A mixed-methods load segmentation approach for informing flexibility services in energy systems

Abstract

This study introduces a robust methodology for residential household archetyping to support Demand Side Management (DSM) in evolving energy systems. Using granular 30-minute smart-meter data over a period of two years, we generate daily load profiles for more than one hundred UK households (with and without ownership of low-carbon technologies (LCTs)) and through K-means clustering, characterise seven representative daily demand patterns. The resulting archetypes highlight the growing fluidity of household energy use, driven by the increasing adoption of LCTs such as battery storage, electric vehicle charging, and residential solar generation, as well as household responses to time-of-use electricity tariffs. These trends contribute to greater variability in daily demand, making it essential for DSM strategies to capture not only typical behaviours but also the intrinsic flexibility, inherent in modern residential energy consumption. Beyond intrinsic flexibility provided by low-carbon technologies, the proposed methodology integrates behavioural attributes to reflect extrinsic flexibility, offering a more holistic representation of household responsiveness. Collectively, this approach provides an accurate and adaptable framework for modelling residential energy demand and supports the design of effective DSM interventions in a rapidly decarbonising energy landscape.