 |
Center for Nonlinear Studies |
As significant variable generation is added to the power grid, operators have to make decisions in the presence of power output uncertainties at multiple locations throughout the network. The cumulative effects of all the uncertainties on the system are non-obvious, and incorrect assessment of the cumulative effect could affect the reliability and stability of the power system. This paper proposes a method of combining confidence intervals of short term (one to eight hours ahead) wind forecast errors with the existing unit commitment on the system to determine possible operational impacts. The output is a visualization that can alert system operators to the potential for transmission line overloads and indicate whether changes to the existing commitment schedule should be considered. In addition, a modified confidence interval estimation method has been developed that, based on simulation results, outperforms existing methods for short term wind forecasts.
Host: Misha Chertkov, chertkov@lanl.gov, 665-8119