Solar for me... sounds shady!

[woodgeek]

Well, I think after 20 years of thinking about it, I am about to pull the trigger on a rooftop solar system.

I can benefit from your wisdom.

My lot is on the north side of an East-West ridge about 60' tall, with 100' mature trees along the ridge. My whole lot is covered with mature trees, so that there is little sun on my roof and I can't grow veggies anywhere in my yard.

BUT, the emerald ash borer came, and I lost ~ 14 Ash trees in the last 5 years, many close to my house. I have a few smaller trees growing over my house that I am happy to take down or cut back. I am also thinking that there is no law that my tiny back yard needs to be dark, gloomy and damp all summer with minimal plantings... we will still have LOTS of trees, but can clear the region around the house.

The proposal (best of 2 I got) is to squeeze 28 panels on the back roof of my house (facing SE, 35* tilt), total rating 12.88 kW_DC. The cost is $27,900, or $2.14/W. 25 year design lifetime, parts under manufacturer warranty, installer provides 25 year repair labor and roof penetration warranty. Micronverters, all equipment made in USA, great reviews on company in business for 8 years.

My electricity costs $0.25/kWh (having doubled in the last 8 years), and PA SRECs would add another 2 cents/kWh (variable). My annual usage in an all electric house with 2 EVs is ~20,000 kWh

My existing roof is in great shape, 14 years old, I have no doubt it will last 25 years more.

So, if I got 1000 hours of DC production per year (as for a modestly shaded lot) that would be 12,880 kWh/year, $3500/yr return, 8 year simple payback w/o inflation. I currently believe we will have a few more years of kWh inflation... that could easily drop to 5 years. I will likely live in the house for at least 10 more years, health willing.

As you might expect, I am not too worried about payback (I'm happy with 10-12 years honestly), I just don't want to flush money away.

Two issues for consideration:

(1) Shading. Two companies flew drones over my house, Ran a 3d canopy model. They got production figures for current canopy in the 5000 - 7000 kWh/yr range. That is around 500 hours of DC rating output per year, very shaded. For the record, I have a 60W hobby panel, below my roof, and I get 400 hours of DC rating out of it per year (about 25 kWh/yr, LOL). That blows up simple payback to 16 years, no inflation.

The range 5000-7000 comes from their branch shading model. Those distant tall trees (not on my property) give diffuse sun when the leaves drop. A conservative branch shading model give 5000 kWh. An optimistic one give 7000 kWh.

Two teams digitally removed selected trees. Just removing two small one growing over my house and one tall one to the SE, projections get to 10,000-12,000 kWh/year. 10,000 kWh/yr get me a 10 year simple payback. And I'm happy to pull the trigger. I will remove those three trees.

(2) Financing. Biden era incentives for resi installs are gone, but are NOT gone for commercial projects. So, installers in 2026 have a loophole. The prepaid lease. They can install the system, and have it bought and owned by a bank. The bank then sells it to me with a 25 year loan. The bank gets a rebate, and my loan principal is lower as I get part of the rebate. After 1 month I pay off the balance of the loan in full, no penalty.

The rub is that the bank retains ownership (with no claim to power) for 60 months, as required by the feds for rebate. After 60 months they send me an email that I now own the system, no cash required. The only rub is that warranty claims during the 60 months have to be made by the bank. I have to call the bank, and then they authorize the work by the installer. After 60 months, the warranty tranfers to me as the owner.

The company also produced a cash quote. It used non-US panels to control cost. It was 5% less power, and cost 10% more to me. But I'd own it free an clear on day one.

So for the additional step of the 60 months of shadow ownership, I get $3000, and 15% faster (simple) payback.

Thoughts? Cash or prepaid lease, PPL? I am leaning to the higher power, PPL system.

Or would you can the whole thing afraid that the production is too low. The $2.14/W price is excellent for this area, which is $2.50-2.80/W these days.

 

[stoveliker]

I don't trust the drone (or satellite) exposure estimates.
Any reputable company puts a fish eye camera with gps and orientational info on all corners of the roof, and in fact measures (plots) where the sun is in the sky at height of summer and depth of winter. That shows you exactly what trees are shading where, when power cuts out, and allows for a much more accurate prediction.
Basically I don't trust anything that did not show me photos with plotted sun paths.
I had those, and they were exactly on target. 6.2 MWh per year. (yes, note the decimal that was predicted correctly).

