Google's Project Sunroof Tells You How Well Solar Would Work On Your Roof

An anonymous reader writes: Google's Project Sunroof aims to make the task of installing solar panels easier by providing financial advice and stats on what solar energy could do for you. The project is only available in San Francisco, Boston, and Fresno for now. Techcrunch reports: "To get started, you simply plug in your address and some data about your monthly electricity bill, and the tool will tell you what the recommended solar installation size is and how much it would cost to buy or lease the hardware. In case you want to go ahead with a solar install, the tool also lets you reach out to local solar providers. Google says these listings are sponsored, so chances are it'll get a bit of a kickback when it generates a sales lead for these companies."

Comments

Not available yet...

[Firethorn]

Darn it: "Sorry, Project Sunroof hasn't reached this address yet."

Can't say much about it then.

Re:Not available yet...

[thinkwaitfast]

Try pvwatts. It's very accurate in my experience.

http://pvwatts.nrel.gov/

Strange limitations

[pla]

Annual insolation, even after considering weather, counts as a well-documented stat across the entire US. Why would they limit this to just a few key cities?

Google says these listings are sponsored, so chances are it'll get a bit of a kickback when it generates a sales lead for these companies.

Oh, riiight! "We don't have any partners outside those cities yet, so the rest of you can go fuck yourselves". Got it.

Re:Strange limitations

[peragrin]

Local, state and federal subsides vary drastically only a few miles apart. That database would tax googles servers.

Re:Strange limitations

[bondsbw]

If you aren't willing to pay for alternative energy, then don't be surprised when you don't get it. Google helps advance that cause and all you can do is complain that they are a business.

Re:Strange limitations

[Mr D from 63]

Local, state and federal subsides vary drastically only a few miles apart. That database would tax googles servers.

That's right. The subsidies are required to make this look good. It is, after all, sponsored by installers. The calculator naturally assumes everything is optimum, and doesn't get very detailed when it comes to additional expenses that may be incurred during installation. It also assumes 2.2% per year rate increases, which is not necessarily going to happen.

I did a spot check on their Redwood Ca numbers. They calculate using solar insolation equivalent to 5 hrs/day, but real numbers are about 4.7 hrs/day.

Re:Strange limitations

[blueg3]

Shade, local terrain, building codes, subsidies, power company buyback policies and rates. There are also a lot of odd business arrangements that are localized that can dramatically reduce the solar capital cost.

I have no idea if they account for any of these factors, but there are certainly a lot more factors than weather-adjusted annual insolation.

Re:Strange limitations

[hawguy]

Annual insolation, even after considering weather, counts as a well-documented stat across the entire US. Why would they limit this to just a few key cities? .

I think it's because they do additional processing to take into account shadows and roof slopes to better estimate the viability of solar on a building. It pretty accurately shows the that north half of my roof is in shadow, and it captures the shadow from the large building to my south that hits part of my roof.

Re:Strange limitations

[PopeRatzo]

Oh, riiight! "We don't have any partners outside those cities yet, so the rest of you can go fuck yourselves". Got it.

Because Google owes us, big time.

Re:Strange limitations

[swillden]

Annual insolation, even after considering weather, counts as a well-documented stat across the entire US. Why would they limit this to just a few key cities?

Because this provides dramatically more detail than regional average insolation. It tells you how much insolation each portion of your roof receives, accounting for local geography, flora and other buildings. That takes some moderately-detailed 3D models and heavy number crunching. The 3D models come from Google's project to build 3D models of all population centers using low-flying aircraft with angled cameras, so Sunroof will only be available in regions where the models are available (zoom in in Google Maps in your area to see if it's already 3D-ifiied) and even then it will take time to crunch all the data.

Re:Strange limitations

[theIsovist]

Having seen this technology presented at DIVA Day last year, the difference is that this technology combines the well known annual insolation data with lidar data, so that you know "exactly" how much solar radiation is falling on a specific roof surface. It's a simple trick, but a clever one that no one has done so far. Google's data should help expand this database pretty quickly.

