UPS for 50HP 480V Motor

[mull982]

I have a customer that is looking for a solution for running a critical 50HP 480V motor (currently fed from MCC) for a duration of 2-3hrs in the event they loose utility power. Has anyone ever done something similar using a UPS & VFD combination (I don't think motor and UPS direct connected would cooperate together) or some other similar solution?

I've heard of this solution being used for short time temporary durations (in seconds) to bridge gap until backup gen power comes online but never have seen anything for this long of a duration? It seems like adding a backup gen for this one 50Hp motor would be overkill.

 

[Cow]

Why do you think a backup gen for a 50hp motor is overkill?

Why do think a UPS would be better suited to run a large motor load for 2-3 hours?

 

[kwired]

Typical VFD's rectify the incoming AC volts to DC, then they use semiconductor devices to create a pulsed output that the motor sees as being very similar to AC current. I would think one could easily tie a battery network into the DC bus and it would easily take over if AC input is lost. That said it will take just a little bit of battery to handle 50 hp for 3 hours, but see no reason why it can't be done.

Back up generator makes sense if you don't want to have such a large battery - the battery can run it long enough to start the generator and get it on line, then you have no loss of load function during the transition.

 

[drcampbell]

Back-of-the-envelope calculation:
50 Hp motor needs about 50 kW.
3 hours = 150 kW*hr
300 (label, 20-hour rate) kW*hr of battery energy will be needed. (Peukert's law)
300 Group 31 recreational-vehicle/marine batteries
$60,000 and 10,000 kg for batteries alone
-
If this is the only critical machine there, would it lend itself to a mechanical solution?
Say, a 50-Hp engine on the other side of a double-shaft motor. When power's lost, the engine starts, the clutch engages and the engine turns both the dead motor and the critical machine it's connected to. No high-voltage battery pack, no UPS, no VFD, no generator.

 

[kwired]

Back-of-the-envelope calculation:
50 Hp motor needs about 50 kW.
3 hours = 150 kW*hr
300 (label, 20-hour rate) kW*hr of battery energy will be needed. (Peukert's law)
300 Group 31 recreational-vehicle/marine batteries
$60,000 and 10,000 kg for batteries alone
-
If this is the only critical machine there, would it lend itself to a mechanical solution?
Say, a 50-Hp engine on the other side of a double-shaft motor. When power's lost, the engine starts, the clutch engages and the engine turns both the dead motor and the critical machine it's connected to. No high-voltage battery pack, no UPS, no VFD, no generator.

Depends on how critical it is that the load doesn't slow down or stop during transition when utility power is lost.

Something like I described above certainly seems likely to be able to ride through the transition without the load seeing any effects. Don't know how common it is or if it is even done. Might not be so easy to do with "off the shelf" general purpose VFD but don't see why one can't be custom made for the application.

 

[zbang]

I have a customer that is looking for a solution for running a critical 50HP 480V motor (currently fed from MCC) for a duration of 2-3hrs in the event they loose utility power. Has anyone ever done something similar using a UPS & VFD combination (I don't think motor and UPS direct connected would cooperate together) or some other similar solution?

AFAIK, virtually nobody sets up a solid-state UPS to run 50kw for more than an hour. OTOH, it's incredibly common to set up a UPS with 15-30 minutes of battery and a generator. Is that not an option?

You should engage the UPS manufacturer about running a large-ish motor and not much else; they may have better ideas.

 

[tom baker]

Our hospital had a 1 MW Emergency generator with a rotary UPS, the ups maintains the load until the generator id on line.
A battery ups would not be practible. As mentioned, have the load connected to a motor and diesel engine. I worked on an old (1935) water pumping station that had this type of set up.

 

[synchro]

Typical VFD's rectify the incoming AC volts to DC, then they use semiconductor devices to create a pulsed output that the motor sees as being very similar to AC current. I would think one could easily tie a battery network into the DC bus and it would easily take over if AC input is lost. That said it will take just a little bit of battery to handle 50 hp for 3 hours, but see no reason why it can't be done.

