Autotransformer Motor Start

[davva]

I have been asked to review some voltage drop calculations for an autotransformer start induction motor scheme. The system is an island network with 2 GTA sets (circa ~100MW capacity combined) and the motor a 11kV, 15MW induction motor.

The calculations have been performed by the machine vendor and they have been given the minimum system fault level from a third party. The fault level calculations have been generated by a power system model which I do not have access to.

The voltage drop calculations are also generated by the Vendor’s software and only give the headline results of study. The fault level of the network is low and the voltage drop calculations show that the machine can be started via the autotransformer and keep within the voltage tolerance of the network which is +/-10%, only JUST though 9.9% dip!. The +/-10% tolerance is a steady state tolerance.

I have a few general questions relating to the above.

1. If the machine vendor has used the min fault level data which has been provided to them to calculate the voltage dip then they are probably using the equivalent synchronous reactance (Xd) data for the network generators, is this correct? If so, given the margin, should they be modelling this more accurately with transient/sub-transient data for the generators?
2. A schematic diagram seems to suggest the scheme offered is an open transition auto-transformer. I am not aware of the transient tolerances on the network but wonder if a closed transition option should have been considered or whether it is considered economically unfeasible at 11kV due to the additional switchgear and components.
3. I’ve read the effects of open transition switching include light flicker and
possible information data loss and circuit breaker tripping. Would generator AVRs be effected by very large inrush currents of short duration or is the open transition time so short that they wouldn’t have time to react anyway?


Many thanks for any responses in advance.

 

[dpc]

If they used synchronous reactances of the generators, which I would doubt, then calculations using the sub-transient would show less voltage drop, since the reactance will be about 10 times lower. During a prolonged voltage dip, the voltage boost function of the AVR would basically trump the generator's natural response characteristics to some extent. There would be an initial dip, then the regulator would go balls out and the voltage would recover (to some level) while the motor was starting.

An open transition auto-xfmr starter does create a noticeable transient on transition. This could translate into lamp flicker. I'm not sure the AVR would respond but that depends on how it's tuned.

A 15,000 kW Auto-xfmr starter will be one big boy. Are they doing this with circuit breakers?

I assume they've considered the available taps on the auto-transformer and evaluated the lowest tap that would reliably accelerate the motor without stalling.

 

[davva]

Thanks for response dpc.

Yes I don't know where I was coming from after re-reading my question 1 yesterday!!

Regards your question on CBs I only have a sketch but they are using VCB technology for 11kV switchgear elsewhere so I guess they are using VCBs/vacuum contactors.

Yes there is a motor starting study that shows the motor starting on the mid range tap (75%). Yet the voltage drop calcs have used the current from the lowest tap range 68% which is more favourable to the voltage dip calculations.

The study shows the motor starting in under 20secs.

Would people generally expect this to be closed transition?

 

[dpc]

At lower voltages, yes, but at 11 kV and using circuit breakers, the closed transition probably gets more complicated and expensive.

 

[LionelHutz]

I would think the AVR will not have time to respond to the switching transients.

Closed transition would require another vacuum contactor and 11kV rated resistors adding a lot of cost and size. I doupt the unit proposed is closed transition. Closed transistion doesn't seem too common except with chillers because chillers tend to damaged by transients.

That autotransformer would be huge. I would also consider a soft-starter solution. There are no switching transitions. The bigger advantage, however, is that the current can be gradually increased allowing the AVR to respond to the extra load and boost the voltage.

There are a couple of manufacturers of soft-starters that can do this size. One can provide a power acceleration which controls the power drawn which in turn means the extra starting load on the prime mover is controlled.

 

[Marke]

Hello davva

1. The calculations are probably based on the start current being limited by the auto transformer starter to a low value equal to the full voltage start current multiplied by the squar of the voltage reduction. This will only be correct if there is sufficient torque available at the reduced voltage to accelerate the motor and driven load up to full speed. If the torque is too low, the transformer will switch over at part speed and this will result in DOL current. see

http://www.lmphotonics.com/m_start.htm#AutoTransformer

It is important to ensure that at the reduced voltage, the motor can produce enough torque to accelerate the motor and driven load to full speed before stepping to full voltage.

2. An open transition starter will produce a very high current and torque transient and should be avoided. The transient can cause oscillations in the AVR system, depending on the type of AVR used. The transient can also cause nuisance tripping of protection equipment.
With a single stage autotransformer starter, it is no more expensive to build a closed transition starter than an open transition starter. You connect the "low voltage" contactor in series with the star point on the transformer rather than to the low voltage tap.
I very rarely see open transition autotransformer starters, never built by those who know what they are doing!!

Best regards,

Mark Empson

http://www.lmphotonics.com

 

[LionelHutz]

You're right Mark. The resistors and a contactor are used to get a closed transition on a Y-Delta starter. When the Y point of the autotransformer is opened it basically becomes a reactor/resistor in series with the motor which keeps it in phase with the line power.

 

[Modula2]

Davva, it's the transient impedance, for the initial voltage dip, not subtransient, not synchronous.

 

[davva]

Thanks for responses.

I've since found out that it will be a closed transition arrangement. Which is one less worry.

Modula 2, yes that is where I was coming from on my original question. I've seen in text books to use either X'd or (X''d + X'd)/2. I don't have information on the transient voltage tolerances only the steady state. So I can only check the calculations for the steady state (Xd) period of the start, i.e. after 1-2 seconds.

I'm still trying to find out what the network transient tolerances are