Google

Internet Content and Information

Climate impact reported for 

2019

Carbon Change

53.2

%

CO2e increase since

CO2e decrease since

2015

Total Annual Emissions

18

 Million

Metric Tons of CO2e

Emissions Intensity

96.7

Kg

CO2e Per $1000 in Revenue

Goal

Become the first major company to operate on carbon-free energy 24 hours a day, seven days a week, 365 days a year.

What are they doing about it?

Renewable Energy Use

In 2012, we set a long-term goal to purchase enough renewable energy to match all the electricity we consume globally on an annual basis. In 2017, we achieved it, becoming the first company of our size to match our total annual electricity consumption with purchases of energy from sources like solar and wind. In 2018, we did it again. In 2019, for a third consecutive year, we matched 100% of the annual electricity used by our global operations— including offices, data centers, and networking infrastructure—with renewable energy (see Figure 7). This amounted to more than 12 million megawatt-hours (MWh) of energy in 2019 alone—more electricity than the state of Maine uses annually.

Renewable Energy Generation

As of December 2019, we’d made commitments to invest nearly $2.7 billion in renewable energy projects with an expected combined capacity of 4.6 GW. In 2020, we made the commitment to invest in and help deploy 5 GW of new clean energy by 2030 in our key supply chain regions (which includes our previous commitments of renewable energy in our key manufacturing regions), bringing our combined commitments to 8.7 GW. Once online, this 5 GW supply chain commitment will avoid global emissions equivalent to taking more than 1 million cars off the road each year and catalyze the additional investment of more than $5 billion in new wind, solar, and other clean energy technologies.

Energy Efficiency

On average, a Google data center is twice as energy efficient as a typical enterprise data center.10 Compared with five years ago, we now deliver around seven times as much computing power with the same amount of electrical power. Much of this improvement has come from new innovations with accelerators, such as our Tensor Processing Units (TPUs)—highly efficient computer chips we designed specifically for machine learning applications.