Painting old light fixtures

[Jerramundi]

If the paint were to be 100% insulating, then I wouldn't worry because that would be similar to housing live wires within a plastic/PVC junction box.

My concern is that the paint would act as a semi-conductor and allow just enough current to flow to leave the fixture housing energized, but not enough to trip the breaker.

 

[ActionDave]

With all due respect, NO.. my logic does NOT say that. I'm hard pressed to see how you come to that conclusion using my logic.

Paint is non-conductive, wire insulation is non-conductive. Code requires paint to be scraped away to make a good ground fault path where the screw is engaged. I'm not even sure if code requires insulation to be stripped off the green wire before it's put under the screw but common sense says it does. So once the paint is scraped, the wire is stripped and hooked around the screw and the screw is tightened the installation is code compliant and safe. We are good to go.

Except, you went on to assert, "if it's required regarding "nonconductive coatings such as paint" at the point of making the bonded connection to the EGC, it would stand to reason the same logic applies to the entirety of the ground fault path".

Well if the entirety of the ground fault path includes a wire type EGC then, by your logic, it should not be insulated.

 

[ActionDave]

If the paint were to be 100% insulating, then I wouldn't worry because that would be similar to housing live wires within a plastic/PVC junction box.

My concern is that the paint would act as a semi-conductor and allow just enough current to flow to leave the fixture housing energized, but not enough to trip the breaker.

The same possibility exists for bare metal, paint has got nothing to do with it. Like I said earlier I've made big blue zots with a hot wire on unpainted metal and not tripped the breaker, I've done it with painted metal too. Either there is enough of a fault to energize the metal and the breaker trips or it does not.

 

[AC\DC]

Let's also assume the paint DOES NOT act as a 100% solid insulator. What if there is enough resistance on the paint layer to allow current to flow to the EGC, but not enough current to trip the breaker? You would end up with an energized piece of metal (i.e. the fixture body). I admit this is a worst case scenario, but I'm just trying to be thorough.

If you Are this worried just GFCI the circuit. Would solve this situation you proposed.

 

[Jerramundi]

Paint is non-conductive, wire insulation is non-conductive. Code requires paint to be scraped away to make a good ground fault path where the screw is engaged. I'm not even sure if code requires insulation to be stripped off the green wire before it's put under the screw but common sense says it does. So once the paint is scraped, the wire is stripped and hooked around the screw and the screw is tightened the installation is code compliant and safe. We are good to go.

Except, you went on to assert, "if it's required regarding "nonconductive coatings such as paint" at the point of making the bonded connection to the EGC, it would stand to reason the same logic applies to the entirety of the ground fault path".

Well if the entirety of the ground fault path includes a wire type EGC then, by your logic, it should not be insulated.

Alright, fair enough. I suppose my choice of wording, "the ground fault path," could be construed this way. Poor choice of words.

What I was attempting to convey is that if you're required to scrap the paint away to make a solidly bonded connection between the enclosure and the EGC, that the NEC is saying the paint is problematic and negatively affecting continuity.

Given that, I would argue that the same logic should apply to any point at which a ground fault could potentially occur. In other words, you wouldn't want the paint to inhibit the continuity between the point at which the fault occurs and the point at which the EGC is bonded to the metal enclosure.

At a minimum, I would argue this applies to the portion that houses the wires, but perhaps not the exterior.

But I acknowledge that, to my knowledge, there is nothing in the NEC explicitly stating this except as to how it applies to the point at which your establishing a bonded connection.

 

[LarryFine]

If the paint prevents the contact, it also prevents the energization. If and when the conductive contact does occur, a proper EGC path will prevent the energization (other than momentary EGC voltage drop).

As I said, or tried to say, earlier, if the current is too low to trip the OCPD, then most of the voltage will be across the arc, and the housing will barely have any voltage on it, as any neutral with a load would.

 

[Jerramundi]

If the paint prevents the contact, it also prevents the energization. If and when the conductive contact does occur, a proper EGC path will prevent the energization (other than momentary EGC voltage drop).

As I said, or tried to say, earlier, if the current is too low to trip the OCPD, then most of the voltage will be across the arc, and the housing will barely have any voltage on it, as any neutral with a load would.

I get what you're saying... but I would hesitate to say that a "proper EGC path will prevent energization." A proper EGC path should encourage energization in order to subsequently prevent it. An EGC path needs to be low impedance. I know you know this. I'm just saying I would hesitate to use the phrasing that you're using for the aforementioned reasoning.

It comes down to dealing with absolutist rhetoric and the differences between insulators, semi-conductors, and conductors... and whether the paint is acting as an insulator, semi-conductor, or a conductor. Given the above code reference, it seems the NEC considers paint to be non-conductive or at the very least, that it inhibits continuity to a degree.

I am of the opinion that unless paint could meet the qualifications of acting as an effective insulator, that it should be conductive in portions of the fixture that house live electrical components. If not the paint itself, then some additional layer on top of it should be. The paint acting as a semi-conductor could create a problem.

I think I'm going to follow your suggestion of just painting the outside and leaving the portion that houses the electrical as bare metal.

 

[Jerramundi]

I would still be curious to see if UL or some other NRTL has standards that touch on this topic. I would imagine how fixture manufacturer's coat their products is governed by someone, somewhere.

Given how frequently these standards often go really, really in-depth, I would be shocked if they didn't have something to say about this.
Initial Google searches reveal UL 1570 governs fluorescent light fixtures, but I haven't dived into it as of yet.

 

[LarryFine]

It comes down to dealing with absolutist rhetoric and the differences between insulators, semi-conductors, and conductors... and whether the paint is acting as an insulator, semi-conductor, or a conductor.

I say it's acting as a colorant and corrosion inhibitor, that happens to have insulative properties.

 

[Jerramundi]

I say it's acting as a colorant and corrosion inhibitor, that happens to have insulative properties.

Alright Merriam, haha. Would something with insulative properties to the degree that said properties are skewed more greatly towards insulator than semi-conductor, even if not the primary function, not still be defined as an insulator?

 

[LarryFine]

Well, we agree on something: The OEM paint is, if it must be defined, an insulator.

That's why I say adding more paint won't change anything important.