Underground cable fault locating field guide for electrical crews

Cable fault locating, and why it is a find-the-spot job

Cable fault locating is the work of finding where a buried or concealed cable failed, close enough that you dig one hole and fix one splice instead of trenching the whole run. The cable is in the ground. You cannot see the damage, and the fault could be anywhere across hundreds or thousands of feet. The job is to turn an outage somewhere on the run into a mark on the ground you can put a shovel through.

It splits into two questions that get answered with different tools. First, how far down the cable is the fault, measured from the end you are standing at. That is prelocation, and it gets you to within a few feet. Second, exactly where on the surface is that point so you know where to open the trench. That is pinpointing, and it gets you to within inches. Prelocate first, pinpoint second. Reverse that order and you are walking a microphone over a quarter mile of grass hoping to get lucky.

This guide picks up after the cable has already failed. Confirming that the insulation is bad and grading its condition is the megohmmeter's job, covered in the insulation resistance guide. Catching a hot termination before it ever faults is the infrared guide. Here the cable is already down, and the only question that matters is where.

Why locate the fault instead of replacing the run?

Because the dig is the expensive part, not the splice. The cable is usually the cheapest thing in the equation. The cost lives in the excavation, the pavement you cut and restore, the traffic control, the duct bank you might damage getting in, and the hours the load sits dead while you work. Locate the fault to a single point and you open one hole. Guess, and you open a trench.

Digging blind is not a real option on most runs anyway. A direct-buried cable under a parking lot, a feeder in a duct bank under a road, a campus loop crossing landscaping and walkways. You cannot trench all of it, and on a duct bank you often cannot even see the cable until you are most of the way down. The locate is what makes the repair possible at all.

There is a time cost to getting it wrong, too. The faulted cable is the outage, and the clock on that outage runs until the cable is back. A clean prelocation and pinpoint can put you on the spot in an hour or two. A blind dig in the wrong place adds a day and a hole you have to backfill for nothing. The method pays for itself the first time it keeps you from opening the wrong ground.

What are the types of underground cable faults?

The fault type drives the method, so name it before you reach for a tool. A cable can fail open, with the conductor broken so current cannot pass. It can short, phase to phase, where two conductors touch. It can fault to ground, where a conductor or the shield finds earth, which is the common underground failure as moisture works into a damaged spot. And the insulation can fail as a sheath or jacket fault, where the outer covering is breached but the conductor is still intact.

The split that matters most for choosing a method is resistance. A low-resistance or bolted fault is a near-dead short, a clean low-ohm path the instruments see easily. A high-resistance fault still holds off most of the voltage and only breaks down when you push enough voltage across it to make it arc. The high-resistance fault is the one you will meet most often on aged underground cable, and it is exactly the one a basic low-voltage instrument walks right past.

Then there are the faults that will not sit still. A wet fault changes resistance as moisture moves. An intermittent fault clears when the cable cools or dries and comes back under load, so it reads fine on the bench and fails in service. These are the frustrating ones, and they usually force you to put real voltage on the cable to make the fault show itself long enough to find.

The prelocate-then-pinpoint workflow

Cable fault locating runs in a set order, and the order is what keeps the job efficient. Skip a step and you either waste time or miss the fault. The framework is the same on a utility feeder and a building feeder, even when the voltages and the gear are different.

First, diagnose. De-energize, ground, and prove the cable dead, then confirm the fault and grade its type with a megohmmeter and continuity checks. You want to know which conductors are involved and roughly what kind of fault you have before you pick a method. Second, prelocate. Get the distance to the fault from one end, using a TDR for opens and low-resistance faults, or a surge method, arc reflection, ICM, or decay, for the high-resistance faults a TDR cannot see. Third, pinpoint. Take the distance to the ground, trace the cable route, walk to the prelocated spot, and use the thumper with an acoustic and electromagnetic detector to mark the exact point. Fourth, trace and verify the depth so the dig opens cleanly onto the cable.

The single most common way crews lose time is collapsing prelocation and pinpointing into one step, dragging a thumper and a microphone over the whole route with no distance to aim at. Prelocation narrows the search from the length of the run to a few feet. Do it first, every time.

