The Carbon Footprint of Email (is quite large!)

Submitted by: Phil Riebel 02/07/2017

This blog first appeared in PIWorld on February 1, 2017

How much carbon dioxide (CO2) could possibly be emitted from sending an email? A tiny puff perhaps! But when multiplied by the number of emails worldwide the carbon footprint of email becomes a significant size. A recent article in The Switch of the Washington Post[1]  draws attention to how our tech habits affect how much power is used and the environmental consequences of email.

A carbon footprint is the total amount of greenhouse gases produced to directly and indirectly support human activities, usually expressed in equivalent tons of CO2.

Establishing the exact amount of the CO2 produced by sending an email includes many variables: the energy it takes to move the email across the Internet, process it, view it, store it, reread it and, after some time, delete it. This in turn requires computers, servers and routers spread around the world that operate with differing levels of energy efficiency. It also involves in some small way, the manufacture of all that equipment.

In his 2010 book “How Bad are Bananas? The Carbon Footprint of Everything”, Mike Berners-Lee estimated that an ordinary email that you click on, open, forward or reply to emits 4 grams of CO2.  If that email has an attachment of, for example 1 MB, the CO2 goes up to 19 g and if that attachment is then forwarded or filed, the footprint for that one email could be as high as 50g! If the email is spam and blocked by a filter, then the number is only 0.3 g[2]. Every email has a unique footprint that reflects its size and the number of times it is moved around.

The footprint also reflects the kind of energy used by the infrastructure which allows that email to be sent. Data centers, for example, store our emails along with all their attachments, photos etc.[3] In 2014, data centers in the U.S. consumed an estimated 70 billion kWh or about 1.8% of total U.S. electricity consumption. Data centers require a lot of electricity because rooms full of servers generate heat that must be cooled if the servers are to be kept running. This cooling comes from large air conditioning systems which vary widely in their source of energy and efficiency; it could be coal-fired power plants or renewable energy (like hydro, wind or solar) or some mix of both. This affects the footprint of the email that travels through that data center.

But should we be worried about the carbon footprint of a few emails? Well, it is not just a few! According to the 2015 report on email statistics by The Radicati Group, a technology market research firm,[4]  the number of emails sent and received worldwide per day is estimated at 205 billion. This figure is expected to grow at an average annual rate of 3% and reach about 246 billion by the end of 2019. This means almost 2.4 million emails are sent every second and some 74 trillion emails are sent per year. 

Assuming all emails emit the lowest estimate (0.3 g CO2 each), the total worldwide CO2 generated by emails would be 22 million metric tons of CO2 per year. This is equivalent to the amount of greenhouse gases produced by almost 5 million cars. If all emails emit the highest estimate (50 g of CO2 per email), roughly 4 billion metric tons of CO2 would be generated each year by emails sent around the world. This is equivalent to the CO2 emitted annually by 890 million cars. The answer no doubt, lies somewhere in between.

What is not in doubt however, is that one email may indeed produce an insignificant amount of CO2 but when all those tiny footprints are measured at a global scale, the footprint becomes astonishing in size. The government of France has already noticed this and has recently released a list of ‘écogestes’ included a suggestion to ‘avoid sending numerous emails with attachments’ during peak demands for electricity. [5]

As our dependence on email and other aspects of the digital economy grows more extensive - and it will—we need to be aware of the energy demands that will follow.


[1] Tsukayama, 2017

[2] McAfee, 2009

[3] Shehabi et al. 2016

[4] Radicati Group, 2015

[5] RTE, 2016

Source: Printing Impressions