Testing power cables

There are two main reasons for testing power cables: to determine the condition of the cable and to locate a fault on the cable. This article provides an overview of a systematic approach for each of these cases and then goes on to discuss fault classification.

The objective of testing the cable condition is typically to check the quality of cable before installing it, or to detect and remedy potential defects in the cable, which might otherwise jeopardise reliable operation. When testing a cable for potential defects, a common technique is to generate flashovers at the sites of the defects, which can then be located using the standard fault location techniques mentioned later.

Depending on the type of cable insulation and the type of test object (cable or accessory), the following types of test voltage should be used: paper insulated lead-covered cable (PILC) - DC voltage, AC voltage 50/60 Hz, VLF (0.1 Hz); PE/XLPE cable - AC voltage 50/60 Hz, VLF (0.1 Hz); components (joints, terminations, etc) - DC voltage, AC voltage 50/60 Hz.

Alternatively, the cable can be tested non-destructively using dielectric diagnosis and partial discharge techniques. The first diagnostic technology is the Tan Delta technology, based on 0.1 Hz VLF voltage for most insulation materials. In cables with PE and XLPE insulation, dielectric diagnosis based on IRC (isothermal relaxation current) analysis makes it possible to determine how much the cable has aged, while with PILC cables, RVM (return voltage measurement) analysis allows the moisture content of the dielectric to be accurately assessed. Partial discharge measurement is used for recording, locating and evaluating partial discharges in the insulation and fittings of medium-voltage cables, and can reveal a wide range of actual and potential defects.

Cable fault location

The steps needed for determining cable fault locations can be divided into five main categories: fault classification - identifying the type of fault; pre-location - determining the distance to the fault; route tracing - determining the route of the cable; pinpointing - determining the exact position of the fault; cable identification - determining which of several cables is faulty. The main diagram shows the outline procedure for identifying and locating cable faults.

Measuring techniques for cable fault location:

Basic tests

DC test to determine flashover voltage; sheath fault test; VLF test to determine flashover voltage.

Pre-location

TDR pulse reflection measurements; ARM (Arc Reflection Method); ARM plus; ARM power burning; decay-plus (ARM - igniting the fault with using a DC generator); decay (travelling wave method, oscillation method); impulse current decoupling (ICE); three-phase impulse current decoupling (ICE); ICE plus (low-voltage networks only); high-voltage bridge method (pre-locating sheath faults); voltage-drop method (pre-locating sheath faults).

Fault conversion

Burning; power burning.

Route tracing

Line location; line routing.

Pinpointing

Audio frequency generator (twist field and minimum turbidity/ distortion methods); surge wave discharges (acoustic field method, acoustic pinpointing); sheath fault pinpointing.

Cable and phase identification

Phase identification on grounded systems, phase identification and phase determination on live systems.

Fault classification

The first step in locating a cable fault must be very thorough. Accurate plans of the cable route, and knowledge of the cable network and of any civil engineering work that may affect it can often provide the first clues to the location of the fault and also help to guard against misinterpretation of test results. Insulation testing and resistance measurements can then supply information about the characteristics of the fault.

After these measurements have been made, a reflection measurement device (time domain reflectometer - TDR) is used to determine the cable length and to detect the presence of joints/splices and other changes in impedance.

It is good practice and important always to compare the results from faulty conductors with the results from good conductors. The more information that is gathered during these preliminary tests, the easier and more reliable will be the overall fault location process.

Insulation testing

By measuring the insulation resistance between the conductors and the cable shield (phase-to-phase and phase-to-shield), the insulation test indicates the type of fault. The results of the insulation test are important in deciding how to proceed with the fault location process. They can be classified as follows: no fault (no deviation between the resistance values); high-resistance (flashing); cable fault (measurements in the kilohm or megohm range); low-resistance fault (contact between conductors or between conductors and shield/screen).

Insulation test sets with an analog display have proved to be particularly suitable for cable testing as they make it easy to see; for example, anomalies during charging which can indicate faults due to the presence of moisture. With very high resistance faults, DC must be used to establish the breakdown voltage of the cable fault.

Measuring the resistance of a fault

When choosing which pre-location method to use, it is important to have accurate knowledge about the resistance and the phase relation of the fault. The results of the resistance and distance measurements should be carefully recorded. With multiple faults, the faults are often in parallel with each other. A drawing often helps with evaluation. In low-voltage cable networks employing only plastic-insulated cables without shielding or armouring, it is recommended that the PEN conductor is disconnected on both ends and then a check carried out to see if there are any faults indicated by excessive leakage current due to contact with the earth.

Overview measurements with a TDR

The following overview measurements should be carried out with a time domain reflectometer (TDR): comparative measurement - measured length compared with length on cable plan; joint/splice calibration; comparison of faulty and fault-free conductors; storage of reflectograms for future comparison.

Fault classification results

Short circuit - 0 Ω

A short circuit is a direct metallic connection between conductors, indicating that the conductors are touching one another or have fused together. This means that acoustic pinpointing techniques will fail. Because of the direct metallic contact, no acoustic noise will be produced at the fault. On the other hand, a short circuit can be very easily seen using classic TDR reflection measurements. Nevertheless, an attempt should be made to change the fault to higher resistance, using a high energy surge wave generator, so that acoustic pinpointing can be undertaken.

Low-resistance fault greater than 0 Ω

Low-resistance faults make it impossible to charge the cable. These faults are, however, visible with almost any surge wave generator-based, high-voltage pre-location or pinpointing method.

Very high resistance faults

The resistance of many of these faults is so high that ignition is not possible using the normal voltage of a surge wave generator. These faults can be charged up to their flashover (breakdown) voltage. The entire energy stored in the cable capacitance is discharged via the fault. Decay and decay-plus pre-location methods can be used, as well as acoustic pinpointing.

Faults due to contact with earth

Pre-location can be carried out using the bridge method and/or the voltage drop method.

Which faults can be seen with a reflectometer?

The faults that can be seen with a reflectometer are: all impedance changes below the cable impedance - for example, parallel and series resistive faults; joint/splices; strong reflections caused by impedance changes; damaged areas and pressure points; water ingress; changes in cross-section; contact problems caused by corrosion.

These faults are not visible:

• Faults with resistance many times the impedance of the cable. In theory, these faults should be visible, but the change in the TDR trace is so small that it gets lost in the normal noise or disappears due to attenuation. With modern cables and correctly fitted joints/splices, the changes in impedance can be so small that they are not visible, especially in telecom cables.
• Faults that normally have a near infinite resistance, such as faults that behave like spark gaps. These faults are ignited by applying a DC or VLF voltage. The breakdown voltage depends on the distance between the conductors.
• Where a cable of unknown length has been severed, it is possible to confuse the end of the cable with the point at which it has been severed. When there is doubt, short-circuiting the far end of the cable will quickly confirm whether this type of fault is present.

In all cases, depending on the type of fault, the reflections can sometimes be so small that they are inconspicuous and, therefore, undetectable