Parallel contingency analysis in a high performance computing environment

Abstract

This paper proposes a parallel processing approach to contingency analysis (CA) for power system security evaluation. Full AC power flow analysis is done for each contingency and violations of bus voltages and line flows are evaluated. All possible line outages are considered to avoid overlooking of certain critical cases. Parallel Processing is employed to increase the speed of execution. Parallelism is achieved by sharing each contingency across processors. The proposed methodology is implemented in a Linux Cluster. Data communication is performed through the Message Passing Interface (MPI). The effectiveness of parallelism is demonstrated by performing contingency analysis on standard IEEE Systems. The performance of the algorithm is analyzed in terms of computation speed and efficiency in comparison with the sequential approach.