HPC for power systems in the framework of PEGASE project

The PEGASE project focuses on the development of new algorithms and prototypes software in the field of state estimation, optimization and time domain simulation of continental/very large systems. To comply with the TSOs new operational requirements, a strong reduction of the computation time for large scale networks is needed. In addition to improve sequential optimizations, the introduction of parallelization at different levels (from very fine to coarse grain) is one of the key elements to reach this target. This paper describes different parallelization levels and methods introduced in various power system application (from Time Domain Simulation to Security Constrained Optimal Power Flow). It outlines the main challenges and achieved speedup improvements strongly dependent on the parallelization level.