Inverter size when maximum system is 11.56 kW (34 panels)

[preilly44]

If money is your limiting resource, then that's not an efficient choice. Go with the smaller inverters.

If roof space is your limiting resource, and money is not an issue, then going for more inverters (lower DC/AC) is a good choice.

If roof space is not your limiting resource, and you want more energy produced than you'll get from a 35 kW DC / 26.6 kW AC system, it's not clear whether you are better off adding more panels, or bumping up the inverter size. Probably more panels, but a more detailed analysis is required.

Cheers, Wayne

Thanks for your help Wayne (and others). My roof is maxed out with panels, that is my limiting factor. I've decided not to push for 4 7.6 and 1 3.6 but I think in the long run I'll be happier if I do go with four 7.6 instead of three 7.6 and one 3.8....even though I recognize it won't make a huge difference. If Tesla doesn't want to do that then I'll go with their recommendation. Thanks for helping!

 

[DrSmile]

It depends on your panels and your layout. I have 100% unobstructed south facing panels and my 13.2kW system is clipping significantly (for over 3 hours a day) even now in February with a 11.4 kW Solaredge inverter. If you have shaded or multiple panel orientations then your system may not clip at all.

 

[SoCalSolarRoof]

It depends on your panels and your layout. I have 100% unobstructed south facing panels and my 13.2kW system is clipping significantly (for over 3 hours a day) even now in February with a 11.4 kW Solaredge inverter. If you have shaded or multiple panel orientations then your system may not clip at all.

I have a 7.68kW solar roof with a delta M8 8kW inverter.

Could someone explain DC to AC ratio to me? Why would a system clip? I'm not sure what my ratio is, and if it matters in my case.

 

[jrweiss98020]

DC to AC ratio is the maximum AC power produced by the inverter, compared with the DC power produced by the solar panels. Clipping is when the inverter cannot convert all the power produced by the panels.

A 340 W panel only produces 340 W under ideal conditions. Up here in the Seattle area that may happen for ~2 hours/day for 4 months of the year, IF there are no clouds. So, it is not necessary to have a 10 kW inverter for 9 kW of solar panels, since they will seldom produce more than 8 kW.

 

[DrSmile]

I have a 7.68kW solar roof with a delta M8 8kW inverter.

Could someone explain DC to AC ratio to me? Why would a system clip? I'm not sure what my ratio is, and if it matters in my case.

Your DC to AC ratio is less than 1 so your system will never clip. Although your inverter may not be operating at top efficency with the wattage your panels supply the difference is likely negligible. The advantage for you is that you can add more panels without needing a new inverter.

 

[Ampster]

DC to AC ratio is the maximum AC power produced by the inverter, compared with the DC power produced by the solar panels.

That is it in a nutshell. The numbers used for that are the STC panel rating (sum of all panels) divided by the inverter capacity.

 

[Southpasfan]

So I have a 16.32 system. Two 7.6 SE inverters, equally balanced. Today, due to cloud edging, I see the system went all the way up to spikes of 15.6 kw.

I didn't think that was possible. Especially in March.

I thought that 16.32 would never really produce more than 15, so 15.2 of inverters was fine.

Is it just a spike? Is it because today the panels are also cool?

Or is this system going to clip like crazy in the summer?

It can't technically go over 16.32, correct?

 

[nwdiver]

It can't technically go over 16.32, correct?

16.32 is probably 'STC' Standard Test Conditions which is 1kW/m^2 @20C. If sunlight is >1kW/m^2 (such as with cloudedge effect) @20C you'll be >16.32kW. If sunlight is ~1kW and it's 0C... you'll be >16.32kW. Lot's of things can cause your panels to exceed their STC but it's rare.

 

[Southpasfan]

16.32 is probably 'STC' Standard Test Conditions which is 1kW/m^2 @20C. If sunlight is >1kW/m^2 (such as with cloudedge effect) @20C you'll be >16.32kW. If sunlight is ~1kW and it's 0C... you'll be >16.32kW. Lot's of things can cause your panels to exceed their STC but it's rare.

Thanks. That's not only an answer, but a great technical answer well explained.

What is interesting is that today, for a few minutes, reported energy exceeded the inverter ratings. Another thing I was not sure of.

 

[Southpasfan]

16.32 is probably 'STC' Standard Test Conditions which is 1kW/m^2 @20C. If sunlight is >1kW/m^2 (such as with cloudedge effect) @20C you'll be >16.32kW. If sunlight is ~1kW and it's 0C... you'll be >16.32kW. Lot's of things can cause your panels to exceed their STC but it's rare.

Thanks. That's not only an answer, but a great technical answer well explained.

What is interesting is that today, for a few minutes, reported energy exceeded the inverter ratings. Another thing I was not sure of.

 

[nwdiver]

What is interesting is that today, for a few minutes, reported energy exceeded the inverter ratings. Another thing I was not sure of.

That generally shouldn't happen but the inverters are usually capped more by current than power. So a 7.68kW inverter is 7.68kW at 32A and 240v (32A)(240v) = 7.68kW; But if your line voltage is 250v some inverters can take advantage of that (32A)(250v) = 8kW.

 

[DrSmile]

That generally shouldn't happen but the inverters are usually capped more by current than power. So a 7.68kW inverter is 7.68kW at 32A and 240v (32A)(240v) = 7.68kW; But if your line voltage is 250v some inverters can take advantage of that (32A)(250v) = 8kW.

This. I've found that line voltage is the most important factor when calculating max output. My line voltage sags on Summer afternoons because I'm one of the last houses on the line and I can watch my solar output drop as the voltage drops below 240 volts. If it gets low enough the entire system trips off. Also this can be a safety hazard because the amps typically need to increase as the voltage drops and this can cause some serious wiring / fuse issues.

