New 20kw Solar Install - just interconnected - Did we get screwed?

[neweclipse]

Why would it be welcome on lower production days? Makes ZERO since.
For most locations a system like this is going to clip most days. It is easily modeled using TMY data and there is no reason to wait with solar so easily modeled on monthly and annual basis.

SOLAREdge must be Magic stuff.
My sun goes behind clouds and some days I cannot even see the sun, Also I'm pretty sure the modeled clipping is just a sliver up on top. Am I to believe that that sliver can never sag during overcast skies? I imagine that with a little sag then 100% of the then REDUCED production is NOW all collected and that is no clipping in my book. Taking this to the next level, if there were never NO clipping and then a overcast sky or clouds come where is the FULL100% going to come from?
My last line above says it best..."One cannot have production from capacity that does not exist"

OP is looking for a NET 100% offset, not an instant 100% offset.

I conceded, you got me there, Butch

not sure of the situation around the change here. Were the PV modules installed just cheaper for the installer to get? why was the change made?

IDK

[ButchDeal]

SOLAREdge must be Magic stuff.
My sun goes behind clouds and some days I cannot even see the sun, Also I'm pretty sure the modeled clipping is just a sliver up on top. Am I to believe that that sliver can never sag during overcast skies? I imagine that with a little sag then 100% of the then REDUCED production is NOW all collected and that is no clipping in my book. Taking this to the next level, if there were never NO clipping and then a overcast sky or clouds come where is the FULL100% going to come from?
My last line above says it best..."One cannot have production from capacity that does not exist"

you are talking nonsense. You stated before :

there can be many days where when there can be zero clipping...then over paneled becomes a welcome feature.

WHY!? would over paneled be a welcome feature because of a FEW cloudy days. It is like saying you would rather have a sports car with a tiny engine because you would not need a large engine on days you don't drive...

Do you get that with a system that has a FLAT efficiency curve you don't get any MORE production using a smaller inverter even on cloudy days?

Now OP has more than a sliver of clipping, a lot more than a sliver of clipping.

[neweclipse]

Show me where all the additional production came from had this been 19kW inverter.

Clipping.png

[ButchDeal]

Show me where all the additional production came from had this been 19kW inverter.

I suggested OP would be far better off with SE11400H and SE6000H which would be 17.4kW (not 19kW) and both inverters capable of working on 3phase which the two installed are not (SE10000H and SE5000H)

a few things here.
First, OP has TWO inverters and one has a much higher ratio than the other. You are looking at a graph of both combined so the one is hidden.
Second, The production on this day would be exactly the same (except for the clipping) if the inverters were larger, so there is no gain with the smaller inverter, but there is a gain with the larger inverters.

So why do you claim that a smaller inverter is welcome on overcast days? it makes no since.

Third, right here, this is where the 5kW inverter is clipping on the day of the graph you showed...


And here is a point when the 10kW inverter is clipping: Layout2.jpg

[Ampster]

My takeaway from that picture and my own system is that the clipping is done at the Optimizers. If that is the case, then as long as the manufacturers recomended maximum AC to DC ratio is not exceeded, the Inverter is not being stressed. I understand the DC voltage is increased but from earlier posts it is only being increase by 20 to 50 volts. I presume that is the MPPT algorithm in operation to effectively doing the clipping.

[ButchDeal]

My takeaway from that picture and my own system is that the clipping is done at the Optimizers. If that is the case, then as long as the manufacturers recomended maximum AC to DC ratio is not exceeded, the Inverter is not being stressed. I understand the DC voltage is increased but from earlier posts it is only being increase by 20 to 50 volts. I presume that is the MPPT algorithm in operation to effectively doing the clipping.

Yes and no. The inverter ALSO does part of the clipping. The optimizers start kicking the voltage up higher so the inverter has to buck that down.

No it is not the MPPT algorithm though that kicks the voltage up. Each optimizers has a buck/boost so the DC line voltage is independent of the MPPT on the other side of the optimizers voltage regulators.

Worse though is that the optimizers on these cloudy days have a hard time resetting to a steady state. with every cloud that passes they go out of the high voltage range and and back in, where for several minutes they are fighting over which ones should be shedding load. Eventually they settle down. Normally this is just a small part of the time and even then not that bad but OPs set up is pretty extreme.

