AC disconnect sizing for dual voltage inverter

[JohnP]

I am in the midst of an install and I am questioning the sizing of the AC disconnect for the inverter.

Inverter is an SMA SB20000TL, 277/480V, 3-phase. AC disconnect is 30A.

Inverter output is rated at 24A
Inverter Maximum permissible fusing is at 50A.

Our utili connection is 277V.

If we calc the short circuit current for the inverter at 277V, I get 90.25A
If we calc the short circuit current for the inverter at 480V, I get 52.08A

Both values are above the permissible fusing of 50A.

Note that the inverter has a listed output of 24A, which applied to the 1.25 s.f., gives us 30A, which I am sure is why the 30A disconnect was chosen and installed.

Please advise.

[inetdog]

When you are dealing with three phase electrical systems, you have to distinguish between the line-to-line current, at the line to line voltage, and the line current at the line-to-neutral voltage.
A fuse or circuit breaker by its nature must be sized for the line current, since it will interrupt that line.

How are you calculating the "short circuit current"? You cannot assume that the inverter is a linear power source.

If we look at the 20,000 watt nominal output (I assume that from the model name), that is 6,667kW per line if the output is symmetric across the three phases.
With a nominal line-to-neutral voltage of 277V, that would be 24A. A GTI does not normally have a "surge" rating, although the NEC calculates as if it could deliver 1.25 times its nominal power in the event that the input DC is above the nominal panel power for some reason (cloud lensing and snow + cold weather, for example).
There is no way that you will see a higher current than that in any of the three line connections to the inverter.
Since the NEC considers the current to be continuous (lasting longer than three hours at a steady value), the rule for sizing the wiring and the overcurrent protection will be 1.25 x 1.25 = 1.56 times the nominal current output. The NEC minimum fuse or CB would be > 37A so a 40A would be the minimum. SMA says that the maximum allowed by them is 50A. No conflict there.

I really think that your problem is in the way you are determining the short circuit current. If you short the output of a GTI, it will stop producing output because the reference grid voltage will have gone away.

[carultch]

I am in the midst of an install and I am questioning the sizing of the AC disconnect for the inverter.

Inverter is an SMA SB20000TL, 277/480V, 3-phase. AC disconnect is 30A.

Inverter output is rated at 24A
Inverter Maximum permissible fusing is at 50A.

Our utili connection is 277V.

If we calc the short circuit current for the inverter at 277V, I get 90.25A
If we calc the short circuit current for the inverter at 480V, I get 52.08A

Both values are above the permissible fusing of 50A.

Note that the inverter has a listed output of 24A, which applied to the 1.25 s.f., gives us 30A, which I am sure is why the 30A disconnect was chosen and installed.

Please advise.

The concept of "short circuit current" DOES NOT apply to inverter outputs. These devices are current-limited. They will adjust performance to avoid exceeding that nameplate current listed on the datasheet, continuously.

The extra 1.25 enhancement factor, applies to modules that are not current-limited. It is in anticipation of extra sunlight striking the surface, beyond the standard 1000W/m^2 at which they are tested in the factory.

The value listed on the inverter datasheet as output current, is the current you use in ALL of your calculations. It is treated as any other continuous load by the NEC.


The 20000TL inverters at 277/480 Volts are rated at 24 amps. You use 24 amps*1.25 for selecting just about all material and equipment that will be connected downstream. Fuses, breakers, disconnects, wire, etc. The exception is devices that are rated for continuous duty, such as typical unfused disconnects (but you do have to check datasheets to be sure). These can be used with a continuous current up to their full nameplate rating.


277 Volts is the voltage from phase to neutral.
480 Volts is the voltage from phase to phase.

The conversion of power to current can either be calculated as:
20000/277/3 = 24A
or
20000/480/sqrt(3) = 24A


Fuses, breakers, disconnects, wiring, for this particular inverter, only need to be 30 A. Because 24A * 1.25 = 30A. This is the continuous load factor, and is the only factor you need to add onto inverters.

You are permitted to connect it to higher amps of breakers or fuses, if you want to, up to 50A listed as the maximum fuse size. Maybe because of an existing spare breaker in the panelboard. You do need to consistently upsize the wire, if you do.

For this particular family of inverters, SMA's manual requires #8 wire (section 6.3.1) for compatibility with terminals. This is a curve ball to watch out for, because the NEC default size would be to use #10 wire at this amps.

