Barney's Blog no 10 - Cable Sheath Fault Location

Methods of finding cable sheath faults, and damage to the outer plastic covering of a cable

The outer plastic sheath covering of both telecom and power cables is the final protection against moisture ingress.


In copper telecom cables, used for local network, moisture entering via damage or poorly sealed joints leads to degraded communications, and eventually the customers complain and investigation starts.
Power cables, depending on operating voltage, may show intermittent faults, unpredictable and frustrating and the most difficult to find, and eventually the cable fails, trips the fuse or circuit breaker and becomes permanently faulty, generally an easier fault to find.

Sheath faults can occur during initial construction and installation, and testing, investigation and elimination of these should be part of the installation procedure before handover and commissioning.
After handover, sheath faults in directly buried cables can be caused by general traffic pressure and sharp stones in the sand used as bedding material, and inevitably as a result of third party damage caused by excavations during street works.

In any case, if the problem is not found and repaired, eventually the cable will have reliability problems and or loss of service.

Cables without plastic outer covering such as "belted" armoured cables don't have this problem, nor can they be tested by methods outlined here.

Sheath faults in ducted cables, both telecom and power, used to be difficult or impossible to locate, but there are now interesting alternative methods to solve this problem.

Communication cables – ducted cables

With increasing use of Fibre Optic cables, many of the problems caused by moisture ingress are over. The copper network now uses cables that are jelly filled, so moisture penetration along the cable is much less of a problem. In this case, the cable system is often ducted, so there is the option of cable replacement available.
Moisture in joints will always be a problem, and in places where the classic paper insulated cables with lead sheath and plastic coating are used, water ingress is a serious problem.

Generally these cables are installed in ducts and originally were used as high capacity cables between telephone exchanges. Now a fault of this type may be difficult and uneconomic to repair, so in many instances installation of new Fibre Optic system is the solution.

The traditional paper lead cables are generally pressurised with "dry air". The idea is that the pressurised, moisture free air, will dry any moisture, and also where the cable is damaged and the dry air is escaping can be detected.

The dried air is supplied by a special compressor and/or pressure bottles installed in the telephone exchange. Air Pressure sensors at the exchange measure the air pressure and flow. In joint bays/pits there are always pressure test points, generally fitted with a pressure sensor, and these pressure sensors outputs are connected using the phone network to the telephone exchange. Pressure loss in a section of cable suggests some damage and starts an investigation.
.
Meanwhile there are solutions to locating the point of air pressure loss; the pressurisation system indicated the problem section, and there are pressure test points nearby.

So it is simple enough to use the Swedish Sensistor system that uses a non flammable tracer gas mixture of 95% nitrogen and 5% hydrogen, and this is injected into the cable via pressure test point. The tracer gas passes along the cable and exits at the damage point, and Sensistor developed a clever way to pinpoint the position of the leak from the cable in the duct; their trace gas sensor is installed at the end of a reel of fibreglass pushrod. The rod is unreeled and pushed up the duct and detects the tracer gas.

If the gas has filled the duct, then it will be difficult to determine the damage point. Sensistor developed a simple solution using a small fan to drive air into the duct and push the trace gas in one direction. This blows the excess gas away and leaves a defined position between gas and no gas. The operator then pulls out the rod and lays it on the surface to see where the fault is.
The one disadvantage of the method is that the tracer gas sensor must not be used in water as it will cause the sensor to fail, so the duct system needs to be pumped clear of water and dry.

Another option using the same general process is the GOK system from Seba KMT.
The GOK rod is manufactured in various lengths, wound onto a reel, and is usually used for water leak detection from inside water service pipes. Its sensor head uses a hydrophone to detect leak noise and a transmitting sonde to determine position.
At the reel end there is a connector to attach the Seba Ground microphone amplifier HL500 or HL5000 (discussed in previous blog) and these provide amplification and display for the leak noise.
By pushing the rod along the duct until the leak noise, bubbles if submerged, and hiss of escaping gas is heard, then locating the position with the sonde head allows precise location and pinpointing on the surface of the sheath fault/gas leak.

