For simple reference:



The picture shows a commercial building with a gutter and two meters. The gutter drops to each meter/main and on the left side is the circuit panel that feeds sub panels inside the two buildings. The tenents in building A have complained and I have varified that their voltage drops considerably and their computers shut off. This might happen 2-5 times a week. When I tested, I saw the voltage at 100V. I then went to the inside subpanel and tested several times however each leg was 120V and the screw side the the breakers was good too. I tightened all the neutrals I could opened the panels and checked for anything loose but didn't see anything. What I did notice was the other meter/main (Left side) had a burning hot spot . That meter/main is connected to the subpanel on it's left with a small chaise nipple. On both sides of the chaise nipple where the lock nut tightens against the panel there were at least one glowing hot spot that came and went every so often. I did have the power company come out and do a through inspection of their side and they said the voltage was steady. I may have been a bit permature in doing that but I was out of ideas.
I would appreciate some other ideas on what to check before I go out there next.

Let me know if you need more information.

Many thanks,
Byron

edited to add picture

[This message has been edited by electure (edited 06-23-2006).]

I ran into a similiar situation years ago. Low voltage condition intermittently at some receptacles in one of three shops.

What my problem turned out to be was some prior electrician had bonded the grounded and grounding conductors to each sub-panel via installing lugs and bolting them to the main disconnect enclosure, rather than installing a single bus with enough openings for all the conductors.

The neutral current flow would depend upon the connection to the enclosure for low impedance. I see the parallel with yours in the "burning hot spot".

I fixed the problem by replacing all the lugs with multi-opening bus, and connecting all enclosures together with a copper bonding conductor, directly connected to each grounded and grounding conductor.

SHUT IT DOWN...

POCO, may have only checked thier side of the equipment, most wont go any further. "NMJ" But wise to rule it out...

"prior electrician had bonded the grounded and grounding conductors to each sub-panel via installing lugs and bolting them to the main disconnect enclosure, rather than installing a single bus with enough openings for all the conductors."

Nothing wrong with that... So long as ground on ground bar, and neutral seperate in main panel, and bonded there, even in multiple services.

Anyway, I'm lost as to what feeds the affected area? But either way it sound to me that the panel on the left is either carrying load or fault current on the ground. This is a potentialy dangerous situation if left to continue for any period. In my professional opinion, it should be shut down until located. As it is a high fire or shock risk.

You could locate this a few different ways...

The hot spots are resistive, and creating heat due to current flow, but only a symtom of the real problem. But will need to be corrected any way. Don't fix this yet, it will aid you in locating the fault.

By now the whole kit and cabootle of the affected panels should be off.... Not kidding. Wait for the hot spots to cool! If they don't you have other problems, like current flowing out toward branch areas = BAD! (Bigger problems - loss of GEC, and the system found a better one for you...) If they do cool off... Attach a solidly connected insulated #6 between the panels with the hot spots, or better between the the best grounded portion of the panels closest to the GEC, and the other side of the affected panels. Then turn breakers on one at a time, and then back off while checking amperage on that #6 you installed. The one that gives you an amperage reading jump is the fault, if all of them give an amperage reading to some degree, the neutral is toast. Or a ground and neutral are connected down the line, and this is parralel current coming back.

I'll opt for neutral is toast! (with secondary screw up's.)

Good luck, but get on it quick.....

Perhaps I did not make myself clear.
In my case, the neutral current was flowing through the main enclosure, through the equipment, and then back to the transformer. There should not be any current flow normally through the locknuts on your conduit. Only in fault conditions. With normal voltages sometimes, and low voltages other times, I'm guessing a high resistance neutral connection somewhere.

The "burning hot spot" says it all. Kill it NOW!

Well I have been reading post hear for almost a year but never given my thoughts. This one I will put my 2 cents in. Last fall I got a call from a Handy man that I do electrical work for he told me that the house he was working on had a bad GFIC in the grudge and when he replaced it half the house lost its power. I went over tested power at plug got a low volt reading I went to meter got same on one side of main checked buss coming from meter and same low volt. I called POCO they came out and found power to meter was fine I told them meter must be bad they pulled meter I looked in and seen the buss had been burnt look a little farther and seen were the manufacture left out the bolt that connects the meter socket to the buss we clean it all up and put in a bolt.This house was about 25 years old. I was amazed it lasted that long with the two metals just touching. So what I want to say is check all connections by manufactures as well as others. Cal

Quick update: POCO went out there and put their testing equipment on the load side of the meters and left it for a couple days. They said it varied only 2 volts.

