Electrical Transmission and Distribution--Relay Protection (part 1)

1. INTRODUCTION

Switchgear, cables, transformers, overhead lines and other electrical equipments require protection devices to safeguard them during fault conditions.

In addition, the rapid clearance of faults prevents touch and step potentials on equipment from reaching levels that could endanger life. The function of protection is not to prevent the fault itself but to take immediate action upon fault recognition. Protection devices detect, locate and initiate the removal of the faulted equipment from the power network in the minimum desirable time. It is necessary for all protection relays, except those directly associated with the fault clearance, to remain inoperative during transient phenomena, which may arise during faults, switching surges or other disturbances to the network. Protection schemes are designed on the basis of:

- Safety.

- Reliability.

- Selectivity.

The requirements for current and voltage transformers (CTs and VTs) associated with relay protection are described in Section 5, and fuse and MCB protection devices in Section 11. Standard reference texts are provided in the references section at the end of this Section. They very adequately cover protection theory and particular relays in UK ,US and general practice. Graphical symbols for switchgear, control gear and protective devices are given in IEC 60617-7. This Section, therefore, concentrates on the principal relay protection schemes and typical applications with practical calculation and computer assisted examples.

While the earliest relays were electromechanical in construction, techno logical developments led to the introduction of solid state or static relays using discrete devices, such as transistors, resistors, capacitors, etc. Advent of microprocessors led to the development of microprocessor-based relays and this culminated with today's state-of-the-art system of numerical relaying where the measurement principles themselves changed from analogue to numerical.

Other recent advances are discussed in the following sections.

2. SYSTEM CONFIGURATIONS

2.1 Faults

All power system components are liable to faults involving anomalous cur rent flow and insulation breakdown between conductors or between conductors and earth. The insulation material may vary from air, in the case of a transmission line, to oil, SF6 or a vacuum, in the case of switchgear. The transmission and distribution engineer is concerned with symmetrical faults involving all three phases with or without earth, and asymmetrical faults involving phase-to-phase and one or two phase to earth faults. In addition, interturn winding faults also occur in transformers and electrical machines.

Section 1 describes computer assisted methods of deriving fault levels in power system networks and Section 28 describes the basic fundamentals involved.

2.2 Unearthed Systems

Such arrangements are only found in small isolated networks. At first sight the earth fault current would seem to be negligible with this connection. In practice, for all but the smallest networks the capacitive current becomes significant and dangerous transient overvoltages can occur due to low power factor arcing faults to earth. Unearthed systems therefore require high insulation levels and are limited to low voltage distribution where insulation costs are less significant. The main application is for very critical systems where continuity of supply is of paramount importance; two separate faults are required before an outage occurs and the first earth fault simply causes alarms which enables damage to be located and repaired before the critical supply is lost.

2.3 Impedance Earthed Systems

In this configuration a resistance or reactance is placed between the trans former neutral and earth. The earth fault current may be limited by the sizing of the impedance. This has the advantage of limiting:

- possible damage to equipment from the fault current;

- interference to control and communication circuits from the resulting induced currents.

High insulation still has to be incorporated in the impedance earthed system since voltage to earth levels on the unfaulted phases during a phase to earth fault will exceed 80% of the normal system phase-to-phase voltage. This is not normally a problem at system-rated voltage levels of 145 kV and below.

The impedance earthed system is known as a 'non-effectively' earthed sys tem. Normally a resistor rather than a reactor is used and the value is chosen still to ensure sufficient fault current to reliably operate the protection under all fault conditions. For a single supply point infeed the protection sensitivity may be set at, say, approximately 10% of the associated transformer full load rating and the earthing resistor to give a fault current equal to the trans former rating for a solid earth fault close to the transformer terminals.

2.4 Solidly Earthed Systems

Solidly earthed systems have the transformer neutral connected directly to earth. This has the advantage that it limits the likely overvoltages during fault conditions and is applied by most electricity supply companies for rated voltages above 145 kV. The voltage to earth levels on the unfaulted phases should not exceed 80% of the normal system phase voltage with the solidly earthed arrangement. The system is then known as 'effectively' earthed and is considered to be satisfied for ratios of X0/X1 , 3 and R0/X1 , 1 throughout the system under all conditions. In practice, these ratios will vary according to the network switching conditions and connected generation. The disadvantage is that the earth fault current can exceed the three phase fault current depending upon the ratio of zero-to-positive sequence impedance (see Fig. 1). Substation equipment must be rated accordingly. Sufficient cur rent to operate the protection relay equipment is, however, not normally a problem. In addition, it should be noted that a high earth fault current will lead to high touch and step potentials during the fault conditions. This must be limited to safe levels by adequate substation earthing. Further, control and communication circuits must be protected against induced currents and possible interference resulting from the earth fault.

