Transient Temperatures of Underground XLPE Power Cables - Comparison of Results from Different Calculation Methods and Full-Scale Experiment
MetadataVis full innførsel
- Institutt for elkraftteknikk 
By gaining an better understanding of the limiting thermal mechanisms of the various power system components, work can be performed to increase the grid transfer capability. Of these components the underground power cable plays an crucial role, and an increased knowledge of transient temperatures is therefore important. The main problem definition of this master thesis was:-How does results from calculation of transient temperatures of underground power cables by means of numerical modeling, as well as by method specified in IEC 60853, compare to measurements from a full-scale experiment? In addition, the following problem definition was of interest to this study:-How will a change in the thermal characteristics of cable sand, used as bedding and initial backfill, influence the transient and stationary temperature conditions of underground power cables? The main problem definition was solved by developing numerical models in COMSOL, as well as MATLAB scripts for the IEC 60853 method. In which results from the two calculation methods were compared to data obtained from an full-scale test facility. Moreover, the problem definition related to the cable sand characteristics and the influence relative to transient and stationary conditions, were solved by means of employing an numerical approach. Where an realistic change in the thermal characteristics of the cable sand was obtained by compressing a sand sample, and experimentally determine the change in the characteristics. When it comes to the main problem definition, the report can based on the results conclude that both calculation methods were in good accordance with measured values, excluding no- load periods beyond an duration of ~70 and ~10 hours for which the numerical and IEC 6083 method respectively gave discrepancies. When comparing the two methods, the numerical approach showed marginally better correlation with measurements than the IEC 60853 method. Moreover, the report can also conclude that the IEC 60853 method have two main shortcomings compared to an numerical approach. Where the first shortcoming is the lacking ability to account for non-homogeneous soils, and the second shortcoming being the representation of ambient temperature in which is implemented instantaneously in the solution, thus large variations in ambient temperature cannot be accounted for in an suitable way. With regard to the second problem definition, the report can based on the results conclude that the thermal conductivity of cable sand have an significant influence relative to stationary conditions. However, the added thermal resistance associated with cables in pipe will reduce the benefits compared to cables directly in ground. When it comes to the dynamic conditions, an neglectable reduction in temperature was obtained by an increased heat capacity, due to the small mass of the cable trench relative to the native soil. The dynamic conditions with regard to cable sand characteristics is thus mainly governed by the thermal conductivity.