Relays: Plug setting; Microprocessor vs electronic vs electro-mechanical/traditional [Industrial Electrical Power Systems]

Home | Sitemap/Articles

AMAZON multi-meters discounts AMAZON oscilloscope discounts

Plug setting--factors influencing choice

1. Load conditions: Must not trip for healthy conditions, i.e. full load and permissible overloads, re-energization and starting surges

2. Load current redistribution after tripping

3. Fault currents: High fault currents can cause saturation of CTs. Choice of CT ratio is important

4. CT performance: Magnetization curve. Its internal resistance

5. Relay burden: Increases at lower taps on electromechanical relays

6. Relay accuracy: Better at top end of curve. Attempt to use in tight grading applications.

The new era in protection -- microprocessor vs electronic vs traditional (electro-mech)


Electromechanical relays of various types have been available from the earliest days of electrical power supply. Some of these early designs have been improved over the years. One of the most successful types of electromechanical protection relays has been the previously discussed inverse definite minimum time (IDMT) overcurrent relay based on the induction disk. With the introduction of electronic devices such as the transistor in the 1950s, electronic protection relays were introduced in the 1960s and 1970s. Since then, the development of relays has been related to the general development of electronics.

AMAZON multi-meters discounts AMAZON oscilloscope discounts

By the late 1960s, extensive experience in the use of electronics in simple protection systems enabled the development of many quite advanced protection schemes and the first high-voltage substations were equipped with static protective relays. Over a period, these have been extended to cover other equipments such as transmission lines, motors, capacitors and generators. New measuring techniques have been introduced, measurements that are more accurate can be performed and high overall quality, reliability and performance of the protection system for high-voltage power systems have been reached.

AMAZON multi-meters discounts AMAZON oscilloscope discounts

Developments in the 1970s concentrated on improving reliability through improved design of printed circuit boards leading to integrated circuits and general improvements in substation designs, particularly grounding. In general, most static protective relays of that time were designed to match or improve on the basic electromechanical performance features. Improvements introduced included low-current transformer burden, improved setting accuracy and repeatability as well as improved speed. Also, during this period, experiments were conducted in Europe, Japan and the USA to test computer-based protection systems based on the availability of digital electronics.

This is particularly true with IDMT overcurrent relays, where it was both difficult and expensive to provide the inverse time characteristics by means of analog electronic circuits. However with the advent of microprocessors, it’s much easier to provide the most commonly used characteristics such as definite time, normal inverse, very inverse, extremely inverse and thermal characteristics using different algorithms stored in the microprocessor's memory.

The overcurrent relay is undoubtedly the most common type of protection relay used by electricity supply authorities for protection on distribution systems. This section concentrates on the various features of modern static overcurrent protection relays in relation to the older electromechanical relays, which are still commonly used on distribution systems today. The purpose is to clarify some of the arguments for and against static protection relays, particularly for medium-voltage applications.

What is a static protection relay?

Static relays are those in which the designed response is developed by electronic or magnetic means without mechanical motion. This means, that the designation 'static relay' covers the electronic relays of both the analog and digital designs. The analog relays refer to electronic circuits with discrete devices like transistors, diodes, etc., which were adopted in the initial stages. However, the digital designs incorporate integrated chips, microprocessors, etc., which had been developed subsequently. In recent years, very few relays of the analog type are being developed or introduced for the first time.

Most modern overcurrent relays are of the digital type. There are many reasons for this, the main ones being associated with cost, accuracy, flexibility, reliability, size, auxiliary power drain, etc. Many of these reasons will become evident during the course of this section, which will concentrate on relays of the digital type. Microprocessor relays are of the digital type.

The main objective of using static relays is to improve the sensitivity, speed and reliability of a protection system by removing the delicate mechanical parts that can be subject to wear due to vibration, dust and corrosion. During the early development of static relays, the use of static components were particularly attractive for the more complicated relays such as impedance relays, directional relays, voltage regulating relays, etc. On the other hand, the early static IDMT overcurrent relays were expensive because it was difficult to match the inverse time characteristic using analog protection circuits.

The battery drain associated with these static IDMT relays was too high and this discouraged the use of this type of relay for medium-voltage applications. The general developments in the field of electronics and the introduction of digital circuits have overcome many of the above problems. Using modern microprocessor relays, almost any characteristic is possible and economical, even for the simplest applications such as, overcurrent relays and motor protection relays.

What is a microprocessor relay?

A microprocessor relay is a digital electronic relay, which derives its characteristics by means of a pre-programmed series of instructions and calculations (algorithms), based on the selected settings and the measured current and/or voltage signals. For example the formula used to derive the inverse time characteristics in an overcurrent relay that comply with IEC 255 and IEC 142 is mathematically defined as follows:

Where: T = operating time in seconds, K = time multiplier, I = current value, I> = set current value.

