POWER TRANSFORMER FAULT CLEARING

A faulted transformer can be separated from its power source by devices such as circuit breakers, power operated disconnect switches, circuit switchers, and fuses, or by remote tripping of fault-interrupting devices.

In addition to separating the transformer from its power source, due consideration should be given to tripping oil pumps and fans to reduce their possible adverse effects in sustaining or  spreading transformer oil fire and to halt circulation of contaminants in the oil resulting from the arc.

Determination of the type of fault-clearing devices to be used should consider factors such as:

a)Installation and maintenance cost
b)Fault-clearing time relative to fire hazard and repair or replacement costs of the transformer
c)System stability and reliability
d)System operating limitations
e)Device interrupting capability

Relay tripping circuits
Usually a transformer protective relay operation requires a careful inspection for the cause of tripping before any attempt is made to reenergize the transformer. Usually two or more breakers or switching devices must be tripped. Therefore, tripping is usually done by a lockout relay that also blocks closing circuits and must be reset manually.

For a large transformer having several protective relays, two lockout relays and dc power supplies are often used. If both lockout relays trip the same breakers, the differential relays may operate one lockout relay and the sudden-pressure relay and overcurrent relays, including instantaneous units, may operate the second lockout relay for the greatest redundancy.

If the two lockout relays perform different tripping functions, then a different assignment of protective relays to each lockout relay may be desired. The lockout relay contacts may initiate breaker failure relaying and may be supplemented by a small self-reset relay operated in parallel with the lockout relay coil for redundancy.

Relays should be connected to trip fault-interrupting devices that will clear faults in the zone that the relay is intended to protect. For example, the relay devices 87G in Figure 8, 67G in Figure 9, and 51GB and 51GT must trip the transformer high-side (source) circuit breaker in order to clear a ground fault on the low-voltage side of the transformer.

TRANSMISSION LINE PROTECTION LENGTH CONSIDERATIONS BASICS

Short lines (SIRs > 4)
Typical protection schemes and communication channels used are as follows:
— Current differential (FO, PW, AT, MW)
— Phase comparison (FO, PW, AT, PLC, MW, R)
— POTT (FO, PW, AT, PLC, MW, R)
— Directional comparison blocking (FO, PW, AT, PLC, MW, R)

Short lines result in a small current magnitude difference and minimal voltage drop difference between close-in and remote faults. Non-pilot overcurrent relays usually cannot be set to discriminate between internal and external line faults. Non-pilot distance relays may be able to discriminate between internal and external line faults for lines with an SIR as high as 20.

However, it may not be possible to coordinate non-pilot distance relays on adjacent long lines with the short line distance relay zones. Therefore, pilot relay schemes utilizing communication channels are used. The most effective forms of pilot protection for short lines include current differential, phase comparison, POTT, and directional comparison blocking schemes.

None of these schemes require distance elements to be set for less than the line impedance. Schemes utilizing Zone 1 distance elements should provide the capability to handle arc resistance or fault impedance, which can be significant compared to the line impedance.

Current differential and phase comparison systems do not provide remote backup for adjacent system elements. However, non-pilot overcurrent and distance relay schemes can be applied as backup protection, provided time delays are adjusted to provide coordination.

Medium lines (SIRs < 4, but > 0.5)
Typical protection schemes used are as follows:
— Phase comparison (AT, PLC, MW)
— Directional comparison blocking (AT, PLC, MW)
— Permissive underreaching transfer trip (FO, MW)
— POTT or unblocking (FO, MW, PLC)
— Step distance
— Step or coordinated overcurrent
— Inverse time overcurrent
— Current differential

Lines with SIRs less than four allow more effective relay discrimination. Relay schemes utilizing under reaching elements can now be set. Zone 1 can be set to underreach the remote end and still protect 80 90% of the line.

The last 10–20% of the line can be protected by a Zone 2 overreaching element. Since selectivity and sensitivity can usually be met, the speed with which a fault needs to be cleared may require the application of pilot relaying. If slow fault clearing is acceptable, simple step distance or overcurrent relays can be applied.

