Static Relays

Electromechanical relays were used successfully for many years, but these mechanical devices are subject to drift and fatigue, and since they lie inoperative for long periods, periodic testing – and in some cases cleaning – is necessary in order to be sure that a given relay will operate correctly when called upon to do so. In the 1960s and 1970s electromechanical relays began to be replaced with electronic relays using analogue circuits to synthesise the required protection characteristics. These were known as static relays since they had no moving parts. Whereas electromechanical relays were powered by the CTs and/or VTs in the circuits they protected, static relays require a secure and reliable source of power to operate. The substation DC tripping supply is used for this purpose. This potential is provided by a battery and it generally lies in the range between 24 and 250 volts DC, with 125 volts commonly used.

Digital and Numerical Relays

The advent of the microprocessor in the 1970s and 1980s saw analogue relays replaced by digital devices in which currents and voltages are represented digitally, and an algorithm is used to implement the protection function. Early digital relays generally performed the task of a single protection element and multiple devices were required to provide a range of protection functions. Advances in technology, the arrival of digital signal processors and parallel processing, have enabled many functions to now be performed by one device, known as a numerical relay or a multi‐function relay. Numerical relays can therefore perform the tasks of multiple protection elements and may have spare processing capacity to enable other parameters to be displayed, including active and reactive power flows and a harmonic analysis of voltages and currents. Capture of current and voltage waveforms is available from some numerical relays, so that faults can be analysed after the event. In addition, self‐diagnostic testing is generally provided, so that in the event of an internal fault an operator can be alerted.

Check Relays

In addition to the main protection relay, high risk schemes such as bus‐zone or sensitive earth fault protection, generally include a separate check relay in order to provide an independent confirmation of a fault or to provide discrimination. (For example, this may be an overcurrent relay looking for a residual current flowing towards a transformer neutral terminal.) In such cases both the main and the check relay must concur for a trip to be initiated. Operation of the check relay will never in itself cause a trip.

In modern numerical bus‐zone protection, the main and check elements often occur within the same relay, although its architecture usually ensures that decisions are made by two independent algorithms. For schemes such as sensitive earth fault protection the check relay is a separate device, even in today’s numerical environment.

Backup Protection

Backup protection is provided to clear a fault should the primary scheme(s) fail to operate. This may be through the operation of a nearby graded overcurrent relay or perhaps by a circuit breaker failure relay that has detected current flow after the breaker concerned has been sent a trip command.

Backup schemes may also be located in another substation, remote from the primary protection system, so that no item of equipment is common to both. (An example of this is the Zone 2 setting of a remote substation’s distance relay.) Backup protection tends to be slower than the primary system it protects and it generally interrupts considerably more of the network.

Reliability

The reliability of a protection system is a function of its design, the equipment used and the care taken in its construction. The simpler and more robust the design the more likely it will be to operate correctly when required to do so. Modern numerical relays provide multiple protection functions and apply any necessary phase shifts digitally. As a result, they require relatively simple current and voltage connections, although the design of the relay itself is more complex. On the other hand complicated connection arrangements are frequently required with static and electromechanical relays to provide the correct phase relationships between the parameters being measured. This introduces the possibility of wiring errors which may not become apparent until the relay fails to operate correctly in service. The commissioning of protection schemes is therefore particularly important. The required protection functions are simulated during commissioning by injecting appropriate currents and voltages into each relay to ensure that it operates as expected in all scenarios.

Unit Protection

Time/current grading is inherently slow in its operation since relays closer to the source of supply are delayed in their operation in order to permit those closer to the fault to operate. Generators, transmission lines and busbars all require fast protection for the reasons already described. In such cases unit protection is employed to provide a fast fault clearance in a defined zone of protection. Current transformers are required at each zone boundary, from which the associated protection relay compares the current entering with that leaving. Should an in‐zone fault occur, the substantial residual current created will be seen by the relay, resulting in the tripping of all associated circuit breakers. An important requirement of a unit protection scheme is its ability not to operate in the event of a fault external to the zone of protection. The scheme must remain stable for external faults (often called through faults).

