| Abstract: |
With the increasing demand, aging infrastructure and integration of highly intermittent renewable energy sources, modern electrical power distribution systems are confronted with a greatly increasing challenge to comply with ever more stringent reliability expectations. This empirical research article proposes a thorough, database case study in which we scrutinize the structural and operational refinements of an urban electricity distribution network with a view to optimize its basic reliability indicators. Focusing on implementing new network modernization strategies automated reclosers, optimal sectionalizer placement, and targeted vegetation management we analyze a rich dataset of thousands of historical interruption events collected over a multi-year baseline period via machine learning. Using statistical analysis, the core reliability metrics (SAIFI, SAIDI, CAIDI and ASAI) are computed from both pre- and post-amendment of the system operations. The working definitions show a decrease in the number and significantly less time for sustained sustained outages at the primary test feeders. Notably, SAIFI ameliorated by 25.4%, and SAIDI reduced by 31.8% which directly correlates with enhanced efficiency in operations and confirmed cost savings for the utility provider through mitigation of unserved energy costs. In addition, this research conducts an important comparative review with historic literature and traditional predictive models to show that localized empirical data analysis with strategic automation application provides one of the best efficiencies for capital versus generic full system retrofits. The verified results provide a solid method for distribution engineers to prioritize capital expenditures, improve grid resilience and respond financially to high regulatory performance requirements found in modern smart grid environments. |
