UPS Failure's

Status
Not open for further replies.
Scenario: 480V 3/4 system with 3 Phase generator back up on ATS.
Situation: We have a couple of instances where UPS's were blown either during or after transfer.

I have seen UPS's / surge protection strips with MOV's fail if neutral bonding has been done at both the transformer and the panel.

I realize the above is sketchy but any thoughts?
 

iceworm

Curmudgeon still using printed IEEE Color Books
Location
North of the 65 parallel
Occupation
EE (Field - as little design as possible)
Scenario: 480V 3/4 system with 3 Phase generator back up on ATS.
Situation: We have a couple of instances where UPS's were blown either during or after transfer.

I have seen UPS's / surge protection strips with MOV's fail if neutral bonding has been done at both the transformer and the panel.

I realize the above is sketchy but any thoughts?

Yes, it is sketchy.

Where is the transfer switch? Does the generator charge the UPS batteries? Or does it take over the UPS load?

What is the definition of "UPS's were blown"? Did the inverter output semiconductors (or other output parts) fail? Did the DC Fuse feeding the inverter open? Did the charger fail?

The system design and the failure mode(s) would certainly have a lot to do with picking the root cause.

ice
 

Besoeker

Senior Member
Location
UK
Scenario: 480V 3/4 system with 3 Phase generator back up on ATS.
Situation: We have a couple of instances where UPS's were blown either during or after transfer.

I have seen UPS's / surge protection strips with MOV's fail if neutral bonding has been done at both the transformer and the panel.

I realize the above is sketchy but any thoughts?
Was the ATS opened while the UPS was on load?
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Was the ATS opened while the UPS was on load?
When you say transfer, are you talking about line to UPS (or constant power dual conversion UPS?) or are you talking about UPS to generator or generator back to UPS?

In many cases if a constantly on UPS needs generator backup it will be better to put the ATS between the POCO supply and the UPS so that either power source runs through the UPS. Is this what you have?
 

iceworm

Curmudgeon still using printed IEEE Color Books
Location
North of the 65 parallel
Occupation
EE (Field - as little design as possible)
Was the ATS opened while the UPS was on load?

What good is a UPS that can't tolerate the supply being opened when on load?

I can't figure out how anybody knows what the one-line looks like.

Is the gen and utility feeding an ATS, and the ATS output is feeding the inverter charger? Or is it something else?

IDK:dunce:

Could one of you enlighten me?:?

ice
 

iwire

Moderator
Staff member
Location
Massachusetts
Well I have certainly not seen it all but have worked with many UPS and all have been wired pretty much the same.

Utilty feeds ATS, generator feeds ATS, ATS feeds UPS.

UPS carries the load between utility failure and transfer to generator.

Often the UPS will be an online double conversion unit with a 480 volt input and a 208 volt output.

Maintainance bypass or redundant generators may also be in the mix but the end result is regardless of what happens on the supply side of the UPS the output of the UPS stays stable unless the batteries die.

How else do you picture it?:)
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Excellent answer. Foolish question..

ice

Only if you are willing to assume that the OP's installation was rationally designed. And if that is the case, why are the UPSs blowing? :)
And, of course, as asked earlier, how are they failing? What does the postmortem show the damage to be?
 

copper chopper

Senior Member
Location
wisconsin
I had an issue like this a couple years ago, it turned out that the gen set side even though they had the colored wires in the same spot as the utility side, the a ,b,c phases from the gen where reversed from the utility at the gen.( dumbasses). I switched them and it all worked fine.other than that I dont have any other experience with ups systems.
 

templdl

Senior Member
Location
Wisconsin
Does the ATS have in phase monitoring optional module installed?

This has my vote. If there is an out of phase transfer there is going to be a big time issue. The transfer process must be conducted in phase or there will be a significant voltage spike which most likely will fry the UPS.
 

iceworm

Curmudgeon still using printed IEEE Color Books
Location
North of the 65 parallel
Occupation
EE (Field - as little design as possible)
iceworm said:
Excellent answer. Foolish question..

ice

Only if you are willing to assume that the OP's installation was rationally designed. And if that is the case, why are the UPSs blowing? :)
And, of course, as asked earlier, how are they failing? What does the postmortem show the damage to be?

No - I'm not assuming anything. That's why I don't have a clue as to what is going on. And since I can't help, I need to stay out of the way.

However, I am enjoying the responses. I'm waiting for telluric currents, back to back cap switching, harmonics, and my all-time favorite ---- Ground Loops :eek:hmy:

ice
 
All - thanks for the input. The UPS does not have phase monitoring but there are other 3 phase loads switched with no problems - UPS has bumpless transfer. Utility and gen are both connected to the ATS.

I mentioned the bonding because I am leaning towards ground loops.
 

Besoeker

Senior Member
Location
UK
What good is a UPS that can't tolerate the supply being opened when on load?
The output of a UPS is usually from an IGBT inverter.
Drop the whole load if it is inductive and the whole 0.5Li2 has to go somewhere
Not to mention the switching transient voltages.

We do some UPS systems but our main experience is with variable speed drives of various sorts.
These days, most are inverters. Disconnecting the load under power is not generally recommended.
For safety requirements, we usually have to hard wire emergency stop buttons to an output contactor.
An early break contact shuts off the inverter but the primary consideration is to remove power to the load.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
All - thanks for the input. The UPS does not have phase monitoring but there are other 3 phase loads switched with no problems - UPS has bumpless transfer. Utility and gen are both connected to the ATS.

Resistive loads will not notice an out of phase transfer.
Inductive (motor or transformer) loads will react to an out of phase transfer with a voltage spike which might damage insulation as well as a torque bump in the motor output. And maybe a loud noise. :)
Capacitive loads (like the AC to DC converter in a UPS) will react to an out of phase transfer with current spikes easily several times the normal full load current. This could fry the input circuitry, since semiconductors do not handle massive pulse overloads very well (at least not when compared to copper wires.)
The similar but smaller AC pulses on turning on a switch or breaker on the input are often handled with soft-start circuitry like varistors, but in this case the varistor is already hot when the surge hits.
If the input of the UPS is via a transformer, an out-of-phase transfer might also generate an overvoltage into the next stage.
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
What good is a UPS that can't tolerate the supply being opened when on load?
Ah, but the reconnection, on the other hand.....
1. If there is no load, the DC filter capacitors after the input rectifiers (assuming that is the design) will remain reasonably charged, limiting any current surge on reconnection.
2. If there is load, the DC filter capacitors could be drawn down to the point where the re-connection surge pulled excess current.
 

iwire

Moderator
Staff member
Location
Massachusetts
Ah, but the reconnection, on the other hand.....
1. If there is no load, the DC filter capacitors after the input rectifiers (assuming that is the design) will remain reasonably charged, limiting any current surge on reconnection.
2. If there is load, the DC filter capacitors could be drawn down to the point where the re-connection surge pulled excess current.

I think you are reaching for some reason.

The function of a UPS is to provided an uninterrupted power supply to the load.

If the UPS cannot handle powering on and off on the supply side it is worthless.
 
Status
Not open for further replies.
Top