Clipping- What would my number have been?

[DavidT]

I’m curious what you all think my numbers would have been if I had 250 watt micro-inverters. You can see from my graph below that the other day was dang near a perfect solar day for me. Obviously plenty of clipping mid-day. I have 33 mono 260 watt panels. So my DC max is 8.58 kW. My system uses Enphase M215 micro-inverters, so my max theoretical AC @ 105% is about 7.45 kW. My actual max is about 7.41 kW before I clip. In the graph, if we drew an imaginary arch to complete the line that’s above the clipping line, it looks like it would peak around my 8.58 kW DC max. So I’m wondering what my max AC output would have been if my micro-inverters weren’t clipping? What is the percentage loss when converting from DC to AC? I know that’s only one of the components that make up a derate number, but isn’t it a big component? I wouldn’t think my AC max could be anywhere near 8.58.

I know it’s best to over-size the panels for the micro-inverters, so I’m not suggesting Enphase M250s would have been a better choice, I don’t think they would have been. But I’m curious as to just how much I’m losing to clipping. I don’t think it’s as bad as the graph would suggest. Thoughts?


Output - 031314.jpg

[Naptown]

Since I live in Maryland I can tell you that yesterday was butt cold with a strong wind blowing. This has a lot to do with what you saw as you were probably getting an extra bunch of watts out of each panel. This doesn't happen that often. but watch your graph. This should not be a regular occurance.

[Ian S]

I’m curious what you all think my numbers would have been if I had 250 watt micro-inverters. You can see from my graph below that the other day was dang near a perfect solar day for me. Obviously plenty of clipping mid-day. I have 33 mono 260 watt panels. So my DC max is 8.58 kW. My system uses Enphase M215 micro-inverters, so my max theoretical AC @ 105% is about 7.45 kW. My actual max is about 7.41 kW before I clip. In the graph, if we drew an imaginary arch to complete the line that’s above the clipping line, it looks like it would peak around my 8.58 kW DC max. So I’m wondering what my max AC output would have been if my micro-inverters weren’t clipping? What is the percentage loss when converting from DC to AC? I know that’s only one of the components that make up a derate number, but isn’t it a big component? I wouldn’t think my AC max could be anywhere near 8.58.

I know it’s best to over-size the panels for the micro-inverters, so I’m not suggesting Enphase M250s would have been a better choice, I don’t think they would have been. But I’m curious as to just how much I’m losing to clipping. I don’t think it’s as bad as the graph would suggest. Thoughts?


[ATTACH=CONFIG]3875[/ATTACH]

Just a rough eyeball of my output suggests that if I had clipping for the most productive four hours of the day like you did, it would amount to about 6-8% loss of output.

[DavidT]

Since I live in Maryland I can tell you that yesterday was butt cold with a strong wind blowing. This has a lot to do with what you saw as you were probably getting an extra bunch of watts out of each panel. This doesn't happen that often. but watch your graph. This should not be a regular occurance.

My system is only two months old so I don't have a lot of history yet. Clipping is not unusual for me lately, but I know March is the big month for it. But for a day like the day in my graph, I'm curious as to just how much potential production was lost due to clipping. If we filled in the bell curve above the clip line, that upper area would be the lost production. But it doesn't seem logical that that whole area could be actual production even if I had higher powered inverters. There's still DC to AC conversion... wire resistance... So I'm curious as to where my AC number would peak - all other factors remaining the same - if my inverters were not maxing out.

[DavidT]

Just a rough eyeball of my output suggests that if I had clipping for the most productive four hours of the day like you did, it would amount to about 6-8% loss of output.

But if you're not clipping then we can't really compare the two, correct? Unless I'm misunderstanding something. I would think that even if I wasn't clipping I wouldn't see that full bell curve in my production. If that was possible, wouldn't that imply my derate factor was practically zero?

[Ian S]

But if you're not clipping then we can't really compare the two, correct? Unless I'm misunderstanding something. I would think that even if I wasn't clipping I wouldn't see that full bell curve in my production. If that was possible, wouldn't that imply my derate factor was practically zero?

If you weren't limited by your microinverter capacity, you would see the entire Bell curve (not really a Bell curve but close enough). I looked at my data for a clear sunny day, calculated roughly what the area was above the line cutting through the curve parallel to the time axis with a line length of about four hours like yours. The percentage of my entire day's production represented by the area above that line was around 7%. I'm not sure why you bring derate into the discussion.

[DavidT]

If you weren't limited by your microinverter capacity, you would see the entire Bell curve (not really a Bell curve but close enough). I looked at my data for a clear sunny day, calculated roughly what the area was above the line cutting through the curve parallel to the time axis with a line length of about four hours like yours. The percentage of my entire day's production represented by the area above that line was around 7%. I'm not sure why you bring derate into the discussion.