Because the hydrophone/microphone only detects noise, not specific tracer gas, it is possible to use any dry air product from gas cylinder or compressor as appropriate for the cable.

Communication cables – direct buried

Although recent installations use ducted cables, for very obvious reasons to simplify and speed up any necessary repair, in older suburban areas much of the local network from the nearest cabinet to the property is often directly buried in the ground. If the telecom joints themselves are also direct buried, any fault locating and pinpointing is made difficult by multiple connections and difficulty interpreting TDR traces.

Water ingress at cable joints, or damage to the cable since installation, can usually be pinpointed using sheath fault location techniques, because where moisture enters the cable, there is a conductive path to the earth for the fault location signal.

The operator can connect the locator transmitter such as the Vloc Pro 5W or 10W transmitter system that has a fault find output, across the affected sheath or pair of the cable and to ground. The fault find signal uses both a medium frequency signal (8 kHz) for locating the cable route, and a very low frequency signal (4.8 Hz) that conducts into the ground at the fault and returns to the transmitter ground stake.
The VlocPro2 receiver is used to locate the 8kHz signal from the cable and an accessory A frame is attached to provide fault location pinpointing using the 4.8 Hz signal.
On the Vlocpro2 display the operator sees green and red arrows corresponding to the colours around the ground spikes of the A frame. Green arrow means move forward along the route, and red arrow means that you passed the fault; numbers on the display indicate the (decibel millivolts dBuV) voltage across the 2 spikes, and these are highest near to the transmitter ground stake and also when close to the fault. So the user needs to judge the distance from the ground stake that the voltage reduces then make a test at the same interval as the distance from the ground stake regularly along the length of the cable. If the A frame's spikes are equidistant from the fault, the numbers are low, but moving backward or forward cause them to increase, then the fault is directly below the centre of the A frame. The fault can then be excavated and repaired.

Power cables

Sheath faults on power cables are now causing interest to Electricity distribution companies, now that they have most of their cables installed by contractors and sheath fault test is required before handover.

Historically, armoured or belted cables could not be checked for sheath faults, because the length of the cable's armour /belt was generally in contact with the ground. This contributed to the general grounding of the electric system because of equi-potential along the cable section.

When plastic cable sheaths were introduced, sheath fault testing was carried out to ensure its integrity and check for external damage, mainly on HV primary cables because water/dampness ingress could lead to insulation failure.

With XLPE insulation, sheath testing became more important, because dampness and moisture in the XLPE could lead to water trees and cable failure.

With low voltage cables for local distribution, there is no possibility to carry out sheath fault tests, because of multiple grounding points in every building on the network. The only possibility to test is after cable installation and before final connections.

Aluminium cable with aluminium neutral/earth screen (consac) is often a problem if the sheath faults are not found because moisture ingress leads to corrosion of the aluminium and cause neutral faults as the screen became open circuit, a serious problem and generally difficult faults to pinpoint unless all grounds are disconnected.

Sheath fault testing procedures – standardised testing??

Measuring the electrical resistance between sheath and ground indicates the condition of the cable sheath. If resistance is very low, then there is a severe fault that can be easily located using a standard pipe and cable locator.
If the resistance is high and is outside the range of the standard multimeter and an insulation tester is used, then equipment with high voltage output must be used to test and locate the fault.

In Britain there seems not to be a national standard for sheath fault insulation, and sometimes sheaths were never tested, and also with the splitting up of the power companies, now many with foreign/multinational owners, who have different standards and use different cables. In Germany the VDE have issued standards that are adopted by most of the rest of Europe, the table below indicates the general requirements.

Testing should be dependent on the sheath material used, and obviously too much voltage and current used during testing could cause more faults than already exist.

