Characterization of Consumer Flexibility and Its Impact on Voltage Regulation in Distribution Systems

Modern distribution networks face growing challenges from demand growth, electrification, and distributed energy resources, complicating voltage regulation and service quality. While demand response has been proposed as a complementary tool, most existing studies assume predefined flexible loads or rely on synthetic demand models rather than deriving flexibility directly from real consumption data. This study addresses that gap by proposing a data-driven methodology that characterizes consumer flexibility from 48 months of real advanced metering infrastructure (AMI) data using statistical variability metrics and non-negative matrix factorization (NMF)-based clustering, and integrates the resulting flexibility trajectories into time-series power-flow simulations of three real distribution feeders with contrasting loading conditions. Targeting only the most flexible consumer segment, representing 38.8% of the population, yields peak demand reductions of up to 5.7%, loss reductions of up to 1.9%, and measurable voltage improvements without infrastructure changes; the extent of the benefit depends on the feeder’s initial loading condition, with flexibility alone insufficient to resolve severe undervoltage in the most heavily loaded feeder. A series of complementary sensitivity analyses, covering spatial load allocation, optimization-based redistribution, reactive-power support, and integration with distributed generation and electric vehicles, shows that these findings hold consistently across the alternative modeling assumptions examined. These results demonstrate that trajectory-based demand management, grounded in real metering data and validated across diverse network conditions, constitutes an effective complementary strategy for voltage regulation alongside conventional regulation devices.