High amps on new TL series inverters 11

[epm]

The older SMA Sunny Boy 8000 always had the benefit of being installed with a 40 amp breaker on 240 volt single phase system with no issues. The maximum output of 32 amps multiplied by the code mandated fudge factor of 1.25% equaled exactly 40 amps.
Well, the newer 8000TL has a maximum output of 33.4 amps at 240 volts, and a 40 amp breaker no longer is compatible when the 1.25% is added.
I don't know why SMA would do this. That was the main advantage of the 8000. it could be installed on a 40 amp breaker in a 200 amp panel without isssues and be code compliant.

 

[SolarPro]

The same issue has come up in the past with the 4kW vs 3.8 kW models. Typically this means a product engineer or engineering team in Germany didn't listen to or forgot to take into account feedback provided by the US applications engineering team. In the past, SMA has added a current-limited SKU at a later date that neatly works with the NEC.

 

[epm]

Interesting you mention the current limiters for the US. The "US" designation on the 8000TL-US is supposedly for the US markets, but the rated output is still 33.4 amps.
This little bit higher output is going to have people very frustrated once inspectors catch on.

 

[SolarPro]

Keep in mind that the maximum inverter output current is really just a software set point. It's not like the hardware changes in an inverter rated for 33.4A vs. 32A. It's just a software change (along with a new or modified part number and a new or modified product listing.)

There are tradeoffs with either set point. The nice thing about choosing 33.4A as your set point is that you have nice round number for the inverter model number and rated power (8,000 W). The nice thing about a 32A set point, is that 7,680 W maxes out the load side connection allowance in a residential 200 A panel, which is a pretty common design scenario.

Either way SMA decides to rate the inverter, the components inside the box are likely identical. All else being equal, manufacturers generally prefer to have the higher power rating, because that reduces the $/watt.

 

[iwire]

Not real familure with recent NEC PV rules but does the NEC actully require the breaker in the panel for the backfeed from the inverter to be the max size allowed by the inverter?

In other words if I have an inverter with 33.4 amp output why can't I treat it as a 32 amp output?


For instance I can certainly code compliantly install a 100 amp generator to power a 50 amp load using a 60 amp breaker. There is no rule that must use a 100 amp breaker.

 

[Garrison]

Not real familure with recent NEC PV rules but does the NEC actully require the breaker in the panel for the backfeed from the inverter to be the max size allowed by the inverter?

In other words if I have an inverter with 33.4 amp output why can't I treat it as a 32 amp output?


For instance I can certainly code compliantly install a 100 amp generator to power a 50 amp load using a 60 amp breaker. There is no rule that must use a 100 amp breaker.

2011 NEC 690.8(B)(1)(a) [2014 NEC 690.9(B)] requires that overcurrent devices carry not less than 125% of the maximum currents calculated in 690.8(A), which for the inverter output circuit is the current listed on the data sheet.

To the OP, if you want to stick with SMA you could use (2) 3800TL. 2 inverters will cost more to install, but 4 MPPT could offer some design flexibility. Otherwise there are quite a few 7600W inverters out there, Power-One/ABB 7.6TL being one good example.

 

[epm]

The NEC requires the breaker to be sized to the rated output of the inverter, regardless of what the actual output is. The actual output is a moving target due to the nature of solar. A lot of systems are oversized on the DC side to account for fluctuations in solar intensity due to factors such as the time of the year, angle of the modules, ambient temperature, etc. However on a perfect day and a prefect senario the system could produce more wattage than accounted for, and the actual output of the inverter could approach the rated output.

 

[ggunn]

The NEC requires the breaker to be sized to the rated output of the inverter, regardless of what the actual output is. The actual output is a moving target due to the nature of solar. A lot of systems are oversized on the DC side to account for fluctuations in solar intensity due to factors such as the time of the year, angle of the modules, ambient temperature, etc. However on a perfect day and a prefect senario the system could produce more wattage than accounted for, and the actual output of the inverter could approach the rated output.

That is of course true, but I have wondered as well why it is that way. In a properly designed system the inverter is incapable of endangering the conductors; the OCPD is there to protect the conductors from drawing too much current from the service in the event of a fault. Why should the NEC care if the breaker is smaller than 1.25X the maximum rated output current of the inverter when the only possible consequence is nuisance trips of the breaker?