Step one: confirm the fault and its type

Before any locating gear comes out, prove what you are dealing with. De-energize the cable, lock it out, ground it, and discharge it. Then read it. A megohmmeter tells you whether the insulation is actually bad and, between conductors and to ground, which path is faulted. A continuity check tells you whether a conductor is open. The insulation resistance guide covers the megger work in full; here it is the first move, not the whole story.

What you are after is a classification, not just a pass or fail. Is the fault a clean low-resistance short the TDR will love, or a high-resistance fault that reads tens of megohms cold and only fails under voltage? Is one conductor open? Is the shield involved, pointing at a sheath fault? The answer steers you to the right prelocation method and saves you from running the wrong one and trusting a bad trace.

It also catches the trap of an intermittent fault that has gone quiet. A cable that dried out or cooled can read clean on the megger and still be the failure. If the history says it faults under load but the bench says it is fine, do not declare it healthy. You will likely need a surge or a controlled overvoltage to make the fault reappear long enough to locate it.

What is a TDR (time-domain reflectometer)?

A TDR is the radar of cable work. It sends a low-voltage pulse down the cable and listens for the reflection that comes back when the pulse hits a change in the cable's impedance. A break, a short, a splice, a water-soaked section, the far end, anything that is not smooth cable bounces part of the pulse back. The instrument times that round trip and turns it into a distance.

The reading comes off the shape of the trace. An open conductor reflects the pulse back with the same polarity and shows as an upward spike. A short reflects it inverted and shows as a downward dip. The clean, sharp reflection of an open or a hard short is what a TDR reads best, and on those faults it will put you within a few feet of the spot in one shot. You also learn to read the normal features, joints and the cable end, so you do not mistake a splice for the fault.

Distance comes from the timing and one setting you have to get right: the velocity of propagation. The TDR knows the round-trip time. It needs to know how fast the pulse travels in that particular cable to convert time into feet. Set the velocity to the cable and the distance is trustworthy. Set it wrong and every reading is off by a proportional amount, which is its own section below.

Why can't a TDR find every fault?

Because a low-voltage TDR pulse passes straight through a high-resistance fault and never sees it. The pulse is small, a few volts. A high-resistance fault is still mostly insulation. It only conducts, and only reflects, when you put enough voltage across it to break it down into an arc, and the TDR pulse is nowhere near that. So the trace comes back showing a clean cable all the way to the far end, on a cable you already know is faulted.

This is the trap that bites crews who own a TDR and nothing else. The high-resistance ground fault is the most common underground failure, and it is invisible to the one instrument they reach for. They run the TDR, see a clean trace, and conclude the cable is fine or the fault is past the end, when really the pulse just sailed past the fault.

The TDR earns its keep on opens, hard shorts, and finding joints and the cable length. For the high-resistance fault you have to make the fault visible, and that means adding voltage. Either you raise the voltage at the fault with a surge so the TDR can catch the momentary arc, which is arc reflection, or you switch to a surge method that works off the breakdown itself. The TDR does not become useless. It just needs help to see the fault that matters most.

Arc reflection for the high-resistance fault

Arc reflection is how you make a TDR find a fault it otherwise cannot. You connect a surge generator and the TDR to the cable through a filter that lets both work at once. The surge generator hits the cable with a high-voltage pulse that breaks the fault down into an arc. For the instant that arc exists, the fault is a near short, a sharp impedance change, and the TDR sees a clean reflection right at the fault.

The instrument captures the trace at the moment of the arc and freezes it. Better units overlay the healthy trace and the arc trace on the same screen, so the fault jumps out as the point where the two diverge. You read the distance the same way you would a hard short. It is prelocation, so the answer is a distance down the cable, not a spot on the ground yet.

The reason this is the workhorse prelocation method for high-resistance faults is that it locates with relatively low energy into the cable. The arc lasts a moment. You are not pounding the cable repeatedly to get the distance, the way you would if you pinpointed by thumping alone from the start. Arc reflection narrows the search first with little stress on the cable, and you save the heavier surging for the pinpoint, where you have already cut the search down to a few feet.

