Analysis of Simultaneous Faults Using Short Circuit Simulations and Fault Records

Author: Yujie I. Lu and Jeffrey Pond, National Grid, USA

Case Two

On May 27, 2007 the System Control Center reported that at 23:33:32 a large oak tree fell on the S197 Line at Brittingham Hill Road near Tower 126 during a thunderstorm. At Bear Swamp 1019 the 197 and T31 circuit breakers (CBs) tripped, and at Deerfield 4 the 3T40 CB tripped (figure 4). At 23:33:33 the 1205W and 1205E CBs opened. At 23:33:37 the 197 CB closed auto and tripped immediately. All breakers were locked out. See table 3 for the reported targets for this disturbance. (Figure 4)

The S197 Line fault was confirmed when the tree was found on the line. However, there was no evidence of a fault on the TR1 Line. The TR1 Line operation was a suspected improper operation of the directional ground (67N) protective relay. Was this an improper operation or a separate incident? Prior to the start of the investigation it was confirmed that the power system was operating normally prior to the fault and the generators at Bear Swamp 19, which is a pump storage facility, were off-line. The initial analysis using the one-line diagram in figure 4 agreed with the theory that the TR1 Line protection did mis-operate for the line fault because the faults were behind the protected zone of the TR1 directional ground relay. However, the Control Center report stated that the 1205W and 1205E tripped after the 197 and T31 CBs and before the 197 CB auto-closed suggesting this may be a separate event. The next phase of the analysis was to correlate the reported targets and the fault records retrieved from the DFR at Bear Swamp Switchyard to determine the actual sequence of events for the fault.

Analysis of the fault records determined the initial fault occurred at 23:33:29.610 when the tree fell on the S197 Line causing a B-C-phase fault. The 197 and T31 CBs at Bear Swamp and the 3T40 CB at Deerfield 4 tripped on DD Z1 S1 (figure 5). It was determined that the TR1 Line protection did not operate for the first S197 fault or immediately after.

The fault record triggered at 23:33:34.802 shows the 197 CB auto closed, after five seconds, into a B-G fault, caused when a transmission tower was damaged as a result of the fallen tree, and tripped immediately (figure 6 & 8) on S197 IOC B-ph. At the same time the TR1 Line DG relay operated initiating transfer trip to the Bear Swamp Powerhouse, and tripped the 1205E and 1205W CBs (figure 7 & 8). The S197 and TR1 Line CBs were at lockout after the second fault occurred. At this point it was clear the TR1 Line operation was not a separate incident but was the result of the auto-close attempt to re-energize the line. The investigation now focused on what caused the TR1 Line directional ground relay to operate. Further analysis of the fault record in figure 7 determined that the TR1 Line supplied zero-sequence fault current at the time the 197 CB auto-closed. The record shows that the TR1 Line A-, B-, and C-phase currents are additive and are summed in the TR1 Line neutral-current. The results obtained from the analysis now indicate that the TR1 Line directional ground relay operated properly for this fault condition.

To confirm the TR1 directional ground relay operated properly for the fault, a short circuit simulation was performed. If the relay operated for the conditions in the fault simulation would prove the operation was correct, while verifying that National Grid's system model was accurate. The simultaneous fault module would not allow placing two faults on the same line segment, therefore, it was decided that in order to create the simultaneous fault in the short circuit model, first, a phase to ground fault was placed on the 115kV Bus at Bear Swamp because tower 126 where the B-phase to ground fault occurred was approximately one mile from the Bear Swamp Switchyard. Second, a three phase to ground fault was placed on the S197 Line near Bear Swamp (figure 9). The simulation confirmed that the Directional Ground relay, which is a Westinghouse IRQ-9 with negative sequence polarizing, had sufficient polarizing volt-amperes in the forward direction and neutral current to operate. As a final check the TR1 line negative sequence quantities were calculated from the fault records to be 2.88VA, well above the 0.76 VA minimum pickup of the relays directional unit, and the phasors were compared to the IRQ directional unit characteristic. As can be seen in figure 10 the calculated negative-sequence components are similar and match the relay characteristic.

Based on the analysis it was determined that for the initial fault the S197 Line directional distance protection operated properly at both Bear Swamp Switchyard and Deerfield 4 for the B- to C-phase fault caused by the falling tree. After the initial phase to phase fault the tree came to rest across all three phases of the S197 line creating a permanent three-phase to ground fault and causing damage to Tower 126 which is lattice steel construction. As a result of the damage caused by tree falling on the line a permanent B-phase to ground fault was created on the S197 Line at tower 126. When the 197 CB closed auto in five seconds the B-phase to ground fault, caused by the damaged tower, was supplied by the S197 line at Bear Swamp. The TR1 Line supplied zero-sequence current to the three-phase to ground fault through the Generator 1 GSU at Bear Swamp 19 as was determined from the fault record (figure 7). Based on the analysis the TR1 Line DG relay operation was determined to be a correct operation even though the two faults were behind the relay.

PDF Version

Ad: Navigate the Sea of Data
Ad: Do you think the demands on your substation have changed in 50 years? Then why are you still using 50-year old  test switch technology?
Ad: ABB Relion®. The perfect choice for every application.