Faultfinding AG

8/8/2019 Faultfinding AG

1/44

FAULT FINDINGSOLUTIONS

WWW.MEGGER.COM


2/44

Table of Contents

Fault Finding Solutions

Section Page

I Cable Characteristics

Good Cable Insulation . . . . . . . . . . . . . . . . . . . .2

When Cable Insulation is Bad . . . . . . . . . . . . . .2

Why a cable becomes bad . . . . . . . . . . . . . . .3

Cable Faults Described . . . . . . . . . . . . . . . . . . . .3

II Fault Locating Procedures

Locate Faults in Buried Primary Cable . . . . . . . .4Test the cable . . . . . . . . . . . . . . . . . . . . . . . . .4

Fault resistance and loop test . . . . . . . . . . . .4

TDR tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

DC hipot test . . . . . . . . . . . . . . . . . . . . . . . . .5

Analyze the Data . . . . . . . . . . . . . . . . . . . . . . . .5Fault resistance and loop test . . . . . . . . . . . .5

TDR tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6

DC Hipot test . . . . . . . . . . . . . . . . . . . . . . . . .6Cable Route . . . . . . . . . . . . . . . . . . . . . . . . . .6

Localize - prelocate the fault . . . . . . . . . . . . . . .6

Locate - pinpoint the fault . . . . . . . . . . . . . . . .6

Locate Faults in Above GroundPrimary Cable . . . . . . . . . . . . . . . . . . . . . . . . . . .6

III Cable Route Tracers/Locators

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

Selecting a Locator . . . . . . . . . . . . . . . . . . . . . .8

Hookups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9

Using the Receiver . . . . . . . . . . . . . . . . . . . . . .10

IV How to See Underground Cable Problems

Methods of Operation . . . . . . . . . . . . . . . . . . .12Time domain reflectometry . . . . . . . . . . . . .12

Differential TDR/radar . . . . . . . . . . . . . . . . .13

Descriptions and Applications . . . . . . . . . . . . .13Low-voltage TDR/cable radar . . . . . . . . . . . .13

Faults that a low-voltageTDR will display . . . . . . . . . . . . . . . . . . . . . .13

Landmarks that a low-voltageTDR will display . . . . . . . . . . . . . . . . . . . . . .13

Controls and Inputs to the TDR . . . . . . . . . . . .14Velocity of propagation . . . . . . . . . . . . . . . .14

Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16Gain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Cursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Zoom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Pulse width . . . . . . . . . . . . . . . . . . . . . . . . . .16

Distance Measurements . . . . . . . . . . . . . . . . . .17Three-stake method . . . . . . . . . . . . . . . . . . .17

Section Page

V Surge Generators, Filters and Couplers

Surge Generators . . . . . . . . . . . . . . . . . . . . . . .19Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19

Capacitance . . . . . . . . . . . . . . . . . . . . . . . . .20Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20

Basic Surge Generator Operation . . . . . . . . . .21Proof/Burn . . . . . . . . . . . . . . . . . . . . . . . . . .21

Surge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

Ground . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21

Single Point Grounding . . . . . . . . . . . . . . . . . .22

Arc Reflection Filters and Couplers . . . . . . . . .22

VI Localizing Methods

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24Sectionalizing . . . . . . . . . . . . . . . . . . . . . . . .24

Resistance ratio . . . . . . . . . . . . . . . . . . . . . .24Electromagnetic surge detection . . . . . . . . .25

Single phase, coaxial power cablewith neutral bridges over splices . . . . . . .25

Single phase PILC cable with bondedgrounds in conduit . . . . . . . . . . . . . . . . . .26

Three-phase PILC . . . . . . . . . . . . . . . . . . . .26

DART Analyzer/High-Voltage Radar . . . . . . . .27Arc reflection . . . . . . . . . . . . . . . . . . . . . . . .27

Differential arc reflection . . . . . . . . . . . . . . .28

Surge pulse reflection . . . . . . . . . . . . . . . . .28

Voltage decay reflection . . . . . . . . . . . . . . .29

VII Locating or Pinpointing MethodsAcoustic Detection . . . . . . . . . . . . . . . . . . . . . .30

Electromagnetic Surge Detection . . . . . . . . . .31

Electromagnetic/Acoustic Surge Detection . . .31

Earth Gradient . . . . . . . . . . . . . . . . . . . . . . . . .33

VIII Solutions for Cable Fault Locating

Underground Utility Locatingand Tracing Equipment . . . . . . . . . . . . . . . . . .34

Time Domain Reflectometers . . . . . . . . . . . . . .35

Cable Fault Pinpointing Equipment . . . . . . . . .36

High-Voltage DC Dielectric Test Sets . . . . . . . .37

Suitcase Impulse Generator . . . . . . . . . . . . . . .37

Cable Analyzer . . . . . . . . . . . . . . . . . . . . . . . . .38

Power Fault Locators . . . . . . . . . . . . . . . . . . . .38

Impulse Generators . . . . . . . . . . . . . . . . . . . . .40


3/44

Table of Figures

Fault Finding Solutions 1

Figure Page

1 Good insulation . . . . . . . . . . . . . . . . . . . . . . .2

2 Equivalent circuit of good cable . . . . . . . . . .2

3 Bad insulation . . . . . . . . . . . . . . . . . . . . . . . .2

4 Ground or shunt fault on the cable . . . . . . . .3

5 Fault region simplified diagram . . . . . . . . . . .3

6 Open or series fault on the cable . . . . . . . . . .3

7 Test for insulation (fault) resistanceusing a Megger insulation tester . . . . . . . . .4

8 Loop test for continuity using aMegger insulation tester . . . . . . . . . . . . . . . .4

9 TDR test for cable length . . . . . . . . . . . . . . . .5

10 TDR test for continuity . . . . . . . . . . . . . . . . . .5

11 How cable locators work . . . . . . . . . . . . . . . .7

12 Cable under test . . . . . . . . . . . . . . . . . . . . . . .713 Using an ohmmeter to measure

resistance of the circuit . . . . . . . . . . . . . . . . .8

14 Hookup showing ground rod atfar end of cable under test . . . . . . . . . . . . . .9

15 Hookup with far end of cableunder test isolated . . . . . . . . . . . . . . . . . . . . .9

16 Current coupler connection toneutral on primary jacketed cable . . . . . . . . .9

17 Inductive coupling to neutral onprimary jacketed cable . . . . . . . . . . . . . . . . .10

18 Use of return wire to

improve current loop . . . . . . . . . . . . . . . . . .10

19 Circling path with receiver . . . . . . . . . . . . . .10

20 No interference, no offset betweenmagnetic field center and centerof cable . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11

21 Depth measurement using null methodwith antenna at 45-degree angle . . . . . . . .11

22 Offset caused by interference fromnontarget cable . . . . . . . . . . . . . . . . . . . . . .11

23 Aircraft radar . . . . . . . . . . . . . . . . . . . . . . . .12

24 TDR reflections from perfect cable . . . . . . .13

25 TDR used to measure length of cablewith far end open . . . . . . . . . . . . . . . . . . . .14

26 TDR used to measure length of cablewith far end shorted . . . . . . . . . . . . . . . . . .14

27 TDR measuring distance to alow-resistance fault to ground . . . . . . . . . . .15

Figure Page

28 TDR used to measure distance toopen in conductor . . . . . . . . . . . . . . . . . . . .16

29 TDR used to localize distance to splice . . . .17

30 TDR used to localize distance to T-tap . . . . .18

31 TDR used to localize distance tofault relative to a landmark . . . . . . . . . . . . .18

32 Three-stake method . . . . . . . . . . . . . . . . . . .18

33 Block diagram of surge generator . . . . . . . .19

34 Energy vs. voltage for a 4-F, 25-kVsurge generator . . . . . . . . . . . . . . . . . . . . . .19

35 Energy vs. voltage for a 12-F, 16-kVsurge generator . . . . . . . . . . . . . . . . . . . . . .20

36 Energy vs. voltage for a constantenergy 12-F, 16/32-kV surge generator . . .20

37 Acoustic shock wave from arcing fault . . . .2138 Single point grounding . . . . . . . . . . . . . . . .22

39 Inductive arc reflection diagram . . . . . . . . .23

40 Resistive arc reflection diagram . . . . . . . . . .23

41 Sectionalizing method . . . . . . . . . . . . . . . . .24

42 Basic Wheatstone Bridge . . . . . . . . . . . . . . .24

43 Murray Loop Bridge application . . . . . . . . .24

44 Application of Bridge/TDR . . . . . . . . . . . . . .25

45 Coaxial power cable with neutralbridges over splices . . . . . . . . . . . . . . . . . . .25

46 Electromagnetic detection in single-phasePILC cable with bonded grounds . . . . . . . . .26

47 Electromagnetic detection of faults onthree-phase power cable . . . . . . . . . . . . . . .26

48 Arc reflection method of HV radar . . . . . . .27

49 Arc reflection and differential arcreflection methods of HV radar . . . . . . . . . .27

50 Surge pulse reflection methodof HV radar . . . . . . . . . . . . . . . . . . . . . . . . . .28

51 Decay method of HV radar . . . . . . . . . . . . .29

52 Acoustic surge detection . . . . . . . . . . . . . . .30

53 Electromagnetic pinpointing . . . . . . . . . . . .31

54 Acoustic/electromagnetic pinpointing . . . . .32

55 SD-3000 display at positions 1, 2, & 3 . . . . .33

56 AC voltage gradient . . . . . . . . . . . . . . . . . . .33

57 DC voltage gradient . . . . . . . . . . . . . . . . . . .33


4/44

Cable Characteristics

SECTION I

2 Fault Finding Solutions

GOOD CABLE INSULATION

When voltage is impressed across any insulationsystem, some current leaks into, through, andaround the insulation. When testing with dc high-voltage, capacitive charging current, insulation

absorption current, insulation leakage current, andby-pass current are all present to some degree. Forthe purposes of this document on cable faultlocating, only leakage current through the insula-tion will be considered.

