Weather outages affect power systems all over the world. In January 1998 Canada experienced the 'Great Ice Strom' which hit an area spanning from Eastern Ontario to Southern Quebec. 15 years later, it experienced another ice storm which left 600,000 people without power at the peak. Wind storms can also cause severe damage to power systems such as wind storms Lothar and Martin that hit France in December 1999. Due to never experiencing winds this high, nearly 3.4 million homes were left without power for up to 17 days. In the future outage rates may increase but we are unaware of what the current effects are, for example is there is any relationship between durations and the intensity of the weather or what the probabilities of these events causing an outage are. However, the weather experienced by each country, even areas within a country can vary. This thesis will present a five stage methodology that was developed for the analysis of both the current effects of weather on power transmission systems and also, the future effects. It also applied this methodology to the GB transmission network as a test case to identify the outcomes interms of network reliability to possible changes in climate. To complete this analysis outage datasets, provided by the transmission companies of GB, were acquired to help understand what the current effects of weather are. It was determined that the three main weather types that cause weather related outages on the GB transmission network were, Lightning, Snow, Sleet, Blizzards & Ice and Wind, Gales and Windborne Objects'. It also compared observational weather data to reanalysis data and lightning strike data and a lightning strike proxy to verify reanalysis data as a suitable replicator of past weather data. A correlation analysis between weather variables and weather related outages was completed to identify the weather indicator for the associated outage. From this fragility curves of failure rates for the dominant weather-related outages based on the weather that is occurring were developed. Using climate projections seven weather test cases were developed for the state sampling model and eight for the sequential simulation. In each case a different weather variable was changed to understand the effect they will have. It was concluded there are varying levels of confidence in climate change predictions making it difficult to assess the risk that poses to transmission networks. Reanalysis was found to be an acceptable replication of past observation data and CAPE was determined as a good proxy for lightning strikes. It was found there was an overall increase in failure rates and that there was an increase in all system indices, suggesting there would be increases in the magnitude and number of load shedding events.
|Date of Award||14 Sep 2017|
- University Of Strathclyde
|Sponsors||National Grid, Scottish and Southern Energy Plc SSE PLC, Central Networks, EA Technology Ltd & EDF Energy Networks ltd|
|Supervisor||Keith Bell (Supervisor) & David Infield (Supervisor)|