Performance of Distributed OPF Algorithm with ZIP Loads and Different PV Inverter Modes

High penetration of inverter-associated renewable energy resources, especially solar photovoltaic (PV) panels, increases the number of controllable grid-edge devices in an electric grid. Optimal power flow (OPF) problems are often formulated to ensure such electric grids' economic, secure, and reliable operation. To overcome the scalability and computation challenges related to centralized OPF methods, distributed algorithms are adopted for medium and bulk distribution grids. However, the convergence of distributed algorithms is highly impacted by the network architecture and operating characteristics of the grid edge resources. Studies are abundant on the variation in network architecture, but no study has investigated the impact of load models and PV inverter operating modes on the performance of distributed OPF algorithms. To address this, in this article loads are modeled as both constant power and voltage-dependent ZIP loads, and PV inverters are modeled considering their four distinct operating modes viz. grid following (GFM), volt-var controlling (VVCM), volt-watt controlling (VWCM), and voltage control (VCM). The load and the PV inverter models are integrated with the consensus Alternating direction method of multiplier (ADMM)-based distributed OPF algorithm as constraints for minimizing total network loss for an unbalanced distribution network. The performance characterization is performed on the IEEE 123 bus test system concerning network loss values, solution time, and total number of required macro iterations.