Consistency in thermal conductivity measured via lab-, field-scale test, and numerical simulation for newly developed backfill materials for underground power cable system

Abstract

This study investigates the thermal properties of newly developed backfill materials (normal and prepacked concrete) for an underground power cable system (UPCS) through different methods (laboratory test, in-situ horizontal thermal response test (TRT), and numerical analysis). First, the thermal conductivities of small-scale backfill materials were measured in a laboratory using a hot disk sensor. Afterward, a 5 × 3 × 2 m full-scale backfill was prepared for the TRT. The infinite line source (ILS) model was employed to estimate the thermal conductivity of the backfill materials using data from the TRT. Finally, a numerical analysis was conducted to simulate the heat transfer process during the TRT and back-calculate the thermal conductivity of the proposed backfill materials. All tests were also conducted using conventional backfill materials, namely natural sand, for comparison with the proposed backfill materials. The results indicated that the prepacked concrete type has the highest thermal conductivity among the three backfill materials regardless of the testing method employed. In addition to a high thermal conductivity, this newly developed material also has good workability (382 mm for the grout), low bleeding, and satisfactory excavation ability for the backfill material (1.56 MPa). Furthermore, the ILS model and numerical model slightly overestimated the thermal conductivity (less than 8 %) compared to the measured values from the laboratory test involving normal and prepacked concretes; however, natural sand exhibited a significant difference (1.365 W/(mK) and 1.8 W/(mK), 32 %) owing to the influence of the water content on its thermal conductivity. In addition, thermal conductivities calculated by the ILS model and numerical simulation were in very good agreement (with the difference being lower than 3 %), demonstrating that the proposed ILS model is appropriate to use to estimate the thermal conductivity of large-scale backfill material using horizontal TRT data.