Multiple (unintentional) grounds in subpanel

[techntrek]

I've split out 11 circuits from my 2 service panels to a subpanel fed by a hardwired 8kva UPS (double-conversion). I've found that about 2/3 of the utility outages - when testing and real - end up with a UPS failure. It maintains load for about a second and then fails hard, w/o even a message in the onboard log. The 22kw battery bank is all new and very oversized for the load, so DC voltage sag or old batteries aren't the problem. After more research I started thinking it may be due to a grounding issue. Apparently UPSs are sensitive to ground problems.

After disconnecting the subpanel feed including the ground, I got no continuity between neutral and ground so that rules out multipath neutral return due to a panel or circuit miswire/failure. But, when testing between an adjacent genset subpanel's ground and the UPS sub ground, I got continuity. So then I disconnected all the circuit grounds and tested each one individually. I found 4 of the 11 circuits had continuity. Some of the re-routed circuits didn't have enough length to be pulled directly to the sub so I ran stubs between the service panels and the subpanel; I double-checked to be sure the ground lines on all re-routed circuits do not touch any ground line or the service panel chassis. However, one of the 4 problem circuits does not double-back, it goes directly to the furnace. So I'm left with the primary ground from the 4-wire feeder, and these 4 additional grounds that somewhere, somehow feed back to ground independently. My house was built by the original owner and I've already corrected other structural and wiring issues I've found, I don't know how it passed the various inspections. It would not surprise me if he ended up bonding grounds from different circuits elsewhere in the house.

The question: before I pull my hair out trying to track down the locations of these additional ground points, should I even worry about it from a code or saftey standpoint? Anyone have any experience with UPSs and think this might be the source of my failures?

 

[petersonra]

All the grounds (green wires) should be bonded together so they should show continuity.

In normal operation all the neutrals (white wires) will show continuity because they are all supposed to be bonded to a green wire somewhere in the circuit.

If any ground or neutral does not show continuity to any other ground or neutral, it is not wired correctly.

I don't see how improper grounding (or neutraling) could case a UPS to fail. It might trip it, but I don't see how it could cause a failure.

 

[techntrek]

I'm not sure you understand the problem. All of the grounds are bonded at the subpanel, likewise all of the neutrals, with the only common bonds at the 2 service panels. The problem here is when the subpanel is disconnected - all 4 wires - I'm still finding that its connected to ground. Specifically by 4 of the 11 circuits I moved to that panel. The other 7 circuits do not show any continuity when tested against a known ground (when tested individually).

"Trip", as in trip a breaker? In my case no breakers trip, it triggers the automatic bypass to "utility" and immediately shuts down.

 

[John25]

For clarity, please add -ing or -ed to ground.
Grounded conductors or grounding conductors?

 

[hillbilly1]

I've split out 11 circuits from my 2 service panels to a subpanel fed by a hardwired 8kva UPS (double-conversion).

However, one of the 4 problem circuits does not double-back, it goes directly to the furnace.

If I read this correctly, the furnace is on the UPS? Most UPS units do not like motor loads, The startup current is usually too long for the UPS to pick it up without going into bypass.

 

[techntrek]

For clarity, please add -ing or -ed to ground.
Grounded conductors or grounding conductors?

I'm equating "neutral" to "grounded conductor" on the branch circuits. The overall system has 3 "grounding conductors" (one from each service panel, one from the genset) running to two ground rods, about 4' from each other.

 

[techntrek]

If I read this correctly, the furnace is on the UPS? Most UPS units do not like motor loads, The startup current is usually too long for the UPS to pick it up without going into bypass.

It doesn't fail or go to bypass during normal operation. Double-conversion UPSs run the load on the inverter all the time (completely isolating the load from the utility). With an 8kva/6kw capacity, it can go to 150% for 10 minutes or 200% for 10 cycles. The blower registers about 300 watts on the built-in ammeter with no noticeable bump on startup. I also run my 15 amp chain saw and 15 amp log splitter from the inverter and they briefly jump to about 110% on startup without triggering bypass.

 

[benaround]

tnt,

It is not agianst the NEC nor is it uncommon for the EGC's of different circuits to be

spliced together.

 

[techntrek]

tnt,

It is not agianst the NEC nor is it uncommon for the EGC's of different circuits to be

spliced together.

True inside all panels, but in my case this may be happening outside of a panel, somewhere else on the circuit. I just want to be crystal clear this is the situation that you are confirming.

For instance, the furnace is propane supplied so the gas line should have a grounding conductor back to one of the service panels, which would explain why that circuit is showing an extra ground connection somewhere.

 

[don_resqcapt19]

All EGCs within a building or structure should have continuity to each other, no matter what their source is.

 

[benaround]

I just want to be crystal clear this is the situation that you are confirming.