Surge prelocation: ICM, decay, and differential

Arc reflection is not the only way to prelocate a high-resistance fault, and on long cables or stubborn faults the other surge methods earn their place. They all work off the breakdown of the fault under high voltage, but they read the result differently than arc reflection does.

The impulse-current method, often shortened to ICM, fires a surge into the cable and records the current transient that travels back and forth between the surge source and the fault after the breakdown. The instrument times those reflections to get the distance. It tolerates very high-resistance and flashing faults that arc reflection can struggle to catch, and it reaches well on long cables. The decay, or voltage-decay, method is built for cables you can charge to a high voltage that then breaks down, common on long runs and on cable types that need higher voltage to flash; it reads the traveling wave that the breakdown launches. Differential methods compare traces to sharpen the read.

You do not need every one of these in your head. The practical point is that arc reflection handles most high-resistance faults at modest energy, and when the cable is long, the voltage required is high, or the fault is intermittent and flashing, the impulse-current and decay methods are the tools that still get a distance. Pick the method to the fault and the cable, and lean on the equipment manufacturer's guidance for which mode their instrument wants for which case.

What is a cable thumper?

A cable thumper is a surge generator that discharges a high-voltage capacitor bank into the faulted cable in repeated pulses. Each discharge drives enough voltage and energy across the fault to break it down and arc, and that arc makes a sharp mechanical thump at the fault point. The cable, the soil, and the duct carry that thump to the surface as a sound and a thud you can hear and, with a microphone, measure.

It is the classic pinpointing tool, and the name is exactly what it does. You set it to fire on an interval, a discharge every few seconds, and you walk the cable route over the prelocated area listening for where the thump is loudest and the ground pulse is strongest. Directly over the fault, the sound is loudest and arrives soonest after the discharge. The thumper does not give you a distance the way a TDR does. It makes noise at the fault so a person on the surface can stand on the spot.

Output runs into the tens of kilovolts, with the exact rating set by the cable class and the instrument; medium-voltage distribution work needs a far higher rating than a low-voltage building feeder. Treat the thumper as the high-voltage, high-energy hazard it is. It stores lethal energy in its capacitors even between discharges, and that is covered hard in the safety section. The thump is the friendly part. The machine making it is not.

Acoustic pinpointing with the ground microphone

Pinpointing is listening for the thump and timing it. You carry an acoustic ground microphone, a sensitive contact mic on a plate or spike you set on the surface, paired with a receiver that also reads the electromagnetic pulse the surge launches down the cable. The surge fires, the magnetic pulse arrives almost instantly, and the sound of the thump arrives a moment later after it travels up through the soil. The receiver shows you the delay.

That delay is the trick that beats sound alone. Soil carries sound oddly, and the loudest spot to your ear is not always the fault. The shortest time between the electromagnetic pulse and the acoustic thump is the fault, because that is where the sound has the least distance to travel up to the surface. You walk the route, take a reading every few feet, and converge on the point where the sound follows the magnetic pulse by the smallest delay. That point is the fault, usually within inches.

Depth and soil change the feel of it. A deep cable or a cable in a duct gives a duller, later thump, and you may have to turn up the gain and slow down. A direct-buried cable in firm soil cracks clean and sharp. Either way the method is the same: let the timing, not just the volume, tell you when you are standing on it. Mark the spot, confirm with one or two more discharges, then stop thumping. You are done surging the moment you have the spot.

Sheath and jacket faults: the earth-gradient method

A sheath or jacket fault is a breach in the cable's outer covering where the metallic shield or sheath finds earth, while the conductor itself is still good. These do not show as a conductor fault and they will not pinpoint well by thumping, because there is no conductor arc to make the thump. They get their own method, and it works off the voltage the fault sets up in the surrounding earth.

You isolate the cable, then apply a pulsed DC voltage between the sheath and a ground rod. Current flows out of the cable into the earth at the breach, and that current creates a voltage gradient in the soil that is centered on the fault. The fault becomes a point source of voltage in the ground, and you find it by reading that gradient on the surface.