 

[bevo]

My thoughts on inverter sizes.
Summary - if your panels are predominantly S, SW, SE facing panels, get the largest size inverter you can to match your total panel wattage.

As others have noted here, you will be clipping if you oversize your panels. I have the standard 8.16kW system with 7.6kW SE inverter. It's March and I'm already clipping. That's OK as I'm only 7% oversized and there's no 8.16kW inverter anyways.
Additionally, while pvwatts is good to get estimates, I find it inaccurate for my real world observations in this instance. Even if I enter in my system at 180 Azimuth, it doesn't show much clipping. Perhaps there are other parameters which need to be entered more accurately. Not sure.

The above article also states that the smaller inverters are more efficient at lower wattage. Maybe so, but being 10% more efficient (the graph shows 85% efficiency perhaps vs a 95% efficiency as an example) on a small number won't make up for the clipping of 1-2K on the upper range (say a 10k inverter with a 12k system).
I think it's disadvantageous to oversize. If I'm designing a new system with Tesla and they tell me it's OK to oversize by 20%, I would definitely try to up that inverter size, or get multiple inverters to better match the system. Obviously if you have North facing panels, then some oversizing is warranted. My $0.02.

 

[nwdiver]

My thoughts on inverter sizes.
Summary - if your panels are predominantly S, SW, SE facing panels, get the largest size inverter you can to match your total panel wattage.

As others have noted here, you will be clipping if you oversize your panels. I have the standard 8.16kW system with 7.6kW SE inverter. It's March and I'm already clipping. That's OK as I'm only 7% oversized and there's no 8.16kW inverter anyways.
Additionally, while pvwatts is good to get estimates, I find it inaccurate for my real world observations in this instance. Even if I enter in my system at 180 Azimuth, it doesn't show much clipping. Perhaps there are other parameters which need to be entered more accurately. Not sure.

The above article also states that the smaller inverters are more efficient at lower wattage. Maybe so, but being 10% more efficient (the graph shows 85% efficiency perhaps vs a 95% efficiency as an example) on a small number won't make up for the clipping of 1-2K on the upper range (say a 10k inverter with a 12k system).
I think it's disadvantageous to oversize. If I'm designing a new system with Tesla and they tell me it's OK to oversize by 20%, I would definitely try to up that inverter size, or get multiple inverters to better match the system. Obviously if you have North facing panels, then some oversizing is warranted. My $0.02.

Shouldn't it just be cost-benefit? If investing $500 in more panels yields another 730kWh/yr but investing $500 in a bigger inverter yields 300kWh/yr.... seems like more panels make more sense because numbers.

PLUS... those additional 300kWh are the least valuable. If you're capturing otherwise clipped kWh that's energy being produced when there's already a lot of solar energy being produced.

 

[h2ofun]

Shouldn't it just be cost-benefit? If investing $500 in more panels yields another 730kWh/yr but investing $500 in a bigger inverter yields 300kWh/yr.... seems like more panels make more sense because numbers.

PLUS... those additional 300kWh are the least valuable. If you're capturing otherwise clipped kWh that's energy being produced when there's already a lot of solar energy being produced.

So I assume with more panels one widens the energy curve, which is better than worrying about a little clipping?

 

[nwdiver]

So I assume with more panels one widens the energy curve, which is better than worrying about a little clipping?

Yes. Energy produced at 7pm is far more valuable than energy lost between 11am and 3pm on a day sunny enough to saturate the inverter.

Here's some real world data. I normally would have used a 5kW inverter but this customer wanted his system expanded after he bought an EV so he ended up with ~5.7kW on a 3.8kW inverter. Oversized by ~50%. The inverter he has cuts back String A when the inverter is saturated. String B never gets cut back so you can see how much production was lost last year due to inverter saturation. It's only ~4%. And this is being oversized by 50%. The last 2 panels on the end of each string outperform the others because they're 310s and the others are 295s.

 

[h2ofun]

Yes. Energy produced at 7pm is far more valuable than energy lost between 11am and 3pm on a day sunny enough to saturate the inverter.

Here's some real world data. I normally would have used a 5kW inverter but this customer wanted his system expanded after he bought an EV so he ended up with ~5.7kW on a 3.8kW inverter. Oversized by ~50%. The inverter he has cuts back String A when the inverter is saturated. String B never gets cut back so you can see how much production was lost last year due to inverter saturation. It's only ~4%. And this is being oversized by 50%. The last 2 panels on the end of each string outperform the others because they're 310s and the others are 295s.View attachment 643875

I still assume there is nothing wrong with having a larger inverter? In my case, it basically was free to use a 11.4K SE like my first. Now, if I had to pay for the larger, .. But the positive is I am looking to add maybe 5K and the inverter is already sized to handle without having to buy a new one.

 

[nwdiver]

I still assume there is nothing wrong with having a larger inverter?

Not if you don't mind spending money and not getting much for it... it's just a question of economics.

 

[Two-rocks]

I still assume there is nothing wrong with having a larger inverter? In my case, it basically was free to use a 11.4K SE like my first. Now, if I had to pay for the larger, ..

Wrong, no - but you may be loosing a bit of efficiency as the inverter spends less time at it most efficient operating range (maximum output).

Kind of like you car has charging overhead (loss), so charging at higher rates is more efficient use of the energy.

 

[h2ofun]

Wrong, no - but you may be loosing a bit of efficiency as the inverter spends less time at it most efficient operating range (maximum output).

Kind of like you car has charging overhead (loss), so charging at higher rates is more efficient use of the energy.

If one knew they were never going to upgrade their solar, either more panels or larger, I agree. But in my case, I am looking at adding a lot more. If I had just put in inverter to the size I have now, I would have been forced to buy new ones.