2.5 Network Arrangements

FIG. 1 Maximum line-to-earth voltage at the fault for earthed neutral systems under any fault conditions.

FIG. 2 Typical radial distribution system.

FIG. 3 Typical parallel feeder arrangement.

FIG. 4 Typical ring system showing use of directional relays.

Notes: Arrows represent current flow direction upon which relays will act. A, B, C, etc. are circuit breakers operated by associated relay. 1, 2, 3, etc.: busbar identification.

2.5.1 Radial

A simple radial feeder is shown in Fig. 2. The fault level is highest closest to the source and limited by the impedances from source-to-fault location.

Clearance of a fault near the source will result in loss of supply to down stream loads. Protection selectivity must be such that a fault on busbar A must be isolated by only tripping the circuit breaker X via relay R1 and maintaining supply to load busbar B.

2.5.2 Parallel

A parallel feeder arrangement is shown in Fig. 3. A fault on one parallel feeder should be cleared by suitable protection such that it is quickly isolated from the supply. There should be no loss of supply via the remaining healthy feeder to the load.

2.5.3 Ring

A ring feeder arrangement is shown in Fig. 4. Two routes exist for the power inflow to a faulted feeder in a closed ring system. It is therefore necessary for the protection devices only to isolate the faulted section and not disconnect the whole system from the source. Often such ring systems use directional relay protection which requires both VT and CT connections.

Alternatively, they may be operated with a mid-point feeder circuit breaker open in order to simplify the protection arrangements.

2.5.4 Interconnected

This is a more complex arrangement of interconnected parallel and radial feeders, often with multiple power source infeeds. More sophisticated protection schemes are necessary in order selectively to disconnect only the faulted part of the system.

2.5.5 Substations

Busbars, transformers, cables and other important plants are all involved in the different substation layouts described in Section 3. The switchgear arrangements will help to dictate the types of relay protection devices used throughout the particular substation.

3. POWER SYSTEM PROTECTION PRINCIPLES

3.1 Discrimination by Time

For simple radial circuits discrimination is achieved by giving the minimum tripping time setting to the relay furthest away from the power source. A small time delay is then added to each relay in turn, moving nearer to the source each time. This ensures that the relay closest to the fault trips first, and as a result leaves the rest of the system between the source and the faulty section in service.

It is necessary to allow a minimum grading interval or delay between successive relay settings in order to take account of:

1. Circuit breaker tripping times - typically from 150 msec for an older oil circuit breaker to 50 msec for the latest vacuum or SF6 switchgear.

2. Relay time delay errors - variation from the characteristic time delay curve for the relay as allowed by the appropriate specification standard, say, 150 msec.

3. Relay reset time - the relay must definitely fully reset when the current is 70% of pick-up value. Electromechanical relays reset at 90-95% of setting and a figure of 85% is taken for calculation purposes. Solid state relays have an even better improved characteristic in this regard.

4. Relay overshoot - an electromechanical relay must stop all forward movement or overshoot of the induction disc within 100 msec of the removal of current. Again solid state or numerical relays have an advantage over electromechanical types in this regard.

If all these items are additive then for discrimination to be achieved typical time grading intervals of 0.4-0.5 s are used for electromechanical relays with oil circuit breakers and 0.25 s for modern solid state or numerical relays which are tripping vacuum or SF6 switchgear. The effect of current trans former errors on relay operating times is not expected to be additive and such errors (say 65%) are normally neglected when establishing a discrimination margin.

The disadvantage of discrimination by time delay alone is that the longest tripping times are those nearest the source. The fault current will be the highest and most damaging at this point in the circuit. Therefore shorter tripping times near the source would be an advantage.