Figure 12.4 shows an example of unit protection in the form of restricted earth fault (REF) scheme surrounding the phase windings of a star connected transformer. The REF relay is essentially an overcurrent relay operating on the residual current arising from the CT connection shown. This will only occur in the event of a ground fault within the protected zone and because unit protection is not required to grade with any other, it can be made fast acting. (See Section 12.6.2 for more detail.).

Pick‐up/Reset Ratio

The pick‐up current is the minimum current that will cause the disc to turn. A relay will not pick up for secondary currents equal to the plug setting, but will do so for currents between 1.05 and 1.3 times this value. Should the load current just exceed the pick‐up current, the disc will slowly begin to turn, or in the case of a digital relay, the tripping delay timer will commence. If the current subsequently falls below the pick‐up value the relay should reset, allowing the disc to return to its rest position. There is a little hysteresis between the pick‐up and the reset currents, defined by the pick‐up/reset ratio which typically lies in the range 0.9–0.95. Therefore the current setting must lie sufficiently above the maximum load current to enable the relay to correctly reset in such circumstances, avoiding a nuisance trip, i.e.

(12.1)

Grading of Overcurrent Relays

In the case of IDMT relays a grading margin is provided between relays in adjacent zones through the choice of appropriate time multiplier settings so that there is little chance of any relay operating before that immediately downstream has had an opportunity to clear the fault. Grading margins include allowance for relay timing errors and overshoot, CT composite error and circuit breaker operating time. A typical grading margin for an electromechanical relay is about 0.4 seconds, or 0.3 seconds for a digital relay. It is the additional time required before the upstream relay operates, should the downstream one fail. The grading margin is applied at the maximum fault current seen by each relay. Provided that all relays use the same tripping curve, this will ensure that satisfactory discrimination will also exist for lower fault currents as well.

The TMS or TD can be estimated from the tripping curves, or obtained more precisely from the equations:

(12.2)

Example: Determine the trip time of a standard inverse (SI) overcurrent relay with a plug setting of 1.5 A and TMS of 0.4 when connected to a 1000:1 CT with a primary fault current of 10,000 A.

Solution: The CT secondary current will equal 10 A which equates to 6.67 multiples of the plug setting (i.e. M = 6.67). The trip time can be read from Figure 12.6, or it may be calculated using the SI equation. By either method the trip time seconds.

Overcurrent Relay Connections

Overcurrent relays may be installed on three‐wire circuits where, under normal conditions , as depicted in Figure 12.7, allowing both phase and earth faults to be detected. Traditionally the arrangement of Figure 12.7a would have been adopted using three relays of the type shown in Figure 12.1. Two of these would be used to detect phase overcurrents, one in the A phase and one in the C phase, while the third would see the residual current and detect earth faults. Since the third element only sees zero‐sequence current, it can be provided with quite a low current setting (20–30% of the phase current setting is typical), necessary to limit damage from earth faults or to operate on the small earth fault currents that flow in impedance‐earthed networks.

A phase‐to‐phase fault will result in overcurrents on at least two phases, and these will be seen by at least one phase overcurrent element. However, in the special case of a delta‐star transformer protected by an overcurrent relay on the HV side, a phase‐to‐phase fault on the LV will result in twice the fault current appearing on one primary phase, which may be the one without an overcurrent element. While such a fault will be cleared, this may not occur as fast as if the heavily faulted phase had its own overcurrent element. These days all phases are provided with a dedicated overcurrent element from either a digital or a numerical relay in the arrangement shown in Figure 12.7b.

Transient Over‐Reach

The transient DC offset early in a fault current waveform will be seen by an instantaneous relay and may cause it to operate prematurely and overreach into the next zone.

The transient overreach percentage is defined as:

(12.3)

where I_{RMS pick up} is the relay’s steady state RMS pick‐up current and is the steady state RMS current which, when fully offset, causes the relay to pick up.

To avoid transient overreach the instantaneous current setting must be increased by this percentage, which in the case of high X/R ratios, can be large. This may result in a setting that is unusable should it exceed the maximum steady state in zone fault current, or result in the relay overreaching into the adjacent zone.

