How cable fault location works

A buried cable fails somewhere under a street. Finding it takes four measurements, each one narrowing the search: from the whole cable, to a distance in metres, to the spot where you dig.

Everything on this page is high-voltage work. Isolate, earth and test under your utility’s switching and safety rules before connecting any instrument.

What a cable fault is

A point where the insulation between the conductor and the earthed screen has failed, or where the conductor itself is broken. Its resistance decides how you find it.

In each sketch the top line is the conductor and the bottom line is the earthed metallic screen.

  • Open circuit

    The conductor is broken. The pulse reflects back with the same polarity it was sent with.

    Low-voltage TDR alone: visible

  • Short circuit

    A low-resistance path from the conductor to the screen or another phase. The reflection comes back inverted.

    Low-voltage TDR alone: visible

  • High-resistance

    A leakage path of hundreds of ohms or more. It barely changes the cable’s impedance, so a low-voltage pulse passes it by.

    Low-voltage TDR alone: hidden, condition first

  • Intermittent

    Holds at low voltage and flashes over only above a certain voltage, often only in service. Between flashovers it looks healthy.

    Low-voltage TDR alone: hidden, condition first

Rule of thumb: power cables have a characteristic impedance of roughly 10 to 50 Ω. A shunt fault well below that reflects strongly. One many times higher, from about a kilohm upwards, reflects too little to see, which is why many faults in medium-voltage cables need step 2 before they can be measured.

Inside a medium-voltage cable

A test pulse does not run along the copper alone. It travels between the conductor and the metallic screen, guided through the insulation. So the insulation sets its speed, and every measurement starts with knowing the cable.

Conductor
Stranded copper or aluminium. Carries the load current and the test pulse.
Conductor screen
Semiconductive layer that smooths the electric field at the strands.
XLPE insulation
Holds the voltage. Its permittivity sets how fast the pulse travels.
Insulation screen
Semiconductive layer that keeps the field inside the insulation.
Metallic screen
Copper wires or tape, earthed. The return path for fault current and for the pulse.
Outer sheath
PE or PVC, often red. Keeps water out. A damaged sheath is a fault of its own.

Velocity of propagation. In XLPE cables the pulse travels at about 160 m/µs, a little over half the speed of light. Reflectometers use half that figure, v/2, because the pulse goes out and back.

A typical v/2 for XLPE is 80 m/µs; paper-insulated cables are usually a little slower. Confirm the value on a healthy phase or a known length before you trust a distance.

Cross-section of a single-core XLPE medium-voltage cable From the centre outwards: a stranded copper conductor, a thin black conductor screen, a thick milky-white XLPE insulation, a black insulation screen, a ring of copper screen wires on a bedding tape, and a red outer sheath. Outer sheath Metallic screen Insulation screen XLPE insulation Conductor screen Conductor Cross-section of a single-core XLPE medium-voltage cable Outer sheathMetallic screenInsulation screenXLPE insulationConductor screenConductor
Single-core XLPE cable, schematic. Layer thicknesses are drawn for legibility, not to scale.

Four steps, one sequence

Each step hands a narrower search to the next. Skip one and the next is guessing.

  1. Step 1: Pre-locate: measure the distance

    A time domain reflectometer (TDR) sends a short low-voltage pulse into the faulty phase and records what comes back. Wherever the cable’s impedance changes, part of the pulse reflects: at joints, at the cable end, and at the fault.

    The time to the start of the reflection, multiplied by v/2, is the distance. An open circuit reflects with the same polarity as the pulse; a short circuit reflects inverted. The further the fault, the smaller and wider its reflection, because the cable attenuates and spreads the pulse.

    Fault type
    Reflection after
    16.0µs
    v/2, XLPE, typical
    80m/µs
    Distance to fault
    1,280m

    80 m/µs × 16.0 µs = 1,280 m to an open circuit.

    Move the fault along a 2,000 m XLPE cable. Top: the cable, with the pulse going out (solid) and coming back (dashed). Below: the trace the TDR draws. The time axis and the distance axis are the same axis, scaled by v/2.
    What you measure
    Round-trip time to the reflection, converted to metres with the cable’s v/2.
    Typical pitfall
    A wrong v/2. Five per cent off in velocity is five per cent off in distance: 64 m on a 1,280 m fault.
  2. Step 2: Condition: make the fault visible

    High-resistance and intermittent faults do not show on a low-voltage trace. A surge generator charges a capacitor to several kilovolts and discharges it into the cable. At the fault the voltage breaks down the weak insulation and an arc forms. While it burns, the arc is close to a short circuit, so for a few milliseconds the fault reflects.

