Cyber Resilient Remedial Action Schemes Vignesh V Contributors : - PowerPoint PPT Presentation

Cyber Resilient Remedial Action Schemes Vignesh V Contributors : Ren Liu, A.Askerman, A. Srivastava and D. Bakken The School of Electrical Engineering and Computer Science Washington State University Patrick Panciatici RTE France March 29, 2017 1

Closed Loop Control in Power System Power system control and operation is highly dependent on • communication and data technologies. Any cyber failure can be hazardous to the reliable operation • of the power grid. Control in power system can be at: • Substation –Protection • At multiple Substations : RAS/ Special Protraction • Schemes At control Center : Economic Dispatch, etc • Main focus is on faster control at multiple Substations

Closed Loop Control in Power System Sensors – PMUS, DFRs, RTU, Weather sensors • Network - LAN, WAN,VPN • Computational Framework – Control Center • Actuators - Relays, motors • Communicati Power System Data on Network Sensors Controller/ Decision Relay Making Actuator

Control Architectures • Traditional Implementation –Prone to node failures. • Decentralized and distributed scheme -Resilient to any single node failure and scalable. • Need for cyber resiliency in decentralized and distributed architectures. 4

Distributed Computational Architecture Divide the power system into various logical groups • Assign a leader node for each group. • – Collects measurements from buses from its group – Exchanges data from leaders of other groups – Triggers the actuators in its own group 5

Distributed Computational Architecture The algorithm is mathematically distributed over the computing nodes. • Each area solves a part of the overall problem. • Exchanges data such as border bus information, tie lie flows etc, with • the neighboring area Solution of distributed algorithm converges to “master” centralized • solution under certain mathematical assumptions. Can be made to converge to “near” desired solution even under the • failure of one of the computing nodes. 6

Case Study- Optimal Wind Curtailment Remedial Action Scheme (RAS) In order to minimize the wind power curtailment and • protect the transmission lines from overload problem or stability problems, RAS can be deployed for faster control action. Existing RAS are hard coded based on pre-determined • control actions. New fault-tolerant computational approach needs to be • developed to provide the fast and correct data to dynamic and adaptive RAS. 7

Case Study-Mathematical Formulation The RAS goal is to minimize the wind power curtailment and • mitigate the overload condition on the transmission line at the same time developed based on linear programming and DC power flow The objective function of this problem can be written as: • 8

Case Study: Validation 9

Validation of Cyber Resilient Control 10

Results • Demo Video 11

Conclusions To improve the resiliency of cyber physical system • decentralized and distributed architecture is proposed. The power system is divided into logical areas and control • center for each area to improve architectural resiliency The mathematical computation is distributed over multiple • computer nodes to achieve computational resiliency. A demonstrative distributed RAS application is being • developed. Preliminary real time results are shown. • Additional failures and cyber threats are being modeled • and simulated to validate the cyber-resilient control architecture. 12