Or move the DAS supply to a separate enclosed OCPD on a tap on one of the incoming or outgoing feeders?
You can specify a 250A busbar. They are common in the 277/480V world.
True. Not sure off hand about other brands, but for Siemens, their smallest 480 panel board will have a 250A bus.
One thing to consider is that the rule is there to protect the busbars, so if there are one and two pole breakers in the panel they only contribute to the breaker count on the bus(es) they are connected to. This has been a minor controversy in here but I have used it in several systems and it has never been challenged by an AHJ.
Interesting, but I don't see how this could reduce the busbar rating needed. Wouldn't you find the busbar with the highest number of CB connections and that would set the minimum busbar rating? Since all the busbars will have the same rating that would set the overall busbar rating for the panel.
I see, that makes sense to me. If I have a two pole 20A breaker on A-B phase and a single pole 20A on C phase that would be 20A per phase/busbar. Thanks for that input!
Would not in the case that all inverters are 3 phase. But you could have, say, a 208Y/120V panel with (7) 208V 2-wire inverters, plus a single pole 120V circuit for monitoring, etc. Then if you balance 6 of the inverters, and put the single pole breaker on the leg unused by the 7th inverter, you will minimize the required busbar rating for the "sum of all breakers" rule. Which rule should be applied with consideration that the breakers sum vectorially, i.e. (3) 40A double breakers balanced around the 3 legs would sum to 40 * sqrt(3) A per pole.
The only thing that the rule considers is the ratings of the breakers. Total them up on each busbar, and the busbar with the most connected total breaker rating sets the minimum for the panel. On a three phase panel that combines single phase two pole inverter breakers it makes a big difference from simply totaling the breaker ratings in the panel.
That language "on each busbar" is not present in the text of 2023 NEC 705.12(B)(3). Rather it is implicit because that is clearly what is necessary based on the physics to protect the busbars.
And I'm further saying that you only need an 87A busbar, because 50 ∠ 0° + 50 ∠ 60° = 86.6 ∠ 30°.
Good luck finding a panel with an 87A busbar, but I wouldn't expect an AHJ to go for that, anyway. If the rule was about current it would be counting 125% of inverter currents as in the other 705.12 rules and your vector math might apply, but the rule is about breaker ratings.
Of course, 87A wasn't meant as a practical example. A practical example would be a 125A 3-phase panel with (3) 2-pole 70A breakers.
The breaker rating is a current rating, so breaker ratings add the same way currents add.
Breaker ratings on a bus add arithmetically when relating to 705.12(B)(3). I know that common practice in here is to beat a dead horse until it is a greasy spot on the ground, but I am done with this. I have said my piece and I hope it is of benefit when calculating compliance when combining single phase inverters into a three phase service.
That is an interpretation of 705.12(B)(3), but vectorially is another interpretation. Just like the words of 705.12(B)(3) are silent on the question of "per pole" or not, they are silent on the question of arithmetically vs vectorially. In both cases, we can be guided by the physics in choosing the proper interpretation.
Unfortunately no, because the particular rule under discussion references overcurrent devices. For all the other rules that go by 125% of inverter output current I agree with your approach.