3.2 Discrimination by Current Magnitude

The impedance of the power circuit between source and fault limits the fault current flowing at any point. Therefore by suitably selecting the current set ting at which the particular relay operates discrimination can be achieved. In practice this is quite difficult for transmission and distribution feeder circuits because various interconnection arrangements significantly alter the fault level at any point in the network. The method works well for power trans former protection where instantaneous high set overcurrent relays can be used to protect the HV windings. Similarly, but for different reasons, instantaneous earth fault relays can be applied to the delta winding of a delta star (Dy) power transformer. Here the zero sequence currents generated in the secondary star winding during earth faults in that winding or system do not appear in the primary delta winding. In this case an instantaneous earth fault relay on the transformer primary delta will not respond to LV earth faults.

3.3 Discrimination by Time and Fault Direction

It is possible to add directional sensing elements to the relay protection sys tem such that the relay responds to both the magnitude and one particular direction of the current flow. Typical applications are for closed ring feeder systems, parallel feeders and parallel transformers. It is vitally important during commissioning of such protection schemes that the polarity of operation is properly checked or maloperation and lack of discrimination may result.

3.4 Unit Protection

In these schemes the CTs located at either end of a feeder, transformer or 'unit' of plant to be protected (the protected zone) are interconnected. A comparison of magnitude and phase angle of the current entering the protected zone with that leaving is made. Two requirements are checked:

1. If the currents entering and leaving the protected zone are equal, operation of the protection must be prevented - this is known as the through fault stability requirement.

2. If the currents entering and leaving the protected zone are unequal the protection must operate - this is known as the sensitivity to internal faults requirement.

A number of unit protection schemes rely on a current balance principle.

Obviously for power transformer differential unit protection the primary and secondary CT connections have to be arranged to take into account the different phase relationships associated with different power transformer connections (Dy1, Dy11, etc.) as explained in more detail in Sub-Section 5.1 (below).

Because the correct current balance performance relies upon CT characteristics it is essential that the associated CTs are matched and dimensioned correctly. Stability under fault conditions outside the zone of protection is vital. Therefore it is necessary to ensure that the spill current is minimized by limiting the degree of CT saturation or to use high impedance relays which are designed to remain stable even under saturated CT conditions.

Conventionally, this is achieved by ensuring that the voltage which appears across the relay circuit with one CT fully saturated is insufficient to operate the relay at a given current rating. Modern numerical relays are also avail able in a low or medium impedance version with differing operating principles to combat problems related to CT saturation, remanance, ratio differences, etc.

The advantage of unit protection is that it provides a very fast (typically 200 msec or less) disconnection only of the plant being protected. The disadvantage is that the interconnection between the relays requires communication systems which make the overall schemes more expensive than simple time graded schemes for long feeder lengths.

3.5 Signaling Channel Assistance

Rapid protection operation may be necessary for system stability reasons as explained in Section 1.3, Section 1. The speed of response of a protection system may be enhanced by the use of interconnecting signaling channels between relays. For example this enhancement can be applied to a distance protection scheme for improving the fault clearance time over the last 15-20% of the feeder length, as explained in more detail in Section 10.6.

Such signaling channels may be by the use of hard wire circuits (dedicated pilot wires, rented telephone cables, etc.) using on/off or low frequency signals. Alternatively, signal information superimposed upon carrier frequencies of several hundred kHz may be used over the power circuits (power line carrier (PLC)) to convey the information. A more modern development is to use fiber optic cables which may, for example, form an integral part of an over head line earth wire. The transmission times are essentially instantaneous but delays associated with interposing relays and electronics must all be considered when checking for correct selective grading.

4. CURRENT RELAYS

4.1 Introduction

The following types of current relays are considered in this section:

- Plain overcurrent and/or earth fault relays with inverse definite minimum time lag (IDMTL) or definite-time delay (DT) characteristics.

- Overcurrent and/or earth fault relays as above but including directional elements. Note that a directional overcurrent relay requires a voltage connection and is not therefore operated by current alone.

- Instantaneous overcurrent and/or earth fault relays. For example a high set overcurrent (HSOC) relay.

- Sensitive earth fault (SEF) relays.

Differential and unit protection schemes that require connections from more than one set of current transformers are described in Section 10.5. Current operated relays are applied almost universally as the 'main' or only protection on power distribution systems up to 36 kV. For very important feeders and at transmission voltage levels above about 36 kV, current operated relays tend to be used as 'back-up' protection to more sophisticated and faster acting relay systems. Current relays may also form part of special schemes such as circuit breaker failure protection.