Bank financing thing: if you trust your bank to step in for you when needed, without further charging you (check the contract), I'd be okay. If you're just a number for your bank, I'd be hesitant.
In the end you paid for it but don't own it for 5 years. That's an issue in my book.
3k$ on 28k$ is not that much, relatively speaking, but 3k$ down the drain for mere peace of mind is a high price. So it depends on how much peace of mind you get for that 3k$ (hence my questions about trusting your bank to behave in your interest if SHTF).

Don't can the whole thing if you trust the production estimate. Payback of 6-8 years is great.I agree prices will keep going up, and that's going to cut that timeframe significantly.
For me (Oct. '18, 22k$, 55% tax credits paid for by y'all, so 10k$ my price, and an oversized system, so I added mini split heating (and cooling) - saving oil (I was still burning significant oil then, now it's wood) - have long paid for my system. Each month I'm saving $150 or whatever the bill would have been by now, and with AC/heat that would likely have been closer to $200.
It's one of the better financial decisions I've made.

And you're primed to add a battery if your EV batteries are not sufficient at some point. Independence, baby!

So, do it, ask for a fish eye lens production estimate, think about whether you trust your bank for 3$k.

 

[stoveliker]

Example of what I mean.

My system is two rectangular arrays, on E and W facing roof sides (ridge is N/S, unfortunately).
Each side (E/W) had pics made at the 4 corners of each rectangle. So 8 pics in total. Below is one of them.

With the orientation and gps, they have the solar tracks on the sky, know the angle of incidence and when the roof is shaded by trees or the earth itself (..).
From that they calculate (rather than estimate!) the production.
Of course using average weather data.

View attachment 347944

 

[woodgeek]

Example of what I mean.

My system is two rectangular arrays, on E and W facing roof sides (ridge is N/S, unfortunately).
Each side (E/W) had pics made at the 4 corners of each rectangle. So 8 pics in total. Below is one of them.

With the orientation and gps, they have the solar tracks on the sky, know the angle of incidence and when the roof is shaded by trees or the earth itself (..).
From that they calculate (rather than estimate!) the production.
Of course using average weather data.

View attachment 347944

I recall the fisheye systems. Nowadays everyone uses drones and claim it gives a more accurate result. They showed me (and rotated) a 3-d model of my roof planes and the entire tree canopy as it exists today (versus a stale satellite image). They can then generate sun tracks going behind canopy for every one of the 28 panels. The estimate is an annual average with sun tracks in that dataset, which is all GPS anchored.

They can also delete individual trees in 3d, and see the effect. If there is a tree behind the deleted tree in line of sight, its effect is smaller.

I trust that calculation, done by two companies with two different packages giving similar results.

The uncertainty is that my shading is so high (<40% of unshaded production if canopy was 100% opaque), that a significant amount of my production is filtered by tree branches when the leaves are off. Eff is also higher in those cold months. Shading is to south and southwest. Clearest region is East to southest, but that is not 100% clear either.

The system uses microinverters and panels with diode strings internally to maximize partially shaded output.

The 3d model measures canopy geometry, but doesn't measure opacity of branches or how that interacts with panel output. So for that portion of the output, they use a **model** to guess the opacity (lower near the top of the canopy and higher lower down), and a factor for its effect on the panels (in my case, a lot of the branches are far away and the shading is diffuse.. good). And that is why different models give me different output. Some predict I will get squat in the winter, as the tracks are below the canopy line ALL the time (conservative) others give a guess that I will get some.

I know that I get solar production in the winter from my hobby panel. I get more in the month before leaves come out (April) than I do after (May). Their conservative models (branches are opaque) don't predict this. Their other models (with branch filtering) DO predict this.

I should've done a fishweye survey in the winter with the leaves off... they could've gotten that number.

