Characterising Namibian residential load profile archetypes using Self-Organising Map (SOM) - based temporal segmentation and seasonal clustering

This study investigates the fluidity of residential electricity demand profiles within Namibian households to assess their flexibility potential. Traditional profiling methods typically generate a single average load profile per household over a given period, overlooking the temporal and spatial variability inherent in energy consumption patterns. To address this gap, we employ a Self-Organising Map (SOM)-based approach to characterise typical Namibian residential load profiles (shape and magnitude), capturing intra-household behavioural diversity as well as spatial heterogeneity across households. The analysis reveals distinct seasonal patterns and notable differences between weekday and weekend consumption behaviour. These insights offer valuable implications for demand-side management strategies, assessing demand flexibility potential, and energy system optimisation, particularly in the context of emerging variable and decentralised energy resources.