Home / Technical Articles / Transformers / 3 Bad Conditions That Cause False Differential Current In Power Transformer

Transformer differential protection

When applying differential protection of power transformer, three bad conditions must be considered which can lead to unbalance to the currents applied to the relay (compared with the expected currents when power flow into the transformer is equal to the power flow out of the transformer).

3 Conditions That Cause False Differential Current In Power Transformer
3 Conditions That Cause False Differential Current In Power Transformer (photo credit: Geoff Collins)

Let’s talk about following conditions:

  1. Magnetizing inrush current
  2. Overexcitation, and
  3. Current transformer saturation

1. Magnetization Inrush

When system voltage is applied to a transformer at a time when normal steady-state flux should be at a different value from that existing in the transformer, a current transient occurs, known as magnetizing inrush current.

This phenomenon is illustrated in Figure 1 for a transformer with no residual flux.

In the figure, the transformer is energized when the system voltage is zero. With the highly reactive circuit involved, the flux Φ should be at or very near negative maximum, but the transformer has no flux. Thus, the flux must start at zero and reach a value of in the first cycle period. To provide this flux, excursion requires a large exciting current, as shown.

Transformers are normally operated near saturation for best efficiency, so values of flux greater than normal Φ result in severe saturation and a large exciting current.

Magnetizing inrush current phenomenon (no residual flux initially in transformer)
Figure 1 – Magnetizing inrush current phenomenon (no residual flux initially in transformer)

If a transformer has been energized previously, there is a high possibility that on de-energization some flux ΦR was left in the iron. This could be positive or negative. If in Figure 1, a residual flux of R had existed from an earlier energization, the flux maximum required would have been 2Φ + ΦR, resulting in a higher maximum magnetizing inrush current. If ΦR had been negative, the maximum required flux would be 2Φ – ΦR with less inrush current.

This is a random phenomenon. If the transformer had been energized at or near maximum positive voltage (see point d in Figure 1), the flux requirement at that time is zero.

Thus, normal exciting current would flow with negligible or no transient inrush. Normal exciting currents for power transformers are on the order of 2%–5% of full-load current.

The maximum initial-magnetizing current may be as high as 8–30 times the full-load current. Resistance in the supply circuit and transformer and the stray losses in the transformer reduce the peaks of the inrush current such that, eventually, it decays to the normal exciting current value. The time constant varies from about 10 cycles to as long as 1 min in very high-inductive circuits.

The factors involved in the inrush, in addition to the time point of energization with relation to the flux requirements, are:

  • Size of the transformer,
  • Size and nature of the power system source,
  • Type of iron in the transformer, previous history, and
  • L/R ratio of the transformer and system.

In a three-phase circuit, some inrush will always occur in one or two and generally all three phases, with the voltages at 120° apart, although it may or may not be maximum or zero in one of the phases.

Figure 2 shows a typical magnetizing inrush current trace when a transformer bank is energized from either the wye- or delta-connected terminals.

Typical magnetizing inrush current to transformers: (a) A-phase current to wye-connected windings; (b) A-phase current to delta-connected windings.
Figure 2 – Typical magnetizing inrush current to transformers: (a) A-phase current to wye-connected windings; (b) A-phase current to delta-connected windings.

Some years ago, studies indicated that the second-harmonic component of the inrush wave was 15% or more of the fundamental current. In recent years, improvements in core steel and design are resulting in transformers for which all inrush current harmonics are less, with possibilities of the second harmonic being as low as 7%.

Magnetizing inrush can occur under three conditions and are described as:

  1. Initial,
  2. Recover, and
  3. Sympathetic

a. The initial-magnetizing inrush

1. The initial-magnetizing inrush may occur when energizing the transformer after a previous period of de-energization. This was described earlier and has the potential of producing the maximum value.

b. Recovery inrush

During a fault or momentary dip in voltage, an inrush may occur when the voltage returns to normal. This is called the recovery inrush.

The worst case is a solid three-phase external fault near the transformer bank. During the fault, the voltage is reduced to nearly zero on the bank; then, when the fault is cleared, the voltage suddenly returns to a normal value.

This may produce a magnetizing inrush, but its maximum will not be as high as the initial inrush because the transformer is partially energized.

c. Sympathetic inrush

A magnetizing inrush can occur in an energized transformer when a nearby transformer is energized. A common case is paralleling a second transformer bank with one already in operation. The DC component of the inrush current can also saturate the energized transformers, resulting in an apparent inrush current.