Long lines (SIRs < 0.5)
Typical protection schemes used are as follows:
— Phase comparison (PLC, MW)
— Directional comparison blocking (PLC, MW)
— PUTT, POTT, or unblock (AT, PLC, MW)
— Step distance
— Step or coordinated overcurrent

Long lines have small SIRs, and most long lines are extra high voltage (EHV) or ultra high voltage (UHV) lines. EHV and UHV lines almost always require high-speed tripping of all terminals for stability purposes and to minimize damage caused by the fault.

Long lines can also have other system elements included in the line, such as series capacitors. This makes the total line impedance variable under certain conditions and introduces transient behavior that makes selectivity difficult. Phase and directional comparison protection schemes over PLC or MW are well suited to this type of application.d to provide coordination.

PER UNIT CALCULATION - FAULT CALCULATION TOOL

Power system calculations can be done using actual voltages and currents or using per-unit representations of actual quantities. While performing a calculation in actual quantities makes sense occasionally, the vast majority of calculations are done in per-unit.

The discussion in this assumes a familiarity with the per-unit method; but, to avoid confusion, definitions of important parameters are given in Table 1. Equivalent three-phase values are usually used in practice, but an understanding of the mathematics presented in Table 1 relies on a careful interpretation of base values as single-phase quantities.

Table 1 - Per unit base parameters


Example A: Calculate the maximum three-phase current measured by an instantaneous relay, fuse, or series trip device on the 13.8 kV feeder serving the 2 Mvar power factor capacitor on Bus 4.

Figure 2 depicts a portion of the system one-line diagram and shows a fault in the 13.8 kV, 2 Mvar capacitor feeder on Bus 4. The calculated current magnitudes are

Distribution of instantaneous currents for fault on capacitor feeder in sample system

— 19 203 A, rms asymmetrical
— 12 410 A, rms symmetrical

A moderate asymmetrical offset (X/R = 15.65) is present due primarily to the close proximity of the generator, which contributes 4080 A of symmetrical current. The largest contributor, of course, is the utility, with other sources making up the difference.

The capacitor itself does not contribute current to the fault. Present practice ignores any capacitor contribution to system short-circuit currents on the basis that it occurs so quickly and is out of phase with system currents.

In this instance, also, a protective device in the capacitor feeder would detect currents flowing from the system to a fault on the feeder, or current flowing from the capacitor to the system, but not both, again illustrating that the analyst should carefully consider what constitutes the branch current measured by the protective device of interest.

Figure 3 shows this same fault condition, but with the impedances adjusted to calculate the long-time fault magnitude. The current has decayed to less than 9185 A symmetrical and the generator contribution is down to 2721 A. Also, the contributions from motors on the system have dropped to zero.

Figure 3 Calculated distribution of long-time relay current to fault on capacitor feeder in sample system

NOTE—The current is purely symmetrical and the contributions from sources other than the generator and utility have decayed away.

PHASE TO PHASE SHORT CIRCUIT BASIC INFORMATION

Phase to phase short circuit is also known as line to line fault.

Faults involving abnormal conduction from one phase to another without involving ground may be represented using the interconnection of Figure below.

Add caption

Sequence interconnections for a phase-to-phase fault not involving ground

The zero-sequence network is not involved and may be ignored. As for the three-phase condition, Zf would normally be the fault impedance, but it could also be a phase-to-phase load impedance if the problem of interest is the response of the system to a phase-to-phase-connected single-phase load.

An interesting and often useful relationship develops from Figure.

In the special case of a zero-impedance phase-to-phase fault, if the negative-sequence impedance of nearby synchronous machines is approximately equal to the positive-sequence impedances of such machines or if no synchronous generators are nearby at all, then the phase-to-phase fault current will be 0.87 times the corresponding three-phase fault current.

This relationship can be proven by calculating sequence currents in Figure for the specified condition and then converting them to phase currents using symmetrical components equation.

GROUND FAULT PROTECTION OF POWER TRANSFORMERS

Faults in Δ-connected transformer windings
A residual relay, device 51N, as shown in Figure 8 and Figure 9, will detect ground faults within the Δ winding of the transformer and in the phase conductors between the CTs and the winding, when an external source of zero sequence current is available. Instantaneous overcurrent relays may be used, but sensitive settings will probably result in incorrect operations from dissimilar CT saturation and magnetizing inrush.

