Utility customer segmentation based on smart meter data: Empirical study

We develop statistical techniques for analyzing the energy information in the 15-min and daily household electricity consumption data. The results provide a good understanding of how usage is affected by environmental, structural and customer features. The analytics yield productive results for a small region, and perform well in other areas and in different seasons. These are versatile tools for identifying specific lifestyle and defining customer segments that can yield measurable results and high returns for energy programs. The data analytics also explore the changes in statistical properties of individual versus aggregated data when customers are clustered.