Measurement-Based Operational Reliability Tools
PhD student Kai Van Horn with advisers A. Domínguez-García and P. Sauer
Real-time operational reliability assessment is a key component of system operators’ strategies for maintaining grid reliability. Such assessment requires tools that can provide system operators with up-to-date information about current operating conditions, e.g., transmission lines at their thermal limits, as well as information about the impacts of potential unplanned outages, e.g., whether or not the current operating point N-1 is reliable. To this end, the real-time security-constrained economic dispatch (SCED) is a widely used market-scheduling tool that seeks to economically balance electricity system supply and demand, and provide locational marginal prices (LMPs) while upholding system reliability standards. The SCED is usually formulated as a convex optimization problem with a piecewise-linear or quadratic objective, typically the minimization of generator costs. The constraints capture equipment limits, the power balance, network-flow capacity limits, and preventative security considerations. These are typically formulated using a linearized power flow model. In this work, we propose a measurement-based approach to the real-time SCED. Furthermore, we propose additional constraints in the SCED that will enable dispatching the system while considering the impacts of line outages on the voltage angle across the line’s terminals, which, if large, can prevent the reclosure of the line—a reliability issue exposed by the 2011 San Diego Blackout.
Our approach, an overview of which is shown in Fig. 1, utilizes power system sensitivities estimated from phasor measurement unit (PMU) measurements to reformulate the model-based power balance and network flow constraints. We estimate the loss factors and injection shift factors from PMU measurements of active power consumption and production, loads, and generators in the system, and active power flow on the lines. Furthermore, we utilize the estimated sensitivities in contingency analysis to identify potential contingency-driven equipment overloads and to formulate corresponding security constraints in the SCED. Additionally, we derive from the sensitivity estimates a sensitivity that allows for the prediction of the voltage angle across a line when it is outaged and show how such sensitivities can be integrated into the SCED to guard against large outage angles, such as those that precipitated the San Diego blackout. We also show how these angle sensitivities can be used to compute an angle component of the LMP, which provides economic signals as to the locations at which active power injection exacerbates/relieves large outage angles. As is borne out in our studies on standard power systems test systems, the resulting measurement-based real-time SCED is robust against undetected topology changes, changes in the system operating point, and the existence of incorrect model data. Moreover, the dispatch instructions and LMPs calculated with our proposed measurement-based real-time SCED accurately reflect real-time system conditions and operational reliability needs and can capture line outages impacts on the angle of a line’s voltage. Our next steps will be to test the sensitivity estimation algorithms using real PMU data and implement the measurement-based SCED in a larger test system to demonstrate its scalability.This research is supported by UT- Knoxville and Grainger Fellowships.