The tool is an A-frame, two earth probes on a frame feeding a center-zero meter. You push the probes into the ground along the route and watch the needle, which points toward the fault. As you step along the cable the deflection grows, then directly over the fault it drops to zero and swings to the opposite polarity once you step past it. That zero-with-a-polarity-flip is the fault. Sheath fault location is also how you find the breach in a cable jacket that is letting water in and slowly creating the conductor fault you will be locating next year if you leave it.

Trace the route before you walk the fault

Know where the cable runs before you try to pinpoint a fault on it. A route tracer puts a tone on the cable, the conductor or the shield, with a transmitter, and you follow it on the surface with a receiver that reads the signal. You get the path of the cable and, on most tracers, an estimate of the burial depth. Do this before the pinpoint, because the acoustic search only makes sense if you are walking along the cable, not across the lawn at random.

The route also keeps the dig honest. A prelocated distance of, say, 240 ft down the cable is only useful if you can lay 240 ft of cable route out on the ground, and underground cable rarely runs in the straight line the print shows. It bends around obstacles, dips through pull boxes, and takes the slack the crew left when they pulled it. Trace the actual path and measure along it, and the prelocated distance lands where the fault really is.

Depth matters for the dig and for the acoustic read. A cable at 18 in sounds different than one at 4 ft, and knowing the depth tells you how deep to open and how hard the thump will be to hear. Trace first, mark the route in paint, then bring out the thumper and the microphone and walk the marks. It is the step crews skip when they are in a hurry, and skipping it is how a good prelocation still ends in a hole in the wrong place.

What is velocity of propagation, and why does it set your accuracy?

Velocity of propagation, or VoP, is the speed a pulse travels in a specific cable, expressed as a fraction of the speed of light or as a distance per microsecond. It is the conversion factor a TDR uses to turn the round-trip time it measures into a distance. The instrument times the reflection precisely. The distance is only as right as the VoP you told it to use.

VoP is a property of the cable, set mostly by the insulation material, so it varies by cable type. A common range for power cable falls well below the speed of light, and the right figure comes from the cable manufacturer's data or a known good length of the same cable. Get the number from the cable, not from a default left in the instrument from the last job on a different cable.

The error is proportional, which is what makes it sneaky. Set the VoP 10 percent high and every distance reads 10 percent long. On a short run that is a few feet and you find the fault anyway. On a long feeder that is tens of feet, and you dig in the wrong place with a trace that looked perfectly clean. The fix is cheap insurance: when you can, calibrate the VoP against the known length of the cable by reading the reflection from the far end and adjusting the setting until the measured length matches the real length. Then the fault distance is trustworthy.

Safety: high voltage, stored energy, and discharge

This is the part of the job that kills people, so it gets the bluntest section in the guide. Fault locating means putting high voltage and real energy onto a cable on purpose. The surge generator and the thumper store lethal energy in capacitors, and that energy stays there after the machine stops firing. A thumper that is switched off is not safe until it is discharged and grounded. Treat every cable and every instrument as charged until you have proven otherwise with your own hands.

The sequence is not optional. De-energize the cable and lock it out at the source. Verify it is dead. Ground both ends and any parallel cables that could be induced. When you connect the surge gear, the cable is energized by you, so the same discipline that applied to the utility voltage applies to your test voltage. When you finish a surge step, discharge the cable through a proper grounding stick and leave a ground applied before anyone touches the conductor. A charged cable holds voltage like a capacitor long after the source is gone, and the discharge is what makes it safe to handle. This is the same discharge discipline the megger guide stresses, and on a thumper the energy is far higher.

Qualified people only. High-voltage fault locating is not a task you hand to whoever is free. The operator has to understand the equipment, the cable, the grounding, and the failure modes, and has to keep everyone clear of the cable and the connections while it is energized. If you are not trained and qualified on the specific gear, you do not run it. The cable does not care that you were in a hurry.

Over-thumping damages the cable

The thumper that finds the fault can also create new ones, and crews who treat it as a blunt instrument pay for it later. Each high-energy discharge stresses the cable insulation everywhere, not just at the fault. On modern extruded cable, the repeated high-voltage surging drives electrical treeing in the insulation, tiny branching damage that grows into the next fault. The fault you pinpoint today by thumping for an hour can be paid for with a second fault a few months out.

