Intelligent Methods for Evaluating the Impact of Weather on Power Transmission Infrastructure
Pawel Maksymilian Pytlak
Weather has a significant impact on human society, both in driving the lifegiving physical processes that allow humans to meet their most basic needs, and as an adversarial force, often causing significant losses to property and life. The electrical power industry is particularly subject to significant weather influence due to the wide-scale exposure of its infrastructure to nature's elements. Severe storms can cause damage in the millions of dollars, and even directly or indirectly cause
... ties. Weather patterns and their impact on the industry are hard to predict using simple statistical measures, and thus more complex methodologies must be used to provide accurate forecasts and impact assessment. The increasing global awareness of climate change is driving the power industry to adopt more green energy sources. Unfortunately, these sources cannot be constructed at will; they must be harnessed where they are available. Consequently, the power industry is not always ready to incorporate these sources into the existing grid without costly infrastructure upgrades and/or expansion projects. To help alleviate these concerns, this thesis presents intelligent methodologies that can use either modern Numerical Weather Prediction (NWP) models or direct weather observations to solve some of the challenges faced by the power industry. It describes the optimization and verification of an ice accretion forecast system that is tuned to increase its predictive accuracy us-ing computational intelligence techniques. The performance of the system is also evaluated in a true forecast simulation. This thesis also describes the enhancement of an industry standard line rating model that is expanded to include the cooling impact of precipitation. Studies are presented that discover the optimal configuration of weather-based dynamic thermal rating systems, evaluate the accuracy and risk of forecasting line ampacity ratings using NWP models, and assess the reduction in emissions by using dynamic ratings to incorporate more green energy into the transmission grid. Finally, this thesis describes intelligent systems aimed at assisting and supporting planning decisions in transmission infrastructure construction and expansion projects.