State estimation and flexibility activation in distribution feeders within a digital twin framework

• Joint state estimation-optimization resolves sparse observability in active distribution networks. • Warm-started genetic algorithm uses an augmented state vector embedding distributed energy resource controls. • Framework successfully validated via 21 k recursive cycles on a live medium-voltage feeder. • State estimation achieves a robust voltage error range of 0.971 V to 4.636 V on monitored buses. • Flexibility analysis shows 14.5 % peak energy shed with a stability-performance trade-off. The challenge of reliable state estimation (SE) in sparsely measured Active Distribution Networks (ADNs) with high distributed energy resource (DER) penetration is addressed through a recursive digital twin (DT) execution kernel. The core numerical contribution is the re-formulation of SE as a joint SE-optimization problem, solved by a warm-started genetic algorithm (GA), which augments the state vector to include control setpoints, ensuring the estimated state is physically feasible. Data ingestion fidelity was rigorously confirmed by comparing power quality analyzer (PQA) streams and historical data, yielding a minimal mean difference of 0.470 kW and a strong Pearson correlation of 0.641. The framework was validated on a live medium voltage (MV) feeder, executing approximately 21,000 recursive cycles over three months (average 6.9 min per cycle). SE accuracy, assessed across four monitored buses, demonstrated an average absolute voltage error ranging from 0.971 V to 4.636 V. Furthermore, flexibility analysis across twelve scenarios, focusing on the controllable 625 kW pump, established a clear performance-stability trade-off, with the optimal aggressive configuration (Scenario 7) achieving a peak flexibility share of 14.5 % (632 kWh shed energy). This active control resulted in a measurable reduction in feeder imports and improved local energy balancing through targeted setpoint modifications in 24 % of all timesteps, effectively mitigating reverse power flow (RPF) while respecting asset constraints.