Breaker Sizing for Solar Project

[GoldDigger]

You may not have to use the first breaker size to calculate feeder breaker sizes but you do have to use them when applying the 120% rule. I should have been clearer on that.

 

[Carultch]

Hmm, what about this situation? A 20kw inverter at 480V comes out to 24.08A.
I've seen it listed as 24A and 24.1A-- 24.1 *1.25 is a 30.125A breaker. (The maker says 24A, so is 24.1A even worth considering?)

It is the value that is listed in datasheet, the manual, or on the unit itself that matters. If there is a discrepancy, I would go by which ever one is largest. Additionally, check the manual to see if the manufacturer requires a size larger than what the NEC alone requires (this is rare, but it can happen). It doesn't matter what the nameplate Watts calculates that the current should be, although this usually gives you a good estimate.

This goes without saying, but if the inverter can operate at multiple possible voltages, the rated current is a function of that voltage. So given an inverter that can operate at 208V, 240V, or 277V, it has a different current for each of the voltages and corresponding recommended breaker sizes to each.

So are you saying that with 3 of those inverters @ 90.375A, a 90A breaker would be ok, but if you had 5 of them @120.5A, the AHJ might require a 200A breaker, as 120.5A * 1.25 is 150.625A?
That does seen a bit ridiculous.

Given three inverters that are 20 kW on the dot in both power and current at nameplate 480V (i.e. 24.08A).

Each inverter is 24.08A.
Each inverter requires a 35A breaker.
The sum total of 3 inverters is 72.24A.
The AC combiner for 3 inverters requires a 100A main OCPD within the circuit. This is per my interpretation.

The sum total of 5 inverters is 120.4A.
The AC combiner for 5 inverters requires a 175A main OCPD within the circuit. Note that this is a fraction of an ampere above 150A.

These fraction of ampere inconsistencies are very frustrating, especially on much larger systems. Imagine if you were a fraction of an ampere above 1200A, and the rounding error means the difference between a panelboard and a switchboard. If you do end up making a judgement call and rounding that ampere fraction down to the nearest OCPD, worst case scenario is that you nuisance trip the breaker.

Also, what did you mean by "taking advantage" of 240.4(B) "where you can"?
Thanks.

240.4(B) is the next size up rule. The way I understand it is that for circuits where it is permitted to apply as discussed in that section, you only need as much wire ampacity as (A) required for the load with the 1.25 factor as is applicable, and (B) as required to be greater than the previous standard size breaker from the one that you are actually using.

For example, consider a main feeder for a circuit with a continuous load of 104A.
1.25*104A = 130A
Therefore, a 150A breaker is needed.
130A of wire & terminal ampacity is needed for the load alone.
The wire must be protected by a 150A OCPD.
Since 130A is not a standard OCPD, the next size up is 150A. The 130A of wire is considered protected by the next standard size breaker (i.e 150A) per 240.4(B).

So where the 240.4(B) rule can apply to this circuit, you can use #1 Cu wire. When it classifies as a feeder, you can use it. When it classifies as a tap or a transformer secondary, you'd need #1/0 Cu wire.

It is a very non-intuitive rule, I agree. I don't know why we have it, and there have been proposals to remove it. I would ordinarily think that you'd need at least as much wire ampacity as you have breaker amps, such that the breaker trips before the wire overheats.

 

[PVfarmer]

1. Additionally, check the manual to see if the manufacturer requires a size larger than what the NEC alone requires (this is rare, but it can happen).

2. This goes without saying, but if the inverter can operate at multiple possible voltages, the rated current is a function of that voltage. So given an inverter that can operate at 208V, 240V, or 277V, it has a different current for each of the voltages and corresponding recommended breaker sizes to each.

3. Given three inverters that are 20 kW on the dot in both power and current at nameplate 480V (i.e. 24.08A).
Each inverter is 24.08A.
Each inverter requires a 35A breaker.
The sum total of 3 inverters is 72.24A.
The AC combiner for 3 inverters requires a 100A main OCPD within the circuit. This is per my interpretation.
These fraction of ampere inconsistencies are very frustrating, especially on much larger systems.... worst case scenario is that you nuisance trip the breaker.

It is a very non-intuitive rule, I agree. I don't know why we have it, and there have been proposals to remove it. I would ordinarily think that you'd need at least as much wire ampacity as you have breaker amps, such that the breaker trips before the wire overheats.

1. You've lost me there- the manufacturers always seem to *recommend* a "max output OCPD size", but I don't see anything as far as requirements there. For the 24A/480V example- a 30A breaker would seem to be the one, but they say 50A max OCPD. Perhaps the extra 20A is for if/when there's another inverter with battery bank in the loop before the breaker or something... as in, with 24A of GTI output and another 10A of potential battery inverter output, you'd have to go with 50A OCPD.

2. Arrgh! I just lost the section that's mentioned in. The only place I see (saw) the variable voltage thing mentioned was input *to* stand alone PV systems. I'm thiniking the current for GTIs won't vary too much once you select the output voltage.
edit: found it- 690.8(4) mentions the current at lowest input voltage.