The unit includes four IEC 142 specified characteristics with different degrees of inverse. The degree of inverse is determined by the values of the constants a and ß.

Degree of Inversity of the Characteristic a ß Normal inverse Very inverse Extremely inverse Long-time inverse

A description of a typical microprocessor (or numerical) relay follows, which includes:

  • • Introduction
  • • The simplified block diagram
  • • Handling of the energizing signal
  • • The microprocessor circuits
  • • The output stages
  • • The self-supervision.

Introduction to the numerical relay

The measurement principle is based on sampling of the energized currents or voltages, analog to digital conversion and numerical handling, where all settings are made in direct numerical form in a non-volatile memory. Setting can be performed either manually on the relay front or by serial communications using either a personal computer or a control/monitoring system. In addition, the operation of the self-supervision is described.

++++ A typical microprocessor relay

++++ Rear view of a microprocessor relay.

Simplified block diagram is shown.

++++ A typical simplified block diagram for a relay

The operation of a protective relay can well be described by using a simplified block diagram as shown above. Here we can recognize the input signal path with the signal processing parts, the output circuits for trip and signal, and the self-supervision circuits.

Handling of the energizing signal

The basic function of the relay is to measure the input and assess its condition. A digital relay comprises of sensitive devices and hence it’s necessary that they don’t fail because of the input changes. This is taken care by the isolating transformer and the limiter used in the relay.

The energizing currents or voltages are brought into the relay using the internal matching transformer (1). After the transformer there is a voltage limiter (2), which cuts the voltages entering the relay at a safe level, should there be an extremely high-current or voltage input. The reasons for this limiting are only to protect the internal circuits of the relay from being destroyed by too high an input voltage. Together with the limiting circuit, a filter can also often be implemented. In this case, the reason is that the harmonics contents of the energizing signal is not wanted and is filtered out. A typical filtering level here is that the third harmonics are attenuated by a factor of 10 and the fifth harmonics or higher are attenuated by factors of 100 and more.

The next stage is to measure the signals, which are to be monitored. In an AC circuit, the voltage, current and power undergo changes in relation to the supply frequency. The multiplexer (3) selects the signals that are to be measured. A sample of each signal is measured once per ms. If the measuring module is of type C, with setting knobs, the setting values are also read through the multiplexer. Finally, the multiplexer also selects a reference channel once per second to have a condition check for the input circuits.

The analog-to-digital-converter (A/D) (5) measures the level of the measured samples and transforms the analog values into a numerical form. The resolution of the A/D converter is 8 bits, representing numerical values 0-255 (20 -1 to 28 - 1). Because the dynamic range of the signal levels to be measured is quite high, the 8-bit conversion is as such not enough to give a good accuracy over the whole current or voltage span. Therefore, a programmable attenuator (4) is needed between the multiplexer and the A/D converter to enable an accurate handling for both low and high current or voltage levels.

For low signals, the amplification is 1, passing the signal directly to the converter. For phase current measurements, e.g. at a current level of 1.00 × In, the numerical value from the A/D is 100. When the signal is too high to be handled directly, i.e. for values above 200-255, an attenuation of ×5 is put on the signal. Furthermore, if the signal is still too high, the attenuation is switched up to ×25. This means that the overall range of measuring capability is 0-6375, corresponding to more than a 12-bit conversion. When the numerical value 100 represents a current of 1.00 × In, the highest measurable current is thus 63 × In.

The sampling rate is typically 1 ms, which means that every half-period of the 50 Hz sine wave is measured by ten samples. This gives a very accurate measure of the peak value during the half-period. For the worst case of the two top samples hitting evenly on both sides of the exact peak, a maximum error could theoretically be about 1%. As all signal handling is based on the mean value of two consecutive half-waves, and the sampling is non-synchronized, the probable theoretical fault is less than 0.5% and is partially compensated in the unit calibration.

As the measured current now is available in a numerical form, several things can be made. For example, the problem with handling of a signal with a DC-component is now very straightforward. When the mean value of two consecutive half-waves is calculated, the DC-component is eliminated to almost 100% without any need for non-linear air gap transformers or similar components. On the other hand, the calculation of the mean value consumes 10 ms, which is not wanted for very high short-circuit current levels where an instantaneous trip is called for. For this case another trip criteria is simply added. If the current in the first half-wave exceeds twice the setting, it’s obvious that the mean value of the two half-waves will exceed the set level and therefore a trip can be carried out instantaneously without the need to wait for the next half-wave.

++++ Sampling of the sine wave at 1 ms intervals

++++ Elimination of the DC-component by mean value calculation of two consecutive half-waves.