[carultch]

Since the NEC considers the current to be continuous (lasting longer than three hours at a steady value), the rule for sizing the wiring and the overcurrent protection will be 1.25 x 1.25 = 1.56 times the nominal current output. The NEC minimum fuse or CB would be > 37A so a 40A would be the minimum. SMA says that the maximum allowed by them is 50A. No conflict there.

What you said earlier was all correct, up until this point.

The 1.56 factor, which is the product of the enhancement factor and the continuous load factor, ONLY applies to the DC side of the system. Modules, source circuits, and combiner box outputs.

It does not apply to inverters, or DC-to-DC power optimizers. It is only a 1.25 continuous load factor that applies to these.


30A is the minimum OCPD for this inverter output.

[inetdog]

The value listed on the inverter datasheet as output current, is the current you use in ALL of your calculations. It is treated as any other continuous load by the NEC.

This may be the case for those who are under the 2014 version of the NEC. For those who are on earlier versions it will vary, but a safe assumption is that you have to size your wires and breakers at nominal output current time 1.25 x 1.25 and that the backfed amps you need to use when applying the 120% rule at all panels between the GTI and the service will be the rating of the breaker closest to the inverter of the sum of those breakers if you have more than one inverter. This does NOT count any output breaker that might be part of the GTI itself, since that is considered "supplemental" overcurrent protection and is ignored by the NEC.

[carultch]

This may be the case for those who are under the 2014 version of the NEC. For those who are on earlier versions it will vary, but a safe assumption is that you have to size your wires and breakers at nominal output current time 1.25 x 1.25 and that the backfed amps you need to use when applying the 120% rule at all panels between the GTI and the service will be the rating of the breaker closest to the inverter of the sum of those breakers if you have more than one inverter. This does NOT count any output breaker that might be part of the GTI itself, since that is considered "supplemental" overcurrent protection and is ignored by the NEC.

Interesting. I never would've expected that.

How long ago was it a requirement to use a 1.56 factor on inverter output circuits? I've been working in solar since 2010, and I've always used a 1.25 factor on all throughout the AC side. I've never even heard of using anything else. (I'm not referring to the 120% backfed bus ampacity rule)

Is anyone still on a 2005 cycle of the NEC?

Some inverters do adapt their current ratings to the low end of the voltage range, instead of the nominal value. I recently used 20kW Solectria inverters that I'd think would be 24A intuitively, and then checking the datasheet, I saw 27.3A. For the gang of 13 inverters, it kicked my breaker panel and master switch from 400A to 600A.

[inetdog]

The rule, to the extent that it was written explicitly, at least for [2008] and[2011], is that the output wires have to be rated for the nominal amperage times 1.25 to allow for higher than nominal output under unusual conditions. Then, when sizing the circuit breaker you have to multiply by another 1.25 because the normal operating current will be continuous. (Unless you use a 100% rated breaker in a 100% rated panel, which is hard to find in residential equipment.)
Which then brings us back to having to size the wire so that it is protected against fault current by the (nominal x 1.56) breaker, unless you can take advantage of the next size up provision, which does not really fit here.

[carultch]

The rule, to the extent that it was written explicitly, at least for [2008] and[2011], is that the output wires have to be rated for the nominal amperage times 1.25 to allow for higher than nominal output under unusual conditions. Then, when sizing the circuit breaker you have to multiply by another 1.25 because the normal operating current will be continuous. (Unless you use a 100% rated breaker in a 100% rated panel, which is hard to find in residential equipment.)
Which then brings us back to having to size the wire so that it is protected against fault current by the (nominal x 1.56) breaker, unless you can take advantage of the next size up provision, which does not really fit here.

I think we are talking about two different things here.

What you are saying sounds like it applies to the "unprocessed" DC circuits, of modules, source circuits, and combiner box outputs.

Grid-tied inverter output circuits by contrast, according to 690.8(A)(3) in NEC2011 and NEC2014, do not require the 125% enhancement factor.

[inetdog]

I think we are talking about two different things here.

What you are saying sounds like it applies to the "unprocessed" DC circuits, of modules, source circuits, and combiner box outputs.

Grid-tied inverter output circuits by contrast, according to 690.8(A)(3) in NEC2011 and NEC2014, do not require the 125% enhancement factor.

I agree.....