Motor Fast Transfer 2

[mls1]

We have a small oil pump with an emergency backup power supply that when the normal supply comes back there is the potential for a very fast restart. Effectively, the backup contactor opens and the normal source contactor shuts. The motor is a 2 Hp, 3-phase, 480V induction motor. We understand the criteria for restart is 1.33 pu voltage between the source and motor bus and that one usually wants 5 to 6 times motor open circuit time constant before restart. What I don't know is a typical value for open circuit time constant for a motor this small. I would expect it would be very short but the contactor transfer should only take a couple of cycles. Is that enough time for motors this small or would it be recommended that we add some additional delay?

 

[davidbeach]

If it is really only 2-3 cycles you are probably good. A good recording meter and a test transfer would confirm. If it stretches out much beyond that, or the recording shows significant current spikes, then you need to put in a good delay. If there's any possibility that the two sources aren't synched, then you absolutely need a good delay; let the motor spin down and then restart.

 

[waross]

Some switches transfer so fast that that they incorporate a sync check circuit.
I have installed a lot of residential standby sets with fast, random, unsync'ed re-transfer to the grid. The refrigerators and freezers seem to take this well. They are small motors with low inertia. Never had a problem 10 or 15 years on.
However, that said, for a critical 2HP motor I would seriously consider a delayed re-start.
The very fast transfer switches typically have very little clearance between the normal contacts and the standby contacts. The clearances are so close and the movement is so fast that there is a danger of flashover from the standby source to the normal source.
Contactors don't drop out and pull in that fast. Many years ago, I had occasion to compare the pull-in and drop-out times of a number of small contactors around NEMA size 0. There was a great range of times but the larger the contactor the more time. I would estimate a time of 100 to 200 milli-seconds for a transfer, but I may be off by 300% or 400%. There was that great a range of times between manufacturers.
Tip; You can generally find the pull-in and drop-out times in the technical section of the manufacturers catalogue.
If your time is worth very much it may be the cheapest course of action to just buy and install a time delay device.
Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[mls1]

Thanks for both replies. The transfer is done with two electrically interlocked contactors such that the an "b" aux switch on the emergency contactor must make before the normal contactor can pull in. The data sheet close time is about 1 cycle so that probably about as long as we can count on.

The time delay relay would fix this but there are six identical installations with very tight enclosures. It would not be cheap. When all is said and done there would be a very low probability of the transfer back to normal happening with the normal source energized but the possibility is there. If anyone has any idea what would be an expected open circuit time constant would be for a motor this size we can at least estimate how much risk we are putting ourselves in. The load, by the way, is an oil pump so I'm not terribly worried about the torque transient, just the electrical transient on whatever field is left after about 1 cycle.

 

[waross]

Don't forget the drop-out time. This may be similar or very different than the pull-in time.
1 cycle is quite fast but is believable.
Automatic transfer switches normally transfer when the normal source is present. If you can find one, I have seen simple two lead time delays that could be inserted in series with a coil to delay the operation for a short fixed period. Some could be spliced in and left supported by the wiring. Quick and easy and not much space.
Can anyone help with a source for these devices?

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

 

[wayne440]

Something like this?

http://www.icmcontrols.com/TIME-DELAY-RELAYS-DELAY-ON-MAKE-DOM-Prodlist.html

 

[mls1]

The ICM device looks to be a very easy, low cost solution. I think we will go that route. Thanks!

 

[waross]

Nice link Wayne. lps

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

 

[itsmoked]

I have to agree with you Bill.

Funny a customer asked me to add a time delay to a shear about 30 minutes ago and I'd just created a folder called "Interesting and useful links" and stuck that site in it this afternoon.

Keith Cress
kcress -

http://www.flaminsystems.com