Nec 690.42(B)2

[innersurffer]

I understand the transliteration of this code.
I understand the inverter is actually pushing power into the main buss bars.
The utility (thru the main OCPD) *is also pushing power into the main buss bars.
Is the reason that we don?t go over this amount due to the fact that the main buss bars (only rated up to 120 percent) *would be put at risk due to the 2 sources of power pushing into the main buss bar and over-rating *it more than 120%?
*
I guess this makes sense, I just don?t ever see this happening. *Loads are created on demand, not on supply right? I?m just curious on your take on this. On the other hand, If you were to compare this to AIC ratings from a transformer, then were talking about potential, and that makes sense. Also I guess if there was a short downstream (between these 2 feeds) and both OCPD?s did not work, then this 120% rule would really come into play.
*

 

[GoldDigger]

I understand the transliteration of this code.
I understand the inverter is actually pushing power into the main buss bars.
The utility (thru the main OCPD) *is also pushing power into the main buss bars.
Is the reason that we don?t go over this amount due to the fact that the main buss bars (only rated up to 120 percent) *would be put at risk due to the 2 sources of power pushing into the main buss bar and over-rating *it more than 120%?
*
I guess this makes sense, I just don?t ever see this happening. *Loads are created on demand, not on supply right? I?m just curious on your take on this. On the other hand, If you were to compare this to AIC ratings from a transformer, then were talking about potential, and that makes sense. Also I guess if there was a short downstream (between these 2 feeds) and both OCPD?s did not work, then this 120% rule would really come into play.
*

Look at it this way:
There is no limit in the code on what the various breakers load breakers in the panel can add up to, so the only protection against a small overload (not a short circuit) overloading the bus is the main breaker or the breaker(s) feeding an MLO panel.
If the only feed is at one end of the bus and the total of the loads is more than the bus rating, then a main breaker rated at or below the bus rating protects the bus.
If you add PV backfeed at the same end of the bus, then the loads further down the bus could cause more current to be drawn through the bus between the backfed breaker and the first load breaker than the bus can handle. Under those circumstances, a 100% rule would be appropriate.
However, once you add the requirement that the backfeed be at the opposite end of the bus, You now have a situation in which you cannot overload the bus in the sense that the current leaving at any one stab will be supplied from both sides and cannot be larger on either side than the largest of the main and backfed sources. Since the CMP is not happy with this argument, they compromised and put in the 120% rule applying when the main and PV are at opposite ends of the bus.

In a sufficiently contrived situation, you might have the total current at one stab exceeding the bus rating, but the stab rating when fed from both sides is what really should count there. And that may not be separately tested and specified. Anybody know for sure? I would not expect that you could install on a bus a single breaker which has a rating higher than the bus rating, as a mechanical concern.

 

[jaggedben]

...
I guess this makes sense, I just don?t ever see this happening. *Loads are created on demand, not on supply right? ...
*

Practically speaking, you are essentially correct. I even had a member of a CMP (not the one that handles these parts of the code) point this out to me. "If the demand has never tripped the main breaker feeding a panel, then you can add any amount of solar up to the size of that main breaker without worrying about overloading the panel." But of course this isn't fool proof from an engineering point of view, nor does it prevent some later electrician from adding more loads, etc. etc.

Now if you ask me why the CMP for 690 picked 120% instead of 125% or 130% or 150%, I can't explain that for you. But GoldDigger is correct that part of the story is that load centers aren't actually tested for this sort of thing.

 

[GoldDigger]

Practically speaking, you are essentially correct. I even had a member of a CMP (not the one that handles these parts of the code) point this out to me. "If the demand has never tripped the main breaker feeding a panel, then you can add any amount of solar up to the size of that main breaker without worrying about overloading the panel." But of course this isn't fool proof from an engineering point of view, nor does it prevent some later electrician from adding more loads, etc. etc.

Now if you ask me why the CMP for 690 picked 120% instead of 125% or 130% or 150%, I can't explain that for you. But GoldDigger is correct that part of the story is that load centers aren't actually tested for this sort of thing.

I find it interesting that some inspectors also apply the 120% rule to feeder wires that do not even have a box from which they could be tapped let alone an existing intermediate load connection. There is no logic behind it except a literal application of the code language.

 

[jaggedben]

I find it interesting that some inspectors also apply the 120% rule to feeder wires that do not even have a box from which they could be tapped let alone an existing intermediate load connection. There is no logic behind it except a literal application of the code language.

