Data analytics is widely used to study distribution system data and extract from it valuable knowledge. This report investigates the use of data analytics techniques in distribution utility applications.
The use of descriptive, diagnostic, and predictive analytics techniques in distribution system applications are demonstrated, and examples of their use on real distribution system data provided by several North American utilities are also presented. The target outcomes identified by the utilities were achieved using the proper analytical techniques on the data they provided. The conducted analyses included:
1) Application of descriptive statistical analysis methods on two datasets to classify the outage cause for each year, for each region, and on one utility dataset to classify the fault events per substation/feeder.
2) Application of diagnostic data analysis methods on utility data to identify voltage sensors failures, and on another dataset to identify capacitor switching failures.
3) Application of predictive data analysis methods on one published dataset to forecast distribution system loads, and on another published dataset to predict the power transformer health index.
The report concludes with a discussion of the future trends in distribution system data analytics and possible applications enabled by smart grid technologies.
Data Mining, Electric Utilities, Statistical Methods