FIG. 5 Normal characteristic inverse definite minimum time lag (IDMTL) relay curve.

4.2 Inverse Definite Minimum Time Lag (IDMTL) Relays

Historically this type of relay characteristic has been produced using electro magnetic relays, and many such units still exist in power systems. A metal disc is pivoted so as to be free to rotate between the poles of two electro magnets each energized by the current being monitored. The torque produced by the interaction of fluxes and eddy currents induced in the disc is a function of the current. The disc speed is proportional to the torque. As operating time is inversely proportional to speed, operating time is inversely proportional to a function of current. The disc is free to rotate against the restraining or resetting torque of a control spring. Contacts are attached to the disc spindle and under preset current levels operate to trip, via the appropriate circuitry, the required circuit breaker. The theoretical characteristic as defined in IEC 60255-3 is based on the formula:

...where...

t=theoretical operating time

G=value of applied current

Gb =basic value of current setting K and a=constants

With K=0.14 and a=0.02 the 'normal' inverse curve is obtained as shown in Fig. 5. This characteristic is held in the memory of modern microprocessor controlled solid state relays. Electronic comparator circuits are used to measure the source current and initiate tripping depending upon the relay set tings. In comparison with grading by time settings alone the IDMTL relay characteristic is such that it still allows grading to be achieved with reduced operating times for relays located close to the power source.

This type of relay has two possible adjustments:

1. The current setting by means of tap 'plugs' on electromagnetic relays or 'DIP' switches on solid state relays for values between 50% and 200% in 100 [...]

4.3 Alternative Characteristic Curves

Alternatives to the 'normal' IDMTL characteristic are available. A 'very' long time inverse curve is obtained with the constants K=13.5 and a=1.0 in the theoretical operating time equation. This very inverse characteristic is useful as a last stage of back-up earth fault protection (for example when used in conjunction with a CT associated with a transformer neutral earthing resistor). The 'extremely' inverse curve characteristic (K=80, a=2.0) is useful for ensuring the fastest possible operation whilst still discriminating with a fuse. The extremely inverse characteristic does not exhibit such a useful definite minimum time and it is difficult to accommodate more than one or two such relay stages in an overall graded protection scheme. A variety of commonly used characteristic curves are illustrated in Fig. 6.

4.4 Plotting Relay Curves on log/log Graph Paper

The characteristic curves shown in Figs. 10.5 and 10.6 are plotted on log/log graph paper with time on the vertical scale and current on the horizontal scale. Three or 4 cycle log/log paper is the most useful in practice for manual relay grading exercises. If a template for the normal characteristic is used based on the same cycle log paper then the 10 or 3 s operating time at 2 or 10xPSM may be used as guide points. The actual circuit current being monitored is transformed by the actual CT ratio and relay PSM being used.

The operating curve for other than TMS=1 relay settings may be approximately drawn by moving the template vertically up the log paper so that the 10xPSM mark coincides with an operating time in seconds equivalent to 3 multiplied by the actual TMS in use.

Relay characteristics are now available on disc for use with microcomputer relay grading programs. An example of such a computer assisted relay grading exercise is given in Sub-Section 10.8.

FIG. 6 Typical IDMTL relay characteristics.

4.5.2 IDMTL Back-up Protection

In more important networks and at higher voltage levels current relays are used as a 'back-up' to more sophisticated and faster acting 'main' protection systems. The back-up protection should be graded to achieve selective trip ping if the main protection fails to operate. However, it must be noted that this is not always possible on highly interconnected networks involving widespread generation sources. If discrimination is not possible throughout the network then it must at least be certain that the back-up protection ensures overall circuit breaker tripping times within the thermal capability of the plant being protected.

If all the IDMTL back-up relays use the same characteristic curve and time multiplier then they will all tend to grade naturally. The faulty circuit will normally carry much more current than the many other interconnected circuits supplying the fault and the relays associated with the faulty feeder will therefore tend to trip faster. An example is shown in Fig. 7a. In this arrangement the IDMTL relay gives a greater measure of selectivity than if definite minimum time relays were being used.

