Rapid deployment of distributed generation resources such as photovoltaic systems, wind farms, and other power electronics-based loads has emphasized the requirement for tap-changing transformers that minimize voltage variations and phase imbalances. Thus, various power-flow algorithms able to model such a tapchanging transformer have been proposed. The algorithms include a method that uses the bus impedance or admittance matrix. However, particularly the previous algorithms that use the impedance matrix did not take tap-changing transformers and P-V buses into account. Thus, this study models tap-changing transformers for a power-flow algorithm using the impedance matrix. For this purpose, it proposes a unified steady-state T shaped transformer model with a tap changer on the primary or secondary side. The proposed power-flow calculation algorithm also models P-V buses by the sensitivity impedance matrix. Since the proposed method reduces the inversion of matrices, it can be faster than the Newton-Raphson and fast decoupled methods (eg, the modified IEEE 30-bus test system added by 3000 or more nodes), not losing the accuracy. K E Y W O R D S bus impedance matrix, P-V bus, power-flow algorithm, sensitivity impedance matrix, tapchanging transformer 1 | INTRODUCTION Distributed generation (DG) resources such as photovoltaic systems, wind farms, microturbines, and other power electronics-based loads inject electric power to distribution feeders, often causing voltage variations and phase imbalances. Since voltage variation can be minimized by tap-changing transformers, power-flow algorithms that analyze DG resource-based distribution feeders should be able to model tap-changing transformers. Thus, the previous studies have