Overcurrent Protection: Current Transformer Requirements

Because IDMT overcurrent protection is inherently slow the transient performance of the CTs used is generally not critical, and some CT saturation in the first few cycles can be tolerated since the relay will usually operate well, after the transient component has died away. Class P transformers are generally used in such applications, but the accuracy limit factor chosen must be sufficient to permit the steady state fault current to flow in the connected burden. For example a 500:1, 15 VA, 10P20 core will deliver a maximum of 20 A to a 15 Ω burden with a composite error less than or equal to 10%; it will therefore be suitable for steady state primary fault currents of up to 10,000 A at rated burden. Where the total connected burden is less than the rated burden, the accuracy limit factor increases according to:

(12.4)

and the maximum primary fault current increases in proportion. Since the burden presented by modern numerical relays tends to be very small, the cabling burden is usually a significant component of the total burden seen by the CT, which may be considerably less than the rated burden.

In the case of instantaneous protection however, where fast operation is required, the transient performance of the CTs concerned is important. Due allowance must therefore be made for any DC transient component of the fault current. This will require the CT to generate larger CT secondary voltages, increased by the transient factor (1 + X/R). In such cases IEC class PX or TP or IEEE class C CTs should therefore be used.

Transformer Magnetising Current

The transformer’s magnetising current must also be catered for in a differential scheme. In the steady state, this is generally only a few per cent of the transformer’s rated current, but as discussed in Section 7.5.4, at switch‐on the magnetic inrush current may become very large indeed, momentarily exceeding the rated current. Such an inrush current will appear as an internal fault, and the relay must be restrained from operating while the steady state flux is established. The DC component present during magnetic inrush (see Figure 7.18) means that the current waveform does not possess half wave symmetry, and therefore contains even harmonics (see Section 13.2.2). Of these the second harmonic is usually the largest, and its presence is frequently used to temporarily restrain the relay’s operation.

Figure 7.18 also shows that the inrush current is close to zero for a considerable fraction of each cycle. This feature can also be used to inhibit relay operation until a steady state flux is established in the core.

Zero‐Sequence Current

An earthed star or zigzag connected winding on one side of the transformer can deliver zero‐sequence currents to an out‐of‐zone earth fault which will be misinterpreted by a differential relay as an in‐zone fault. It is therefore necessary to prevent the passage of zero‐sequence currents in the associated CT secondary circuit. In the case of electromagnetic or static relays this is achieved by including a delta connection, either in the CT’s windings themselves or in those of an associated interposing CT. Digital and numeric relays have the software capability of filtering zero‐sequence currents in addition to making phase and ratio corrections. Figure 12.12 illustrates this for a Dyn11 transformer where HV zero‐sequence line currents are blocked by the transformer’s delta winding. Zero‐sequence currents can flow in the star connected LV windings but are excluded from the secondary circuit by the delta connected CTs, which are also required for the phase correction.

Tap‐Changers

Where a tap‐changer is present, CT ratios are usually set to suit the principal tap (i.e. that for which the nominal primary voltage will generate nominal secondary voltage, on an unloaded transformer). A tap‐changer dynamically alters the transformer’s effective turns ratio, partially upsetting the primary–secondary current balance, which can lead to a spill current flowing in the relay. However, correct operation is ensured through the use of a biased differential relay in which the spill current required to operate the relay increases with the magnitude of the current flowing.

Biased Differential Relays

Magnetising current, the presence of a tap‐changer or CT ratio errors can all contribute to an erroneous spill current flowing in the relay’s operating winding. These effects could be negated by setting the relay’s differential operating current above the maximum spill current likely to arise from such sources; however, this would unnecessarily limit the sensitivity of the relay. Instead, the relay operating current is made to increase in approximate proportion to the magnitude of the fault current. This is achieved through the use of bias windings to restrain the relay’s operation, as shown in Figure 12.13. This significantly improves the low fault sensitivity by permitting a smaller differential current to trip the relay under light load conditions.

The bias characteristics for an Areva KBCH transformer differential protection relay appear in Figure 12.14. The bias current (sometimes called the restraint current) is equal to the average of the magnitudes of the CT secondary currents I₁ and I₂, while the minimum differential tripping current, can be set in the range between 0.1I_n and 0.5I_n, where I_n is the rated CT secondary current. The tripping characteristics lie above the likely magnetising current of the transformer concerned and have a slope of 20% up to rated current, to allow for operation throughout the transformer’s tapping range and any CT mismatch. Beyond I_n, the slope rises to 80% to allow for the effects of CT saturation at high through currents occurring due to the transient DC fault current component.

