New Roof With Solar - Finding the right contractor

[Steven Fahey]

It's time for my house to get a new roof. Not only are the shingles worn out, but the skylight has developed a leak, too. Re-roofing the house would be an ideal time to find a roofer who is PV capable. Or, rather, should I be looking for a PV installer who knows how to nail down shingles? Anybody who has followed my projects will know that I've been building wind turbines and putting up solar panels around my house for years, but only at the "hobby" scale so far. The roof, on the other hand, calls for a pro.

One important factor: I live in southern Alberta. Lots of sun here, just outside of Calgary, but the electricity prices are very very low. All that coal makes electricity less than 10 cents per kWhr, you see. My electric bills are now 66% fees, 33% consumption. Weak economic incentive; instead I want to have less brown-outs and black-outs. I think I walk around the house to re-set all the clocks every 2 months or so. Lost a freezer of food. The well pump needs 240VAC or there's no water. Lots of issues like this, so I will need a competent partner who can manage all of these issues.

So far I've come up with Goose Creek, and Skyfire. So far, the other AB members on the CANSIA website are either commercial building engineers, or defunct.
Do any of you folks know anyone else I should contact?

[Rob Beckers]

Steven, so you're looking to install PV for charging batteries, with an inverter to supply power when the grid goes down?

In general, the two trades of roofing and installing PV don't overlap too much. Because of the goldrush here in Ontario with the feed-in-tariff there are a number of roofers that took up PV installation as well, but most of those I can't recommend. They are generally clueless about electrical matters and solar panels, and their installations reflect that. Most of our customers that need new shingles get a roofer to do that part first, then we come in and install PV on the new roof.

Why do you say your roof calls for a pro to install solar? If you've been making and installing wind turbines it shouldn't be all that much of a leap to install panels on the roof. Or is it a particularly difficult roof, with steep pitch etc.? If you're interested, I've helped quite a few people install PV on their roof as a DIY project (we provide materials, design, and instructions on how it all goes together). Sure, you have to be comfortable working on the roof, but other than that it is not rocket science. There are a number of details you need to get right for a good installation, and that's where advise comes in. Depending on the installation, there's electrical work that I would have an electrician do, such as replacing a meter base (something we need to do here for the FIT program). Hooking up panels is pretty easy though, the wires just click together.

Give it some thought...

Cheers,

-RoB-

[Andy Rhody]

My shingles are only 10 years old so I don't think that they would need replaced before putting up solar panels but what scares me a bit is attaching the racking to the roof without any leaks.

Is there any tricks to that?

Thanks.

[Rob Beckers]

My shingles are only 10 years old so I don't think that they would need replaced before putting up solar panels but what scares me a bit is attaching the racking to the roof without any leaks.

Is there any tricks to that?

Thanks.

Yeah, use flashing!
We use FlatJack flashing since our insurance company would take a dim view of us causing leaky roofs. There are a number of brands and types, they all work by shoving partially under the next higher row of shingles, so water cannot run under it. The L-foot attaches to the top of the flashing, see these pictures of an install (http://www.greenpowertalk.org/showthread.php?t=11638) for example.

Just to be on the extra-safe side we also use tri-polymer caulking between the foot of the flashing and the roof, where the holes are. Many installers use just caulking between L-feet and roof, and for the most part this stuff will keep those leak-free, so doing both caulking and flashing swings the odds of decades of leak-free service definitively in your direction.

By the way, tri-polymer caulking is pretty special stuff: We use something called Proflex (which also goes under the name Geocel 2300, depending on the country you're in), it's rated at 50 years of use under UV exposure. Under the flashing it won't be exposed, do this lasts essentially forever. It sticks to just about anything, including wet (!) asphalt. It also sticks very well to tools, don't ask how I know...

-RoB-

[Steven Fahey]

Hi Rob,
Thanks for all your ideas.

