"Interesting" question

[ggunn]

We have a system under design that uses 480V inverters connecting to a 208V service through a transformer. It just so happens that converting the inverter current using 480V/208V comes out to 401.5A, which bumps our OCPD up to 600A, which then makes us increase the conductor size. My question is this: Can I legitimately use the transformer efficiency to justify the current being less than 400A?

 

[gadfly56]

We have a system under design that uses 480V inverters connecting to a 208V service through a transformer. It just so happens that converting the inverter current using 480V/208V comes out to 401.5A, which bumps our OCPD up to 600A, which then makes us increase the conductor size. My question is this: Can I legitimately use the transformer efficiency to justify the current being less than 400A?

It sounds reasonable to me, as long as the transformer output doesn't already account for the losses.

 

[ggunn]

It sounds reasonable to me, as long as the transformer output doesn't already account for the losses.

It doesn't, since when we use the 480V/208V current conversion, we are assuming an ideal transformer.

 

[pv_n00b]

I don’t know of anything in the NEC that would allow this. Maybe if you had documentation from the manufacturer to show the loss and the AHJ allowed an AMM then you could do it. Rounding is your friend. I’ve always questioned the NEC allowing rounding 5.4A down to 5A but not allowing 501A to round down to 500A. Rounding the smaller number is a much larger percentage change.

 

[pcanning87]

Isn't the next standard size OCPD 500A? Unless you're talking before 125% in which case yes you'd be at 600A OCPD.

But can't you use the next size up OCPD without increasing conductor size since you're under 800A, or am I missing something?

Also, I've definitely seen slight overages like this stamped by PEs. Call it "engineering discretion".


Sent from my iPhone using Tapatalk

 

[Carultch]

Isn't the next standard size OCPD 500A? Unless you're talking before 125% in which case yes you'd be at 600A OCPD.

But can't you use the next size up OCPD without increasing conductor size since you're under 800A, or am I missing something?

Also, I've definitely seen slight overages like this stamped by PEs. Call it "engineering discretion".


Sent from my iPhone using Tapatalk

Actually, the next standard size after 400A is 450A. And the OP is including the 125% factor, as is required for standard OCPD used in PV systems.

Assuming the transformer is ideal, which we know no transformer is, that brings the current to just slightly above 400A, which causes the OP to round up to a 600A OCPD on the 208V side instead of a 500A OCPD that would more closely match the load. Rounding the current down to 400A on the 208V side is not completely unreasonable, as few (if any) transformers are 99.6% efficient or greater. Assuming both voltages are ideal at 480V and 208V, you are going to get less than 400A out of this transformer in practice. Even though no NEC rule allows you to take credit for this.

One way to make your design work with a smaller OCPD and conductors, is to use a continuous duty rated breaker at 450A.

 

[electrofelon]

Interesting quandary. I worked on a a bunch of ~400 kw systems summer before last , just as an installer not designing but I always like to double check what other people come up with, and it turns out that several of the systems had a similar situation where the current was slightly over a standard size and they had just "rounded down" and called it good. I dont remember how much over they were, but it was more than your value of a few amps I do know that. Anyway, I remember almost posting a thread on that topic of "how close would it have to be for you to just round it down", or "would you have rounded down".

You could investigate whether the inverter(s) can be programmed for a different power setting. I had never heard of this or considered it before the last 1.3 meg system we did, but there the utility was super strict that the inverter outputs not be over 1 meg on the money. The factory values was a hair over. These were Huawei inverters, and they sent someone out to reprogram the power setting.

 

[electrofelon]

Another idea that just came to mind, what about adjusting the taps 1 notch?

 

[jaggedben]

Another idea that just came to mind, what about adjusting the taps 1 notch?

Won't that throw off the voltage the inverter sees and effectively narrow it's operating window? And/or won't it clip the amount of power that can be exported? I'm not practically familiar with transformer taps, but from a theory standpoint I don't see how that works.

 

[GoldDigger]

Won't that throw off the voltage the inverter sees and effectively narrow it's operating window? And/or won't it clip the amount of power that can be exported? I'm not practically familiar with transformer taps, but from a theory standpoint I don't see how that works.

If looking at the current on the grid side of the transformer, then IF the inverter output is power limited you cannot reduce the output current on the 208V side by changing transformer taps.
If the inverter output is strictly current limited, then by changing the transformer ratio you can decrease the inverter output voltage to get less power which will correspond to less current on the fixed voltage grid side.