For shielded cable, insulation is used to limit cur-rent leakage between the phase conductor andground or between two conductors of differingpotential. As long as the leakage current does notexceed a specific design limit, the cable is judgedgood and is able to deliver electrical energy to aload efficiently.

Cable insulation may be considered good whenleakage current is negligible but since there is noperfect insulator even good insulation allows somesmall amount of leakage current measured inmicroamperes. See Figure 1.

The electrical equivalent circuit of a good run ofcable is shown in Figure 2. If the insulation wereperfect, the parallel resistance RP would not existand the insulation would appear as strictly capaci-tance. Since no insulation is perfect, the parallel orinsulation resistance exists. This is the resistancemeasured during a test using a Megger InsulationTester. Current flowing through this resistance is

measured when performing a dc hipot test asshown in Figure 1. The combined inductance (L),series resistance (RS), capacitance (C) and parallelresistance (RP) as shown in Figure 2 is defined asthe characteristic impedance (Z0) of the cable.

Amps

HVTestSet

Figure 1: Good insulation

Series Resistance

RS

ParallelResistance

RP

CapacitanceC

Inductance

L

Z0 Z0

Figure 2: Equivalent circuit of good cable

mAmps

HVTestSet

Figure 3: Bad insulation

WHEN CABLE INSULATION IS BAD

When the magnitude of the leakage currentexceeds the design limit, the cable will no longerdeliver energy efficiently. See Figure 3.

Why A Cable Becomes BadAll insulation deteriorates naturally with age,especially when exposed to elevated temperaturedue to high loading and even when it is not physi-cally damaged. In this case, there is a distributedflow of leakage current during a test or whileenergized. Many substances such as water, oil andchemicals can contaminate and shorten the life ofinsulation and cause serious problems. Cross-linkedpolyethylene (XLPE) insulation is subject to a con-dition termed treeing. It has been found that thepresence of moisture containing contaminants,irregular surfaces or protrusions into the insulation

plus electrical stress provides the proper environ-ment for inception and growth of these treeswithin the polyethylene material. Testing indicatesthat the ac breakdown strength of these treedcables is dramatically reduced. Damage caused bylightning, fire, or overheating may require replace-ment of the cable to restore service.


5/44

Cable Characteristics

SECTION I


mAmps

Fault

HVTestSet

Figure 4: Ground or shunt fault on the cable

Phase Conductor

Shield or Neutral

Spark GapG Fault Resistance

RF

Phase Conductor

Shield or Neutral

Spark GapG

Fault ResistanceRF

Figure 5: Fault region simplified diagram

Figure 6: Open or series fault on the cable

CABLE FAULTS DESCRIBED

When at some local point in a cable, insulation hasdeteriorated to a degree that a breakdown occursallowing a surge of current to ground, the cable isreferred to as a faulted cable and the position of

maximum leakage may be considered a cata-strophic insulation failure. See Figure 4. At thislocation the insulation or parallel resistance hasbeen drastically reduced and a spark gap hasdeveloped. See Figure 5.

Occasionally a series fault shown in Figure 6 candevelop due to a blown open phase conductorcaused by high fault current, a dig-in or a failedsplice.


6/44

Fault Locating Procedures

SECTION II


LOCATE FAULTS IN BURIED PRIMARY CABLE

After all clearances have been obtained and thecable has been isolated in preparation for cablefault locating, it is strongly recommended that afixed plan of attack be followed for locating the

fault. As in diagnosing any complex problem, fol-lowing a set step-by-step procedure will help inarriving at the solution or, in this case, pinpointingthe fault efficiently.

At the very start, it is a good idea to gather asmuch information as possible about the cableunder test. Information that will help in the faultlocating process is:

Cable type is it lead covered, concentric neu-tral (bare or jacketed), tape shield?

Insulation type is it XLPE, EPR, Paper?

Conductor and size is it CU, AL, stranded,solid, 2/0, 350 MCM?

Length of the run how long is it?

Splices are there splices, are the locationsknown?

T-taps or wye splices are there any taps, arethe locations known, how long are branches?

After obtaining the cable description the acronymTALL can help you remember the procedure forfinding cable faults in buried cable.

TEST ANALYZE LOCALIZE

LOCATETEST THE CABLE

Fault Resistance and Loop Test

Although most faults occur between phaseand ground, series opens also occur such asa blown open splice or a dig-in. Phase-to-phase faults can also occur on multi- phaseruns. Helpful information can be gatheredwith a Megger Insulation Tester that hasboth megohm and an ohm (continuity)range.

Make a series of measurements as follows:

At end A, connect the instrumentbetween the faulted conductor andground as shown in Figure 7. Using aninsulation resistance range, measure andrecord this resistance reading.

Figure 8: Loop test for continuity using a Megger insulationtester

At end A, connect the instrument between eachof the other phase conductors, if any, andground and record the insulation resistancereadings.

After connecting a short between the phase andneutral at end B (Figure 8), do a loop test forcontinuity at end A using the ohms or continuityrange on the instrument. If a reading of greaterthan 10 ohms is obtained when the cable has aconcentric neutral, test the conductor and neu-tral independently by using a nearby good cableas a return path. This will help to determinewhether it is the conductor or neutral that isthe problem. A reading in the hundreds of ohmsis a good indication of corroded neutral if work-ing on a bare concentric-type cable. If no nearbygood cable is available, use a long insulated con-ductor to complete the loop from end B. If areading of infinity is measured either the phaseconductor or the neutral is completely openbetween end A and end B which could becaused by a dig-in or a fault that has blownopen the phase conductor.

Repeat all tests from end B and record allreadings.

Fault

End A End B

MEGGERInsulationTester

Fault

End A End B

Shorting Strap or Grounding Elbow

MEGGERInsulationTester

Figure 7: Test for insulation (fault) resistance using aMegger insulation tester


7/44

Fault Locating ProceduresSECTION II


Figure 9: TDR test for cable length

Figure 10: TDR test for continuity

TDR Tests

Refer to Section IV for details on the use of theTime Domain Reflectometer.

At end A, connect a TDR or DART Cable

Analyzer (use the TDR mode) between the fault-ed conductor and neutral or shield as shown inFigure 9. Look for an upward reflection from theopen end of the cable and measure the lengthto the open using the cursors.

After connecting a short between phase andneutral at end B (Figure 10), look for the down-ward indication of a short circuit at the cableend on the TDR. If the TDR shows an alternatingopen and short when alternately removing andapplying the ground at the end of the cable, thephase and shield are continuous to the cableend. If the short does not appear on the TDRand a high resistance was read during the looptest, either the phase or shield is open at somepoint before the cable end.

If a downward reflection is observed on the TDRand the fault resistance measured less than200 in the test, the fault has been found. If adownward reflection is observed on the TDR andthe fault resistance measured greater than 200 in the test, there is likely a T-tap or wye splice atthat location.

DC Hipot Test

After a surge generator is connected to the cableunder test, do a quick dc proof test to be sure the

cable is faulted and will not hold voltage. Make anote of the kilovolt measurement when the faultbreaks down. This will be an indicator of whatvoltage will be required when surging in orderbreak down the fault when doing prelocation orpinpointing. If there are transformers connectedto the cable under test, a proof test will alwaysindicate a failure due to the low resistance path toground through the transformer primary winding.A dc proof test in this case is not a valid test.

ANALYZE THE DATA

Fault Resistance and Loop Test

If the insulation resistance of the faulted conduc-tor is less than 50 or more than one M, thefault will be relatively easy to prelocate but maybe difficult to pinpoint. For values between 50 and 1 M, the fault may be more difficult tolocate. Some reasons for the difficulty with thesefaults is the possible presence of oil or water inthe faulted cavity or the presence of multiplefaults.

Fault

End A End B

TDR

Fault

End A End B

Shorting Strap or Grounding Elbow

End A

TDR

If tests indicate insulation (fault) resistance valuesless than 10 ohms, it may not be possible to createa flashover at the fault site when surge generatormethods are used. This type of fault is oftenreferred to as a bolted fault. A TDR can be used tolocate this type of fault.

If a measurement of very low resistance in ohms is

made from one end and a high resistance inmegohms from the other end, it is likely that thephase conductor or a splice is blown open.