 

[stoveliker]

okay. Clearly the industry improved over what it was in 2018

My system only produces 60% of what it would if it was facing optimally. (Whatever that may mean, because unless you have a rotating orientation, one never keeps perpendicular incidence all day long.) I have ridges E and W of me, so I start late and stop early. (after 3 pm I don't make much, summer or winter...)
I did notice that I do make a bit more at 4 pm when the sky is cloudless but hazy. Cloudless is not ideal in all cases!

I would not worry about 40% efficiency or whatever - what matters is whether it produces enough kWh to make financial sense. I think you are stating that it does. Who cares that the neighbors have a nicer car have a better producing PV system. If you gain financially, it's a go.

I have microinverters too. No diode strings tho here.

 

[Graphex]

Is community solar a thing in your state? It is here in Maine and I 'own' .5% of a 3-4 acre solar farm that's on a farmer's pasture about 15 miles away. My share of what's generated shows on my utility bill as a credit against my useage. Credits carry forward 12 months and then I lose the remaining excess - they actually go into a pool for low income power customers. I can transfer to a new account if I move or transfer it to the new owner if I move out of state. After 4 years I can say payback is under 8 years. When I signed up I looked at it as a 12.5% CD with it paying for my power for 25 years. I do pay a monthly connection charge that's now up to about $30/month and an annual fee of $200-ish to the company that admins the owner group and does maintenance on the array.

 

[stoveliker]

I like excess going to help low-income folks.
My connection charge is around $17 now (was $12 in 2018...)
$200 a year is about one month of electric bill, I presume, so community solar does cuts into the profitability by order 10%.
I think on your own roof is better - and that community solar is great for folks in apartments or places where otherwise PV can't be installed.

 

[begreen]

You are getting good suggestions. Our house is in a somewhat similar location. on the north slope behind a gradually sloping E/W ridge. Our house didn't work out well for shading so the arrays are field mounted. One on the ground and one on a pole.

Questions:
Are all the shading trees deciduous? (Here we have a mix, but mostly conifers. Winter gain when leaves have fallen doesn't help due to persistent shading and low arc of the sun at 47.3º N.)
Did the payback calculation include declining performance? (Our arrays are 10 & 13 yrs old. The decline is notable.)

As far as ROI, there are other gains besides bottom line. Free fuel for the cars has value, so does getting off of the fossil fuel teat.

 

[stoveliker]

Interesting.
I'm starting my 9th year in October and the decline is not beyond statistical yearly weather average fluctuations:

first full year 2019, yearly totals in MWh
6.3
6.3
6.2
6.2
6.2
6.3
6.2

 

[begreen]

Our original panels on the ground array are old tech, 2011 panels. They are all made in WA which was required to get the maximum rebate from WA state @ 54¢/kWh. The company went out of business and some of the panels are starting to outgas. They are performing with a decline of about 21% so far. The second array is also made in WA. They were installed in 2015 and are a more conventional design. So far their decline has not been too bad, less than 10%.

 

[stoveliker]

Just to have some reference data, I looked up my old quote.

My system has an estimated "4.6 sun hours", which means a daily power production that is the same as if each day the sun was 4.6 hours at noon height.

In fact, I see that they predicted 6.144 MWh production per year. So less than what it has been doing.
Note that tree growth will decrease current estimates too over time.

The production estimate in my case says they assume 0.05% annual PV performance reduction.
That is less than what the PV panel mfg warranty guarantees.
The warranty says year first 5 years no less than 95% of labeled power.
From year 6 to 25, they say 0.4% per year, meaning at year 25 I still have 87% guaranteed. We'll see about that, but a 13% decline in 25 years is 0.5% per year, which is great imo.

 

[woodgeek]

No community solar around here.

My panels have a warrantied max decline of 8% in 25 years FWIW.

For ROI, I am just computing SIMPLE payback, no inflation or anything. Just cost divided by year one revenue. That looks like 10-12 years. My power inflation recently has been close to 6-8% per year, so that will (i'm spitballing) get me closer to 8 years positive NPV if it continued for a few more years. I can look at it as a hedge... power stays cheap I win, power gets expensive, I'm hedged.

And dropping my total home + 2cars CO2 emissions by about 50%, from ~8 tonnes COeq/year to ~4 tonnes/yr will be nice. 2005-2008 home + 2 cars was ~24 tonnes/year for reference.

Not counting wood CO2, for the record.