An Exact Convex Formulation of Optimal Power Flow in Radial Distribution Networks Including Transverse Components

TL;DR

This paper presents an exact convex formulation for the optimal power flow problem in radial distribution networks that includes transverse components and line ampacity constraints, with conditions ensuring solution feasibility and optimality.

Contribution

It introduces a convex formulation that accounts for transverse parameters and ampacity constraints, providing verifiable conditions for solution exactness in radial power grids.

Findings

Conditions for exactness are mild and applicable to real networks.

The proposed formulation ensures feasible and optimal solutions.

The method correctly accounts for line ampacity constraints.

Abstract

The recent literature has discussed the use of the relaxed Second Order Cone Programming (SOCP) to formulate Optimal Power Flow problems (OPF) for radial power grids. However, if the shunt parameters of the lines, composing the power grid, are considered the proposed methods do not provide sufficient conditions that can be verified ex ante for the exactness of the optimal solutions. Additionally, the same formulations have not correctly accounted for the ampacity constraint of the lines. Similar to the inclusion of upper voltage-magnitude limit, the SOCP relaxation faces difficulties when the ampacity constraints of the lines are binding. In order to overcome these limitations, we propose a convex formulation of the OPF problem applied to radial power grids for which the AC-OPF equations, including the transverse parameters, are considered. We augment the formulation with a new set of more conservative constraints to limit the lines current together with the nodal voltage-magnitudes. Sufficient conditions are provided to ensure the feasibility and optimality of the proposed OPF solution. Furthermore, the proofs of the exactness of the SOCP relaxation are provided. Using standard power grids, we show that these conditions are mild and hold for real distribution networks.