Differential Protection: CT Requirements

IEC class PX, TP or ANSI class C current transformers are generally used in transformer differential protection applications, and it is important that they do not heavily saturate during through‐fault conditions. The biased relay characteristic will accommodate some saturation, but if this is to be avoided, the knee point voltage must be chosen to accommodate the transient DC component of the fault current. The knee point voltage must exceed the maximum secondary winding voltage. A safety factor of 2–4 is often applied, therefore:

(12.5)

where E_k is the knee point voltage, I_s max is the maximum symmetrical secondary current under fault conditions, R_ct and R_b are the CT winding and the burden resistances respectively, including the cabling resistance, X and R are the primary network reactance and resistance up to the point of the fault and K_td is the transient dimensioning factor (typically 10–25).

In the case of IEC class PX or TP transformers the knee point voltage forms part of the specification and appears on the nameplate, while with ANSI class C CTs it is related to the rated secondary terminal voltage, and can be easily found from the corresponding magnetising characteristic.

Differential High Impedance Bus‐Zone Protection

Figure 12.17 shows two arrangements for differential bus‐zone protection. In Figure 12.17a circulating CT currents balance on an instantaneous basis, and since the single overcurrent relay only sees the zero‐sequence current, it will only operate for ground faults located between the incoming and outgoing CTs.

The three‐element scheme in Figure 12.17b operates in a similar fashion, but since a relay is provided for each phase, it has the advantage of also being able to detect phase‐to‐phase faults as well as ground faults. Modern relays provide overcurrent elements for all phases and therefore this arrangement is now preferred.

Through‐Fault Stability

During out‐of‐zone faults, the fault current will be supplied by all the bus incomers (current sources), delivering it to the faulted outgoing feeder. Consequently one current transformer may see the entire fault current, the transient component of which may be sufficient to cause it to saturate, while the others carrying considerably less current, remain unsaturated. The result of this will be to generate a spill current in the relay, which may result in an unwanted trip. One way to avoid this is through the use of a high impedance relay. This may be a low impedance current calibrated relay with a series connected stabilising resistor, making it behave as a voltage sensitive device, or a high impedance voltage calibrated relay, in which the stabilising resistance is fitted internally.

Figure 12.18 shows such a situation in which a ground fault occurs on an outgoing feeder and the fault current is delivered by two incomers; the faulted feeder CT is assumed to fully saturate as a result. In this state, this CT sees no flux change throughout much of each cycle, and it consumes most of its current internally as magnetising current, therefore as a first approximation it can be represented as a short circuit, as shown in the figure.

The reflected fault current in the secondary circuit, divides between the relay, including its stabilising resistance R_R, and the wiring resistance R_wC in series with saturated CT winding resistance R_ctC. The currents flowing into the CT () and the relay (I_R) therefore become:

Therefore by choosing R_R appropriately, the spill current can be made as small as necessary to preserve the through‐fault stability of the scheme. The voltage developed across the relay and its stabilising resistance (V_R) is therefore determined by the maximum fault current and the maximum wiring and winding resistances of any saturated CT. This is referred to as the stability voltage and is given by:

(12.6)

Where a voltage calibrated high impedance relay, like the MFAC14 is used (one including an internal stabilising resistor), the setting voltage V_S, must be greater than V_R. On the other hand, if a current calibrated low impedance relay such as the MCAG14 is used instead, together with an external stabilising resistor, the setting current , must be chosen so that:

(12.7)

Finally, the assertion that a fully saturated CT can be replaced by a short circuit is in practice not quite correct. Every half cycle the flux density must swing from its saturation value in one direction to that in the other, creating a rapid change of 2B_sat. During this interval the CT produces a short pulse of current, one that is rich in harmonics. Many circulating current relays tune the relay winding to the fundamental frequency, making it insensitive to the harmonics created by saturated CTs. The inclusion of a capacitor in the relay circuit also makes the relay less sensitive to the transient DC component.