So you minimize it. Thump at the lowest voltage that still produces a clear thump, and lean on the instrument's design to deliver energy at that lower voltage rather than cranking the voltage up. This is why crews condition or burn the fault first, often using a VLF set or the burn function, to bring its breakdown voltage down so the thumper can pinpoint at a lower, less damaging level while still delivering the energy that makes the thump. And you thump only as long as you need to nail the spot, then you stop.

The other half is doing as little pinpoint surging as possible by prelocating well. Every foot you cut off the search with a good TDR or arc-reflection distance is surging you do not have to do over the rest of the cable. Prelocate to a few feet, walk straight to that area, and pinpoint with a handful of discharges instead of marching a thumper down the whole run. Fewer thumps, lower voltage, shorter time. That is how you find the fault without making the next one.

Medium-voltage versus low-voltage work

The method is the same shape at both voltages, but the gear and the stakes are not. Medium-voltage underground distribution, the shielded cable in a utility loop or a campus duct bank, is the home ground of the surge generator and the thumper. The faults are usually high-resistance ground faults on aged cable, the runs are long, and the locating set is a real high-voltage instrument with the energy to break the fault down. This is where prelocate-then-pinpoint with arc reflection and acoustic pinpointing is the daily routine.

Low-voltage work, branch and feeder cable in a building or a parking lot, is often simpler. A TDR alone finds many of these because the faults are more often hard shorts or opens than the high-resistance ground faults that plague aged MV cable, and the runs are shorter. You may never bring out a thumper. But the same trap applies in miniature: a high-resistance fault on a low-voltage cable is just as invisible to a plain TDR, so do not assume LV means TDR-only.

Data centers and large campuses sit in between and bring their own pressure. The underground feeders and bus that tie utility, generator, and the building together are medium-voltage, the runs cross under occupied space you cannot dig freely, and the cost of the load being down is enormous. A fault there is exactly the case the full workflow is built for: confirm, prelocate to a few feet, trace the route under the slab or the yard, and pinpoint precisely, because opening the wrong spot under a live facility is not a mistake you get to make twice.

Acceptance testing versus fault locating

Do not confuse proving a cable with finding a fault in one. They use overlapping voltage gear and they get mixed up on the job, but they answer different questions. Acceptance and maintenance testing asks whether the cable is good enough to put in or keep in service. Fault locating asks where a cable that already failed went bad.

Acceptance work on shielded power cable is increasingly done with VLF, very low frequency, withstand testing, often paired with diagnostics like tan delta or partial discharge, with the framework in the IEEE 400 series for field testing of shielded power cable systems. The older DC hipot is still around but has fallen out of favor on aged extruded cable because high DC voltage can itself damage that insulation. The point of all of it is to prove the cable, or to find that it cannot hold voltage, before it carries load.

Where the two worlds touch is conditioning. The same VLF or hipot set that proves a cable can be used to break down and burn a fault, lowering its breakdown voltage so the thumper can pinpoint it at a gentler level. That is the useful overlap. What you should not do is treat a fault-locating thump as if it proved the cable, or treat an acceptance test as if it located anything. Find the fault, repair the splice, then prove the repair with the acceptance test. Different jobs, run in that order.

The fault-locating toolkit

A full underground fault-locating kit is a set of instruments that each do one part of the workflow, not a single magic box. You diagnose with a megohmmeter, prelocate with a TDR and a surge generator, pinpoint with the thumper feeding an acoustic and electromagnetic detector, find the route and depth with a tracer, and chase sheath faults with an A-frame. Smaller combination units fold several of these into one case for distribution work, but the functions are still the same.

Match the gear to the cable. The surge generator's voltage and energy rating has to suit the cable class and length, which is why an MV distribution kit is a different animal than a low-voltage building kit. The TDR has to handle the cable type and have a velocity setting you can adjust. The route tracer and the A-frame are cheap next to the surge gear and they are what keep a good prelocation from ending in the wrong hole. Buy or rent the full chain, because a kit missing the route tracer or the pinpoint receiver is a kit that gets you a distance and then strands you on the surface.

