The freezing temperature of water in soil is not constant but varies over a range determined by soil texture. Consequently, the amounts of unfrozen water and ice change with temperature in frozen soil, which in turn affects hydraulic, thermal, and mechanical properties of frozen soil. In this paper, an Am-241 gamma ray source and time-domain reflectometry (TDR) were combined to measure unfrozen water content and ice content in frozen soil simultaneously. The gamma ray attenuation was used to determine total water content. The TDR was used to determine the dielectric constant of the frozen soil. Based on a four-phase mixing model, the amount of unfrozen water content in the frozen soil could be determined. The ice content was inferred by the difference between total water content and unfrozen water content. The gamma ray attenuation and the TDR were both calibrated by a gravimetric method. Water contents measured by gamma ray attenuation and TDR in an unfrozen silt column under infiltration were compared and showed that the two methods have the same accuracy and response to changes of water content. Unidirectional column freezing experiments were performed to apply the combined method of gamma ray attenuation and TDR for measuring unfrozen water content and ice content. The measurement error of the gamma ray attenuation and TDR was around 0.02 and 0.01 m3/m3, respectively. The overestimation of unfrozen water in frozen soil by TDR alone was quantified and found to depend on the amount of ice content. The higher the ice content, the larger the overestimation. The study confirmed that the combined method could accurately determine unfrozen water content and ice content in frozen soil. The results of soil column freezing experiments indicate that total water content distribution is affected by available pore space and the freezing front advance rate. It was found that there is similarity between the soil water characteristic and the soil freezing characteristic of variably saturated soil. Unfrozen water content is independent of total water content and affected only by temperature when the freezing point is reached.

Simultaneous measurement of unfrozen water content and ice content in frozen soil using gamma ray attenuation and TDR

[1] Gas hydrate formation and dissociation in sediments are accompanied by changes in the bulk volume of the sediment and can lead to changes in sediment properties, loss of integrity for boreholes, and possibly regional subsidence of the ground surface over areas where methane might be produced from gas hydrate in the future. Experiments on sand, silts, and clay subject to different effective stress and containing different saturations of hydrate formed from dissolved phase tetrahydrofuran are used to systematically investigate the impact of gas hydrate formation and dissociation on bulk sediment volume. Volume changes in low specific surface sediments (i.e., having a rigid sediment skeleton like sand) are much lower than those measured in high specific surface sediments (e.g., clay). Early hydrate formation is accompanied by contraction for all soils and most stress states in part because growing gas hydrate crystals buckle skeletal force chains. Dilation can occur at high hydrate saturations. Hydrate dissociation under drained, zero lateral strain conditions is always associated with some contraction, regardless of soil type, effective stress level, or hydrate saturation. Changes in void ratio during formation-dissociation decrease at high effective stress levels. The volumetric strain during dissociation under zero lateral strain scales with hydrate saturation and sediment compressibility. The volumetric strain during dissociation under high shear is a function of the initial volume average void ratio and the stress-dependent critical state void ratio of the sediment. Other contributions to volume reduction upon hydrate dissociation are related to segregated hydrate in lenses and nodules. For natural gas hydrates, some conditions (e.g., gas production driven by depressurization) might contribute to additional volume reduction by increasing the effective stress.

/pdf/volume-change-associated-with-formation-and-dissociation-of-1e8pfe6yg3.pdf

Volume change associated with formation and dissociation of hydrate in sediment

We review notable advances published in the year 1999 in the area of phase change and phase growth in porous media. Phase equilibria thermodynamics, particularly in micropores, and growth kinetics, emphasizing the pore-network structure, are highlighted. Advances reported include the effects of confinement in phase transitions in micropores, and of the pore microstructure in the growth and dissolution of gas and liquid phases with applications ranging from capillary condensation to drying to gas evolution to condensation.

Phase change in porous media

The Stefan equation is one of the simplest approximate analytical solutions for the thaw-freeze problem. It provides a useful method for predicting the depth of thawing/freezing in soils when little site-specific information is available. The limited number of parameters in the Stefan equation makes possible its application in a multi-layered system. We demonstrate that a widely used algorithm (JL-algorithm), which has been frequently used in permafrost regions, was derived by an incorrect mathematical method. It will inevitably result in systematic errors in the simulation if this algorithm is used in a multi-layered soil. We present another simple thaw-freeze algorithm (XG-algorithm) for multi-layered soils. The new algorithm can be used to determine the freeze/thaw front in multi-layered soils no matter how thick each layer is and how many layers the soil profile contains. Simulation results of the JL-algorithm and the XG-algorithm are compared using hypothetical soil profiles, and the XG-algorithm is also used to simulate the thaw depth at three permafrost monitoring sites on the Qinghai-Tibet Plateau and one on the Loess Plateau, China. These applications show that the XG-algorithm could be readily used to analyse the factors that affect active-layer thickness. It can also be coupled with hydrological or land surface models to simulate the freeze-thaw cycles in permafrost regions and for related engineering applications. Copyright © 2013 John Wiley & Sons, Ltd.

Short Communication A Simple Thaw-Freeze Algorithm for a Multi-Layered Soil using the Stefan Equation

The Stefan equation is one of the simplest approximate analytical solutions for the thaw-freeze problem. It provides a useful method for predicting the depth of thawing/freezing in soils when little site-specific information is available. The limited number of parameters in the Stefan equation makes possible its application in a multi-layered system. We demonstrate that a widely used algorithm (JL-algorithm), which has been frequently used in permafrost regions, was derived by an incorrect mathematical method. It will inevitably result in systematic errors in the simulation if this algorithm is used in a multi-layered soil. We present another simple thaw-freeze algorithm (XG-algorithm) for multi-layered soils. The new algorithm can be used to determine the freeze/thaw front in multi-layered soils no matter how thick each layer is and how many layers the soil profile contains. Simulation results of the JL-algorithm and the XG-algorithm are compared using hypothetical soil profiles, and the XG-algorithm is also used to simulate the thaw depth at three permafrost monitoring sites on the Qinghai-Tibet Plateau and one on the Loess Plateau, China. These applications show that the XG-algorithm could be readily used to analyse the factors that affect active-layer thickness. It can also be coupled with hydrological or land surface models to simulate the freeze-thaw cycles in permafrost regions and for related engineering applications. Copyright (c) 2013 John Wiley & Sons, Ltd.

J.T. Holden

Papers

Heat and mass flow associated with a freezing front