Case Studies

The Challenge

The UK electricity network transfers electrical energy from generating power plants to individual consumers. The electricity is first transferred on a high voltage transmission network from power plants to substations near to population centres. The distribution network then transforms the electricity to a lower voltage and transfers it between the substations and consumers.  Both networks are made up of overhead lines, underground cables, substations and transformers. Both are susceptible to faults, which can be attributed to a variety of causes such as deterioration of equipment, fire, third party interference or weather.

Faults caused by weather can have a large impact on customers; in 2008/2009 weather-related faults caused approximately 1.9 million UK customer interruptions (an interruption to customers lasting more than 3 minutes). Not only do the network operators have to manage the short-term operational implications and costs of weather-related faults, but due to the long lifetime of the network, they must also consider the potential longer-term implications of climate change. Understanding how climate change could affect the frequency of weather-related faults on the network is vital for building adaptation plans into long-term network investments such as creating new design standards or replacing existing equipment. Whilst projections of future climate are available (for example how temperature, rainfall or wind is projected to change in the future), this information is difficult to incorporate into any long-term plans without a thorough understanding of how these changes may affect the frequency of weather-related faults.

The Approach

Historical time series of weather-related faults and weather can be analysed to investigate how the electricity networks are affected by weather.  Examining the instances of weather-related faults in the past reveals that the three main causes of weather-related faults are wind and gale, snow, sleet and blizzard (SSB) and lightning. A more detailed exploratory analysis of the data highlights important key findings about the type of weather that causes these faults.  For example:

• Wind and gale faults are caused by high wind gusts and can be exacerbated by trees and debris being blown on to the lines which, in turn, is affected by the length of growing season and the schedule of maintenance work carried out on the trees.
• SSB faults are not only caused by snow, but a combination of snow and high-wind gusts that causes ice to build up on the lines.
• Lightning faults are usually caused by cloud-to-ground lightning.  Since future projections of lightning are not available, a measure of atmospheric convection is used as a proxy.

Using this information, it is possible to formalise a statistical relationship between each type of weather-related fault and its primary weather predictors (for example, the predictors for SSB faults are whether or not snow has fallen and the daily maximum wind gust). Projections of weather-related faults can then be estimated by applying the statistical relationships to future projections of the weather predictors from the Met Office Regional Climate Model.

The Value

Future projections of weather-related faults give the electricity network operators an indication of how the frequency of these faults may change in the future.  Statistical methods can be used to show that whilst future projections of wind and gale faults indicate that they may remain the same, increase or decrease in the future (due to the large uncertainties associated with wind gust projections), lightning faults are projected to increase in the future. SSB faults are projected to decrease in the future due to a reduction of snow days, but when snow does fall the intensity of the event may remain the same or increase.

By using statistical techniques, it is possible to provide the industry with projections of future weather-related faults that can be used to help prioritise which adaptation measures to implement and when. For example, given the projected increase in lightning faults, one possible adaptation option would be to increase the number of lightning arrestors on the electricity network. Arrestors protect the insulations on the network by diverting the current from a lightning strike or surge around the insulation to the ground.