If the loop test indicates a resistance reading inthe 10 to 1000 ohm range and particularly if thereading varies during the measurement, there isvery likely neutral corrosion on the cable. Thiscould affect success when performing localizingand locating procedures. If the loop test measure-ment is infinity, indicating an open circuit, eitherthe phase conductor or a splice has blown open ora dig-in has occurred.

TDR Tests

If the TDR tests indicate a shorter than expectedcable length with no change of reflections when ashort is applied to the cable end there is likely ablown open splice or phase conductor or a dig-inhas occurred. If the TDR tests indicate a longerthan expected cable run, a thorough route tracemay be in order to detect additional cable notindicated on maps.


8/44

Fault Locating Procedures

SECTION II


DC Hipot Test

If the cable holds voltage during the dc hipot test,the cable may be good. If the cable is faulted,burning may be required to reduce the breakdownvoltage required when surging or you have con-nected to the wrong phase.

Cable Route

At this point, it is recommended that the cableroute be determined or confirmed by consultingaccurate maps or actually tracing the cable route.See Section III. When attempting to localize orlocate the cable fault, prelocation measurementsand pinpointing techniques must be made overthe actual cable path. Being off the route by as lit-tle as a few feet may make the locate an extreme-ly difficult and time-consuming process.

LOCALIZE - PRELOCATE THE FAULTSelection of a localizing technique is based, atleast in part, on the character of the fault. Severaltechniques are fully described in Section VI. Theyare as follows:

Sectionalizing

Bridge single faults

TDR/low-voltage radar faults measuring lessthan 200 and all opens

High-voltage radar methods all faults arcreflection, surge pulse reflection and decay

Electromagnetic impulse detection all shortsand some opens

LOCATE - PINPOINT THE FAULT

Locating, often referred to as pinpointing, is nec-essary before digging up direct buried cable. Afterthe fault has been localized, a surge generator isconnected to one end of the faulted cable andthen listening in the localized area for the telltalethump from the fault. When the thump is not loudenough to hear, it may be necessary to use a surgedetector or an acoustic impulse detector to pin-point the fault.

Voltage gradient test sets are effective in pinpoint-ing faults on direct-buried secondary cable but themethod depends on the fault existing betweenconductor and earth. When the cable is in conduit,a different method must be used. When a single

conductor is contained within a plastic conduit,shorts cannot occur unless water gains accessthrough a crack or other entry point. When a faultdevelops, leakage current flows from the conduc-tor through the break in insulation, and then fol-lows the water to the break in the conduit toearth. If voltage gradient is used, the location ofthe crack in the conduit could be found, but thelocation of the fault in the insulation wouldremain unknown.

LOCATE FAULTS IN ABOVE GROUND PRIMARYCABLE

Some faults can be found by searching for obviousphysical damage to the cable especially if the cablesection is short. If necessary, connect a surge gen-erator and walk the cable and listen for the dis-charge. If the cable is very long it might take agood deal of time to walk the cable while thesurge generator is on. To reduce the total timespent and to minimize high-voltage exposure tothe cable, use a localizing technique beforeattempting to pinpoint the fault.

Once the fault is localized, a listening aid is usedto zero in on the thump that occurs when thesurge generator breaks down the fault. For metal-to-metal (bolted) faults on non-buried cable, an

electromagnetic impulse detector may help to pin-point the fault. The use of electromagneticimpulse detectors is discussed in detail in SectionVI.


9/44

Cable Route Tracers/Cable Locators

SECTION III


7 MEGGER

OVERVIEW

Before attempting to locate underground cablefaults on direct buried primary cable, it is neces-sary to know where the cable is located and whatroute it takes. If the fault is on secondary cable,

knowing the exact route is even more critical.Since it is extremely difficult to find a cable faultwithout knowing where the cable is, it makessense to master cable locating and tracing and todo a cable trace before beginning the fault locat-ing process.

Success in locating or tracing the route of electricalcable and metal pipe depends upon knowledge,skill, and perhaps, most of all,experience. Although locating canbe a complex job, it will very likelybecome even more complex asmore and more underground

plant is installed. It is just asimportant to understand how theequipment works as it is to bethoroughly familiar with the exactequipment being used.

All popular locators/tracers consistof two basic modules:

The transmitter an ac generatorwhich supplies the signal currenton the underground cable or pipeto be traced.

The receiver detects the electro-magnetic field produced by thetransmitted ac current flow. SeeFigure 11.

Before starting, it will be helpfulto obtain the following informa-tion:

What type of cable is it?

Is the cable the same type all theway along its length?

Is the target cable the only cablein the trench?

Are there any taps?

Is the cable run single phase ormultiphase?

Figure 12: Cable under test

Is the cable shielded or unshielded?

Is the cable direct buried or in conduit?

Are there metal pipes or other undergroundstructures under, over or near the target cable?

Is the target cable connected to other cables orpipes through grounded neutrals?

This information will help to select the mostappropriate locator and to prepare to locate thecable successfully. See Figure 12.

Transmitter

AC Current Flow Electromagnetic FieldProduced by Current Flow

Receiver Antenna

Nearby Cables and/or Pipes

Current Return Paths

Tee or Wye Splice

Interfering Cables and Pipes

Cable to be traced

Figure 11: How cable locators work


10/44


SECTION III


Many transmitters are equippedwith some means of indicating theresistance of the circuit that it istrying to pump current throughand can indicate a measurement

of the current actually being trans-mitted. Output current can bechecked in several ways as follows:

By measuring the resistance ofthe circuit with an ohmmeter.When the resistance is less thanapproximately 80,000 , therewill typically be enough currentflowing in the cable to allow agood job of tracing. This is noguarantee that the transmittedcurrent is passing through thetarget cable. The measuredresistance may be affected byother circuits or pipes electrically connected tothe target cable acting as parallel resistances.See Figure 13.

By observing the actual signal strength beingtransmitted by the transmitter. Many transmit-ters provide a measurement or some indicationof output current. A loading indicator on thePortable Locator Model L1070 blinks to indicatethe approximate circuit resistance. A rate of fourblinks per second indicates a low resistance,almost a short circuit providing a very traceablesignal. A rate of one blink every three secondsshows a high resistance and a weaker signal.

By observing the signal power detected by thereceiver. Signal level indicator numbers are dis-played digitally on most receivers and oldermodels may display signal power with analogmeters. The L1070 has both an analog style sig-nal strength bargraph plus a digital numericreadout. Tracing experience gives the operatorthe ability to judge whether or not the numbersare high enough. This is the most practical wayto check signal current flow.

Remember, the more current flow through theconductor the stronger the electromagnetic fieldbeing detected by the receiver and the further

from the conductor being traced the less field isbeing detected.

SELECTING A LOCATOR

Cable locating test sets, often referred to as cabletracers, may be grouped as follows:

Low frequency usually less than 20 kHz some-times referred to as audio frequency (AF).

High frequency usually higher than 20 kHzand in the radio frequency (RF) range to about80 kHz.

60 Hz most tracers provide this mode to allowtracing of energized cables.

Low frequency (AF) is considered the general-pur-pose selection because it is more effective in trac-ing the route of cables located in congested areasdue to less capacitive coupling to everything elsein the ground. Low frequency can be more effec-tive over greater distances due to less capacitiveleakage and because higher signal power isallowed by the FCC. The use of high frequency (RF)is typical in non-congested areas on relatively shortlengths of cable or when a return path cannot beprovided from the far end. If a proper return pathis provided, either low or high frequencies can beused effectively for very long distances. The L1070allows selection of AF, RF, both AF and RF, or 60 Hz

as required by the specific application.

Ohmmeter

If the resistance is too high, ground the far end

If the resistance is still too high, connect an insulated jumper wire for the return path

Figure 13: Using an ohmmeter to measure resistance of the circuit


11/44


SECTION III


HOOKUPS

When a direct-buried secondary cable is to betraced, the transmitter is connected to the conduc-tor. When coaxial types of primary cable aretraced, the signal may be transmitted along either

the phase conductor or the neutral.

Whenever possible use the direct connectionmethod with the test leads supplied with the loca-tor. This is often referred to as the conductivemethod. Connect one output lead (usually red)from the transmitter to the conductor under testmaking sure that the alligator clip is making good

contact. Connect the other lead (usually black) to atemporary metal ground probe and check that thepin is making good contact with the earth. Whenthe earth is dry it may be necessary to use a longermetallic ground stake or to pour water on the

ground rod to give it better contact with theearth. Place the ground rod off to the side as faraway from the target cable as practical, but try toavoid crossing over neighboring cables and pipes.It may be necessary to vary the location of theground rod to obtain suitable results.

For best results, install a temporary ground con-nection to the far end of the conductor being

traced. See Figure 14. In this caseeither AF or RF can be used. If aground cannot be applied to thefar end, use RF and expect that theeffective traceable length may beas short as 200 feet. See Figure 15.The only current flow in this situa-tion is due to capacitive currentflow and after some point the sig-nal disappears.

If a direct connection is impossible,a clamp coupler can be used toinduce the signal current onto thetarget cable. See Figure 16. If trac-ing a primary cable, place the looparound the neutral. When tracingsecondary, connecting jumper wiresfrom the conductor to earth atboth ends of the cable may be nec-

essary to obtain an adequate signalcurrent flow through the targetcable. Remember that for sufficientcurrent to flow to produce a strongtraceable field there must be a loopor return path provided back to thesource.

