Question about inverters and clipping

[SolPowered]

I must be misunderstanding something about the DC to AC conversion going, or how much a set of panels can actually produce compared to its given rating.

I'm going to have a set of active solar tiles at roughly a 160 degree azimuth (SSE), and I'm not sure about the roof pitch, but I'll call it "average". The set of tiles is rated at 5.6 kW. The weather data source long/lat for PVWatts is 41.73, -72.54.

They're planning on using a M4-TL-US for that set of tiles (datasheet at https://filecenter.deltaww.com/Products/Download/05/0505/M series datasheet rev13-20190703.pdf). When I inquired about clipping, and moving to an M6-TL-US, the project advisor checked with engineering and got back saying they confirmed there wouldn't be any clipping on the M4 so they're going to stick with that.

The M# units from Delta has a maximum DC/AC ratio of 1.3. And its namesake 4000W is maximum for the inverter - its nominal is 3840W. That's a DC/AC ratio of 1.46. Now I understand the vast majority of the time, it's not going to be maxed out. But in the summer months when the solar radiation is near the rest of the country, it seems like the set of tiles would be producing well over the limit of the inverter to convert.

What am I missing here that would mean there'd be no clipping?

 

[CrazyRabbit]

They are correct, 5925 for both. I ran at 20 degrees pitch, 150 degree azimuth, and just changed the ac/dc ratio.. your south of the equator with panels tilted away from sun.

7367 if facing north...

 

[SolPowered]

They are correct, 5925 for both. I ran at 20 degrees pitch, 150 degree azimuth, and just changed the ac/dc ratio.. your south of the equator with panels tilted away from sun.

7367 if facing north...

Sorry, I had that backwards - Lat, Lon: 41.73, -72.54

But I also don't understand how yearly production would factor in to the electronics aspect of an inverter's clipping a set of panels at somewhere nearly 5.6kW, when that inverter's limited to 3.8KW. Does the yearly production near 7,055 kWh/Year mean it wouldn't clip somehow?

 

[CrazyRabbit]

you can put in the ac/dc ratio, 1 would be a perfectly matched inverter and panels. if the yearly production total is the same for 1.0 and 1.4 then no clipping.

5,056 vs 5059. the smaller inverter will will gain you 3 kw for the whole year!

 

[BrettS]

Sorry, I had that backwards - Lat, Lon: 41.73, -72.54

But I also don't understand how yearly production would factor in to the electronics aspect of an inverter's clipping a set of panels at somewhere nearly 5.6kW, when that inverter's limited to 3.8KW. Does the yearly production near 7,055 kWh/Year mean it wouldn't clip somehow?

Clipping has been covered here a number of times, but the short version is this:

Even though your panels are rated at 5.6kW, for the vast majority of the time your panels will be producing much less than 5.6kW. Solar production is a bell curve that ramps up in the morning, peaks in the afternoon, then ramps back down in the evening. So even during the brightest, sunniest day your panels will only be outputting the maximum amount for a short time during the day. And then when you factor in things like panel orientation, geographical location, and weather that further reduces the amount of time that the panels will be producing even close to their maximum output. If you have panels facing different directions that can also affect it as each bank of panels will peak at a different time, so you will never have all of your panels producing their maximum output at the same time.

So as @CrazyRabbit said, the best way to figure out if this could be a problem for you is to use a calculator like PVWatts. If you run it once at a DC to AC ratio of 1.0 (5.6kW panels and 5.6kW inverter) and once at a DC to AC ratio of 1.4 (5.6kW panels and 4kW inverter) and compare the annual production numbers they will likely be very close, probably less than a 1 or 2% difference. However, if there is a significant difference then that would indicate that it would be beneficial to move to the bigger inverter.

 

[SolPowered]

Thank you both for that explanation, and yes, changing the inverter ratio there does correspond with what you say. I guess I'm just surprised by how low the panels max out compared to their rating. I would have hoped they'd be near max for more of the summer, but I'm just not in the right area for that.

 

[wjgjr]

For what it's worth, we have a ~8.2 kW solar glass roof split half on a south (187 degree) and half on a north-facing roof. It seems to max out at 5.4 kW of instantaneous power (occasional spikes higher due to cloud edge effect are ignored.) I, too, initially was concerned that the numbers seemed low (even knowing the north-facing roof would have lower production.) But the thing to focus more on is the total production over the course of at least a month, in order to help even out weather issues.

Look at the annual numbers Tesla provided, and the PVWatts numbers (which for most people are higher.) Because PVWatts gives monthly numbers, you can use that to calculate what the monthly numbers for Tesla would be. (For example, if Tesla says 5,000 kWh for the year and PVWatts says 6,000, then multiply each PVWatts monthly number by 5/6 to get Tesla's estimate.)

For us, August was the first full calendar month of post-PTO operation. PVWatts provides an estimate of 874 kWh*, while Tesla would imply 808 kWh. Actual production for the month was 863 kWh, so that gives me some comfort that the system is operating normally and we are getting what we paid for from Tesla, which is the annual number they put in the contract.

* Since we have a small system with SRECTrade, they actually calculate this for us using PVWatts, so I plan to use their numbers as they likely best know how to correctly input all the data. I got similar, though not identical, numbers.

