Power transformer inrush current 7

[JPetersen]

Hi,

Could anyone tell why a 3-phase power transformer shows a so nonlinear and high inrush current characteristica, when induction motors do not.
Of course induction motors do have a high (4-8 x In)starting current, but it seems low compared to transformer inrush of 10-20 x In.

The electric equivalent circuit diagrams for induction motors and transformers seem quite the same - however those diagrams are apparently only valid in steady-state conditions

Brgds

 

[electricpete]

Some differences:
1 - A power transformer more closely resembles an ideal transformer than a motor. That is the leakage reactances are smaller and magnetizing reactance larger compared to motor.
2 - Transformers operate further into saturation than motors.
3 - Upon energization a motor acts like a transformer with low-resistance-shorted secondary while transformer is most cases has open circuited secondary upon energization.


I think #2 and #3 are the biggest factors. In both cases you will have a dc component with a theoretical maximum equal to peak of the fundamental ac component (of locked rotor current for motor or magnetizing current for transformer). That decays with L/R time constant. L/R is much lower for the motor due to rotor resistance in the circuit and the dc component decays away very fast always virtually gone within two or three power cycle in the traces I have seen. In contrast for energization of generator stepup transformer from the high side with low side open-circuited, I have seen the dc decay transient and associated huge unidirectional current spikes last for minutes.

So, relating to the non-linear part of the inrush, #3 makes the dc last longer. #2 means the dc pushes the core further into saturation and so current spikes from entering saturation are much larger.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[dpc]

electricpete,

Wouldn't the presence of the air gap in the motor make a big difference? I guess this is really your #1 point.

dpc

 

[Skogsgurra]

Great reasoning.

There are also design considerations. The core is designed to absorb a certain flux and this flux is based on a +/- flux swing, i.e. with a zero mean flux. Depending on where the closure happens (on top of sine is best) the assumption regarding zero mean flux is more or less fulfilled and zero crossing will make the zero mean flux invalid during the first few cycles and thus lead to saturation of the core.

The air gap in the motor prevents saturation while there is no air gap in a normal transformer, which makes satiration more likely.

Stars for Pete for his lucid summary of what is going on in motor and transformer.

 

[JPetersen]

Considering pete's #3:

If transformer is loaded on secondary when primary is energized, this may in fact reduce the inrush current magnitude and/or reduce the inrush time?

Brgds
Petersen

 

[Skogsgurra]

Yes JP,

The loading reduces available voltseconds across the core and reduces the saturation. You can test this easily by connecting a transformer to mains and look at the inrush current (clamp and scope). Loading the secondary reduces inrush. Yes.

 

[ppaya]

JP:
I don't agree that equivalent motor network is made for steady state condition only. There is sub-transient and transient reactance for model. (X" and X')
Very good thread guys.

 

[tinfoil]

Skogsgurra:

Unless I am mistaken, you stating that adding load (finite impedance) to an open secondary (near infinite impedance) will REDUCE the inrush to a xmfr?!

Taking this to its limit, this infers shorting the secondary terminals will further reduce the inrush!

 

[davidbeach]

tinfoil,

I don't know for sure about shorting a transformer to reduce inrush, but that conclusion is not really that far fetched. When one wants to do a short circuit test on a generator, the way to do so is to short the generator while it is not rotating and then bring it up and apply the field with the output shorted. This avoids the sub-transient and transient response and goes directly to the steady-state response. This long term steady-state response will typically be less than full-load current.

With a transformer, the current to a shorted secondary is limited by the impedance of the source and the impedance of the transformer, the 3-phase bolted fault current. So you would have currents equal to maximum fault current but with minimal transient inrush current on top of the fault current. Not really a practical way of limiting transformer inrush.

 

[JPetersen]

davidbeach,
just some thoughts from here:

consider a transformer-starting induction motor. With such installation (large machines) it could be useful to energizer transformer and motor in one step? If this is practiced in real transformer starting motor installations, I can not answer you. Anyone? But it could be a practical way of limiting transformer inrush.

Brgds
Petersen

 

[davidbeach]

I don't know off hand what would happen if one were to energize a transformer with an induction motor directly connected to the secondary, there is pretty much always a starter in between. Because the energization of an induction motor is not a constant impedance load, the motor will have some impact on the transformer inrush, but the question of what that impact is will have to be answered by someone with recorded data or good EMTP studies.

 

[Skogsgurra]

No tinfoil, there are also non-linear relationships...

 

[electricpete]

tinfoil -


My thoughts to your question:

I say the answer is yes if the shorting is resistive. There is a dc component which must decay away before the inrush dies. It represents a portion of the magnetic energy in the core which decays using L/R. The transformer by itself is mostly inductive with very little resistance... very little capability to dissipate energy as resistive losses. It should not suprise us that adding a resistor on the secondary which will consume tremendous heat we can dissipate the energy faster. The lower the resistance the more heat power is dissipated (P = V^2/R where V is roughly constant as a first approxiamtion), the faster the dc portion of magnetic field energy is dissipated, the faster the inrush goes away.

Unfortunately a motor load is mostly inductive and from that standpoitn wouldn't help as much as a resistive load during energization.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I did not mean to imply that all of the energy dissipated on the secondary would be associated with decay of the field... but a portion would.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

last sentence should have been decay of the dc portion of the field.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[rbulsara]

A star from me too.

 

[Reactorman]

The main factors which determine transformer inrush current are as follows:

1)The residual magnetism in the core this will vary from zero to approximately 70% depending mainly on the instantaneous value of the voltage when the previous shutdown occured (this does not happen in a motor as the air gaps prevent any residual magnetism).

2)Depending on the voltage point when the breaker closes there could be a total swing from minus voltage to plus and therefore the core can saturate quickly. Once the core saturates the the only thing limiting the current is the air core reactance of the primary winding.

The worst case can be easily calculated, however the air core reactance of the primary winding must be known.

Reactorman