I went out there today to check on a few things. I only looked at one meter/main the sub panel to it's left and the three sub-panels that are fed on the inside. I did find a neutral that was loose in one of the inside sub-panels. I figured this was it. I then opened all panels and tightened everything. I then checked my voltage at a few outlets. I found that the voltage was high this time (128); a problem the tenant told me about. Infact he said if it wasn't around 100V it was up around 140V.
Still collecting data, I decided to check the meter/main case to the sub-panel case it fed to the left. I got around 30V. I then decided to shut the sub-panel breakers off (all of them) and test the ground to the load side of those breakers. Again, around 30V. Finally, I decided to shut the meter/main off and test it's case to the sub-panel case; around 30V.
The one thing I did notice is there isn't a good bond between the meter/main case and the sub-panel case. They rely on a chase nipple and lock nuts. I know to neutralize the case to case voltage I could put in a bonding jumper, but this doesn't fix the real problem.
I did examine the GEC and it seemed to be in fine shape. The wire went underground so I couldn't see if there were any further problems with it.

I'm open to suggestions.

Thanks

I've said it, Reno said it again, I'll say it again...
"The "burning hot spot" says it all. Kill it NOW!

The two cans should NOT be carrying current! This is NOT a time to be collecting data in a situation like this. A high/low voltage problem is indicative of a lost neutral, and should not be left in this manner for any amount of time...

We can not go troubleshoot it for you, but it is obvious that the neutral is gone! And it sounds like a second connection of the grounded conductors (Neutrals) to the ground(s) of the sub-panel.

I hope your liability insurance is top notch.... The joint could be on fire right now, some items react that way with the voltages you've described.

Heres a simular example . Ignore the roaches and needle containers, focus on what happens when a ground becomes a current carrier, and choosing paths on its own.

Byron,
I am somewhat baffled that, as a licensed EC, you don't seem to feel a bit of urgency in the fact that you're describing a metal box and fitting GLOWING RED HOT!! It's simple to figure out that you've lost a neutral somewhere downstream (It's not the poco's issue if they already checked their side under load,) and current is finding a HIGH impedance path back to the main service. I certainly hope no one bumps their arm on that service and gets nailed !!

SHUT THE POWER OFF!!! If you can't find an obvious place where the neutral/ground are being tied together, other than at the main service point, then
turn all the branch breakers off, disconnect your neutral feeders to the subpanels at the main service disconnect and ring them out to ground. (Open noodle test) You shouldn't get anything. Likely from your described scenario, at least one will give a reading.. Follow it until you find the point downstream where it has faulted and/or been tied together with your grounding. (a toner might be helpful here incase the wire itself might have failed underground)Follow it from point "A" to point "B". If youre lugs were tight and not corroded at each end, you'll find it along the way with the toner when you lose tone.

I might also suggest replacing the parts that became "burning hot", and even though it may not be required, pull a properly sized ground wire to help eliminate "glowing fittings" should another fault occur down the road. You stated you found loose fittings here and there... It's an easy "peace of mind".

Of course, this is all dependant on your jobsite not becoming a "Structure Fully Involved" call to your local fire department before you decide to lock the power out!!

I did mention shutting the power off to the owner of the complex and she flipped. Believe me, I have expressed the importance of getting this fixed but she says it has been going on for over a year and she seems to be more interested in tenents paying rent and being happy. This problem is noticed by only one tenant. She only hired me about a week ago to check it out. I'm not sure how many electricians have looked before me. You know how that is.

If I shut off the both meter/mains and then go from the case to the water/ground below and get a reading of 30Amps don't you think that should be POCO's problem??

Thanks

If you're getting an amperage reading, that means there is a load someplace that wasn't shut down... Might check that a pole of one of the disconnects isn't siezed/fused closed. If you meant volts, then that difference in potential between the metal equiptment casing and the ground rod, water pipe or whatever you're measuring to is a bonding issue. Chances are you have a bonding jumper/bonding screw, etc. missing or corroded. Your equiptment casing and bonding wouldn't typically be a PG&E issue, unless youre in equiptment that they supplied.

No, not nessesarily.... Although that, if an amperage? does seem a little high. The gorund there, as with all connected to that tranformer serving the area are grounding the POCO side, some less equally than others. But in the situation you describe, it could or would sound much more like what might be found with the lost neutral load on....

Rule of thumb for PG&E, voltage on thier side of metering OK, "NOT thier problem"....

I will throw this question out there, 'cause I know a number of people who don't break seals. Did you open the By-pass section of the metering? And inspect the neutral there? If any? Or the gutter above? (I'm sure many of us would have ripped the whole thing apart day-one.) What is that can to the far right? (It doesn't look like a regulation termination can for UG service, or is this OH service) Is ther any other service on, or in any of the 'other' buildings? Are there any other voltages, or transformers in the building?

