Fault finding
Image credit: Stock

Problems in medium voltage grids are usually handled by redundancies within the network setup, along with the implementation of appropriate switching measures, which generally ensure a relatively uninterrupted continuation of the supply.

This article was originally published in Smart Energy International issue 2-2020. Read the full digimag here or subscribe to receive a print copy here.

A lengthier wait until a fault can be corrected mostly results only in an increased risk due to the possibility of consequentially occurring disturbances. In low voltage networks, which usually have no redundant power supply, the time intervals until a customer can be reconnected to the power supply greatly depend on the speed in locating the fault. Longer waiting periods due to a more remote fault location system are very problematic.

However, a low voltage installation also has advantages. The distances are relatively short and easily manageable. In many cases, the joint position can also be very clearly delineated based on the known positions of the private connections. Since 80–90% of cable faults occur in joints, an anticipatory localisation of the fault is possible.

Transient faults

  • Transient Irregular, short-term voltage drops without fuse triggering
  • Intermittent Irregular triggering of fuses at longer intervals

Many LV cable faults change from transient to permanent (‘flickering lights’ are possible signs of a transient fault). Also, LV cable faults are often unstable/non-linear and can therefore only be located when the cable is conducting live voltage. Only once a fault has become permanent can it be located using conventional techniques with the cable in a voltage-free state.

All unstable LV cable faults require a change in their fault state to be located. The only way of doing this – with consuming devices connected – is to reconnect the mains voltage.

If the period between re-energising and the fault occurring is too long, the more efficient and easier method is re-energising via a mains fuse. If the intervals are shorter, and where there is sufficient space, automatic reconnection devices such as PowerFuse can be used to maintain the mains supply and effect a change in the fault state.

The technical problem of fault location in branched networks

Since, in order to locate high-resistance cable faults, DC and surge pulse voltages must be used, the private connection fuses must be removed. The problem of access to the service box is not always a given. A real problem in pre-locating faults on cables with many T-branches arises from the pronounced attenuation of the reflection measurement signals and the complexity of the reflectogram due to the jumps in impedance on the joints and branchings. Often, faults arising after the third or fourth T-joint are no longer recognisable due to these effects. Even more challenging is the situation with faults in branch joints, since these create a strong self-generated reflection themselves. Even the tried and tested arc reflection method (ARM) is equally affected by these limitations.

In view of this, even experienced technicians today must often locate the fault by measuring various endpoints of the branched cable. In some instances, the cable is actually cut in order to limit the test stretch.

Basic principles

The use of T-branches in low voltage networks makes the evaluation of reflectograms considerably more difficult. Only through comparative measurements of fault-free and defective wires can evaluable results be achieved. By utilising a Telefl ex reflectometer, the test pulses are partially reflected on the T-joint with a negative algebraic sign while test pulses that continue are simultaneously reduced in amplitude. The amount of reflection depends on the impedance in the mainline and the continuing line. The T-joint is, according to the theory of transmission lines, a parallel switching of the impedance of two conductors.

Figure 1: T-Branch

With identical impedance of both continuing lines, Z is reduced by 50% at the T-joint.

However, this seldom occurs in practice.

As a rule, the mainline has a larger crosssection than the secondary line and thus a different impedance.

The reflection factor ‘r’ can be derived using the following equation:

The result shows that with identical impedance, 33% of the test pulse is reflected with a negative algebraic sign, and 33% of the pulse continues in each of the two continuing lines.

Due to the different cross-sections of the main and secondary lines and, therefore, varying impedance, the reflections at the T-branches are principally between 10% and 30%.

Actual test results show that in singly branching networks, positive results can still be achieved even after 10 T-joints. In networks with multiple branchings, however, the situation is more complicated.

This diagram shows a branched LV network with 12 T-branchings and two connecting joints.

The end of the red line corresponds to the distance of the fault, which could be located on the straight line or in the T-branch.

Intermittent fault location.

Intermittent faults are very difficult to locate.

Due to numerous joints and connections, these faults frequently occur in low voltage and street lighting networks. Corrosion in lamp masts and poor connections in joints lead to these faults.

Megger’s Digiflex Com and Teleflex MX are equipped with ‘IFL mode’ and both devices perform continuous measurements and record them.

