Chasing a ground issue - siding seems the culprit

Status
Not open for further replies.

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
Looking for some advisement here on how to proceed. Doing a job which includes rewiring a kitchen. The rewiring means installing a new subpanel and removing a few old circuits made of cloth nm.

While chasing circuits (old home so various things are connected that would be absolutely stupid by modern logic), found one circuit with no grounding conductor in the panel, but that showed ground at the first recepticle. It was a basement circuit primarily, so I could follow it and disconnect various junctions and outlets. Another circuit had a bootleg ground, so I was initially setting out to just find that... But there were none.

Of course the last point of the entire wiring was the culprit... And that is an outdoor light that is screwed in via aluminum siding. I used a fluke 179 to check continuity everywhere, and with the conductor disconnected at the switch, I still get continuity on the outdoor end of the conductor. So it's disconnected but since the conductor is bonded to the lighting box, which is screwed to the aluminum siding, its a ground path to the panel.

The disconnected ground conductor to the neutral busbar is 0.1 ohm. I then exposed some metal on the siding, and also got 0.1 ohm resistance to the neutral busbar. Further continuity testing proved to me that the nm-b is NOT damaged along it's run.

A detailed look-over did not show any indication of any attempts to specifically ground the siding. It is heavy, thick 1950's wide aluminum.

So what recommendation is best? Gfci/AFCI breaker (there is none currently)? Unground that light and allow it to use the siding while the rest of the circuit uses it's dedicated ground (logic being that we don't want stray currents using the siding as it's best path ever for loads on the circuit)?

Thanks!
 
Last edited:

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
The siding may be touching something like a water pipe that has been deliberately bonded.
There is no reason under the NEC to avoid parallel EGC paths, just parallel neutral paths (especially those involving EGCs.)
Where you have one reliable ground and one unreliable one, do not remove the reliable one!
 
Last edited:

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
Thanks for that clarification. The danger here is that a circuit tester shows that the circuit is "good" though it has no copper wire equipment grounding path to the panel. If it were a matter of sharing some other ground from another circuit that caused the show of continuity, I'd be a bit less worried.

But my plan is to add a proper 12ga grounding conductor back to the panel (the very first section of wire had a ground conductor but it was cut off), a gfci at the first point possible, with this load connected via the gfci. then the siding will remain a 0.1ohm ground path, but the copper connection will also be present.

Thanks!
 

electricalist

Senior Member
Location
dallas tx
I think that if this siding is conducting this good , then it might be a good thing that its grounded.if you could re hang the light so that the siding isnt a better path to ground than the egc, you would be on the right track. This is only my opinion.. its a interestingpoint. Another example could be the work on freezer tops , mounting any electrical to it may make it conductable. But ive never seen a ground specific for its siding per say. How about in a older warehouse were they used a groundrod and cw all the way at the water heater support i beams bolted to concrete. Then u pipe in some lights securing it with anything ul listed but zip ties. Now the structure is is capable of being conductive . Itll be ok. We are taught the importance of bond and ground sizing etc. Just like ur siding alot of things are grounded once or twice before we touch it.....Many are not. The code book is the final answer
 

JHZR2

Member
Location
New Jersey
Occupation
Power Systems Engineer
The meter is about 15ft (not straight/same panel) from the light, and the panel is just inside below the entrance. The main house ground is right there too, but I've not seen it grounded or even in contact.

frankly I'm just surprised that this is what it turned out to be, after all my effort taking apart all the connections and boxes looking for bootleg grounds and g-n shorts.
 
Last edited:
Status
Not open for further replies.
Top