Arcing between subpanel and concrete wall

[Gene B]

There's a subpanel mounted directly to a basement concrete wall which shows signs of arcing between the box and the wall. What are the possible causes?

Regardless of the cause, is it a bad idea to mount a metal box directly on a concrete wall because of the possibility of an unintentional earth connection?

 

[ptonsparky]

Lightning could be a cause of that evidence

i would hazard a guess that there are millions of panels mounted to concrete walls.

Add: We want earth connection.

 

[big john]

Yeah, I can't imagine a circumstance where a low-voltage system would show visible arcing damage to concrete.

You sure this isn't corrosion related?

Mounting directly to concrete is normal and only regulated in wet locations.

 

[don_resqcapt19]

...
Regardless of the cause, is it a bad idea to mount a metal box directly on a concrete wall because of the possibility of an unintentional earth connection?

The only issues with unintentional earth connections are when they involve the neutral or an ungrounded conductor.

There are no issues with multiple points of connection for the grounding conductor(s) or conductive object connected to grounding conductors.

 

[K8MHZ]

I would go with lightning as well, or perhaps a surge from a primary coming in contact with a secondary and energizing the panel with thousands of volts. I don't think it's possible to make an arc on concrete with 120 volts.

 

[Gene B]

Yeah, I can't imagine a circumstance where a low-voltage system would show visible arcing damage to concrete.

You sure this isn't corrosion related?

I'm sure. If it's lightning, wouldn't it be better to have only one earth connection? That way the lightning wouldn't be using the equipment ground as a path.

 

[K8MHZ]

I'm sure. If it's lightning, wouldn't it be better to have only one earth connection? That way the lightning wouldn't be using the equipment ground as a path.

Lighting will take any and all paths, including paths of air, to get to it's charge center in the earth. Providing more low impedance (crucial) metal paths to earth will reduce the likelihood of a strike taking an air path and making big sparks.

If the single earth connection isn't sufficient to shunt the bolt of lightning, the result is the bolt using other paths to get back to the charge center.

 

[mgookin]

There's a subpanel mounted directly to a basement concrete wall which shows signs of arcing between the box and the wall. What are the possible causes?

Regardless of the cause, is it a bad idea to mount a metal box directly on a concrete wall because of the possibility of an unintentional earth connection?

Can we get some pictures?
We like pictures

Panels are mounted on concrete walls all the time.

 

[big john]

I'm sure. If it's lightning, wouldn't it be better to have only one earth connection? That way the lightning wouldn't be using the equipment ground as a path.

Can you get photos?

When it comes to lightning not much is an effective insulator, so generally it's far better to ground and bond the heck out of stuff to reduce voltage gradients, than it is to try and stop lightning from causing current flow through isolation.

 

[Gene B]

Here's a photo:

 

[big john]

You sure that didn't exist prior to the installation of that panel?

The exterior of the panel looks pretty unaffected. What does the interior look like?

 

[kwired]

Here's a photo:

What does inside of box look like? Or even a picture that shows side of the box better?

Some hot conductor possibly dangling there and touched the box at some point, leaving the mark on the wall?

 

[Dennis Alwon]

IMO, that was existing unless you have some marks on the inside of the panel

 

[brian john]

I'm sure. If it's lightning, wouldn't it be better to have only one earth connection? That way the lightning wouldn't be using the equipment ground as a path.

Lightning can and does anything it wants to, even the best of elaborate grounding systems with TVSS (SPD's) and Lighting protection can sustain damage if the lighting hits in the right location and is of long enough duration and magnitude.

Any chance there is re-enforcing steel exposed (or was exposed) at this location.

 

[Travisat419]

I'm near Philly, we always mount a back box to a minimum of plywood to keep from corrosion of the metal, I prefer 7/8 metal strut and 1/4 20 spring nuts and the sort if suitable. I deal with wall walls that are 100 years old sometimes it is a mission to get a 2' x 4' cut of plywood to firmly attach. In your pic it looks like a hot conductor came in contact with the top of your panel making a flash. wipe it of, check the inside underneath, clean?

 

[Speshulk]

If the panel is mounted directly to the concrete, how would there be an arc?

 

[Gene B]

If the panel is mounted directly to the concrete, how would there be an arc?

Suppose that a panel screw just happens to be close to rebar. The path would be rebar->arc->screw->case->EGC.

 

[Speshulk]

Suppose that a panel screw just happens to be close to rebar. The path would be rebar->arc->screw->case->EGC.

We're looking at a picture of an "arc" between a cabinet and a concrete wall that it's mounted to. If two conductive surfaces are touching each other, the path of least resistance between them is going to be through the points of contact, thus, no arc.

 

[Strathead]

On top of everything else, that panel looks new enough that it must have the code required raised mounting feet that means that spot is 1/4" from the actual concrete. Doesn't that take like 3,000 volts minimum to jump?

 

[iwire]

Regardless of the cause, is it a bad idea to mount a metal box directly on a concrete wall because of the possibility of an unintentional earth connection?

That is not an issue at all.