All measured numerical values can of course easily be transferred over the serial communication, be stored in memory banks, etc: for later retrieval e.g. when fault reasons are being investigated. The sampling is also used in another good way i.e. to minimize the transient over-reach. When the operating time for a stage has elapsed and the trip order is to be carried out, the stage will wait for still one single sample exceeding the set level before the trip is linked to the output relay. In this triggered state, the relay will wait for a short time and if no further samples are detected, the relay will reset. This means that the retardation time or the transient over-reach is very short, less than 30 ms.

The microprocessor circuits

In the entire signal handling stages, the microprocessor or more accurately the microcontroller (6) of course takes part as every operation is controlled by this component. Furthermore, all protective decisions are made here, the operating time is counted for every stage and after that, the outputs are linked to the output relays.

The display used for the MMI is also served by the controller.

External components to the controller which are not shown in the figure are also the three different memory components that are used:

1. The random access memory (RAM), which is used as a scratchpad for measuring and calculating results, storing of memorized values, etc.

2. The read-only memory (ROM), which contains the program firmware for the module.

3. The electrically erasable programmable read-only memory (EEPROM), which is being used as parameter storage memory, e.g. for all the setting values.

The output stages

The outputs from the module is linked from the microcontroller, via a power buffer amplifier (8) to the electromechanical output relays (9) on the output card.

To prevent a disturbance condition from causing a false output, a double signal arrangement is used. This means that an enable signal (ENA) must be sent together with the output command before any output relay can be activated. Furthermore, the outputs can also be inhibited by self-supervision. The mechanical operating times of the output relays, typically about 10-15 ms are subtracted from the operate times for the definite time protective stages. Thus, very accurate timings can be achieved for these operations.

For IDMT operations, this is however not possible as you cannot decide when you have reached the trip instant - 10 ms for all possible variations of current.


In order to avoid false operations due to relay internal faults and to maximize the overall availability of the protection, a set of auto-diagnostic circuit arrangements have been implemented in the relay modules. Generally, all tests are performed during a period of 6 min.

The different memory circuits, i.e. the RAM, the ROM and the EEPROM are all continuously tested by different methods at a speed of about one byte per 10 ms. Thus all memory is cyclically checked out.

The microprocessor and the program execution are supervised by a watchdog (12) once every 5 ms. The multiplexer, the switchable attenuator and the A/D converter are tested by measuring of a very accurate reference voltage once a minute and always before tripping. This is to ensure that a measured signal is real, and not caused by a fault or disturbance in some input circuit, all to avoid false outputs.

The settings are tested once a minute by the use of a checksum arrangement. It’s obvious that the settings must be ensured under all conditions of power breakdown, etc.

Any kind of battery backup must also be avoided as all batteries have too short a lifetime for this purpose. A relay is designed to have a lifetime of 20 years but the best life batteries cannot survive much more than about 5 years.

The solution is to have all settings stored in a non-volatile memory (EEPROM) and in two identical subsets. For both the setting memory areas, the microcontroller automatically keeps track of the exclusive or checksum for the whole block. This checksum is stored in the last memory place of the block and is used as a reference of the correctness of the contents in the memory block. If a fault is detected in one block, the controller will directly check the other one and replace the contents of the faulty block with the contents of the correct one. This means that the settings are self correcting, for e.g. ageing/refreshment faults due to the properties of the EEPROM. Normally a refresh cycle should be carried out every 10 years, in this way the relay module makes the refreshment when needed.

++++. The two memory areas allocated for keeping the relay settings secured

The internal supply voltages from the power supply module are all tested once a minute by measurement of the different supply voltages, +8 V, +12 V and +24 V. This test is performed by the voltage limit detector (13). The trip output path, the output amplifier and the output relay are checked once a minute by injecting a 40 ms voltage pulse into the circuit and checking (10) that a current flows through the trip circuit.

If the self-supervision detects a fault, the output buffer amplifier (8) is immediately blocked to ensure that no false signals are carried out due to the fault condition. After this, the watch-dog tries to get the microcontroller to work properly again, by resetting the process three times. If this attempt is not successful, a signal about the internal relay fault (IRF) is sent after a time delay (14) linked to the output relay (15). Furthermore information about the fault is sent as an event over the serial communication RxTx and a red LED on the front of the module is activated. If the module is still in an operative condition, also an indication about the character of the fault is shown as a code number in the front display.

Even a full breakdown of the relay, e.g. by loss of power supply will be detected as the IRF relay (15) operates in a fail-safe mode, causing a signal when the relay drops off. In addition, the serial communication will indicate loss of contact to the module and later on the module goes into a reset condition.

Top of Page PREV: Construction / operation electro-mechanical IDMTL relay NEXT: Universal microprocessor overcurrent relay Index

Wednesday, January 16, 2013 0:10