It's the professional roofer I need, much more than the pro PV installer. Sorry that wasn't clear enough in my OP. Last summer I replaced the roof on the garage, that took me 3 days longer than it should have, and I'm not satisfied with the result. The shingles I put on are fine, it's all the edges and details that I think could have been done better. But it's just a garage.

The house's roof is much more complicated. The south-facing slope is interrupted every 3 feet, the north-facing slope is one continuous expanse... now what? Eliminate or move the things that are in the way. Skylights need total re-thinking if I want PV or hot water up there. Other things like vents look like they could be moved or consolidated, to make even more room for PV, but I need an expert who knows right from wrong.

I am personally hoping to find a roofing "pro" so that I don't have to worry too much about the underlying structure, instead focus on the PV installation. It would be nice to have a roofer able to understand PV systems, but your opinion sounds like that could be very hard to find.

And the flipside is that even among the reputable roof-top PV installers, I shouldn't expect them to want to re-roof the house first. Is that what you're saying?

Okay, forewarned is to be four-armed.

The system is, as you summarized it, to add PV and grid-tie to an existing wind system, with battery backup. The batteries are already here; a modest stack of AGM's that I can stretch for a couple of days of furnace heat, water, lights and microwave oven, should the grid go down (not theory: actual snowstorm experience). The batteries are not currently adjacent to the house, just close enough that an umbilical was run across the yard. A permanent set up in the house, properly wired into a sub-panel, would yield more efficiency and convenience.

If you are interested in consulting on this, I'll do the internet equivalent of walking around to your "front door". There are PV suppliers in Calgary that shrugged when I visited and labelled myself a "DIYer". Then they give me the canadian tire mark-up price per watt quote on solar panels, so it doesn't seem to win me points with some guys. I am going to talk to the local PV installers first, in case they do have local roofer recommendations.

Thanks again, I will be in touch.

[Rob Beckers]

Can you post a picture of the south side of the house/roof. Let's see what you're up against...

By the way, the age of the shingles is an item on our survey checklist. We tell people that they don't want to re-roof for at least the next 20 years. So, yes, we do ask some to put new shingles on before we install PV.

-RoB-

[Steven Fahey]

You read my mind, Rob.
I can't get far enough away from the house to take the roof in one shot. From a ladder I did it in 7... but I have Photoshop, so here it all is! (reduced to a reasonable resolution I hope)

The photos on the left are the south-pointing addition, so the slopes face east and west there. Ignore that.

The photos from the center to the right face directly south, at a ridiculously shallow slope of 12 degrees (1:5) given the snow loading I've seen on it. Slope is okay for summer here at 51 degrees latitude, not so great for winter production during our 4-hour long days. Yes, there are lots of trees around the house but they are mature (not getting any taller) and far enough away that in the winter, shadows from only a few of the tallest (and bare) branches fall on the roof. The house is 19 years old - do I get a rebate on the 20 years shingles? :unsure:

Pretty cluttered area. Lots of little protrusions. The skylights started leaking. I tried patching last year but that didn't work, as evidenced by the deluge in the kitchen after a heavy hailstorm last week. Also populated by a furnace chimney, a water heater chimney (natural gas), bathroom fan exhausts, drain sewer vents, and roof space ventilation ports too.

Since it just rained 2mm today, you can see how water is shed (and sadly absorbed). The critical areas are the roof joint, and surrounding the skylight.

I have some detail shots if you want to see specific things. I think you're going to tell me it's a tough-luck case and to get a ground-mount. :laugh:

[Steven Fahey]

Yeah, that's been overly reduced during the upload process. I'll try again after supper.

