Effective Parallelism for Equation and Jacobian Evaluation in Power Flow Calculation

TL;DR

This paper explores advanced parallelism techniques for power flow calculations, demonstrating significant speedups in equation and Jacobian evaluations through multi-threading and SIMD vectorization on large-scale power systems.

Contribution

It introduces two levels of parallelism—inter-model and intra-model—for power flow calculations, enhancing computational efficiency and scalability.

Findings

Equation evaluation speed increased by ten times.

Overall Newton power flow improved by 20%.

Effective parallel workflows demonstrated on 70,000-bus system.

Abstract

This letter investigates parallelism approaches for equation and Jacobian evaluations in large-scale power flow calculation. Two levels of parallelism are proposed and analyzed: inter-model parallelism, which evaluates models in parallel, and intra-model parallelism, which evaluates calculations within each model in parallel. Parallelism techniques such as multi-threading and single instruction multiple data (SIMD) vectorization are discussed, implemented, and benchmarked as six calculation workflows. Case studies on the 70,000-bus synthetic grid show that equation evaluations can be accelerated by ten times, and the overall Newton power flow advances the state of the art by 20%.