If a current coupler is not available,the transmitter module itself can beused to couple the signal inductive-ly from an antenna in the base ofthe transmitter into the cable. SeeFigure 17. The transmitter is set on

the earth directly over the targetcable with the arrow on the toppanel in line with the cable. Usethe RF frequency selection andkeep the transmitter and receiverat least 25 feet apart to avoid inter-fering signals generated directlythrough the air.

Transmiter

BIDDLEBIDDLE

RR

L1070

PortableLoc

ator1070Po rtab leLocator

TMM

Transmitter

BIDDLEBIDDLE

RR

L1070

PortableLoc

ator1 070Por table Locator

TMM

Figure 14: Hookup showing ground rod at far end of cable undertest

Figure 15: Hookup with far end of cable under test isolated

Transmitter

Jacket

Neutral

BIDDLEBIDDLE

RR

L1070

PortableLoca

tor1 070Port ableLoc

ator

TMM

Figure 16: Current coupler connection to neutral on primaryjacketed cable


12/44


SECTION III


Keep in mind that the best technique is to connectthe isolated far end of the target cable to a tem-porary ground rod beyond the far end of thecable. This will reduce the loop resistance, increasethe transmitted current flow, and maximize the

strength of the signal to be detected by the receiv-er. See previous Figure 14.

When the far end is parked and isolated, loop cur-rent is entirely dependent upon capacitive cou-pling through the insulation or jacket of the cableand through any faults to ground that may bepresent. See previous Figure 15.

If all else fails and in a very congested area, com-plete the current loop by using a long insulated

jumper wire connected between one side of thetransmitter and the far end of the cable undertest. This technique has limitations as to length

but will definitely limit current flow to the targetcable. See Figure 18. Remember to keep the routeof the return wire well off to the side to avoidinterference.

Direct buried concentric neutral cable can betraced by connecting the transmitter to the con-ductor or the neutral. Remember that when con-nected to the neutral, the signal can more easily

bleed over to other cables and pipes thatmay be connected to the ground. Astronger tracing signal can sometimes bedeveloped when the transmitter is connect-ed to the neutral. This is particularly truewhen using a current clamp or coupler asshown previously in Figure 16.

USING THE RECEIVER

To begin the tracing process, start by circlingthe connection point to the target cable ata radius of 10 feet or so to find the positionwith the strongest signal when using thepeaking mode. See Figure 19. The L1070receiver allows pushbutton selection ofeither the peaking or nulling modes of trac-ing. See Figure 20. Some older models

require a change in position of theantenna head from horizontal to

vertical. Most receivers now alsoprovide an automatic depth meas-urement, usually with the push of abutton. Older units require posi-tioning of the antenna head at a45-degree angle and following theprocess shown in Figure 21.

In the peaking mode of operation,a maximum signal level is obtainedwhen the receiver is positioneddirectly over the target cable. In thenulling mode, a minimum signal isdetected when directly over the

target cable. Some units provide asimultaneous display of bothmodes. In general, if the object oftracing is simply to locate theapproximate path of the targetcable, the peaking mode is recom-mended. If a more accurate trace isrequired such as prior to secondaryfault locating or splice locating, thenulling mode may be the better

Transmitter

Neutral

Figure 17: Inductive coupling to neutral on primaryjacketed cable

Transmitter

BIDDLEBIDDLE

RR

L1070

PortableLoc

ator1 07 0Port able Loc

ato r

T MM

Figure 18: Use of return wire to improve current loop

Transmitter

BIDDLEBIDDLE

RR

L1070

PortableLoc

ator1070Po rtab leLocator

TMM

Figure 19: Circling path with receiver


13/44


SECTION III


Figure 20: No interference no offset betweenmagnetic field center and center of cable

choice. An analog bar-graph display, a digitalnumeric readout, a variable volume audible toneor all three may indicate the receiver signal level.

While walking along the route with the strongestsignal level, note the value of signal strength. Alsowhile tracing, periodically check the depth. If thesignal level numbers drop as you proceed alongthe path away from the transmitter, there shouldbe a corresponding increase in depth. If the signallevel increases as you proceed along the path,there should be a corresponding decrease indepth. If signal level decreases, even though thedepth does not increase, it could mean that youhave just passed a fault to ground or a wye splice.

The transmitter current flow beyond a fault maybe significantly reduced to only capacitive leakageso the resulting drop in signal level may beenough evidence to conclude that a fault to

ground has been passed.When no interference is present, the combinedantennas in the receivers of newer locators willsense both a null and a peak magnetic field at theidentical spot directly over the target cable.Interfering conductors and pipes can cause themagnetic field around the target cable to becomeoval, or egg-shaped rather than circular and con-centric. This will cause an offset between thedetected and actual location. See Figure 22. Thisproblem is often not possible to detect at the timethe locating is being carried out and is only discov-ered when digging begins. To prevent this, every

effort should be made to prevent signal currentfrom bleeding or leaking onto other conductors inthe area, which is often impossible.

Antenna

Conductor or Pipe

Peak Mode Null Mode

X Feet X Feet

X Feet

True Location ofTarget

Indicated Position ofTarget

Conductor or Pipe Interfering Conductoror Pipe

Figure 21: Depth measurement using null methodwith antenna at 45-degree angle

Figure 22: Offset caused by interference fromnon-target cable


14/44

How to See Underground Cable Problems

SECTION IV


METHODS OF OPERATION

Cable analyzers provide a visual display of variousevents on electrical cable and may serve as thecontrol center for advanced cable fault locatingtest systems. Displays include cable traces or signa-

tures which have distinctive patterns. Signaturesrepresent reflections of transmitted pulses causedby impedance changes in the cable under test andappear in succession along a baseline. Whenadjustable markers, called cursors, are moved toline up with reflections, the distance to the imped-ance change is displayed. When used as a TDR,approximate distances to important landmarks,such as the cable end, splices, wyes and transform-ers can also be measured.

Time Domain Reflectometry

The pulse reflection method, pulse echo method

or time domain reflectometry are terms applied towhat is referred to as cable radar or a TDR. Thetechnique, developed in the late 1940s, makes itpossible to connect to one end of a cable, actuallysee into the cable and measure distance tochanges in the cable. The original acronym,RADAR (RAdio Detection And Ranging), wasapplied to the method of detecting distant aircraftand determining their range and velocity by ana-lyzing reflections of radio waves. This technique isused by airport radar systems and police radarguns where a portion of the transmitted radiowaves are reflected from an aircraft or groundvehicle back to a receiving antenna. See Figure 23.

The radar set, other than the electronics to pro-duce the pulses of radio frequency energy, is basi-cally a time measuring device. A timer starts count-ing microseconds when a pulse of radio frequencyenergy leaves the transmitting antenna and then

stops when a reflection is received. The actual timemeasured is the round trip, out to the target andback. In order to determine simply distance out tothe target, the round trip time is divided by two. Ifthe speed of this pulse as it travels through the airin microseconds is known, distance to the targetcan be calculated by multiplying the time meas-ured divided by 2 times the velocity.

Distance = Vp time2

The speed or Velocity of Propagation (Vp) of thispulse in air is nearly the speed of light or approxi-mately 984 feet per microsecond.

This same radar technique can be applied to cablesif there are two conductors with the distancebetween them constant for the length of the runand a consistent material between them for thelength of the run. This means that a twisted pair,any pair of a control cable, any pair of a triplexcable, or any coaxial cable are radar compatible.When applied to underground cable, 10 to 20 volt,short time duration pulses are transmitted at ahigh repetition rate into the cable between thephase conductor and neutral or between a pair ofconductors. A liquid crystal or CRT display showsreflections of the transmitted pulses that are

caused by changes in the cable impedance.Any reflections are displayed on the screen withelapsed time along the horizontal axis and ampli-tude of the reflection on the vertical axis. Sincethe elapsed time can be measured and the pulsevelocity as it travels down the cable is known, dis-tance to the reflection point can be calculated.Pulses transmitted through the insulation of typi-cal underground cable travel at about half of thespeed of light or about 500 feet/s. Movable cur-sors when positioned at zero and a reflectionpoint provide a measurement of distance to thatpoint in feet.

The TDR sees each increment of cable, for exampleeach foot, as the equivalent electrical circuitimpedance as shown in Figure 24. In a perfectlength of cable, all of the components in everyfoot are exactly like the foot before and exactlylike the next foot.

Distance

D = Velocity of Propagation (Vp) X Time (s)

2

Figure 23: Aircraft radar


15/44

How to See Underground Cable ProblemsSECTION IV

This perfect run of cable will produce no reflec-tions until the end of the cable appears. At theend of the cable the pulses see a high impedance(an open circuit), causing an upward reflection. Ifthe cable end is grounded (a short circuit), thepulses see a low resistance and a downward reflec-tion is caused. A low-voltage TDR is an excellent

tool for the prelocation of series open circuits andconductor to conductor shorts. For cable shunt orground faults with a resistance higher than 200ohms the reflection is so small it is impossible todistinguish from normal clutter reflections on thecable. Unfortunately, almost all faults on primaryunderground distribution cable are high resistancefaults in the area of thousands of ohms or evenmegohms. Due to the reflection characteristics ofthese high resistance faults, they are impossible tosee using only the low-voltage TDR. An alternatetechnique such as arc reflection must be utilized toprelocate these faults as discussed in Section VI.