Might as well ...

[PPH]

... just contact a local system installer. The insolation data for a particular region is already known and publicly available. What will affect your particular system are things like local shading, roof pitch and orientation and cost of installation as affected by your house and lot particulars. Local installers will also be familiar with your utilities solar programs.

Re:Might as well ...

[Anonymous Coward]

Yeah, but the most important number is the cost of electricity, which is _assumed_ to hold constant, rather than go up in price. Given the shit to natural gas which is historically cheap right now, and the difficulty of going back to coal for baseload, that's not a very good assumption on balance.

We opted not to get solar, but we live in San Francisco. Our home doesn't even have an A/C, just an old furnace. The one thing that almost persuaded us was that if we got solar this year we'd be locked into a favorable, 20-year contract with PG&E which would require them to purchase our excess capacity at market prices. That program will end sometime this year or next, when the number of participants hits the statutory cap.

Here in California electricity prices are relatively low. They've remained constant for almost a decade, and are just beginning their inexorable rise as PG&E begins the decades-long replacement of ancient infrastructure. If we lived outside San Francisco, especially in South Bay or East Bay in a home with A/C, we'd get solar without hesitation.

As we live in SF, and our biggest expense is for gas for the furnace during the winters and spring, the risk just wasn't worth it. Our average electricity bill is less than $75/month, and our peak electricity use in the winter is at night (when we use a space heater upstairs). So our exposure to increased prices isn't that great. Our payback at current prices was in 20 years. However, it would have increased the value of our home substantially (because of the "eco" factor, because of the PG&E lock-in), so including increased equity the payback would have been almost immediate. But we have other, higher priority home improvements to worry about right now. Solar installation contractors are almost universally shady; it's exceptionally difficult to find a reputable one. I just didn't like the idea them causing damage to our roof, however minimal the risk, given our current budget.

Re:This already Exists

[hawguy]

There is already a the PVWatts calculator at on NREL's website. You input your location, the type and placement of the solar panels and it tells you how much power to expect based on local weather measurements. Since these are the people gathering the data, I can't imagine google's project does anything than access this same database.

http://pvwatts.nrel.gov/

The NREL site just looks at location and weather to estimate solar availability -- The Google site attempts to calculate usable roof area and take into account shading from nearby trees and other structures.

Real numbers

[FlyHelicopters]

It will cost, give or take, about $40K to install a 10 kilowatt system on my home.

Returning the 30% federal tax credit back to me puts me at a cost of $28K.

Based on my location in Texas and my current utility rate (total cost) and that I have net-metering, I'll save about $1,400 a year in electricity with such a system.

That puts me at an even 20 year payback period. Now, in fairness, electric rates are not likely to stay the same, adding solar does add something to my home's value, so there is that.

Lets say that electric rates will rise with the rate of inflation, which the government currently says is nearly zero, but will probably rise, then add something to the value of my home, and you get about a 12 year payback period, if you use numbers that favor solar and 17 years if you don't.

What those numbers DON'T take into account is the loss of net-metering, which is a real risk. If too many people go to solar, it will have to go away. If everyone installed solar on their roof and ended up with no electric bill, the power companies would go out of business. Clearly they would actually go out of business long before then, maybe at 20%. You can talk about batteries all you want, but the reality is they likely will get lawmakers to remove net-metering before then.

Regardless, it is a terrible investment, it makes no sense whatsoever from a financial point of view, at least for me. If you pay more than I do for power, then it might make sense for you. I have family in Australia who recently installed solar because they pay more than 25 cents per kWh, so the numbers are quite different there.

Re:Real numbers

[craighansen]

Sunroof doesn't operate in Texas, and perhaps because it's not worth it due to the structure of electricity rates. In California, we've got tiered rates that punish high usage - buying your electricity in Costco-sized bundles costs more than buying it in convenience-store tiny bundles. The statewide solar initiatives (CSI) is done, kaput, played out, but used to pay as much at $2.50/installed-watt or 39c/kwh - but it ended at 20c/installed-watt or 2.5c/kwh. The 30% federal tax credit's the sole big remaining inventive, until the end of 2016.

