Inverter working clearance

[jaggedben]

... But my ignorance of different products comes into play here, for something like a SolarEdge 7.6kW inverter, would you typically do that or need to do that? Or would you just hook everything up, power it up, and check what the inverter tells you? Then if there's a problem, depower, adjustment connections, and repeat.

..

I agree with petersonra. Plus, if it's a Solaredge inverter, it's particularly likely to require servicing.

 

[electro7]

Thanks Jaggedben. Somehow I got onto thinking the inverters were to high with some of what was being posted.

It is very hard for me to believe that 3" of the ledge protuding out past the front of the inverter which is approx 20" below the inverter is going to impede servicing- and makes it not readily accessible. Im running with it and hoping to not catch that 1 out of 20 inspector, lol

Sent from my SM-N960U using Tapatalk

 

[ggunn]

Thanks Jaggedben. Somehow I got onto thinking the inverters were to high with some of what was being posted.

It is very hard for me to believe that 3" of the ledge protuding out past the front of the inverter which is approx 20" below the inverter is going to impede servicing- and makes it not readily accessible. Im running with it and hoping to not catch that 1 out of 20 inspector, lol

Sent from my SM-N960U using Tapatalk

I think that as far as the ledge is concerned you are covered by 110.26(A)(3).

 

[tortuga]

I was not referring to how often it would need to be accessed while live. If the chance you need 1 or more live accesses over the life of the product is, say, 90% or more, that's obviously "likely". But suppose the chance is 1 per 100 product-years, with a 20 year life span for the product. Is that likely? Or suppose it's 1 per 1000 product-years, is that likely? "Likely" is undefined.


I agree with you that if checking string voltages live while commissioning is SOP, then that equipment obviously needs to meet the 110.26(A) clearances. But my ignorance of different products comes into play here, for something like a SolarEdge 7.6kW inverter, would you typically do that or need to do that? Or would you just hook everything up, power it up, and check what the inverter tells you? Then if there's a problem, depower, adjustment connections, and repeat.

The latter, with all AC disconnects, DC disconnects, and OCPD located elsewhere, was what I had in mind for an inverter than might not require 110.26(A) clearances.

Cheers, Wayne

I was wondering about this back in 2019, and I basically argued that an inverter should be treated like a transformer
provided

• It contains no overcurrent devices
• you have a DC disconnect and AC disconnect in 110.26 locations

the inverter needs not comply with 110.26.

Why they put a DC disconnect in an inverter is still beyond me, but most do, so 110.26 usually applies, unless you add an extra DC disconnect in a 110.26 compliant area.

 

[jaggedben]

...

Why they put a DC disconnect in an inverter is still beyond me, but most do, so 110.26 usually applies, unless you add an extra DC disconnect in a 110.26 compliant area.

Well, when it started being done about a decade ago it saved a ton of money because off-the-shelf DC rated knife switches were comparatively expensive. Once people started selling the rotary switches as separate components then maybe it was overkill. But for resi systems it generally saved a lot of labor and supply management because only 10% or so of jobs have a problem with 110.26. Depending on where you worked.

Also you needed discos grouped with the inverter so most of the time if you didn't have a 110.26 compliant location then you were screwed anyway.

 

[tortuga]

I just dont get the DC disconnect in the inverter for lots of reasons;
First the NEC 690 has so many defined circuits to dance around it:

• 'PV source circuits'
• and the 'PV Output circuit'
• and the all encompassing 'PV system DC circuit'

I think of these as DC taps or 'service conductors' as they can get their protection at the 'load' end (inverter) or no over-current protection at all [690.9(1)]

I have been on more than a dozen PV inverter swap out service calls where its impossible for one man to swap, I'll admit I can't lift what I used to in my 20's, but those inverters are tanks, they could loose a few pounds and shed the DC disconnect.

Then on said service call the DC disconnect is built into the 'high voltage' device (inverter) you need to safety off to service.
While its unlikely the rotary switch has melted and cracked and will arc when I take the cover off, I have no idea till i do, I am there because the inverter "quit working and smells funny in the garage".
I cant lockout the thing I am removing /opening.

300V - 1000V DC per lockout tag-out policy involves going on a roof and unbolting PV panel(s) and unplugging wires...


For both those reasons I think it would make sense to have the DC disconnect (rotary switch if you will)
at the point of transition from the 'PV source circuits' and the 'PV Output circuit', then you could simplify the code and end article 690 at that DC disconnect, delete 'PV system DC circuit' and 'PV Output circuit' and you just now have a article 215 feeder from the DC disconnect/AFCI/GFCI/rapid shutdown device to the article 705 inverter or Article 408 DC panel:

PV Panel(s) ------- source circuit(s) ----->/----> DC disconnect / rapid shutdown / G/AFCI ----->'PV output circuit' (becomes DC feeder).

Once the DC disconnect is part of the rapid shutdown box (ahead off the shutdown relays) and 'accessible' per NEC at the edge of roof installs so, if you need to replace an inverter (or rapid shutdown box) you just need a ladder pop up to the edge of the install and flip a rotary switch.
If a siding guy damages the DC feeder conduit, pop up to the roof flip a switch.
Just my two cents

 

[greens]

The SMA inverters we use are not approved for vertical stacking, The can be stacked with offsets. The installation manual covers this and we were told it is a convection cooling issue.

Also the rotary disconnects used, at least on SMA inverters, are not designed for repeated disconnect. Guidance is to shut down the inverter by disconnecting the AC side, and then operate the DC disconnect. The are not rated for routine disconnections. One engineer told us the switch life, if routinely used can be counted on fingers. They are effective for keeping the DC feed open.