advice on UPS.

Status
Not open for further replies.
electrofelon

electrofelon

Senior Member
Location
Cherry Valley NY, Seattle, WA
I have a mass spectrometer that I need to provide a UPS for. The machine consists of 3 components, and take a 230/400V supply:

1. The main machine is 8A @ 230V (L-N)
2. RF generator is 10A @ 230V (L-N)
3. Vacuum pump which is 7A three phase (400V)

IF I put the main machine and RF generator on different phases, I get a max phase current of 17A. I am just going to assume a UPS cant put more than a third of its rating on one phase, so assuming 17A on ALL phases I get 12KVA. We were originally looking at a 20KVA/18KW UPS, however I am a bit concerned about start up current of the pump, PLUS someone we were speaking to from the UPS manufacturer said this:

We escalated this to our 3 phase expert and he provided this feedback:


I am afraid we do not have a solution for this application.


First, our UPSs are not really designed for inductive load applications. Any loads containing, pumps, motors, compressors and/or any components that produce high inrush in a repetitive cycling which is very harsh on the internal components of the UPSs. The customer may end up purchasing an expensive system that will experience premature wear resulting in system failure.


In terms of unbalanced loads, to be honest there is not a perfect customer out there that has a perfect balanced loads, however our engineering team calls for the recommended line-to-line load unbalance to be less than 50%. That would be another concern for this application. Remember that an unbalanced system produces more heat in the wirings , rise in neutral current, stress on the transistors etc.
Click to expand...

I am a little skeptical of some of that. For one thing, the pump will not be repetitively cycling, probably just on and off a few times per day. Also I find the comments on unbalanced systems to be kinda misleading. More heat in the wiring? Rise in neutral current? Stress on transistors? Well if you are designed and spec'ing around a known imbalance and are well within all specs, I am not seeing that as an issue.


My thinking is if we go with a 30KVA/27KW unit, that should cover the startup current of the pump (its nameplate is actually 4.8, not the 7 shown on the general spec sheet). and should provide plenty of extra capacity for any inductive loading of the pump. So I just wanted to see if anyone had any comments or experience. I know there is no definite answer here as they just dont provide any specifics of allowable types and percentage of loading, other than their comment "its not really designed for inductive loads". Seems to me oversizing the unit which would also account for the inrush, would provide plenty of robustness and make the inductive load component a relatively small portion of the nameplate output. I have attached a picture of the vacuum pump.
 

Attachments

  • pump.jpg
    pump.jpg
    241.4 KB · Views: 13
My guess is the manufacturer is giving good advice to be cautious about the LRA, motor starting load. I recall seeing specs for a large UPS, it had a fault clearing mode rating of 200 % for 2 seconds.

What the output transistors can handle is one issue but the UPS would be expected to go into self protect mode much faster than the damage to the output transistors would happen. If they say it will not like the motor, likely they are right about that.

The load imbalance would not concern me and I would expect it should not concern the UPS either. But if the manufacturer warns away from that also, I would listen to them.

Perhaps the customer can live with the vacuum pump not on the UPS and keep the large savings, would be what I would suggest to them. Most outages are less than a few seconds and the vac pump could ride through those easily, while the UPS would be there to clean power for the truly sensitive load. The customer could choose to save the money as a good bet.
 
__dan said:
My guess is the manufacturer is giving good advice to be cautious about the LRA, motor starting load. I recall seeing specs for a large UPS, it had a fault clearing mode rating of 200 % for 2 seconds.

What the output transistors can handle is one issue but the UPS would be expected to go into self protect mode much faster than the damage to the output transistors would happen. If they say it will not like the motor, likely they are right about that.

The load imbalance would not concern me and I would expect it should not concern the UPS either. But if the manufacturer warns away from that also, I would listen to them.

Perhaps the customer can live with the vacuum pump not on the UPS and keep the large savings, would be what I would suggest to them. Most outages are less than a few seconds and the vac pump could ride through those easily, while the UPS would be there to clean power for the truly sensitive load. The customer could choose to save the money as a good bet.
Click to expand...
For this unit they state 130% for 1 second. So 43 amps X 1.3 = 56 amps available to start a 4.8A motor. That seems safe to me.

