Any smart guys here able to estimate what I'm loosing with Clipping

[RSQ]

I have been going back and forth with my installer, and wanted to get this addressed before any more time goes past, but I do understand I don't have a hot summers day for sample yet. Here are my past two days. I have a 11.05 KW system 34 REC 325NP and 34 IQ7's My question is how much power on these days am I loosing. Can this be calculated with math somewhow, So far my curve looks like this on all days that have been full sun. I understand in the summer when its HOT that my production may decrease but for the cost difference from the 7 to 7+'s if its only a year or two to return my investment I'm going to be upset if I'm not getting everything I can. Thanks in advance.

enphase1.pngenphase.png

[Ampster]

I had the same reaction you did when I first saw my system charts. My system has a DC to AC ratio of 1.5 to 1. Based on the capacity of the 325 watt panel and the capacity of the IQ7 at 240 continuous Watts, your ratio is 1.35 to 1.
What is important is the area under the curve, not the energy clipped. My system is performing well compared to a PVWatts estimate and is exceeding the installers garanty. There are a lot of factors that go into system design and there are constraints such as power company limitations and service panel limitations that may have driven the design. A DC to AC ratio of 1.35 is not unusual. Enphase has tables that you can look up and see if your installer matched those panels with the recomended inverter.

[Skiman26]

I have a similar problem my system is 10.4 kw and is making substantially more power than the installer figured it would. The system has the IQ7 micro inverter and is maxing out at 8.4 kWh. Please let me know what you think

[solardreamer]

I have been going back and forth with my installer, and wanted to get this addressed before any more time goes past, but I do understand I don't have a hot summers day for sample yet. Here are my past two days. I have a 11.05 KW system 34 REC 325NP and 34 IQ7's My question is how much power on these days am I loosing. Can this be calculated with math somewhow, So far my curve looks like this on all days that have been full sun. I understand in the summer when its HOT that my production may decrease but for the cost difference from the 7 to 7+'s if its only a year or two to return my investment I'm going to be upset if I'm not getting everything I can. Thanks in advance.

enphase1.pngenphase.png

iQ7's are maybe on the small size for your panels but still within normal range. If you really want to know you have to drill down to the array view in Enlighten and look at each microinverter's power graph and estimate the clipping loss for each then sum and average. I have done it for my system and found only ~5% average loss on maybe a dozen days in a year. Not really a concern for me. Also, remember you panels will degrade a bit every year.

[bob-n]

As others have said, you're doing OK. There is nothing wrong with the system design and you are getting plenty of power out of it. You could spend more money and get more power out of the same panels. But that doesn't mean that your system is wrong. It just means that you spent a specific amount and (we all hope you) got what you paid for.

Again, as others have said, figuring out how much more you would get with larger inverters is complicated. As a rough estimate, you can try to fill in the plot and see how much would be created if there were no clipping, with a sketch something like below, and then try to guestimate how much would be in the open area. Is it 15% more? 20% more?

Then you have to determine when you would get that extra energy. On cloudy days, iQ7 is more than adequate. Perhaps also for hot days, when the panels put out less, and for winter, when the sun is lower in the sky. That's what makes the math complicated. Also, as others have also said, you'll get less as the panels age.
missing.png
You asked for a break-even time for upgrading to iQ7+. I wish I were smart enough to do that math, but I can try a very rough estimate. To do it accurately, you'd first need to know how many days would be like those shown here. Let's take a wild guess and say 50 per year. So in one year, you're getting perhaps 20% more power for 50 days with the upgrade.

You're getting roughly $10.30 worth of power from 34 panels, or $0.30 per panel per day. If we can add 20% to that for 50 days, we will get $3.03 per year more per panel with iQ7+.

What is the cost to upgrade? Wholesale, they may only get charged $20 extra for an iQ7+. But they mark that up to you, so figure $40 each. Add to that the labor to swap, the loss of value for the iQ7 that they are removing, and then the labor to reconfigure your system. Let's wildly add another $40 for that. You could argue any of these figures, but we're just roughing it for now.

