Help with 705.12(D)(2)(3)(b)

[Pinesburgh]

Can someone please help me out with the theory of 705.12(D)(2)(3)(b).
Question- Why do we add 20% to the rating of the busbar when the the two sources are at opposite ends of the bus?
Question- Or maybe a quick lesson in the difference between 705.12(D)(2)(3) a,b and c?

 

[Pinesburgh]

Sorry, I'm in Florida so working with 2014

 

[jaggedben]

We only add 20% because the code making panels have been very conservative. My understanding is that part of the reason is that UL standards for panelboards don't really test for the effects of having two sources, where the total current flowing in the panelboard could exceed its rating, although not at a single point. (a) and (c) both limit the total current in the panelboard to its rating, in different ways. I guess that's considered safe per the usual testing. (c) was put in to legitimize 'AC combiner panels' for multiple inverter outputs.

 

[ggunn]

Can someone please help me out with the theory of 705.12(D)(2)(3)(b).
Question- Why do we add 20% to the rating of the busbar when the the two sources are at opposite ends of the bus?
Question- Or maybe a quick lesson in the difference between 705.12(D)(2)(3) a,b and c?

a) is simple. It says if the total ratings of all OCPD feeding a busbar* add up to less than the busbar rating, you can put them anywhere.

b) says if the total ratings of all OCPD feeding a busbar* add up to between 100% and 120% of the busbar rating, then you have to put them at opposite ends of the bus. This is because Kirchoff's Law says that if you do that no part of the busbar will carry more than the busbar rating, although the 120% number is arbitrary. It also is implicit that if they* add to more than 120%, you either cannot do it or you must qualify the bus under (c).

c) says that if you disregard the OCPD feeding the bus from the service, the total of all breakers' ratings** attached to a busbar cannot exceed 100% of the busbar rating.

* In these cases if one of the breakers is from a grid interactive inverter, then the number is 125% of the rated max output current of the inverter(s), not the breaker rating.
** In this case it IS the breaker rating fed by the inverter(s).

 

[pv_n00b]

It's all based on the fear that loads attached to panels will pull more current than the panel is rated for. In a normal panel with one current source the main OCPD would open if this happened but on a panel with two or more current sources the loads could draw up to the sum total of the sources. Of course, if the panel loading is done correctly this would never happen, but there is a fear that people are not doing panel loading correctly to code and then there is no main OCPD to enforce it on the dunderheads.

The reason for the 20% allowance has to do with heating in the panel from the extra power going through it. Remember that panels have a given amount of resistance and some small percentage of the power flowing through gets converted to heat. The code makers on the CMP decided that 20% over would not cause excessive heating in the panel if it happened so that much was allowed. There was a lot of pressure to allow some reasonable overage or the residential PV market would die. How many PV installs would be stopped short if a service upgrade were required for everyone? Quite a few.

The bottom line is that all this assumes some dunderhead put too much load on a panel and no one ever noticed until the PV system was installed or someone came along after the system was installed and added loads that only operate when the PV system is producing power to draw more current than the panel was rated for. I have never talked to anyone who has found a case of this happening in the field, but it has the CMP scared so we get all these Rube Goldberg type rules.

 

[ggunn]

The bottom line is that all this assumes some dunderhead put too much load on a panel and no one ever noticed until the PV system was installed or someone came along after the system was installed and added loads that only operate when the PV system is producing power to draw more current than the panel was rated for. I have never talked to anyone who has found a case of this happening in the field, but it has the CMP scared so we get all these Rube Goldberg type rules.

In defense of the CMP I'll just point out that there is no limit to the amount of load that can be put into an MDP; I have seen load breaker ratings total to 2X or more the bus rating. If a PV backfed breaker were to be added at the same end of the bus as the main breaker in one of those panels, the loads could draw more through the bus than the rating of the bus before the main would trip.

 

[pv_n00b]

In defense of the CMP I'll just point out that there is no limit to the amount of load that can be put into an MDP; I have seen load breaker ratings total to 2X or more the bus rating. If a PV backfed breaker were to be added at the same end of the bus as the main breaker in one of those panels, the loads could draw more through the bus than the rating of the bus before the main would trip.

There is a limit on the load, but not a limit on the OCPDs in the panel. That's a crucial difference. When I'm talking about load I'm talking about the loads that are calculated using NEC chapter 2, not the rating of the OCPDs in the panels.

For a PV back feed to overload the bus there has to be a code violation relating to the load connected to the panel. If the NEC violating overload occurs when the PV system is producing power, and the PV output plus the main OCPD rating are greater than the overload then the bus will be overloaded. That's a narrow set of requirements to get an overload. And in designing 705.12 the CMP is assuming that the loading requirements in NEC chapter 2 will be violated and the PV connection requirements need to take this into account. Personally, I don't think code should be written assuming other parts of the code are going to be violated and a backstop to that violation needs to be put in place.

 

[jaggedben]

...

For a PV back feed to overload the bus there has to be a code violation relating to the load connected to the panel. If the NEC violating overload occurs when the PV system is producing power, and the PV output plus the main OCPD rating are greater than the overload then the bus will be overloaded. That's a narrow set of requirements to get an overload. And in designing 705.12 the CMP is assuming that the loading requirements in NEC chapter 2 will be violated and the PV connection requirements need to take this into account. Personally, I don't think code should be written assuming other parts of the code are going to be violated and a backstop to that violation needs to be put in place.

To disagree...

Calculated load has demand factors. It's possible that in a fluke situation (occupant turns on everything at once) or negligent overuse (grow house?) that a panel can draw more than the calculated load. So it is not the case that an overcurrent necessarily involves a code violation on the part of the installing electrician or supervising engineer.