specifying the Roof on a Custom House for maximum solar utility

[GeoMetry]

roof.png
I am building a custom home in the northern hemisphere. The footprint has been determined and the height is limited by the local code. It is a simple two sided roof and one side faces directly south. The location is 37 degrees North latitude. I can pitch the roof at any angle I want. I could use the rule of thumb and make the angle match the latitude or I could go as low as 20 degrees. Because the overall height is restricted, the higher the angle of the south side of the roof the less south facing surface area I get. The south facing surface area would be about 43 percent more if I make the roof slope 20 degrees instead of 37 degrees. How do I decide?

[peakbagger]

What sort of net metering do you have, it makes a big difference. Is there annual or seasonal reset?. Is it 1 to 1 or biased to the power company?. Ideally if you have an ideal net metering plan you optimize the generation for total annual production and use the grid as a "battery". If you on the other hand have no net metering or a bad plan with seasonal or annual reset and crappy reset date that precludes you from building up a surplus you may want to optimize the angle to produce more or less power at certain times of the year. Batteries are great for day to day storage but useless for seasonal.

Not applicable to you at 37 degrees but If you were in a snow zone that also factors in, the steeper the better for fixed panels that are subject to snow and that also plays into improved winter production.

You also need to pick out a panel size as you need to optimize the surface area to match the array sizing. In most cases the least cost will be the biggest panels you can get but if you end up leaving a strip of roof because another row does not fit you have less than optimal layout. You also need to determine if your AHJ/fire department will waive the frequently required strip of no PV near the ridge, many will waive if there is access to the back of the roof, some will not but that can limit the number of rows in an array.

[Mason Parsons]

If you pitch it at 20 degrees instead of 37, you'll get more surface area facing the sun. Plus, it won't be too steep either.

[Mike 134]

Use this program to test various roof pitches. Peakbagger has some good points

PVWatts Calculator (nrel.gov)

[bcroe]

Moving the fixed height roof ridge as far north as possible will
allow collecting the most energy for higher sun. It will also
do better under clouds. But it will not give maximum panel
efficiency, a smaller number of panels raised to around your
latitude will run at best efficiency, but lower total output.

You might want to consider that a steeper panel angle will
better reject snow. good luck, Bruce Roe

[GeoMetry]

What sort of net metering do you have, it makes a big difference.

I will admit to being somewhat ignorant about net metering in my area. The house is going to be located in Williamsburg VA so generally not much in the way of snow to worry about. According to the Dominion Power web site it is 1:1 net metering but I think they are not telling the whole story. Just a little looking around seems to indicate that there is a system capacity limit of 25KW and some restrictions with systems as small as 15KW. Also they have some other limitations such as a state wide net metering cap of 6% of the utilities peak-load from the previous year. They also don't want me to overproduce so maybe I need to run the house for a year on a smaller array at first and upgrade after the usage is known. If I want to stay below the 25kW limit then I can probably just have the roof slope be 37 degrees because I would still have plenty of room. I would guess that 25kW is more than enough for a well constructed 4500 square foot house even with a couple EVs. I would say that I like the look of a lower pitched roof and I would think installation and maintenance would be much easier on a walkable roof.

[peakbagger]

10KW is big household system. If built using modern energy efficient techniques a house that size should be able to run off a 10 KW. Unless you need a lot of EV daily charging I would not envision going over 15 KW.

The devil is in the details with net metering. 1 to 1 is good but just as important is if there is annual or seasonal "true up" date where the utlity pays you a pittance for any power credit you may have built up. PV systems usually generate more power in the summer and carry that into the winter where less power is usually needed. Then again in your area you may have highest bills in the summer due to AC use. Ideally the new metering contract has no required true up and then you can build and maintain a surplus. If you do have true up date than you may be better off keeping the system smaller and keeping any surplus low. The worst case is if the wide net metering cap is already met and they do not allow you to net meter. New Hampshire hit a lower limit and new slar installs were held up for months until a new solar metering program came in that far more in the utilities favor.

[bcroe]

The starting point might be, just what do you want to do, then
what will accomplish that? Here the objective was primarily
heating the house thru a 42 deg Lat winter, that takes more
than a 10KW system. Bruce Roe

[J.P.M.]

There are several things to consider and they are somewhat interrelated.

1.) First off, a lot of this depends on the load the application is expected to meet and its pattern.

2.) But maybe even before that, have you considered other goals along with alternate energy ?
Sometimes, folks not familiar with dwelling design focus on one or two things (like roof angle) and make other design considerations that are as or even more important more difficult or impossible to achieve.
If you haven't done so already, get a good book on solar design or designing for climate.

3.) What's your application, solar PV or solar thermal ?
If solar flat plate thermal (or vacuum tube thermal for that matter) and depending on your local climate and site conditions (hills, trees, building obstructions, local building codes, solar irradiance patters as date/season, other stuff), you might consider a collector pitch of something like latitude + 20 degrees or so depending on those and other things. PV array orientation is another story.

4.) As a general rule, if your annual load is somewhat fixed yr./yr. and you can identify the load on something like a monthly level or tighter, the optimal array orientation is somewhat easy to determine, but that's not to say that any orientation is cast in stone.
As a matter of reality, most array orientations - even with new designs with few restrictions of any kind - are less than optimal from an energy production or cost effective standpoint (and note too, those two things are usually not the same) because of necessary design compromises or workarounds.
BTW, while most residential arrays are mounted parallel to a flat roof (note, flat isn't the same as horizontal - most roofs are flat, some are mostly horizontal, most are tilted to the horizontal), there's nothing from a design standpoint that mandates an array be parallel to a roof, or even on the roof for that matter. it's just usually easier and more cost effective - depending on the application.

