safety with high voltages in ground-mount "string" systems

[RShackleford]

Systems that utilize string inverters tend to have lethal voltages (as high as 600vdc) in the wiring among the panels in a "string" - as multiple panels (or multiple optimizers connected to panels) are wired in series before being connected to the string inverter.

I'm curious what safety precautions are required - both by good practice and by NEC (specifically 2017 for my area) - particularly for ground-mount systems. I don't know exactly what the wiring methods used are, but from what I've read I'm pretty sure they aren't conduit (except for the final dive into the ground for the run to whatever building contains the inverter, etc), so it seems like anyone could come up and expose themselves to hundreds of DC volts without too much trouble.

My inspector stated that he does not require a fence around such installations. Maybe he's just flat wrong, and I'd be negligent by not fencing it. Or, do the wiring methods provide sufficient safety ?

Alternatively, suppose I wanted to eliminate the problem by elevating the panels farther above ground. How high is high enough ? Obviously a roof is, since no special precautions such as fencing are required for those installations. But is 6ft (at the low south side) enough ? If the wiring comes out of the panel at the top edge, then those lowest wires would be 9ft or so above ground. Is the '-' end of the string referenced to ground in any way ? If so, and the strings are wired from bottom to top, then lethal voltages would not be present until the second or third "row" of panels (above the lowest row) so even if the lower edge is 3ft off grade, the lethal voltages might be 9ft or more higher.

What do people do ?

[Mike90250]

I think generally, code requires a fence around ground mount panels. Panels that are reachable from the ground, 8' high for energized portions should be OK without a fence. But since he's going to approve lower standards, you can always over do it.

Voltages over 70V are generally going to be lethal, which is why all systems with exposed battery terminals are limited to 50V nominal. 48V is OK, because the nominal voltage is below 50, even though some charging happens at 68V

[RShackleford]

I think generally, code requires a fence around ground mount panels. Panels that are reachable from the ground, 8' high for energized portions should be OK without a fence ...

So how 'bout this: I'll have a roughly 16ft-square 3x5 array of panels, at latitude tilt, so if the low edge is a couple of feet off grade the high edge will be at least 12ft. Put all the optimizers at the top (easy access since tit'll be open underneath) and just run wires from the panels up to the optimizers, and come down from the series-connected chain of optimizers with conduit for the buried run to the house and the inverter. So there's no voltages above a single panel's 40vdc or so accessible anywhere below the top edge of the panel.

Lotta extra wire, but probably less (or no more) expense than a fence. I don't want a fence. I'll just raise the whole ting before I do that.

Seems reasonable to me, and to any inspector with an open mind. Of course, mine seems to say no worries anyhow, but I think I'd worry.

[nwdiver]

Systems that utilize string inverters tend to have lethal voltages (as high as 600vdc) in the wiring among the panels in a "string" - as multiple panels (or multiple optimizers connected to panels) are wired in series before being connected to the string inverter.

I'm curious what safety precautions are required - both by good practice and by NEC (specifically 2017 for my area) - particularly for ground-mount systems. I don't know exactly what the wiring methods used are, but from what I've read I'm pretty sure they aren't conduit (except for the final dive into the ground for the run to whatever building contains the inverter, etc), so it seems like anyone could come up and expose themselves to hundreds of DC volts without too much trouble.

My inspector stated that he does not require a fence around such installations. Maybe he's just flat wrong, and I'd be negligent by not fencing it. Or, do the wiring methods provide sufficient safety ?

Alternatively, suppose I wanted to eliminate the problem by elevating the panels farther above ground. How high is high enough ? Obviously a roof is, since no special precautions such as fencing are required for those installations. But is 6ft (at the low south side) enough ? If the wiring comes out of the panel at the top edge, then those lowest wires would be 9ft or so above ground. Is the '-' end of the string referenced to ground in any way ? If so, and the strings are wired from bottom to top, then lethal voltages would not be present until the second or third "row" of panels (above the lowest row) so even if the lower edge is 3ft off grade, the lethal voltages might be 9ft or more higher.

What do people do ?

I've done two state inspected ground mounts. Both were approved without a fence. I've seen plenty of residential ground mounts that were unfenced using ~500vdc string systems. In any real world scenario the plug that's powering your laptop is more dangerous than a ground mounted solar array. You can shock yourself with a paper clip. You'd be hard pressed to shock yourself on a properly installed ground mount....

[peakbagger]

My pole mount does not have a fence. I know of a similar install with a fence. I believe the code at the time of both systems was that the wiring needed to be "inaccessible" unless "tools" are used within so many feet of the ground. The local AHJs do not have a lot of experience with solar and most are roof mounts and tend to go with whatever a licensed installer puts up. My pole mount is 2 rows high of panels. I installed the top row of panels "upside down" with the junction boxes on the bottom instead of the top. I have some horizontal support struts running close behind the panels so I drilled and tapped the struts to hold some galvanized mesh on top of the strut. The interconnecting cabling and the junction boxes sit behind the mesh. It keeps things neat and would require tools to remove. I like it better than a fence as a fence requires more maintenance with lawn care equipment which introduces IMHO a high risk to the panels.The other install tried to keep the fence clear for several years but he finally gave up dealing with the vines that like to grow up in the fencing and now its green wall. The biggest safety concern with my pole mount is two sharp corners on the lower edge. Unless there is special risk like active kids, IMHO no code need for fence if the cabling is protected.

