Digital, real-time communication technology has provided a platform for the development of a fully numerical, decentralized busbar and breaker failure protection system. Dubbed REB 500, it is based on the concept of remote, bay-oriented data acquisition and preprocessing and is suitable for all kinds and sizes of substations. REB 500 offers distributed func tionality plus enhanced reliability and performance. The decentralized arrangement of the protection saves space and reduces wiring, making it easier to upgrade or extend substations. With their enhanced capability, the reduced hardware, and the many other advantages of numerical tech nology, protection schemes based on REB 500 offer both technical and cost benefits over their lifetime. Experience with protection systems al ready operating in the field has been very positive and confirms the flexi bility and exceptional performance of REB 500.

T

 he architecture of the numerical bus bar protection (NBP) system provides greater flexibility than conventional bus bar protection schemes. Commissioning and maintenance therefore require less time and are more user-friendly than in the past. 

Additional features of the numerical system are breaker-failure protection, disturbance recording and diverse bay and back-up protection functions. The bay-oriented functions work autonomously even when the central unit is blocked.

The system can be modified or ex tended simply by adding extra bay units and CPU capability. Using the man-ma chine interface, the new configuration and settings can then be transferred to the system and stored.

Busbar protection

The busbar protection has to satisfy the highest requirements with regard to re liability, security, selectivity, speed, measurement and versatility, under con sideration of all possible busbar switching conditions.

The disconnection of a busbar which is in operation can have a strong effect on the stability of a power system, and in the worst case even cause total failure of the power supply to a complete region. Short-circuits on a busbar can moreover cause substantial damage to the installed equipment, since the fault currents can reach values of several ten thousand amperes, depending on the source of the supply and actual circum stances. For this reason, busbar protec tion systems have to provide a much higher level of security and work at a much faster speed than other types of protection.

Busbar protection systems The operation and zone selectivity of the busbar protection are based on compari son of the current in all incoming and out going feeders of a busbar.

This principle of measurement has been used successfully for many years, and systems employing it are referred to as current differential protection or cur rent balance protection. It can be em ployed in low- as well as high-impedance differential protection. Both have their justification, and specific technical char acteristics allow optimized applications in each case. Other factors influencing the choice of protection can be national or traditional practices, and the configur ation of the primary system being pro tected.

Security is often increased by choos ing two different measurement prin ciples, each with its own independent criteria. A ‘two out of two’ basis is also a guarantee of safe and selective pro tection when a numerical platform is used.

High-impedance protection The differential protection arrangement has a positive effect on the system stabil ity due to the high-ohm resistor at the input. The voltage drop across this resis tor is used in the evaluation.

In high-voltage applications where the highest priority is given to security, the usual practice is to install an additional high-impedance protection system and to connect it to the protection circuit as a so-called ‘check zone’. This provides a second trip criterion, but also requires an additional set of suitable current trans former (CT) cores as well as hard-wired connections for each feeder. The CT cores have to meet the following require ments: 

 • They must be dedicated, ie not shared with other protection relays. 

 • They must have identical CT trans formation ratios. 

 • They must have a low-resistance sec ondary winding. 

 • The magnetizing current must be low. In general, the CTs have to comply with BS 3938, class X, or the new class TPS stipulated by IEC. It is a tradition in some countries to use high-impedance protection for double busbars. Unfortunately, however, to provide selectivity for every possible busbar configuration, the secondary CT circuits have to be switched via the auxiliary contacts of the disconnector (and/or additional interposing relays). Switching has to take place very fast, as under certain circumstances the CT core could become open-circuited and subsequently destroyed due to the high voltage across the secondary wind ing.

Also, since a relatively high voltage (several kV) lies across the high imped ance when an internal busbar fault occurs, the high-impedance relay itself has to be protected using external volt age dependable resistors (VDR) and short-circuit contacts.

Thus, in addition to a high-impedance protection system requiring a consider able investment in time for maintenance and project engineering, a substantial amount of additional equipment also has to be installed. All in all, it cannot be described as user-friendly, although the initial hardware costs appear to be low. Another factor is the cost of the addi tional CT core sets, which has to be in cluded in the total cost calculation for a specific protection system.

