Three-Phase Unbalanced Load Flow

Unlike balanced positive-sequence models, unbalanced load flow explicitly represents the neutral conductor and its associated grounding impedance. This is critical because return currents in the neutral can cause neutral-to-earth voltages that create safety hazards and equipment stress. The full 4x4 admittance matrix captures self and mutual impedances between all three phases and the neutral, accounting for Carson's equations for earth return effects.

Distribution lines are not transposed, meaning the phase conductors occupy fixed physical positions on the pole or tower. This creates unequal self-impedances and significant off-diagonal mutual impedance terms in the phase impedance matrix. An unbalanced solver must retain the full 3x3 primitive impedance matrix rather than applying sequence component decoupling, which is only valid for balanced, transposed transmission lines.

Distribution feeders routinely serve single-phase tap lines and V-phase (two-phase) laterals that are absent in transmission systems. These create inherent structural unbalance that cannot be averaged out. The load flow algorithm must handle missing phases at nodes, representing them as zero-injection constraints while maintaining numerical stability in the Jacobian matrix during the Newton-Raphson iteration.

Loads are modeled as constant power (PQ), constant current (I), or constant impedance (Z) on a per-phase basis, often in a ZIP combination. In unbalanced systems, a single-phase air conditioner on Phase A creates a fundamentally different loading condition than the three-phase industrial motor on Phases A-B-C. The solver must handle phase-specific load models and update the injected current vector accordingly at each iteration.

Rooftop solar inverters are predominantly single-phase 240V devices connected line-to-line or line-to-neutral. This injects power asymmetrically, potentially causing reverse power flow on one phase while the other two phases import power. The unbalanced solver must model the inverter's control mode—whether it operates in constant power factor, Volt-VAR, or Volt-Watt mode per IEEE 1547-2018—to accurately predict voltage rise on the energized phase.

Distribution transformers introduce phase shift and voltage unbalance based on their winding configuration. A delta-grounded wye transformer blocks zero-sequence currents from propagating upstream, while an open-wye open-delta bank creates inherent voltage asymmetry. The load flow must include detailed admittance matrix models for each transformer connection type, accounting for tap ratios, phase shifts, and leakage admittances on a per-phase basis.

A mathematical transformation that resolves an unbalanced three-phase system into three balanced sets of phasors: positive, negative, and zero sequence. Developed by Charles Fortescue in 1918, this method converts coupled phase-domain equations into decoupled sequence networks, dramatically simplifying fault analysis. In unbalanced load flow, sequence components help quantify the degree of asymmetry and are essential for modeling the response of synchronous machines and transformer connections to unbalanced conditions.

Two competing modeling paradigms for unbalanced systems. The phase frame (ABC domain) models each conductor explicitly, preserving the physical topology and mutual coupling but requiring the solution of a 3N×3N matrix. The sequence frame (012 domain) diagonalizes balanced sections of the network but introduces coupling at points of unbalance. Modern distribution solvers predominantly use the phase frame to handle arbitrary asymmetries without iterative sequence conversions.

A set of infinite series formulas published by John R. Carson in 1926 that calculate the self and mutual impedances of overhead conductors with an earth return path. These equations account for the frequency-dependent skin effect in the soil and are fundamental to building accurate primitive impedance matrices for distribution feeders. Carson's corrections are applied to every conductor pair in a line segment before Kron reduction eliminates the neutral and ground wires.

A network reduction technique that eliminates specified nodes from a system of linear equations while preserving the electrical characteristics at the retained nodes. In three-phase load flow, Kron reduction is applied to remove the neutral conductor and ground wire from the primitive impedance matrix, folding their effects into the equivalent phase impedances. This produces the compact 3×3 or 4×4 phase impedance matrices used in the final admittance formulation.

A widely adopted power flow formulation where the nonlinear bus power equations are converted to a linear current-mismatch form using a fixed-point iteration. At each iteration, complex bus currents are computed from the specified power injections and the latest voltage estimate, then the linear system YV = I is solved. This method avoids the computationally expensive Jacobian matrix updates of Newton-Raphson and is particularly robust for the high R/X ratios typical of distribution feeders.

The classical power flow algorithm adapted for three-phase systems by expanding the Jacobian matrix to include partial derivatives of phase power mismatches with respect to phase voltage magnitudes and angles. Each bus contributes a 6×6 Jacobian block capturing intra-phase coupling. While exhibiting quadratic convergence near the solution, this method requires a good initial guess and can struggle with the flat voltage profiles and high R/X ratios of distribution networks.