You are, but it is just one of 5 options to protect the panel busbar.
Ease of administration: just look at all the breakers and see that it complies. The inspector doesn't have to verify which ones are loads and which ones are supplies. And FWIW, a supply breaker could have a high resistance fault that turns it into a load breaker as far as the busbar loading.
One thing about this rule is that the letter and the spirit can be different if there are breakers that do not touch all the busbars. If, for example, you are feeding a three phase AC combiner panel with single phase inverters, an interpretation that makes sense to me is that each busbar can only have breakers landing on it with the summed ratings up to 100% of that of the busbar. I used that interpretation many times when filling combiner panels and never had an inspection fail for it.
This rule could be one way to qualify a service panel bus but I don't believe it is commonly if ever used that way. It was added to the NEC in 2011 or 2014 (or thereabouts) to legitimize what we in solar were already doing with PV AC combiner panels.
I commonly use it when instead of adding a "PV disconnect" I had another "normal" 230.40 exc #2 or 3 main breaker panel board for the PV (really the same thing as a PV disconnect and combiner panel but just combing into one piece of equipment).
I guess that works in some jurisdictions but most around here require a separate bladed, labeled, and lockable AC disconnect for a PV interconnection. OCP is not necessary in the disco if the conductors are otherwise protected at the point of interconnection.
Physics-wise, it is clear that 2017 NEC 705.12(B)(2)(3) should be applied on a per busbar basis. So in a 3 phase panel, with only 3 pole breakers, you could just add up all the breakers to compare it against the busbar. If you have single pole and double pole breakers as well, you should do the comparison 3 times, once for each busbar, counting up only the breakers on that busbar each time.
20A per busbar. A 2P 20A breaker connected to the A and B phases adds 20A only to the A and B busbars.
Then you would use the highest ampacity of the 3 busbars. Got it-- I assumed that, but there was no code citing how it should be done. This way is the most obvious.
If you mean the busbar with the most total attached load and source breaker rating as the upper bound, then yes.
I my experience, that is mostly a utility requirement. The two major utilities here in upstate NY have an exception for small residential PV systems and dont require it.
Ok thanks. Seems silly....and LOL, "dont want to complicate the text" thats like the funniest thing ive heard all year, coming from the "can-hardly-make-a-coherent-sentence CMP's
Well, that way the same rule can cover combiner panels and panels that are underloaded with breakers. Also you don't have to argue about whether a breaker is a load or a source, which is helpful when it's both, such as the breaker to the micro-grid. Also what Wayne said.
It was added in 2014 and I've used it dozens of times to qualify service panels since it became code here in 2017.
I guess I am still not understanding why its not "sum off all load devices". IT seems very simple and more consistent with electrical theory. IS there a situation I am missing when it fails? Also that would have covered the common meter main with panelboard with feed thru lugs scenario.
Not mention the busbar in a MLO service panelboard (which although not allowed now, that change had nothing to do with concerns of overloading the busbar)
<Insert my well worn comment about having to comply with the NEC whether or not we agree with or understand the reasoning behind it>