$80 investment paying back $3.03 per year is 26 years to break even. Ouch. Change the assumptions and do the math again, and I think you'll still have a hard time paying for the new inverters.

[J.P.M.]

Well, according to the Enphase spec sheets I'm looking at, the EQ7's max output is 250 VA. The EQ7+ has a max. output of 295 VA.

To a first approx., what that means with either inverter is that anytime a panel's output exceeds something like 250 W or 295 W for the EQ7 or EQ7+, respectively, they will clip that panel's output. That's what appears to be happening on your included graphs.

A reasonable quality panel that's ideally oriented for this time of year can be expected to produce VERY roughly about 85 to MAYBE 88-90 % of it's STC output, so that rough approx. would make a 325 STCW panel's max. output ~ 275 - 290 STC W.

A cursory looks shows maybe the EQ7+ might prevent all or most of the clipping.

Another question is: What's a rough guess as to how much clipping is occurring with each micro ?

The bigger question is: How does that penalize a typical year's output using each micro ?

The rough answer I use for that first question involves using PVWatts.

Do a run with proper inputs for your array and use a 10 % system loss factor and a DC/AC ratio of one.
Get the hourly output option.
On the same spreadsheet as the hourly output option, for each hour where the modeled output is greater than either (250 W/panel )*(34 panels) or (295W/panel)*(34 panels), depending on which micro your looking at, subtract (250*34) or (295*34) from the modeled for each hour.
Sum all the differences for each such hour.
That sum will be a pretty good 1st approx. of annual clipping loss as a result of using each micro. The difference in doing that for one micro or the other will be an indication of how much more clipping the EQ7 causes over the EQ7+.

As for what the clipped (lost)output is worth, with that difference and some idea of what your POCO's rates are by hour and plugging that hourly rate into the PVWatts that have the clipping by hour, you'll be able to get a SWAG on what the clipping for each micro might be costing you.

Looking at your included graphs, it's hard to get very precise, but it looks like clipping occurs at around 2.1 kW or so per 15 min. period or 8.4 kWh/hr.
Note, 34EQ7 micros *250 w/micro = 8.5 kW.

Pertinent to what your 2 clear day output may be showing you, once you do the PVWatts runs, if you look at the PVWatts hourly output for clear days within a few days of 3/19 and 3/20 and compare that to your actual output, you'll get a rough approx. of how much energy was clipped those days as a result of the microinverter output clipping. But I'd caution, it's only a rough approximation and won't apply to an entire year, primarily because most days are not entirely clear.

I'd also suggest to intimidate yourself by PVWatts. It's designed to be user friendly and it mostly is. Just read ALL the help screens a couple of times, including the part about DC/AC ratios, and use a system loss parameter of 10 % rather than the 14 % the model will default to.

About 20 minutes of quiet study of the help screens and then a few runs will probably result in enough fluency with the model to give you enough confidence to talk to your installer. Beyond that, you'll have a better understanding of what's happening with your system.

FWIW, using the model as described above is actually a bastardization of the model's intended purpose, but doing so seems to produce pretty good 1st approx. numbers when I've used it so.

[DanS26]

Clipping is NOT a bad idea.......here read this........

https://www.gses.com.au/wp-content/uploads/2016/03/GSES_Oversizing.pdf

[bob-n]

Dan: Thank you for the article. That is an excellent read, and clearly shows that an optimum system has some clipping.

In very broad terms, optimization is deciding how much to spend on panels & racking versus how much to spend on inverters. Is it a better investment to buy one more panel and one more small microinverter, or to buy larger microinverters?

Curiously, this is a moving target. Panel cost will drop with time, racking cost will go up with time, installation cost will go up with time, and inverter costs will go down with time. My crystal ball suggests that the optimum point will go towards larger inverters and less clipping in the future, but time will tell.

[J.P.M.]

Clipping is NOT a bad idea.......here read this........

https://www.gses.com.au/wp-content/u...Oversizing.pdf

Except that it has the likely potential to waste money and resources on too large an array which doesn't sound to me like a good way to start a design.

