SYNOPSIS: Electrical systems developed to allow power generated at a central large power station to be distributed to industry. Within each industry, a local distribution network is required to supply power to individual items of plant. These electrical systems are designed with a typical expected life of 25 years. However. failure of some parts of these networks is inevitable and for this reason electrical protection systems developed to minimise the damaged associated with such failures.

As the protection systems are themselves liable to failures, management of protection is necessary to obtain the most cost effective supply of power to an industrial plant.



Electrical Relay Protection consists of interconnected current transformers, voltage transformers, breakers, tripping batteries and a variety of relays whose main duty is to monitor the electrical supply system and trip the appropriate breaker/s which will isolate an electrical fault or system condition which would result in damage to plant.


An industry’s final profit will depend on the most efficient usage of the capital invested in factory or mine. This requires that the electrical power is effectively available for 100% of the plants operating time, however, no equipment can be guaranteed to be 100% reliable.

Protection systems cannot prevent faults, thus outages will or may occur. But the extent and the duration of the outage must be minimised, which can be accomplished by using adequate protection. As improvement of the protection installed will i9ncur a higher cost , management needs to assess the risk of failures and their cost against the cost of the protection which could be installed, including it’s LIFE CYCLE cost, for maintenance and replacement.

No perfect protection philosophy design exists as each industry has it’s own peculiar problems and each case must be judged on it’s own merits.

The philosophy of protection for a particular industry needs to be set out at the design stage of a project to ensure that the industries requirements are covered and that expensive modifications are not required later, typically:-

• the size of motors being started DOL relative to transformer size,
• the number of breakers in series in a radial system.
• the fault clearing capability of breakers.
• the number of redundant transformers of feeders
• allowance on rating and space for reasonable future expansion of the system


Any fault which occurs, will generally result in some damage, e.g. ionised air at the point of fault burns insulation and causes explosions, excess current creates thermal stresses, system overvoltages create stress on the insulation.

Whereas in the electrical distribution system, where overhead lines are a major component of such systems with faults on them being usually transient and causing minimal damage, most faults in electrical equipment of industrial systems are usually damaged by such faults. Besides transformers and switchgear, industrial systems consists of cables and motors which are more easily damaged by faults.

A fault in industrial plant thus results in equipment down time and loss of production.

Adequate protection can reduce the duration of the fault and thus the amount of damage caused. This allows equipment to be repaired more easily and at a lower cost as well reducing the time to return the equipment to service and production.


A similar result of adequate protection is the reduction of the energy created at the point of fault and thus the possibility of explosions or expulsion of very hot gases. This will reduce the risk of injury to personnel in the immediate vicinity of a fault.

This aspect of protection has moral, social and legal implications for management, who could be held responsible under the OHS Act if insufficient protection is installed and installed protection is not correctly maintained because of inadequate management or staff training.


Even if protection removes a faulty piece of equipment rapidly, reliability of electrical supply can be impaired if healthy equipment is unnecessarily isolated at the same time.

Protection philosophy should ensure that the minimum of equipment is isolated when a fault occurs so that as much of the plant as possible remains in service after a fault is cleared.

As many industries have processes which are sequential and interdependent, loss of any component of the process may require the shutdown of the entire line. Thus at the design stage of the plant, redundancy of equipment and the associated electrical protection must be considered together to ensure the best overall performance.

In some cases, the simplest protection may be adequate as the entire process may have to shutdown for the loss of any section of the process. Here reliable high speed operation of protection may be the best solution to minimise damage and loss of production.


Even though protection is essential, the probability that it will ever be needed is low. This leads to protection being in service for many years before it is required to operate. However, when required to operate it must operate as per design. This requirement can only be ensured with equipment designed and tested to IEC 255 or equivalent standards. The use of less reliable equipment and inadequate initial installation and commissioning can result in unnecessary and expensive shutdowns. The protection equipment itself needs maintenance, especially the tripping batteries which are at the heart of the protection system.

Protection systems can be compared to Life Insurance. It always seems to be a cost which does not result in any tangible benefit until an accident / fault occurs. After such an accident / fault, the perceived need for protection is very high, but this perception diminishes as the inverse of the cube of the time since the last accident / fault.


Protection equipment needs to be correctly commissioned and subsequently maintained like any other plant. The importance of this is highlighted by the fact that CIGRE initiated an investigation, which resulted in Working Group 34-06 producing a report “ Maintenance and Management of Protection Systems”

Unfortunately, this commissioning and maintenance is often neglected, as the protection usually does not affect plant operation when equipment is in a healthy state. Only maltripping will cause the protection to be checked, thus inoperative protection may exist for many years before failure to operate results in checks being carried out.

Initial design of the protection should allow for ease of maintenance and maintainability. This involves the use of test blocks for injection during testing and maintenance, and the use of rack-in relays to allow for easy and rapid replacement of any faulty relay.

