HomeDigitalizationBig DataGoogle's big data calculates US rooftop solar potential

Google’s big data calculates US rooftop solar potential

In an application of big-data capabilities to the decentralized energy sector, a project by Google has found that almost 80% of rooftops in the US are suitable for solar systems.

Since its inception in 2015, the company’s Project Sunroof has analyzed around 60 million buildings in all 50 US states, determining overall that 79% have enough unshaded area to install photovoltaic (PV) panels.

In sunnier states such as Hawaii, Arizona, Nevada and New Mexico, the analysis found that over 90% of rooftops could support PV, while rooftops in more northerly states such as Pennsylvania, Maine and Minnesota are only around 60% suitable. à‚ 

Among cities, Houston in Texas has the most solar potential, with an estimated 18,940 GWh of rooftop solar generation potential per year. Other sunny cities such as Los Angeles in California, Phoenix in Arizona and San Antonio in Texas follow in the rankings, with northern and often snow-bound (yet roof-plentiful) New York City in fifth place.

Google said the Project Sunroof tool uses imagery from its Google Maps and Google Earth in combination with 3D modelling and machine learning. For every building included in the data, Project Sunroof calculates the amount of sunlight received by each portion of its roof over the course of a year, taking into account weather patterns, the sun’s position in the sky at different times of the year, and shade from trees and tall buildings. This estimated sunlight is translated into energy production using industry standard models for solar installation performance.

According to Google, if the top 10 cities reached their full rooftop solar potential, they could produce enough energy to power eight million homes across the US.