This can be avoided by using a short-time overcurrent relay with a sensitive setting. The scheme is particularly valuable in plants or systems where the transformers are remote from the circuit breakers. By using CTs at the circuit breaker, sensitive detection is obtained for cable, bus, Δ winding, and bushing faults.

A single window or doughnut CT supplying an instantaneous relay (as commonly used in motor protection) is secure, but is limited to cases of low and medium voltages where all three conductors can be fitted through the CT window.

Figure 8—Complete ground fault protection of a Δ-Y transformer using a residual overcurrent and differentially connected ground relay

Figure 9—Complete ground fault protection of a Δ-Y transformer using a residual overcurrent and directional relay


Faults in grounded Y-connected transformer windings
To successfully detect faults in grounded Y-connected transformer windings, the relay system should discriminate between faults internal and external to the protected zone. The ground differential relay, device
87G in Figure 8, typically an overcurrent relay, or the directional ground relay, device 67G, connected as in Figure 9, is satisfactory. Both relay schemes will operate correctly for any internal ground faults with the circuit breaker in the circuit to the grounded Y winding open or closed.

They will operate correctly with an external zero sequence current source, and they will not operate for external ground faults. The auxiliary CT is necessary if the phase and neutral CTs are of different ratio. Both schemes are particularly applicable where the ground fault current is limited and phase differential relays may not respond.

The device 67G operating coil or element current is zero for an external fault with CT ratios matched. Therefore, it is wise to select the auxiliary CT ratio to give positive nontrip bias to device 67G for an external ground fault (auxiliary CT secondary current slightly greater than the transformer neutral CT secondary current). Unequal CT action can produce residual error current during external phase faults. No transformer neutral current is produced and sensitive relays could misoperate.


Case ground
On a grounded neutral system, it is possible to isolate the transformer case from ground except for a single point. A CT and overcurrent relay at this grounding point would detect any internal ground fault or bushing flashover.

Although effective, several problems are encountered. The system should be tested periodically to determine that no accidental grounds have been added.

Incorrect operations can result from accidental grounds from power tools and transformer auxiliary equipment. Careful coordination between auxiliary equipment circuit breaker or fuse curves, arrester characteristics, and a time overcurrent trip relay can minimize this danger.

Impedance-grounded system
Transformer differential relays may not be sensitive enough to operate on ground faults where the transformer bank or system is grounded through an impedance. In these cases, it may be necessary to apply a sensitive time overcurrent relay in the transformer impedance-grounded neutral or a time overvoltage relay connected across the neutral impedance.

These relays should be coordinated with any feeder and line protection relays that they may overlap. It is possible to provide high-speed protection and to avoid the need for coordination by using sensitive product-type relays, which are connected to trip only for ground in the protected zone.

An overcurrent relay connected to a neutral CT is torque-controlled by the blocking contacts of a plunger-type instantaneous relay in the neutral of the main breaker CTs. Since the transformer differential relays may not operate for such ground faults within the differential zone, these ground fault relays must trip the source-side circuit breakers.

Ground relays also used for sensitive ground fault protection
The primary advantage of ground relays over phase relays is their sensitivity. In systems where the ground fault current is purposely limited, their use may be vital. Ground relays can normally be applied with sensitivities of 10% or less of full load current.

This compares very favorably with differential relays, whose pickup current may be from 20–60% of full load current under the most advantageous conditions. It is common practice in the United Kingdom and other countries influenced by the U.K. to protect all power transformers with the restricted earth relay.

The term restricted earth is an expression referring to a sensitive ground relay system that is designed to detect ground faults within a well-defined protective zone (similar to that in Figure 8)

MAGNETIZING CURRENT INRUSH - PROTECTION CONSIDERATION

This is a phenomenon that causes the violation of the basic principle of differential relaying since the magnetizing branch of the transformer can have a very low impedance without a transformer fault. Current produced by magnetizing inrush can reach many times the transformer rating, and these currents appear in the differential relay.

The inrush current duration can range from a few cycles up to many seconds. Although usually considered only in conjunction with the energizing of a transformer, magnetizing current inrush can be caused by any abrupt change of magnetizing voltage.

Such transients include the occurrence of a fault, the removal of a fault, the change of character of a fault (for example, the change from a single phase- to-ground fault to a two-phase-to-ground fault), and out-of-phase synchronizing.