Back up generator makes sense if you don't want to have such a large battery - the battery can run it long enough to start the generator and get it on line, then you have no loss of load function during the transition.

This company apparently sells equipment that utilizes a battery to power a VFD DC bus during an outage for at least a short while, and can then transition to generator power:

Bonitron Complete UPD Solutions

www.bonitron.com

Bonitron UPD Battery

www.bonitron.com

Loading

www.bonitron.com

I've no experience with them. But it could be worthwhile contacting them to see if they have something for the OP's application.

 

[Todd0x1]

The other issue with the batteries is the replacement every X years. Seems like TCO would be lower long term with 15 min of battery + genset. Whats this for? Does the backup scheme need redundancy?

 

[Jraef]

Yet another problem with putting batteries on the DC bus of a VFD is that when you open the disconnect of the VFD to work on it, you DO NOT KILL ALL SOURCES IF AVAILABLE ENERGY so it can be VERY DANGEROUS. You would have to put a separate DC rated disconnect switch on the battery pack that is somehow linked to the VFD enclosure so that you can’t open the VFD unless the batteries are disconnected. Not impossible, but it adds even more cost and complexity. Then if the motor is 460V, the DC bus required will be a minimum of 650VDC, which puts you into a 1000VDC rated disconnect switch; more money on top of more money. (Those Bonitron battery systems take care of all of this by the way, you just don’t see it in the brochure because each one ends up having to be engineered to your system. And they are NOT inexpensive!)

That backup generator looks better and better all the time...

 

[kwired]

Yet another problem with putting batteries on the DC bus of a VFD is that when you open the disconnect of the VFD to work on it, you DO NOT KILL ALL SOURCES IF AVAILABLE ENERGY so it can be VERY DANGEROUS. You would have to put a separate DC rated disconnect switch on the battery pack that is somehow linked to the VFD enclosure so that you can’t open the VFD unless the batteries are disconnected. Not impossible, but it adds even more cost and complexity. Then if the motor is 460V, the DC bus required will be a minimum of 650VDC, which puts you into a 1000VDC rated disconnect switch; more money on top of more money. (Those Bonitron battery systems take care of all of this by the way, you just don’t see it in the brochure because each one ends up having to be engineered to your system. And they are NOT inexpensive!)

That backup generator looks better and better all the time...

But you may at least still need a lower capacity battery system if you can't take the short downtime between utility loss and generator startup.

 

[drcampbell]

Depends on how critical it is that the load doesn't slow down or stop during transition when utility power is lost. ...

Yes. Some thought and engineering is required, starting with knowing what the specific requirements are.
With a large flywheel and an engine maintained at operating temperature when it's standing by, the engine could be turning the machine less than a minute after losing electric power, without coming to a stop. Whether that's good enough or not, I'll leave for the folks familiar with the system to say.

Typical VFD's rectify the incoming AC volts to DC, then they use semiconductor devices to create a pulsed output ... I would think one could easily tie a battery network into the DC bus and it would easily take over if AC input is lost. ... see no reason why it can't be done. ...

Many of the VFDs I've worked on will raise a "phase loss" fault when the AC power is lost, even if the DC buss voltage is maintained.

 

[kwired]

Yes. Some thought and engineering is required, starting with knowing what the specific requirements are.
With a large flywheel and an engine maintained at operating temperature when it's standing by, the engine could be turning the machine less than a minute after losing electric power, without coming to a stop. Whether that's good enough or not, I'll leave for the folks familiar with the system to say.


Many of the VFDs I've worked on will raise a "phase loss" fault when the AC power is lost, even if the DC buss voltage is maintained.

Like I said typical off the shelf drive may not be so simple to do this with, but I'm sure something could be custom built to do this, may be able to disable the fault detection as well - pretty much have to in cases where using single phase input to the drive.

 

[Besoeker3]

AFAIK, virtually nobody sets up a solid-state UPS to run 50kw for more than an hour.

