enphase s280 with LG315N1C-G4

[sensij]

There is another CEC calculator that I thought ncs55 might be referring to, that could conceivably be more accurate (as opposed to the version that is based on PVWatts V2). Link here:

http://www.gosolarcalifornia.org/too...ator/index.php

It would not be very smart to size a system to produce exactly what they are using from day one, as it would not take long before that system would be to small to cover their consumption.

Anyway, the idea that a system must be sized the achieve 100% offset to be successful is another fallacy perpetuated by solar installers. "Too small to cover their consumption" is in some cases the most cost-effective way to size a system.

[J.P.M.]

From what I see of systems that I check out and or repair,(not my design or installs) The CSI calculator is pretty close to the data that I see from the customer before I optimize the system.

I see the opposite, especially with new systems. Also, my very limited experience is that 4-6 yr. old systems produce about the same as when new, but that's more anecdotal than documented. Even so, given that the CSI calculator seems to oversize as much as it does - both for my system and others I've seen, usually by 10+ % or more - I believe it will take a long time for system degradation to catch up with such low estimates.

Yes, PV Watts seems conservative until you start making parameter changes. The problem I see in my area is that sales people are doing just what you said to sell more equipment.

Caveat Emptor. Knowledge is power.

The other thing to consider is that, in no way can any calculator present what a system will actually produce. It cannot predict the weather or how much irradiance will hit the modules, how much dirt accumulates on the array vs cleaning etc.

Of course not. But most folks, including most peddlers, do not understand that.

You can get as detailed as you want and I see that you did, and none of the calculators were 100% correct. In the end you will only get a range of production.

100 % correct is not a term that's applicable to this field. If that comment is meant for me, be advised your preaching to the choir.

[ncs55]

That is well taken and interesting. Seems your system is kickin some booty! From what I see of systems that I check out and or repair,(not my design or installs) The CSI calculator is pretty close to the data that I see from the customer before I optimize the system. Yes, PV Watts seems conservative until you start making parameter changes. The problem I see in my area is that sales people are doing just what you said to sell more equipment. The other thing to consider is that, in no way can any calculator present what a system will actually produce. It cannot predict the weather or how much irradiance will hit the modules, how much dirt accumulates on the array vs cleaning etc. it only can estimate production from the values programmed into it. I usually size my systems a little larger to account for degradation and hedge the customer, and they are told this. It would not be very smart to size a system to produce exactly what they are using from day one, as it would not take long before that system would be to small to cover their consumption. When we present the estimated system output we use both calculators to give the customer an idea of what they can expect. It kind of gives them a lowest to highest estimate and is usually pretty close. Some years are on the low end and some years are on the high end and some years are in between. You can get as detailed as you want and I see that you did, and none of the calculators were 100% correct. In the end you will only get a range of production.

[J.P.M.]

[QUOTE=ncs55;n314526] I was referencing to the system performance overall, and the CEC calculator gives much more accurate results than PV Watts for this region.

I have a different opinion about your version of the accuracy of what you call the CEC calculator.

Probably not as extensive as your data base/experience, and without diving into what PVWatts is, and how its estimates of long term performance can be interpreted (or misinterpreted), from what I've seen, PVWatts seems to give a reasonably reliable estimate of likely performance, at least in 92026, provided the system loss parameter is adjusted to something like 10% or so, at least in my neighborhood.

Early on, some of us railed about how PVWatts seemed to be underestimating likely performance. Others seem to have noticed the same thing because Paul Gilman and the folks at NREL redid PVWatts. alluding to comments from users that it was too conservative.

According to the text in the CSI incentive calculator page, the CSI- EPBB calculator (assuming that's what you're calling the CEC calculator) is itself based on PVWatts. When I plug in the LG module and inverter to the CSI calculator, it pukes out 474 kWh/yr. per module. PVWatts for the same system with 10% system losses gives about 541 kWh/yr. per module. PVWatts needs the system loss parameter to be forced to about 19% to get down to the CSI #. of 474 kWh/yr. per module. That,19%, IMO only, is unrealistic unless conditions warrant it. Not many do.

Also stated on the CSI calculator page: "...while this calculator relies on industry standard-assumptions, and is driven by NREL's PV Watts v. 2 calculator..." . Ignoring the incongruity that sentence from NREL introduces into your statement: "the CSI calculator gives much more accurate results than PV Watts for this region", I'd note the part about industry standard assumptions. Given how conservative the CSI calculator seems to be, seems to me that vendors just might be tempted to use the CSI calculator and/or manipulated PVWatts output estimates via running up the system loss parameter, as tools to sell extra equipment, using those tools' implied accuracy to give the solar ignorant the impression that they (the potential customers) need a bigger system than is actually necessary. I'd bet those "industry standard assumptions" might just err on the side of oversizing. No salesperson in any industry I know of ever got fired for over sizing equipment.

