New Protection Method for Automated Distribution Feeders
Written by Thierry F. Godart and Andre Smit
The constant variations in line and source topologies found in a typical automated distribution feeder can create design challenges for protection engineers, especially those using traditional overcurrent protection methods. A novel approach using differential protection recently debuted on an A&N Electric Cooperative feeder in Virginia, demonstrating the ability to reduce the number of required settings groups without compromising feeder flexibility or the speed, accuracy, and selectivity of protection.
Coordinating an automated distribution feeder can be a complicated undertaking. Even on a simple feeder comprising four protection zones, the number of settings required to achieve coordination can quickly outstrip the processing capabilities of contemporary relays. To resolve the issue, protection engineers limit the number of possible feeder configurations to reduce the number of required protection settings.
A&N Electric Cooperative of Tasley, Va. recently automated a feeder to supply the only hospital in its service area. The consulting firm hired to coordinate the protection said it couldn't be done—the number of potential line configurations and source variations made coordination seem an impossible task.
In automating the feeder, A&N elected to implement conventional time-coordinated overcurrent protection. To make sure faults were not fleeting or spurious, an overcurrent trip was followed by an instantaneous reclose cycle. If the fault was still present, a three-second delay (dead time) was instituted in advance of another reclose onto the fault using a short overcurrent time. If after the second reclose cycle the fault was still present, a five-second delay was instituted in advance of a third reclose onto the fault using a long overcurrent time to give the line fuses time to clear the trouble. If the problem persisted, a trip was initiated followed by a lockout.
Although the consulting engineers believed coordination to be impossible, A&N engineers decided to look into the matter. In doing so, all possible feeder topologies were taken into account, such as those caused by fault-isolated line sections, loss-of-source transfers, and shifting feeder open points to correct load imbalances. Through a laborious process of elimination, four sets of coordination settings were developed that would work with all topologies. However, the coordination became quite tight and in some cases the allowable error tolerance of the relays was nearly reached.
When the automation system was commissioned, engineers noted that whenever a fault was not cleared at the first reclose cycle and persisted through the following cycle, emergency generators at the hospital would start when the fault was upstream. A&N engineers determined that a loss-of-source exceeding 25 cycles would activate the generators. This scenario was not acceptable to A&N management and they requested a solution to the problem.
To resolve the issue, the engineers decided to apply the automation system's differential fault location tool—Siemens' j-Diff Fault Location Technology—as a protection function to isolate faulted sections along the feeder. Subsequent actual operations in the field proved the speed, accuracy and efficacy of the approach. Although not a traditional differential protection function in that it does not correlate complex phasor information, the approach does incorporate controllers that compare line currents among the feeder sections to locate "jumps," or current disturbances in the line. To promote operating speed and security, the controllers use IEC 61850 GOOSE messaging to share line status information among intelligent devices in the system.
Since the application of the differential approach, faults are located quickly and accurately, which allows affected line sections to be rapidly isolated, and service is restored to unfaulted sections, minimizing the need for backup power. Faulted sections are then reclosed to determine whether faults are temporary and if so, service continues nearly uninterrupted in these sections. If faults are permanent, the affected areas are isolated pending repairs.
In addition to quickly detecting and isolating faults, the j-Diff Fault Location & Protection Technology drastically simplifies protection coordination, obviating the need for multiple settings groups. When a fault occurs and a reclose cycle is implemented, a simple, proven time-current curve is applied only to the faulted feeder section. Because of the similar nature of the sections making up the feeder, a single uniform curve may be used on all protection points on the line. The coordination only needs to consider the source feeding the section and downstream fuses.
The new automation system has been in service on A&N Electric for over a year. Although there have been few operations, the system has performed as designed for all faults and reconfiguration actions, including during Hurricane Irene, which directly affected the cooperative's service area. For emergency protection, a backup overcurrent system is always in place.
In addition to initial field tests conducted by Siemens, A&N subsequently commissioned extensive protective relay systems testing (RTDS) to implement logic to formally convert the automated fault location function into a true protection function. The latter testing was performed by Quanta Technology of Raleigh, N.C., which was tasked with identifying any shortcomings that might have existed in the new approach. Test results showed the method to be absolutely reliable, which has been the case on A&N with both passive and active loads. Results of the testing and a report on actual field performance will be published in a forthcoming paper.
Moving forward, A&N Electric has gained the confidence to automate additional distribution feeders. As a relatively small cooperative, the cost and complexity to automate and coordinate a feeder are very important considerations. The application of smart grid thinking, NIST-approved technologies, and the differential protection approach has helped A&N to streamline coordination and rollout, minimize costs, reduce complexity and increase system reliability.
Thierry F. Godart is president of the Siemens Smart Grid Division in North America. A graduate of the École Supérieure d'Electricité (SUPELEC) in Paris, France, he has more than twenty years experience in the application of information technology to the power industry. Godart earned a doctoral degree in electrical engineering and master's degrees in applied mathematics and electrical engineering at Georgia Tech. He is a member of the IEEE Power & Energy Society and CIGRÉ.
Andre Smit has worked in the protective relaying field for more than two decades. He studied at Vaal Triangle Technikon in South Africa and graduated in 1987 with a degree in electrical power engineering. He joined Siemens in 1989 and in 1998 moved to Siemens USA, where he developed a system engineering and production capability for protection systems. A member of the IEEE Power Systems Relaying Committee, Smit directed the Siemens USA protective relay business unit from 2004 to 2008.