Symbolic Factorization Re-utilization for Contingency Analysis

As the power system evolves and is operated closer to its operational limits, accurate and real-time contignency analysis becomes more critical. The speed of power flow calculation is a limiting factor of contingency analysis real-time applications. In this work, graph theory and NICSLU, a state-of-the-art sparse and parallel circuit solver, are used to analyze how post-contingency power flows can be solved more quickly. Graph theory is used to show that the base case symbolic factorization can be reused to solve contingency cases. Statistical analysis demonstrates that this method is preferable to recomputing or updating the symbolic factorization for contingencies involving few branch elements. Experimental results on Sichuan and MP10790 systems show that symbolic factorization reuse greatly reduces computation time.