This transient current, when added to the inrush current of the bank that is energized, provides an offset symmetrical total current that is very low in harmonics. This would be the current flowing in the supply circuit to both transformer banks.

Go back to contents ↑

2. Overexcitation

The flux level within a transformer is proportional to the voltage applied to the transformer and inversely proportional to the frequency of the applied voltage. When overexcitation conditions that are above transformer design limits occur, the transformer core becomes saturated resulting in a buildup of heat with eventual damage to the transformer.

Generator transformers are especially subject to overexcitation as such transformers are connected directly to the generator terminals. Voltage and frequency conditions at the generator terminals are subject to voltage and frequency variations, especially during startup of the generator.

Transformer overexcitation concerns, however, are not limited to generator transformers.

Magnetizing current at overexcitation, where I1 is the fundamental frequency current, I5 is the fifth harmonic current, Im is the total magnetizing current and In is the nominal current.
Figure 3 – Magnetizing current at overexcitation, where I1 is the fundamental frequency current, I5 is the fifth harmonic current, Im is the total magnetizing current and In is the nominal current.

Overvoltage and underfrequency conditions can occur anywhere on the power system, especially when disturbances cause portions of the system to operate as isolated islands. Bulk transmission systems are also subject to high voltage conditions during light load periods.

This is because such systems often contain long transmission lines, which contain significant capacitance. During light load periods, the effect of line capacitance predominates voltage drops caused by load flowing through the inductive reactance of the line, resulting in increased voltage levels on the system.

Voltage levels can increase to the point where the ratings of system facilities, including transformers, are exceeded.

Harmonic content of transformer excitation current is predominantly odd harmonic. Typical transformer excitation current will contain a fundamental component, which is 52% of nominal, a third harmonic component equal to 26% of nominal, a fifth harmonic component equal to 11% of nominal, a seventh harmonic component equal to 4% of nominal, and so on.

Overexcitation protection should be considered for all large transformers utilized as generator unit transformers or those that are connected to portions of the power system conducive to causing transformers to become overexcited.

Such protection should consist of relaying that is capable of directly responding to the level of excitation that exists such as volts=hertz relaying.

Transformer differential relays are subjected to operation on high transformer excitation current. However, the operating characteristic of the relay on such current does not correlate well with transformer overexcitation limit characteristics. As such, it is not practical to use differential relaying as a means to protect transformers against overexcitation.

On the downside, transformer differential relays are subject to operating on overexcitation current at levels below that which may cause damage to the transformer.

Moreover, operation of differential relaying caused by overexcitation could cause confusion to post-disturbance investigations. Larger transformers, for which overexcitation is a concern, should be equipped with dedicated overexcitation protection and associated differential relaying should be blocked from operating on excitation current for reasons cited earlier.

3. Current transformer Saturation

Saturation of current transformers associated with transformer differential relaying causes several concerns with regard to such relaying:

  1. CT saturation on external faults can cause incorrect operation of differential relaying due to operating current that can result from the distorted secondary current waveforms that exist during such conditions.
  2. The harmonics contained in secondary currents of a saturated current transformer may delay operation of transformer differential relaying on internal transformer faults.

Proper selection of current transformers can minimize exposure to the problems listed earlier. Design features of transformer differential relays also address these concerns.

Go back to contents ↑

Reference // Protective Relaying Principles and Applications by H. Lee Willis and Muhammad H. Rashid

SEARCH: Articles, software & guides

Premium Membership

Premium membership gives you an access to specialized technical articles and extra premium content (electrical guides and software).
Get Premium Now ⚡

About Author


Edvard Csanyi

Electrical engineer, programmer and founder of EEP. Highly specialized for design of LV/MV switchgears and LV high power busbar trunking (<6300A) in power substations, commercial buildings and industry fascilities. Professional in AutoCAD programming. Present on

One Comment

  1. celso hoffmann jr
    Dec 27, 2017

    I already had this problem. saturation. it is difficult to find and solve. You have to fool the relay with keeping the security. because no one is going to change a bushing CT…

Leave a Comment

Tell us what you're thinking... we care about your opinion!

Premium Membership

Get access to premium electrical guides, technical articles and much more!