Microinverters, DC to AC ratio and Clipping oh my!

[gigabit]

I would like to share my findings on a recent installation with microinverters. I had 21 LG 335W panels installed with IQ7 microinverters. On a sunny day the system would come up to about 5.14kW AC and stay there from 11:45am to 4:15PM. Makes sense the maximum output of the IQ7 is 250VA (assume a Power factor of 1 and you get 250W). This gave the system a DC to AC ratio of 1.34. Pretty high, but my initial understanding was clipping is good to an extent and it could possibly yield more power earlier and later in the day (Wider curve vs taller curve). Fast forward and the same system has IQ7+ which has a maximum continuous output of 290VA. This gives the system a DC to AC ratio of 1.15. I have found the ramp up output between the IQ7 and IQ7+ are near identical but one peaks at 5.14kW (IQ7) and the other 5.92 (IQ7+). Now the area under the curve between the two was much more on the IQ7+ yielding approximately 2.4kWh more over the entire day. From the simplistic modeling from PV watts the clipping was much less in PVwatts than what I was seeing with the IQ7 real world.

Now given this is in the summer time with no winter output data but this had me thinking, are microiverters to a degree exempt from the conventional DC to AC ratio criteria? Possibly the higher the ratio the more you will lose due to clipping since the microinverters have the same startup voltage and efficiency along the curve.

Maybe I am bringing out the jump to conclusion mat too soon but what do you think?

[J.P.M.]

Pretty much if a panel's output at any point in time is greater than the microinverter's max. rated output, the difference will be lost.

If the goal for an inverter is to utilize the max. amount of a panel's output, why would you think clipping that is caused by not doing that is a good thing ?

[JSchnee21]

??? So I take it you were unhappy with the IQ7 output so you replaced them with the IQ7+ Did I read that right? Was there a cost associated with this upgrade?

Clipping, in general, is never "good" and will always lead to some loss of production. But, that said, a little clipping is not always bad. The question is which solution is more cost effective in the long run -- MWh produced / total system cost. Presumably the IQ7+ Micros cost more.

It is true that panels rarely produce at their STC rated wattage. So, in general, there is usually no harm and little (if any) production loss from minimal to moderate over-subscribing. BUT, if your IQ7's were clipping for 5 hours per day, surely you'll have some loss compared to the IQ7+, a properly sized Solar Edge solution, or more, smaller panels (assuming they fit on the roof).

Putting more DC STC Wattage on the roof (by using higher wattage panels and over-subscribing the Micros more) will broaden your curve and increase production all year. But once again its a cost / benefit question. In the end, you want the most MWh produced for the absolute lease cost.

So while buying bigger Micros or more panels will improve total production, the cost effectiveness of said production is reduced with each upgrade.

I don't think PVWatts is able to model Clipping at all. Last I checked it didn't understand the differences in harvest capabilities of different inverter / micro solutions. Rather it only models the average production based on PV size (DC STC), "system losses", weather, temp, latitude, orientation, inclination, etc. I think the expectation is that production x (1-System Losses) = Amount Harvested. Where "System Losses" is essentially wiring and conversion inefficiencies.

[ButchDeal]

I would like to share my findings on a recent installation with microinverters. I had 21 LG 335W panels installed with IQ7 microinverters. On a sunny day the system would come up to about 5.14kW AC and stay there from 11:45am to 4:15PM. Makes sense the maximum output of the IQ7 is 250VA (assume a Power factor of 1 and you get 250W).

250 is PEAK, continuous is 240W

This gave the system a DC to AC ratio of 1.34. Pretty high, but my initial understanding was clipping is good to an extent and it could possibly yield more power earlier and later in the day (Wider curve vs taller curve).

based on short peak value but should be constant value which is lower thus worse ratio of 1.4
clipping rarely results in more power and almost always results in less overall generation.

Fast forward and the same system has IQ7+ which has a maximum continuous output of 290VA. This gives the system a DC to AC ratio of 1.15. I have found the ramp up output between the IQ7 and IQ7+ are near identical but one peaks at 5.14kW (IQ7) and the other 5.92 (IQ7+). Now the area under the curve between the two was much more on the IQ7+ yielding approximately 2.4kWh more over the entire day.

as expected.

From the simplistic modeling from PV watts the clipping was much less in PVwatts than what I was seeing with the IQ7 real world.

Model is only as good as the set up.

Now given this is in the summer time with no winter output data but this had me thinking, are microiverters to a degree exempt from the conventional DC to AC ratio criteria?

