Computational analysis of adjoint network-based methods applied to contingency evaluation

Abstract

Contingency evaluation methods are typically based on algebraic manipulation of the load flow equations around a known base-case solution. Alternatively, it is possible to make use of an adjoint network-based approach to evaluate the state of a power system for simulation of single or multiple branch outages and single or multiple power outages. In this paper we analyse the most promising variants of adjoint network-based methods applied to state variable evaluation for steady state power system contingency analysis, including the case of network equivalents. The computational requirements of the methods for medium- and large-scale systems are quantified. From the computational point of view, these methods are then compared with conventional algebraic methods. In view of this comparison, we make an alternative proposal for contingency evaluation studies which is more reliable and computationally less demanding.