Distributed Traveling-Wave Location for Surge Arrester Operation Counting And on Transmission Lines

Surge arresters on transmission lines are exposed to a dual burden: they must discharge lightning and switching surges, and their metal-oxide varistor (MOV) elements degrade cumulatively with every operation and with sustained leakage current. The conventional instrument for tracking this burden is the surge counter—a local, discharge-triggered register attached to the arrester ground lead. Its limitations are well documented. Counters are read during periodic patrols, so the temporal resolution between inspections is coarse. They count impulses above a manufacturer-defined amplitude threshold, yet switching operations on a 400 kV line can produce traveling-wave overvoltages that trigger counting with a probability as high as 98.6% in simulation, falling to about 4% once corona is properly modeled. This discrepancy means that counter records alone are an unreliable proxy for the arrester's actual electrical duty.

Distributed traveling-wave (TW) fault-location devices offer a structurally different observation layer. These terminals are installed along the overhead line—on conductors, towers, or cable sections—typically spaced 20–30 km apart, and they report to a central station as a line-deployed monitoring network. Each terminal captures the arrival time of the initial traveling-wave front launched by a lightning stroke or switching event. Because the wave propagates at a speed close to the speed of light in the line medium, the arrival-time difference between two adjacent terminals localizes the event to within hundreds of meters or better. This localization is independent of load current, fault resistance, and system impedance, making it robust in weak-network conditions.

For arrester operation counting, the TW layer supplies something the counter cannot: an independent, time-stamped event record. When a traveling wave is localized to a span containing a line arrester, the event can be associated with that arrester's discharge path. This association transforms counting from a local register read into a network-level correlation. A counter reading of "one operation since last patrol" can be matched to a specific stroke time, location, and estimated current class. Commercial monitoring systems already classify discharge events into amplitude bands—for example, 0.1–1 kA, 1–10 kA, and 10–100 kA—and store timestamped records in non-volatile memory. The TW layer adds the spatial coordinate and the traveling-wave signature, allowing the central station to distinguish a genuine arrester discharge from a switching transient that happened to exceed the counter threshold.

The early-warning function rests on combining this operation history with a second indicator: resistive leakage current. Aging of metal-oxide elements—driven by surface pollution, moisture ingress, and electrothermal stress—raises the resistive component of leakage current, and if the process reaches thermal runaway, the arrester can fail prematurely. The fundamental resistive leakage current amplitude I_R and the third-harmonic resistive component I_{3rd} are established condition indicators; their ratios against total leakage current provide a normalized measure of degradation. IEC 60099-5:2013 provides the diagnostic framework, including expanded discussion of performance diagnostic tools and end-of-life considerations for arresters in service.

A distributed TW architecture turns these two data streams—event-localized operation counts and leakage-current trend—into a predictive signal rather than two disconnected records. The operating logic proceeds in three steps. First, each TW-confirmed discharge increments a location-specific operation counter with a timestamp and a current-band label. Second, the same central station archives leakage-current measurements from monitoring outputs at the arrester, tracking I_R and I_{3rd} normalized by total leakage current. Third, the warning engine evaluates the joint trajectory: an arrester whose TW-confirmed operation count is accelerating while I_R or I_{3rd} shows an upward drift is flagged for inspection before the next scheduled patrol. Conversely, a stable leakage trend with a moderate operation count supports continued service. The value of the TW layer is that it removes the counting uncertainty that otherwise contaminates the operation-side input to this logic.

What distinguishes this approach from conventional condition monitoring is that the traveling-wave device is not repurposed as a surge counter. It remains a fault-location and disturbance-analysis instrument. Its contribution to arrester management is a byproduct of its primary function: every localized event carries a time and a place, and those two attributes are precisely what a local counter lacks. The incremental engineering effort is therefore modest—a data-association rule in the central station and a joint warning criterion—while the information gain is a continuous, network-wide record of where and when the arresters are being exercised, set against the slow drift of their leakage-current health indicators.