The testing of cable sheaths is regulated in different norms, IEEE xxxx, IEC 60229, VDE 0276 part 620 and part 632, also their Harmonised document HD 60602 and 60632
The VDE specifies the following details:

Voltage tests at the cable sheath
Test Voltage Test Level Test Duration
DC at cables acc. to VDE 0276 Part 620
(Extruded cables from 6 to 36 kV)

PVC–Sheath 3 kV PE – Sheath 5kV

Not specified
DC at cables acc. to VDE 0276 Part 632
(Extruded cables above 36 to 150 kV)
5 kV 1 min

User definitions

Several local power utilities have already defined own regulations for sheath testing, where the main idea was to define measuring criteria.

Connected with this test is a "quality test", which has to be proven by external contractors.
This factory norm operates outside the normal regulations and defines more details which are shown in the following table below. The test duration for this test is defined with 10 minutes.

Cable length Leakage current Leakage current
Metres Feet  PVC  PE
 50  164  0,04 mA  0,001 mA
100  328   0,08 mA  0,002 mA
250  820   0,2 mA  0,005 mA
50c  1640   0,4 mA  0,01 mA
750  2460  0,6 mA   0,015 mA
1000  3280  0,8 mA   0,02 mA
2000  6560  1,6 mA  0,04 mA 
5000  16400  4,0 mA   0,1 mA

 

 
  
 164  
100 328  
250 820  
50C 1640 0,4 mA 
750 2460 0,6 mA 
1000 3280 0,8 mA 
2000 6560 1,6 mA 
 

Whatever standard is used, there needs to be a simple enough field test method once the cable is installed to record the test and save it to memory as reference. So the usual high voltage insulation tester requires a tester and a witness to both sign the test results. Mis-transfer of results to paper record could lead to false results.
Also to achieve a good test, the testing instrument needs to be capable of producing both sufficient voltage and charge for the length of cable being tested, and also have means to record the results, fail or pass, into a memory store. So providing the cable sheath holds the specified voltage for the specified time, it shows that the cable is well installed and will give a long reliable life, and there is a time/date stamped record of the test result.

If the cable section passes the appropriate tests, with the new Seba KMT MFM10 sheath fault testing and fault location equipment, the results can be saved for future reference on a protocolling USB memory stick with reference to date and time from the internal real-time clock, and later uploaded to PC.
Testing is carried out by connecting the MFM 10 to sheath and ground. Negative DC voltage is applied to cable sheath, with maximum voltage and current regulated according to the sheath material, and customers' test parameters, and the time duration.
An easy to understand graphic display shows the history of the voltage/time as a rising ramp, and current, indicating a failure, should remain very small, allowing for the cable section length. Bargraphs show voltage and current and values.

If the cable sheath is perfect, the operator sees the voltage ramp up to the specified voltage and the seconds tick by until the test ends automatically and the cable is discharged. At the end of the test the operator can choose to save the result to the USB memory stick, and this will be saved with date, time, and a file number in accordance with the number of tests carried out on that day.

If the cable fails during the test, then the voltage ramping up will show a drop and current will rise. The MFM10 can be set up to recognise this and stop the test automatically at this point. As before the result can be saved to the USB memory stick.

With the vintage Bicotest popeye/magpie equipment the operator needed to increase the voltage by rotating the voltage transformer control and at the same time observe an analogue needle on the voltage dial to see the sheath breakdown point. Further increase of voltage may cause the fault to disappear as moisture is burned off.

Sheath fault location.

When the sheath test fails, the fault needs to be located and pinpointed so that it can be exposed and repaired.

Traditional sheath fault locating methods use a DC pulse injected into the cable sheath. The DC leaves the cable sheath at the sheath faults and returns through the ground to the grounding of the DC pulse transmitter. The operator of the fault locator, equipped with ground probes, has to walk the length of the cable to find the fault by detecting the flow of current returning from the fault through the ground.

