UPS on a Kitchen GFI outlet

[pkinkc]

In a retail facility, client has a food service counter for selling pizza by the slice, sandwiches, espresso, etc. Per NEC 210.8 it is a "kitchen", so all 20A receptacles are gfci, including the outlets at the Point of Sale.

Client intends to plug in a power-strip style point-of-use UPS so the computer has uninterrupted power.


The concern: EMI filters in the UPS may leak enough current to ground that it trips out the GFCI. If that happens, or if the GFCI trips out for another ground fault event, the UPS will immediately re-introduce 120V current back into the system.

I find nothing in the NEC addressing this. It seems problematic, but is it a violation of a code? Engineers can design a code-compliant system, but cannot control what end users plug into outlets.

 

[qcroanoke]

In a retail facility, client has a food service counter for selling pizza by the slice, sandwiches, espresso, etc. Per NEC 210.8 it is a "kitchen", so all 20A receptacles are gfci, including the outlets at the Point of Sale.

Client intends to plug in a power-strip style point-of-use UPS so the computer has uninterrupted power.
View attachment 10230

The concern: EMI filters in the UPS may leak enough current to ground that it trips out the GFCI. If that happens, or if the GFCI trips out for another ground fault event, the UPS will immediately re-introduce 120V current back into the system.

I find nothing in the NEC addressing this. It seems problematic, but is it a violation of a code? Engineers can design a code-compliant system, but cannot control what end users plug into outlets.

The UPS will not feed any outlets but those in the UPS. With or without utility power.
Are you asking if the UPS will try to back feed the circuit it is plugged in to?

 

[pkinkc]

I know it won't backfeed into the circuit it's plugged into. I'm saying that the outlets on the UPS would still be hot, even though the GFCI outlet it's plugged into tripped. The computer, printer, monitor, and anything else plugged into that UPS will still be hot.

 

[qcroanoke]

I know it won't backfeed into the circuit it's plugged into. I'm saying that the outlets on the UPS would still be hot, even though the GFCI outlet it's plugged into tripped. The computer, printer, monitor, and anything else plugged into that UPS will still be hot.

That's true.
I see what you are saying, That could pose a problem code wise but the only option may be to remove the UPS if GFI protection is required where the computer is located.
Maybe others will give their thoughts on it.

 

[kwired]

Interesting problem, that said when this unit is powered from the inverter - what kind of voltage reference is there to ground? Chances are there is no reference to ground and it essentially is an ungrounded system when running from the battery, and the risk of shock is greatly diminished.

That or the unit may even have it's own ground fault detection to protect the inverter.

 

[mgookin]

...

That or the unit may even have it's own ground fault detection to protect the inverter.

That's interesting. Anyone ever plug a GFCI tester into a UPS and press the button?

And what do we do with the kid with a pc in his bedroom? No AFCI there, right?

 

[kwired]

That's interesting. Anyone ever plug a GFCI tester into a UPS and press the button?

And what do we do with the kid with a pc in his bedroom? No AFCI there, right?

I didn't necessarily mean it would have class A GFCI protection, but could easily see it having similar ground fault protection as VFD's have.

 

[iwire]

The UPS will not feed any outlets but those in the UPS. With or without utility power.
Are you asking if the UPS will try to back feed the circuit it is plugged in to?

In my opinion the NEC does not apply to outlets on listed equipment.

 

[kwired]

In my opinion the NEC does not apply to outlets on listed equipment.

I would have to agree with that as well.

NEC applies to the premises wiring, what someone happens to plug in later is not covered - though often people try to make it apply.

There are situations where fastened in place utilization equipment may or may not be permitted to be cord and plug connected, but I don't see the UPS in the OP fitting in to what NEC applies to.

 

[dereckbc]

In my opinion the NEC does not apply to outlets on listed equipment.

Ditto to that. Appliances fall under UL domain. Not certain the picture shown is a UPS due to its very small size. If it is a UPS I doubt it would be of much use.

Looks more like a Power Strip, and most likely with most power strips they only have Class C or D Surge Protection Devices (SPD's) like MOV's and SAD's. In Class B, C, and D SPD are installed L-G and N-G by design and if used on a GFCI circuit will certainly operate the GFCI interrupter if the surge has enough energy sustained long enough to be detected.

If by chance the device has any has any RFI circuitry, they may well be installed between L-G and N-G which again may cause false GFCI operation especially in cheaper models which are known to leak normal line frequencies to the ground circuit.

 

[kwired]

Ditto to that. Appliances fall under UL domain. Not certain the picture shown is a UPS due to its very small size. If it is a UPS I doubt it would be of much use.

According to the label on it - it is an uninterruptable power supply. It also seems to indicate the far side receptacles are connected to the "supply" only and the near side receptacles are both supply and backup powered.

Probably not a pretty low reserve unit though.

 

[tom baker]

Use a portable GFCI plugged into the UPS.

 

[kwired]

Use a portable GFCI plugged into the UPS.

Great, if you know electrical theory and codes, most that don't probably are not going to do this or even have a clue there is any potential hazard.

 

[dereckbc]

Probably not a pretty low reserve unit though.

Just the physical size tells you it has very little reserve capacity. At best if it uses Lithium batteries is no more than 100 watt hours of capacity. If NiCd or lead acid 1/4th of that.

 

[K8MHZ]

The main purpose of small boxes like the OP's is simply to keep a computer on long enough for an orderly shut down. When power fails, software starts shutting the computers down. There is a USB connection between the UPS and the computer.

Most of these devices have a run time of 6 - 9 minutes.

I don't think there really is a use for a GFCI for that short of a length of time.

My unit has been UL un-approved by the addition of a 100 AH external battery. Run time is several hours. I don't know for sure, but I don't think there is a grounded conductor when the unit is running off DC. Some day I will check. So, if there is no grounded conductor, there is no potential between earth and either of the 120 volt conductors, making a GFCI nearly useless.

 

[kwired]

Originally Posted by kwired Probably not a pretty low reserve unit though.

Just the physical size tells you it has very little reserve capacity. At best if it uses Lithium batteries is no more than 100 watt hours of capacity. If NiCd or lead acid 1/4th of that.

Strike the work "not" in my post and replace with "is". I did not mean what was written there, and likely started to say one thing and then changed it, but did not read the end result all that carefully.

The main purpose of small boxes like the OP's is simply to keep a computer on long enough for an orderly shut down. When power fails, software starts shutting the computers down. There is a USB connection between the UPS and the computer.

Most of these devices have a run time of 6 - 9 minutes.

I don't think there really is a use for a GFCI for that short of a length of time.

My unit has been UL un-approved by the addition of a 100 AH external battery. Run time is several hours. I don't know for sure, but I don't think there is a grounded conductor when the unit is running off DC. Some day I will check. So, if there is no grounded conductor, there is no potential between earth and either of the 120 volt conductors, making a GFCI nearly useless.

I would think the EGC terminal is still connected to the incoming EGC of the premises wiring system, but the EGC likely has no solid reference to the inverter output. The inverter may have fault sensing components built into it though that could shut the output down, similar to how VFD's detect ground faults. A simple plug in tester may show no EGC when plugged in while running on backup power, but would show its presence when running on normal power.