Conditional demand analysis: A technique to estimate end-use load shapes

As more utilities incorporate end-use peak demand load forecasting techniques, the need for dependable appliance load shapes is becoming increasingly important. Given the expense of direct end-use metering and the lack of appropriate borrowed data, conditional demand analysis (CDA) is a viable alternative for providing these load shapes. This paper describes the CDA methodology used by East Kentucky Power Cooperative (EKPC) to extract appliance-specific load shapes from whole house metered data. The purpose of this study was to develop reasonable appliance load shapes to incorporate into EKPC`s peak demand forecasting model. A straightforward method utilizing ordinary least squares regression analysis on a small sample of metered residential households and their accompanying surveys was employed.