is my project economically viable ?

[RShackleford]

No one should be using the possibility of increased home value in a ROI evaluation. If the ROI looks so bad that you have to start adding variables like "increased property value" to make it seem logical, you should not be going solar.

I'm not sure it looks that bad. I need to talk to the POCO coop again and make sure I understand the way the net metering works better - it really kinda sucks if I only get credit for as many peak-hour kwh's as what I used, and not for all I generate. Makes a big difference: PVWatts has me generating about 150 peak-kwhs per month for the 6 months of the "summer" rates, whereas I use about 50 in each such month.

Anyhow, what would you all say is the ratio of system cost to expected annual payback that makes sense ? Perhaps a range, between the number where it's definitely a good idea, and the number where it's definitely not. I suppose it depends on the time value of money, or inflation.

[foo1bar]

I suppose it depends on the time value of money, or inflation.

Somewhat.
I ignored that. And ignored the likely increases in cost of electricity from the POCO.
I figure those two will both be small percentages and in opposite directions so at least partially cancel each other out.

IMO 7-10 year is OK - probably worth doing. <7 year is definitely where I'd do it. >10 year I would be hesitant (but might still go for it after more analysis.)
I think it also depends on how certain you are you'll keep the property for 7 (or 10 or whatever) years.

[RShackleford]

Anyhow, this is an important thing to consider, but ... does anyone have any answers/opinions for the questions asked in my OP here: https://www.solarpaneltalk.com/forum...for-diy-system ... from which thread this current thread got spun off ?

[RShackleford]

I ignored that [the effect of inflation]. And ignored the likely increases in cost of electricity from the POCO.
I figure those two will both be small percentages and in opposite directions so at least partially cancel each other out.

Good point. To the extent that electricity rates track with inflation, both can be ignored.

I think it also depends on how certain you are you'll keep the property for 7 (or 10 or whatever) years.

Pretty sure.

[RShackleford]

I decided to get more serious about figuring out my shading losses. One simplifying thing: I can model my shading as though my panels will sit at the bottom of a bowl with a very irregular lip, in other words, for any given azimuth the sun is blocked up to a certain elevation angle, and then it's clear above that. So I built this little gizmo (photo below) which I can aim in a certain direction, giving me an azimuth within a degree or so, and then aim the tube at the top of the treeline, giving me an elevation within a couple of degrees. I aimed it for every 10 degrees of azimuth from 120 to 240 (no chance I'm gonna get any yield from directions outside that range), and measured the elevation angle of the top of the treeline.

Then, in the pvwatts_hourly file that PVWatts generates (with shading loss set to zero), I added columns in which I compute the azimuth and elevation of the sun for every hour of every day. I put my measurements into a little lookup table. Indexing into that table for each sun azimuth position to grab the treeline elevation, I compared the sun's elevation to see if I'd be shaded. I also added a column for the degree of sun transmittance if the sun angle was below the trees; for now, I set that to 0 for all days except Dec 1 through March 15, for which I set it to 50% (optimistic perhaps). Finally, I computed a column for the overall percentage of transmittance (complement of shading) I'd see for each hour of each day: 1 if the sun's elevation is above the treeline, 0 if it's below the treeline for any days except Dec1-Mar15, and 0.5 if it's below the treeline and it's one of those winter days.

Finally I did a dot-product (Excel SUMPRODUCT function) of the "AC system output" column with this "percentage of transmittance" column, and came up with a kwh/year figure adjusted for my measured shading. Pretty grim result: just a little over 50% of the PVWatts number (which already has all the system losses except shading incorporated). Specifically, PVWatts gave 7315 kwh/yr with 0% shading (but all the other system losses). PVWattts gives 5107 kwh/yr with the 30% shading number the installer used for their guaranteed-production quote. And with my modeling of shade I get 3757 kwh/yr.


IMG_20191109_172619430.jpg

[hardwired]

Is this still accurate - "I calculate a return of about $368 per year"

For $30 clams per month forgettaboutit,

[bcroe]

I can model my shading as though my panels will sit at the bottom of a bowl with a very irregular
lip, in other words, for any given azimuth the sun is blocked up to a certain elevation angle, and then it's clear above that.