The decision on the choice of the relay manufacturer and the local supplier is important and should be based on:

A) long term product support especially with software in numeric relays

B) ability to provide rapid repair or replacement facilities

C) provision of information to assess reliability and LIFE CYCLE costs

D) provision of back up support to assess correct relay operation

The older types of electromechanical protection have given very reliable service even with minimal maintenance. However, tarnishing of contacts, corrosion of leads and deterioration of insulation does occur so that checks on Relays, CTs, VTs, Batteries, etc. should be carried out at typically two yearly intervals.

The more modern electronic relays have improved the available functionality and range capabilities of the protection systems. But at the same time these relays have introduced disadvantages, namely:-

a) higher DC loads on batteries and chargers

b) many more components which even with their higher individual reliability can have an overall lower Mean Time Before Failure (MTBF)

c) the use of electrolytic capacities in power supplies results in a need for systematic change / replacement of these components after about eight years

d) failure of a component leads to a higher loss of protection facilities as more functions are integrated into one relay unit

The above results in an increased maintenance requirement for electronic relays i.e. once yearly while the other components of the protection system and their maintenance have remained substantially the same (CTs, VTS etc.).

The currently designed NUMERICAL relays being offered, have attempted with various success to provide built in diagnostic / testing facilities. These facilities can reduce the need for regular maintenance, which in itself can lead to “finger trouble” when the maintenance is carried out by inadequately trained personnel and subsequently results in protection malfunctions or inadvertent trips.

These diagnostic systems in numeric relays cannot prevent protection failure. They must, however, be inherently failsafe in terms of the alarms that they provide. The alarms must be made available at a suitable point where management can be certain that action will be invoked to correct the problem.

Too often alarms have been considered spurious or unimportant and have been ignored or blocked so that protection has remained out of service for months or years – Chernobyl!! Adequate training of operating staff is thus very important.

Alarms should also be grouped in such a way that essential alarms are easily recognised compared to those used for information only. Flags and alarms for protection must not operate during normal plant start ups eg. undervoltage and underfrequency, so that they have to be reset when the start up is complete. Such flags and alarms are often left operated, leading to later confusion in analysing fault operation, or result in operators mistrusting the alarm system when incorrect flagging occurs during motor starting or other switching sequences.

Modern computer controlled testing equipment has improved the ability of personnel who test the protection systems, to repeat the actual tests in a consistent manner. The use of computer control allows fixed test routines and predefined control values to be used for tests and the results can be compared to previous results by the same program, highlighting only those results which are out of specification.

Use of this modern equipment can reduce the technical skills required of the maintenance personnel. However, such equipment is expensive and requires the initial use of well trained staff or contractors to set up such a comprehensive maintenance system.

The need for spare protection relays being held by the Plant’s stores has to be linked to:-

a) the cost of holding spares

b) the guaranteed holding of spares by the supplier (roll over of stock)

c) the future availability of the specific relay

d) the cost of extended down time to find an alternative relay and modify circuits

e) the time and cost to repair older relays


Protection relay settings should be calculated for the system conditions defined by the Plant Operation Manager and the ratings of the available equipment.

This information gathers together important data which needs to be readily available for future fault analysis, change of operating methods, modification of the plants electrical sytems and comparison to the latest recorded values obtained during maintenance.

This information has usually been recorded manually on paper and is often filed on a random basis which makes future access difficult.

Associated with this data are the schematic diagrams for the switchgear and installed relays and the maintenance manuals for relays and switchgear. For assisting the protection maintenance personnel and the operating staff in the plant, a functional operating diagram of the protection should be available. The functional diagrams should be available “AS BUILT” and may be supplied by the switchgear manufacturer as part of his group of schematic diagrams.

The schematic drawings need to be “AS BUILT” and MASTER DRAWINGS must be maintained as up-to-date drawings, which is part of the protection maintenance system. This leads to the problem that the switchgear manufacturer generally produces his drawings for the actual switchgear and does not include detailed information of interconnection for SCADA, metering, connection to field contacts and monitoring units. This detail is often spread out on numerous other documents, making an integrated set of drawings difficult to obtain and thus they are often not available.

Some of the modern relays have multiple relay outputs, which can be selected by means of a software matrix to allocate the internal protection functions which will operate each relay. The matrix selection is often perceived to be settings for the person doing the settings of the protection relays, to define. This is considered by the author to be incorrect as this matrix determines whether the relay can perform the tripping as required and defined in the functional specification. For this reason, the designer who lays out the schematic diagram has to check during the design stage how he is able to use the matrix to provide the tripping functions, which implies that the matrix is in fact part of the “wiring”. The Project Engineer’s representative should thus insist that the switchgear manufacturer includes the matrix setting as part of the schematic diagram, as the “wiring” is incomplete without this matrix setting information.