Differential TDR/Radar

When a TDR such as the Megger Model CFL535Fwhich has two inputs and is programmed to allowa display of the algebraic difference between twoinput traces, a technique referred to as differentialTDR can be used. If the two traces (L1 and L2) areidentical, the display will show a totally flat line.When using differential TDR, any differencebetween the two phases (L1 minus L2) will be easi-ly identified on the display. This can be usefulwhen fault locating on a three-phase systemwhere the faulted phase can be compared to agood phase. The fault is likely where the differ-

ence is and the cursor can be positioned to meas-ure the distance to that point.

DESCRIPTIONS AND APPLICATIONS

Low-Voltage TDR/Cable Radar

A low-voltage TDR is an appropriate method tolocalize faults and other impedance changes on

electrical cable such as twisted pair, parallel pair,and coaxial structure. TDRs are available in smallhand-held, larger portable, and rack mount con-figurations for a broad variety of applications.Low-voltage, high-frequency output pulses aretransmitted into and travel between two conduc-tors of the cable. When the cable impedancechanges, some or all of transmitted energy isreflected back to the TDR where it is displayed.Impedance changes are caused by a variety of dis-turbances on the cable including low resistancefaults and landmarks such as the cable end, splices,taps, and transformers. See Figures 25 through 31for typical reflections or cable traces.

Faults That a Low-Voltage TDR Will Display

Low resistance faults of less than 200 betweenconductor and ground or between conductors aredisplayed as downward reflections on the screen.Series opens, since they represent a very highresistance, are displayed as upward going reflec-tions. See Figures 27 and 28.

Landmarks That a Low-Voltage TDR WillDisplay

A TDR can localize cable landmarks, such as splices,wye or T-taps, and transformers. See Figures 29

through 31. The TDR helps to determine the loca-tion of faults relative to other landmarks on thecable. This is especially true on complex circuits.Traces of complex circuits are necessarily also verycomplex and difficult to interpret. To make senseof these complex traces, it is extremely helpful toconfirm the position of landmarks relative to thefaults observed. See Figure 32.

For every landmark that causes a reflection, thereis slightly less transmitted pulse amplitude travel-ing from that point down the cable. This means ona cable run with two identical splices, the reflec-tion from the first splice will be larger than that ofthe second down the cable farther. No conclusionscan be drawn based on the size or height ofreflections at different distances down the cable.


Figure 24: TDR reflections from perfect cable


16/44


SECTION IV


CONTROLS AND INPUTS TO THE TDR

Velocity of Propagation

Certain information must be provided to the TDRbefore it can provide distance information. Most

important is velocity of propagation (VP), thespeed at which the transmitted pulse travels downthe cable under test. This value is used by the ana-lyzer to convert its time measurement to distance.This velocity is primarily dependent on the type ofcable insulation although technically is also affect-ed by conductor size and overall cable diameter.The table to the right shows typical velocity valuesfor various primary cable types.

An alternate method to determine an unknownvelocity value is to:

1. Set the right cursor to the upward-going reflec-tion at the end of the cable section.

2. Determine the true length of the section ofcable under test.

3. Adjust the velocity until the correct distance isdisplayed.

If a known length of cable is available on a reel,the above procedure may be used. The longer thesample of cable the better for an accurate deter-mination of velocity.

Open End

DART Analyzer orlow-voltage TDR

BIDDLE RRR DART

ANALYSISSYSTEM

Shorted End


BIDDLE RRR DART

ANALYSISSYSTEM

Figure 25: TDR used to measure length of cable with far end open

Figure 26: TDR used to measure length of cable with far end shorted


17/44


The units of velocity can be entered into the DARTAnalyzer or TDR in feet per microsecond (ft/s),meters per microsecond (m/s), feet per microsec-ond divided by 2 (Vp/2) or percentage of the speedof light (%).

The values in the Velocity of Propagation Table areonly approximate and are meant to serve as aguide. The velocity of propagation in power cablesis determined by the following:

Dielectric constant of the insulation

Material properties of the semiconductingsheaths

Dimensions of the cable

Structure of the neutrals, integrity of the neu-trals (corrosion)

Resistance of the conductors

Additives in the insulation

Propagation characteristics of the earth sur-rounding the cable

With such a large number of variables and a num-ber of different manufacturers, it is impossible topredict the exact velocity of propagation for agiven cable. Typically, utilities standardize on onlya few cable types and manufacturers and have soilconditions that are similar from installation toinstallation. It is highly recommended that faultlocation crews maintain records of propagationvelocities and true locations. Using this informa-

tion, accurate, average propagation velocities canbe determined.


Open EndDART Analyzer orlow-voltage TDR

Low resistance fault to ground

BIDDLE RRR DART

ANALYSISSYSTEM

Figure 27: TDR measuring distance to a low-resistance fault to ground

Insulation Wire Vp Vp Vp Vp

Type kV Size Percent Ft/s M/s Ft/s

EPR 5 #2 45 443 135 221

EPR 15 #2 AL 55 541 165 271

HMW 15 1/0 51 502 153 251

XLPE 15 1/0 51 502 153 251

XLPE 15 2/0 49 482 147 241

XLPE 15 4/0 49 482 147 241

XLPE 15 #1 CU 56 551 168 276

XLPE 15 1/0 52 512 156 256

XLPE 25 #1 CU 49 482 147 241

XLPE 25 1/0 56 551 168 276

XLPE 35 1/0 57 561 171 280

XLPE 35 750 MCM 51 502 153 251

PILC 15 4/0 49 482 147 241

XLPE 0.6 #2 62 610 186 305

Vacuum 100 984 300 492

Velocity of Propagation Table


18/44


SECTION IV


Range

Range is the maximum distance the TDR seesacross the face of the display. Initially, select arange longer than the actual cable length so theupward-going reflection from the end can be

identified. Move the right cursor to that upwardreflection and measure the total length. Does themeasurement make sense? A range can then beselected that is less than the overall cable run butthe TDR will only see out the distance of the rangesetting.

Gain

Gain changes the vertical height or amplitude ofreflections on the display. It may be necessary toincrease the gain on a very long cable or a cablewith many impedance changes along its path toidentify the end or other landmarks. Gain adjust-

ment has no effect on measurement accuracy.

Cursors

For all TDR measurements, the cursor is positionedat the left side of the reflection, just where itleaves the horizontal baseline either upward ordown. Move the right cursor to the reflection ofinterest just as it leaves the base line so that theTDR can calculate its distance. If the left cursor isset to the left of the first upward-going reflection,its zero point is at the output terminals of theinstrument. If you do not recalibrate, it will benecessary to subtract your test lead length from alldistances measured. Remember, the TDR measures

every foot of cable from the connector on theinstrument to the reflection of interest.

When the test leads are especially long (such as125 feet long on most high-voltage radar systems),it is often desirable for you to set the left cursor tothe end of the test leads. When this offset is cali-brated, the distance indicated by the right cursor

will not include the length of the test leads. To dothis calibration in the field simply touch the endsof the test leads and position the left cursor at thetoggle point as the TDR sees an open and then ashort. Press the Save Offset to set the left cursorzero to that point.

Zoom

When you have set the cursor at the reflection ofinterest, the distance to that point on the cablerun will appear in the distance readout. When azoom feature is provided, the area centeredaround the cursor is expanded by the zoom factorselected: X2 (times 2), X4 (times four), etc. It isoften possible to set the cursor to a more preciseposition when the zoom mode is activated and thereflection is broadened.

Pulse Width

The width of the pulses generated by the TDR typ-ically ranges from 80 nanoseconds up to 10microseconds. As range is changed from shorter tolonger, the pulse width is automatically increasedin width so that enough energy is being sentdown the cable to be able to see some level ofreflection from the end. The wider the pulse themore reflection amplitude but the less resolution.

The narrower the pulse the more resolution butless reflection amplitude. For the best resolution orin order to see small changes on the cable, a nar-row pulse width is required and in order to see the

BIDDLE RRR DART

ANALYSISSYSTEM


Open conductor fault

Figure 28: TDR used to measure distance to open in conductor


19/44



end a wide pulse width may be required. Thepulse width may be changed manually to overridethe automatic selection. An effect termed pulsedispersion widens the pulse as it travels down along run of cable so resolution may be worse

toward the end of a long cable.

DISTANCE MEASUREMENTS

Three-Stake Method

Measurements to a fault using a low-voltage TDRare strictly a localizing technique. Never dig a holebased solely on a TDR measurement. There are toomany variables that include:

The exact velocity

The exact cable route

The accuracy of the TDR itself

The three-stake method is a means to get a rea-sonably accurate fault pinpoint using only theTDR. The method consists of making a fault dis-tance reading from one end (terminal 1) of thefaulted line and placing a marker (stake 1) at thatposition as shown in Figure 32. With the TDR con-nected at the other end of the line (terminal 2),find the fault distance for a second marker (stake2). In actual practice, stake 2 may fall short ofstake 1, may be located at the same point, or maypass beyond stake 1. In any case the fault will liebetween the two stakes. It is important that thesame velocity setting is used for both measure-ments and the distance measurements are made

over the actual cable route. This may mean tracingthe cable.