Low-impedance protection This type of differential protection is the preferred type for single and multiple bus bars, with or without a transfer bus. The main reasons for this are: 

 • No special CT cores are required. 

 • The CT transformation ratio can be dif ferent for each feeder. It is normally sufficient to use 30-VA protection cores designed to Standard 5P20. 

 • The CT core can be shared with other protection relays. 

 • The secondary circuits are never switched, ie they cannot be opened. 

 • The CT wiring is monitored internally by means of differential current alarm measurements. 

 • The disconnector positions are moni tored internally to check their plausibil ity.

The low-impedance ABB busbar protec tion systems (INX5 and the new REB 500) employ advanced, proven measurement principles featuring a stabilized differential current measurement and a directional phase comparison evaluation (each per formed individually per phase). By com bining these two measurements in the tripping logic, a very high level of security is achieved without having to duplicate or add check zones.

The application of low-impedance busbar protection systems is com paratively simple, since all of the modules required for the protection are integrated in the system together with clearly de fined interfaces for the connection to the substation, and the system as a whole can be fully tested in the factory prior to delivery. The result is a substantial reduc tion in secondary costs, eg for the over all substation planning, cubicle wiring, documentation and commissioning and maintenance.

Another advantage is that system monitoring is simple, with either an inte grated automatic test facility or, as in the REB 500 numerical system, internal supervisory algorithms being used. Other possible features include dis turbance recording, breaker-failure pro tection and earth-fault protection, plus overcurrent protection as a back-up func tion.

Application of numerical protection principles Many power plants and substations are currently undergoing modernization. Hard-wired connections are being re placed by fiber-optic links, while person al computers are being installed to allow operators to automatically monitor and control the substations. Information about the installation and its operation is transmitted over signal links to different levels in the substations or to the national dispatching center. The result is more economic operations control and man agement plus lower life cycle costs for the installation. The REB 500 numerical busbar pro tection supports the extension of primary systems by reducing the cabling and dis tributing the functionality, thereby minim izing the changes that need to be made to the hardware. Station downtime is also reduced.

REB 500 decentralized numerical busbar protectio

System structure Depending upon the user requirements, the REB 500 numerical busbar protection can be either bay-oriented (ie decentra lized) or installed in the conventional way (in a central location). The latter is mainly used for retrofit projects [1, 2].

The decentralized installation of the bay units (BUs) in the bay control and protection cubicle close to the primary equipment results in short connections to the high-voltage apparatus 1 . Each BU is connected via fiber-optic cables to the central unit.

The central unit collects data from the BUs via the process bus and runs the busbar protection algorithms. The bundles of copper wires used previously have been replaced by fiber-optic links, thus solving the problems that can be caused by electromagnetic interfer ence.

Bay unit The BU is the interface to the switchgear installed in a bay and acts as a data ac quisition unit, since all the information about the currents and voltages, as well as about the disconnector and circuit breaker positions, is provided here. In ad dition, it represents the point at which the power system is separated electrically from the control equipment.

Every BU has the current inputs required for the busbar protection scheme. Additional voltage inputs are available for disturbance recording. All the analogue input signals are filtered and converted to digital signals by the ana logue input and preprocessing (AIP) mod ules. The sampling rate of 2,400 Hz also allows the detection of saturated current signals.

The binary input/output modules (BIO) detect and process the positions of dis connectors and bus couplers, blocking signals, start signals for the breaker-fail ure protection, external resetting signals, etc. The BIOs operate in a wide range of auxiliary voltages.

The BUs convert the sampled current values into current phasors for each phase and transmit them to the central unit for further processing at intervals de fined by the protection algorithms. Short tripping times are ensured by fast, deter ministic communication between the BUs and the central unit as well as by the effi cient algorithms.

Central unit The central unit is responsible for the overall processing and the trip decisions. When it detects a busbar fault, it trans mits a trip signal to all the BUs in the re spective busbar zone.