The matrix settings must also be included in the relay setting sheets or computer files so as to provide a complete list of setting details for the person applying the settings to the relays.

Without an integrated set of drawings, even a well trained technician will struggle to correctly carry out commissioning, fault finding and maintenance.

The management of data / drawings must thus start at the design stage where the most comprehensive drawing produced, usually by the switchgear manufacturer, must be linked in the supply contract to the requirement of producing a completely integrated and “as built “ set of drawings. The Responsible Person for the project needs to co-ordinate the final production of comprehensive drawings and an additional cost allowed for the contractor to include other detail in his drawings. This cost is actually small compared to later costs of extended plant down time when only inadequate drawings are available for fault finding and maintenance!

A further problem, which can be cost effectively resolved at the design stage, is that the tender documents must stipulate that the drawings are produced on a CAD system suitable to the Plants data system. Otherwise the manufacturer will hold the originals for a limited time period after which they will be archived or discarded. Reverse engineering drawings for which no comprehensive “AS BUILT” copies or “MASTERS” exist will be very expensive. Often, if the Plant Manager is lucky, at least one hand marked up copy exists on site. If this is lost or destroyed, fault finding and maintenance become extremely difficult or impossible within reasonable time periods.

Computers have made the possibility of keeping records much easier. Hard copies of manufacturer’s data can be relatively easily captured for computer records and can be stored on CD’s. Where possible, manufacturers should supply data in electronic format for archiving on CDs. Master copies of such data must be managed and kept safe. The fire, a few years back, in the Pretoria Municipal offices, which destroyed their drawing office records, highlights this problem.

Often the settings for protection are done by the Plant’s Project personnel or their Consultant / Contractor during commissioning. Their initial philosophy could differ from that of the Plant Manager and operating staff and the basis of the calculations may not even be defined in detail for any assessment by the future operating staff. Thus detailed information relating to such settings and equipment ratings must be obtained from the project personnel who acquire very specific knowledge during the commissioning of new plant. This information must not be lost to those who have to manage the plant for the next 20 – 30 years.

The difficulty in maintaining this data base is increased if the project personnel do not have a consistent way of presenting the data, especially when various contractors are held responsible for only their part of the plant.

As protection can be supplied from various manufacturers, even computer generated data and files may be incompatible, i.e. not even using the same soft ware which may result in plant maintenance personnel having to acquire various software packages and having to develop a basic proficiency in these various programs. Compatibility with “Windows” (a system standard??) can reduce this problem to some extent. It may be necessary to retain older versions of software as well if newer versions cannot pull in data for older versions, or if alternative CAD and word processing software become the norm.

The insistence that manufacturers produce drawings for a specific CAD package used by the Plant will allow drawings to be maintained more easily, if the plant itself does not change their CAD supplier.

An overall management system of keeping a record / index of all data needs to be installed so the finding of specific data is simplified and can be accessed from different starting information, e.g. :-

1) Panel No

2) Plant Item NO

3) Schematic Drawing No

4) Manufacturer’s equipment description (where used in plant)

5) etc.

The installation of such a data system will require much initial thought, but will make future maintenance and design modification easier as correct information will be more easily accessible.

An associated problem with information is the rapid change of software used in a relay. The relay manufacturer needs to have a well organised record system in order to track the software installed in a particular relay. The preferred minimum requirement is that the software version can be down loaded from the relay to assess that the software is a version without errors or bugs, according to the supplier’s information. Haphazard changing of soft ware in a relay has already led to incompatibility between settings, manufacturer’s literature and relay final operation. This incompatibility has a cost implication for all involved and can result in incorrect operation of protection.


Most medium sized plants are unable to employ staff specifically for maintenance of protection. Staff used for normal electrical maintenance must be used when required to carry out the necessary maintenance of protection equipment.

Traditional methods of testing would use highly trained staff using manual test injection gear to inject relays and prove operation. All results were (are) logged manually. These highly trained staff members are often not available in small plants.

As protection is often in the back ground and seldom thought about, as mal-operations should be minimal, the staff often do not have time or incentives to learn about complex relays and systems. Thus when a fault occurs and protection operates, the staff may not be able to rapidly assess the problem correctly.

Besides the maintenance personnel, the operating staff also need to have a working knowledge of the protection as there may be no maintenance personnel immediately available when the fault occurs. To minimise down time of plant, the operators must be able to assess the problem rapidly from the information available in order to return plant to service or call out the correct maintenance personnel to deal with more problematic failures. Where the “production at all costs” concept prevails, additional damage is often done by operating staff trying for force equipment back into service, as they may not understand the implications of resetting the protection relays and restarting equipment.