The difference in DC voltage across the two ground probes is measured, and is shown as a directional kick on the instrument. As the operator moves towards the fault, the difference in voltage becomes greater, and is shown by a higher indication on the instrument. Once past the fault the current flow direction is reversed and the operator sees this easily. Finally if both ground probes are equidistant from the sheath fault there is no difference in voltage, and no reading is shown, and the fault has been located.
While this is easy enough on say 200M of cable, walking several kilometres is time consuming, and also means that the cable sheath is being charged and discharged through the fault for some time until the fault is located, and this may change the fault condition, and stress the cable sheath more than necessary.

Using a simpler low frequency AC transmitter and A frame locator generally uses less voltage to provide the same result, but possibly will run out of power/signal for long distance and multiple faults. It is quick and easy for looped in street lighting cable systems when there is only 100-200 between lighting columns and the cables can easily be disconnected from ground for the sheath test faultfinding method. The Vloc Pro locator system with a 10 watt transmitter and Accessory A frame works well for these faults. If the fault is within a cable joint and there is no leakage to ground then the A frame method is not the solution.
Older lighting systems using a Teed off connection system are much more difficult as there are multiple earths that need to be disconnected before testing and fault location.

Using DC systems is preferable for long cable lengths and with DC it is generally possible to use various bridge methods to prelocate faults; but that is more difficult for accurate results if there is more than one fault.

Sheath fault prelocation

The usual cable fault prelocation methods, such as a TDR don't work on cable sheaths as there is only one effective conductor – TDR's need two – and TDR's don't reveal high resistance faults.

So bridge methods are used. Bridges are frequently used to prelocate earth faults in the metallic cores of telecoms, signal, control, pilot cables and also umbellical cables used to operate undersea ROV's. Seba KMT produce automatic bridge KMK7, and the KMK8, automatic bridge combined with TDR for exactly this purpose.
Using Bridge method for power cables means that for best results the voltage used to measure should be the same or just slightly greater than the breakdown voltage of the sheath fault, and this ensures that the worst faults (lowest resistance) are easily found, and the higher resistance faults that do not break down at lower voltage do not affect the measurements. It also means that the equipment used for power cables needs a high voltage source which the instruments for telecom cables don't need.

All bridge methods require a "help wire" so that the resistance to ground can be compared from the far end of the cable as well as the near end – so it's more useful to use the core of the cable being tested as the "help wire". Simplest bridges provide distance to fault as a percentage of the length – so the technician has to calculate the actual position and distance to fault.
To achieve any results on longer power cables, sufficient current must be passed along the cable and into the fault, but of course greater current causes local heating at the fault that may skew the results. The connections need to be made so that they do not change the results - poor connections mean higher resistance and incorrect prelocate distances.

The Seba KMT MFM 10 equipment that I have been using for 18 months, has an extremely useful automatic voltage drop bridge that prelocates the distance to the sheath fault. Seba chose to include this bridge method because it is generally more user friendly and gets good accurate results without the necessity to have perfect connections and uniform "help wires". The MFM10 also uses a Bi-polar measuring system, so the faults are measured from both ends of the sheath using positive and negative polarity for the tests. The automatic measuring system only applies the voltage for long enough to get the measurement and doing this reduces local heating that can change resistances and affect the results.

The operator needs to make the connections, set the voltage and current limits and dial in the cable length, then start the measurement. When measurement in finished, the MFM10 displays the distance to the fault in Metres and the percentage of the length dialled in.
The skilled user, familiar with the MFM10 equipment can observe the sheath test and see the voltage starting to drop and the current increasing – so this sheath breakdown voltage is used for the voltage setting in the prelocation test, as lower resistance faults are seen immediately. If the operator then makes a prelocate test at a higher voltage and that gives different distance to sheath fault, this suggests that there is another fault breaking down at higher voltage.

In this case it is generally easier to pinpoint and expose the lower resistance fault, clean and dry the cable, and retest to find the next fault. If the sheath test now shows that the fault is clear, cable is repaired, retested and the task is completed.