In the case of in‐zone faults, all CTs carrying a fraction of the fault current drive the relay and its stabilising resistance. Since the latter has been chosen to preserve through‐fault stability, it is quite possible that with the fault current from several CTs flowing, the burden presented may cause an excessively high voltage to develop, one which may threaten the CT winding insulation. In such circumstances while it is likely that some saturation will occur, it is usual to fit a non‐linear zinc oxide resistor (known as a Metrosil), in parallel with the relay and its stabilising resistor to limit the amplitude of any voltage developed. This device consumes very little current when the relay voltage is low, and effectively clamps the peak voltage by consuming large amounts when the voltage rises, by which time the relay will have operated.

Effective Primary Current Setting

During fault conditions all parallel connected CTs are excited to the setting voltage (or possibly beyond), and therefore the effective setting current should also include the magnetising current consumed within each CT which, depending on the magnitude of the setting voltage, may not be negligible. The effective primary current required to operate the scheme is increased in proportion, and is given by:

(12.8)

where I_P is the primary tripping current, I_S is the setting current, N is the CT turns ratio, n is the number of parallel connected CTs in the scheme and I_eS is the excitation current at the relay setting voltage.

In order to obtain a fast fault clearance time the effective primary tripping current should be less than 30% of the minimum phase‐to‐earth fault current. In large switchboards, with many parallel connected CTs, the cumulative effect of the magnetising currents may lead to an excessive tripping current, which is another reason why the three‐element arrangement of Figure 12.17b is preferred.

High Impedance Bus‐Zone Applications

High impedance busbar protection requires the secondary windings of all current transformers to be connected in parallel, together with the relay and its stabilising resistor. These CTs must be virtually identical and dedicated to the scheme; they cannot be shared with other protection functions. The relay effectively becomes an over‐voltage device, sending trip signals to all breakers within its protection zone when its setting voltage is exceeded. The wiring for such a scheme is straightforward, and its cost is relatively low. These features make it attractive for simple bus arrangements like the sectionalised bus in Figure 14.7 which can be divided into two zones, one on either side of the bus coupler. For more complex arrangements where feeders can be switched from one bus to another, a secondary replica of the primary feeder arrangement is required and in such cases low impedance bus‐zone protection is preferred.

Low Impedance Bus‐Zone Protection

Low impedance differential relays can also be used in bus‐zone protection schemes and are arranged in a similar manner to the transformer differential protection discussed earlier. Stability is similarly assisted through the use of bias windings or through the use of a software implemented bias characteristic generated within a numerical relay.

The choice between high impedance and low impedance bus‐zone protection depends to a large extent on the complexity of the particular bus application. In the case of complex reconfigurable busbars, where incomers and outgoing circuits can be switched from one bus to another, low impedance bus‐zone protection is generally preferred. The reason for this choice is partly due to the differences between these protection schemes and partly due to the flexibility of modern numerical relays.

Consider applying high impedance bus‐zone protection to the double bus arrangement in Figure 14.8. In this configuration, incoming and outgoing circuits can be easily moved from one bus to another, by placing the associated circuit breakers into the appropriate cubicles, or by closing the appropriate isolator. Since separate zones of protection are required for each bus, the CT secondary currents must be reconfigurable to exactly mirror the primary circuit arrangement. This effectively demands that the CT currents be automatically switched whenever the bus is reconfigured. While it is possible to achieve this using the auxiliary switches within the circuit breaker cubicles (activated when the breaker is inserted) the switching of CT currents is fraught with danger. The risk of a switch mal‐operation may be relatively low, but its consequences for the protection scheme are high. An open circuited CT will generate dangerously high voltages and its absence from the scheme will result in a spill current likely to trip the relay.

Low impedance schemes can be configured in a way that does not require the switching of CT currents. While these appear to operate on a similar principle to high impedance schemes, the significant difference between them is that a low impedance relay is sensitive to current and not voltage. While fault current can be detected by conventional low impedance relays like those in Figure 12.17, it need not be. So long as the vector sum of the CT secondary currents can be calculated, the presence of an in‐zone fault can be determined with certainty. This fact and the flexibility of numerical relays has enabled low impedance bus‐zone protection to be applied to reconfigurable busbar arrangements without the need to switch CT currents.

