The recent launch of the online magazine “Branch” has opened up the topic of sustainable digital technology to a wider audience. This first issue also provides a better understanding of the position of specific tech actors regarding digital sustainability. However, what about other perspectives? Sustainable digital technology is a complex subject that will require international cooperation, but the French community publish in French and rarely translate its reports and papers, as do Germans. This article tries to provide an understanding of the French perspective on digital sustainability and its recent progress.
Disclaimer: I do not represent the French sustainable digital community, I am just trying to synthesize the action of many individuals and collectives in French.This contribution is a rework and an updated version of an article published in November 2020 on my personal website.
One of the biggest differences between France and practices described in the previous issue of Branch is the way we select and assess environmental impacts.
In France, we tend to follow a Life-Cycle Assessment (LCA) approach that includes four factors: Greenhouse Gas emissions (CO2eq), water consumption (Liters), abiotic resources consumption (Sbeq. Or Antimony equivalent) and primary energy consumption (MJ). LCA methods follow international norms such as International Standard of Organisations (ISO) 14040 and ISO 14044 life cycle assessment framework. Like everybody else we measure these impacts on three different poles: data centers, networks and end-user equipment, from manufacturing to operations. One of our leading LCA experts when it comes to digital services, Julie Orgelet, manages to integrate End of Life impacts (EoL) by adding a volume of generated e-waste to her LCAs. But today EoL is rarely included in the measurement of the environmental footprint of digital technologies. Thus, the reference tool that we generally use in the community is based on GreenIT’s LCAs and therefore calculates the impacts during manufacture and use: Ecoindex.
So what’s the difference with the approach and tools described in the previous issue?
The main focus seems to be on reducing carbon emissions through the vector of electricity. This implies that there is only one factor of environmental impact: carbon. When we use the vector of electricity to calculate carbon emissions we are only looking for operations and not including equipment manufacturing.
Therefore, this approach focuses more on the carbon intensity of the energy mix. This implies that the main lever for action is the integration of renewable energies into the energy mix and secondly, the reduction of the data transfer to reduce the electricity cost for transmission (GB/kWh). The most popular measurement tool appears to be Wholegrain Digital’s Website Carbon. It measures the electricity consumption linked to data transfer and the carbon emissions linked to the nature of the energy mix. These two approaches therefore give very different results. For example, my website emits 0.08gCO2 per visit according to Website Carbon and 1.51g eqCO2 according to Ecoindex.
The difference is linked to the inclusion of manufacturing impacts.
Different measurements implies different design methods
Do these different approaches influence digital design methods? It seems to me so and I will try to explain it with an example given in Branch.
The Organic Basics website promotes a low impact website approach. Indeed, they have incorporated certain good practices. They also tried an approach that I find interesting: the user experience of the site depends on the carbon intensity of the energy mix that powers their server.
The more carbon there is per kWh and the fewer assets are called up (the site can theoretically close), the less carbon there is per kWh the more elements they load (HD photos, animation, etc.). We can call it carbon-responsive design. One of the ideas I really like is the vector processing (SVG) of product thumbnails and loading images only when clicking on the product sheet. However, a fundamental problem with Organic Basics’s approach is that the less carbon the electricity has, the more the site uses. Two problems then emerge, firstly, this is an example of a rebound effect: the drop in carbon intensity leads theoretically to an increase in electricity consumption.
Second, under the pretext of reducing solely carbon impact, we increase the impact of other factors that we do not look at. For example, in France, we know that the production of 1 kWh involves the use of 4 liters of water (production of electricity via steam). The focus on carbon emissions during operations (via the vector of electricity) can become very misleading, even counterproductive over time.
What’s happening now in France
The French community lookat digital sustainability is relatively old, from the creation of the GreenIT collective in 2004, of the research cluster Ecoinfo in 2006 to the “newcomers” like the NGO HOP in 2015 (Stop Planned Obsolescence) or the Shift Project.
These actors, between many others, had a considerable impact on lawmaking in France regarding digital sustainability. Several bills have been pushed since 2019 to account for environmental impacts of the digital sector.
In February 2020, the Anti-Waste and Circular Economy Law was passed. It requires software editors to provide software updates for 2 years and to dissociate functional updates from evolutive updates. This law also established the repairability score, the declaration of data transmission GHG footprint for ISPs, purchasing of second-hand hardware for public administration and preference for “less-energy intensive ” software for public purchasing. A complete assessment of the environmental footprint of digital technology in France has been commissioned by the French Agency for Ecological Transition (ADEME) and the Electronic Communication Regulatory Authority (ARCEP),where results will be published in one year and will be the official assessment that will inform public policies.
Finally, the ”Reducing the environmental footprint of digital technology in France” bill, proposed by the Senate, will go through its first reading in Parliament next week. The main goal of this bill is to transform the concept of “digital sobriety”» into law. That means buying less-powerful machines, replacing them less frequently and not using energy-intensive approaches where possible. We now have to see what will be left of these proposals after Parliament’s amendments. At the moment the community is not very optimistic but has allied parliamentarians who will defend the bill.
Where we might be going
This legislative wave is a semi-victory as it is the fusion of business as usual thinking that the digitalisation allows to reduce the GHG emissions by default (which is not demonstrated), and of a new way of thinking that the environmental footprint of the digital sector must drastically reduce. The next step might be to ask the question much more frankly: what does a digital ecosystem (infrastructure and services) look like in a world stabilized at +2°C?
To answer this question, we will have to greatly improve our knowledge of the environmental impacts of digital technology. Many years, and international collaboration, will be necessary to stabilize our knowledge on the topic. It will also be necessary to have a much better understanding of the possible positive impacts of digital technology in specific contexts. It seems dangerous to presuppose that more digitalisation is necessarily better, especially when we are facing planetary limits.