No. But over higher ratios are just old rules of thumb based on OLD design criteria for older inverters and pricing of older inverters.
There are some reasons to do it but not as a general rule of thumb which is just lazy.

Possibly the higher the ratio the more you will lose due to clipping since the microinverters have the same startup voltage and efficiency along the curve.

Micro-inverters do not share inverter capacity like strings is more of the answer here.
Micro inverter manufactures push the idea that clipping is good when they do not have a product that doesn't clip and give results based on keeping the inverter loaded, which makes little since, you want to harvest the most, no one cares about keeping an inverter maxed out.

[J.P.M.]

???
I don't think PVWatts is able to model Clipping at all. Last I checked it didn't understand the differences in harvest capabilities of different inverter / micro solutions. Rather it only models the average production based on PV size (DC STC), "system losses", weather, temp, latitude, orientation, inclination, etc. I think the expectation is that production x (1-System Losses) = Amount Harvested. Where "System Losses" is essentially wiring and conversion inefficiencies.

On the clipping and PVWatts: A SWAG for a no, or not much shading situation:

If you take the hourly output option for the PVWatts model, add a column, and in that column subtract (hourly system output)/(#of panels) for each hour from the micros rated output, and then sum all the negative hourly results, and then multiply that result by the # of panels, you'll get a rough approx. of the annual system clipping.

[gigabit]

??? So I take it you were unhappy with the IQ7 output so you replaced them with the IQ7+ Did I read that right? Was there a cost associated with this upgrade?

Clipping, in general, is never "good" and will always lead to some loss of production. But, that said, a little clipping is not always bad. The question is which solution is more cost effective in the long run -- MWh produced / total system cost. Presumably the IQ7+ Micros cost more.

It is true that panels rarely produce at their STC rated wattage. So, in general, there is usually no harm and little (if any) production loss from minimal to moderate over-subscribing. BUT, if your IQ7's were clipping for 5 hours per day, surely you'll have some loss compared to the IQ7+, a properly sized Solar Edge solution, or more, smaller panels (assuming they fit on the roof).

Putting more DC STC Wattage on the roof (by using higher wattage panels and over-subscribing the Micros more) will broaden your curve and increase production all year. But once again its a cost / benefit question. In the end, you want the most MWh produced for the absolute lease cost.

So while buying bigger Micros or more panels will improve total production, the cost effectiveness of said production is reduced with each upgrade.

I don't think PVWatts is able to model Clipping at all. Last I checked it didn't understand the differences in harvest capabilities of different inverter / micro solutions. Rather it only models the average production based on PV size (DC STC), "system losses", weather, temp, latitude, orientation, inclination, etc. I think the expectation is that production x (1-System Losses) = Amount Harvested. Where "System Losses" is essentially wiring and conversion inefficiencies.

No difference in cost from me between the IQ7 and IQ7+. I think panels get closer to PTC and most definitely get above 280 watts continuous from 12pm to 4pm (Summer time). PVWatts "Models" clipping by reducing the output of the array at or near solar peak but it is much shorter duration that what I was seeing in real life (I understand its only a model so expected) PVwatts was showing with my inputs only 1 hour of clipping on the best of days (output hourly) by changing the "advanced options" DC to AC ratio. Real world I was a lot more clipping.

The most interesting thing I witnessed was not the clipping but the curve width being unchanged. From what I understand from Utility sized solar did not seem to apply to microinverters. The power produced in the morning and evening seemed unchanged from the IQ7 and IQ7+. It looks like the starting voltage on the micros are the same between the IQ7 and IQ7+ so im suprised Enphase has two models maybe they are two different bins of the microchip inside to handle the difference, im not sure. I was wondering if this relationship might change in winter time or there is no reason to not go with the IQ7 if your panel can output more.

[jflorey2]

Pretty high, but my initial understanding was clipping is good to an extent

It's not "good" in that you lose generation. There are two cases where it might have some benefits:

1) Panel limitations. In many cases you are limited to a 40 amp breaker, which means max output of 32 amps at 240 volts (7680 watts.) So when choosing between a 7700 watt inverter that will clip and a 10kW inverter that will not - your installation may be a lot cheaper with a 7700 watt inverter since you will not need a new panel or a line side tap or anything.

2) Inverter pricing. If the 7700 watt inverter is much cheaper than the 10kw inverter then it may make sense to stick with the 7700 watt inverter even if you will lose (say) 5% of your production on sunny/cool days.