    With the arc reflection method (ARM), a filter couples the TDR to the cable and keeps the high voltage away from it. The TDR records one trace before the surge and one during the arc. Where the two traces part is the fault. When the arc is too short to catch, or on long cables, the impulse current method (ICM) reads the current transients of the discharge instead.

    Arc reflection method Top: a surge generator feeds the cable through an ARM filter, with the TDR connected to the filter; an arc burns at the fault. Bottom: two TDR traces. The reference trace, taken before the surge, runs flat to the cable end. The trace taken during the arc dips down at the fault. The point where they part marks the fault distance. PG-32 ARM Filter TDR-901 Cable under test Arc at the fault Before the surge: reference During the arc Traces part here 0 m Fault Cable end Arc reflection method PG-32 ARM Filter TDR-901 Arc at the fault Reference, before the surge During the arc Traces part here 0 m Fault Cable end
    The reference shows only the cable end: the high-resistance fault is invisible to it. During the arc the fault behaves like a short and reflects inverted.
    What you measure
    The distance at which the arc trace departs from the reference trace.
    Typical pitfall
    Surging higher than the fault needs. Raise the voltage in steps, and stay within the test voltage allowed for the cable.
  3. Step 3: Pinpoint: find the spot to dig

    A pre-located distance still carries the uncertainty of v/2 and of the route, often several metres. Digging needs the exact spot.

    With the surge generator firing at a steady rate, every discharge makes a sharp thump at the fault, and the sound travels up through the soil. Walk the traced route around the pre-located distance with a ground microphone. The thump is loudest directly above the fault and fades on either side.

    Compare loudness, not just presence: on hard surfaces and along ducts the sound can carry several metres.

    Acoustic pinpointing A cable is buried about one metre deep. Sound waves rise from the discharge at the fault to the surface. Bars above the ground show the loudness heard at ten stops along the route: the bar directly above the fault is the tallest, and the bars fall away on both sides. Loudness at each stop Loudest: dig here Discharge at the fault about 1 m −4 m −2 m 0 +2 m +4 m Distance along the route from the loudest point Acoustic pinpointing Loudness at each stop Loudest: dig here Discharge at the fault about 1 m −2 m 0 +2 m
    Bar heights follow 1/r for a fault 1 m deep, ignoring soil damping: a metre off the spot the thump is already about 30% quieter.
    What you measure
    The loudness of each discharge at stops along the route. The maximum marks the fault.
    Typical pitfall
    Listening while walking. Footsteps and traffic mask the thump: stand still for each shot.
  4. Step 4: Trace route: put the distance on the map

    The TDR measures along the cable, not across the street. A cable that turns a corner can be 1,280 m away along its length and well under 1,000 m away in a straight line. Without the route, the pinpointer walks the wrong line.

    A transmitter puts an audio-frequency signal on the cable. A hand-held receiver picks up the magnetic field above it and follows the cable along the street, corner by corner. Pace the distance along the traced path, and allow for slack at joints and terminations.

    In practice you trace before you walk with the pinpointer. It comes last here because it turns a distance into an address.

    Route tracing over a street plan A street plan seen from above. The traced cable leaves a substation, runs east along one street and turns south down another. The fault lies 1,280 metres along the cable, but only about 952 metres from the substation in a straight line. 952 m in a straight line 500 m 1,000 m Substation 1,280 m along the cable Traced cable route 200 m Route tracing over a street plan 952 m in a straight line 500 m 1,000 m Substation 1,280 m along the cable Traced cable route 200 m
    Same fault, two distances. Only the one along the traced route leads to the spot the pinpointer is listening for.
    What you measure
    The cable’s path at the surface, and on most receivers its depth.
    Typical pitfall
    Pacing the distance in a straight line, or along the kerb instead of above the cable.
Inside the cable test van: an instrument bench with a surge set and two orange instrument cases, a control panel with meters, a UPS, and a pull-out drawer holding a hand-held receiver and headphones.

Put the whole chain in one vehicle.

The cable test van carries every step, fitted and wired: a reflectometer, a surge generator, an acoustic receiver and a route tracer, with the high-voltage connections at the rear doors. The crew arrives ready to measure.

Price the chain for your network.

Tell us the cable voltage and length, and the faults you see most. We recommend the kit and quote it.