Thus, a desensitizing scheme that is operative only when energizing a transformer may not be adequate. There are several conditions that cause particularly severe magnetizing inrush phenomena.

One involves the energizing of a transformer at a station at which at least one other transformer is already energized. The inrush phenomenon will involve transformers that are already energized as well as the transformer being energized.

This inrush transient may be particularly long in duration. It is important to realize that the inrush into the transformer being energized occurs during the opposite half-cycle to that of the already energized transformer.

Thus, the net inrush into all transformers may approximate a sine wave of fundamental frequency, and therefore not operate the harmonic restraint unit of a differential relay if it is protecting both parallel transformers. Another inrush phenomenon involves the energizing of a transformer by means of an air switch.

Arcing of the switch can result in successive half-cycles of arc of the same polarity. Thus if the first half-cycle results in substantial residual magnetism in a transformer core, succeeding half-cycles can cause a cumulative increase in residual magnetism, each time resulting in a more severe inrush.

Two important characteristics of magnetizing inrush current are
a)It contains substantial harmonics, particularly the second harmonics. These harmonics are not always present in high quantities in all three phases.

b)That there is always a time during each cycle where the current magnitude is almost zero. This time is always greater than a quarter cycle.

The harmonic content of the inrush current depends on various factors such as remnant flux in the core, switching angle, and load on the transformer. Harmonic analysis of the inrush current during the events in

Figure below shows that the second harmonic content of the inrush current is sensitive to all these conditions. The second harmonic content reduces appreciably with increasing load at lagging power factor conditions.

Typical magnetizing inrush current wave

PROTECTIVE RELAY CURRENT TYPES - BASICS

Relay current
Two characteristics of power transformers combine to complicate detection of internal faults with current operated relays

a)The change in magnitude of current at the transformer terminals may be very small when a limited number of turns are shorted within the transformer.

b)When a transformer is energized, magnetizing inrush current that flows in one set of terminals may equal many times the transformer rating. These and other considerations require careful thought to obtain relay characteristics best-suited to the particular application.

Minimum internal faults
The most difficult transformer winding fault for which to provide protection is the fault that initially involves one turn. A turn-to-turn fault will result in a terminal current of much less than rated full-load current.

For example, as much as 10% of the winding may have to be shorted to cause full-load terminal current to flow. Therefore, a single turn-to-turn fault will result in an undetectable amount of current.

Maximum internal faults
There is no limit to the maximum internal fault current that can flow, other than the system capability, when the fault is a terminal fault or a fault external to the transformer but in the relay zone. The relay system should be capable of withstanding the secondary current of the CT on a short-time basis.

This may be a factor if the transformer is small relative to the system fault and if the CT ratio is chosen to match the transformer rating.

Through-faults
Fault current through a transformer is limited by the transformer and source impedance. While current through a transformer thus limited by its impedance can still cause incorrect relay operations or even transformer failure, CT saturation is less likely to occur than with unlimited currents.

The above favorable aspect may disappear if the transformer protective zone includes a bus area with two or more breakers on the same side of the transformer through which external fault current can flow with no relationship to the transformer rating. An example is a transformer connected to a section of a ring bus with the transformer protection including the ring bus section.

SYMMETRICAL COMPONENTS TUTORIALS FOR PROTECTIVE RELAYING

Symmetrical components are applied to calculations of unbalanced fault currents and voltages and in rotating machine analysis. Theory of symmetrical components can be briefly stated thus: a coplanar vector is defined by the position of its terminal and length and has 2 degrees of freedom.

A three-phase balanced system has 2 degrees of freedom because the current or voltage vectors (phasors) are displaced from each other by equal angles of separation of 120◦ and are of equal length. A three-phase unbalanced system of currents or voltages has 6 degrees of freedom because the vectors are of varying length at varying displacement angles from each other.

Such an unbalanced system can be resolved into three symmetrical systems, each system having three vectors with 2 degrees of freedom. Positive-sequence system is a set of balanced three-phase components of the same phase sequence as the original unbalanced set.

Negative-sequence system is a set of three-phase components of opposite phase sequence to the positive sequence system but vectors (phasors) of the same magnitude. Zero-sequence system consists of three single-phase components of the same magnitude and cophasial.