5.) So, how do you decide roof tilt ?
As far as roof tilt is concerned, and again as a rough 1st approximation, an array with an approximate south facing azimuth (maybe 20 deg or so +/- true south) and tilted at latitude will produce more energy per m^2 than a lower or higher tilt, again depending on site particulars.

Once the annual system load is defined with respect to magnitude and temporal pattern, and again as a rough 1st approximation, any off-optimal orientation penalty can be estimated for preliminary design purposes by ratioing the cosigns of the considered tilts.
That means that close to south array azimuths off optimal tilt orientations will need more array area (and so cost more) for the same production as an optimally orientated array.

But - and this a big but - depending on the annual load and its temporal distribution, a 20 deg. tilt may be more cost effective if your air conditioning load is expected to me a major or the dominant component of the annual load.
Reason: the sun is higher in the sky (a larger zenith angle) when A/C is usually needed most.
Still, a lower (or, depending on the design a higher) tilt for the array (and probably the roof) will give you more area for panels, but that perhaps off-optimal tilt (and still assuming the array is parallel to the roof) will require more array area to produce as much energy as a smaller array at the more optimal orientation for no other reason than the roof tilt angle.
Start with the optimum orientation for both tilt and azimuth from a production standpoint and modify the design from there.
for any design and particulaly a new one, there will be other considerations that may make an off optimal orientation necessary of preferable.
Design is a set of compromises and a balancing act between priorities that have nothing to do with a minimum energy design.

Anyway, and only FWIW, often/usually a true south orientation is not optimal for production and in spite of the B.S. nonsense floating around like a big chunk in a cesspool of placing arrays facing west for T.O.U. considerations, a SLIGHTLY west orientation of maybe 195 to 210 deg. azimuth is better than a 270 degree azimuth. For almost any reasonably sunny location under a T.O.U. tariff that has large mid/late into evening afternoon rates a 195 - 210 degree array azimuth will yield a larger return on investment than any 270 deg. azimuth.

6.) Do this for a start, if you haven't done so already, get familiar with PVWatts after you estimate your annual load and its temporal distribution by month. Read all the info and help screens a couple of times and only then make a few runs.
Then, download and read a free PDF of "Solar Power Your Home for Dummies".
It's a bit dated but most of the info is still valid.
Then, come back here and ask questions. Chances are

Not trying to be an asshole about it, but it's not possible to give intelligent comments to your questions with just the info you've provided so far.

[GeoMetry]

Other things to specify might be to put all plumbing vents or other penetration's on the north side of the roof. I was also leaning towards a standing seam metal roof. I suppose some prewiring or at least some conduits might be very helpful. Is it worth hiring a consultant when I am confronted with those decisions? I suppose I could research and try to learn everything I need to know to get this right but sometimes it's just better to hire a pro. I don't think I'll be able to figure this out on my own after just watching a bunch of YouTube videos.

[Mike 134]

Other things to specify might be to put all plumbing vents or other penetration's on the north side of the roof. I was also leaning towards a standing seam metal roof. I suppose some prewiring or at least some conduits might be very helpful. Is it worth hiring a consultant when I am confronted with those decisions? I suppose I could research and try to learn everything I need to know to get this right but sometimes it's just better to hire a pro. I don't think I'll be able to figure this out on my own after just watching a bunch of YouTube videos.

Only if they know what they are doing and even having an engineering degree is no guarantee.

My own personal experience......contacted a Structural engineer I knew from work who had done some drawings for me when I wanted to remove a load bearing wall. He didn't do structural reviews for solar but gave me a name he'd met at a professional conference who was local. That guy (another SE) did some informal back of envelope calculations with a quick phone call, told me the truss roof would need to be reinforced yada yada yada did I want to hire him to do drawings? Just didn't seem right to me since my understanding of local codes requires a roof to carry 2- 235LB layers of shingles and my roof only has one.

Through work contacts found a firm that does solar structure reviews as their bread and butter. WOW what a difference. 7 page report detailing live loads, dead loads, wind uplift and wind pressures, expected worst case snow loads, all the code sections called out and footnotes to them. In addition, wet stamped my drawings.

Long story short no additional work needed to the roof. Good luck.

[J.P.M.]

Other things to specify might be to put all plumbing vents or other penetration's on the north side of the roof. I was also leaning towards a standing seam metal roof. I suppose some prewiring or at least some conduits might be very helpful. Is it worth hiring a consultant when I am confronted with those decisions? I suppose I could research and try to learn everything I need to know to get this right but sometimes it's just better to hire a pro. I don't think I'll be able to figure this out on my own after just watching a bunch of YouTube videos.

I'm not trying to rain on your parade but:

Are you doing this design all on your own ?

If so and for a lot of reasons, from what I've read from your posts to this thread, it sounds (reads) to me like you're already in over your head but don't know it.
So far, I see you have a footprint and a cartoon and are wondering about things that should have been addressed much earlier in the preliminary design (if you even have a preliminary design).
I'd bet you don't even have a preliminary HVAC design load yet and you're wondering about a PV system size.
I'd respectfully suggest you get some professional help with your design, not just for the PV, but for the entire project.
Anyone relying on the new idiot's bible (u-tube) for anything besides a joke or two is pretty scary.
To answer your question, I'd get someone who knows how to design structures (a P.E.) and a competent builder to do and administer the project.
Without professionals to do it, I don't see this project coming to a good end.

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