But doesn't the 2014 code allow for the maximum output current of the inverter to be used in the calculation of the 120% rule rather than the rating of the backfed breaker? If that is the case the 33.4A rating of the inverter would be used and it would therefore fit into a 200A panel.

 

[GoldDigger]

Yes, and that might be a factor in SMA's decision to, for now anyway, market just the 8000 and not a 7600.

 

[SolarPro]

But doesn't the 2014 code allow for the maximum output current of the inverter to be used in the calculation of the 120% rule rather than the rating of the backfed breaker? If that is the case the 33.4A rating of the inverter would be used and it would therefore fit into a 200A panel.

No. Here's what the 2014 Code says at the start of 705.12(D)(2):

(2) Bus or Conductor Ampere Rating. One hundred twenty five percent of the inverter output circuit current shall be used in ampacity calculations for the following:

 

[jaggedben]

That is of course true, but I have wondered as well why it is that way. In a properly designed system the inverter is incapable of endangering the conductors; the OCPD is there to protect the conductors from drawing too much current from the service in the event of a fault. Why should the NEC care if the breaker is smaller than 1.25X the maximum rated output current of the inverter when the only possible consequence is nuisance trips of the breaker?

Nuisance tripping can be a safety problem because often the 'solution' is to upsize the breaker to something that's more than the ampacity of the wire. (I once surveyed a house with a 240V 10awg oven circuit. There were four or five 30/20 quad breakers lying around the vicinity of the load center. The circuit was on the 40 of a 40/20 quad. It was clear what had happened.)

And another thing... I really don't want my customers solar systems to be nuisance tripping and them losing the money I'm supposed to be saving them. On many systems the customer has no occasion to notice until they get their next bill.

So given that, the problem is where do you draw the line? A 40A breaker on a max 33.4A inverter probably would never be a problem. But how about 36A? 38A? The latter probably has pretty decent likelihood of tripping after a few hours of continuous output. And then if the load center is in the sun or something...

 

[Zee]

OP:
Don't get me started on the SMA offering of perfectly round, 1,000 Watt, German-engineered, inverter increments:rant:
Except for the 3800, (which is nice) every single one of their 3, 4, 5, 6 and 8 kw inverters are just 1 amp or less (@125%) over the relevant breaker increments. (15, 20, 25, 30, 40)

Garrison had an excellent idea: use 2 @ 3800 TL's. I second.
Plus, you will add significant value with those inv.s. They are excellent inverters with amazing features (like a solar power outlet) ...... and an actually usable V window.
(Have you ever been able to fit your strings (on a resid.job) into the impractically narrow V window of a 6kw+ SMA TL?)

 

[Zee]

check the latest spec sheet. SMA usually corrects these mistakes eventually......it may now be 32A??

 

[epm]

The reason that the breaker has to be sized at 125% of output is that breakers are not designed to carry the full rating for a continuous load. They are designed to carry 80% load which is the same as output multiplied by 1.25. The exception would be 100% continous rated breakers. A continous load is considered 3 hours of more.
Same for conductors. Any circuit is really not supposed to have a continous load above 80% of its rating. Happens all the time though.

 

[jaggedben]

OP:
Don't get me started on the SMA offering of perfectly round, 1,000 Watt, German-engineered, inverter increments:rant:
Except for the 3800, (which is nice) every single one of their 3, 4, 5, 6 and 8 kw inverters are just 1 amp or less (@125%) over the relevant breaker increments. (15, 20, 25, 30, 40)

That's not an accident. If nominal voltage around here were 250 instead of 240 then all would be perfect.

But I completely agree with how annoying it is.

 

[Zee]

That's not an accident. If nominal voltage around here were 250 instead of 240 then all would be perfect.

But I completely agree with how annoying it is.

Always good to share the pain

Wait..what? Is it 250 VAC in Germany?

 

[GoldDigger]

Always good to share the pain

Wait..what? Is it 250 VAC in Germany?

Much more likely that they just are not subject to the 80% rule for breakers (or current x 1.25) and as was suggested earlier they just forget about that lesson every now and then.