Medium Voltage Motor Re-Acceleration 1

[franyoud]

​

Dear community,
I was wondering is someone can help regarding a technical assesment on a topic I'm working on. Is it relating to ANSI 66 protection on Switchgears (Number of Start per hours).
The one we are using to protect motor is SEPAM, and in our case have the protection defining number of start per hour in cold and hot mode. In addition to the number of Start also delay between start and stop can be set (either at 0 or 1min to hour).
In our case we have considered to implement the 1min delay between start and stop to avoid motor re-acceleration (while still spining) effect which can lead to much inrush than normal starting. Have you ever face this matter with the delay between start and stop in your protection study of HV motors involving ANSI 66 ? Does it sound relevant to you to keep this 1min delay for MV motors (from several hundred kW to MW) ?

 

[wcaseyharman]

I’ve applied the requirement of NEMA MG1 for number of starts per hour, but I don’t think I’ve applied a delay time between the stop and the start. Seems like a good idea though, especially if you have a clutch like the motors I was working on.
Edit: now that I think about it I think that, but not 100% sure, the control system logic requires the system to be at zero speed before another start is given due to the nature of the clutch attached to the motor. Other sites that had clutches that could engage while spinning did not have those limitations.

 

[waross]

That sounds as if it could work.
The closest that I have come to that issue was a large compressor motor limited to three starts per hour.
Depending on air usage, the motor easily cycle more than 3 times per hour.
We added a time so that the automatic stop would not function until the motor had been running at least 20 minutes.
The compressor had an electrically operated unloader and the unloader was used to stop compressing air when the pressure was up to cutoff.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Gr8blu]

The motor usually has a "start" nameplate on it somewhere. The wording typically goes something like this:

"With rated voltage and frequency - within NEMA limits - at motor terminals and with connected load
inertia not exceeding value shown below, the following starting duty should not be exceeded.
Load inertia = NNN lb.ft2
Motor COLD = XX consecutive starts
Motor HOT = YY consecutive starts
Subsequent starts with motor running between starts = AAA minutes
Subsequent starts with motor standing between starts = BBB minutes
Starts should not average more than DDD per day throughout the life of the motor"

A "cold" start means that everything in the motor (e.g., all the steel AND all the copper) is at the prevailing ambient temperature. If anything is warmer than ambient, the attempt is a "hot" one. This is why the allowable number of successive starts for the "cold" condition (e.g., XX) is higher than for hot (e.g., YY).

The second part of the plate refers to the dwell time between one series of starts and the next. The duration for the motor "running" (e.g., AAA) is lower than for "standing" (e.g., BBB) because the combination of convection and radiation provides better heat transfer than radiation alone. Note that to meet the "running" condition requires that the rotor remain spinning at a significant percentage of rated speed for the entire duration.

As to applying a delay between the a "start" command and the next "stop" command: that is an excellent idea. If the start sequence was successful, the setting should correspond to the value listed for "motor running between starts". If the start sequence was unsuccessful, the setting should correspond to whichever condition applies (i.e., "motor running" or "motor standing").

Lastly - in certain instances, it may be advantageous to try to reclose the breaker or contactor "on the fly" while the rotor is still spinning. Some machine types can do this easier than others - in particular, the constraint for a synchronous machine is different than for a squirrel cage induction. It is usually better (for the life of the machine) to allow it to come to a stop before attempting to restart. How long this takes is dependent on the mechanical characteristics of the system (friction, inertia, the ability to "decouple" components, etc.).

Converting energy to motion for more than half a century

 

[LionelHutz]

On induction motors you just need to pause long enough to let the residual field decay which is generating a higher terminal voltage. Usually lasts around 5 to 20 cycles. Then, the initial inrush is the same regardless of the rotor speed.

Gr8blu - it's very optimistic to expect to find anything like that on the motor. I don't recall ever seeing that level of detail on a a motor rating plate before, and I've been doing motors for 25 years. On the motor data sheets, sure sometimes, but not usually even then.