These are related by the following equations:

 

and

 

Where Va, Vb, and Vc are the original unbalanced voltages; a is a unit vector operator that rotates 120◦ in the counterclockwise direction; and V+a, V −a, and V0a are the positive, negative, and zero sequence components of the original unbalanced set.

Characteristics of Sequence Components

In a three-phase wye connected and ungrounded system, no zero sequence current flows. If the wye point is grounded, neutral carries the out-of-balance current. In a delta connection, no zero sequence currents can appear in the line currents.

In a balanced three-phase system with balanced loads, only positive sequence currents can flow. Negative sequence currents are set up in circuits of unbalanced impedances and voltages.

In symmetrical circuits, currents and voltages of different sequence do not affect each other (i.e., the positive sequence currents produce only positive sequence voltage drops and the theorem of superposition applies). Sequence impedance networks must be constructed for unbalanced fault current calculations and data input to digital computers.

As an example, the single-line-to-ground fault is given by the expression:

 

 

Where Ig is the single-line-to-ground fault current; E is the line-to-neutral voltage; and Z+, Z−, and Z0 are the positive, negative, and zero sequence impedances to the fault point.

THE NATURE OF SHORT CIRCUIT CURRENTS – PROTECTIVE RELAYING CONSIDERATION

Under normal system conditions, the equivalent circuit of Figure 2-1 may be used to calculate load currents. Three impedances determine the flow of current.

Zs and Zc are the impedances of the source and circuit, respectively, while Zl is the impedance of the load. The load impedance is generally the largest of the three, and it is the principle determinant of the current magnitude.

Load impedance is also predominantly resistive, with the result that load current tends to be nearly in phase with the driving voltage. A short circuit may be thought of as a conductor that shorts some of the impedances in the network while leaving others unchanged.

This situation is depicted in Figure 2-2. Because Zs and Zc become the only impedances that restrict the flow of current, the following observations may be made:

a) The short-circuit current is greater than load current.

b) Because Zs and Zc are predominately inductive, the short-circuit current lags the driving voltage by an angle approaching the theoretical maximum of 90°. 

 

The change in state from load current to short-circuit current occurs rapidly. Fundamental physics demonstrate that the magnitude of current in an inductor cannot change instantaneously. This conflict can be resolved by considering the short-circuit current to consist of two components:

— A symmetrical ac current with the higher magnitude of the short-circuit current

— An offsetting dc transient with an initial magnitude that is equal to the initial value of the ac current, but which decays rapidly

The initial magnitude of the dc transient is directly controlled by the point on the voltage wave at which the short circuit occurs. If the short circuit occurs at the natural zero crossing of the driving voltage sinusoid, the transient is maximized.

However, the transient is a minimum if the fault occurs at the crest of the voltage sinusoid. At any subsequent time, the magnitude of the dc transient is determined by the time constant of the decay of the dc, which is controlled by the ratio of reactance to resistance in the impedance limiting the fault.

Equation (2-2) can be used to calculate the instantaneous magnitude of current at any time. For the protection engineer, the worst case initial current includes the full dc transient.

The driving voltage depicted in Figure 2-1 and Figure 2-2 is the Thevenin equivalent opencircuit voltage at the fault point prior to application of the short circuit. This voltage includes sources such as remote generators with voltage regulators that maintain their value regardless of the presence of a short circuit on the system as well as nearby sources whose voltages decay when the short circuit is present.

The amount of decay is determined by the nature of the source. Nearby generators and synchronous motors with active excitation systems sustain some voltage, but because the short circuit causes their terminal voltage to drop, the current they produce is gradually reduced as the fault is allowed to persist.

At the same time, induction motors initially participate as short-circuit current sources, but their voltages decay rapidly as the trapped flux is rapidly drained. Figure 2-3 shows the generic tendencies of various kinds of short-circuit current sources and a composite waveform for the symmetrical ac current decay.  

 

Figure 2-4 depicts the most realistic case of the decaying symmetrical ac current combined with the decaying dc transient. From this figure, a generalized short-circuit current may be described in the following terms:

— High initial magnitude dc transient component of current, which decays with time

— High initial magnitude symmetrical ac current, which diminishes gradually with time

— Symmetrical ac current lags driving voltage by a significant angle, approaching 90°