IMO only, there's a lot in that piece that is vague and not well explained, including graphs like fig. 5, and over simplified statements like most of the 2d paragraph. Lots of other stuff like the statement that array oversizing leads to higher yield under low light conditions. More honestly perhaps, that might better have been written as something like "a little more yield under very poor conditions with a lot of effort cost for a small reward." or such like. The whole piece just seems less than objective to me and full of half truths and arguably incorrect or at least misleading information and innuendo. Seems to me that's more of a backdoor deceptive pitch to sell more panel area than a discussion of DC/AC ratios.

Also, before I started down the array oversizing route, I'd be careful to define just what the term "oversizing" refers to. A 5.2 STC kW array driving a 5 kW inverter probably isn't oversized. If that 5.2 kW array drives a 4 kW inverter, the array may be considered oversized or, depending on the application, the inverter may just as well be considered undersized.

Seems to me the whole subject of inverter/array size ratios would be more productive if framed with the overall goal (after safety considerations) of best engineering while getting most bang for the buck. Such designs would have inverters sized to the array in terms of the cost of losses by undersizing inverters rather than any possible and harder to justify (and specious IMO) energy gains and cost savings from oversizing an array.

As in: Wait ! - I spend more time/$$ by making my array bigger while not changing inverters and so lose more of what I gained due to more to clipping. What's wrong with this picture ?

FWIW, I did this when designing my array: I sized the array for the most cost effective load fraction of my annual load, and then sized a "theoretical" inverter so that it equaled the max. array power output under modeled conditions. Then, I compared the cost of the next smaller and next larger actual available inverters, and compared the life cycle cost differential of the inverters to the life cycle cost of the lost energy by using the smaller inverter. It was close, so as a matter of macho, I chose the larger inverter. The actual array size is 5.232 STC kW. The max. modeled output was 4.863 kw to the inverter. The inverters were either 4 kW or 5 kW. The 4 kW inverter clipped ~ 150 - 250 kWh/yr. or ~ 200 kWh*$0.28/kWh ~ = $56/yr. The 12 yr. lifecycle cost differential was ~ $700 less tax credit. Also, and more FWIW, the max. actual (5 kW) array output under cloudless skies has been 4.89 kW. The max. measured output of the 5 kW inverter is 5.032 kW whenever cloud reflections boost array output for a few sec. to a minute or so. Max. power to the inverter under those conditions has been 5.44 kW making me think my panels are closer to 340 STCW output rather than the 327 STC W as shown on the nameplate or beyond the +5 % nameplate tolerance, but that's another topic.

If one design goal is the most cost effective system for a residential application, rather than worrying about or having array oversizing as a design goal (which I'd suggest may be analogous to having a very powerful and unnecessarily large engine in a vehicle and then setting a hard and permanent limit on output by throttling the engine output so the wheels never break traction when driving in slippery road conditions), what might be a better design approach would be to size the array to the most cost effective load fraction (which may well be less than 100% of annual load) with one of the (system) costing parameters being the cost of the clipping using a comparison of the cost of the lost energy vs. the cost of a larger inverter as one parameter.

[Ampster]

I agree that a high DC to AC ratio should not be a design goal. However based on the cost of panels there are situations where overpaneling makes sense. Our own @bcroe is one example and I may be another. This whole discussion reminds me of the classic question of whether the glass is half full or half empty? Instead of answering the question, some people will chose to say that the glass is the wrong size.

I am on a TOU with two EVs and leverage my rates to that at True Up I have consumed more than I generate but I still have a dollar credit. To me the extra generation I get from overpaneling helps me build that dollar credit. To be clear, I am not saying you are wrong, I just think everyone should hear all the arguments and make the decision that suits them. The OP asked for opinions. He was focusing on what he was losing and not what he was gaining. It all depends on where your are standing.

[J.P.M.]

I agree that a high DC to AC ratio should not be a design goal. However based on the cost of panels there are situations where overpaneling makes sense. Our own @bcroe is one example and I may be another. This whole discussion reminds me of the classic question of whether the glass is half full or half empty? Instead of answering the question, some people will chose to say that the glass is the wrong size.