The hardware structure of the REB 500 busbar protection system is shown in . Each of the units (AIP, BIO, CMP, CSP) has at least one microprocessor. This keeps the software module distribu tion flexible. The software is split into ap plication software, which provides the system functionality, and diagnostic modules, which are responsible for the start-up, self-supervision, event record ing, system blocking and shut-down functions. The central unit takes care of the system configuration, busbar replica, assignment of the bays within the system, operating parameters, manage ment of the process bus, and synchron ization and communication with the station control system. The busbar rep lica and the changes in the protection zones are adapted dynamically on the basis of the process data provided by the BUs. Protection algorithms To ensure secure, dependable protec tion, the measurements are carried out according to two different, independent measuring principles. The first measuring principle makes use of a stabilized differential current al gorithm. The currents are evaluated indi vidually for each of the phases and each section of the busbar protection zone. The differential current IDiff and the re straint current IRest have to fulfil the follow ing two conditions for the detection of an internal fault:

Fundamental frequency according to Fourier in busbar feeder n of line L Stabilizing factor

Limit of stabilizing factor

L Conductor m Number of busbar feeders n SD Busbar feeder Threshold for differential current The calculations take place in the central unit, which also puts the resulting deci sions into effect. The second measuring principle is a directional phase current comparison. This compares the phasors of the cur rents in all the feeders connected to a busbar section. In the event of an internal fault, all of the feeder currents have almost equal angles. If differences of about 180° exist be tween the feeder currents, the system is either in the normal operating mode or an external fault has occurred. If the minimum phase difference ∆ϕ between all combinations of feeder phase angles is smaller than the tripping angle for the phase comparison ∆ϕmin, the algo rithm will have detected an internal fault.  <∆ϕmin Empirically, ∆ϕmin = 74° has been shown to be the optimum threshold value for the tripping characteristic. This angle requires no station-specific settings. The evaluation of the measurements is shared between the BUs and the central.
Additional functions I 1 A I 2 0 I 1 I f max t a BU t BU CU t h t o I 2 Principle of the static maximum-value holding method Blue Primary current, feeder 2 Red Saturated secondary current I2 Green Saturated secondary current I2 after maximum value holding I1 Ifmax ta th Primary current, feeder 1 Max. current value in sampling window Rise time Holding time ms BU Bay unit t0 Operating time of the holding function unit. The BUs perform the preprocessing and all other bay-specific tasks. The algorithms process the complex current phasors obtained by digital filter ing and which are based only on the fun damental frequency. Current transformer saturation CU Central unit the Fourier filtering. 3 shows the signal correction principle for saturated current signals using the maximum-value holding method. This method works in that the maxi Additional functions provided by the sys tem besides the busbar protection func tion can be activated as required. I 2 Overcurrent release An additional tripping criterion can be in tegrated in each BU via the overcurrent release function. The criterion is parame terized on the MMI. Zero current measurement In power systems which are earthed to provide current limiting, the protection functions ‘current comparison with cur rent stabilization’ and ‘phase compari son’ do not always guarantee a correct calculation. In such cases, the zero cur rents are used in the evaluation. Trans mission faults during operation can cause a zero current to be simulated. Selectiv ity is ensured by a Fourier frequency analysis which supports the correct func tion of the zero current evaluation. Breaker-failure protection 3 One of the main requirements a busbar protection scheme has to fulfil is proper behaviour in the case of current trans former saturation. The NBP is largely in sensitive to CT saturation phenomena due to a special method applied prior to mum current value Ifmax is extracted from the sample window and held for a defined holding time th. The current value is held either at the maximum value Ifmax or a pro portional value v · Ifmax. The factor v de pends on the rise time ta, ie the time interval between the last zero crossing and the maximum value of Ifmax. The breaker-failure protection function is integrated in the BU and incorporated as an option in the NBP. The fast-acting overcurrent functions monitor the phase currents independently and are equipped with two timers. The first time step oper ates as a stand-alone function in each BU, ie independently of the central unit. The software logic in the BU is able to handle different breaker-failure configur ations. For example, once the delay of the first timer has expired, a tripping com mand can be applied to a second tripping coil on the circuit-breaker. A transfer trip ping signal is simultaneously transmitted to the station at the opposite end of the line. If the fault is still being measured at the end of the second time delay, the breaker-failure function uses the discon nector replica to trip all the other feeders in the same busbar section. Each breaker has one BU allocated to it in which all the measurements and positions of that breaker are stored.