Location of the third marker (stake 3), the actualfault, may be found by using the proportionalitythat exists between the fault distances, d1 and d2,and their error distances, e1 and e2. To locatestake 3, measure the distance d3 between stakes 1

and 2 and multiply it by the ratio of distance d1 tothe sum of distances d1 and d2. Stake 3 then isplaced at this incremental distance, e1, as meas-ured from stake 1 toward stake 2.

e1 = d3 d1

d1 + d2

Alternatively, stake 3 can be placed at the incre-mental distance, e2, as measured from stake 2toward stake 1.

e2 = d3 d2

d1 + d2

This third stake should be very close to the fault.

A practical field approach (with no math involved)is to make a second set of measurements fromboth ends with a different velocity. If the distancebetween stakes 1 and 2 was 50 feet, by adjustingthe velocity upward the new distance measure-ments may reduce the difference to 30 feet. Withenough tests at differing velocities the distancecan be lowered to a reasonable backhoe trenchingdistance.


Splice

BIDDLE RRR DART

ANALYSISSYSTEM

Figure 29: TDR used to localize distance to a splice


20/44


SECTION IV



Open End of Tap

BIDDLE RRR DART

ANALYSISSYSTEM

Tee (Y) Splice

Figure 30: TDR used to localize distance to a T-tap

Open End

Low Resistance Fault to Ground

Open End

Tee (Y) Splice

Splice

DART Analyzer or

low-voltage TDR

BIDDLE RRR DART

ANALYSISSYSTEM

Figure 31: TDR used to localize distance to a fault relative to a landmark

Figure 32: Three-stake method

d3

e1 e2d1 d2

Stake 1take 1 Stake 3take 3 Stake 2take 2

Faultault

Terminal Terminal 2


21/44

Surge Generators, Filters and CouplersSECTION V


SURGE GENERATORS

The first commercially available surge generatorsfor underground cable fault locating were intro-duced in the late 1940s. The device is basically ahigh voltage impulse generator consisting of a dc

power supply, a high voltage capacitor, and sometype of high voltage switch. See Figure 33.

The power supply is used to charge the capacitor

to a high voltage and then a contact closure dis-charges the capacitor into the cable under test. Ifthe voltage is high enough to break down thefault, the energy stored in the capacitor is rapidlydischarged through a flashover at the fault creat-ing a detectable sound or thump at ground

level. The important specifications of a thumperare the maximum voltage it can develop and howmuch energy it delivers to the fault.

The classical fault locating process has been tohook up the surge generator, crank up the volt-age, and walk back and forth over the cable routeuntil the thump is heard or better yet to feel theearth move. This process pinpoints the fault allow-ing a repair crew to dig a hole and repair thecable. In some cases, it may take hours (or days) towalk the cable and definitely locate the fault.During that time, the cable is being exposed tohigh voltage thumping.

A few years after polyethylene cable began to beinstalled underground, evidence began to surfacethat due to treeing in the insulation, high-volt-age thumping of this plastic cable for long periodsof time was doing more harm than good. Thesame is not true for PILC cables where typically

higher voltage and more energy is required tolocate faults with no damage to the cable. There ismixed opinion as to the treeing situation in EPR.Due to this treeing situation, many utilities issuedwork rules reducing the maximum allowable volt-

age to be used for fault locating.

Energy

The energy output of any surge generator meas-ured in Joules (Watt-Seconds) is calculated as fol-lows:

E = V2 C2

where E = Energy in Joules, C = capacitance in f,

V = voltage in kV

To increase the bang at the fault the only twooptions are to increase the voltage which can bedone by the operator or increase the capacitancewhich must be done by the manufacturer. Figure34 shows the output energy curve of a typical fourmicrofarad surge generator that generates 1250Joules at a maximum voltage of 25 kV. If the faultlocating crew is told that the output voltage ofthe thumper must be limited to 12.5 kV (one halfof 25 kV), the output energy of their thumper isreduced by a factor of four down to 312 Joules.

In a practical world, 300 to 400 Joules is thethreshold for hearing a thump at ground levelwith no acoustic amplification and with very littlebackground noise. If the thump at the fault can

not be heard, the only option is to increase volt-age in order to find the fault, make a repair andget the lights back on.

HV dc

Power

SupplyCap

Proof / Burn Mode

Surge Mode

Discharge

Resistor

Discharge

Switch

HV Switch

HV

HV Return

Safety Ground

Chassis

Figure 33: Block diagram of surge generator

1250

0

200

400

600

800

1000

1200

1400

1 3 5 7 9 11 13 15 17 19 21 23 25

kV

Joules

Figure 34: Energy vs. voltage for a 4-F, 25-kVsurge generator


22/44

Surge Generators, Filters and Couplers

SECTION V


To help in lowering the voltage required to locateunderground faults, the PFL-4000 Surge Generatoruses a 12 microfarad capacitor producing 1536Joules at 16 kV. See Figure 35. This allows thump-ing at lower voltages while still delivering reason-

able energy to the fault. Thumping at 12.5 kV, asabove, now produces a very audible 937 Joules.

Different surge generator energy levels are

required to do fault locating on different lengths

and types of cable constructions. XLPE and EPRinsulated cables typically require much less energyto locate a fault than a lead cable of comparablesize and construction. For long runs or complexsystems of lead cable that cannot be broken downinto small manageable sections, high voltage andenergy may be required to flashover a fault.

There are currently two types of surge generatorsavailable, one referred to as progressive energy asdescribed above and the second, constant energy.The constant energy units contain two or morecapacitors with a corresponding voltage range foreach capacitor. The energy is only constant at themaximum voltage on each range. A typical exam-

ple is a PFL-6000 with two voltage ranges of 0 to16 kV and 0 to 32 kV. When the 16 kV range isselected, a 24 F capacitor is switched in and whenthe 32 kV range is selected a 6 F capacitor is used.In this case, at 16 kV or 32 kV the energy outputwill be a constant 3072 Joules. See Figure 36.

Capacitance

Cables by their very nature are capacitive sincethey consist of two conductors separated by aninsulator. The two conductors in power cable arethe phase conductor and the shield, sheath, or

concentric neutral. These two conductors are sepa-rated by XLPE, EPR, or oil impregnated paper.

Safety is always a priority even when a cable is notenergized because, as any capacitor, the cable willhold a charge until discharged or grounded. Acable must always be grounded before making anyconnections even if the cable has been parked andisolated as it may pickup up a charge from thefield of adjacent energized phases.

The longer the cable or the more complex the sys-tem or network, the higher the capacitance. If thesurge generator capacitor is smaller than the cablecapacitance, the fault will not discharge until cablecapacitance is fully charged which could take mul-tiple surges. If the cable capacitance is smallerthan the surge generator capacitance, the faultwill typically flashover on the first try.

Voltage

Deciding on surge generator voltage levels isextremely important. Without a high enough volt-age, the fault will not break down. Very high volt-age surging for long periods of time may promotethe growth of treeing and reduce cable life. If thefault does not flashover, there will be no thumpthat identifies and pinpoints the fault. A very

important factor to consider is that the voltagepulse doubles in peak-to-peak amplitude on agood cable as it reflects from the isolated open

1536

0

200

400

600

800

1000

1200

1400

1600

1 3 5 7 9 11 13 15

kV

Joules

Figure 35: Energy vs. voltage for a 12-F, 16 kVsurge generator

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Joules

0

500

1000

1500

3000

3500

2500

2000

3072 3072

Figure 36: Energy vs. voltage for a 12-F, 16/32-Fsurge generator


23/44



end. This also applies if the cable is faulted but thevoltage doubling only occurs between the faultthe open isolated end of the cable.

When surging at 15 kV, the cable between thefault and the end is exposed to a shock wave of 30kV peak-to-peak. A hint for fault locating on acable that has several splices and has been workedon from the same end is to look for the currentfault past the last splice. That section of cable wasexposed to voltage doubling during the previousfault locate with a high probability that the pres-ent fault is at a weak spot in that length of cable.

BASIC SURGE GENERATOR OPERATION

Proof/Burn

A proof test is performed to determine whether ornot a cable and accessories are good or bad. The

result of a proof test is on a go/no-go basis. Thevoltage is increased on the cable under test to therequired voltage level and held there for a periodof time. If there is little or no leakage current andthe voltage reading is stable, the cable is consid-ered to be good. If a voltage is reached where thereading becomes unstable or drops with a dramat-ic increase in current, it is considered to be bad.This test should be done initially as described inSection I to help establish that the cable is actuallybad and then to gain some information on thefault condition. A quick check can also be doneafter repair to be sure there is not another faultand to check workmanship on the splice.

The burn mode is used when the fault will notflashover at the maximum available voltage of thesurge generator. This condition is due to the elec-trical characteristics of the fault that may bealtered by applying voltage to the cable until thefault breaks down and then supplyingcurrent flow. This causes conditioningor additional damage at the fault loca-tion that in turn decreases the faultresistance and reduces voltagerequired for breakdown.

