Real-time monitoring of power systems using fast second-order method

Abstract

One of the basic approaches to the static state estimation of power systems involves the solution of the so-called independent equations using conventional load-flow techniques, such as Newton-Raphson (NR) and fast decoupled (FD). Of these two, the FD method is preferred to the NR approach, primarily because of reduced computational requirements. As is well known, the superiority of the former is primarily attributable to the underlying `decoupling? assumptions, among others. However, since these assumptions are not valid for all modes of system operation, the results provided by the FD estimator may be unreliable. Moreover, as the computational requirements of the NR estimator are prohibitively large and the results obtained from the FD method might be suspect, especially in the real-time environment, this paper presents a new steady-state estimator, combining all the attractive features of both the NR and FD methods, while remaining free from their limitations. The new estimator is based on a complete Taylor's series expansion of nodal and line-flow equations in Cartesian co-ordinates. The capability of the proposed method to serve as a real-time monitor is demonstrated by digital simulation studies on a number of sample power systems, the results being compared with those of the FD method. This comparison reveals that, owing to its exact formulation, lower computational requirements, and the ability to provide a reliable system state even during unusual operating conditions, the proposed method is a practically viable and preferred analytical tool for real-time monitoring of power systems.