An experimental study of soil thermal conductivity using a guarded hot plate apparatus

A guarded hot plate apparatus was used to generate comprehensive sets of thermal conductivity for two types of soils, namely Ottawa sand and Richmon Hill clay-loam, for temperature variation from 2 to 92°C and moisture content variation from complete dryness to full saturation with measurement errors of less than 3%. Numerical simulation of heat transfer within the apparatus with sample inside was performed to validate the experimental design and setup. To prepare the samples, a consistent specimen preparation technique was developed for the cases of dry, barely-to-moderately moist, and highly-to-fully saturated moist soils. On the basis of gathered datasets, empirical correlations for soil thermal conductivity were developed as a function of both temperature and moisture content. The proposed correlations produced excellent fit to majority of the experimental data, and could be easily integrated into numerical analysis of underground heat transfer. As an application example, one of the correlations was employed to evaluate soil thermal conductivity in a numerical study of underground heat loss from a basement wall and floor, in order to illustrate the importance of considering the dependence of soil thermal conductivity on soil texture, temperature and degree of saturation.