When applied to paper-insulatedcable, the insulation actually burns

and becomes charred, permanentlyaltering the fault characteristics. Asapplied to XLPE cable, heat producedby arcing at the fault can soften theinsulation but when arcing is stoppedthe insulation returns to a solid condi-tion without changing its characteris-tics drastically. Burning can be effec-tive on a splice failure or a water filledfault.

Surge

In the surge mode, the internal capacitor ischarged up to the level selected with the voltagecontrol and then discharged into the cable. Thisprocess can be automatically repeated on a time

basis by adjusting the surge interval control ormanually by push-button on some models. A surgeof current from the discharging capacitor travelsdown the cable, arcs over at the fault, and returnsback to the capacitor on the neutral or sheath.This rapid discharge of energy causes an audibleexplosion and the sound created travels outthrough the earth and is used to pinpoint the faultlocation. See Figure 37. It is assumed that thesound travels in a straight and direct path up tothe surface of the earth. Sometimes the soil condi-tions are such that the sound travels away in adownward direction or is absorbed and cannot beheard. In this case, some type of listening device or

surge detector may be needed to assist in pin-pointing. If the surge of current sees a high resist-ance path back to the capacitor, as is the casewhen the neutral is corroded, the sound level cre-ated at the fault will be minimal. This current flowback through the earth can also cause a rise inpotential of any metallic structures mounted in theground and a difference in potential on the sur-face.

Ground

If the surge generator safety ground is connectedproperly, the ground mode absolutely and posi-

tively grounds and discharges the surge genera-tors capacitor and the cable under test. Afterturning the main power switch off, which dis-charges the capacitor and cable through a resistor,always move the mode switch to ground beforeremoving test leads.

Faultault

Surge Generator

Shock Wave

Figure 37: Acoustic shock wave from arcing fault


24/44

Surge Generators, Filters and Couplers

SECTION V


SINGLE POINT GROUNDING

For safety, always use the single point groundingscheme as shown in Figure 38 when using a surgegenerator. When making or removing connectionsto a cable always follow your companys safety

rules and regulations.

Check the isolated cable for voltage and ground it.Connect the surge generator safety ground to theground rod at the transformer, switch cabinet orpole. Next, connect the high-voltage return lead tothe shield or neutral as close as possible to thehigh-voltage connection. Leave the neutralgrounded at both ends of the cable. Finally, con-nect the high-voltage test lead to the phase con-ductor. When removing test leads, use the oppo-site sequence by removing the high voltage, highvoltage return and lastly the safety ground. Thelocal ground is only required if company safety

procedures demand it. The safest and lowest resist-ance safety ground connection is system neutralwhich will keep the equipment at zero volts in thecase of a backfeed.

ARC REFLECTION FILTERS AND COUPLERS

In order to reduce the cable exposure to high volt-age surging and thereby avoid the possibility ofsetting the cable up for future failures, somemethod of fault prelocation should be used. Thesurge pulse and arc reflection methods of preloca-tion have been used for many years. In order touse either method, additional equipment isrequired including a DART Analysis System as dis-

cussed in detail in Section VI.

A signal coupler must be added to the surge gen-erator to provide the additional capability of usingthe surge pulse reflection method of prelocation.See Figure 39. The coupler can be an inductive orcapacitive type that is used to pick up reflections

on the cable and send them to the DART Analyzer.Both types of couplers work effectively and theonly difference is that the captured wave shapesvary slightly.

An arc reflection filter is necessary to provide thecapability of using the arc reflection method. Thisfilter allows a TDR developing 10 to 20 volt pulsesto be connected to the same cable that is alsobeing surged at 10,000 volts. The filter also doessome pulse routing to make sure both high- andlow-voltage pulses are sent down the cable undertest. The primary purpose of the filter is to allowthe TDR or analyzer to look down the cable whileit is being surged and, of course, to allow thiswhile not destroying the analyzer in the process.The filter may also contain the coupler necessaryto provide surge pulse capability.

There are two types of arc reflection filters, induc-tive and resistive. Both types are placed in the cir-cuit between the surge generator and the cableunder test. The inductive filter, as shown in Figure39, uses a choke that slows the surge generatorpulse down, extending it over time. This makes thearc at the fault last longer and reflects more TDRpulses, providing a higher probability that a down-ward going reflection will be captured. The induc-tance of the choke also blocks the TDR pulses from

High voltage

High-voltage return

Safety ground

Local ground

(supplemental)

Figure 38: Single point grounding


25/44



going back to the surge generators capacitor andbasically being shorted out. One advantage of theinductive filter is that it helps to clamp or limit thevoltage applied to the cable under test to only thelevel required to breakdown the fault. The choke

in the inductive filter also absorbs less energy cre-ated by the surge generator, letting more downthe cable to arc over at the fault.

The second type of filter, as shown in Figure 40,uses a resistor to do the job of pulse routing andhas the benefits of lower cost and smaller size. Theresistor still blocks the TDR pulses and changes thesurge generator pulse slightly but does not limit orclamp the voltage. The resistive filter tends toabsorb somewhat more of the surge generatorenergy than the inductive filter.

Analyzer/TDR

Analyzer

Inductive surge

pulse coupler

roo urn mpu se

Arc Reflection

HV

HV return

Figure 39: Inductive arc reflection diagram

Analyzer

Analyzer/TDRInductive surge

pulse coupler

Proof burn impulse

Arc Reflection

HV

HV return

Figure 40: Resistive arc reflection diagram


26/44

the relatively short bad section remaining canbe replaced. If the cable is in duct or conduit, thebad section can be replaced. This relatively primi-tive method is usable on most types of phase-to-ground and phase-to-phase faults.

Resistance Ratio

Often called thebridge method, avariation of theWheatstoneBridge is anexample of theresistance ratiomethods. SeeFigure 42.

When using aWheatstone

Bridge, B1, B2,and C2 representknown resistanc-es. C1 representsthe unknownresistance.

At balance, typi-cally by adjusting the resistance values of B1 andB2 when the zero center null detector D indicateszero, C1/C2 = B1/B2

Therefore, C1 = (C2 x B1)/B2

A variation on the Wheatstone Bridge is theMurray Loop Bridge. Figure 43 shows that theadjacent resistances, RC1 of a faulted cable in aloop with RC2 of a good cable can be made torepresent C1 and C2 of the Wheatstone Bridge.Similarly, corresponding portions of a slidewireresistor RB1 and RB2 can be made to represent theresistances B1 and B2. At balance in the MurrayLoop Bridge, RC1/ RC2 is equal to RB1/RB2.

C2B2

C1B1

D

Battery

Figure 41: Sectionalizing method

Locating Methods

SECTION VI


OVERVIEW

Localizing or prelocation methods provide anapproximate distance to a fault. With cable in con-duit or duct, an approximate distance is all that isrequired because the bad cable will be pulled outand new back in. With direct buried undergroundresidential distribution loop fed circuits, localizingcan be used to isolate a bad section between twopad-mount transformers. The bad section can thenbe parked and the loop fed from both ends. In thecase of a radial feed localizing must be followedby an appropriate pinpointing method. Some earlylocalizing methods are as follows:

Sectionalizing

The earliest sectionalizing method has been calledthe Cut and Try Method or the Divide andConquer Method. This was among the first tech-

niques to be used for fault locating on direct-buried cable. Hopefully, its use today is limited tothat of a last resort. See Figure 41.

After isolating both ends of the cable sectionunder test, a Megger Insulation Tester is connect-ed between the conductor and neutral or ground.A faulted cable will have a lower insulation resist-ance than a cable with no fault. After measuringthe fault resistance, a hole is dug half way downthe length of the cable section. The cable is cut atthat location and a resistance measurement ismade on each half. The bad half of the cablewith the fault will a have lower resistance than the

good half and the resistance value on the badhalf should be the same as the fault resistancemeasured on the complete length of cable. A sec-ond dig is made half the distance down the badhalf. Again, the cable is cut and a resistance meas-urement is made on each section to identify thefaulted part of the remaining section. Eventually,

L

First Cut at 1/2 L

Second Cut at 3/4 L

Third Cut at 5/8 L

Fault

Figure 42: Basic WheatstoneBridge

RB2

RB1

D

Battery

RfFault Resistance

Good Conductor

Faulted Conductor

L

RC1

Shorting

Jumper

RC2

Lx

Figure 43: Murray Loop Bridge application


27/44

Locating MethodsSECTION VI

When it is assumed that the resistance of a uni-form conductor is linear and proportional to itslength, and the total length of the cable sectionunder test is L, the distance to the fault, Lx, is cal-culated as follows:

Lx = 2L RB1/RB2

When using the Murray Loop method the seriesresistance and length of the good phase and fault-ed phase must be identical. If the resistances aredifferent, as would exist if one phase contains asplice and not the other, the resulting accuracy isdrastically affected. This is a localizing method nota pinpointing method. Figure 44 illustrates thepractical application of an instrument thatcombines a TDR and Murray Loop Bridge in oneinstrument.

