Development of a Bottom-up Model for Residential End-use Energy Demand

The increasing proliferation of renewables and distributed energy resources makes the traditional top-down forecasting approach not powerful enough to represent the demand side diversity. Therefore, the aim of this project was to develop a bottom-up model that captures the electricity consumption patterns of individual end uses at the appliance level. To achieve this, a time-inhomogeneous Markov chain was built in R to model the end users’ activities, and the activity pattern was then converted to consumption patterns by appliance types. The goals were to offer better insights for demand-side management and to identify opportunities in the space of digital businesses.