We did quite a few UPS systems for emergency lighting in public buildings. Typically, they were up to 60kVA for three hours. I never did get an answer to the why three hours for emergency egress lighting. Three hours to vacate a cinema??

 

[broadgage]

As regards the three hours emergency lighting for public buildings, this is a requirement in the UK for all but the smallest and simplest public buildings.
It is IMHO a sensible requirement for the following reasons.

Firstly, if the power fails but without any other emergency, it would be reasonable for persons to remain in the building for an hour waiting to see if utility power comes back on.
After an hour, evacuation might commence, with only TWO hours of battery run time remaining. Now consider that the evacuation might NOT proceed as smoothly as it does during a fire practice.
Suppose that someone falls and breaks a leg, on an upper floor. Consider the time to fetch a stretcher, apply first aid, and carry the injured person down several flights of stairs. This could take more than an hour.
Consider also that the battery that ran for three hours when last tested, might only achieve two and a half hours during the actual emergency 11 months later.
And if the battery only runs for two and a half hours, when tested, is the venue really going to close down until the new battery is procured and fitted. I doubt it.
And finally consider the consequences of a utility failure earlier in the day that has partly discharged the battery.
Utility failure from 05-00 until 08-00 battery fully discharged.
Theatre or cinema opens for afternoon performance at 15-00. Emergency lighting battery only half charged, 14 hours for a full charge is a common requirement..
At 15-30 utility power fails again, battery run time now only about an hour and a half.

The venue staff might be unaware of the early morning power failure, and even if they ARE aware of it the manager might well refuse to cancel the afternoon performance due to a power cut that "was hours ago"

 

[Besoeker3]

Firstly, if the power fails but without any other emergency, it would be reasonable for persons to remain in the building for an hour waiting to see if utility power comes back on.

Would it?
The lighting the UPS feeds is emergency egress lighting. The show would not go on. So what would be the point of remaining?
Even assuming there is not a mandatory requirement to vacate the building?

I know where I worked, if the power went off we had to leave.

 

[tom baker]

One place I worked when the power went out, they wanted us to replace batteries in e lights. If employer sent workers home they would have to pay for time off.

 

[Jraef]

I was once involved in a project to put UPS units on the back room computers at all of the Costco stores, because at the end of each day they would transmit data back to the HQ office, and if power went out during that process, it screwed up their system. So that involved 62 30kVA UPS units at the time, it was a nice project. But being the greedy Gus that I was, I talked the POS manager into doing a study on putting small 500VA UPS units on all of their cash registers, based on the idea that if there was a power outage during store hours, they could at least finish the transaction in process rather than just send those people home. We had it all worked out on how much this would save for Costco and create fewer angry customers, the avoidance of lost transactions alone would pay for it in under a year. The problem I hadn't foreseen; they did not have back-up generators, so most Fire Marshals required that the buildings had to be cleared out within 10 minutes if the power failed. Made me wonder what happens to their cold and frozen foods then.

 

[kwired]

We did quite a few UPS systems for emergency lighting in public buildings. Typically, they were up to 60kVA for three hours. I never did get an answer to the why three hours for emergency egress lighting. Three hours to vacate a cinema??

Well if the emergency starts early in the event and they don't stop the show, some might want to see the entire show

 

[topgone]

Would it?
The lighting the UPS feeds is emergency egress lighting. The show would not go on. So what would be the point of remaining?
Even assuming there is not a mandatory requirement to vacate the building?

I know where I worked, if the power went off we had to leave.

Some institutions see 4 stages during emergencies. 1) decision time upon knowing the emergency, 2)notification time, for the decision-maker/s to transmit info to evacuate, 3) preparation time and 4) response time. Maybe it's different with time estimates for an individuals noticing the emergency and feeling the need to just scram!

We have had an argument with facility owners before re UPS up time which they decided to be 30 minutes. They laughed at us only to experience difficulty shutting down the facility from the usual DCS consoles. When there was a blackout lasting for about an hour and a half, operators ran to another DCS console in another building to access controls of a process area and shut things down! We were the one's laughing last!