Soft start into a spinning load? 4

[bentov]

At our motor repair shop we have 200a 240V 3 phase service equipment, but because our actual operating loads are low the utility supplies us with a whopping 45kva. We routinely rewind motors up to 500hp 480V. The problem is getting motors up to full speed/full voltage for no-load testing.

We have two existing test panels. One is an old 150hp 4 pole synchronous motor with redesigned rotor winding & gearmotor attached to output shaft, makes a continuously variable autotransformer good for 0-480V out with 240V in over a 1/4 turn (90 degree) rotation. It's really cool BUT can't supply enough current to start larger motors, especially older ones with heavy rotors. The other is a transformer/contactor setup using a couple of 75kva transformers I found, voltages available are 60, 120, 240, 480. That does start the big ones (takes 800a at 60V sometimes for old motors with heavy rotors). So we can get them up to speed, but still limited on full voltage test running; depending on details we can get up to around 250hp at 480V - larger ones generally trip the 200a main breaker at our service entrance. The problem is the setup is necessarily open transition - closing the contacts at full voltage of course is a big current spike we just can't handle (though the rotor is at full speed, inrush is a function of winding impedance/rotor reactance at applied voltage - big is big).

I'm thinking about a third setup, a 480V VFD that can handle the rotor inertia acceleration without all that current, get us all the way up to full speed/full voltage. For proper testing though (vibration, etc.) we should run without a VFD, so my question (finally!): what happens if we have a motor running full speed, turn off the VFD, then re-energize with a soft starter while still spinning close to full speed? Especially when the motor is quite a bit larger than the soft start (like a 500hp motor running no load with a 150hp soft start)?

 

[waross]

Spin a little over speed with the VFD and then catch it on the way down, DOL, with a sync check relay?


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

The soft-starter would catch it but there could be some inrush current to pull the motor to running speed depending on how it's done. I would expect that if you over speed the motor and then start the soft-starter with it limiting current to the maximum current your source is happy with, then the motor would slow down to running speed and start running there just fine. That's assuming the no-load current of the motor is under the current limit, otherwise it would likely never stabilize.

 

[ScottyUK]

Don't forget that a big motor has high magnetising current even if it is running light. The source and the soft start still have to carry the magnetising current - which could be 60% of full load current.

 

[bentov]

Thanks for the replies. My plan is to use a standard soft starter with internal bypass, point is to get to xline (across-the-line) no load running condition for extended test at rated voltage. Two concerns then - what if the soft start (correctly) senses that acceleration is complete, just closes the bypass contacts without SCR ramping? I would think they all are designed so that thyristors MUST function, with closed transition to the bypass state before they are gated off, but I have seen field cases where that didn't happen (i.e. bypass contact closure AFTER thyristors opened, long previous thread about that). The other concern is regeneration, can't visualize that well enough to understand if it's a problem - if we connect a 150hp soft start to a 500hp spinning motor, is there a chance it'll get vaporized by some kind of back EMF?

Will the sync check relay per waross help with either of those scenarios? Will overspeeding per waross & LionelHutz help, or maybe increase the chance of skipping the ramp step?

ScottyUK, we find that newer motors generally have significantly lower idle currents (iron core flux densities lower with higher efficiency, lower eddy current and hysteresis losses thus proportionately lower magnetizing current) - flip side of that is significantly higher inrush (partly due to the extra copper in the windings - rewind business is getting way tougher, right when the next generation of winder trainees wonders why anyone would want to work so hard!).

 

[waross]

When starting a motor from standstill there are two inrush factors.
1. Normal starting and accelerating current.
Duration, several seconds. Often single digits with DOL starts.
Longer with reduced current starting on high inertia loads.
This typically starts at about six time FLA and decreases as the motor comes up to speed.
2. Magnetizing transients.
Duration, a few cycles.
This is similar to transformer inrush.
It depends on the residual magnetism of the motor, the X/R ratio of the motor winding and the point on wave that the power is applied.
In the worst case it may approach the peak to peak value of the starting current.

Applying current to a spinning motor adds another two factors.
Back EMF, and the frequency and phase angle of the back EMF.

The back EMF of a spinning motor may be both at a different frequency and out of phase with the applied power.
1. If the motor is spinning at either more or less than synchronous speed, there will be a current inrush and a torque inrush as the motor is either slowed down or sped up to match the grid frequency.
2. If the applied power is out of phase with the back EMF, then depending on the point on wave of the power application there will be an additional current transient. In the worst case, when the back EMF opposes the applied voltage, this may be more than the transient from a normal DOL start.
Square that factor to get the torque transient.
This is one of the reasons why open transition star:delta starting has fallen out of favour.
As the over speed motor is coasting down, there will be a point where both the frequency and the phase angle is very close to matching the grid frequency and phase angle.
A closure when both the frequency and phase angle of the back EMF are close to the grid frequency and phase angle will be fairly benign.
Conditions of exact frequency and phase angle match may never coincide.
Like the game of horse-shoes, close counts.
As the out of phase closure can multiply the out of frequency transient, I would set the sync check relay for a very close phase match and be more forgiving on the frequency setting.
Absent the ability to over-speed the motor, a sync check device to match the phase angles before closing may avoid the worst transients.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[bentov]

Thanks for thinking about this. I have never used a sync check relay, only read about them for paralleling generators. Will one work in this case (with an unpowered spinning motor)? Will it detect (to sync with) back EMF arising from residual magnetism?