One other concern is that the magnetizing current of the transformer will be furnished from the grid side and that will increase the current on that side of the transformer beyond that corresponding to the inverter output.

 

[electrofelon]

If looking at the current on the grid side of the transformer, then IF the inverter output is power limited you cannot reduce the output current on the 208V side by changing transformer taps.
If the inverter output is strictly current limited, then by changing the transformer ratio you can decrease the inverter output voltage to get less power which will correspond to less current on the fixed voltage grid side.

Yes the latter as I believe inverters are current limited AND power limited. For example a solectria 36TL puts out 43.5 amps max at 480 volts. Increase the voltage and the current goes down. Decrease the voltage and the current stays at 43.5.

Yes your voltage window would certainly be smaller and your power would get reduced, but the latter is really what you need to do unless you are going to fudge the numbers and call it close enough, or go with the larger equipment. I am guessing the smallest tap change (2.5%?) would be much more than you need and would a deal breaker in terms of power lost and voltage window. If there was a .5% tap then that would be perfect

 

[ggunn]

Update: I bit the bullet and put in the 600A fuses and larger conductors/conduits. Better safe than sorry.

 

[jaggedben]

Yes the latter as I believe inverters are current limited AND power limited. For example a solectria 36TL puts out 43.5 amps max at 480 volts. Increase the voltage and the current goes down. Decrease the voltage and the current stays at 43.5. ...

This must depend on the inverter. All the inverters I'm most familiar with are primarily current limited and will (if the sun is there) output power above nameplate when grid voltage is higher than nominal.

 

[ggunn]

This must depend on the inverter. All the inverters I'm most familiar with are primarily current limited and will (if the sun is there) output power above nameplate when grid voltage is higher than nominal.

Are you sure about that? All of the inverters that I have seen spec that condition show lower current at higher voltage so that the power is the same. I guess that would make them technically power limited rather than current limited.

 

[Carultch]

Are you sure about that? All of the inverters that I have seen spec that condition show lower current at higher voltage so that the power is the same. I guess that would make them technically power limited rather than current limited.

I think they are both power and current limited. At less than nominal voltage, they produce less than nominal power, if the rated current matches the nominal power at nominal voltage. And at greater than nominal voltage, they produce less current in order to maintain their power rating.

Some inverters are rated to produce greater current when the voltage is lower than nominal.

 

[ggunn]

I think they are both power and current limited. At less than nominal voltage, they produce less than nominal power, if the rated current matches the nominal power at nominal voltage. And at greater than nominal voltage, they produce less current in order to maintain their power rating.

Some inverters are rated to produce greater current when the voltage is lower than nominal.

That's what I mean, and it's really power limiting within a voltage range. Not all inverters are that way; some (SMA comes to mind) are truly current limited and produce less power at lower voltage.

 

[Smart $]

That's what I mean, and it's really power limiting within a voltage range. Not all inverters are that way; some (SMA comes to mind) are truly current limited and produce less power at lower voltage.

Wouldn't matter. Code requires using the nominal values for the calculations.

 

[ggunn]

Wouldn't matter. Code requires using the nominal values for the calculations.

Sure, but that wasn't the topic of discussion. For wire sizing and OCPD, use the nameplate maximum current.

 

[Smart $]

Sure, but that wasn't the topic of discussion.

Perhaps not directly, but it is part of drawing the proper conclusion for....

... My question is this: Can I legitimately use the transformer efficiency to justify the current being less than 400A?

For wire sizing and OCPD, use the nameplate maximum current.

The actual phrase is "the inverter continuous output current rating" in 690.8(A)(3), then applied as provided for in 690.8(B) for Conductor Ampacity and 690.9(B) for Overcurrent Protection

 

[jaggedben]

Are you sure about that? All of the inverters that I have seen spec that condition show lower current at higher voltage so that the power is the same. I guess that would make them technically power limited rather than current limited.

I'm absolutely sure of it.

To be clear, I'm not talking about different nominal grid voltages, e.g. the spec current at 208V vs. 240V. I'm talking about what happens when the grid voltage fluctuates higher than nominal, e.g. you're measuring 247.6V on a nominal 240V service. An inverter with max output of, say, 32A, will still output 32A at the higher voltage and thus output a few percentage points more power. Just the other day I was looking at a 3800W nameplate inverter outputting over 4000W (at least according to it's own monitoring data).

Enphase inverters will output higher than nameplate watts if the voltage is high, but also have a power limit that they hit at somewhat above nameplate. Their approach at how they've labeled this has varied a bit on different models.