In any case the prelocation speeds up fault location and greatly reduces the time that the cable sheath is subjected to high voltage. Prelocation and indeed all the stages of sheath fault location require that the cable is buried directly in the ground, so that the current applied to find the sheath fault has a circuit/return path via the ground – so it would suggest that cables in ducts generally cannot be sheath tested.

Cables in ducts

Power cables installed in ducts generally will not have measurable sheath faults as there is no leakage to ground if the duct is dry. The only part of the installation where sheath faults could be detected is where the cables have their joint positions and generally the joints will be directly buried at this position. Sheath testing will show if there is a problem and prelocation will identify the suspect joint bay.
Pinpointing using the ground probes and earth fault locator will verify the joint bay position, so that the joints can be excavated. Once the area is cleaned and dry the cable is tested again, and if it passes test, then the sheath fault is in the area cleaned off. If the fault is still present, pinpointing it may lead the operator to the duct entrance, and if the duct is sufficiently damp to be conductive, then the sheath fault is somewhere up the duct. It is also worthwhile to check at the next nearest joint bay to confirm direction to the sheath fault. But if the duct is dry and non-conductive then pinpointing cannot give a result at that point.

Ducted graphite coated cables

High voltage (132kV) network in some electricity regions use graphite coated 3 single core cables in separate ducts with a parallel separately ducted ground wire. The graphite coating is used to create an earth path in the event of a cable fault, and if there is a sheath fault then the shield is connected to the earth via the graphite coating.
At the cable joints, generally between 500M and 600M distance apart, there are usually chambers with removable links for cross-bonding the cable shields. The cable joints are close to these, and the shields of the cables are connected or disconnected within the link box chamber that can also be used for sheath testing, assuming the next set of links are also removed.
If there is a current path to ground it will be at the point where the cable leaves the duct and is formed into the cable joint. So testing will show if a fault exists, but conventional pinpointing will show results at points where the cable exits the duct, and provided no information about the actual sheath fault location.
Prelocation is possible with graphited cables - in practice the graphite is equivalent to the ground, and the point where the fault crosses the sheath to the graphite can be prelocated using the MFM10 equipment. But as the cable cores are not accessible at the joints, best solution is to use the shields of the other two phases, and in practice this is shown to provide good prelocate results even if these sheaths are also faulty. The connections are conveniently available in the same and next link boxes.
One of the advantages of the voltage drop bridge prelocate method is the help wires do not have to be uniform – but it is always worth double checking the results by connecting from the other end of the cable section under test and comparing the distance results. If they are not the same position, then it is always possible that there is more than one fault, always possible if the cable
has been hauled around a bend in a duct.

Pinpointing using ground probes in not possible as there is no connection to the earth apart from the ends of the ducts.

Recent developments using duct section couplings made of conductive material offer a means to pinpoint the fault to a certain duct section – but obviously this can only be used on new projects and is not practical for a retrofit.

Sheath fault pinpointing using pulsed DC

The pool of potential method as used by the traditional Bicotest popeye device uses DC pulses transmitted into the cable shield that leak to ground through the sheath faults. Faults are detected by the ground fault detector and this had to be synchronised with the transmitted pulse by having the ground probes in the earth relatively close to the transmitter, where the returning pulse was near maximum and the direction of the potential is back towards the grounding point for the transmitter. Away from the transmitter, there is less voltage difference across the ground probes and approaching the fault the voltage again increases to a maximum when the leading ground probe is closest to the fault.