[RShackleford]

I like it better than a fence as a fence ...

I absolutely will not do a fence. Pretty sure my AHJ is gonna be ok with that. If not, there is my idea of putting all the optimizers up high, so there's only one panel's worth of voltage near ground level. But I really like your idea of some kind of mesh (if I understand you correctly) that's underneath the panels and prevents access to the wiring. I intend to use the area under the panels for storage of stuff like trash cans and bicycles; it'd really suck if some moron hung something from a cable carrying 500vdc !

[bcroe]

So far I have not found any fried deer trying to eat my wiring, though they can reach it. If it
were a problem I would use some heavy screening over wiring channels. The danger underneath
panels, is trying to positions some stuff, and accidentally ramming a metal pipe through a live panel.
Bruce Roe

[RShackleford]

The danger underneath
panels, is trying to positions some stuff, and accidentally ramming a metal pipe through a live panel.

Maybe I'll use a finer mesh like hardware cloth (with 1/2"-square openings).

[peakbagger]

I used 1/2" mesh galvanized hardware cloth, held on with SS hardware. Note I was attaching to galvanized unistrut. I would be careful about using galvanized on aluminum mounts due to galvanic corrosion.

[RShackleford]

I used 1/2" mesh galvanized hardware cloth, held on with SS hardware. Note I was attaching to galvanized unistrut. I would be careful about using galvanized on aluminum mounts due to galvanic corrosion.

Actually, and this was something I wanted to ask about (maybe a separate thread), I was thinking about building the entire structure from pressure-treated wood. Four 6x6 or 8x8 posts on concrete piers, and a grid of beams & joists that makes sense. Doesn't seem like anyone ever does this, and I wonder why. This might be cheaper, and would leave the entire area under the panels free for storage of bicycles, trash cans, and the like.

[SunEagle]

Actually, and this was something I wanted to ask about (maybe a separate thread), I was thinking about building the entire structure from pressure-treated wood. Four 6x6 or 8x8 posts on concrete piers, and a grid of beams & joists that makes sense. Doesn't seem like anyone ever does this, and I wonder why. This might be cheaper, and would leave the entire area under the panels free for storage of bicycles, trash cans, and the like.

As long as the structure meets your local wind and building codes having a solar roof over a storage area should work.

[peakbagger]

I have seen several solar pergola projects in the media that look good and are constructed out of standard pressure treaed framing. The one big caveat is its not weather tight as there are gaps between the panels. I have seen photos of at least one install where the owner installed polycarbonate panels under the solar panels to redirect water away from the space under the panels. Polycarbonate is clear is eventually turned milky by UV. I expect since its mostly covered by the PV panels that it will take longer to turn. Most attempts at finding a resilient method of filling the gap directly between the panels do not appear to end well.

As Sun Eagle stated its got to meet local wind and building codes. That can get interesting in areas with high wind.

[RShackleford]

I have seen several solar pergola projects in the media that look good and are constructed out of standard pressure treaed framing. The one big caveat is its not weather tight as there are gaps between the panels. I have seen photos of at least one install where the owner installed polycarbonate panels under the solar panels to redirect water away from the space under the panels. Polycarbonate is clear is eventually turned milky by UV. I expect since its mostly covered by the PV panels that it will take longer to turn. Most attempts at finding a resilient method of filling the gap directly between the panels do not appear to end well.

I guess something like polycarbonate panels would kinda eliminate the advantage of easy access to the wiring, optimizers, etc - but would at least alleviate the high-voltage safety concerns.

I can't understand why good silicon caulk between the panels wouldn't work well. But I've never actually seen a panel in person yet, so I don't know what the mating surfaces would look like.

[J.P.M.]

I guess something like polycarbonate panels would kinda eliminate the advantage of easy access to the wiring, optimizers, etc - but would at least alleviate the high-voltage safety concerns.

I can't understand why good silicon caulk between the panels wouldn't work well. But I've never actually seen a panel in person yet, so I don't know what the mating surfaces would look like.

Once you see a common array, you'll understand. Primarily, it has something to do with the 2-4 cm. gap between panels. Even if butted with as little gap as possible, caulking tends to quickly deteriorate. It's difficult to maintain an effective way to keep water from penetrating/leaking through on mostly horizontal surfaces without layering.

Or, try it, but I'd suggest you consider not putting anything underneath that can be damaged by H2O.

[RShackleford]

It's difficult to maintain an effective way to keep water from penetrating/leaking through on mostly horizontal surfaces without layering.

Well, I'll probably tilt at my latitude, 36 degrees, so not that horizontal.

Or, try it, but I'd suggest you consider not putting anything underneath that can be damaged by H2O.

Trash cans, and bicycles. Miscellaneous crap that is sitting around as an eye-sore now. The installation will sit a little north of the house, so main view will be of the south side, so that stuff will be hidden; can also do cheap lattice stuff around the sides.