TDR and Cable Fault Location: Finding Faults After VLF Identifies Them

Locating a fault in medium-voltage cable is a multi-stage process. The first stage, pre-location, uses dielectric testing methods such as VLF tan delta testing or DC high-potential testing to confirm that insulation has degraded or failed and to characterize the fault impedance. Pre-location does not yield a distance to the fault, only a pass/fail on the cable condition. The second stage uses time-domain reflectometry (TDR) or bridge methods to calculate a distance to the fault from one end of the cable. The third stage, pinpointing, uses a surge generator (thumper) to create a repeating acoustic pulse at the fault site and an audio-frequency electromagnetic locator to trace the cable route and listen for the discharge through the ground surface.

The boundary between pre-location and rough location, and between rough location and pinpointing, depends on the fault impedance. A low-impedance fault (shorted phases or phase-to-ground resistance in the hundreds of ohms range) is directly measurable with a TDR. A high-impedance fault, common with XLPE insulation that has degraded at a water tree but not yet fully bridged, may resist TDR measurement and require conditioning with an impulse generator before it will produce a measurable reflection.

TDR fundamentals: how it works

A time-domain reflectometer injects a fast-rise-time voltage pulse into one end of the cable and measures the time for reflections to return. When the pulse traveling down the cable encounters any impedance discontinuity, a splice, a fault, the far end, or a damaged section, part of its energy reflects back to the source. The TDR measures the elapsed time between the transmitted pulse and the received reflection, and uses the known velocity of propagation for the cable type to convert elapsed time into a distance.

Distance = (velocity of propagation × elapsed time) / 2. The factor of two accounts for the round trip, the pulse travels from the test point to the fault and back. For solid dielectric cables such as XLPE (cross-linked polyethylene), the velocity of propagation is approximately 50% of the speed of light in air, giving a value of approximately 150 m/µs. For paper-oil insulated cable, the value is slightly different, typically 50–55% of the speed of light depending on insulating compound. Entering the correct velocity of propagation is the most common source of TDR distance error; using the value for one cable type on a splice run that includes a different cable construction shifts every distance by a proportional error.

Reading a TDR trace

A TDR trace displays amplitude (the reflected signal) versus distance (calculated from time). On an undamaged cable, the trace shows a transmitted pulse at distance zero, a flat baseline through the cable length, and an end reflection at the far end, a positive reflection for an open-circuited far end, a negative reflection for a short-circuited far end. Splices show as small amplitude discontinuities at the splice location, the impedance mismatch at the splice interface creates a partial reflection and a slight change in the baseline level on either side.

A fault reflection appears as an impedance discontinuity between the launch point and the end reflection. A low-impedance fault (shorted) produces a negative reflection similar in polarity to the end reflection from a short; a high-impedance fault produces a smaller positive reflection. The distance reading from the TDR is a rough location, it indicates a section of cable, not a specific point. TDR distance accuracy is typically ±1–2% of cable length, meaning a fault in a 500-meter run will be located to within approximately 5–10 meters of its actual position. That's sufficient to identify the right manholes on either side of the fault for subsequent pinpointing.

High-impedance faults and conditioning

Many real cable faults, particularly in aged XLPE cable with water tree damage, present as high-impedance faults in the megohm range under TDR test voltage. A high-impedance fault does not produce a measurable TDR reflection because there is insufficient impedance mismatch at the fault point to create a detectable reflection within the noise floor of the instrument. The fault must be conditioned, reduced to a low-impedance state, before TDR rough location can proceed.

Conditioning is accomplished by applying a surge voltage from a high-energy impulse generator at sufficient voltage to break the fault down from a high-impedance state to a low-impedance arc. The process is called burn-down or fault conditioning. The surge generator applies repeated high-voltage impulses across the fault, typically 10–25 kV for distribution cable, until the fault resistance drops below a level that the TDR can measure, typically below a few kilohms. After conditioning, the TDR is reconnected and the fault reflection becomes visible. The trade-off is that conditioning further damages the insulation around the fault site, which is acceptable because the cable requires splicing or replacement regardless.

Surge generator (thumper) pinpointing

Once the TDR has identified the rough location, a section of cable within two adjacent manholes, the surge generator is used to pinpoint the exact fault position. The surge generator charges a capacitor bank to a high voltage (typically 10–30 kV for MV cable) and then discharges it through the cable at the fault point. The discharge creates a loud acoustic impulse at the fault site, a thump that is audible above grade as a rhythmic knocking sound if the fault is in a duct bank under a road or sidewalk, or as a sharp crack if the cable is directly buried.

The surveyor walks the route between the two manholes identified by the TDR, listening with a ground microphone or electromagnetic sensor for the characteristic discharge sound. When the thumping is loudest, the fault is directly underfoot. Because the exact cable route may not be known or documented, particularly in older installations with handwritten as-built drawings of questionable accuracy, the acoustic survey also serves to trace the cable path. The surge generator pulses at a slow rate (typically one discharge per 2–5 seconds) to give the surveyor time to move and listen between discharges.

Audio tone tracing

For cable route tracing where a fault is not yet involved, for example, to verify cable routing before excavation, or to trace a cable that has been incorrectly documented, audio frequency current injection is used in place of the surge generator. A tone generator injects a continuous AC signal at a frequency typically in the 8 kHz range into the cable, using the cable shield or armor as the return path. A detector clamp or probe carried by the surveyor picks up the magnetic field radiated by the current in the cable, producing an audio tone in headphones that reaches maximum intensity when the detector is directly over the cable route.

Audio tracing can locate a fault indirectly by identifying the cable route and narrowing the fault to a section identified by TDR, but it cannot identify the precise fault point within that section. Its primary use in fault location work is route tracing, confirming which manholes the cable passes through and its approximate depth, before the thumping phase begins. In congested duct banks where multiple cables of similar size are present, audio tracing also discriminates which cable is being tested by comparing the trace on the energized cable against the dead cables on either side.

Floating and intermittent faults

Some cable faults resist the entire fault location sequence. A floating fault, a fault where the damaged insulation reforms its dielectric under the applied test voltage and presents as no fault at all, will not produce a TDR reflection and will not condition under surge voltage. This type of fault is most common in cables that failed momentarily during a switching transient or lightning event and appear to have recovered. The cable tests clean under DC or VLF, and the TDR shows no anomaly. These cables often fail again in service under a subsequent transient and may test clean again afterward. Trending tan delta over multiple test cycles, looking for an increasing tip-up that indicates deteriorating insulation, is the most reliable diagnostic for a cable heading toward failure that has not yet produced a persistent fault.

An intermittent fault, one that appears and disappears depending on cable temperature, voltage, or moisture conditions, requires conditioning at the conditions under which it manifests. A fault that only appears when the cable is at operating temperature may require the cable to be loaded before the TDR will see it, or conditioning at elevated voltage that forces the fault to a permanent low-impedance state. For intermittent faults in critical circuits, the practical answer is often replacement of the suspect section rather than extended diagnostic effort on a cable that has demonstrated unreliability.