estimating the hydraulic conductivity of fine-grained soils: a geometrical and mathematical approach

Purpose

Determining the hydraulic conductivity of low permeable fine-grained soils is difficult and time-consuming. This work develops a new method with an eye to the pore morphology to correlate hydraulic conductivity with pore-size distribution (PSD) parameters obtained from mercury porosimeter data. In order to realize this method, calculating percolation loss along the flow paths in pore channels and quantifying the spatial morphology of pore channels by proposing a cavity-throat connecting model is necessary.

Materials and methods

In order to establish the standard process of the new method, a kind of sedimentary mucky clay with regular dual-structural PSD has been collected. The samples are divided into three series: (a) vibrated with variable frequencies; (b) frozen at variable temperatures and unfrozen, making the freezing-thawing effect as the variable; and (c) remolded with different water contents. The PSD of freeze-dried samples at the end of each process is obtained by mercury intrusion porosimetry. After that, the method is demonstrated with application to 12 series of fine-grained soils.

Results and discussion

Deduced from mercury porosimeter data, the volume-based PSD curves of fine-grained soils are bimodal, due to the presence of inter-aggregate and intra-aggregate pores. Two important hypotheses have been proposed: (i) one is that in the smaller pore scales, the experimental extrusion curve controlled by the hysteresis loop has a really approximate part compared to the theoretical overall retraction curve, making the experimental extrusion curve characterize the pore cavity size approximately, and (ii) the pore system consists of a series of multistage cavity-throat connections. Accumulating the effects of single connection on the percolation can be used to measure the overall effects of pore system on the percolation. Based on fluid-driven path analysis of percolation, the pore system is quantified by a series of cavity-throat connections and the percolation loss has been derived to estimate the hydraulic conductivity.

Conclusions

The permeable parameter (κ) representing the overall effects of pore connections on the hydraulic conductivity (K) is suited to correlate the microstructure and hydraulic conductivity by the linear relationship with the fixed slope in semilogarithmic coordinate for the fine-grained soils. It is the destruction and recombination of cavity-throat connections that are dominant during the treatments like freezing, remolding, and reinforcing.

     

Effect of extreme acid and alkali treatment on pore properties of soil samples

Samples of six Polish and three Korean soils were acidified and alkalized with elevated concentrations of hydrochloric acid or sodium hydroxide from 0.001 to 1 mol dm^—3. The pore system of the studied soils was investigated using mercury intrusion (”︁macropores”) and water vapor adsorption (”︁mesopores”) experiments. The characteristics of the pores and their surfaces were very sensitive against acid and alkaline treatments. The macropore volume and radii increased in general with increasing of the concentrations of the treatment solutions. The macropore surface of all the clay rich soils exhibited a fractal behavior. The ranges of fractal scaling differed for particular soils. Macropore surface fractal dimensions changed under both treatments. For a given soil, the macropore fractal dimensions changed in the same direction under the effect of acid as under alkali. Both treatments decreased the mesopore volumes of Polish and increased the mesopore volumes of Korean soils. However, this was not true for two of Korean soils after extreme alkali treatments. In general, the average mesopore radius increased under lower treatments concentrations and decreased under higher concentrations. The mesopores were also fractal. The range of their fractal scaling was different for the Polish and Korean soils and roughly the same within each group. The fractal dimension of the mesopores decreased after acid treatment showing a smoothing of the material porous surface. Alkaline treatment affected the mesopore fractal dimension to a lower extent. A slight rise of the mesopore fractal dimension was noted in most cases.<?show $6#>     

用显微CT研究不同植被恢复模式的土壤团聚体微结构特征

为了更好了解不同植被恢复模式对土壤团聚体微结构的影响,该研究采用显微CT技术扫描3~5 mm土壤团聚体,获取了3.25μm分辨率的二维图像,并应用数字图像处理软件对团聚体孔隙结构进行三维重建,定量研究了黄土丘陵区不同植被恢复模式下(自然草地、人工灌木和坡耕地)土壤团聚体微结构特征。结果表明,两种植被恢复模式均显著提高了土壤有机碳含量和团聚体水稳性(P0.05),降低了土壤容重。与坡耕地处理相比,自然草地土壤团聚体总孔隙度、大孔隙度(100μm)、瘦长型孔隙度分别增加了20%、23%和24%,而分形维数和连通性指数欧拉特征值分别降低了2%和75%,且各指标二者间差异均显著(P0.05)。人工灌木土壤团聚体的上述各项孔隙参数均优于自然草地(较坡耕地分别增加了70%、88%和43%以及降低了4%和92%),且除欧拉特征值外,差异均显著(P0.05)。分形维数和连通性对土壤结构变化的响应相当敏感,可作为该地区植被恢复过程中土壤质量评价的指标,研究结果可为黄土高原土壤质量评价提供科学参考。