Pretty grim result: just a little over 50% of the PVWatts number

I see my open area as a bowl as well surrounded by shading trees. To minimize shading, the
array facing east is placed at the west edge of the bowl, looking over the maximum length of
clear space. Facing south are at the north edge, etc.

This PATHFINDER is one method of checking shade for all season in a single measurement.
Yes that is rather grim, would call for a chain saw here. Bruce Roe

PathFd1.jpgPathf2.jpg

[J.P.M.]

I decided to get more serious about figuring out my shading losses. One simplifying thing: I can model my shading as though my panels will sit at the bottom of a bowl with a very irregular lip, in other words, for any given azimuth the sun is blocked up to a certain elevation angle, and then it's clear above that. So I built this little gizmo (photo below) which I can aim in a certain direction, giving me an azimuth within a degree or so, and then aim the tube at the top of the treeline, giving me an elevation within a couple of degrees. I aimed it for every 10 degrees of azimuth from 120 to 240 (no chance I'm gonna get any yield from directions outside that range), and measured the elevation angle of the top of the treeline.

Then, in the pvwatts_hourly file that PVWatts generates (with shading loss set to zero), I added columns in which I compute the azimuth and elevation of the sun for every hour of every day. I put my measurements into a little lookup table. Indexing into that table for each sun azimuth position to grab the treeline elevation, I compared the sun's elevation to see if I'd be shaded. I also added a column for the degree of sun transmittance if the sun angle was below the trees; for now, I set that to 0 for all days except Dec 1 through March 15, for which I set it to 50% (optimistic perhaps). Finally, I computed a column for the overall percentage of transmittance (complement of shading) I'd see for each hour of each day: 1 if the sun's elevation is above the treeline, 0 if it's below the treeline for any days except Dec1-Mar15, and 0.5 if it's below the treeline and it's one of those winter days.

Finally I did a dot-product (Excel SUMPRODUCT function) of the "AC system output" column with this "percentage of transmittance" column, and came up with a kwh/year figure adjusted for my measured shading. Pretty grim result: just a little over 50% of the PVWatts number (which already has all the system losses except shading incorporated). Specifically, PVWatts gave 7315 kwh/yr with 0% shading (but all the other system losses). PVWattts gives 5107 kwh/yr with the 30% shading number the installer used for their guaranteed-production quote. And with my modeling of shade I get 3757 kwh/yr.


IMG_20191109_172619430.jpg

That's a pretty serious attempt and, if I understand what you are writing correctly, it appears, to me anyway, well thought out. FWIW, nice work so far.

I've done things similar to what you describe.

A comment meant as constructive criticism: If you want to perhaps improve on your method, read up on solar irradiance, particularly the direct and diffuse components. It's not complicated, but a bit messy to understand because a lot of the information is either semi- or entirely empirical.

Precise: Under a "clear" sky, diffuse irradiance makes up ~ 20 % of the total irradiance reaching the ground. Of that 20 %, and depending on the method used to allocate beam/diffuse irradiance, very roughly, about 2/3 - 3/4 of the diffuse will be incident on a shaded surface such a shaded array if the shading object is solid. And of course, that solid shading will be changing constantly.

Because no two are alike, and making things more complicated, shade from trees or semi transparent objects is nearly impossible to model. That's most of the reason all this is so empirical. Then there is also the idea of how an array works under partial shade.

The shade modeling I did back in the day (and back east) was slightly before the solar pathfinder got to market. When available, I borrowed one from my last alma mater to compare results with what I'd done. I was also under the yoke of a hand held irradiance meter which was OK, but probably less than fit for purpose. What I found was that the solar pathfinder gave results that were about in line with what I had done in terms of shading penalties. However, and perhaps different to what you write of here, irradiance measurements were a bit higher than the what the solar pathfinder would lead me to believe if I'd discounted the contribution of the diffuse component. The two got closer, but still off a fair amount when I adjusted (and still a big SWAG) the P.O.A. irradiance to account for some of the diffuse portion of the irradiance hitting a surface that is in shade.

All that was pre-spreadsheet days so I was working in Fortran IV and later in BASIC. Looking back, all that seems, and probably is archaic compared to current tools.

