From an article in the IAEI magazine written by John Wiles who wrote the code, "This section has been revised to specifically require that multiple inverters in a single PV system shall be connected to the existing premises wiring system at a single dedicated circuit breaker or fusible disconnecting means. This section no longer allows multiple connections to a load center or panelboard where there are multiple inverters involved. Multiple inverters must first be combined in an AC combining panel and the output of the panelboard is then connected to the single point of connection in the distribution equipment through one circuit breaker or fusible disconnection means."
I'll be blunt: It really bothers me that Wiles is writing nonsense in IAEI magazine.
It's nonsense on the face of it, because the code makes no distinction between combiner panelboards and other panelboards. So if this section is supposed to prohibit multiple breakers in a service panelboard, then in effect it prohibits combiner panels as well, and therefor in effect prohibits multiple inverters
at all in many situations.
It's also nonsense because
nowhere in the
Report on Proposals or
Report on Comments is there
any talk of multiple breakers in a panel presenting any kind of issue or of any intent to require combiner panels. As best I can tell, the change in the 2014 code was due to Bill Brooks' proposal 4-394 Log #2647 NEC-P04 (see page 813 at the first link above). The purpose of which was "to clarify that the interconnection point of a utility-interactive inverter must not contain loads between the inverter output and the overcurrent protection device", and to allow "Multiple inverters [to] be connected to the same circuit, as with micro-inverters or ac modules." So somehow a proposal intended to
allow multiple inverters to be on the same
branch OCPD gets flipped around to
require all inverters to be on the same
feeder OCPD? Again, nonsense. I'd say the CMP botched it up, but what's really botching it is people (including Wiles, apparently) not carefully reading either the plain language of the code or the ROP or ROC.
Yes, John Wiles wrote the original version of 690 many, many years ago. That doesn't make the final authority on code interpretation. Lately the CMP has been rejecting most of his proposals.
Ask Bill Brooks about this one. He's just as much an authority, and it was his proposal which was accepted in principle.
To avoid the expense of installing a dedicated AC combiner panel?
Just for a two inverter system.
For inverters in separate buildings to avoid the need to run an inverter output circuit from one building to the other.
Thank you. These are important cost considerations in a competitive business, especially the latter. (The former is not trivial either, saving time and money on a 20/20 quad vs a subpanel for an 18 panel micro-inverter system where both the extra expense and the profit are measured in the hundreds.). We don't need onerous and clearly erroneous code interpretations adding costs when there is no safety issue.
Then what does the rule apply to?(705.12(D)(1)
The purpose of the rule is to prohibit loads and sources from being on the same branch circuit. That's really the entire purpose, IMO.