In 2013 our net carbon footprint was zero. Before offsets, our gross 2013 carbon footprint was 1,766,014 metric tons of CO2e including green power purchase agreements (Scope 1: 41,373, Scope 2: 1, 245, 253, Scope 3: 479,388) and 2,351,902 metric tons of CO2e excluding green power purchase agreements (Scope 1: 41,373, Scope 2: 1,831,142, Scope 3: 479,388).
Per industry guidance, as of 2012, we deduct our green power purchases from our footprint, which accounted for the drop in emission from 2011 to 2012. We've also increased our renewable electricity purchases. Starting in 2013, we now report our greenhouse gas emission inventory both with and without green power purchase agreements.
Our carbon intensity values include Scope 1 & Scope 2 emissions, where as previously, they included Scope 1, Scope 2, and Scope 3. We made this change to be consistent with the CDP definition of carbon intensity, which is "Scope 1 and 2 emissions for the reporting year in metric tonnes CO2e per unit currency total revenue.
This calculation includes three components: renewable electricity native to the grid, renewable electricity from green power purchase agreements and renewable electricity from on-site facilities. To calculate renewable power from the grid, we multiplied the electricity consumption of our various offices and data centers around the world by the percentage of renewables in that location. For US locations (excluding those associated with green power purchase agreements), our grid renewables data are from the EPA’s eGRID. For non-US locations, our grid renewables data are from the International Energy Agency. For our green power purchase agreements, we assume they are 100% renewable. Similarly, we assume our on-site renewable power generation is 100% renewable. We sum these three totals and divide them by our total power consumption, including on-site renewables.
In 2014, Google’s long-term agreements for renewable power produced 1,056,433 MWh. These contracts covered 28% of our total electricity consumption. We also obtain green power from the grid and on-site renewables, making the total share of renewables in our mix over 37%.
We have made agreements to fund nearly $2.5 billion in renewable energy projects. The total nameplate capacity of all our renewable energy project investments is over 3.7 GW. These projects are expected to generate over 11 billion kWh annually—far more electricity than we consume (although we don’t purchase or consume this electricity ourselves). To provide context, this electricity is equivalent to that consumed by approximately 1,000,000 US homes. This assumes that the average home uses 10,932 kWh per year, which comes from the US EIA. Here are a few other ways to view this power in everyday terms:
We are actively searching out and signing new deals for renewable energy investment projects. Note that our renewable energy project investments are separate from our green power purchase agreements, as discussed in the previous question.
Data centers use energy to run the facility as well as to power the servers that do the work. Most data centers use almost as much non-computing or “overhead” energy (like cooling and power conversion) as they do to power their servers. At Google, we’ve reduced this overhead to only 12%. Additionally, we reduce the energy our servers use by designing our own servers, which we have optimized for the work they need to do. We’ve pulled out the components we don’t need, such as peripheral connectors and video cards. We've also efficiently redesigned the components we do need, like power supplies, voltage regulators modules and backup power systems. Combining our efficient servers and facilities, our data centers use 50% less energy than a typical data center.
We used an average commute distance for Googlers, the U.S. Department of Transportation's average fuel economy estimate of 23.5 miles per gallon and the Greenhouse Gas Protocol's estimated carbon content of 8.81 kg CO2e per US gallon of gasoline.
We estimate that Google uses the CO2e equivalent of driving one mile to serve an active user for one month. To calculate this, we used the following assumptions:
To estimate the savings from Google Apps for Enterprise, we took our active users of this product and estimated the emissions before and after switching to Google Apps. We used the public Cloud Energy and Emissions Research Model developed by Lawrence Berkeley National Laboratory and Northwestern University. Except for the number of users, we used the same assumptions as those described in this paper.
We produce and track a small amount (less than 1% of our footprint) of biogenic emissions—emissions from biologically based sources like plants, animals or waste. These emissions are released via the landfill gas cogeneration facility we operate on our Mountain View campus and the biodiesel we use in our shuttles. Biogenic emissions are often excluded from an entity's compliance footprint (e.g. these California regulations) because they are part of the natural carbon cycle. The natural carbon cycle takes carbon out of the atmosphere on the front end (as opposed to using fossil fuels) and results in emissions that would have happened anyway. For example, the landfill gas we use to generate power would have otherwise been flared, resulting in the same emissions. Or it would have vented directly as methane which would have an even higher greenhouse gas impact. While we track our biogenic emissions, we do not include them in our footprint.