[neweclipse]

Butch, You keep saying how bad 5kW inverter is clipping, any chance you can post pic of this 5kW inverter in same format as I did for that same day as 2 combined?
It's sure hard to understand the severity of clipping on 5 kW since it sure doesn't drag down the combined graph I attached by much.

Is it possible maybe to move a few panels from 5kW over onto the 15 kW, I just don't see why all the boo hooing...just needs some tweaking.

[ButchDeal]

Butch, You keep saying how bad 5kW inverter is clipping, any chance you can post pic of this 5kW inverter in same format as I did for that same day as 2 combined?

You keep saying how good it is though...or "welcome" ?

I am NOT the OP and only have access to the same shared public reporting that you do. OP can graph it but you can clearly see that the 5kW inverter is maxed out in the photo I sent.
Further because they are so maxed out you can play through the day and see them flailing around trying to reduce the load, look at the first photo at the larger array. See how the modules are all over the map. That is not a cloud covering part of the array like that, that is them trying to come to consensus on which of them is going to shed load...

It's sure hard to understand the severity of clipping on 5 kW since it sure doesn't drag down the combined graph I attached by much.

OP has access to the charts section and can show (as he did for the one day earlier) a graph of the two individual inverters amp and DC voltage. The issue is very clear there.
If the arrays were shadowed or with two azimuths it would be a more practical arrangement with less clipping and more production but with everything south facing with no shadows, the clipping is pretty drastic. There appears to be no limitation from the PoCo to the 15kW size, and thus no reason at all not to go with slightly larger inverters. There would be NO loss in production with larger inverters, only gain. of course the limiting factor is that this is what was installed, appears to be what was contracted as well, so OP getting installer to make a change is going to be difficult.
Possibly get something in writing from installer that should anything fail they will replace with larger inverters.

Is it possible maybe to move a few panels from 5kW over onto the 15 kW, I just don't see why all the boo hooing...just needs some tweaking.

There are two inverters a 10kW and a 5kW. you can move some from the 5kW to the 10kW ( on a different building) but that will just make the 10kW have more of a problem.

[neweclipse]

Okay, tell me why not just replace the 5kW with a 10kW and everyone walks away grinning...

[Ampster]

Okay, tell me why not just replace the 5kW with a 10kW and everyone walks away grinning...

I think you are talking for the sake of talking. If you want to discuss high DC to AC ratios, start another thread. It is a topic that interests me but I dont want to hijack this thread when the issue here is not just about clipping.

[LCF]

I agree guys. I would like to replace the 5kw with a 10 based on everything Butch has said. The installer told us it is not possible and that our power company National Grid will not let them. They are saying the power company is capping us based on prior 3 years energy usage. We don't believe this and are trying to find out more details.

[Ampster]

And get one that is 3 phase capable. Would a reasonable workaround for the power company be to limit that one to self consumption. Or maybe you would only need a building permit if that one was limited to self consumption. I dont know the timing of your loads or configuration of your service panels to know if that would be optimal. I am just trying to think long term and out of the box and address you long term goals. However, it may affect your rebates.

[neweclipse]

I think you are talking for the sake of talking. If you want to discuss high DC to AC ratios, start another thread. It is a topic that interests me but I dont want to hijack this thread when the issue here is not just about clipping.

Understand and will cease that discussion point. Just trying to understand the hurdles that exist and get to the lowest common denominator towards a solution. Sorry for the hijack...

[J.P.M.]

I agree guys. I would like to replace the 5kw with a 10 based on everything Butch has said. The installer told us it is not possible and that our power company National Grid will not let them. They are saying the power company is capping us based on prior 3 years energy usage. We don't believe this and are trying to find out more details.

Have you checked the POCO website for particulars ? I saw max. size limits on the Nat. Grid site, but I couldn't find anything limiting array size as f(annual usage).

[ButchDeal]

I agree guys. I would like to replace the 5kw with a 10 based on everything Butch has said. The installer told us it is not possible and that our power company National Grid will not let them. They are saying the power company is capping us based on prior 3 years energy usage. We don't believe this and are trying to find out more details.

Well since they are on different buildings. The cheapest solution is to purchase an SE11400H. Replace the current SE10000H with the new SE11400H and replace the SE5000H with the SE10000H. The SE10000H is overkill for that array but will not hurt anything.

I would also like to see the stringing of these as currently the application does not have any string information.

It also sounds like your past usage was high enough that this shouldn't be limiting you. Usually if your future usage is changing for some reason (new purchase or change in service) you just have to send in some documentation to indicate said change.