Unfortunately this is a 50 HZ only pump and we are we are using the UPS to make 50 hertz from the 60 hertz so we can't run the pump separately from the UPS. A new 60 HZ pump is about the same cost as the UPS (12k)
 
The UPS has maintenance and battery lifetime issues (expensive), which may be avoided as not necessary. You can make 50 Hz without needing the batteries and the maintenance that goes along with that.

So you have 5 kW of other load. I would try to avoid the extra 25 kW of UPS just to start the vac pump motor. The alternatives must offer some money savings, It is the customer's call if they want to spend the substantial extra money to UPS the vac pump. The battery maintenance and battery lifetime issues make it uneconomical unless absolutely necessary. Possibly they do not need any UPS for the vac pump, I would advise.
 
electrofelon said:
Unfortunately this is a 50 HZ only pump and we are we are using the UPS to make 50 hertz from the 60 hertz so we can't run the pump separately from the UPS. A new 60 HZ pump is about the same cost as the UPS (12k)
Click to expand...

If the mechanical load presented by the vacuum pump allows it, perhaps a VFD could be used to create the 50Hz and also reduce the peak current required for starting the motor. A sine wave filter would be needed reduce any conducted switching noise. A UPS will have a PWM inverter inside of it like a VFD does, but it has filtering to provide a sinewave output. The manufacturer of the mass spectrometer might be consulted whether there could be a possible interference issue when the vacuum pump was operating on a VFD with a sine wave filter. If the vacuum pump just needs to pull a vacuum initially before the equipment is operated, then interference might not be a concern.

Perhaps the sensitive loads of the main machine and RF generator could both be placed on one single-phase UPS. And another UPS (single or 3-phase) could then feed a VFD to provide isolation from the other equipment.
 
I get the desire to pull off the vacuum pump, I would like to also. However they seem rather set on having the pump also on the UPS. I am not sure how long the machine can ride thru without the vacuum pump, but there is definitely value in having the machine be able to keep running without interruption during a short outage. My understanding is this is something like a half million dollar machine. Maybe I can convince them to reconsider, but we stil have the frequency issue then, so a VFD would be needed. We could possibly run the pump at 480 60 hz per the data plate, but that is for the motor itself, and not sure whether that is acceptable to the pump itself.
 
The vacuum pump probably works with a small storage tank or chamber. Probably takes some time to pull down, but that would also give it some storage if the pump happened to start and stop. Probably, there is some monitoring of vacuum pressure, so it throws an error and shuts down the rest of the system if it detects the vacuum is out of spec.

It's some additional engineering questions to investigate and consider, probably I would guess non paid. They would have to weigh the cost of installing the UPS and maintaining it compared to what it cost them to have their experiment interrupted, and how often that happens. Then of course it's their money to spend. I am sure 9 out of 9 guys would just say 'yup, great idea' to sell the job. Seen that often enough.

If they can ride through the 99% (wag) of interruption less than a few seconds, their cost difference would be a 5 hp drive compared to the 25 kW extra UPS. They could decide to try the drive to see what the actual interruptions are, then it's easy to upgrade later because they have no money into the play at that point (not a material cost error). Going the other way, from a 25 kW UPS down to a drive for the same effect, that I would say could be a material error, but not so big that that I've not seen huger ones elsewhere.
 
If you wanted to supply a motor from a UPS source, you have to size the UPS according to the locked rotor amps of the motor. Since you have other loads, just get the highest line amps plus the locked rotor amps of one phase. In your case, say you consider the higher single-phase load plus your locked rotor amps = 10A + 41.4 (LRA of a 3/4 HP motor, FLA=7) =51.4A! That would be 35.6 kVA! ~40 kVA, 400V!
 
I would actually expect the vacuum pump to run continuously, and thus would not worry about the pump cycling and stressing the ups system.

