Feeder Reconfiguration

The classic optimization objective for feeder reconfiguration is the reduction of active power losses (I²R losses). By shifting load from a heavily loaded, high-resistance path to a parallel feeder with lower impedance, the algorithm minimizes the total copper losses across the system.

• Mechanism: Solves a combinatorial, non-linear optimization problem to find the radial topology with the lowest aggregate resistive loss.
• Constraint: Must maintain radiality (no closed loops) and keep all node voltages within ANSI C84.1 limits.
• Impact: Typical loss reductions of 10-30% are achievable in heavily interconnected urban networks.

Reconfiguration prevents thermal overloading of individual feeders and substation transformers by equalizing the load factor across adjacent circuits. This is a critical operational strategy to defer capital-intensive infrastructure upgrades.

• Objective: Minimize the loadability index or the maximum feeder loading deviation from the mean.
• Trigger: Activated when real-time SCADA measurements indicate a feeder is approaching its continuous thermal rating.
• Result: Balances the stress on aging assets, reducing hot-spot temperatures and extending transformer life.

Following a permanent fault, reconfiguration algorithms execute a self-healing sequence to isolate the faulted section and restore power to healthy downstream customers via an alternative path.

• Process: Fault Detection, Isolation, and Recovery (FDIR) logic identifies the fault location, opens boundary switches, and closes normally-open tie switches.
• Speed: Modern IEC 61850 GOOSE messaging enables restoration in under 60 seconds, dramatically improving SAIDI metrics.
• Constraint: Must verify that the backup feeder has sufficient residual capacity to accept the transferred load without causing a sympathetic trip.

By shortening the electrical distance between the substation source and end-of-line customers, reconfiguration can flatten the voltage profile along a feeder and mitigate under-voltage violations.

• Method: Transfers the tail-end load of a long, heavily loaded feeder to a geographically closer substation via a tie switch.
• Coordination: Works in tandem with Volt-VAR Control (VVC); reconfiguration handles the gross topology adjustment while capacitor banks and regulators handle fine-grained voltage smoothing.
• Benefit: Reduces the need for conservation voltage reduction (CVR) curtailment.

A fundamental operational rule in distribution systems is that the network must operate in a radial (tree) structure, even though it is physically built as a meshed network. Reconfiguration algorithms must strictly enforce this topological constraint.

• Why Radial?: Simplifies protection coordination, fault current calculation, and voltage regulation.
• Algorithmic Check: Uses Graph Neural Networks (GNNs) or spanning tree algorithms to verify that the resulting switch configuration contains no closed loops.
• Complexity: The combinatorial explosion of switch states makes this an NP-hard optimization problem, often solved via metaheuristics like genetic algorithms.

Practical reconfiguration engines must balance competing objectives simultaneously, such as minimizing switching operations while maximizing loss reduction, to avoid excessive wear on switchgear.

• Pareto Front: Algorithms generate a set of non-dominated solutions, allowing the operator to choose the optimal trade-off between switching cost and efficiency gain.
• Weighted Sum Method: Combines objectives like loss minimization, load balance, and voltage deviation into a single scalar fitness function.
• Dynamic Reconfiguration: Uses Model Predictive Control (MPC) to schedule a sequence of switching actions over a time horizon to accommodate shifting renewable generation.