I may be barking at the wrong squirrel here anyway (with the soft start idea), just read this about starting multiple high voltage synchronous motors with a VFD and step up transformer (

https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6963463);

(sounds like they had some, uh, "exciting" initial results with transitioning to DOL).

The comparison of various starting methods there got me thinking we might be better off (for "flicker") with a closed transition RVAT to manage magnetizing inrush after the VFD takes care of acceleration current. Do you suppose RVAT at 50% tap (to re-energize the spinning motor) might be less taxing to our supply vs. the full voltage chopped waveform from a soft start?

 

[waross]

How much will the magnetizing current be on those 500 HP motors?
You may find that even without inrush current that your 200 Amp breaker can not supply enough current to handle both base loads and the magnetizing current of a 500 HP motor.
Re Sync Check relays. You will have to check the specs of individual relays.
Not all models may be capable of detecting residual back EMF.
The magnitude of residual back EMF may vary from motor to motor.
You may have to "Roll your own".


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[bentov]

What we know is we can run any of them at 60V and 120V without tripping the breaker. Occasionally we trip at 240V (on a 500hp), for sure we trip above 250hp at 480V - but all those voltage steps are open transition with the current setup (just energizing the transformers dims the lights). So maybe what we should be rolling up is a custom autotransformer and contactor arrangement for closed transition steps, with taps at 25% (120V), then maybe 50% & 75% on the way to 100% . . .

It does seem like fiddling with Sync Check relays and/or soft start ramp settings will be unpredictable at best (explosive at worst), given the wide range of motors we'd like to test run.

 

[waross]

Check your base load and the expected magnetizing currents of the larger motors before spending any money.
The initial transient is related to the point on wave that the switch is closed relative to the back EMF.
If you can never get a large motor energized at full voltage the magnetizing current may be too great even without a transient or surge.
You may be able to use some capacitors to reduce the magnetizing current through the main breaker if this will not interfere with your testing.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

I believe you are over thinking this. All the trickery required to pull off the above machinations is likely too hazardous for everyday operations across large variations of motors. Sooner or later someone will 'slip-up' with some kind of nasty repercussions.

You should be able to do your characterization tests with a VFD running the motor. The VFD should be superior as a test platform as it may be able to hand you lots of running data not otherwise available.

Run your normal tests with smaller motors you have no problem running DOL with a follow-up test with a VFD and prove it to yourself. Assuming you can't do something that you actually may well be able to, would be a shame in this situation.


Just say "No to transitions".




Keith Cress
kcress -

http://www.flaminsystems.com

 

[bentov]

Transitions can definitely be interesting! We're probably overthinking the back EMF thing, should ask EASA but you guys are more fun. I should specify we're talking exclusively standard squirrel cage induction motors, which I understand generate current if driven above synchronous with voltage applied (induction generators) but otherwise . . . not excited, they just don't put out (like, uh . . . never mind). The residual magnetism decays rapidly, is maybe how to think of it. Maybe not always though, right? The oddball circumstance (high residual), un-watched for, creates the "nasty repercussions".

We've certainly been cavalier about test running lots of motors these many years, no accidental shocks yet. I've occasionally measured voltage on de-energized spinning motor leads, seen nothing so called it safe (big ones coast a while, meantime we're disconnecting from the panel, masking for paint, getting on with the day). Also we routinely jog reverse to stop the spinning (multiple test runs for balancing, end play adjustment, etc.), typically at one step lower than the last-used forward voltage. Kind of fun muscling the big rotors to standstill, never had a breaker trip doing that, always sounds very smooth . . . so my gut says back EMF is a non-issue, but I remain suspicious of soft starts (have more commissioning and troubleshooting problems with them vs. VFDs).

What I should do is stop yapping and start measuring. We have a couple of very fast Fluke 1750 data loggers that just don't miss anything - plan is to dedicate one to the new test panel anyway for recording our run results (you're right, Keith, standard VFDs tell us way more than staring at panel meters and writing things down on greasy job cards - soon as I ask, "Are you sure about that (the volts, amps, whatever)?" I see the doubt in their eyes). Guess I should start by recording the heck out of what we already do, zoom in on the details to see for myself . . .

 

[itsmoked]

Hey, if you find someone who can write a little code you can even ask the VFDs for all sorts of data and compile a standard test result. Obviously you can also have the code run up the motor and exercise it in some useful manner. Of course, you can also have the VFD bring it to a non-plugging quick stop.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Search this site for "star delta starting".
You will find a lot about transition surges due to out of sync transitions.
You will see graphs of current spikes due to bad transitions from star to delta. The switching transient may be greater than the initial starting surge.
The issue is the phase angle difference between running on the star connection and the phase angle at the moment of connection to delta.
There is also a 30 deg difference between star and delta. With luck and favourable connections, the motor may drop back those 30 degrees during the time that the open transition takes and there will be little surge. The emphasis is on luck.
A different load on the same motor and starter may cause a different deceleration curve and the transients may increase.
I remember a bank of 30 HP fans that were often shut down by tripping the breaker.
The back EMF would hold in the magnetic contactor for one or two seconds after the power was cut.

Bill
--------------------
"Why not the best?"
Jimmy Carter