Abstract
Load and temperature data collected from an onshore wind farm export cable during a nine-month period was used to test and evaluate models for transient calculations of cable temperatures. The temperature data was measured using a distributed temperature sensing (DTS) system. A single location along the cable, where the cable is directly buried in a trefoil formation, was chosen for analyses. The objective was to test models based on thermal equivalent circuits (TECs), which are simple and computationally effective models and promising candidates for affordable and efficient system-wide ampacity rating of cables. The TEC models were compared to finite-element analysis (FEA) models and tested on real data. The work was divided into three case studies with increasing complexity. Firstly, the steady state temperature of the cable was calculated for two different loads using both TEC and FEA. The results corresponded well, with a maximum deviation of 1.51 °C between the two model types. In the second case, the transient temperature response to a step increase of the load was calculated for two different loads. In this case, there was a larger deviation between TEC and FEA, especially for short times. Finally, the models were tested on real data from the cable installation. In this case, the FEA outperformed the TEC model with a mean error standard deviation) of 1 17 1 00 °C, while TEC had 2 44 1 27 °C. The TEC showed systematic errors that followed the load and ambient temperatures, which indicates errors or room for improvements in the model itself. As for the FEA model there were no apparent systematic errors. It is hypothesized that the errors in this model arose due to the assumption that the thermophysical properties of the soil were constant in time. This is, in most cases, not a valid assumption because the weather, especially precipitation, has a large influence on the thermal properties of the soil. Further work should be done in terms of improving the TEC model, especially regarding modelling the soil.