While low impedance schemes do require an image of the primary circuit to be generated (often called a dynamic bus replica), this is not in the form of switched CT currents. Instead it is sufficient to supply a numerical relay with a digital indication as to which bus a given breaker is connected, from which it can build a software replica of the primary circuit. Position information is obtained from auxiliary switches associated with the isolator connecting the breaker to the bus, either from within the circuit breaker cubicle in the case of metal clad switchgear, or directly from within the isolator, in the case of an aerial switchyard.

The secondary current from each CT is also supplied to the relay as shown in Figure 12.19, and from the dynamic bus replica the relay vectorially sums all the CT currents appropriate to each zone. When a circuit breaker is relocated to another zone its CT current will appear in the associated summation, therefore any residual current arising indicates the presence of an in‐zone fault, for which all the breakers in that zone must be tripped.

Should an auxiliary switch mal‐operate and provide the relay with incorrect position information, a false trip could result. This contingency is mitigated by the use of both normally open and normally closed position switches, the logical states of which must agree if a valid breaker position is to be inferred. When these disagree, the relay alarms, and switching operations must cease until the problem has been rectified. As mentioned earlier, bus‐zone protection schemes also generally include a check relay to confirm the existence of a fault; this will prevent a protection mal‐operation in the presence of a position error.

Because a low impedance relay has access to the current from each CT, it can implement any necessary bias characteristic, generating a bias (or restraining) current from the algebraic sum of the CT currents, and the operating (or differential) current, from their vector addition. This information is used to determine a dynamic trip threshold for the relay of the kind shown in Figure 12.14.

Other functions such as breaker fail protection can also be provided by a low impedance numerical relay, where if a post‐trip current is detected in any circuit breaker, the tripping of all breakers supplying the failed device is triggered. Numerical relays also permit the use of current transformers having different ratios and because each CT is only burdened by the relay, it can if necessary be used for other functions as well.

Current Transformer Requirements

High impedance schemes require a greater degree of CT performance and in order to preserve the through‐fault stability, the setting voltage for a high impedance scheme V_S must exceed the stability voltage V_R. As a result, the CT knee point voltage (E_k) must also exceed the setting voltage. Relay manufacturers generally require a safety margin of at least a factor of two, yielding:

(12.9)

IEC class PX or TP current transformers are suitable for use in both high and low busbar protection applications as are ANSI class C transformers, since the knee point voltage and the secondary winding resistance are readily available.

Arc Flash Busbar Protection

The requirement to insulate the entire switchboard frame and cable glands from ground is expensive and must be maintained throughout the life of the switchboard. This has led to a decline in the use of frame leakage schemes. In recent years fibre optic sensors capable of detecting the light from an arcing fault have become available. Together with arc flash monitor relays, these provide a simple alternative to frame leakage and circulating current based bus protection systems. They can be retro‐fitted to existing metal clad switchgear or fitted to new switchgear during manufacture and can detect both phase‐to‐ground and phase‐to‐phase faults and provide very fast fault clearance times.

Arc flash sensors fall into two categories: the first is a naked optical fibre, capable of collecting light from an arc flash anywhere along its length and delivering it to an arc flash relay. This type of sensor can detect a fault anywhere throughout the bus chamber of a switchboard. The second type of sensor detects light from an arc flash through a lens at the end of an optical fibre. These are used to detect arc flash energy in specific locations such as in circuit breaker chambers or in cable termination enclosures.

Arc fault detection protection schemes also require some form of check protection to prevent nuisance trips. The check function is usually provided by a neutral earth fault or phase overcurrent relay or an overcurrent element within the arc monitor relay itself. This element is not required to time out or to generate a trip, but simply to pick up, confirming the existence of an internal fault. The arc monitor relay will issue a trip command when an arc has been detected and a fault current has been detected within the switchboard. In a similar fashion to frame leakage and high/low impedance bus protection schemes, arc fault protection can also be applied to sectionalised busbars, so that the minimum equipment is tripped in the event of a fault.

Co‐ordination of Distance Relays

The actual reach achieved by a distance relay depends upon the accuracy of the information supplied to it. Errors in current and voltage information, the impedance values programmed into the relay and even those in the relay itself, all contribute to the tolerance on the reach achieved. Satisfactory coordination of distance relays is achieved by employing a multi‐zone approach to reach and trip delay setting. Three zones of protection are usually employed, although modern digital and numerical relays offer the possibility of more.

