High-Performance Computing-Based Open-Source Power Transmission and Distribution Grid Co-Simulation

As the electric power grid becomes more and more complex due to the penetration of various distributed energy resources (DERs), co-simulation of power transmission and distribution (T&D) grid in one unified environment emerges as an effective approach to accurately account for the changes and impact of DERs on the power system. Fast and scalable T&D co-simulation leveraging high-performance computing (HPC) techniques is key to the successful deployment of large-scale grid monitoring and assessment to support online operation. In this article, two open-source power grid simulators — GridPACK (for transmission grid simulation) and GridLAB-D (for distribution grid simulation) are coupled together on Hierarchical Engine for Large-scale Infrastructure Co-Simulation (HELICS) — an open-source scalable co-simulation engine, to enable fast parallel dynamic simulation in transmission grid with enriched system dynamics captured from its interconnected distribution grids that are simulated in a distributed way. Experiments on a synthetic co-simulation study comprising one 3000-bus transmission grid model with detailed generator models and various interconnected IEEE-123-node distribution systems. The results show the faster-than-real-time performance of parallel dynamic simulation for the transmission system and high scalability with multiple distribution grids when running the co-simulation on both commodity-level multi-core computers and high-end supercomputing clusters.