For many years Seba KMT manufactured a variety of MMG and MFO DC sources, as testers and pulse generators. Generally a pulse /space timing of 1.5 seconds on and 3.5 seconds off was the best to work with analogue instrumentation as the operator could easily determine the directional swing of the needle. The slow sequence also allowed the operator to move on along the cable between pulses.
The ESG-80-2 ground fault locator, used to detect the return current could be used as a pure DC sensing device, or with an amplifier with stepped gain control to improve the sensitivity. For adjustment this device has a control to centre the analogue display so that it is easier to see the direction of the needle in response to the DC pulse returning from the sheath fault back through the ground. Increasing the distance between the ground probes, also improves sensitivity, but may mean that at least two operator are needed for speedy work. Once a needle response is observed, the operator can bring the ground probes closer together, readjust the gain, and pinpoint the sheath fault.
Because the ESG 80-2 is measuring the voltage difference between its ground probes, there is a higher reading when one or other of the probes are closest to the fault, and a minimum when either of the probes are equidistant from the fault. Fault pinpointing is best achieved by keeping the ground probes at the same distance apart whilst moving along the cable's route and testing every few paces. When the indicator needle changes swing direction the fault has been passed, and reversing route whilst keeping the ground probes in the same orientation will show the minimum, and at that point the probes are equidistant from the fault. During testing adjustments need to be made to centre the needle and also to the gain.

It is always worth checking the correct position by placing the probes at 90 degrees, across the cable route, then test the swing as operator crossed the cable route. A second minimum should be found when the centre of the cable route, unless the cable route has deviated, or a cable joint is at a different line from the general cable route. In any case pinpointing along the cable and also across the cable confirms the sheath fault position exactly before excavations start.

Seba KMT's new earth fault location ESG-NT unit is more sensitive than the ESG 80-2 and with its "Easygo" rotary control it is really straightforward to use. Once set up no gain adjustment is required and the bargraph on the LCD colour display shows direction of the fault and voltage between ground probes. To make it really intuitive, the red bargraph colour means go in the direction of the red probe lead, or black bargraph means go in the direction of the black probe lead. As the device needs no gain adjustment, it needs 2 or 3 pulses to adjust, but a quicker DC pulse rate can be used effectively as the operator can see the change in the bargraph more easily that the small swing of an analogue instrument's needle. The set-up also includes a history reading method, so the recent results are displayed like results on a rolling drum. Also various filters can be used if the test area has high interference.

In all cases, used together with the MFM10 system, that has different pulse/space timings built in, the user can choose their preferred rate.

Difficult sheath faults

Sometimes the cable has been tested and there is confirmation that the sheath is faulty to ground. Prelocation provides a repeatable result from both ends of the system – but for some reason there is no response at or around the fault position with the ground fault locator ESG 80-2.

Various possibilities cause this; there is no DC pulse leaving the cable into the ground at the prelocated point – but it must be leaving at this point so there is a possibility that the DC is passing on to an adjacent metal pipe or another cable.
First confirm that the MFM10 connections are correct and check that there is a detectable return signal near to the grounding of the MFM10.
If the ESG 80-2 or ESG-NT ground probes are being used on a hard surface, wet the ground to ensure good ground contact, and if there is still no response, find out what other utilities are in the same area, because there is possibility of the DC pulse being conducted away through one of them, and of course pipes and power cables are grounded together in every building so it is advisable to stop DC high voltage output as soon as possible.

Clues to this type of fault are low resistance to ground when the sheath is tested.

What to do?
The MFM10 system also has very low frequency 4.8Hz AC output in addition to the range of DC pulse timings, for use with Vivax and Metrotech sheath fault location systems. Frequently AC will provide successful results because DC needs the shortest, lowest resistance path and AC has more leakage to earth by capacitance rather than conduction. So with the MFM10 using AC output and a Vivax/Metrotech locator and sheath fault locating A frame, pinpointing results may be possible that are not possible with DC pulsed systems. The locator display automatically switches to sheath fault locating mode with green or red forwards/backwards direction arrows and a dBuV number showing the voltage across the A frame spikes. When the standard 10watt transmitter is used, a cable locating frequency is also applied to the cable.

There is the alternative method if the sheath fault has very low resistance and that means using a normal Vivax/metrotech locating transmitter connected to the faulty sheath of the cable, and locate where the signal decreases because it leaves the cable where the sheath fault is located.