Every impedance change, short-circuit and interruption is automatically saved and represented in a reference curve.

Advantages of the IFL mode:

  • No time synchronisation required, every change is automatically recorded.
  • The operator can perform the measurement unassisted and determine the end of a line. The reflectometer can be connected with the faulty cable over a lengthier period. All events are graphically shown.
  • In ‘Difference mode’, even small changes in the impedance are visible.
  • No triggering device is required.
  • No high voltage is required.

Measuring low voltage networks while under voltage

Separation filters (400V) enable the direct connection of a reflectometer on the low voltage network while under voltage.

The measurement should always be performed from the cable end or service box in the direction of the supplying station. Transformers, switchgear and distribution boxes generate large reflections, which overlap the measurement signals. In addition, the test pulse travels through all outgoing lines and also receives multiple returning signals from them. This greatly increases the difficulty in evaluating the actual fault reflections. A measurement from the end that is distant from the distribution generally has only a defined direction of diffusion.

In some countries, such live measurements are used to detect illegal consumers. The requirement for this is a comparative measurement with previously recorded reference patterns. A measurement made in the immediate vicinity of the service box and the meter contains so many reflections that the detection of additional lines and consumers is only possible via a comparative measurement. Such measurement requires, however, the observance of certain safety criteria, e.g. of connection lines. These safety criteria are described below.

Safety for measurements made on live networks

Measurement circuits are subject to load due to the operating voltage and the transient loads of the electrical system to which they are connected during the measurement.

The use of measuring devices on live networks requires certain constructional safety measures and a corresponding label. These are defined by the engineering standard VDE 0411 / IEC 61010 and divided into the categories of CAT 1 to CAT 4.

The defining element here is the hazard arising from surges and peak voltages in the corresponding CAT range. The insulation in the device and the corresponding measuring lines must reliably insulate these voltages.

In the case of an arc igniting due to a surge, several thousand amperes could be generated, depending on the connection zone, before the upstream safety elements trigger.

Category IV Three-phase connection to the LV voltage source as well as the low voltage overhead lines: suitable for measurements at the source of the low voltage installation.

Examples are meters and measurements on primary overcurrent protection devices and ripple control devices.

Category III Three-phase distributions as well as public/industrial single-phase lighting systems: suitable for measurements in the building installation. Examples are measurements on distributors, circuit breakers and cables.

Railway distributors, distribution boxes, switches, sockets of the fixed installation, devices for industrial use and other equipment as well as fixedly installed motors.

Category II Single-phase plug-operated applications: suitable for measurements on circuits that are directly linked electrically with the low voltage network. Examples are measurements on household devices, portable tools and similar devices.

Category I Electronic system: suitable for measurements on circuits that are not directly linked with the network. Examples are measurements on circuits, which are specially protected circuits that are diverted from the network.

ARM method

High-resistance cable faults in low voltage networks can be localised with the arc reflection method. This method requires both OK and fault patterns for evaluation.

The disadvantages in branched networks, such as attenuation of the test pulses at the T-joints and additional reflections from the cable ends, also apply to this method of pre-location. The cable must be disconnected, and all fuses must be removed from the service box. If the cable fault is located after several T-joints, the difference between the OK and the fault pattern is very slight. By enlarging the pulse width, the difference can be made more noticeable. The fault location system should be implemented as needed.

In principle, practically every measuring system can be employed with the ARM method. The lowest available surge voltage level presents a restriction.

Three to maximally 4 kV is optimal, whereby ‘less is more’!

If, for example, the voltage in an 8 kV level needs to be reduced to 3 kV, only limited surge energy remains available. (W = 0.5 x C x UÇ).

For pre-location, this is not as important, but for pinpointing at least 300–500 J should be available.

EZ Thump

A compact, practical system for this application is the EZ Thump, which offers a complete fault location system.

The EZ Thump has a 4 kV level (alternatively, 12 kV as well), which is used for testing, breakdown detection, pre-location and pinpointing.

An automatic routine allows the locating of faults with ‘hardly any’ expertise with the equipment. The system guides the operator automatically through the diverse applications, detects the situation in the test object and informs the user accordingly.

