Reaching our solar potential, one rooftop at a time
The sun produces an astonishing amount of energy. In one second it generates enough to meet the needs of civilization for 500,000 years. More important, enough sunshine hits the earth in a single hour to power the global economy for a full year. Unfortunately, only a tiny fraction of that energy is captured. If all the suitable rooftops in the U.S. had solar panels, they could supply 39% of the nation’s energy needs. In short, humanity is sitting on — or rather, under — some massive potential with solar energy.
Why is solar still so undervalued? It’s partly an infrastructure problem; the traditional power grid isn’t ideally suited to renewable power sources. It’s a marketing problem as well: Way too many people still think of solar for their home as an expensive luxury when in fact solar production costs are at an all-time low, and behind-the-meter solar is now often cheaper than grid power.
But ultimately it’s an information problem. For most people, the road to solar power is paved with questions: about affordability, weather and light patterns, usable roof area, angle and tilt, government incentives. The information is scattered across the web, from databases by the U.S. Department of Energy to solar suppliers’ websites. Going solar, in other words, isn’t anywhere near as easy as it should be, and could be.
But several years ago Carl Elkin, an engineer in Google’s Cambridge, MA, office, had a realization: If you pulled together a bunch of different data streams and did a bit of tricky math, you could produce, for any given address, a solid estimate of how much solar energy that rooftop could provide.
The idea started with the rooftop satellite imagery displayed in Google Earth: Those shots provide a digital surface model showing the direction a roof faces (south or southwest exposure is best), the angle of its tilt, and the presence of shady objects like trees. From there you can model how much sunlight hits a rooftop surface by tracking the light through the day, using 3-D geometry. Add in data about weather patterns, calculate the averages over the course of a year, convert from sunlight to kilowatt-hours, and boom: a baseline estimate of that roof’s solar potential. From there, it’s just a few extra calculations to provide a cost estimate customized to that address.
Elkin shared the idea with colleagues. Dozens of people in five offices volunteered to pursue it. “People started coming out of the woodwork,” Elkin says. As the project team evolved the tool, other ideas were added. Sunroof averages 20 different energy usage scenarios, and people can input their exact costs and usage to make it even more accurate. The team also employed machine learning to help Sunroof do things like better distinguish roofs from trees and measure light from every area of the sky.
Sunroof processes roughly 1 petabyte (1,000 terabytes) of data: height and color for 43 million homes; weather information; about 1,000 state and local incentives; and hundreds of local electricity rates.
Over the past 3 years, Sunroof has grown from a part-time project to a full-time job for Elkin and his team. Initially launched to drive consumer awareness and education, the service now also makes it easy for interested homeowners to connect with solar providers in their area. Sunroof covers 43 million rooftops in the U.S. — which is more than 50% of all households — and in the coming months will be available in all 50 states.
he team is also exploring international expansion and recently launched Data Explorer, a tool that gives researchers, community advocates and local policy makers access to more aggregated data of solar potential to help them make the case for larger solar deployments at the state, county, city, and neighborhood levels.
One of the industry’s biggest financial impediments has actually been customer acquisition cost; competition is so tight that providers spend almost 44% of the cost of an installation just to win new contracts. To offset this problem, Sunroof gives referrals to solar providers for free. We can’t think of a better way to help let the sun shine in.
Google’s “Project Sunroof”
Google developed „Project Sunroof“ by connecting Google map that provides interactive maps and satellite view with global solar map that provides average annual GHI (global horizontal irradiance). Solar map is a subset of a larger global solar dataset created by company 3TIER, specialized in assessment and forecasting of renewable energy. Averages are based on more than 10 years of hourly GHI data derived from visible satellite imagery using a combination of in-house research and algorithms published in peer-reviewed scientific journals.