I just reran my system using the current CSI calculator with the same inputs my vendor used in Aug. of 2013 when they calculated my CSI incentive. I got 8,009 kWh/yr. today - the same as they got in Aug., 2013. That would lead me to guess that PVWatts v.2 revision didn't change my estimate much.

For some comparison, SAM calculates my ave. long term system output at 9,469 kWh/yr.with zero shading and 3% fouling using TMY3 Miramar data.

I've got self written software that estimates my system's long term ave. output at about 9,450 kWh/yr, using my exact location and elevation, zero shading and 3 % fouling.

Other folks in my HOA I've monitored and estimated have outputs that are much closer to the SAM estimates than the CSI estimates, and need PVWatts system loss parameters usually somewhere between 8-10% range to get close to actual clear day totals.

For my system, for PVWatts to match the SAM output, I need to force the PVWatts system loss parameter to something like 6.5 %

My actual, annual day over day running 365 prior day output since 10/18/2014 (that is, the prior 365 day running total output, updated each day) has ranged from 8,974 to 9,520 kWh, with a mean of 9,229 kWh/prior 365 days, and a population std. dev. of 182 kWh, making that data maybe a bit leptokurtic if anyone cares.

That actual recorded data includes what I estimate as somewhere between 3% and 5% shading penalty in the late afternoon, depending on the season, and an estimated ~ 0.6 % difference between the Sunpower monitor and the SDG & E meter. I believe the S.P. monitor reads about 0.6% high(er) than the SDG & E meter. Long story on how I got that #. Some other time maybe.

[wes@SH]

This is my favorite seemingly unbiased resource I've found for the Enphase vs SolarEdge debate: http://www.redstonecleanenergy.com/e...-vs-solaredge/

Besides being a very close side by side comparison, it also lets you see the an active system for each company in case you have a GUI preference. By my calculations for the slight difference in panels vs orientation of arrays, the SolarEdge system, with essentially one less panel and a slightly worse roof pitch for that latitude, is surpassing or equaling the Enphase output.

[ncs55]

sensij commented

Today, 01:58 PM




I don't understand your question. Inverters only rarely run at the maximum, that is how a smaller inverter (with a higher installed DC STC array rating) can outperform a larger inverter (with a lower installed DC STC array rating). Clipping is the whole point of what we are discussing, if there was no clipping, there would be no question that a system built from 320 W panels would outperform a system with the same number of 300 W panels.

Ok, factoring the clipping and everything else. The system with the 300W modules are coupled with 305W ac output micros. How can a system with 280W ac output micros produce more power overall at maximum output ratings. I understand that micros rarely operate at maximum but for comparison of these two systems I am referring to the rated output of the products at maximum for even comparison.

[sensij]

I don't understand your question. Inverters only rarely run at the maximum, that is how a smaller inverter (with a higher installed DC STC array rating) can outperform a larger inverter (with a lower installed DC STC array rating). Clipping is the whole point of what we are discussing, if there was no clipping, there would be no question that a system built from 320 W panels would outperform a system with the same number of 300 W panels.

[ncs55]

sensij commented

Today, 01:07 PM




Yes, the instantaneous output of the 300 W inverter would be higher than the 280 W inverter under the right irradiance conditions. However, over the course of a year, the system built with 320 W panels with 280 W inverters would still outproduce the system built with 300 W panels and matched inverters... there just aren't enough clipping hours in the year to offset the outperformance during all the non-clipped hours.

Based on inverter output alone and who cares about the clipping, how is it possible for the 280W AC output inverter to produce more than a 300W AC output inverter annually when they are both running at maximum?

[sensij]

Yes, the instantaneous output of the 300 W inverter would be higher than the 280 W inverter under the right irradiance conditions. However, over the course of a year, the system built with 320 W panels with 280 W inverters would still outproduce the system built with 300 W panels and matched inverters... there just aren't enough clipping hours in the year to offset the outperformance during all the non-clipped hours.

[ncs55]

sensij commented

Today, 12:47 PM




Perhaps I am misunderstanding your comparison. If the total system size is held constant, say, 5 kW, then yes, a system with larger inverters may outperform the system with smaller inverters, if the clipping is worse than the low light inverter efficiency hit. However, if the number of panels is held constant, and the choice is between 320 W panels with a 250 W microinverter vs 300 W panels with 300 W microinverters (or some other design that minimizes clipping), the 320 W panel will outperform under almost all installation scenarios. Any reasonably accurate modeling system will agree with this.