Perhaps I'm not straight yet on all the solar terminologies and calculations. My interpretation of derate is that it's a group of factors that represent, as a percentage, the ratio of AC energy a system is expected to produce when compared to its nameplate DC number. Is that correct? In either case, if clipping wasn't a factor for me the other day and my output did reach the top of that imaginary line on my graph, that would suggest my 8.58 DC system was capable of producing about 8.58 of AC energy. That doesn't seem possible so what am I getting wrong? I would assume that even if everything else were perfect there would still be a percentage drop of about 5 points when converting from DC to AC.

[pleppik]

Enphase has a technical paper on the effect of clipping on production: http://enphase.com/global/files/Enph...ightsizing.pdf

This isn't precisely the question you're asking, but it seems interesting and relevant. Their bottom line is that a small amount of clipping is acceptable, and they conclude that an optimally-sized microinverter will occasionally clip because if you never get any clipping you probably spent too much money buying inverter capacity that you almost never use.

Of course, they're writing with an agenda (sell more microinveters), but it seems like a reasonable conclusion given that the inverter is a major part of the total system cost.

[DavidT]

Enphase has a technical paper on the effect of clipping on production: http://enphase.com/global/files/Enph...ightsizing.pdf

This isn't precisely the question you're asking, but it seems interesting and relevant. Their bottom line is that a small amount of clipping is acceptable, and they conclude that an optimally-sized microinverter will occasionally clip because if you never get any clipping you probably spent too much money buying inverter capacity that you almost never use.

Of course, they're writing with an agenda (sell more microinveters), but it seems like a reasonable conclusion given that the inverter is a major part of the total system cost.

I had read that Enphase report and found it quite interesting. I don’t think there would be a hidden agenda with Enphase though since they also have the M250 that they sell. If the M250s each cost $25 more than the M215s, then my math shows that I’d have to be clipping more than 480 hours a year on avg, over the next 20 years before I could have justified stepping up to M250s with my system. That’s being very conservative too since a lot of times while I’m clipping I wouldn’t be 1 whole kW above the clip line. To be even more conservative, the M250s require more daylight to stay awake, so they wake up later and shut down earlier. Maybe not a tremendous difference, but nevertheless that would factor in every day, not just days where there’s clipping.

[Ian S]

Perhaps I'm not straight yet on all the solar terminologies and calculations. My interpretation of derate is that it's a group of factors that represent, as a percentage, the ratio of AC energy a system is expected to produce when compared to its nameplate DC number. Is that correct? In either case, if clipping wasn't a factor for me the other day and my output did reach the top of that imaginary line on my graph, that would suggest my 8.58 DC system was capable of producing about 8.58 of AC energy. That doesn't seem possible so what am I getting wrong? I would assume that even if everything else were perfect there would still be a percentage drop of about 5 points when converting from DC to AC.

You must remember that your nameplate rating is obtained at a particular temperature (probably 20C but I don't recall for sure.) If your conditions are significantly colder than that - and Naptown says they were, you can exceed your nameplate number. I know I have. It may also be that you got an especially good lot of panels.

[DavidT]

You must remember that your nameplate rating is obtained at a particular temperature (probably 20C but I don't recall for sure.) If your conditions are significantly colder than that - and Naptown says they were, you can exceed your nameplate number. I know I have. It may also be that you got an especially good lot of panels.

That's a very good point and perhaps that does explain it. On that particular day it wasn't extremely cold... just a bit below freezing. But it was very windy, so it felt bone chilling. Does wind chill also impact panel performance?

[russ]

Does wind chill also impact panel performance?

Not at all

[Ian S]

Does wind chill also impact panel performance?

Wind can affect panel performance to the extent that it increases air flow past the panels and removes heat from them. This can keep panel temperatures lower than they otherwise would be in the absence of wind and thus enhance the solar output. Between the wind and the the sub-freezing temperature, your panels could have been 30C lower than they would have been under nameplate test conditions. Based on a typical temperature coefficient of 0.4 to .5%/degC, your output could have been enhanced by 12-15%.

[russ]

Wind can affect panel performance to the extent that it increases air flow past the panels and removes heat from them. This can keep panel temperatures lower than they otherwise would be in the absence of wind and thus enhance the solar output. Between the wind and the the sub-freezing temperature, your panels could have been 30C lower than they would have been under nameplate test conditions. Based on a typical temperature coefficient of 0.4 to .5%/degC, your output could have been enhanced by 12-15%.

New science from Ian? Wind chill does not affect a solar panel. Cool temperatures OK but not wind chill.

[DavidT]

New science from Ian? Wind chill does not affect a solar panel. Cool temperatures OK but not wind chill.

I get the point of what Ian was saying. He didn't say "wind chill", he said "wind".