3. Why not a 30A? When 24.08A * 1.25= 30.1A, I don't see how 35 is necessary- if the output voltage is only going to vary + or - 3%, so will the current, and you won't really have to worry about nuisance trips perhaps.

 

[jaggedben]

It is the value that is listed in datasheet, the manual, or on the unit itself that matters. If there is a discrepancy, I would go by which ever one is largest. ...

The label on the unit has the final say. Some datasheet or manual might not be up-to-date. Not that I've ever had this problem where the product number matches in all three places. If I did, that would be a mark against that manufacturer.

Of course if I planned a whole project around a datasheet with a higher value then the point will most likely be moot.

 

[Carultch]

1. You've lost me there- the manufacturers always seem to *recommend* a "max output OCPD size", but I don't see anything as far as requirements there. For the 24A/480V example- a 30A breaker would seem to be the one, but they say 50A max OCPD. Perhaps the extra 20A is for if/when there's another inverter with battery bank in the loop before the breaker or something... as in, with 24A of GTI output and another 10A of potential battery inverter output, you'd have to go with 50A OCPD.

2. Arrgh! I just lost the section that's mentioned in. The only place I see (saw) the variable voltage thing mentioned was input *to* stand alone PV systems. I'm thiniking the current for GTIs won't vary too much once you select the output voltage.

3. Why not a 30A? When 24.08A * 1.25= 30.1A, I don't see how 35 is necessary- if the output voltage is only going to vary + or - 3%, so will the current, and you won't really have to worry about nuisance trips perhaps.

1. You can't mix other circuits on a GTI's first fuse or breaker. GTI's need a dedicated circuit. One reason you might want a breaker higher than minimum, is if you are unifying the layout of your panelboard to minimize chance of error.
2. I'm talking about configured nominal voltage. In otherwords, bear in mind that with a different nominal voltage, the manufacturer specifies a different current.
3. I can't say why. I'm just saying what the code says, and the code doesn't tell you that you can round down amperes in this calculation. I'm aware that voltage can be lower than nominal, and therefore so will output power if the maximum current is based directly off of nameplate power. The inverters are current limited and power limited. Some inverters have a maximum current that is built in anticipation of this issue, and is therefore higher than your intuition would tell you.

 

[PVfarmer]

1. You can't mix other circuits on a GTI's first fuse or breaker. GTI's need a dedicated circuit. One reason you might want a breaker higher than minimum, is if you are unifying the layout of your panelboard to minimize chance of error.
2. I'm talking about configured nominal voltage. In otherwords, bear in mind that with a different nominal voltage, the manufacturer specifies a different current.
3. I can't say why. I'm just saying what the code says, and the code doesn't tell you that you can round down amperes in this calculation. I'm aware that voltage can be lower than nominal, and therefore so will output power if the maximum current is based directly off of nameplate power. The inverters are current limited and power limited. Some inverters have a maximum current that is built in anticipation of this issue, and is therefore higher than your intuition would tell you.

1. Here's something from an installation manual-
Loads installed between the inverter and the circuit breaker must be fused separately.

I read that as- you can put in say a (3ph) well pump or something between an inverter and its breaker. Is there something preventing that in code? If not, you'd have to go with the 50A breaker protection for a 24A inverter plus 10A of water pump?

It says the above italics right after this-
In PV systems with multiple inverters, protect each inverter with a separate three-phase circuit
breaker. Observe the maximum permissible fuse protection (see Section 14 "Technical Data",
page 69). This will prevent residual voltage from being present at the corresponding cable
after disconnection.

Also- could you expand on this? Not quite sure what you mean by unifying.
"if you are unifying the layout of your panelboard to minimize chance of error."

2. Right, I'm with you on that one.

3. But does code say that 30.1A has to be rounded up to a 35A breaker? Isn't it 30.5A would be rounded up to 31A, so then a 35A breaker required?

 

[PVfarmer]

1. You can't mix other circuits on a GTI's first fuse or breaker. GTI's need a dedicated circuit. One reason you might want a breaker higher than minimum, is if you are unifying the layout of your panelboard to minimize chance of error.

Isn't it possible to have multiple PV inverter outputs going into an AC combiner, but run a hybrid inverter back-up system off of just one of those multiple outputs before the combiner?
If so, that's when the "oversized" PV output OCPD would be required, I suppose.

 

[jaggedben]

Isn't it possible to have multiple PV inverter outputs going into an AC combiner, but run a hybrid inverter back-up system off of just one of those multiple outputs before the combiner?
If so, that's when the "oversized" PV output OCPD would be required, I suppose.

It's possible, but it really has nothing to do with the discussion, and your conclusion is overly broad.

 

[ggunn]

The label on the unit has the final say. Some datasheet or manual might not be up-to-date. Not that I've ever had this problem where the product number matches in all three places. If I did, that would be a mark against that manufacturer.

Of course if I planned a whole project around a datasheet with a higher value then the point will most likely be moot.

When I see a discrepancy (and yes, it happens, and many times it's in the design phase where it will be a while before I can lay eyes on the gear), I call the manufacturer's tech support. Case in point: SMA made some changes to some entries in their inverter line where the max current is the same whether the AC voltage is 208 or 240. I had one data sheet that said one thing and another one for the same inverter that said something else. A call to tech support got it straight.