Just for refferance is this 3-phase service?

Actually, yes day one I pulled the cover off the gutter and checked the connections there. That is where they connected the GEC to the neutral. The box to the far right is just a pull thru for underground feed. There are no transformers in the building. The one transformer that feeds this building feeds two others as best I can tell. Yes, it is a 3-phase service with no high leg. Tomorrow I'm going to shut off both mains and see if there is any voltage neutral to ground. If there isn't I'm going to take Lostazhell's advice and ring out the neutral and feed wires. If I'm still getting voltage I'm just going to tell the owner to call someone else.

Many thanks

Byron,
Why lumber someone else with this problem?.
You are Johhny on the spot, you should be able to work this out for yourself.
As far as the Owners comments about "this has been going on for about a year" go, I think it is just rubbish, no way would an installation in my opinion sustain a fault like this over a period of time, without some sort of burning occuring.
As has been mentioned in the Chat rooms here, look to get the place shut off during the weekend, there is no such thing as "Can't turn the power off!".
You need to systematically track and and trace all of the neutral conductors, to make sure that there is nothing amiss.
Get an extra guy if you have to.
A lot of my time is taken up with investigating fires of Electrical origin.
If this place goes up, you could be looking at man-slaughter charges through in-action.
Just bear that in mind.
Just my $0.02 worth.

If tomorrow, you have ALL MOCP's (Main CB's) OFF, and you have voltage, then the fault current is from before the mains. If this is the case, it is even more dangerous!! A cause of this may be an arching phase to ground or neutral. Something like that could have created all the other problems. Locating it live will expose you to an arch blast hazard - quite likely - will be the moment of discovery. At that time, if there is voltage with ALL MCOP's OFF, I would get PG&E back out to kill the service before it kills you or anyone else. I wouldn't go looking for it without proper protective gear, for that matter I wouldnt still, unless the service was dead, completely dead. A PG&E short goes off like a grenade - NOT FUN! Trust me....

If you have no voltage, tear that baby apart, and make like you were putting the whole thing back together fresh! (Might as well, what ever it is may require replacemment of some or all of the gear.) After the neutral ground connection is found in or past the sub-panels, check out what ever it is, because it may have been carrying a lot of current, for what sounds like a long time. But the key to the voltage problem is that neutral connection, no point in turning it back on otherwise. Sometimes not too easy to find, but its there.... Sometimes it can be tricky, as often it will even show the correct voltage, but be unable to carry a load at all, or even to varying degree. Then all of the grounding and bonding repairs....

[This message has been edited by e57 (edited 06-27-2006).]

Problems solved.
I shut down all the sub panels. When I shut down a sub panel on the right meter's load I noticed the sizzle between the left meter case and the sub panel to it's left stopped. I verified this via case to case voltage test. I found the one phase that was causing problems by systematically removing them one at a time and checking the case to case voltage. It turns out that the inside sub panel has an AC on the roof. They only fed the AC panel with two #6 wires and used the conduit as a ground. The real problem arose when some handyman electrician saw he could fit another breaker in that panel and used it for security lighting. Thus the neutral for the 120V lighting circuit was tied into the casing. This caused the voltage through the ground. What is interesting is how that voltage got back to the transformer. It must have found a path through the other meter's conduit system that was better than it's own. I wouldn't have suspected a problem that is on one meter to show symptoms on another meter that is tied together through a hot gutter.
As far as the voltage spikes, that was another problem all together. Someone tied the "neutral" to a metal box where they pulled in a multi wire circuit. They took the neutral and wire nutted it to the neutral for the other leg and pushed them back. They thought they would gain a "neutral" for the receptical by connecting to the box.

Many thanks for your help everyone!

Byron

Whoa, might not be out of the woods yet, need to be thourough, thats 3 neutral to ground connections that could explain current on the ground, but NOT the HIGHER, or LOWER voltages.

When you have a 120/208, or for that matter any system other than just plain vanilia 120, and you get a higher voltage to ground than what is input to it, that voltage, that higher voltage has to come from somewhere. In a high/low voltage scenaio, Phase A and Phase B (Or C) can borrow and steal from each other... One goes up, 1-2 two down, et cetera. The "real" neutral point can "float" between loads. Depending solely on which might have more load than the other at the time. A set of voltages read fluctuating like 140, 100, 90 would mean the balance point of the load is not held in check at one point by solid grounding. Was this High/Low problem on one of the circuits you just described and limited to them or it alone? If this High/Low problem was panel wide, or service wide, it sounds like there is still a problem.....