On Monday, during a public event and as he had done on other occasions, the Minister of Ecological Transition Roberto Cingolani spoke about the environmental impact of the “digital sector”, which “produces four percent of the planetary carbon dioxide”. In his speech, Cingolani cited various online behaviors such as posting and uploading photos, using social media and sending emails, describing them as emitters: “To give you an idea, all air traffic produces two percent of global CO2. “.
Cingolani – who is a researcher active precisely in the digital sector – said that obviously the importance and centrality of the internet are known, but that it should be used with more “sobriety”: when you reply to an email “send the link instead of the 'attachment (the attachment, ed) and in case some useless post, spare it “, he said.
The issue of internet emissions has long been debated, and it became even more so after the coronavirus pandemic, when countless people around the world were forced to spend much more time indoors due to restrictions, and various online services. , such as video streaming, have seen a sharp increase in usage. It is also an extremely complex issue. As always happens when we talk about this kind of thing, any attempt to calculate how many emissions a certain sector produces always involves estimates, which can have quite large variations.
Cingolani's quoted estimate that the internet would produce 4 per cent of global emissions probably comes from a March 2019 study published by the French research center The Shift Project, according to which emissions caused by the digital sector were 3.7 per cent of total in 2015 and 4 per cent in 2020.
However, this estimate has been contested, for example by Carlo Canepa in Repubblica, because among other things the study of The Shift Project has not been subjected to the peer review process (“peer review”), that is the mechanism according to which the results of a scientific research must be verified by scientists unrelated to that study to prove its reliability, before being published.
If we look at the research published in scientific journals, we note however that the studies that deal with the emissions produced by the digital sector are exceptionally few, and that their results show an extreme variability.
For example, an article published in Patterns magazine in September notes how major studies published in recent years have made estimates of emissions from the digital sector (or more precisely the ICT sector, which includes telecommunications systems, computers and technologies that enable digital communications) ranging from 1.8 percent to 6.3 percent of total emissions. The authors of the Patterns study believe that a probable estimate is 1.8–2.8 per cent, which however rises to 2.1–3.9 per cent if the emissions produced by the supply chain are also taken into consideration. “(Literally” supply chain “) of the technology sector.
The studies published in recent years are also very conflicting when it comes to identifying the sources of emissions in the digital sector.
Everyone agrees that the main sources are three: first of all the devices used by users, such as computers and smartphones (and there is much debate as to whether or not to also include connected TVs); then the data centers, that is the places that host the servers that contain the data of the websites and the cloud, and which are very energy-intensive; finally, the telecommunications networks. The percentage of each of these three sources of emissions contributes to producing the total is much debated. Some believe that users' appliances are much more polluting, due to their production process and the fact that they are replaced often; others instead that the three sources are more or less all on the same level.
Thus it is already clear that in order to try to calculate the emissions of the internet and the digital sector, it is necessary to make enormous approximations, which must take into account an exceptional variability in practices and consumption.
Much more accurate estimates have also been made, for example of how much CO2 is emitted by doing an online search, sending an email or watching an hour of Netflix. Cingolani himself said on Monday that sending an email with a megabyte attachment would be equivalent to keeping a 60W light bulb on for half an hour. However, these specific estimates are even more difficult to justify, and are subject to even greater variability that makes them substantially useless, as two researchers from the Royal Society have explained.
One thing they all agree on, however, is that internet emissions are set to increase. This is because the amount of people online is constantly increasing (today there are about 4.6 billion: there is margin), and there is a need for more devices and infrastructures. The amount of data used by each user also continues to grow, and the more data used increases, the more electricity needs for data centers and networks increase (even if, due to the way the internet is structured, this increase is not proportional).
The increase in emissions production has so far been unstoppable, although many internet companies have been trying to reduce their environmental impact for some years, especially in the West, where data centers have been made progressively more efficient, and where many companies have committed to powering them with electricity produced from renewable sources. In the rest of the world, however, things do not go the same way, and a substantial part of the electricity used to run the internet is still generated by burning fossil fuels.
However, it must also be considered that the internet often replaces other behaviors that could potentially produce more emissions. As some studies have shown, for example, it is true that, especially during the pandemic, the hours of streaming for videoconferences have increased considerably, and this has led to an increase in emissions; at the same time, on average, having the same conference in person would have generated more energy consumption.
Another point on which the researches are quite in agreement is that a very large part of the emissions comes from the production of hardware. Although not all opinions coincide, it is generally believed that the emissions generated during the production of appliances (servers, smartphones and so on) are greater than those generated by their operation (which in practice translates into electricity consumption) .
Producing microchips, contained in practically every digital device, requires, for example, enormous quantities of energy, water and other materials. And over the course of a computer's life cycle, two thirds of all energy consumed comes from production, and only one third from use.
Structural actions are therefore necessary to reduce the environmental impact of the digital sector and the internet: at this time, as the Patterns study writes, the sector is unable to meet the emission reduction obligations necessary to maintain the increase in emissions. temperatures within 1.5 ° C. However, these actions must come from internet companies or governments, most of which have no specific legislation regarding emissions in the digital sector.
As for what individual users can do, many experts say the point isn't so much posting fewer online and sending fewer emails. The most impactful thing you can do is change your smartphone, laptop or TV less often.
