Backward/Forward Sweep

The algorithm fundamentally relies on the radiality constraint of distribution feeders. Unlike meshed transmission systems, a radial network has a single path from the substation to each load.

• Tree Structure: The network is modeled as a directed graph with no closed loops.
• Layered Processing: Nodes are organized into layers based on their distance from the source, enabling sequential computation.
• No Jacobian Required: The method avoids the computationally expensive matrix inversion needed by Newton-Raphson, making it ideal for real-time applications in Distribution Automation (DA).

Starting from the terminal nodes and moving toward the substation, the backward sweep calculates the current flowing through each branch.

• Load Current Calculation: At each node, the load current is computed using the specified power and the current voltage estimate: I_load = (S_load / V_node)*.
• Shunt Admittance: Charging currents from line capacitance are included at each node.
• Kirchhoff's Current Law: The branch current is the sum of the load current at the downstream node plus all currents in branches further downstream, moving layer by layer toward the source.

Once all branch currents are known, the forward sweep updates node voltages starting from the known substation voltage and moving outward to the feeder endpoints.

• Voltage Drop Calculation: The voltage at the downstream node k is calculated as V_k = V_i - I_branch * Z_branch, where V_i is the known upstream voltage.
• DistFlow Equations: This sweep directly implements the recursive DistFlow Equations, which are the standard simplified power flow model for radial systems.
• Convergence Check: After the forward sweep, the maximum voltage magnitude mismatch between iterations is checked against a tolerance (e.g., 10^-6 p.u.).

The basic algorithm is easily modified to incorporate Distributed Energy Resources (DERs) like rooftop solar, which act as negative loads or controlled current sources.

• PQ Node Model: A DER with specified active and reactive power output is treated as a negative constant power load in the backward sweep.
• PV Node Model: For voltage-controlled DERs, reactive power is adjusted iteratively to maintain the specified voltage magnitude, requiring a sensitivity-based compensation step within the sweep.
• Bidirectional Flow: The algorithm naturally handles reverse power flow from excess generation without reformulation, a key advantage for modern grids.

The method exhibits linear convergence characteristics and can be adapted for networks with a small number of loops, such as those created during Service Restoration (SR).

• Breakpoint Compensation: Weakly meshed networks are solved by "breaking" the loops at Normally Open Points (NOPs) and injecting compensating currents to simulate the closed loop.
• Robustness: It is highly robust for high R/X ratio lines typical of distribution, where Newton-Raphson often diverges.
• Computational Efficiency: The forward/backward sequence is computationally light, making it the standard engine inside Outage Management Systems (OMS) and real-time Digital Twin simulations.

Backward/forward sweep is the core power flow engine inside heuristic optimization algorithms like the Branch Exchange Method for loss minimization.

• Loss Calculation: After each sweep, total active power loss is computed as sum(I_branch^2 * R_branch).
• Switching Evaluation: When a tie switch is closed and a sectionalizing switch opened, a new radial topology is formed. The sweep rapidly evaluates the losses of this candidate configuration.
• Voltage Constraint Check: The forward sweep output immediately reveals if any node violates the ANSI C84.1 voltage limits, allowing the Network Reconfiguration Algorithm to discard infeasible topologies.

A simplified recursive formulation of power flow specifically derived for radial distribution networks. Unlike the full Newton-Raphson method, DistFlow equations exploit the tree structure to calculate voltage magnitudes and branch flows sequentially from the terminal nodes back to the substation. This mathematical framework provides the theoretical foundation for the Backward/Forward Sweep algorithm, expressing the relationship between active power, reactive power, and voltage drop across each line segment without requiring a large Jacobian matrix inversion.

A fundamental operational rule mandating that distribution networks maintain a tree topology with no closed loops. This constraint is critical because:

• It simplifies protection coordination by ensuring fault currents flow in a single direction
• It enables the use of efficient algorithms like Backward/Forward Sweep that cannot converge on meshed networks
• It guarantees a unique path from the substation to every load node During network reconfiguration, maintaining radiality while transferring load between feeders is the primary combinatorial challenge.

The dominant iterative technique for solving transmission system power flow, which contrasts sharply with Backward/Forward Sweep. Newton-Raphson constructs and inverts a large Jacobian matrix of partial derivatives to simultaneously solve for voltage angles and magnitudes. While robust for meshed transmission grids, it often suffers from convergence issues in distribution systems due to:

• High R/X ratios in distribution lines
• Radial or weakly-meshed topologies
• Unbalanced loading conditions The Backward/Forward Sweep was developed specifically to overcome these limitations.

The process of altering the open/closed status of sectionalizing switches and tie switches to transfer load between feeders. Backward/Forward Sweep serves as the inner-loop power flow engine for most DFR algorithms because:

• Each candidate topology must be evaluated for voltage compliance and loss calculation
• The algorithm's speed allows thousands of configurations to be tested rapidly
• It accurately models the radial structure after each switching operation Heuristic methods like Branch Exchange rely on repeated BFS executions to find lower-loss topologies.

A grid efficiency technique that intentionally lowers service voltage to the lower bound of the ANSI C84.1 standard range to reduce energy consumption. Backward/Forward Sweep is essential for CVR planning because:

• It accurately predicts the voltage profile at every node under reduced substation voltage
• It verifies that no customer falls below the minimum allowable voltage
• It models the ZIP load behavior where constant impedance, current, and power loads respond differently to voltage changes Utilities use BFS simulations to quantify expected demand reduction before implementing CVR control strategies.

Distribution systems inherently operate with unbalanced loading due to single-phase laterals and uneven customer connections. The Backward/Forward Sweep naturally extends to three-phase modeling by treating each phase conductor and the neutral as separate coupled elements. The algorithm computes:

• Phase voltages independently at each node
• Mutual coupling effects between phases in the backward current sweep
• Neutral-to-ground voltage rise caused by return currents This capability makes BFS the preferred method for analyzing voltage imbalance and neutral shift in real distribution feeders.