FIG. 7a IDMTL overcurrent relay discrimination with common relay setting on all feeders.

FIG. 7b Traditional electromechanical IDMTL relay. The illustration shows an earth fault relay (type CDG) with a 'very inverse' characteristic. The plug settings are at the front of the relay and time settings at the top (courtesy GEC Alsthom T & D Protection & Controls).

4.5.3 Instantaneous High Set Overcurrent Relays

The tripping times at high fault levels associated with the IDMTL characteristic may be shortened by the addition of instantaneous high set overcurrent (HSOC) elements. These may be an integral part, for optional use depending upon front panel settings, of a modern solid state relay. Alternatively, the elements may be specified as an extra requirement when ordering electro magnetic type relays and included in the same relay case. The high set overcurrent or earth fault characteristic is such that at a predetermined current level the relay initiates essentially instantaneous tripping. This allows relays 'upstream' towards the power source to be graded with the instantaneous HSOC 'downstream' relay setting and not the maximum fault level that would be required to trip using the normal IDMTL curve.

The relays are often specified for use with transformer protection. The primary side HSOC setting is chosen to be high enough so as to ensure no operation for secondary side faults. This is possible because of the fault limiting effect of the transformer reactance which allows adequate grading with transformer secondary protection. Some manufacturers produce 'low transient over-reach' HSOC elements. These include a means of rejecting the DC component of fault current and improve the grading margin between HSOC setting and maximum transformer through fault current.

HSOC protection may be applied to cable feeders, overhead lines or series reactors where sufficient reactance is available to avoid anomalous tripping for remote faults in the network. Care must also be taken into account to ensure any initial energization or switching inrush currents do not cause the relays to trip anomalously.

FIG. 7c Modern solid state IDMTL relay-type KCGG (courtesy of GEC Alsthom T & D Protection & Controls).

FIG. 7d Modern microprocessor controlled solid state IDMTL relay - type MCGG. A wide variety of characteristics is available. PSM and TSM settings are configured using the small switches accessible from the front of the relay (courtesy GEC Alsthom T & D Protection & Controls).

FIG. 8 High set protection exercise network (all impedances to 100 MVA base).

4.5.4 Sensitive Earth Fault Relays

Under certain high resistance conditions ordinary distance protection and some other sophisticated relays may fail to operate. Such conditions have been experienced with flashovers from overhead transmission lines above bush fires, distribution lines adjacent to overgrowing vegetation and rubber tyred (tire) gantry crane vehicles touching overhead lines. The fault is said to be outside the 'reach' of the relay.

Historically, earth fault relays with very sensitive, low setting ranges, typically 1-5% of circuit rating, have been used to detect such faults. They may be connected in the residual circuit of the overcurrent back-up protection overcurrent CTs as shown in Fig. 9b. When used in conjunction with main protection they usually incorporate a sufficient time delay to allow the main protection and possible associated auto-reclose scheme to operate first.

In addition, when used in conjunction with parallel feeder arrangements the settings must not cause tripping due to unequal load sharing between the same phases of the two line circuits.

However, relays with a high impedance fault detection feature (HIF) have been available for some time now, based on advanced wavelet and statistical algorithms.

5. DIFFERENTIAL PROTECTION SCHEMES

5.1 Biased Differential Protection

5.1.1 Introduction

Basically, unit protection schemes compare the current entering and leaving the protected zone. Any difference will indicate the presence of a fault within the zone. By operation of the appropriate relays the associated circuit breakers can be made to trip thus isolating the faulty equipment from the power network.

====

FIG. 9a VT connections for directional relays.

Five-limb VT or Three Single-phase VTs To directional earth fault relay (a) To directional overcurrent relay

====

FIG. 9b Earth fault relay connected in residual circuit of protection CTs (23o/c, 13e/f protection).

========

With perfect CTs, relays and symmetry of connections, stability should be theoretically possible under steady state conditions and no anomalous tripping due to faults outside the zone of protection occurs. In practice, there will be differences in the magnitude and phase of the currents entering and leaving the zone of protection. CT characteristics will not be perfectly matched and DC components will be involved under fault conditions. There is an upper level of through fault current at which both steady state and transient unbalance rapidly increases as shown in Fig. 10. Restraint features, known as 'bias', are therefore added to the relays to desensitize under through fault conditions. The discrimination quality of a differential system can be defined in terms of a factor given by the ratio of the degree of correct energization of the relay under internal fault conditions to that which occurs (and is unwanted) under external fault conditions at the same time as the specified CT primary current. If the system were perfect the factor would be infinite.