[Steven Fahey]

Too big?

http://www.sparweb.ca/Forum/roof_collage.jpg

[Rob Beckers]

Steven, wow, that is a lot vents...
Am I counting 5 sewer vents alone? If you're going to re-roof it should give you a chance to clean up the roof a bit. It's easy to move the sewer vents to the other side of the roof (and they can be consolidated into a single vent as well). I'm seeing two (what looks like) bathroom vents; same deal, those can be moved to the other side as well. The mushroom vents near the ridge are actually not an issue. We put panels over them (if anything, the chimney effect between roof and panels will pull more air out of them). If that works for you depends on how you install the racking: By using flashing we elevate the L-feet for the racking by about an inch, that puts the bottom of the panels at just about 5" above the roof surface, and well above those mushroom vents (so the rail would run just below the vents). Flashing is a good idea by the way, it guarantees a leak-free installation for decades to come. You could move the vents too though, or put better vents on the north side of the roof and do away with them (mushroom vents do next to nothing when it comes to moving air).

As to the bubble-type skylights, those have terrible thermal performance. Something you may want to consider for your climate. If your heart is set on having skylights for those rooms, install (for example) Velux double-pane skylights. I have put a number of them in my own roof, they work very well. Reasonably priced (Rona sells them), no leaks, good quality. Another option would be to install light-tubes instead of skylights. They don't have better thermal performance than double-pane, but they have a much smaller footprint while still supplying copious amounts of light.

What is the goal in energy production for the PV installation? That would allow the calculation of how many panels it takes, given your roof's pitch and orientation, and that would dictate what amount of space needs to be available on the roof (and what would need to move to accomplish this).

-RoB-

[Steven Fahey]

Thanks for the ideas + suggestions, Rob.

The light tubes and/or double-paned skylights are definitely in the plan. The room lighting seems designed around skylighting, so eliminating all skylights would require additional lights to work comfortably there in the daytime (ironic, eh).

Yes I also want to consolidate the sewer vents if it would make room for un-broken spans of solar panels.

I've been able to discuss this with a PV installer. The initial reactions were that I am on the right track, as far as making provisions for the PV. One warning: since I'm probably going with asphalt shingles then the lifetime of the roof could be less than the lifetime of the PV panels. But overall it's easier for the PV mounts or racking to be installed on asphalt shingle roofs, rather than any other type, in their opinion.

They were concerned about the slope, at 12 degrees, limiting production. I must concur with that, but wind load on tilted panels is also a problem that I'm concerned about. Roof load requirements around here are 50 Lb/sqft. When I walk on the roof I wonder about the safety margins. Lots of bounce.

PV production:
I would prefer to have over 5 MWhr in a year, and in the winter, a minimum of 500 kWhr per month. That will keep the furnace blowing hot, the fridge cold, and the water pumped. The winter is the hard part. Designing for the winter makes the system dramatically over-designed for the summer (As if you don't already know all about this!)

Winter solar energy input on a horizontal surface is only about 200 W/m2, and only lasts about 4 hours per day. Also assuming panels 14% efficient and the balance of system another 90% efficient, the array needs to be 165 square meters! :nuts:

Now that I've confronted the numbers.... roof-mount may not be the way to go.

My weather station's photocell is horizontal. It sees very little sun in the winter.
My pair of PV panels on the barn are 15 degrees from vertical, and they get close to max output on sunny days in January.

An array inclined to a proper winter solstice elevation with the sun would reduce the needed array size by a factor of 5. Still too big at 33 square meters but it would furnish the needed 10-12 kWhr per day. Tracking could help, but battery-back-up is also key to stretching time without grid into multiple days.

I already have 19kWhrs of 'rated' battery storage, though to be fair it should only be used down to about 75% DOD, in a pinch, 14 kWhrs. That's equal to a full day of winter power at minimum consumption. If the array is contributing just 7 kWhrs per day, then it's possible to stretch the power during the blackout period to 2 days.

So that's the workable PV configuration, about 7 kWhrs per day or 15 square meters, if inclined near vertically. I guess I've just negotiated (with myself) my production goal down from 500 KWhr/month to just 200 kWhr/month, by considering the factors that allow me room to compromise.