What to document

A located fault is also a record waiting to be written, and the cable that faulted once at a spot tends to be worth watching there. Capture the prelocated distance, the method you used to get it, the velocity setting if it was a TDR, the pinpointed location on the ground, the depth, the fault type, and the repair you made. Tie it to a measured route, not just a print, so the next crew can find the splice.

Keep the locate with the cable, not in someone's truck. A field tool like FieldOS holds the cable record, the fault location, the photos of the dug-up fault and the finished splice, and the route trace, so the repair history travels with the asset instead of living in a notebook that disappears with the tech. When the same run faults again two years on, the record of where it failed last time and what the cable looked like there is the fastest start on the next locate.

Common mistakes

• Digging blind with no prelocation, opening a trench instead of one hole.
• Running only a TDR and calling a high-resistance cable clean because the pulse passed the fault.
• Leaving the velocity of propagation on a default, so every distance reads long or short.
• Skipping the route trace, so a good distance lands in the wrong spot on a bent run.
• Pinpointing by volume alone instead of the shortest surge-to-sound delay.
• Over-thumping at high voltage for too long and seeding the next fault with treeing.
• Failing to discharge and ground the cable before touching it, with lethal stored energy still on it.
• Putting an unqualified person on the high-voltage surge gear because the crew was short.

Field checklist

Standards and references

Fault locating itself is a manufacturer-and-method discipline more than a code one, so cite carefully. The IEEE 400 series gives the framework for field testing of shielded power cable systems, with IEEE 400.2 covering VLF testing, which is where the acceptance and conditioning side of this work lives. NETA acceptance and maintenance testing specifications set the framework for proving cable in service. The locating methods, arc reflection, ICM, decay, the thumper output and its energy, and the velocity values, come from the equipment manufacturer's instructions for the specific instrument and the specific cable.

Installation and grounding sit under the NEC, NFPA 70, and the safety of the work sits under the electrical safety practices for working on and around energized and high-voltage equipment. The exact thresholds, the test voltages, and the procedures vary by cable class, cable type, and the instrument, so confirm them against the manufacturer's documentation and the standard the project and the utility have actually adopted. Edition and local practice control.

Where this guide gives a number, treat it as a typical figure to confirm, not a spec. Thumper output, VoP, and the voltage to break down a given fault all depend on the cable and the gear in your hands. The three things that do not bend: prelocate before you pinpoint, a plain TDR will not find a high-resistance fault, and the cable and the surge gear carry lethal energy until you discharge and ground them.

Units, terms, and synonyms

Cable fault locating carries its own vocabulary, and the same idea shows up under different names across instrument manuals and field talk. Distances run in feet or meters, surge output in kilovolts and joules, and pulse timing in microseconds.

Prelocation is also called rough locating or distance-to-fault. Pinpointing is also called precise location. The thumper is a surge generator or impulse generator. The TDR is a time-domain reflectometer or echometer. Velocity of propagation is also written VoP, velocity factor, or propagation velocity, and is given as a fraction of the speed of light or as a distance per microsecond. Arc reflection, impulse current (ICM), and voltage decay are the surge prelocation methods for high-resistance faults. The earth-gradient or step-voltage method, read with an A-frame, locates sheath and jacket faults.

Prelocation

Finding the distance to the fault from one end, before pinpointing

Pinpointing

Finding the exact spot on the ground, usually by thumping and listening

TDR

Time-domain reflectometer, sends a low-voltage pulse and times the reflection

VoP

Velocity of propagation, the pulse speed in the cable that converts time to distance

Arc reflection

TDR plus a surge, so the TDR reads the momentary arc at a high-resistance fault

Thumper

Surge generator that discharges into the fault to make an audible thump

Earth gradient

Voltage in the soil around a sheath fault, read with an A-frame to locate it

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is insulation resistance testing used for?
• What does an infrared inspection find?
• What is cable tray?
• How deep does a duct bank go?
• What is a three-way match in accounts payable?