I am on a TOU with two EVs and leverage my rates to that at True Up I have consumed more than I generate but I still have a dollar credit. To me the extra generation I get from overpaneling helps me build that dollar credit. To be clear, I am not saying you are wrong, I just think everyone should hear all the arguments and make the decision that suits them. The OP asked for opinions. He was focusing on what he was losing and not what he was gaining. It all depends on where your are standing.

I'd suggest that the DC/AC ratio need to be checked as part of the design process, but probably shouldn't drive most, or control any designs.

I agree there are situations where overpaneling may make some sense, but those usually apply to situations where POCO tariff structures or site conditions such as shading or other perhaps unusual conditions (part year occupancy being one example) get in the way of cost effectiveness.

If we're talking about the more/most common residential rooftop PV, I'd suggest that while overpaneling is an option, it's one that I'd only look at after I got close to a final design and then check whether or not, and if so under what operating conditions overpaneling made any economic, or practical sense or personal preference (with respect to convenience, maint. or ease of operation standpoints for example).

Off topic, but since you took it up, with respect to Bruce's arrays:

I've discussed at length why, from a cost effectiveness standpoint, I believe Bruce's experience is a good example of misguided design where there's confusion between one optimum orientation that has max. annual cost effectiveness per installed STC kW with respect to NEM operation and POCO rates, and multiple orientations that do little except spread the times of collection over a greater portion of the day and reduce the energy collection and bill offset per installed STC kW revenue over any 365 day period. That's not to say his design doesn't have a lot of good features and innovation built into it. FWIW, I couldn't do what he's done. Hat's off Bruce. I'd help him dig post holes and pull wire if I was nearby. But at the same time, I believe most of what he's done could have just as well been done on one optimally oriented and smaller array that collected as much energy over 365 days and cost less while generating more bill offset revenue (but probably wouldn't have been as much fun).

To that point, and as more of a thought experiment if you or others care to, run PVWatts with the idea of generating 10,000 kWh/yr. or so with a ground mount array in a shade free field. Now, take that array size, and split it in two and orientate the 2 equal half size arrays in any 2 other orientations, around due equator facing or not and try to find any combinations of orientations where those two arrays produce more annual output than the single orientation array. Or, 3 or more equally sized arrays in any orientation.

There may be sound and necessary reasons, including personal choice(s), for multiple array orientations for a PV system, but few, if any of them, have anything to do with maximizing annual energy production per installed STC kW or with cost effectiveness.

[Skiman26]

Is a 10.4kwh system that’s making 67kwh a day seem like more production than normal for a south facing home in the California desert I am being told it’s more than expected as a reason for the clipping on my system which seems to be occurring starting about 6 weeks or so ago and is lasting about 3 hours a day. I just want to know if this is an error or it’s normal. I. Reading the posted article it seems like it may be a good thing? I was also told that going to an IQ7+ inverter would mean that the system would start production a bit later in the morning ans shut down a bit earlier? Any truth to this, if so I don’t understand the reasoning.

[oregon_phil]

Post #6, JPM runs through using PVWatts. Give that a read and input your system variables into PVWatts.

Regarding the IQ7+ comment: Going to an IQ7+ inverter vs what other options?

[J.P.M.]

Is a 10.4kwh system that’s making 67kwh a day seem like more production than normal for a south facing home in the California desert I am being told it’s more than expected as a reason for the clipping on my system which seems to be occurring starting about 6 weeks or so ago and is lasting about 3 hours a day. I just want to know if this is an error or it’s normal. I. Reading the posted article it seems like it may be a good thing? I was also told that going to an IQ7+ inverter would mean that the system would start production a bit later in the morning ans shut down a bit earlier? Any truth to this, if so I don’t understand the reasoning.

That would be 67 kWh/day/10.4 STC kW = 6.44 kWh/day per STC kW. Sounds about right to me for the desert this time of year. I'm in zip 92026 and my modeling shows a clear day 31 day average for this date +/- 15 days time of year of 33.61 kWh/day for a 5.232 STC kW array. 33.61/5.232 = 6.42 kWh/day per STC kW.

What's the panel STC ?

[Skiman26]

My panels are 300 watts each I have 34 of them.