 

[electricpete]

On induction motors you just need to pause long enough to let the residual field decay which is generating a higher terminal voltage.

LPS. Residual voltage can cause a transient but after long enough it decays away. I think NEMA MG-1 says 1.5 times the "open circuit time constant" allows residual voltage to decay to 33% which ensures worse case difference between incoming and residual is 133%. They have a formula for that time constant based on motor equivalent circuit parameters. You could also estimate measuring residual voltage directly at stator terminals during coastdown to estimate how long to get to safe levels... recognizing the voltage decay is product of flux (which decays by open circuit time constant) speed (which decays according to inertia and load torque during deceleration)... for conservatism should measure during conditions when the coastdown is at least as long as expected during future operations. Once the voltage decays away, the motor rotating forwards doesn't cause problems for a restart, if anything it reduces the time needed to re-acceleration to full speed.

Gr8blu - it's very optimistic to expect to find anything like that on the motor. I don't recall ever seeing that level of detail on a a motor rating plate before, and I've been doing motors for 25 years. On the motor data sheets, sure sometimes, but not usually even then.

NEMA MG-1 doesn't require that to be on a nameplate attached to the motor BUT some customers include that in their spec. It's included in some of the specs used for large motor purchase during our plant construction, so it's very common for our large motors.

 

[waross]

Motor starting current, motor starting surge and motor starting transient.
Motor starting current at around 5005 or 600% lasts until the motor is at about 90% of rated speed.
There is an energization transient similar to the transient experienced when energizing a transformer for the same reason.
This has to do with residual magnetism and the point on wave that the motor is energized.
Closing in on out of phase residual potential may cause a worse transient and may do mechanical damage.
Often by the time the residual field decays, the motor speed has dropped to less than 90% hence the current surge.

Then, the initial inrush is the same regardless of the rotor speed.

In practical terms and in almost all cases yes, but the surge is due to the motor speed having dropped to less than 90% rather than the absence of residual effects.
There may be exceptions for very high inertia loads, however such loads will generally require reduced current starting methods that make them special cases.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[LionelHutz]

No Bill, the initial inrush is an energization transient and it will occur no matter what speed the motor is rotating. Residual voltage can make it worse if you don't wait for that to decay enough.

 

[waross]

Understood, Lionel.
I guess that I didn't explain myself well.
Does this fly?
Motor starting current; About 600% of full load current, and lasts till about 90% of rated speed.
Motor energization transient; May approach a peak of 2.8 times starting current. The magnitude of the transient depends on residual magnetism and the point on wave of energization. The transient peak may exceed 2.8 times the starting current if residual potential is present. Again depending on point on wave.
Anecdote alert:
I remember a checking series of starts of a small motor with my brand new digital ammeter.
The peak lock was picking up the energization transient which was higher than the expected 600% of FLC.
With multiple starts, the peak varied over a range of more than 2:1, all higher then the expected 600%.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[electricpete]

On your anecdote. This is nothing you don't know but I'll say it in my own words. The 600% thumbrule excludes the decaying dc component and that dc component adds onto that in a fashion which as you say is variable between starts (especially if you're looking at a single phase). When we talk about peak capture of a motor start on the digital multimeter I assume it's not instantaneous peak but more likely the highest rms that it computes over a half cycle or full cycle and that will still have heavy influence from that variable dc.

Again nothing new so why did I bother? What I was building towards is that I tend not to trust any number anyone gives me about a motor starting transient (because multimeters have differences and can be set up and used in a variety of ways). If it's important I encourage them to capture waveforms (and even that can have pitfalls... is the dc component properly indicated, did you capture enough starts...). I suspect most people here can relate to that skepticism of numbers that someone else gives you about a motor starting transient.

 

[waross]

It was a cheap meter and one of the first digital ammeters.
It measured peaks and scaled them to RMS.
I soon abandoned using it, but I did get some interesting insights before I set that meter aside.
And yes, it was a group of single phase motors.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!