While whole heartedly encouraging you to continue your investigations, I'd respectfully suggest you consider looking into either acquiring or renting a solar pathfinder of the type Bruce has pictured. Most folks need it one time and renting is probably a more cost effective way to go. Not a plug, but I believe the solar pathfinder can give reasonably reliable shading estimates. For most applications, that's more than sufficient with results that are about as accurate as necessary with not too much effort.

Take what you want of the above. Scrap the rest.

[RShackleford]

Precise: Under a "clear" sky, diffuse irradiance makes up ~ 20 % of the total irradiance reaching the ground. Of that 20 %, and depending on the method used to allocate beam/diffuse irradiance, very roughly, about 2/3 - 3/4 of the diffuse will be incident on a shaded surface such a shaded array if the shading object is solid.

So perhaps I'm going to get more irradiance on my panels that a simple shading model (like the one I did) would indicate ?
So maybe in the winter (trees without leaves) my 50% transmission is pretty reasonable. But in the summer, I should allow more than zero when the sun is below the treeline, perhaps 3/4 of 20%, or 15%. But only when it's something like 15 degrees below the treeline. When I make those adjustments, I'm down to 3600 kwh/yr (less, because formerly I didn't have the 15 degrees-below-treeline restriction on my winter 50% transmittance thru the trees.

[RShackleford]

Is this still accurate - "I calculate a return of about $368 per year"

For $30 clams per month forgettaboutit,

Actually, the way my net metering is done makes computing the savings pretty simple. All my generation zeroes out first my peak kwh's, then the off-peak;, and I'll always generate more than my peak usage. This works even if the generation during peak hours is less than the peak usage. This all boils down to a savings of about $300/year. However, right now I am making a voluntary "green power" contribution of $16/month on my electric bill (this somehow funds "blocks" of renewable energy generation); I would stop doing that, so now the savings is about $500.

If I build my system on the cheap, use the $0.36/watt Talesun panels, and no optimizers, I think I can get it up for about $6000, or $4000 after federal credit. So 8 year payoff.

[RShackleford]

This PATHFINDER is one method of checking shade for all season in a single measurement.

So I guess you still have to manually add up all those half-hour numbers for each month ?

[J.P.M.]

So perhaps I'm going to get more irradiance on my panels that a simple shading model (like the one I did) would indicate ?
So maybe in the winter (trees without leaves) my 50% transmission is pretty reasonable. But in the summer, I should allow more than zero when the sun is below the treeline, perhaps 3/4 of 20%, or 15%. But only when it's something like 15 degrees below the treeline. When I make those adjustments, I'm down to 3600 kwh/yr (less, because formerly I didn't have the 15 degrees-below-treeline restriction on my winter 50% transmittance thru the trees.

It's very hard to estimate. That's on big reason why most of the methods are as empirical as they are. To have a detailed discussion is beyond the scope of this venue. That's why I suggested you read up on beam/diffuse irradiance and related concepts. If you do, you'll still not get a simple answer, but you may well understand a bit more of why you won't get a simple answer and learn a bit along the way - if you're curious.

You appear headed in the right direction but, IMO, you need background material and knowledge

Good luck.

[jflorey2]

Or, as I've seen more often on a local level, and similar to a lot of the posters I see around here, residential PV may add perceived value if the buyer is uneducated enough to let others with skin in the game oversell the benefits and swallow the hype - a lot of it from friends and neighbors who are equally ignorant, and/or from sources such as the green wash media shills of the solar peddlers.

Perhaps. But that's true of any feature of a house that doesn't keep the rain off your head, like granite countertops, pools, insulation, efficient lighting, instantaneous hot water heaters, sheds, nic views, proximity to water etc etc.

[bcroe]

So I guess you still have to manually add up all those half-hour numbers for each month ?

I would use it more as a guide, as to where my chain saw would do the most good. For numbers
you might note the sun rise clearing shade times for 6 months, and also times for sun going down.
Compare those dozen times to official sun rise/sun set to see how much time you are losing. But
for the typically south facing array, poor angle production in those intervals is quite small anyway.
It is more important if you have arrays directly facing the rising/setting sun as here. Bruce Roe

[J.P.M.]

Perhaps. But that's true of any feature of a house that doesn't keep the rain off your head, like granite countertops, pools, insulation, efficient lighting, instantaneous hot water heaters, sheds, nic views, proximity to water etc etc.

Agreed, but that doesn't make the problem go away or the deception any more ethical.