Zone 1 is typically set at 70–80% of the length of the line being protected. This is to ensure that the relay cannot see beyond the end of its line. Zone 1 protection is directional, responding only to faults along the line, allowing the Zone 1 element to trip instantaneously.

Zone 2 is also directional and covers the remainder of the line. It is typically set to 120% of the line impedance and therefore it also sees part of the next line. It must therefore be time graded with any other relays that lie within its reach and provides backup protection for the remote busbars.

Zone 3 is often set to reach beyond Zone 2 with correspondingly longer delay times. The Zone 3 impedance circle can also be offset to include the origin as shown in Figure 12.24a. This will permit the relay to provide backup protection for the local busbars immediately behind it.

Lenticular Impedance Characteristic

Figure 12.24a also includes a typical load impedance characteristic. In cases where the Zone 3 reach is large, there is a risk that the load may encroach on the relay characteristic, possibly causing nuisance trips, particularly under emergency conditions when the line may be heavily loaded. To avoid this, a lens‐shaped (lenticular) characteristic may be used instead, as shown in Figure 12.24b. This can be obtained from the intersection of two mho characteristics with one displaced towards the X axis, thereby providing an improved immunity against load encroachment.

Quadrilateral Impedance Characteristic

In addition to the circular impedance characteristics discussed, a quadrilateral characteristic similar to that in Figure 12.25 is also available. This is useful for the earth fault protection of MV and HV lines where vegetation can give rise to highly resistive faults. The relay’s forward reach along the line is determined by the line impedance at point B which can be set independently from its resistive reach at C. Care must be taken in the choice of this parameter so as to avoid problems of load encroachment.

Distance Relay Impedance Measurements

Phase voltage and line current information is supplied to an impedance relay which sees an apparent secondary impedance given by:

(12.11)

The relay’s reach must therefore be specified in terms of the equivalent secondary impedance of the portion of the line to be protected.

Distance relays evaluate the positive sequence impedance Z₁, between the relay and the fault. In the case of three‐phase, phase‐to‐phase and phase‐to‐phase‐to‐ground faults the relays measure this impedance by evaluating the ratio of the line‐to‐line voltage and the current flowing between the faulted phases. For faults involving the A and B phases for example, Z₁ is calculated from:

(12.12)

In the case of phase‐to‐ground faults the fault current must return to the neutral terminal via the ground impedance Z_g (which includes any neutral earthing impedance that may be fitted), as shown in Figure 12.26a. In this case the ratio of the phase‐to‐ground voltage and the phase current will not yield the positive sequence impedance of the line, since . However, modifying the impedance equation as shown below will result in the correct evaluation of the positive sequence impedance.

(12.13)

where K_n is called the neutral compensation factor , which is often approximated as a real number, and 3I₀ is the residual current seen by the relay.

The replica phase voltage V_A is generated in the relay by the sum of two voltage drops: I_A through the positive impedance Z₁ and the residual current 3I₀ through Z_g. The latter is adjusted so that for an earth fault at the maximum reach of the relay, V_A equals the voltage measured at the relay, thus .

By comparing the zero‐sequence networks in Figure 12.26b and remembering that 3I₀ flows in Z_g, we obtain the relationship and therefore . A distance relay requires three impedance comparators for phase‐to‐phase faults and three for phase‐to‐ground faults, for each of its zones of protection.

Voltage and Current Transformers

As mentioned, voltage transformers provide a distance relay with phase‐to‐ground voltage information. The sudden loss of potential from a voltage transformer may suggest a close‐in fault sufficient to instantly trip the relay, or alternatively it may simply mean the failure of either a VT primary or secondary fuse. Voltage supervision circuitry within a distance relay is necessary to prevent spurious trips in the latter case, while ensuring that the relay operates correctly in the former.

Impedance relays evaluate the line’s positive sequence impedance from voltages and currents at the system frequency only, and therefore any transient DC current present will be ignored. However, it is important that the DC current does not saturate the CTs and thereby corrupt the underlying symmetrical fault current information. The knee point voltage E_k should therefore be chosen according to:

(12.14)

IEC class PX or PXR CTs or IEEE class C CTs would be suitable for distance relay applications (see Chapter 9).