I wish I'd thought this through earlier! :o

Somebody should sell me a ground-mount array quick before I get cold feet :laugh:

[Rob Beckers]

Steven, check out the PV-Watts Web site (http://rredc.nrel.gov/solar/calculators/PVWATTS/version1/), it is pretty accurate in putting production numbers to a PV array. Calgary is one of the locations you can select. For the "DC to AC derate factor" use 0.75, that will be fairly accurate for grid-tie. For off-grid this should be lower; I've not worked the numbers, likely around 0.6 (I use a different method to calculate off-grid systems). This because of the additional battery and inverter losses.

Your roof of 12 degree pitch, facing south, still does about 86% of the annual energy production of the 'perfect' roof for your location (this does not take snow cover into account though, but at that pitch there's not much to loose in winter). That's not bad. Each kW of PV would produce around 1,100 kWh of energy annually for grid-tie.

In a larger context, the problem is that it is nearly impossible to take a regular grid-tie house off the grid for any duration. People that live off-grid, and a number on this forum do, live a very different life-style. All the appliances in the house are selected on energy efficiency, they do the laundry when a sunny day is in the forecast etc. Their energy use (from what I've seen) is in the range of 125 - 350 kWh a month, with the median probably around 225 kWh/month. Compare that to a typical grid-tie house with 1000 kWh/month. You would have to work on improving energy efficiency (much cheaper than adding solar and batteries!), installing alternate sources of heat (wood stove for example) so you don't depend on the energy-hogging blower motor so much etc.

Back to your roof.
Producing serious energy in winter with solar on this roof is, well, difficult IMO. As you noted, winter production (if you can keep the snow off such a shallow-pitch roof) is just about 1/5 of summer production. It would take about 10kW to 15kW of PV to make 500 kWh in the dead of winter, and that is if the weather cooperates! A wind turbine will do much better in winter, and since you already have one of those I would focus on wind for that time of the year.

I'd set this up so you net-meter when the grid is working. Export excess to the grid in summer, and use the grid to keep the batteries charged in winter. Then when the grid goes out, whatever you can generate with solar and wind will help till you over a little longer before the batteries run low. Assuming you don't loose power all that often, there's nothing wrong with running the batteries down to 80% DOD on those (few) occasions. You're not going to be cycling them much anyway, they'll be floating 99% of the time, so their life-expectancy is not determined by cycles but simply by age (batteries will die after a number of years even if they are never used and kept at float; the acid eats away at the plates and reduces the active material available).

If you want to use PV to make energy in winter you have to get them tilted at around 60-degree pitch. That will be hard to do on the roof, and if you can't get far enough from your roof to take a picture of the whole thing, I have a feeling there's not a whole lot of space to put them on the ground either.

Assuming you can rid the roof of the many vents, how many panels (1.65m x 1m) would fit on there?

By the way, durability of shingles is not a big issue. Standard 35-year shingles last around 20 - 25 years, and you can get 45-year and even life-time shingles these days. Installing solar on asphalt shingles is definitely easier than doing so on metal roofs. Both can be done of course.

-RoB-

[Steven Fahey]

Rob,
You are right, net metering is important to make the system economical when a grid tied system
is being considered. When attempting an off-grid system, there's probably no payback in that
scenario. I'm considering a hybrid, because I have the grid, and I have the batteries, and have
the inclination to fuss with the complexity. I have had a number of grid outages that lasted for
days, and I have no reason to expect that to end. The fuss of an outage is equivalent to the
work setting up a RE system, spread over the number of years it will be in service.

My goal stated at the outset is to obtain independence from the grid, due to its fallability here,
so my main economic driver is avoidance of interruptions. Net-metering just "helps" a little. I
was afraid you wouln't pick up on that, but it seems you have.

My last posting "evolved" in the writing; I finished it about 2 hours after I started, while
poking away at a spreadsheet and my met station data between bouts of typing. I did make a
mistake in my calculations. I'll try to rectify that here.

