At present, the export cables connecting wind farms back to shore are sized based on a continuous current rating, as is common practice with cables on land. However, the generation profile of the wind farm will certainly not match this steady state assumption due to its dependence on weather conditions. While there will be times when the wind farm is generating rated output, the total duration of such events is likely to be a small percentage of the asset life. Given the thermal capacitance within the cable system and its environment, the conductor temperature may rarely come close to its rated maximum.
The inevitable consequence of taking the steady state rating approach is that the size of the cable is determined by an overly worst case set of assumptions. This has the potential to lead to the cable system being over-sized, forcing up the cost of the export cable system considerably. In some, near shore, cases where ac transmission is still possible, this can lead to more cables being laid than is strictly necessary to deliver energy to shore with the required reliability. This acts to force up the cost of connection and hence the cost of the energy when sold in the power market, reducing the economic viability of some projects while increasing the cost of energy for consumers.
To tackle this problem, this project will develop statistical rating techniques suitable for application to ac, and eventually HVDC, wind farm export cable systems. All thermal sections of the cable will be considered, including examples of: ââ¬Â¢ Burial at sea ââ¬Â¢ Land falls with and without horizontal directional drills ââ¬Â¢ Land cable sections ââ¬Â¢ J-tube installations
In each case, appropriate thermal models will be designed either based on existing algorithms or using techniques such as finite element analysis. Through a synergy with an existing HubNet funded PhD programme, improved models of the subsea environment will be incorporated based on typical UK coastal shelf sediments.
Having created suitable dynamic thermal models for each cable section, the models will be solved to obtain a temperature profile for multiple model-years, based on an input of actual wind farm output levels. This data will be supplied from existing offshore wind sites, plus shorter term predictions of wind farm output. Once typical levels of cable system capacity utilisation are known, the project will examine statistical methods for rating the system. This will allow the cable system to be sized based on the likelihood of needing transmit certain levels of power for certain time intervals, rather than simply taking the steady state approximation. A brief economic review will then be undertaken to determine whether or not the cable size, or number of cables, could be reduced and the likely cost benefit to the wind farm operator as a result.