Electromagnetic Surge Detection

Electromagnetic surge detection techniques havebeen used to localize faults on power cable formore than 50 years. Theoretically, the methods canbe used to locate faults on many types of powercable but they are generally used only to identifyfaulted sections of cable in conduit or duct and


mostly on network lead paper cables. Electromag-netic detection methods are used with a surgegenerator which provides the impulse of currentnecessary to produce the strong electromagneticfield required to make the method practical.

Since the current impulse is polarized, the field isalso polarized which gives the method much of itsusefulness. An iron core sheath coil with a second-ary winding will induce a polarized output whensubjected to this field. When a zero centermicroammeter is connected to the sheath coil sec-ondary winding, the direction of the net electro-magnetic field can be determined. This characteris-tic allows a determination of whether the fault isahead of or behind the detector. See Figures 45and 46. Since the sheath coil is polarized, and tomaintain consistent information, always place thecoil with arrow on top pointing toward the surgegenerator.

Single Phase, Coaxial Power Cable with NeutralBridges Over the Splices

Electromagnetic detection methods can be used tolocate faults on coaxial cable systems with access

points such as manholes, cabinets, andpull boxes as shown in Figure 45. Whenthe cable system is designed with neutralbridges over the splices, the pickup coil isplaced directly on the cable under theneutral. The current impulse producedby a surge generator will produce astrong polarized indication as it passes

through the phase conductor and thedirection of the current impulse can bedetermined. As long as a strong positive-ly polarized electromagnetic field can besensed, the fault is located in the portionof cable still ahead. Since the phase con-ductor is isolated at the far end, the cur-rent impulse flows through the fault andthen back to the surge generatorthrough the neutral. Either a weak or noelectromagnetic field will be sensed pastthe location of the fault. Therefore, it ispossible to determine in which directionthe fault is located.

If the sheath coil is placed over the neu-tral, the electromagnetic field producedby the current impulse in the conductoris balanced exactly by the return currentback through the neutral and the meterprovides no indication and stays at zero.When bonded grounds are present in a

Figure 44: Application of Bridge/TDR

ShortingJumper

Fault

Figure 45: Coaxial power cable with neutral bridges over thesplices

Fault

Access #1

Strong Field

Access #2

Weak Field

+- 0 +- 0


28/44

Locating Methods

SECTION VI


system such as usually found in paper-insulated,lead-covered cable (PILC) construction, it may bepossible to use a fault locating technique involvingelectromagnetic detectors, even though the pickupcoil must be placed over the leaded neutral.

Single-Phase PILC Cable with Bonded Grounds inConduit

As shown in Figure 46, this method only applies tocircuits with good bonded grounds at every man-hole location as commonly provided in networksystems. Without bonded grounds, the surge cur-rent through the phase conductor is exactly thesame magnitude as the return surge currentthrough the neutral. With bonded grounds, thecurrent impulse through the phase conductoris slightly greater than the returning surgecurrent in the neutral. This differential iscaused by the small amount of surge currentthat flows through the neutral beyond thefault and into earth through the bondedground at the next manhole and back to thesurge generator through the bonded groundsbefore the fault. See Figure 46. No currentflows through the second bonded groundafter the fault. When relative readings aretaken with the sheath coil placed on thecable both before and after the fault, theywill all be positive. Readings taken on theconductor before the fault will almost alwaysbe noticeably higher in magnitude than thoseafter the fault. However, the difference isoften too small to instill confidencein the cable fault location. Moreimportantly, readings taken on thebonded grounds before the faultwill become progressively higher asthe faulted section is approached.Also, the reading taken on thebonded ground in the first manholeafter the fault will also be high.Readings taken on the bondedground in the second and succeed-ing manholes after the fault will bezero. This process allows the faultedsection of cable can be identified.

Three-phase PILC

Strange as it may seem at first, it isusually easier to locate a fault inthree-phase PILC coaxial cable thanin single phase. See Figure 47. Thecurrent pulse from a surge genera-tor through one phase of a three-

phase cable will generate a stronger magneticfield at the cable surface closest to the faultedphase. When the detector is placed at various posi-tions around the cable ahead of the fault, thereadings will vary in magnitude. When the detec-

tor is placed at various positions around the cablein the first manhole past the fault, all readings willbe the same. Note that almost all of the currentpulse returns to the surge generator through theneutral at the fault site. A small amount of surgecurrent passes through the neutral past the faultand out of the next bonded ground. This smallcurrent finds its way back to the surge generatorfrom earth through the bonded grounds ahead ofthe fault.

First manhole after fault

Fault

First manhole in system

Surge current to Thumper

Surge Current from Thumper

Figure 46: Electromagnetic detection in single-phase PILCcable with bonded grounds

Phase 1 being surged

Sensor position 1 (Strong)

Sensor position 4 (Weak)Sensor position 3 (Weak)Sensor position 2 (Strong) First bonded ground in system

First bonded ground after fault

Fault toneutral

Cross section of cable at lastmanhole before the fault

Cross section of cable at firstmanhole after the fault

All Detector positionsdetect the same strength

Figure 47: Electromagnetic detection of faults on three-phase powercable


29/44

Locating MethodsSECTION VI

DART ANALYZER/HIGH-VOLTAGERADAR

Because the low-voltage TDR is unable toidentify high resistance shunt faults, itseffectiveness as a fault locator on power

cables is limited. When used in a high-voltage radar system with surge genera-tors, filters or couplers, the TDR/Analyzeris able to display both low and high resist-ance faults. The DART Analyzer providesboth TDR and storage oscilloscope func-tions and is able to utilize all of the cablefault locating methods listed below.

Arc Reflection

This method is often referred to as a high-voltage radar technique that overcomesthe 200 limitation of low-voltage radar.

In addition to the TDR, an arc reflectionfilter and surge generator is required. Thesurge generator is used to create an arcacross the shunt fault which creates amomentary short circuit that the TDR candisplay as a downward-going reflection.The filter protects the TDR from the high-voltage pulse generated by the surge gen-erator and routes the low-voltage pulsesdown the cable.

Arc reflection is the most accurate andeasiest prelocation method and should beused as a first approach. The fault is dis-played in relation to other cable land-

marks such as splices, taps and transform-ers and no interpretation is required.

Arc reflection makes it possible for theTDR to display before and aftertraces or cable signatures. See Figure 48.The before trace is the low-voltageradar signature that shows all cable land-marks but does not show the downwardreflection of a high resistance shunt fault.The after trace is the high-voltage sig-nature that includes the fault locationeven though its resistance may be higherthan 200 . This trace is digitized, stored

and displayed on the screen allowing thecursors can be easily positioned in orderto read the distance to the high resistancefault.

Open End

High Resistance Fault to Neutral

Low Voltage Trace

High Voltage Trace

BIDDLE RR R DART

ANALYSISSYSTE M

Figure 48: Arc reflection method of high-voltage radar


Open End

High Resistance Fault to Neutral

Open End

Splice

Arc Reflection Trace

Differential Arc Reflection Trace

BIDDLER

R R DARTANALYSISSYSTEM

Long test leads

Tee splice

Low Voltage Trace

Figure 49: Arc reflection and differential arc reflectionmethods of high-voltage radar


30/44

Locating Methods

SECTION VI


Differential Arc Reflection

This high-voltage radar method isbasically an extension of arcreflection and it requires the useof a surge generator, an arc

reflection filter, and an analyzer.Using an algorithm, the DARTAnalyzer displays the algebraicdifference between the low-volt-age trace and the captured high-voltage trace on a second screen.See Figure 49 on previous page.As in differential TDR, differentialarc reflection eliminates all identi-cal reflections before the faultand the first downward-goingreflection to appear is easily iden-tified as the cable fault. This sim-plifies the fault prelocation partic-

ularly if the fault reflection is notwell defined or the fault is on a complex systemwith lots of clutter and unidentifiable reflections.

Surge Pulse Reflection

This method requires the use of a surge coupler, asurge generator, and an analyzer. The analyzeracts as a storage oscilloscope that captures and dis-plays reflections from the fault produced by thesurge generator high-voltage pulse. The analyzeroperates in a passive mode and is not acting like aTDR by actively sending out pulses. Surge pulse iseffective in locating faults on long runs of simple

circuits and on faults that are difficult to arc overwhich do not show up using arc reflection. Thismethod will find most of the same faults that canbe prelocated using arc reflection, but usually withreduced accuracy and confidence due to greaterdifficulty in interpreting the displayed signature.The captured trace does not show landmarks onthe cable as arc reflection does.

In this method, a surge generator is connecteddirectly to the cable without the use of a filterwhich can limit both the voltage and currentapplied to the fault. Some faults with water or oilin the fault cavity require more ionizing currentand higher voltage than arc reflection can provide.

The surge generator transmits a high-voltage pulsedown the cable creating an arc at the fault thatcauses a reflection of energy back to the surgegenerator. This reflection repeats back and forthbetween the fault and the surge generator untilall of the energy is depleted. A current couplersenses the surge reflections which are captured bythe analyzer and displayed on the screen as atrace. See Figure 50.

To determine the location of the fault, cursors arepositioned at succeeding peaks in the trace. The

analyzer measures time and calculates the distanceto the fault using the velocity of propagation. Forthe trace

Faultfinding AG

Documents