Getting shocked while washing GROUNDED panels on metal roof

[Norwasian]

The panels were professionally installed, and grounded via their own dedicated ground rod placed at the foot of one of the building's concrete-anchored metal posts with a ground wire run in conduit from the panels. The panels' frame is aluminum, as are the mounting brackets, and the roof is quite possibly unpainted aluminum (silver colored). While washing the panels in the daytime, I rested one hand on the metal roofing using the other hand to wash. When washing and contacting the frame with a wet cloth, I felt a shock from the roof.

I never felt a shock from the panels I installed in the mountains with an ungrounded, off-grid system (no ground at all) which were also installed on a metal roof, albeit that roof's metal was painted. I am now wondering whether there is a problem with the grounding in this on-grid system which was grounded per the common standards.

Should I just cut the ground wire? Why should I be getting shocked?

 

[ricardocello]

Researching electricity in Thailand, I found this informative page:

Electrical Grounding or Earthing

Electrical Grounding or Earthing

www.thailandguru.com

The voltage you feel is likely coming from AC capacitive coupling with your solar charge controller or All-in-one over the DC wires.
Connecting the frames to the earth through a rod will only bleed off static electricity, not prevent shocks.

The ground wire from the panel frames should also be bonded to the electrical ground in the house (if there is one).
Then you won’t get shocked. And you should have a neutral-ground bond somewhere.

 

[Bkord]

Could your roof be live?

 

[AntronX]

Are you looking for help or to pick an argument again against all grounding? Your metal roof/building and panels should be tied to same grounding point to prevent different potentials from developing between them.

 

[Norwasian]

Are you looking for help or to pick an argument again against all grounding? Your metal roof/building and panels should be tied to same grounding point to prevent different potentials from developing between them.

I do not see how this would help. The metal frames of the solar panels are most certainly already contacting the metal roof. That should be already preventing a different potential, which is why it is rather shocking that one still gets shocked.

 

[BarracudaBob]

getting "shocked' is a difference in potential. If your water, solar, utility, and oven are all grounded the same to the main utility and ground rod at your service entrance you wouldn't have a difference in potential.

Put a volt meter on there and find out what is not grounded/bonded and fix it.

 

[Norwasian]

getting "shocked' is a difference in potential.

Most certainly, and I agree.

I noticed that I was shocked by some panels but did not notice the shock with others. As I was working while observing this, I wondered if perhaps I was not shocked by the other panels because the cloth had largely dried and was no longer so wet. Just the same, despite the aluminum rail system holding the aluminum frames of the panels being attached directly to the metal roofing, there is evidently poor contact somewhere, and/or some other cause for a difference in potential.

Hmm....maybe @Bkord is on to something. I wonder if one of the light fixtures on the ceiling just below this roof (only inches apart) was not wired/insulated properly.

I do wonder whether the current was AC or DC. Unfortunately, I don't have an oscilloscope, so I'm not sure how I could know.

 

[hwy17]

I bet the shock is coming from the inverter and not the PV. Did it feel like bzzzz?

 

[BarracudaBob]

there is evidently poor contact somewhere, and/or some other cause for a difference in potential.

Use a volt meter either AC or DC and just start checking. No other way to do it. Identify and repair. O-Scope sounds like a nightmare and over complicated.

 

[Norwasian]

I bet the shock is coming from the inverter and not the PV. Did it feel like bzzzz?

I'm not sure if I have enough experience with DC shock to distinguish it from an AC current, but it definitely felt like a buzz or a vibration. Does DC not feel this way?

 

[Fisherguy66]

I dunno I once stuck my tongue across a 12v batt. Gave me quite a buzz!

 

[Brucey]

Don't some of the AIOs expose ac voltage on the pv lines or some other crazy stuff?

 

[Fisherguy66]

Don't some of the AIOs expose ac voltage on the pv lines or some other crazy stuff?

I wouldn't doubt if my eco-worthy might.

 

[Hedges]

Check PV frames to roof, optionally to ground pin of an extension cord. If they are properly grounded there should be no AC or DC voltage present.

Many of these AIO drive common-mode AC on PV+/-. If PV frames aren't grounded, they cause shocks due to capacitive coupling to PV circuit.
Yours are grounded.

But now you've put another electrode (wet cloth) over the glass, forming parallel plate capacitor to PV cells. AC voltage on cells couples through capacitance to wet cloth, and your body provides a path back. I'd think fairly small capacitance and small current.

Test AC voltage from wet cloth to grounded PV frames and to roof.

 

[Norwasian]

But now you've put another electrode (wet cloth) over the glass, forming parallel plate capacitor to PV cells. AC voltage on cells couples through capacitance to wet cloth, and your body provides a path back. I'd think fairly small capacitance and small current.

I was only shocked when the cloth reached the edge of the panel and contacted the frame. As long as it was only on the glass, I noticed no problem.

I'll have to get up there on the roof and check with a voltmeter, as folk are suggesting, when I have a chance. I'll report back when I've been able to do so.

 

[hwy17]

I'm not sure if I have enough experience with DC shock to distinguish it from an AC current, but it definitely felt like a buzz or a vibration. Does DC not feel this way?