FIG. 10 Biased differential protection-discriminating factor and stability: (a) unbiased relay-discriminating factor5AC/BC; (b) biased relay-discriminating factor A0 C0 /B0 C0 (c) biased relay characteristics under internal (int.) and external (ext.) conditions.

5.1.2 Biased Differential Transformer Protection

FIG. 11 Transformer biased differential protection. CT connections arranged to compensate for vector grouping and turns ratio.

Fig. 11 shows a typical arrangement. It is necessary to take into account the following:

1. Transformer ratio - the magnitudes of the currents on the primary and secondary sides of the power transformer will be inversely proportional to the turns ratio. The CTs located on the primary and secondary sides of the power transformer should therefore be selected to match this difference so that CT secondary currents are as equal as possible and spill cur rents through the relay operating circuit are kept to a minimum. CT ratios are standardized and often do not match the power transformer ratio. In such cases interposing correction CTs are employed.

2. Transformer vector grouping - the primary and secondary power trans former currents will have differing phase relationships due to the vector grouping. The CT connections must be made in such a way as to compensate for this change and bring the CT secondary currents back into phase. This phase correction can be achieved either with the main or interposing CTs.

3. Transformer tap changer - if the transformer has a tap changer then the ratio between the magnitude of the power transformer primary and secondary currents will vary depending upon the tap position. The mean tap position should be taken for calculations and any spill current compensated by the relay bias circuit.

4. Magnetizing inrush current - when a transformer is energized a magnetizing inrush current up to 10x full load current can occur. This current is present only on the primary or source side of the transformer and if not compensated will introduce an unbalance and cause the differential protection system to trip.

In fact magnetizing inrush current has a high second harmonic component which does not specifically appear under fault conditions. A compensation circuit is therefore introduced into the relay to detect second harmonic components. These are recognized as switching transients and the relay is restrained from operating under this condition.

An alternative method of inrush current recognition and relay operation restraint used by some manufacturers is the detection of the zero period in the magnetizing inrush current wave form and blocking relay initialization.

Restraint and/or blocking for other harmonics such as the 5th are also catered to in modern transformer differential schemes.

FIG. 12 High impedance protection under external fault conditions. Rct: CT resistance; R1: cable and lead resistance between CT and relay; Rr: relay resistance (CT magnetizing circuit neglected).

FIG. 13a Balanced earth fault and restricted earth fault protection examples - network configuration (all impedances on 100 MVA base).

FIG. 13b BEF protection CT arrangement.

FIG. 13c LVREF protection CT arrangement and lead-CT resistances.

5.2 High Impedance Protection

Typical applications for high impedance relays include busbar and restricted earth fault protection schemes. Under through fault conditions the differential relay unit protection scheme must remain stable. During such an external fault one CT may become saturated producing no output while a parallel connected CT might remain unsaturated and continue to produce full output.

This condition is shown in Fig. 12. The current from the remaining operating CT will divide between the relay and the other arm of the network comprising of connecting cable and lead impedance and the effectively short circuited saturated CT. By using a relay of sufficiently high impedance (not a problem with modern solid state electronic relays) the proportion of current flowing through the relay is reduced below the operating level of the relay making for stable operation under these heavy through fault conditions. (It must be noted that modern low or medium impedance relays are designed so that they are largely unaffected by some of these issues and cater for larger amounts of CT saturation, remanance, etc.) Because the relay has a high impedance the voltage developed across it during an internal fault may be sufficiently high so as to force the CTs into saturation. It is therefore necessary when applying high impedance relay protection to calculate the relay circuit resistance and CT 'knee point' voltage such that the CT is still able to produce sufficient output under saturated ....

5.4 Pilot Wire Unit Protection

5.4.1 Pilot Cables

Since currents at each end of the protected zone must be compared a signal ling channel is necessary using:

- pilot cables - typically two core, 2.5 mm^2 cross-section;

- private or rented telephone circuits;

- radio links;

- fiber optic links.

