New solar setup - Enphase IQ8s and clipping

What’s obvious above is that I have some clipping. Let me describe my system.

My system is a 9.49kW (dc) setup - 24x 395W Canadian Solar panels with Enphase IQ8+ microinverters, (290VA max continuous, 300VA burst), and the BOP for Enphase Sunlight Backup. More info on that here and here. Short version: The system can island itself with an ATS and keep ~30% of the AC nameplate capacity when grid is lost but sun is shining. I’m in a location where I expect about 12.5MWh/year from this system, 4-4.25 average hours of sun per day in the cloudy Midwest. Adding Sunlight Backup to the proposal added about $3k.

I heavily rely on the grid to be my free infinite battery, especially seasonally, since my home is all-electric and most of our energy use is in the winter and less in the summer. Net metering here isn’t bad, I have monthly true-up and get paid out about half of retail in months that I overproduce. This seems pretty fair and very simple.

This system was energized in January after a long arduous process with an installer that hadn’t ever done and IQ8 system before- papers signed in June, panels up in October, first production in January. Oof. Since my system came online in winter, every day this spring has the potential to set a new production record! Yesterday I made 57kWh which is great!

Clipping is an issue. When the installed proposed putting 290VA continuous-rated microinverters on 395W panels, I was quite hesitant. They told me they’d done the math and showed some familiarity so I went with it. When I observed the clipping starting so early, I asked them to give me their numbers and they sent me this which seems to indicate I’m still in the best position financially:

  • iQ8M (330VA continuous) micros in lieu of iQ8+:

    • Increase in proposal cost for difference in equipment price = $1,179
    • Increase in annual production of 186 kWh. Value @ current $0.116/kwh = $21.576/year
    • ROI = over 54 years
  • iQ8A (366VA continuous) micros in lieu of iQ8+:

    • Increase in proposal cost for difference in equipment price = $1,579
    • Increase in annual production of 209 kWh. Value @ current $0.116/kwh = $24.244/year
    • ROI = over 65 years

Naturally the cost difference is a key factor- they’re calculating with +$49/ea for the IQ8M whereas it’s only a $20 price difference on, and the ROI would technically be different but probably still not enough to make it worth it. I’m not seriously considering asking them to change out the inverters at this point, just making sure I understand the situation.

So it seems like ~25% over-paneling is within economic range now, clipping be damned. What do you all think?

That’s not unreasonable.

I don’t recall what your AC interconnect is, but another issue is often with the AC side interconnect of things.

The IQ8+ are rated 1.21A continuous AC output, so 29.04A - derate at 80% to 36.3A required ampacity for the busbar, and you’re within the 120% rule for a 200A busbar.

If you go to the IQ8M, they’re rated for 1.35A output - 24x is 32.4A, derated, requires 40.5A - so you can’t actually put 24 of those on a 200A busbar, meaning “more costs” (potentially, I’m not sure how everything is connected).

But especially with a 200A busbar, “more panel and a high AC/DC ratio” makes a lot of sense for total system cost.

The other question, though, is how they handle running at 100% rated power for long periods. I don’t like to run power electronics that hard. I know they’re rated for it, but I’m only running my big inverters around 75% of rated most of the year, or less, so I expect some improved longevity from it.

The AC interconnect side is handled via a line-side tap. I was surprised they picked this route since I assumed it couldn’t be possible with the possibility for islanding, but the arrangement they have with an internal ATS in their system controller and a separate small subpanel for backed-up loads works out. They extended the conductors for the backed-up circuits to that panel, landed them on new breakers, and fed those breakers from the AC interconnect bus on the system controller on the load-and-PV side of the ATS along with the inverter strings. They also have contactors between the AC bus and the breakers for shedding capability.

So- no MCB or panel bus ampacity issues, I get to utilize the full ampacity of my service drop, which is probably 200A+ as far as the utility (outside of NEC) is concerned.

100% for long periods remains an unknown. There are a few million microinverters up on the roofs for their nth summer in the sun each year, and I await the news of how they’ll actually do on their 25th year. We shall see- technology always has its risks. I’m cautiously optimistic, but not unwary. I think it’s worth noting that the time these inverters will spend at actual max continuous output is a tiny fraction of what industry would consider continuous duty. It’s a bit irrelevant for thermal soak purposes, but comparing to a device that’s intended to run 24/7/365, these have a cake-walk since they have nights off, most of winter off, mornings and evenings off, etc. My setup is more like a torture test for microinverters in that they’ll spend more hours maxed out each year compared to a setup where they’re paired with a 300W panel, where they’d almost never reach full power in practice. Like anything else predicting the future- we’ll see!

Huh. I do recall you mentioning a line side tap, but their numbers lined up oddly well for the 40A backfeed limit.

Realistically, it’s probably better for the grid to be clipping as well during the solar peak - that solar noon peak comes at an inconvenient time. And it sounds like their numbers work out, just… eh. I do hate throwing away potential power for a grid tie system. My office, meanwhile, dumps most of the potential production, because I’ve got nothing to do with it on sunny days.

And good point regarding actual service life on the microinverters. I just don’t like Enphase very much…

Yeah, we ended up with the system size we did by maxing out roof space. We applied to the city with 26 panels but got knocked down to 24 panels for setback requirements. I mean, looking at my array, I cannot imagine any firefighter wanting to walk on the edge of the roof next to the panels, but okay. I’m going to be overproducing anyway, but I didn’t know that this time last year. The consumption data from the older couple that lived here before was quite high.

Agreed for the grid impact reasons. A wider, flatter production curve is easier to absorb on the grid than one with a peak, which may add up across 1000 (or 1M) customers with similar setups. I don’t think it will do much for intermittency, the overhead between panel capacity and inverter capacity won’t provide meaningful buffer for clouds going over etc.

Enphase says they should care about those things! I remain somewhat skeptical of the firefighter setbacks and such as well, and my view is that if you’ve got another perfectly good roof pane next to the solar covered one, that should work just fine to vent things (this seems to be generally what firefighters say they’ll do as well).

I think it’ll buffer more than you think. Cloud cover vs production isn’t at all obvious to human eyes, given how well they adjust. You’ll likely have “stable production lines” on some partly cloudy days.