If neither of these methods succeed, there is always the powerful Seba 50 watt transmitter, the HPL50/FLG50 and using its low frequency and signal select output with locator and A frame, it is another method using a 491Hz AC frequency – higher frequency, hence more leakage to the ground.

There is an even higher AC frequency that allows the use of a special Seba A frame DEB 10 used with the FLE10 receiver that uses capacitive plates as well as ground contact spikes. The plates allow sheath faults to be found over hard surfaces and don't need to have ground contact; but a high power transmitter is needed, and also higher frequency of 8.44kHz or 9.82kHz is used. Generally the system is best on shorter lengths of cable due to capacitive losses restricting the transmitter range.

Another problem area is where cables are installed together with a nearby bare copper ground wire. Frequently used on new cable installations this method is used to improve the earthing, or to ensure that different parts of the installation are bonded together. Sheath testing shows if there are sheath faults, and using MFM10 to prelocate will be useful in many cases, but pinpointing is unlikely to produce results as any DC pulse leaving the sheath will travel the lowest resistance path - that is the short distance between sheath and ground wire - and not be detectable on the surface. Best methods are to detach the ground wire at each end and try again. If there is still no success, it will be necessary to cut the ground cable and remove it far enough from its next part so that the resistance to ground from the cable to the ground wire is higher than the fault to ground.
Installations of this type have a higher chance of sheath faults; and the close proximity of the ground wire makes sheath faults difficult to pinpoint. So the solution is to bury the ground wire about a metre away from the cable route, or further away if the ground is likely to be damp and conductive.

Pinpointing the sheath fault when cable is exposed

Generally when the sheath fault has been located, pinpointed and exposed the problem can be seen immediately but sometimes it's not visible. In this case the ESG 80-2 and ESG-NT can be used to carefully pinpoint the fault. The trick is to retain a coating of earth/sand above the cable, and using the MFM10 on minimum voltage (100V) repeat the pinpointing technique along the line of the cable, but with the ground probes about 25/30 cm apart. This usually gives the classic maximum-minimum-maximum response. Mark the centre of the minimum response and closely examine the cable in this position.
An alternative method is to have one of the spikes driven into the side of the trench, and probe gently along the cable looking for maximum response. Using this procedure, direction is not important.

If the cable system is 3 single core cables installed as triplex, it's always possible that the sheath fault is either underneath or between the 3 cables. If the fault is at a joint, there is always the possibility of slight leakage out of part of the joint assembly

If the fault is not visible to the naked eye another trick is to clean and dry the cable and retest to see if the fault has disappeared. If the cable test is now perfect, the fault must be in the cleaned area – sometimes only a pinhole causes the problem. If the fault returns when the cable is reburied, then the fault is clearly in this location. Sometimes the condition of close by faults change as the cable is moved, so rechecking with the ESG 80-2/ESG-NT ground probes is also useful.

And if all else fails?

Applying higher voltage to the cable sheath may result in creating more faults, but 10kV is available from the MFM 10 if that is necessary – but it is not advisable. Burning is also available and that will reduce the sheath fault resistance and make it easier to pinpoint, but at the same time it may affect other parts of the cable sheath or reduce its long term life.
So a "scorched earth plan" or "insure and burn" are not advisable, and in any case may also dry out the fault so that it disappears. Best to wait a few days until the ground around the cable is damp enough and probably the sheath fault will have returned.

Conclusions

Sheath fault testing and fault location is an essential part of cable commissioning and regular testing, to ensure long fault free cable life.
Sheath testing needs to be recorded and saved.
Prelocation saves time and stress to the cable sheath.
Sheath fault pinpointing can use DC pulse or AC, but DC is preferable for long distances
Good equipment makes sheath fault location quick and easy
The operator of the good equipment needs to be trained, proficient and able to deal with difficult cases.