The selection is based on economic and reliability factors. With increasing length, or difficult installation conditions, buried hard wire or fiber optic cable links become expensive.

Telephone lines have typical limitations on the peak applied voltage and current as follows:

- Maximum applied voltage - 130 V DC.

- Maximum current - 0.6 A.

All connected circuits must be insulated to 5 or 15 kV.

Telephone pilots may be subjected intermittently and without warning to ringing tones, open or short-circuit conditions. Therefore their use requires more complicated terminal equipment and if rented may have a lower level of reliability.

Most systems use armored and well-screened twisted pair two core cop per cable installed in the same trench as the power feeder cable or under the overhead transmission line in the associated wayleave. This type of pilot cable is known as 'high grade'. The twisted pairs and minimum power-to pilot cable spacing of approximately 300 mm help to minimize pick-up noise and induced currents. At present the typical economic distance for such a unit protection relay and pilot cable scheme is 25 km. After this distance the cost of cable and trenching begins to exceed the cost of a comparable distance relay scheme. The pilot cable may be represented as a complex pi net work of uniformly distributed series resistance and inductance with shunt capacitance and conductance. If the cable is in good condition the leakage conductance will be very small. Unlike a communications circuit (which is terminated with the cable characteristic impedance in order to achieve a resistive load and maximum power transfer) a protection pilot cable is terminated in an impedance which varies almost between open and short circuit (see Fig. 14). On an open circuit the pilot cable has a predominantly capacitive impedance while on short circuit the impedance is inductive. The impedance of the pilot cable is an important factor in the design of the protection relay system and it must match the relay manufacturer's requirements.

====

FIG. 14 Pilot cable equivalent circuits: (a) p representation; (b) short-circuit conditions (inductive); (c) open circuit conditions (capacitive).

====

5.4.2 Summation Transformers

In order to minimize pilot cable costs the three phase primary currents are converted via the matched CTs and summation transformers at each end of the link to a single phase secondary current. This is justified since, as explained , the primary currents bear definite relationships for different types of fault condition. The output from the summation trans former may be set, by altering the proportion of n turns as shown in Fig. 15, to give greater sensitivity to earth faults. This is a useful feature when using a unit protection scheme on a resistance-earthed system which produces low earth fault currents.

FIG. 15 Summation transformer arrangement (three phase to one phase).

FIG. 16a Pilot wire differential protection under "through" and "in zone" fault conditions (circulating current).

FIG. 16b Balanced voltage scheme.

5.4.3 Basic Schemes

The fundamental difference between biased differential protection and pilot wire differential protection is that relays are required at each end of the pilot wire scheme. Only one relay is required for the biased differential schemes described in Sub-Section 5.1.

The long distances involved between the two ends of a feeder cable or an overhead line circuit necessitates the use of a relay at each end of the protection zone. The relays control the associated circuit breakers and minimize the effects of pilot cable characteristics on relay performance. In addition, a single relay scheme is not used because the CTs would have to be impossibly large in order to avoid saturation on through fault current when used in conjunction with a pilot cable burden of approximately 1,000 ohm. The relays are arranged to operate simultaneously with intertripping to provide very rapid fault clearance irrespective of whether the fault is fed from one or both ends of the protected zone.

Practical schemes are based on circulating current or balanced voltage principles as shown in Figs. 10.16a and b. Since the pilot cables may run in parallel with power cables or overhead lines for long distances isolation transformers and non-linear resistors are used in order to prevent unacceptable induced voltages appearing at each end on the relays. The condition of the pilot cables can also be monitored by incorporating pilot wire supervision modules into the relay. Such supervision will not prevent anomalous operation but can be arranged to raise an alarm due to open circuit, short circuit or cross-connected faulty pilots. When there is sufficient fault current anomalous operation due to faulty pilots may be positively prevented by the addition of overcurrent check features into the relay scheme.

FIG. 17 Modern pilot wire protection relay - type MBCI (courtesy GEC Alsthom T & D Protection & Controls).

Fig. 17 shows a modern pilot wire protection relay based on the voltage balance principle which can incorporate all these features.

