Data, an activity with a high environmental impact
An underestimated environmental footprint
The digital sector already accounts for between 2% and 4% of global greenhouse gas emissions.
Today, many companies focus their CSR strategies on actions such as limiting air travel or reducing waste. Although key, these strategies could become insufficient in the face of the growing environmental impact of other activities, such as data management. So it’s important to take stock of our companies’ priorities, and the bonus or malus that each of our actions can represent.
The digital sector already accounts for between 2% and 4% of global greenhouse gas emissions. By way of comparison, civil aviation accounts for 2% to 3%. The transfer and, above all, storage of data, as well as the manufacture of the necessary electronic systems, are energy-intensive processes: it has been estimated that the total energy consumption of data centers in 2018 was 205 TWh, or 1% of global energy consumption1. At the same time, a second study shows that the amount of digital data created or replicated per year worldwide could be multiplied by almost 4.5 between 2018 and 2025. If nothing is done, the environmental impact of data could well explode in the coming years – in France, it is estimated that digital technology will be responsible for 7% of greenhouse gas emissions by 2040.
- Despite an environmental footprint that can no longer be overlooked, data remains “under the radar” and its impact underestimated. There are three possible reasons for this:
The illusion of the immateriality of data and its low financial cost, which leads to exaggerated data generation;
- Storage practices that consume too much energy and are ill-suited to current uses;
- A lack of familiarity with Big Data technologies.
Fortunately, there are solutions that could help to limit the environmental footprint of data, and thus curb or even reverse this trend.
The illusion of dematerialized data
To dematerialize is to materialize differently (Guillaume Pitron)
Since the early 2000s, the vast majority of organizations have sought to replace their physical archives with digital ones. Dematerialized, more easily exploitable and above all simpler and more economical to store, digital appeared to be the ideal solution. It was even seen as more “eco-friendly” than printing on paper.
We have now entered a paradigm in which data management is guided by the maxim “who can do more, can do less”, and in which we are building gigantic datalakes services without moderation and sometimes without any a priori defined goal. We seek to maximize what some still call “the new oil” (irony of fate), not measuring the real impact of these immaterial flows. But “to dematerialize is to materialize differently” (Guillaume Pitron): big data is physical, and it pollutes. To illustrate this, let’s take a simple, everyday example: videoconferencing. Taking part in a one-hour videoconference with two other people, on a platform such as Zoom, Teams or Google Meet, in standard quality, causes an average transfer of 750 MB of data per person. In total, this meeting will emit 55 gCO2eq, roughly as much as a 500 m car journey. These figures increase with screen sharing, and are multiplied by 2 or 3 when the video is broadcast in higher quality (720p or 1080p). Every day, hundreds of millions of people take part in at least one videoconference, especially since 2020. Collectively, this equates to a carbon footprint of around 15 million tCO2eq, or roughly the impact of 15 million passengers making a round-trip plane journey from Paris to New York every day. While telecommuting has the potential to be more responsible than our traditional methods, its gains can therefore disappear if we ignore certain good practices (such as preferring the telephone when convenient).
In fact, the conversion rate between gigabytes of data transferred and grams of CO2e depends on the use made of this data and the means of transfer and storage (due to less efficient data compression and energy-intensive storage, an e-mail could emit 190 times more CO2eq per MB than videoconferencing, for example). But whatever the case, this quick calculation helps us to understand that the data we generate – whether it’s a by-product of our tools, as in the example above, or a genuine asset for our company – does indeed emit greenhouse gases, on a daily basis; and not only when it’s transferred, but also and above all when it’s stored, as we’ll see in the next part of our series.
2 levers to limit the volume of unnecessary data in your company
To limit the impact of data in our companies, the first transformation we need to make concerns our mindset and habits. It is important to adopt sustainable data management today, by acting on two simple levers:
1st lever: limit the generation and storage of non-essential data
The aim here is not to advocate limiting data generation altogether. Indeed, we must emphasize here that Big Data is an opportunity and can be a major driver of innovation and research, for example in the pharmaceutical world, where data generated during clinical trials can be used to discover new treatments or better understand how our bodies work. The challenge here is to get everyone to ask themselves, first and foremost, whether the data they generate is useful; whether it will be in the future; whether it needs to be stored; and, if so, for how long. Integrating these simple questions into our decision-making processes could already considerably limit the impact of digital technology on our businesses, as today we all too easily store an excessive volume of data that does not meet these criteria.
Similarly, the implementation of certain good daily practices (for example, the GreenIT association recommends regular archiving campaigns, among a host of other simple and effective initiatives) and awareness campaigns aimed at our employees could have a strong impact. To take things a step further, we could also try to limit the number of electronic devices we use on a daily basis, or prefer to use the telephone when it’s convenient, for example.
2nd lever: limit superfluous duplication of data
Duplicating data may have a security value, and may therefore be necessary. But every time we duplicate a document, for whatever reason, we also double its environmental impact, since each new MB stored will consume energy every day! So let’s avoid duplicating versions of the same document, or saving multiple copies of the same data in different places.
Digital technology has become a priority environmental issue, and its impact is growing exponentially. Digital sobriety in our day-to-day actions is a simple and effective way to act now and slow down, or even reverse, this trend, by limiting the volumes of data stored to the bare minimum. In the following sections, we’ll explore other levers that can be activated in your company, starting with data storage optimization and how it can help you achieve a more sober approach. Alcimed is at your side to help you with your digital sobriety strategy. Don’t hesitate to contact our team!
 Deluzarche, C. Global warming: the digital sector generates more greenhouse gases than aviation. Futura.
 Gaudiaut, T. (2021). The Big Bang of Big Data. Statista Infographies.
 Sénat (2020). Mission d’information sur l’empreinte environnementale du numérique.
 Singh, R. (2021). How Much Data is Consumed for 1-Hour of Video Conferencing on Zoom, Google Meet, Skype, Microsoft Teams, Slack and Hangouts?
 Derudder, K. (2022). Which video conferencing mobile application to reduce your impact? 2021 Edition. Greenspector.
 The Carbon Footprint of Email vs Postal Mail. (2021). Eco2 Greetings.
About the author,
Matthieu, Consultant in the Alcimed’s Data team in USA