Jon
 
When we provide UPSs for inductive loads, we oversize the UPS to be sure the LRA does not activate the overload condition of the UPS putting it into bypass (guessing this is a high end UPS you are looking at with static bypass, and maybe maintenance wraparound bypass).
We have sized static UPSs to accommodate chilled water pumps and the like, so it just needs to be oversized.
 
ron said:
When we provide UPSs for inductive loads, we oversize the UPS to be sure the LRA does not activate the overload condition of the UPS putting it into bypass (guessing this is a high end UPS you are looking at with static bypass, and maybe maintenance wraparound bypass).
We have sized static UPSs to accommodate chilled water pumps and the like, so it just needs to be oversized.
Click to expand...
Yeah that is my thinking
 
electrofelon said:
Yes, that is what the lab director said, that it basically runs continuously.
Click to expand...
I was under the impression that you needed a UPS to prevent interruptions to your operations! UPS designs are there to provide power during temporary power loss of supply to the equipment. They are rated in minutes, and they will transfer when the backup generation comes online.
If not, what you need is an inverter that supplies whatever you have.
 
topgone said:
I was under the impression that you needed a UPS to prevent interruptions to your operations! UPS designs are there to provide power during temporary power loss of supply to the equipment. They are rated in minutes, and they will transfer when the backup generation comes online.
If not, what you need is an inverter that supplies whatever you have.
Click to expand...
Topgone, I am not understanding what you are saying here. Yes we want the UPS so we can "ride thru" brief power outages and "blips". In this particular case, we also need 50hz and the UPS will do that.
 
electrofelon said:
Topgone, I am not understanding what you are saying here. Yes we want the UPS so we can "ride thru" brief power outages and "blips". In this particular case, we also need 50hz and the UPS will do that.
Click to expand...
Not splitting hairs here. Be aware of snake oil salesmen, they might claim their product provides xxx volts at 50 Hz and leave out info on how long their unit can continually supply the xxx volts @50 Hz that you need.
UPSs are not designed for longer uptimes. Although every UPS has inverters in them, they are only rated for short durations. UPSs need the backup generators to come online and then transfer loads as they are available.
A suitable equipment called an inverter, rated for the loads you are having is the equipment you'll need. Those are designed to carry AC loads 24/7 while deriving its power from a DC source. DC because the AC feed available is 60 Hz and you wanted 50Hz. You cannot get 50 Hz from the AC mains without having to rectify the AC source (which is 60 Hz) to DC and back to AC again at 50 Hz. As a consequence of using an inverter, you also get to enjoy the capacity to ride over blips in the AC supply because your battery that is paralleled to your rectifier output at the DC bus makes it "uninterruptible".
 
electrofelon said:
I have a mass spectrometer that I need to provide a UPS for. The machine consists of 3 components, and take a 230/400V supply:

1. The main machine is 8A @ 230V (L-N)
2. RF generator is 10A @ 230V (L-N)
3. Vacuum pump which is 7A three phase (400V)
Click to expand...

Does component 3. containing a vacuum pump only have the motor that's shown in your picture for an electrical load, and no other internal loads? The reason I ask is that a mass spectrometer will need additional pump(s) to lower the pressure even further than a low RPM roughing pump, such as a turbomolecular pump running at very high RPM or a diffusion pump, and maybe associated controls. If you use a VFD you'd only want to connect it to the motor in your picture and not to other loads that might need a constant 50 Hz.

Bonitron has systems providing "Uninterruptible Power for Drives":
https://www.bonitron.com/complete-systems
https://www.bonitron.com/undervoltage-cs-battery.html
 
zbang said:
Huh? Dual-conversion UPSs run the inverter continuously. Line-interactive and off-line UPSs do not.

BTW, and usual UPS sizing criteria is for 15 minutes of runtime, and that's based on battery capacity.
Click to expand...
That's my point. He might be fooled into believing line-interactive/ offline UPS works like a dual-conversion UPS. There are a lot of sharks prowling the electric equipment market nowadays!
 
Status
Not open for further replies.
Top