5.5 Busbar Protection

5.5.1 Introduction

Busbar reliability is of paramount importance since failure will result in the loss of many circuits. In practice, busbar faults are rare and usually involve phase to earth faults. Such faults may be due to:

- insulation failure resulting from deterioration over time;

- flashover due to prolonged or excessive overvoltages;

- circuit breaker failure to operate under through fault conditions;

- operator/maintenance error (especially leaving earths on busbars after a maintenance operation);

- foreign objects falling on outdoor catenary busbars.

Some electricity supply companies prefer not to employ busbar protection because they do not wish to incur the occasional outage due to protection maloperation which could be more likely than a true busbar fault which might only occur once in 20 years. Such a risk may be reduced by employing a separate 'check' feature in the busbar protection relay scheme which must also recognize the fault before tripping is initiated. In a similar way, in areas of low thunderstorm activity, an electricity supply company may decide not to employ outdoor substation overhead lightning screens because the likelihood that they may fall and cause a busbar fault is considered to be more probable than an outage due to a lightning strike.

5.5.2 Frame Leakage Detection

This is the cheapest form of busbar protection for use with indoor metal-clad or metal-enclosed switchgear installations. Since the probability of a busbar fault on such modern equipment is very small, busbar protection on such equipment is only considered for the most important installations. The switchboard is lightly insulated from earth (above, say, 10 ohm) and currents in a single connection to earth measured via a CT and frame leakage relay.

This arrangement requires care to ensure all main and multicore cable glands are insulated and that bus sections are not shorted by bolted connections through the concrete floor rebar or switchgear steel floor fixing channel arrangements ( Fig. 18a). To avoid anomalous tripping an additional check feature should be incorporated when possible by using a CT on the star point neutral of the supplying transformers. The installation must be in a dry sub station and in practice is not really suitable for retrofitting on existing switchboards. Equally, care must be taken in extending switchboards that do already have frame leakage detection, since some new switchgear types do not permit the necessary insulating features.

The frame insulation resistance is effectively in parallel with the substation main earthing system with typical resistance values to true earth of less than 1 ohm. The earth leakage relay will therefore only 'see' approximately 9-10% of the earth fault current when a 10 ohm switchgear-to-earth insulation level is employed. To ensure stability under external out-of-zone fault conditions the frame leakage relay may be set at 30% of the minimum earth fault current.

FIG. 18a Frame leakage detection.

FIG. 18b CT location arrangements and overlapping zones of protection.

5.5.3 Bus Zone

A comparison is made between the currents entering and leaving the busbar or busbar zone. CTs are therefore required on all circuits and the CT locations are arranged to maximize the required zone of protection coverage as shown in Fig. 18b. Conventionally, main and check high impedance relays are used in conjunction with these CTs to measure the sum of all the currents. Very fast operating times (40 msec) are feasible with such schemes.

An example of the practical application of traditional busbar protection principles is given in Sub-Section 10.9.

The check feature makes tripping dependent upon two completely separate measurements of fault current using separate CTs and different routing for CT wiring to the protection relays. In a double busbar arrangement a separate protective relay is applied to each bus section (zones 1 and 2) and an overall check system arranged to cover all sections of both main and reserve busbars.

In current practice, schemes based on multiple operating principles and sometimes based on a distributed concept, are applied for EHV substations.

These schemes are typically low impedance types and handle differing CT ratios, a high amount of CT saturation, diverse issues such as evolving faults and CT remanance. Continuous self-supervision and the ability to detect abnormalities on CTs, CT wiring and the auxiliary contacts used to provide a bus image make the scheme complete and highly reliable, thus obviating needs for a separate set of check zone CTs and relays.

5.5.4 CT Selection

In order to ensure stability under load, switching transient and external fault conditions the CTs must be all carefully matched up to the maximum fault level with the same ratio and characteristics. As explained in Section 10.5.2 it is the voltage required to operate the relay rather than its current setting which determines the stability level of the scheme. The CT 'knee point' volt age must be kept as high as possible and at least three times the relay voltage setting. The testing of busbar protection schemes will therefore necessitate particular care over CT polarities, correct operation of busbar selector auxiliary contacts and primary operating current at the selected relay settings.

An interruption in CT wiring will cause an unbalance and anomalous busbar protection operation. Wiring supervision is therefore a feature of most schemes in order to raise an alarm with typical settings of unbalance at 10% of minimum circuit rating.

See also Sub-Section 10.10 regarding optical communication links.

cont. to part 2 >>