Maybe. System overall performance, not module vs module performance is the reference. It is pretty obvious that a 320 outputs more than a 300 on the dc side. The ACE 300 has an output rating from the inverter at 305W continuous, AC side. The Enphase coupled with whatever module you want has an output rating of 280W continuous, AC side. Based on the inverter output in this scenario, how would a 280W inverter out produce a 300W inverter in full sunlight? Even if the enphase starts a little earlier and runs a little later it will probably not be enough to outperform the other higher W inverter since the most production is during the solar window say between 10 am and 2 pm, give or take. So regardless of the module output on the DC side, the system with the 300W modules and 300W inverters would spank the 320W modules with the 280W inverters in overall production, given both systems were the same size, right? Hence my point. It is not only about the module wattage in the overall system design and the highest wattage module is not necessarily the best choice depending what they are coupled with. I see what you are saying about the module performance but given both systems are operating at max, (which is not realistic and only for comparison). The system with the lower W module in this scenario, would outperform the other, right?

[sensij]

Perhaps I am misunderstanding your comparison. If the total system size is held constant, say, 5 kW, then yes, a system with larger inverters may outperform the system with smaller inverters, if the clipping is worse than the low light inverter efficiency hit. However, if the number of panels is held constant, and the choice is between 320 W panels with a 250 W microinverter vs 300 W panels with 300 W microinverters (or some other design that minimizes clipping), the 320 W panel will outperform under almost all installation scenarios. Any reasonably accurate modeling system will agree with this.

[38kW]

Solar production May 4.pdf Here's a graph of the production meter outputs downloaded from APS website from May 4 which was a high production day.
Blue 17.36kW DC system with 42 LG 315 panels @ 213 deg azimuth 15 deg tilt & 14 panels @ 169 azimuth 15 deg tilt, Enphase S280 inverters
Red 17.85 kW 70 SolarWorld 255 panels with Enphase M215 inverters, 180 deg azimuth 7 deg tilt installed 2013

[ncs55]

You are probably right about the 350 and the micros, I do not sell those systems. I do install them for other contractors though. True if you are only considering the module performance a higher wattage module should always produce more. I was referencing to the system performance overall, and the CEC calculator gives much more accurate results than PV Watts for this region. But in reference to the 300W with built in micro vs 320W with a 250 -280 micro. The 300 would outperform the 250 or 280 due to the higher output of the inverter, in full sun. So the lower watt module would have the advantage in that scenario.

[sensij]

So in essence that 300W module would give much greater results than coupling even a 350W module with an enphase due to the limits of the inverter.

I'd agree with most of what you just wrote, but take exception to the statement quoted. A simple PVWatts model, using San Diego TMY3, roof mount, premium panels, 10% loss.

300 W panel, 0.9 DC to AC ratio (no clipping) - 511 kWh annually
350 W panel, 1.4 DC to AC ratio (modeling M250) - 583 kWh annually

I doubt you could actually pair a 350 W panel with an M250 due to other electrical incompatibilities, but the point is that higher wattage panels will outperform lower wattage panels, even when their peak is strongly clipped. High power panels just don't spend enough of the day in clipping to counter the outperformance every other hour of the day.

[ncs55]

We are talking about a special product - an LG module with a *built in* inverter. Again, LG, the distributor, or the installer will find some way to keep it from being cheaper for the customer. Which is ok, but just isn't a selling point to the customer. It seems to be sold to installers to save on labor.

Not only to save labor but because it had 305 W output. No other micro even today can beat that. And less labor would equate to less labor charges to the customer. The module would have cost more but with less labor it would likely cost the same overall to install as the other micro system and produced more for the same square footage of the array. So in essence that 300W module would give much greater results than coupling even a 350W module with an enphase due to the limits of the inverter. Like I said earlier, module wattage is not the only focus when you are designing a system. It might be with micros, but compared to strings you can have a more powerful system with lower wattage modules for less money, with no clipping or wasted module production. You tell me which is a better value to the customer. IMO micros are a waste of money and have twice the chance of equipment failure over a string inverter, they take more time to install, and use more wire, I just don't see the value unless there is a multifaceted or shaded roof. Also note that when one of the micros fails in the middle of three or more rows you have to uninstall other modules to get to it. When you have a rooftop with 200-300 modules with micros and have to dig out one or two in the middle that have failed, will make for a costly repair for the installer under warranty or the consumer if it is out of warranty. As far as production in full sun, I see no evidence from anyone that shows micros produce more than strings, they are pretty close either way.