Locating Faults By Using Incremental Quantities: Introduction, Application Considerations, and Performance

Fault locating plays an important role in maintaining a reliable power system, and it has become a significant area of interest for transmission line owners and operators. Obtaining an accurate fault location allows for fast service restoration and facilitates identifying corrective measures to prevent recurring faults. Dependable fault locating ensures that a fault location result is available for all fault conditions. For decades, impedance-based fault-locating methods have proven to be dependable, but they are significantly less accurate than transient-based methods, such as those that use traveling waves (TWs). More recently, TW-based methods have proven to be very accurate for a wide range of faults and system conditions, but they may have minor dependability challenges that are related to factors such as the system topology, line length, fault distance, fault resistance, and fault point on wave. This paper presents a novel multi-ended fault-locating method that provides improved accuracy over impedance-based methods and maintains very high dependability for conditions that may challenge TW-based methods. The new method works in the time domain (i.e., it uses samples from an arbitrary data window rather than full-cycle phasors), and it is based on the observation that the change in voltage at the fault location estimated from the local terminal is the same as that estimated from the remote terminal if the fault location assumed in the calculations is correct. The method uses the measured instantaneous incremental voltage and instantaneous incremental replica current along with the magnitude of the positive-sequence line impedance to calculate the change in voltage at the fault location as a function of the per-unit distance to the fault, m. The method applies the Least Errors Squared (LES) algorithm with an arbitrary fault data window to find the value of m that yields the closest match between the calculated voltage changes. To improve accuracy, the method uses the LES approach to match voltage changes in multiple loops. Additionally, using multiple loops allows the method to work without requiring faulted-loop selection logic. The paper introduces the new incremental-quantities-based fault-locating method and then demonstrates the performance by analyzing several real-world event records captured by ultra-high-speed (UHS) line protective relays in the field. The examples show that the method works very well for evolving and short-lived faults as well as for resistive faults, including arcing faults that make the fault voltage-current relationship nonstationary and nonlinear. The paper discusses accuracy and application considerations, and it provides test results from applying the new method to a large number of fault records captured by UHS relays in the field.