How the project works? First step is entry of home address as base for personalized roof analysis. Project Sunroof computes how much sunlight hits the roof taking into account Google’s database of aerial imagery and maps, 3D roof modeling, shadows cast by nearby structures and trees, all possible sun positions over the year, historical cloud and temperature patterns that might affect solar energy production. Second step is data entry for calculation of solar solution. Project Sunroof recommends an installation size to generate close to 100% of home electricity use, based on roof size, the amount of sun hitting the roof and amount of electricity bill. When discussing solar installation with your provider, the recommended installation size (in kW) is a good starting point for a more fine-tuned estimate of total costs and benefits. Third step is computing of your savings. Project Sunroof uses current solar industry pricing data to run the numbers on leasing, taking a loan or buying solar panels for your house to help you choose what is best for you. It also compiles tax state tax credits, utility rebates, renewable energy credits and net metering to calculate your final costs. The last step is the selection of the installer. The system offers the installers serving your area and those you select get information for making an offer.
The project is still in its early phase with some room for improvement, for example taking into account certain elements such as roof’s inclination or different amounts of shade at different times during the day. So far, the roof exposure to the direct sun can be analyzed by citizens of Boston, the San Francisco Bay area and Fresno. If the prediction turns out to be correct, the service might be available all over the U.S. rather soon, with potential to be spread worldwide. Although we don’t know when the Project Sunroof will be applicable in Europe, G-Solar welcomes this project for several reasons: it is simple to use, it raises the awareness on renewable energy and, most importantly, supports end users by estimating cost effectiveness of PV installation.
Google’s Project Sunroof will peer pressure you to get solar panels
You can now check if your neighbors have installed solar panels on their rooftops, thanks to a new feature in Google’s Project Sunroof. The free online tool has been around for a couple years, allowing users to check whether installing solar panels could cut their energy bill. Now, if you zoom in on the satellite map, a red dot will indicate homes and buildings that have solar panels.
Project Sunroof launched in 2015, even expanding outside the US earlier this year. The tool tells you how much sunlight hits your rooftop, how much you could save by installing solar panels, and gives you information on local panel providers to do the installation. The new feature now tells you whether your neighbors are already harnessing the power of the Sun
The information is gathered by a machine learning algorithm that’s trained to recognize solar panels from satellite images. For now, the information is available on 60 million buildings in the US, but Google wants to add 40 million buildings in the next few years, according to The Atlantic.
The goal of the new feature is to encourage more people to take the plunge. Studies have shown that people are more likely to install solar panels if their neighbors have done so, especially if the panels are visible from the street. That information is now just one click away.
Google Launches Project Sunroof Tool In UK
Teams up with E.ON for tool to work out cost savings for individual roof structures fitted with solar panels
Google has teamed up with energy provider E.ON to launch its Project Sunroof online tool in the United Kingdom.
The tool assists homeowners work out if its worth them installing solar panels, by inputting detailed analysis on Google’s “Solar Calculator”. Google is also working with German software firm Tetraeder on the project.
This tool (found here) estimates uses Google Maps to model how much sunlight will hit a given property, and also estimates how much space there is for solar panels, and the projected cost savings. The tool was initially launched in the United States back in 2015.
Essentially, customers have to enter their exact address details, as well as other other data such as roof area, pitch (angle) etc.
Customers also have to enter thier contact details, so expect a follow up sales call.
The tool then uses machine learning to estimate how much solar potential a house has by examining the property’s features and weather data, such as sun positioning.
The tool is not unique, as other firms also have similar tools such as Ikea (in collaboration with Solarcentury), and Tesla’s own Solar Roof Calculator.
It should be remembered that the UK as a country is not ideally suited to solar as initial cost of installing solar panels can still be fairly high (typically between £4,000-£6,000 for an 3kW unit to power a family home), and the fact that the UK generally tends to suffer from a great deal of cloud cover (which means that solar panels don’t tend to work quite as well as in full sunshine).
Solar panels also do not produce constant electricity and only provide a house with electricity during daylight hours – unless an expensive home battery, such as the Telsa Powerwall (costing at least £5,500), is utilised.
But Google, which is based in the much sunnier climate of California, is a big backer of solar energy.
For example in 2015 it invested $300m (£195m) in a $750m (£486m) SolarCity fund to finance the installation of solar panels onto the roof’s of residential properties.
Besides residential roofs, Google has also invested heavily in solar farms across the world including a solar photovoltaic power plant in Germany, as well a solar energy plant in California’s Mojave Desert.
In 2017 Google revealed that its global operations were powered using only renewable energy sources.