detailed procedures to assure NEMA requirement 1-start hot/2-start cold

[electricpete]

I was asked to detailed easy-to-follow operational procedures which will assure that the NEMA 1-start hot / 2-starts cold requirement will not be violated (for one specific motor). In other words, a set of rules which a non-motor person could use to evaluate any postulated operation scenario start1/run1/stop1/rest1/ start2/run2/stop2/rest2/ start3/run3/stop3/rest3/ etc (with specific durations given for run1, rest1, run2, rest2, run3,rest3 etc) in order to determine whether the scenario meets the NEMA requirement one start hot/2-starts cold.

You'd think that would be easy. Maybe I'm slow today, but it's difficult for me to come up with the specific procedure.

The motor in question takes about 3 hours for stator temperature to cool 66% of the way to its final temperature after being secured. It also takes only one half hour to reach its final temperature after being started (as expected the time constant is much longer while secured than while running). I don't have access to any rotor temperatures of course.

My goal is simplicity and meeting the NEMA guideline. If I end up being a little overly restrictive, that would be ok.

Here's my discussion / conclusion.

Discussion:
NEMA MG-1 specifies one start hot / two starts cold.

The intent of one start hot / two start cold is that we only ever restart the motor once from any condition other than “ambient”. After we use up that start, we have to wait for it to cool to ambient.

For example if motor is initially cold, we start it and it trips, we can attempt restart. (this is our allowance two starts cold). Any further starts cannot occur until after the motor is cooled to ambient.
If motor is running for a long period of time and then is secured or trips, we can restart it before it cools but only once. (one start hot allowance). Any further starts cannot occur until after the motor is cooled to ambient.

How long it takes for the motor to cool to ambient after it is secured is not well defined. My judgement based on review of the stator winding temperature trends is this interval is 3 hours for the motor under consideration.

Conclusion:
In general, all starts of this motor should be spaced at least 3 hours apart. If we find occasion where we need two starts (start #T and start #T-1) to be less than 3 hours apart, we can do that provided the first of these two starts (#T) was separated at least 3 hours from the start before that (start #T-1).

I think this meets the intent of the NEMA requirement. What do you think.

Would you have an alternate suggestion for specific procedures to assure the NEMA guideline is met?

Would you have an alternate suggestion for specific procedures for limiting of large motors when OEM has not provided any repetitive starting limitation information other than the NEMA requirement?


=====================================
(2B)+(2B)' ?

 

[electricpete]

Whoops. That conclusion was gibberish. Let start start again.
Conclusion:
In general, all starts of this motor should be spaced at least 3 hours apart. If we find occasion where we need two starts to be less than 3 hours apart, we can do that provided the motor had been stopped/resting at least three hours before the first of these two starts

=====================================
(2B)+(2B)' ?

 

[electricpete]

In general, all starts of this motor should be spaced at least 3 hours apart. If we find occasion where we need two starts to be less than 3 hours apart, we can do that provided the motor had been stopped/resting at least three hours before the first of these two starts

It seems to me this accomplishes the NEMA limit (assuming 3 hours secured is enough to get the motor back to ambient).

AND YET, if I tell our operators this, and they have an evolution such as a difficult balancing evolution where they want to start/stop the machine several times, using my rules they can achieve more starts per time using pattern A than pattern B:

Pattern A: Start/run-for-1-minute/stop/rest-for-1-minut/start/run-for-1-minute/stop/rest-for-3-hours/ Repeat.
(makes use of two-starts cold pattern to achieve approximately 2 starts per 3 hours)

Pattern B: Start/run-for-3-hours/stop/rest-for-1-minute/repeat
(makes use of one start hot pattern to achieve approximately 1 starts per 3 hours)

It seems to send a message that resting between starts is better than running between starts (and therefore we can fit more starts in if we rest between starts then if we run between starts). That's actually backwards from the way it should be, but there's really no other way to formulate a limit that follows the NEMA rule that I can see.

My head hurts.

Any suggestions (other than aspirin) ?


=====================================
(2B)+(2B)' ?

 

[electricpete]

How about this:
1 - In general, all motor starts should be spaced at least 3 hours apart.
2A - If an additional start is required to respond to unplanned events (motor trip) within three hours of the first motor start, this is allowed provided the motor had been at rest at least 3 hours prior to the first start.
2B - If an additional start is required for planned evolutions (maintenance) within 3 hours of the first start, this is allowed provided that the motor had been at rest at least 3 hours prior to the first start and provided that 6*R+S>180 where R and S are the running and stopped time in minutes between the first start and second start.
[2A and 2B are mutually exclusive]


=====================================
(2B)+(2B)' ?

 

[davidbeach]

How about a motor relay that incorporates a motor thermal model that will allow a start when ever there is enough thermal capacity in the motor for the start?

 

[electricpete]

We have electro mechanical relays.

Digital relays are being talked about for somewhere in our distant future. The timing of that modification is beyond my contol.

For now we have to work with what we have.

=====================================
(2B)+(2B)' ?

 

[davidbeach]

All I can say is that the relay replacement, probably plant wide, will be far less expensive than ever getting it wrong, once, and having to replace/rewind a motor. But that's the way of the bean counter.

The engineer's definition of an accountant - someone who knows the cost of everything and the value of nothing.

 

[electricpete]

I could make an even better case based on the fact that we have had several events where the plant trips due to an electrical event of initially unknown cause and the plant has remain shutdown at a cost of around a million dollars per day waiting for waiting for us engineers to figure it out the cause.

All we had to go on were the relay flags in the plant and some recordings in the switchyard, and whatever inspection and testng we could get done. And trying to read between the lines of voltage magnitude sampled at one second intervals as I posted in another thread once.

Those are pretty crude tools. The timeliness of such investigation could doubtless be improved if we had digital relays providing information about what happened before and during the event. Not to mention the quality of the investigation and various less-critical investigations I do like motor failure investigations. But the timeliness alone at one million dollars a day should justify it.

Sadly as I said, beyond my control. The modification proposal is in the pipeline and has to run its course in the process and compete for money and design engineer resources against other projects that are also good ideas.

=====================================
(2B)+(2B)' ?