MV Motor Start

[Modula2]

We have multiple motors to start, at 6.3 kV, 50 Hz. The power system is weak. We have looked at the following system:

http://powersystemanalysis.com/EasyMVStarting.pdf

My question is if anyone can see some areas of improvement on the switchgear part.

 

[Skogsgurra]

Hi,

I am right now involved in a similar project where we start four refiner motors from a LV frequency inverter. The motors are 6 kV and somewhere in the 5 MW range. We have done measurements on an older system (with GTO inverters) doing the same thing, but we are replacing the old inverters with PWM. Our measurements look OK and we have no doubts that the system will work as intended.

There are a few caveats, though. You need to start with no or very reduced load. And, the acceleration needs to be kept to low values. Do not be surprised if the starting takes a couple of minutes. Or more.

Gunnar Englund

http://www.gke.org

 

[jraef]

Interesting concept.

I think that the switchgear is the least of your problems. First off, in that detailed example they are using a 400HP VFD to start a 1500HP motor and then a 600HP motor. First off you must then assume that the loading of those motors is very light, if not non-existent (especially with the 1500HP). Is your system the same? In other words, that is a tricky endeavor to take on before you even get to the next issues. It can be done, but only if very thoroughly engineered and investigated.

The next issue is closed transition with a VFD (if in fact that is what they are doing). You can ONLY do this if the motor is at exact slip speed when the line power (sinusoidal as they call it) is applied. Any difference in frequency, however slight, will be disastrous for the VFD. Only a few VFD mfrs even attempt synchronous transfer schemes and I have NEVER seen one done with a drastically undersized VFD like this.

If they are instead doing an open transition (as might be indicated by their statement about transitional torque), then again, that will be highly application dependant. That's not to say it too couldn't work under the right circumstances, but you must put a lot of trust in the design to spend that kind of money up front not knowing for absolute sure that it will work or that there will not be costly consequences.

Their statement about Soft Starters and RVATs being in the dark ages is a little bold, because obviously this cannot work for every application. Under the right circumstances though, it has possibilities. I give them credit for marketing it and stating UP FRONT that they have just recently made ONE work! Give it to them for honesty at least!

http:/

http://www.jraef.com

Eng-Tips: Help for your job, not for your homework Read faq731-376

 

[Skogsgurra]

I should perhaps add that the system I am working on have synchronous motors and that excitation is on all the time.

Also, we take the motors slightly overspeed and then let them coast down to sync speed, where a synchronization device switches them to the 6 kV busbar.

So there is no closed transition. I do not quite see how that would work.

Gunnar Englund

http://www.gke.org

 

[Modula2]

JRAEF, it was a closed transition. Transient torques occur because the two power sources, are at slightly different voltage magnitude, frequency, and phase, even though it is closed transition. Graphs I saw of shaft measured torque transients show modest transients. For comparison, an autotransformer transfer could have much larger voltage difference and higher torque transients. The graph I saw of the 4.16 kV output voltage, when the motor had come up to speed, prior to transfer, was a perfect sinusoid. That is on the output of the sinusoidal filter. The VFD must be designed to be able to protect itself. Synch is done like a generator. Because of the VFD control, that should be good control and small window. The starting time was about 1 minute, but less than motor full load amp. The motors need to have a reduced load during start, but that is what we commonly do for reduced voltage starting. The impedance numbers on that diagram, indicate a very weak supply, and a starting flicker limit. Assuming a 6% transformer impedance, I can easily estimate how reduced voltage starting would compare.