Real-time remedial action screening using direct stability analysis methods

Current utility practice consists of offline stability analyses to determine remedial action strategies that are eventually “programmed” into operating procedures. However, these strategies are often sub-optimal with respect to real-time operating conditions. Although utilities perform online contingency analyses, these real-time analyses only involve static conditions. The speaker is engaged in a research project, sponsored by the US Department of Energy, that uses a direct method, based on Lyapunov functions, to rapidly screen, in real time, the set of possible system trajectories to determine which remedial actions will be most likely to result in system stabilization. The project involves several fundamental developments, in the areas of composite Lyapunov functions, homotopy theory, and the accommodation of stochasticity of operating conditions within this computational framework. The presentation will discuss this emerging methodology, and the advances made in the above areas.