Existing underground cable ratings are very conservative and therefore underestimate the full system capacity especially during high times of high power consumption during summer months. There is a need for improved and more accurate rating calculations. An objective of this project has been to investigate the heat and moisture migration initiated by the heating of the medium surrounding buried cables, which when subjected to prolonged heating under extreme conditions can lead to 'thermal runaway'. Identification of conditions under which 'thermal runaway' occurs will allow the improvement in estimation of the external parameters (such as soil thermal resistivity and capacitance) used in the existing cable ratings. A controlled experiment pertaining to heat and moisture transfer has been conducted in order to verify the model developed by Philips and de Vries using the finite element analysis method. Experiments involved the use of soil of various low moisture contents and the results are compared to the model which used variable soil properties. Thermal conductivity and soil suction are two essential soil properties that have been determined experimentally. Whereas, other hydrological soil properties are obtained theoretically using established methods. Soil properties that are used in the model are made variable to temperature and volumetric moisture content. Investigation has revealed that a strong correlation exists between moisture and temperature due to heat transfer. The verified Philip-de Vries model has been used to simulate the effects of drying-out. This involves looking at the changes of moisture content of the surrounding medium, the dry/wet interface and the effects of different heat losses and initial volumetric moisture contents have on the heated backfill.