My house's electricity usage has been reduced significantly since I moved in. Numerous appliances
have been changed, for example. Wind helps, indirectly. The estimate of 500 kWhr/month was a
"padded" number that started at a 380 kWhr figure in early stages of the calculation. Then I looked
at the assumptions and worked out that my battery back-up provides over a day's basics therefore
200 kWhr per month is enough PV production in December to get me through 2 days outage. I do
not view this as excessive, or technically difficult, nor a "bad" off-grid target. I'm not off-grid,
anyway, I'm just trying to use off-grid technology when the grid vanishes.

Since protracted cold snaps are possible, the usefulness of a monthly figure is limited when facing
periods of time that upset the average. On the other hand, a daily usage figure can be very variable,
and attempting to design for the worst day ever busts the budget again. I settled on a monthly
target, with a margin of error, based on my actual consumption. I believe you disagree with it, so
perhaps we can debate this a little further. If you do, I would rather give you more information
before you make any assumptions.

I experimented with PVWatts which confirms at least one thing: I would need 5x more panels on my
roof to achieve my winter production goal. Since the production goal is based on winter sun, it's a
complete mismatch with my roof angle.

Plugging in: 5kW array, at 75 degree incline: I get 350 kWhrs in December, but at 15 degrees incline
only 180 kWhr. A 50% loss! Doesn't meet the target either. And I haven't considered the available
surface area yet.

A 2 kW array covers about 45 square meters or 150 square feet. I'm judging that the size and output
(of a Kyocera 240W module) would require 8 units (actually only adding up to 1.92kW). There is just
BARELY enough room on my roof to fit those on, as a single rack installation, even if I did move the
vents. If I tried to pack in more panels, then 4 could be added, here there everywhere. It would be
a mess because there would be a number of isolated panels, and only gain 1 kW more.

What can I do with a 2KW array at 15 deg roof tilt? In winter that will yield just 70 kWhr per month.
Not nearly enough. It will keep the batteries at float, and sell to the grid 2MWhr in the summer, just
offsetting 25% of my actual consumption, but this drifts away from my stated goal. The Alberta utility
company charges a monthly net-meter fee that eats up the returns! I have to think of net-metering
only as a way to cover the connection fees - this is a major difference between Ontario's FIT and
Alberta's free-market rates.

Bottom line is that I started this thread to evaluate the physical installation of panels on my roof, but
I'm coming to the conclusion that it can't do what I want it to.

[Steven Fahey]

I just went back to my drawing and realized I can turn the panels around
and get 12 on in a contiguous array. Sort-of a horseshoe shape, wrapping
around the main skylight. That's upward of 3kW. Still not enough for the
winter, but plenty to exceed my consumption in the entire summer. Worth
considering, but the benefit of riding through protracted outages when
they typically happen (winter and spring) is too important to skip if I'm
going to all this trouble.

[Rob Beckers]

How about 12 on the roof (with 255W modules somewhat of a sweet spot in price, that's 3,060W total), and some more at a much steeper angle, either ground mounted or as an awning hanging off a wall. They'd need their own charge controller, but that's not a big deal. If neither one by itself gets the job done, maybe the combination will go a long way: It would add to winter production, and still have good year-around production for net-metering. Your wind turbine could add the gravy, especially during winter.

Those 12 panels will give you around 120 kWh/month in the dead of winter (if you can keep the snow off). All it would take to reach 200 kWh/month is another 4 panels mounted at a 60-degree pitch. Your goal seems near!

As you note, these are averages and actual weather can be completely different. No sun for weeks is not uncommon around here in winter. That's why almost all off-grid people I know also have a genset. Wind makes a great addition too, PV and wind are very complementary.

-RoB-

[Steven Fahey]

I like that: two smaller arrays. Each deployed differently for different seasons.
I'll noodle that one for a little while...

Yes clearing snow on nearly-horizontal panels is a chore, too.