Where operating staff are not properly trained, they may also reset alarm information, which may be essential in assessing the actual cause of failure. This could allow a latent problem to re-occur and cause further loss of production and damage. Correct logging of all alarms is necessary to allow assessment of any plant failures.

Protection forms only part of the training required by operators and maintenance staff, but is considered a very important part if maximum availability of plant is to be ensured.

In South Africa, most of the protection training is made available by:-

i) Eskom to it’s own staff

ii) Relay Manufacturer’s as specialised or general courses

iii) Some of the larger Municipalities and Companies

iv) Private Protection companies and Consultants

When plant is initially installed, the plant personnel on site have the best chance of learning about the methods of operating the plant and it’s associated protection. Staff taken on at a later stage, often are unable to gain this type of “training”.

Much of the knowledge obtained by staff in smaller companies is by on job self training when problems occur. Smaller companies must either hire skilled staff or make use of training offered by the Manufacturers, which may unfortunately be biased towards their own products instead of being of a fundamental protection knowledge basis.

Management needs to assess the cost of training their staff and relate this to the cost of extended outages because of incorrect assessment of plant failures.


The cost of the protection installed on a plant can be relatively high.

When faults do occur, the operation of protection with regard to hardware and settings needs to be recorded and evaluated to ensure that correct operation of the protection has occurred, relative to the fault on the system.

The availability of reliable past history, setting data and protection function diagrams is now essential for an educated assessment to be made. This assessment will allow the correct decision to be made for follow up action, which may be necessary, eg. repeated failure of a particular manufacturer’s type of equipment may require total replacement.

The detailed history kept by “data” management will allow an objective decision to be made. Where an assessment is made on “remembered” details, a costly incorrect decision often results.

The older types of protection relays result in operation of alarms and flags being the only information available. This information needs to be recorded manually, but should then preferable be transferred to a data base for initial assessment and historical record keeping for possible future use.

The more modern electronic relays often record the levels of fault current, the function which operated, the level of voltage dips, etc., typically for up to the last five faults. These need to be extracted manually for further use. However, some of the modern relays include Communication facilities to computers and with the appropriate software, far more detailed records can be extracted.

The latest Numeric relays have options for far more detailed fault recording, including records of waveforms during a fault. These waveform recordings can be down loaded and used by modern injection test sets to provide a replay of the fault condition when attempting to assess protection operation or suspected malfunctions.

The above functions would probably be beyond the requirements of most smaller plant management requirements. However, these functions are becoming more standard in the newer relays, thus management needs to define what data should be retained for future use.

Even with numerical relays and user friendly software, the skill burden for plant personnel is increasing as each item of plant and associated protection has it’s own learning curve.

For this reason, manufacturers are designing communication protocols which allow their relays to be coupled to a centralised computer which automatically logs all data from the various relays. This can included real time data of loading (amps, volts watts) etc. and records of operations for faults.

The manufacturers are also experimenting with EXPERT systems, which are intended to sort incoming information and only output details of problems and mal-operations. This will become more and more necessary to reduce the work load on skilled staff.

As a standard protocol for the communication between relays and a centralised computer system has not yet been finalised, the manufacturers have had to develop their own protocols. This lack of a standard results in systems which cannot interchange information or use a common program without separate interface programs being written by a third party. For this reason a plant would need to have interfaces developed for the particular mix of protection employed in their system.

To minimise costs of an effective data capturing system, the designers need to limit the number of different types of relays and manufacturers used. As more standard protocols are developed and commercially available interfaces are written, this restriction may fall away.

With the increasing use of PLCs for distributed intelligent control around a centralised control room for many modern industrial plants, the integration of the relays’ communication systems with this process control system is possible. Many of the numerical relays allow control of the breaker, allow additional contact inputs to be relayed onwards and also supply data such as load current .

Combining process control and protection / data information can reduce costs by eliminating separate control relays and transducers. However, many protection personnel tend to be conservative and do not want other functions in the relays as the relay is the last line of defence for clearing faults and the relay’s reliability can be compromised when used for other purposes as well.


In a modern economy there is a very high reliance on electrical power, which needs to be as close to 100% in reliability.

Protection systems contribute to this high availability of electrical power by minimising the time of a fault remaining on the system, reducing fault damage and limiting the equipment taken out of service in order to remove the fault.

Although a perfect electrical system is not possible, the initial choice of relays, judicial use of maintenance of protection plus associated data and the training of personnel can reduce the Life Cycle cost of protection and thus improve the availability of electrical power.

The mix of older type relays with modern numeric relays means the total self diagnostics of all relays will not be available in the near future and some traditional maintenance procedures for protection will be necessary.

Email Us

3 + 6 =

Contact Us

Physical address: 10 Pieter St, Highveld Park, Centurion
Postal Adress: P.O.Box 68299, Highveld Park, Centurion, 0169

+27 12 665 0545