Agent Based Supervision of Zone 3 Relays to Prevent Hidden Failure Based Tripping 1 Shravan Garlapati, Hua Lin, Santhoshkumar Sambamoorthy, Sandeep K. Shukla, James Thorp Bradley Department of Electrical and Computer Engineering, Virginia Tech Blacksburg, Virginia E-mail address: {gshra09, birchlin, ssan, shukla, jsthorp}@vt.edu Abstract— In this paper, we propose a distributed agent based supervisory scheme to make Zone 3 relays robust to hidden failure induced tripping, facilitated by the communication network -- soon to become an integral parts of the smart grid. Possible elimination of Zone 3 relays (remote backup protection) has been studied in the recent past and these remote backup relays have been adjudged to be essential for power system protection [23]. Even though Zone 3 relays are often overly sensitive to remote line overloading, and are known to cause unwarranted trips during cascading failure scenarios, they are prescribed as acceptable means for remote backup. Therefore, providing robustness to Zone 3 relays to minimize the risk of erroneous trips, especially when hidden failures [10, 1] make them vulnerable to over reaction, is an important problem. In our scheme, a synchronous grid is populated with agents at each relay, and an agent hierarchy is maintained in master/slave relationship. The communication established between relay agents decreases the probability of erroneous Zone 3 trips thereby preventing them from aggravating cascading failure scenarios, and reducing the probability of cascading blackouts. Unlike other agent based relay proposals, ours is a nonintrusive approach. Keywords-component; Zone 3 Relays, Remote Backup Protection, Hidden Failure, Cascading Blackouts, Agents, Communication, TCP/IP, UDP I. INTRODUCTION Transmission systems with protective relaying usually have redundancy in the form of local and remote backups. The idea is to quickly isolate the fault conditions and power system equipments from the whole system in order to ensure stability and security. This needs to be done with minimal disruption of service to customers. It has been observed [3, 23] that local backups could be susceptible to ‘common mode failures’ due to sharing of the same electrical and communication infrastructure. Remote backup or Zone 3 relays are usually located in remote substation, isolated from the substation where the original disturbance occurred, and hence is not susceptible to such ‘common mode failures’. However, remote backup requires longer fault clearing times, and more load may be removed when a remote backup is called upon to isolate a fault. Therefore, unless absolutely necessary (when Zone 1 and Zone 2 relays fail to isolate the fault for various 1 This research was partially funded by an NSF-EFRI grant. reasons), it should be not allowed to trip. After a comprehensive study of a number of historically disastrous blackouts such as the 1965 Northeast blackout, the 1977 New York blackout, and the 1996 western blackout, NERC authenticated that unwarranted zone 3 relay failures have been involved in 70% of the blackout events [16, 2]. But [3, 23] clarified that Zone 3 relays cannot be dispensed with. It is known that Zone 3 relays can actually erroneously trip due to hidden failures [10, 1]. A hidden Failure is usually rare but could happen due to software or hardware errors in the Zone 3 relay. It may go unidentified for a long time. However, such problems may manifest as extra sensitivity of a Zone 3 relay to even remote line overloading. Even though such an overloading might be transient, or might not have reached a level where the Zone 1 and 2 relays need to act, an overactive Zone 3 relay may trip, starting a sequence of other trips which may lead to a cascading failure. In [1] a large number of such scenarios were simulated and many of the cascading failure so discovered matched with historical chains of events in cascading blackouts recorded by National Electric Reliability Council (NERC). Hidden failure may be present in any equipment in the Electric power network, especially as they become more hardware/software dependent. One of the main components of the Smart Grid vision is that the power system will be enabled with communication networking to unprecedented level, and wide area measurements and controls will provide the power system (transmission and distribution) with unprecedented robustness. The architecture, media, protocols for such communication network is still being developed, but it is conceivable that empowered by the wide area visibility, fast communication, and computation, control functions can indeed provide such robustness and prevent untoward incidents such as cascading blackouts. In this paper, we are considering one question in this context. The question is: “How can a communication network and fast computing abilities enhance the functionality of Zone 3 relays so that they can be robust to hidden failures, and over active undesired tripping can be prevented?” In this paper we attempt to answer this question by first developing a distributed hierarchical agent based scheme and then demonstrating a few simulations based experimental data to backup the possible validity of our scheme. However, the problem at hand is as much a problem in the domain of distributed fault-tolerant computing with networked 978-1-4244-6511-8/10/$26.00 ©2010 IEEE 256