QUOTE]

Yes, it is.

 

[wirenut1980]

It doesn't fail or go to bypass during normal operation. Double-conversion UPSs run the load on the inverter all the time (completely isolating the load from the utility). With an 8kva/6kw capacity, it can go to 150% for 10 minutes or 200% for 10 cycles. The blower registers about 300 watts on the built-in ammeter with no noticeable bump on startup. I also run my 15 amp chain saw and 15 amp log splitter from the inverter and they briefly jump to about 110% on startup without triggering bypass.

So it sounds like the UPS only fails (shuts down) during "most" of the switchovers from utility to generator. Does the failure occur right at the time of switching sources? Has it worked fine for some time and then started happening or has this always happened?

 

[Pierre C Belarge]

So it sounds like the UPS only fails (shuts down) during "most" of the switchovers from utility to generator. Does the failure occur right at the time of switching sources? Has it worked fine for some time and then started happening or has this always happened?[/quote]



This would be my first question.

#2. has any work been performed around the same time this issue has popped up?

 

[techntrek]

Well, its not failing during a switchover to a genset, its right after the loss of utility. IF (big if) it holds out after 1 second it will run for days on battery with a light load, or up to a day with normal useage. When utility or genset come online it always makes a smooth transition back. Transition is a better word than switchover, since there is no physical switch via a transfer switch like with cheaper offline inverters.

It didn't happen the first month or so, and no major changes were made anywhere when it started. As I mentioned before, my first reaction - and the solution to most inverter issues - was to replace the battery bank which I completed a month ago.

 

[wirenut1980]

Well, its not failing during a switchover to a genset, its right after the loss of utility. IF (big if) it holds out after 1 second it will run for days on battery with a light load, or up to a day with normal useage. When utility or genset come online it always makes a smooth transition back. Transition is a better word than switchover, since there is no physical switch via a transfer switch like with cheaper offline inverters.

It didn't happen the first month or so, and no major changes were made anywhere when it started. As I mentioned before, my first reaction - and the solution to most inverter issues - was to replace the battery bank which I completed a month ago.

So the UPS shuts down sometimes only when the battery is called on to provide the power. I understand that the batteries are always in the circuit since this is an online UPS, but my guess is there is something wrong with the UPS rectifier or inverter. Although the fact the shutdowns don't happen all the time is troubling. I wonder if the circuit configuration in the UPS is any different when there is source power as oppposed to on battery power? Does the UPS manufacturer tech support have any ideas?

 

[brian john]

It ain't the grounds.

 

[hillbilly1]

It doesn't fail or go to bypass during normal operation. Double-conversion UPSs run the load on the inverter all the time (completely isolating the load from the utility). With an 8kva/6kw capacity, it can go to 150% for 10 minutes or 200% for 10 cycles. The blower registers about 300 watts on the built-in ammeter with no noticeable bump on startup. I also run my 15 amp chain saw and 15 amp log splitter from the inverter and they briefly jump to about 110% on startup without triggering bypass.

Is it possible that the charging circuit is absorbing some of the load while on utility or generator? Have you tried the chain saw or log splitter without the utility or generator supplying the UPS?

 

[don_resqcapt19]

It doesn't fail or go to bypass during normal operation. Double-conversion UPSs run the load on the inverter all the time (completely isolating the load from the utility). With an 8kva/6kw capacity, it can go to 150% for 10 minutes or 200% for 10 cycles. The blower registers about 300 watts on the built-in ammeter with no noticeable bump on startup. I also run my 15 amp chain saw and 15 amp log splitter from the inverter and they briefly jump to about 110% on startup without triggering bypass.

An electric motor draws at least 4 to 6 times its full load current on start-up. It appears that the built-in meter is too slow to catch this bump in current.

 

[techntrek]

wirenut - now that I've ruled out batteries or ground issues (thinking maybe multiple grounds were causing capacitance or inductive problems), tech support may be my only & last option...

hillbilly - the DC supply inside the inverter is always supplying the internal charger and inverter. In double-conversion UPSs it goes utility>PF corrector>RF filters>rectifier>charger & inverter. I've run the chain saw and splitter from battery several times for hours, and as with other loads the running isn't the problem, its the initial transition. (If you wonder why I run them from the inverter, its because it has an option to output at 127V, so it lets me get more than 50' from the house.)

don - actually the furnace blower doesn't register a surge because its a multi-speed model and always starts at a very slow RPM before ramping up. The refrigerators, freezer, chain saw and other normal inductive loads all show a nice bump when they start.

Thanks to everyone for confirming that the ground issue isn't a safety/code problem, and for mulling my UPS issue.

 

[Ibarra]

Tech,

Please check DC voltage. We found this problem in UPSs when we there is an open battery or a false contact.

Ric