Comparative Implementation of High Performance Computing for Power System Dynamic Simulations

Dynamic simulation for transient stability assessment is one of the most important, but intensive, computational tasks for power system planning and operation. Several commercial software tools provide functionality for performing multiple dynamic simulations such as those in contingency analysis simultaneously on parallel computers. Nevertheless, a single dynamic simulation is still a time consuming process performed sequentially on one single computing core as the tools were originally designed. Modern high performance computing (HPC) holds the promise to accelerate a single dynamic simulation by parallelizing its kernel algorithms without compromising computational accuracy. Parallelizing a single dynamic simulation is a much more challenging problem than the contingency-type parallel computing. It requires a good match between simulation algorithms and computing hardware. This paper provides guidance for such a match so as to design and implement parallel dynamic simulation to maximize the utilization of computing hardware and the performance of the simulation. The guidance is derived through comparative implementation of four parallel dynamic simulation schemes in two state-of-the-art HPC environments: 1) message passing interface and 2) open multi-processing. The scalability and speedup performance of parallelized dynamic simulation are thoroughly studied to determine the impact of simulation algorithms and computing hardware configurations. Several testing cases are presented to illustrate the derived guidance.

I. Introduction

Dynamic simulation for assessing transient stability is one of the most important computational tasks that affect the secure operation of the bulk electric power system. It typically solves a large set of differential and algebraic equations (DAEs) many times in a time-stepping manner. These DAEs describe the time-domain trajectories of the electro-mechanical interaction between system loads, generators, transmission lines, and power electronic devices and their controllers, when the system is subject to disturbances. Due to the increasing complexity of today’s power grid, the DAEs used for dynamic simulations are of a fairly large size. For example, there are more than 15,000 buses and 2,000 generators in a model representing the entire Western Electricity Coordination Council (WECC) in North America. The model for the Eastern Interconnection can be even several times bigger. Sequential simulation of such large systems could be very time consuming and prohibited for real-time applications. Several commercial software tools [1]–[4] provide functionality for performing multiple dynamic simulations such as those in contingency analysis simultaneously on parallel computers. But a single dynamic simulation is still a time consuming process performed sequentially on one single computing core as the tools were originally designed. The method of distributing a large number of study scenarios to multiple processors or workstations works well for offline studies and some online studies where the total computation time is less critical. However, in more time-critical online studies and predictive applications for preventive/corrective controls, further acceleration of a single simulation is necessary for stabilizing the power grid.