The test results are given directly as alphanumeric values on the display.

Teleflex LV Monitor – online reflectometer monitoring

The Teleflex LV Monitor serves in locating all faults on low voltage networks, but particularly intermittent faults. In doing so, the LV Monitor operates under-voltage without shutting off the consuming devices.

In the reflectometer/TDR mode, a conventional reflection measurement is performed with a device. The same principles of transmission line theory apply here as in classical reflection measurement.

In contrast to regular reflectometer fault location devices, the Telefl ex LV Monitor is connected online simultaneously with all three phases of an operating low voltage cable and enables the operator to perform the reflection measurement either locally or remotely on any phase combination.

The Teleflex LV Monitor is supplied with power through a line with a three-phase connection in which at least one phase must be under voltage.

With the Telefl ex LV Monitor, the measurement is always made on the cable which is under voltage. After setting all the basic parameters such as amplification, pulse width, measuring range and selection of the defective wire, the LV Monitor sends test pulses continuously to the defective cable. In the case of voltage drops or triggering of fuses, 64 reflectograms are recorded chronologically around the event. Fault location is done by comparing the measurement before the event (OK) and during the event (fault pattern). Since the measurement is performed on cables under voltage, the time intervals for the measurements between ‘OK’ and ‘fault pattern’ must be very brief, since otherwise, the switching on of powerful consuming devices (short-circuit for reflectometer pulse) would lead to false interpretations. The type of voltage and the progress of the fault of individual phases as well as the current curve can be seen during the event time window and can be included for evaluation.

Supplementary methods:

Powerfuse

The Powerfuse serves as an automatic backup fuse and is used in pre-locating intermittent faults in low voltage networks with connected consumers.

Low voltage networks are largely protected with NH fuse elements. If the fuse drops out due to an insulation fault, the customer is disconnected from the network. Reconnection requires manual replacement of the defective NH fuse.

Particularly with intermittent faults, the fuse is tripped in irregular intervals and replacing it requires a greater amount of work.

With the aid of the Powerfuse, the respective cable section is automatically switched back on.

In connection with a reflectometer, cable prelocation per OK/fault pattern can be done at the same time.

After switching on 9 times within 5 minutes, the device shuts down. Tripping current can be set incrementally from 125 to 315A. With faults in IT networks, control lines or, for example, signal lines in railways, the term used for this is short-to-ground rather than a fault.

IT networks are specially protected networks, which are designed so that contact with a voltage-conducting line is harmless (hospitals) and that in the event of a short to the earthing, no current flows (explosion protection).

Especially in industrial systems in which the cables are nearly always in an environment with good electrical conductance, short-circuits are one of the greatest potential hazards faced.

Normally in an IT network, a short-to-ground initially does not trip any fuses and thus does not interrupt any processes.

However, the short causes the formerly unearthed, potential-free system to set itself to the earthing potential which was created by the fault.

As a consequence, the unaffected phases take up a defined potential against earthing.

An additional short of a different phase (double short-to-earth) can now cause a true short-circuit and thus lead to the total failure of the power supply. This could then, for example, halt critical manufacturing processes or create an arc due to the high current flow, which actually poses the greatest danger in an explosion-protected environment.

Such installations have insulation or a short-to-earth monitoring system, which displays this state in the event of a short-to-earth, thus warning the operator.

The operator can thus localise and resolve this short-to-earth as quickly as possible in order to restore the operating safety of the system.

Neutral conductor break – impedance measurement

The N conductor is the most important conductor in the network since it is required by all phases. Flickering lights can be an indication of a neutral conductor break.

Because of a higher phase voltage, damage to users cannot be excluded. The most frequent neutral conductor breaks occur in joints.

Corrosion from humidity at the terminals, incorrect assembly and exterior mechanical damage from civil engineering work can be the catalysts and reasons for this fault. The neutral conductor break is a power disruption and leads to a high imbalance in the network.

Depending on the type of this fault, such as contact to earth, the following test methods for pre-location can be applied:

  • Impulse reflection method (impedance change)
  • sheath fault location for contact to earth
  • Impedance measurement

For more information on these and other methods of fault location on low voltage networks, contact Megger: www.megger.com