Not in my experience. DC burns or stings but does not feel like the buzzing/vibrating of AC.

 

[Norwasian]

Not in my experience. DC burns or stings but does not feel like the buzzing/vibrating of AC.

@BarracudaBob: Well, I made it up there finally with a voltmeter. I measured 32.4 VAC between the panel frame and the sheet metal.

@Brucey, Is this a reasonable voltage to expect in the case of an inverter backfeeding AC to the PV's DC lines?

I suppose it isn't a dangerous voltage, but it's uncomfortable to sustain with wet hands.

Another factor I just thought of, @Hedges:

We are in the dry season right now, and haven't had much rain for quite a while. Could the ground rod, driven into red clay, be inadequate?

 

[Bongbong]

I assume you were standing on the roof. The roof, panel frames, racking and any other metal parts of the structure should be bonded together, it doesn't matter if that is continued to a ground rod or not as their should not be a potential difference between any of the metal parts to give you a shock. As for continuing the ground wire to the main ground rod, just follow local regulations, some places will require a separate ground rod, others will require a main rod and others still will require both.

You should investigate where the 32.4 VAC comes from though.

 

[crossy]

Well, I made it up there finally with a voltmeter. I measured 32.4 VAC between the panel frame and the sheet metal.

You "should" measure zero there if everything is correctly bonded.

Somewhere along the line there's a discontinuity. I'll bet your building steel isn't bonded to your electrical earth. It's not actually required by the Thai regs, but they were written long before solar was a thing.

As it happens our roof steel measures as a better ground than our rod! They are now connected together.

32V AC is definitely enough to feel an unpleasant buzz with wet hands/feet.

 

[Norwasian]

Perhaps an odd question, but I've always wondered what happens to that "potential" when one bonds two points of different potentials like this. Does the voltage passing through the bond have the potential (no pun intended) to damage something somewhere in the equipment connected to it?

For example, suppose that this roof voltage is indeed coming from the roof, and not from the solar panels, nor the inverter. Will it now have the potential to damage something in the inverter if given a solid connection between the roof and the PV frames?

I mean, it's a rule of physics, right?, that every action has an equal and opposite reaction. What is that reaction doing?

 

[Brucey]

@Brucey, Is this a reasonable voltage to expect in the case of an inverter backfeeding AC to the PV's DC lines?

I suppose it isn't a dangerous voltage, but it's uncomfortable to sustain with wet hands.

Yeah sorry can't hep you there my inverters and sccs are seperate so such things are physically impossible

 

[Norwasian]

I assume you were standing on the roof. The roof, panel frames, racking and any other metal parts of the structure should be bonded together, it doesn't matter if that is continued to a ground rod or not as their should not be a potential difference between any of the metal parts to give you a shock. As for continuing the ground wire to the main ground rod, just follow local regulations, some places will require a separate ground rod, others will require a main rod and others still will require both.

You should investigate where the 32.4 VAC comes from though.

If there are local regulations, I may not be able to easily find or read them. I read Thai slowly: it has no word spaces between words and is very difficult to scan for specific bits of information. That said, the installers were professional, and I have a modicum of assurance that they would have followed the codes. They put all of their wiring into conduit--including metal conduit for the DC side, and left things neat and tidy. The man in charge seemed to have an up-to-date understanding of the Thai regulations for solar installations as well. I've been quite satisfied with the installation so far, five months down the road. We recently got the CT for it moved over to the single-phase meter so that the whole building now benefits from the solar production. I guess this stray current is the first issue I've encountered, and I do not yet know if the solar setup has anything to do with it beyond giving me an opportunity to discover the issue.

 

[Norwasian]

Yeah sorry can't hep you there my inverters and sccs are seperate so such things are physically impossible

I've understood that it is the inverter that backfeeds the AC voltage by having a common ground with the DC side. If this is the case, I don't think having a separate charge controller would change anything.

 

[Supervstech]

Perhaps an odd question, but I've always wondered what happens to that "potential" when one bonds two points of different potentials like this. Does the voltage passing through the bond have the potential (no pun intended) to damage something somewhere in the equipment connected to it?

For example, suppose that this roof voltage is indeed coming from the roof, and not from the solar panels, nor the inverter. Will it now have the potential to damage something in the inverter if given a solid connection between the roof and the PV frames?

I mean, it's a rule of physics, right?, that every action has an equal and opposite reaction. What is that reaction doing?

Bonded surfaces negate flow of static and provide a fault path for fault current.

If you dont have a fully bonded path, there will be imbalances somewhere.
PV panels in contact with the roofing panels do not make a good electrical connection, unless they are properly connected due to surface oxidation insulating properties.

If all is properly bonded to the structure bonding system, stray currents should be dissipated.

 

[Brucey]

I've understood that it is the inverter that backfeeds the AC voltage by having a common ground with the DC side. If this is the case, I don't think having a separate charge controller would change anything.

The aios internally have the pv output and ac going to a shared bus running at a higher voltage, this is presumably how ac can end up on the pv lines.

With separates, its simply not possible for ac from inverter to reach the pv panels via the scc. There's only a DC connection to the sccs and then the panels. AC is only available at the inverter ac output terminals.