Assessment of thermal conductivity prediction models for compacted bentonite-based backfill material

This study presents the experimental investigation on thermal conduction of bentonite-based backfill material with various sand combinations (100/0, 80/20, 60/40, 50/50, 30/70, by weight percentage bentonite/sand). The bentonite-based composites were tested employing KD2 Pro dual thermal sensor, and the effect of sand, dry density and saturation were investigated. Despite of the different type of bentonite, the outcome demonstrates that the thermal conductivity (TC) of bentonite-sand material is directly related to its dry density, water content, sand content and saturation. Further, the present experimental and literatures thermal conductive data were assessed with different TC prediction models. The predictions showed a root mean square error of [0.110-0.391] W/mK and mean error of [- 0.06 to 0.37] W/mK for predicting TC of bentonite. The predictions showed a root mean square error of [0.184-0.350] W/mK and mean error of [- 0.07 to 0.30] W/mK for bentonite-sand combinations. Overall, model of the Nikoosokhan et al. performed more satisfactory results followed by model of the Cote and Konrad, Johansen and Lu et al. Moreover, this work offers a possible perspective for the choice of heat conduction predictive model for backfill used in underground power cable systems and other geothermal structures.