I agree, and would go even further to say that the idea that an intermediate load connection presents a real danger has little support.

Say a tap is done with a Polaris type connector that is listed for all the wire sizes involved. What is the difference, from the connector's point of view, between two sources serving one load vs. one source serving two loads in the opposite direction? It is the same amount of current passing through the same terminals and wires, so it should not present a problem. Conceptually I believe this is true for any type of connector that is used according to its listing. The same logic might not apply to panelboards because overcurrent protection is integral to the testing there, but for connectors it should hold true.

 

[innersurffer]

Great responses. I agree

Sent from my SGH-T889 using Tapatalk

 

[GoldDigger]

I agree, and would go even further to say that the idea that an intermediate load connection presents a real danger has little support.

Say a tap is done with a Polaris type connector that is listed for all the wire sizes involved. What is the difference, from the connector's point of view, between two sources serving one load vs. one source serving two loads in the opposite direction? It is the same amount of current passing through the same terminals and wires, so it should not present a problem. Conceptually I believe this is true for any type of connector that is used according to its listing. The same logic might not apply to panelboards because overcurrent protection is integral to the testing there, but for connectors it should hold true.

The connector should not really care, but you do have to make sure that the sum of the two input currents does not exceed the ampacity of the third wire on the Polaris connector, assuming that there is no downstream OCPD to protect it. I think that the one legitimate application of the 120% rule to tap connectors is the similar to the justification for why it is not safe (nor allowed of course) to plug in a non-UL cord-connected inverter to a receptacle outlet on a circuit which also serves other loads. Namely, that the branch OCPD which originally protected the LOAD and its wiring from overcurrent can no longer do that since there are now two sources of current. In that case, however, a 100% rule would be more appropriate than a 120% rule.
It gets frustrating.

 

[innersurffer]

Inverters generally draw little and put out a lot. This is the confusing part that makes no sense. There pushing, not pulling current and they help take care of the immediate demand before they ever push over the panels buss limits.

Sent from my SGH-T889 using Tapatalk

 

[GoldDigger]

Inverters generally draw little and put out a lot. This is the confusing part that makes no sense. There pushing, not pulling current and they help take care of the immediate demand before they ever push over the panels buss limits.

Sent from my SGH-T889 using Tapatalk

You are missing the point.
They will be supplying part of the current to the load(s), whatever they may be. Initially, that means less current through the main (and less current coming through the bus from the direction of the main.)
But it also means that a defective load or an overloaded branch combined with other normal loads could end up getting more total current than the POCO-driven main OCPD would allow to flow. That is what potentially overloads the bus or the feeder conductors in question.

 

[jaggedben]

The connector should not really care, but you do have to make sure that the sum of the two input currents does not exceed the ampacity of the third wire on the Polaris connector, assuming that there is no downstream OCPD to protect it.

Of course there should be a downstream OCPD to protect the third wire, per 240.21(B). I mean, if we are talking about an unprofessional installation by someone who doesn't know to protect wires with an OCPD, then all bets are off anyway IMO. They could use a wire without a sufficient ampacity for the load, or make a host of other mistakes. But someone who follows 240.21(B) is not going to create a dangerous situation merely because the feeder they tap happens to be fed by two sources.

I think that the one legitimate application of the 120% rule to tap connectors is the similar to the justification for why it is not safe (nor allowed of course) to plug in a non-UL cord-connected inverter to a receptacle outlet on a circuit which also serves other loads. Namely, that the branch OCPD which originally protected the LOAD and its wiring from overcurrent can no longer do that since there are now two sources of current. In that case, however, a 100% rule would be more appropriate than a 120% rule.
It gets frustrating.

We have the dedicated OCPD rule (705.12(D)(1)), which amounts to the 100% rule that your talking about. Indeed it amounts to a prohibition on cord-connected inverters even if UL dropped the ball and listed one.

I think that applying the 120% rule to conductors (instead only panelboards) is a mistake. There should be different rules for conductors. Conductors that can be fed by two sources from opposite ends should be allowed to be rated only for the larger source OCPD. Conductors that can be fed by multiple sources from the same end should be rated for the sum of all sources, unless protected at the opposite end by a single OCPD. Something like that at, at any rate, without applying any percentages.

Right now I can't remember what the 2014 code is going to say about this, but I do think it moves in a somewhat more logical direction.