Off grid house bonding sanity check

wwhitney said:
Hmm. If you have no service, and just one electrical source, it is by definition an SDS.

If you have no service, and two electrical sources, then you have 0 or 2 SDSs. 0 SDSs would be the case where the two sources have a solidly connected circuit conductor, e.g. the grounded conductor. While 2 SDSs would be the case where there is no non-grounding/bonding connection between the two sources. ...
Click to expand...

Lol there you have it. 250.4 thru 250.20 apply, but 250.24 and 250.30 do not. Notably, 250.142(B) also does not apply because of how it’s worded. So, codewise you can ground this system anyway and anywhere you want that meets the very generally stated intents in 250.4.

I would still go with the main panel myself. Keep it simple.

I'm pretty sure the Sol-Ark inverter syncs with the generator when it turns it on. I believe it will not behave like a grid-tied inverter in this configuration, but is capable of acting as a parallel voltage source. As Birken will tell you this is not a particularly new. Actually I'm just assuming it has similar capabilities to the old Trace or Sunny Island.
 
jaggedben said:
I would still go with the main panel myself. Keep it simple.
Click to expand...
Agreed
jaggedben said:
I'm pretty sure the Sol-Ark inverter syncs with the generator when it turns it on. I believe it will not behave like a grid-tied inverter in this configuration, but is capable of acting as a parallel voltage source.
Click to expand...
The way 250.30 reads to me is this parallel voltage source would be one SDS.
 
pipe_bender said:
Agreed

The way 250.30 reads to me is this parallel voltage source would be one SDS.
Click to expand...

What Wayne has brought our attention to is that the definition of an SDS isn't written that way.

Actually it really ought to be fixed because it could be read as saying that 250.30 doesn't apply to a system supplied by both a transformer and a grid-tie inverter, but it's industry standard to treat the transformer as an SDS and I'm strongly of the opinion it should stay that way.
 
jaggedben said:
What Wayne has brought our attention to is that the definition of an SDS isn't written that way.

Actually it really ought to be fixed because it could be read as saying that 250.30 doesn't apply to a system supplied by both a transformer and a grid-tie inverter, but it's industry standard to treat the transformer as an SDS and I'm strongly of the opinion it should stay that way.
Click to expand...
Sorry I am confused 250.30 it says "Multiple separately derived systems that are connected in parallel shall be installed in accordance with 250.30". So just for my education what makes them in parallel or not in parallel ? I would think a inverter syncing with a SDS makes it all one SDS?
 
jaggedben said:
Actually it really ought to be fixed because it could be read as saying that 250.30 doesn't apply to a system supplied by both a transformer and a grid-tie inverter
Click to expand...
Well, the second sentence of (2017) NEC 250.30 may reasonably be taken to mean that if you have two or more sources connected in parallel, and they would individually be SDSs if they weren't in parallel, 250.30 still applies.

Cheers, Wayne
 
wwhitney said:
Well, the second sentence of (2017) NEC 250.30 may reasonably be taken to mean that if you have two or more sources connected in parallel, and they would individually be SDSs if they weren't in parallel, 250.30 still applies.
Click to expand...
I see. But subsequently they went ahead and muddled that with 'of the same type.' 🤷‍♂️
 
wwhitney said:
So for an off-grid system of the sol-ark plus a generator, how does the sol-ark control the generator? Is it able to turn the generator on at low battery SOC (and/or possibly at AC demand in excess of the inverter rating), and off at high battery SOC, without interrupting power to the loads?
Click to expand...

Exactly. 2 wire start output to the generator.

wwhitney said:
Seems like this would require a switch (in the ungrounded conductors) that can detect the generator AC timing while it is still disconnected from the loads, (relatively) slowly change the inverter AC timing to sync with the generator, and then cut in generator output while simultaneously changing the inverter from grid-forming mode to grid-following mode.
Click to expand...

I only know for sure with the older systems, Outback, Xantrex, SMA and the like, but that is exactly what they do. They sync up to the generator and then close a relay and at the same instant the inverter shuts off. The generator powers only the load for a moment, then the inverter starts to work in reverse and ramp up charging to the battery. The process reverses once the charge cycle is done.

wwhitney said:
That seems complicated, but now that I think about it, with the Powerwall 2 system I have, I don't believe there's a power interruption when the system is in standalone mode (grid down) and it sees the grid return and cut over to the grid. So it must be doing the same thing, and perhaps this is just standard for hybrid inverters.

Or maybe in both cases there is a power interruption of under 1 cycle, and most loads are able to ride that through.
Click to expand...

On single phase, syncing is not really essential, look at backup computer UPSs, they just sit there dead until an outage, then they kick in, in an instant, and the computer never knows the difference.
 
Twisting in the wind over interpretations of the NEC notwithstanding, what could be wrong with making the N-G bond in the main panel?
 
Last edited:
electrofelon said:
I think I addressed this in post #25. It wouldn't be allowed if it was a service. It's just more stuff connected to neutral and subject to voltage differences from that.
Click to expand...
Maybe I am misunderstanding you, but I have seen N-G bonds in main panels all the time even when there is a disco with OCPD ahead of the panel.
 
ggunn said:
Something makes the decision of what source or sources is/are powering the loads.
Click to expand...
Read further in my paragraph you quoted. The inverter itself is the arbiter. There is nothing external deciding whether to use inverter power or generator power. It is nothing like utility/generator/ATS. Rather the inverter is the heart of the system and the generator is its slave. It powers the loads and it charges the batteries, at the behest of the inverter and its controls.

(Actually some systems like Tesla do use ATSs. This is apparently because Elon did not hire anybody who knows how this has been done for the last quarter century and they came up with their own [dumb] ideas.)
 
Birken Vogt said:
Read further in my paragraph you quoted. The inverter itself is the arbiter.
Click to expand...
OK, then; there is one. If there is no sunlight and the batteries are drained, the generator powers the loads. If there is plenty of power from solar and/or batteries, there is no reason to run the generator.
 
Last edited:
Birken Vogt said:
What is the point of this argument exactly?
Click to expand...
Only that you said that there was no arbiter and then you said the inverter is the arbiter; I agree that there is obviously some sort of switching function at the output of the inverter. Maybe calling it an arbiter is not the correct terminology.
 
ggunn said:
what could be wrong with making the N-G bond in the main panel?
Click to expand...
Nothing wrong
It apply when ATS with or within the hybrid inverter brand
It not apply when a paralleling arrangement in another hybrid inverter brand for generator supply. In that case, ng bond generator retain, ng bond at inverter disconnect on parallel operation, no ng bond main panel and so only one ng bond. On inverter only operate, ng bond at inverter by contactor in inverter and generator with its own ng bond disconnect, no ng bond in main panel and again only one ng bond.
 
ggunn said:
Only that you said that there was no arbiter and then you said the inverter is the arbiter; I agree that there is obviously some sort of switching function at the output of the inverter. Maybe calling it an arbiter is not the correct terminology.
Click to expand...

Therein lies the problem. There is no switching at the OUTPUT of the inverter. The inverter calls for generator and then takes power from the generator into itself. It may pass generator power on to the loads but that is only incidental. The inverter does not switch BETWEEN itself and anything else. The inverter is always the "source".

This thread is about the place of the N-G bond, GES etc. on an off grid system. Some postulated that it maybe should be at the generator.

I thought of another analogy for this. If the generator is a valid place for the N-G bond then the solar array would also be a valid place. But this would be ridiculous. The generator, solar array, some water wheel, are all auxiliary sources that feed into the inverter (or battery system) but the inverter is the one thing that is always there, always running, always required.
 
You must log in or register to reply here.
Top