 

[SolPowered]

I think what was really throwing me that was Tesla said there'd be no clipping, and while it appears that it's really minimal, there would be some. I just couldn't believe that 5.6kW panels would never really produce more than 3.8KW. Which is still disappointing how rarely it gets near that number.

I do now understand how you can use PVWatts to change that AC/DC ratio and see that yearly production varies by about 30 kW by changing those values from 1.46 to 1.

I guess my only other thought is the M6 allows there to be three strings, where the M4 has two, which could help optimize things a little bit better. But since they've already decided that clipping isn't an issue, and they'd submitted the M4 to the power company, they're quite resistant to switching. And I don't want anything else to slow this down so that the project has half a chance of getting down this year.

 

[CrazyRabbit]

with the m6, you would be running at 2/3 capacity, the inverter efficiency would be lower and wash out the 30KWh.

for fort worth texas, in the summer east, west and south panels have the same production numbers. but in winter the east and west panel are about half the south panels. and a 10/12 pitch will give you more level output by 7% to 8/12 pitch panels

 

[BrettS]

If it helps, look at this graph of the solar production from my system the other day:

My system is a 15.12kW system. Not only did it max out at about 10 or 10.5 kW for the day, but it was only at that max production for maybe two hours. Even if I had 9kW inverters, which would be way undersized for my system and they did clip, the power that would have been ‘lost’ over two hours that they were clipping would be about 2kWh or just over 2% of my daily production. By the time that you factor in that in the winter when the sun is at a less optimal angle it might not go over 9kW at all and even in the summer when it could get over 9kW that there will be some days with rain and clouds so even in the summer not every day would have clipping you can see that the impact on annual production will really be minimal.

 

[Doom Man]

A lower-rated inverter will be more efficient than a higher-rated inverter. So you might get some clipping with the smaller inverter, but you can more than make up for that with increased efficiency in converting to A/C.

 

[charlesj]

.... I guess I'm just surprised by how low the panels max out compared to their rating. ...

Consider the panel rating in light of a lab measurement used for rating at exactly 90 degrees on the panel with the max light on the whole surface. And, probably at the best panel temperature since temp has a good impact on production.
Then, you go into the real world with a fixed pitched roof that I would speculate may be only a few days or a week when the sun is in that perfect position, then the real temperature up there.

I have 215W micro inverters. It clips at about 226W. I had 1 such day in the 1st year on line, all 17 panels at 226W. Panels rated at 235W.
Or, at least that is what inverter company said about the inverter but still recommends up to 270W panels with that inverter.
To me, that would mean that while the inverter would clip, it would ramp up faster and down later in the afternoon maintaining max power for much longer.

 

[jrweiss98020]

A lower-rated inverter will be more efficient than a higher-rated inverter. So you might get some clipping with the smaller inverter, but you can more than make up for that with increased efficiency in converting to A/C.

Where did you find that information?

 

[BrettS]

Where did you find that information?

I’ve seen people mention that inverters are typically more efficient at higher loads, so I don’t think that it’s really that a lower rated inverter is more efficient overall than a higher rated inverter, but rather that with the same DC system size the lower rated inverter would be running closer to its max capacity more of the time and as such be more efficient in this situation than a higher rated inverter.

 

[Doom Man]

I’ve seen people mention that inverters are typically more efficient at higher loads, so I don’t think that it’s really that a lower rated inverter is more efficient overall than a higher rated inverter, but rather that with the same DC system size the lower rated inverter would be running closer to its max capacity more of the time and as such be more efficient in this situation than a higher rated inverter.

That's a more correct way to say it than what I said above. It was mentioned in a few articles that more or less said "don't worry about a little clipping; it's good for you in the big picture." It's been over a year, so I can't recall the sources. Another factor was that the extra cost of a higher-rated inverter did not justify the marginally higher production -- payback periods were over 30 years, I think.

 

[holeydonut]

That's a more correct way to say it than what I said above. It was mentioned in a few articles that more or less said "don't worry about a little clipping; it's good for you in the big picture." It's been over a year, so I can't recall the sources. Another factor was that the extra cost of a higher-rated inverter did not justify the marginally higher production -- payback periods were over 30 years, I think.

Enphase has a white paper on this topic. They recommend the array slightly oversized to the inverter as well. And they’d actually benefit from selling bigger inverters, so it’s interesting they’d even publish this.

https://enphase.com/sites/default/files/downloads/support/Enphase-Tech-Brief-Why-Larger-EN-US.pdf

 

[charlesj]

Enphase has a white paper on this topic. They recommend the array slightly oversized to the inverter as well. And they’d actually benefit from selling bigger inverters, so it’s interesting they’d even publish this.

https://enphase.com/sites/default/files/downloads/support/Enphase-Tech-Brief-Why-Larger-EN-US.pdf

Beat me to the punch. The 215IQ7 215W, can take up to a 270W panel. This would allow it to generate max out of 225W for a much longer time period while the system has 17 panels on a breaker.
It may not make sense on a new install but an upgraded older system like I have, I thought about doing this with a few panels to see how well it works but cannot find a black panel that is thick as the ones I have, 1.8".