Supervisors - Prof. Keith Bell, University of Strathclyde and Prof. Adrian Bowman, University of Glasgow
PhD Student - Fulin Fan, University of Strathclyde
Status - Closed
This project will make use of an advanced spatial/temporal model developed within the EEE Department at Strathclyde over the last 5 years. This model makes use of Vector Auto-Regressive representation of hourly meteorological data from 14 sites across the UK to capture the time structure and the spatial correlations between sites. Importantly, the model makes use of temporal de-trending to properly capture seasonal and diurnal (day-night) variations in wind speed, and been comprehensively validated (Ref 1). The model can provide accurate forecasts of the wind speed up to 5 hours ahead and these forecasts are generally more precise than weather forecasts from the met office. The model also provides a good estimate of the uncertainty associated with each forecast, which contrasts with met forecasting.
The proposal is to apply the model to forecast wind speeds along the length of transmission lines so as to provide improved indicators to the TSO of the current carrying capacity of the overhead lines which are significantly cooled by the wind in a manner that is already well understood and included in relevant IEEE Standards. Spatial interpolation techniques will be applied to provide the forecasts at regular intervals along the lines. The model will also be applied to air temperature forecasting as it is the combination of wind velocity and air temperature that determines the local line cooling rate and it this that determines the maximum current that can be carried at any time.