Here in the Bay Area, one of the few regions where Sunroof actually is operating, residential electric rates start at 16c/kwh (Tier 1), but rise to 19c, 28c, and 34c (Tier 4) as your usage increases over "baseline". If you size your system to knock out Tier 3 & 4 usage via net metering, the payout's much quicker.

If you combine net metering with time-of-use metering, the payout time can be even sooner, as the Weekday (Monday-Friday), "summer" (May-October), Tier 4 rate reaches 49c/kwh at times of high demand and 38c/kwh at medium demand. The high demand period runs 1pm-to-7pm, and medium-demand runs 10am-to-1pm and 7pm-to-9pm, so the sweet-spot daylight times for solar are generally net-metered at medium-demand and high-demand rates. Notably, this means that west-facing solar panels get a sweeter payback than either east-facing or south-facing on a typically-sloped roof.

As a rough figure, this probably puts payback times into the 3-to-7-year range, depending on cost of installation and orientation of the panels. It takes a lot longer to break even if you're trying to zero out your utility bill, because if you try to drive your net-usage all the way to zero, you're net-metering down at the 16c level, actually 13c during "winter" season (November-April). If you oversize your system and try to get a net payment from the utility, they only pay about 3c/kwh for excess power.

As an aside, I used this opportunity to check out the TOU rates to see what the average rate is for, say, a continually-running server, and it looks like the TOU schedule is 0.78c/kwh lower than the standard tiered rate plan. For me, that was one one of the questions to consider whether going for a TOU plan was going to hurt my bill.

For other stereotypical Californians (not me, not me, it's for a friend, really), a grow lights for ...umm... plants might be a similar issue - but typical schedules for grow lights & heat cycles are 12 hours/day - perhaps you can reverse day for night, and grow your plants with off-peak power FTW? Or supplement with actual sunlight during peak hours?

Re:This already Exists

[FlyHelicopters]

Once you factor in things like government subsidies, solar make economic sense for almost everybody, no matter how small your budget. A 20 year pay off is still a 20 year pay off, period.

You have an interesting view of economics.

A 20 year payback might as well be forever... The same money could be better spent in many other things. The money isn't either spent on solar, or nothing...

For example, we replaced our 11 year old 13 SEER HVAC with a 16 SEER dual speed, dual stage unit. It was $18K including some ductwork for a pair of TRANE units, a 5 ton and a 3 ton. The payback period is 10 years on that, less if the cost of power goes up. My electric bill 2 years ago in August was $700, this year it was $500. In the winter it saves a ton of natural gas on heat, not quite as much, but it is a good amount.

But the thing is, the old unit was broken and needed $3,500 in repairs, so it wasn't really $18K for a new one, it was $13K for a new one, $1,500 for some needed ductwork, and the rest in money that was going to be spent anyway.

It isn't a matter of "spend money on solar or have no power", we have power now. Without an HVAC, we'd have no AC, so the math works differently there. 10 year payback is pretty darn good considering we need it anyway and the old one was only going to get MORE OLD.

----

Regarding the "20 year payback is a 20 year payback" comment, few people stay in their house for 20 years anymore, if you sell in 5 years, you don't get the 20 year payback. Sure, sure, you say it increases the value of your home, and it probably does somewhat, but that also depends on your market.

Where I live, no one has solar. Or let me be more clear, in a city of 250,000 people, about 150 of them have solar. I have actually NEVER seen solar on anyone's roof, ever. It is that rare around here. So having it doesn't mean as much for property value as it probably does in San Fran (which I'll grant you, probably does do more there).

It simply doesn't make sense for most people, the fact that you do tells me that you see the world quite differently than most people do. Which is fine, but you might want to consider that your blanket statement... might be in error...

Not enough data

[RogueWarrior65]

The only way that this would be accurate is if it looked at the satellite imagery throughout the day and all year long because shadows.