Overcurrent-time protection B U S B A R P R O T E C T I O N External MMI This is a back-up protection function which is calculated locally in the BU and used to provide simple, autonomous feeder protection in each BU. Event recording All events are stored with a resolution of 1 ms in the central unit and in the BUs. System, protection and test events can be generated. A more comprehensive and convenient interface is provided at the second level by a personal computer 4 which is con nected via a fiber-optic link to the central unit or the BU. With this PC, the operator can configure, parameterize and check the entire busbar protection system. Station control Protection events refer to the pro tected busbar and include the changes in the disconnector and breaker positions, trips, etc. Test events are recorded during testing with the test signal generator. System events describe the diagnostic and fault messages of the NBP system as a whole. Disturbance recorder function The three conductor currents, the current in the directly earthed conductor, binary signals, and the voltages, can be rec orded in each BU. Recording can be started, for example, by a protection function or an input signal. The disturb ance recorder data are stored in the REB 500. They can be read out directly on the BU or at the central unit in COMTRADE format and stored on the PC. An evalu ation program (eg, WINEVE) is used for further processing. If the busbar protection has to be tied into a station control system (SCS) or station monitoring system (SMS), an op tional communication module is added to the central unit. At this third communi cation level the MMI functions can be per formed in the station control center. Security concept Self-supervision and diagnostics To ensure maximum security and avail ability, all the protection functions are continuously supervised [2, 3]. Spurious tripping has to be avoided under all cir cumstances. In the case of a fault in the NBP, the system is blocked, an alarm is given and a diagnostic event is generated for analysis. Important parts of the hardware, eg the processors, auxiliary supplies, A/D converters and memories, are tested in various ways when the system is being started and during operation. At the top of the hierarchy is the central unit master processor, at the bottom the inputs and outputs. Each diagnostic module con tinuously supervises the hardware on which it runs as well as the related appli cation software and all diagnostic mod ules at the next lower level. In the event of a critical fault, the entire system is blocked; when the fault is tran sient, the system is automatically de blocked as soon as it is over. The processing of the tripping com mands is also important for the security and reliability. Control of the tripping re lays has to be enabled periodically. Should this periodic function fail due to a fault in the software or hardware, all the tripping relays belonging to a particular zone are blocked and the failed function is signalled. Availability The use of tested and proven com ponents is essential if high reliability is to Bay unit of the REB 500 numerical busbar protection. Operation is via the built-in control unit or a personal computer. 4 Communication The numerical busbar protection is con trolled by means of a modern, user friendly operator interface. The man-ma chine communication takes place at three levels: Local control At this level, control takes place via the control unit incorporated in the BU or the central unit. This has LEDs for alarms, tripping commands and standby, an LCD display for the messages and for showing the measured values and stat uses, and a small number of keys for entering data.

be ensured. The key advantage of a nu merical protection system is its self supervision capability, which allows all system faults to be immediately detected and, when necessary, the protection blocked [2]. Another advantage of numerical bus bar protection is its modular design, which reduces the spare parts inventory to just a few types of module. Nevertheless, individual component failure is possible over the long lifetime of the system. A back-up system therefore has to be available in case repairs be come necessary. In the case of busbars, this is normally the distance protection in remote stations. Security and stability The security and stability of the numerical busbar protection are guaranteed by a range of measures corresponding to the different sources of failure which could be responsible for unselective or sudden tripping. As mentioned, two independent pro tection algorithms provide higher se curity, since both have to come to the same conclusion (eg, an internal fault exists) in order for tripping to take place. Most hardware and software failures are detected by the self-supervision func tion. Attention also has to be given to the auxiliary contacts of the disconnectors. This requirement is well known from pre vious busbar protection systems. Super vision of each disconnector is an integral part of the busbar protection, and the dis connector replicas are realized without mechanical parts. An additional logic is also available that eliminates the time-consuming adjust ment of the auxiliary contacts of the dis connectors. Testing Each NBP system is fully tested in the factory in accordance with the busbar and NBP configuration specified by the customer. A new testing system has been developed for simulating the numerous operating situations of a versatile NBP system and verifying the correct behav iour of the numerical algorithms.