School of Electrical, Computer and Energy Engineering
PhD Final Oral Defense
Reliability Enhancements for Real-Time Operations of Electric Power Systems
by
Xingpeng Li
11/1/2017
8am - 10am
ERC-533
Committee:
Dr. Kory Hedman (chair)
Dr. Gerald Heydt
Dr. Vijay Vittal
Dr. Jiangchao Qin
Abstract
Prior work in literature has illustrated the benefits of treating the transmission network as a flexible asset. However, the flexibility in power system networks is not fully modeled in existing real-time contingency analysis (RTCA) and real-time security-constrained economic dispatch (RT SCED) applications, where the transmission network is traditionally treated as a static network. Thus, corrective transmission switching (CTS) is proposed in this dissertation to enable RTCA and RT SCED to take advantage of the flexibility in the transmission system in a practical way.
In this dissertation, RTCA is first conducted and critical contingencies that may cause violations are identified. Then, for each critical contingency, CTS is performed to determine the beneficial switching actions that can reduce post-contingency violations. In an optimization model, transmission status is typically represented by binary variables, which increases its computational complexity. To reduce computational burden, fast heuristics are proposed to generate candidate switching lists. Numerical simulations performed on three large-scale realistic power systems (TVA, ERCOT, and PJM) demonstrate that CTS can significantly reduce post-contingency violations with the proposed heuristics. Parallel computing is also implemented in both the RTCA routine and the CTS routine to further reduce the computational time.
RT SCED is supposed to eliminate system vulnerabilities, which include the actual overloads and the potential post-contingency overloads identified by RTCA. Thus, RT SCED models both base-case network constraints and contingency-case network constraints. In this dissertation, a procedure that is consistent with current industry practices is proposed to connect RTCA and RT SCED; this procedure is referred to as Procedure-A. The branch limits of network constraints are determined by RTCA. As CTS can reduce post-contingency violations, a higher branch limit (referred to as a pseudo limit in this work) may be available for the same network constraint. Thus, Procedure-B is proposed to take advantage of the reliability benefits provided by CTS. With the proposed Procedure-B, CTS can be modeled in RT SCED implicitly through the proposed pseudo limits for contingency-case network constraints, which requires no change to existing RT SCED tools. Numerical simulations demonstrate that the proposed Procedure-A can effectively eliminate the flow violations reported by RTCA and that the proposed Procedure-B can enhance Procedure-A and reduce most of the congestion cost with consideration of CTS.
Recent work in literature shows that state estimation is subject to false data injection (FDI) cyber-attacks. The system status may be inaccurately estimated due to FDI attacks, which may mislead system operators to adjust the system improperly. This may cause system violations and reduce the reliability. Thus, a two-stage FDID approach along with several metrics and an alert system is proposed in this dissertation to effectively detect FDI attacks. In this two-stage method, the first stage is to determine whether the system is under attack and the second stage would identify the target branch. Numerical simulations demonstrate that the proposed two-stage FDID approach can successfully identify FDI cyber-attacks and would not mistakenly consider a normal load fluctuation as the sign of an FDI attack.