Useful calculations – energy vs carbon

We care about emissions because carbon dioxide and its other greenhouse gas (GHG) equivalents are responsible for climate change. You might therefore think that reporting emissions is all that is necessary. Simply tell everyone how much CO2e you were responsible for and when it hits (and stays at) zero, everything is great.

This is true to the extent that there is a global target. We want to hit global net zero to stabilize atmospheric carbon levels, then keep going to achieve negative emissions and reduce total carbon. This is the macro level.

However, to understand whether and how things are improving on a micro level, we need more details. A reduction in GHG emissions looks good in a report, but how those reductions are achieved varies significantly.

As an example let’s consider the internet, which is made up of many interconnected networks. Energy is used in the form of electricity to enable the transmission of data across the network. There are many different telecoms companies operating networks of varying sizes around the world. Some of these are large enough to publish annual environmental reports, either voluntarily or because their size triggers mandatory disclosure regulations.

Telefónica is one such company. Based in Spain, they operate large networks in Germany, UK, France, Brazil, Peru, US and various other regions. Their annual report says that in 2020 they consumed 6,863,728 MWh of energy, of which 4,918,373 MWh was renewable. Total Scope 1 (directly generated) + Scope 2 (purchased energy, market based) GHG emissions were 212,682 + 530,684 = 743,366 tCO2e.

Another company is Virgin Media. They operate primarily in the UK & Ireland and their annual report says that in 2020 their UK Scope 1 + Scope 2 (market based) emissions were 35,070 tCO2e. They don’t publish figures in energy units.

Ignoring the fact that as of 2021, Telefónica is now a part owner of Virgin Media (this reporting is from 2020), it’s currently impossible to compare how efficient their networks are and whether they are improving.

Telefónica operates across many more regions than Virgin, so it makes sense their emissions are larger. However, with sufficient data that shouldn’t make any difference to the ability to compare them. What we need is an intensity metric – something that will give us the energy consumed per unit of work done. In networking, this is usually energy per unit of data.

Telefónica actually provides this! They report the total amount of data traffic and total amount of energy, then provide an annual figure for MWh per PB of data. In 2015 this was 408.7 MWh/PB and in 2020 it was 78.2 MWh/PB. They don’t provide a breakdown by region, but this is useful to see that they are improving the energy efficiency of their network. This translates directly into emissions reductions on a per traffic unit basis, but also on a total energy consumption basis because such a large efficiency improvement offsets a huge growth in data traffic.

Virgin Media publishes something that looks like an intensity metric, but it is CO2e per TB of data. The calculation is from location-based emissions so is at least linked to the region where the energy is consumed. However, even though it is showing improvements – from 0.06 tCO2e/TB in 2014 to 0.005 tCO2e/TB in 2020, we don’t know whether that is just inheriting the improving carbon intensity of the grid as a whole, or if Virgin Media are investing in energy efficiency in their network. Probably both.

Virgin Media also publish the market-based energy intensity, but this is even less useful because it includes market measures like RECs and offsets that could reduce that intensity to zero without improving anything on a local level.

Both companies do publish tCO2e/revenue, so we can see that on a Scope 1 + Scope 2 basis, Telefónica’s intensity is 17.26 tCO2e/€million and Virgin Media is 36.94 tCO2e/€million, but that is not particularly useful for examining their network practices.

How would you pick a provider on the basis of their sustainability? Without an energy per work done-based metric, you can’t be sure. Ideally you want both energy and CO2e (which is important, but not the only aspect of sustainability). How those CO2e reductions are achieved is also relevant – is a company actively investing or is it just relying on system-level improvements from others? This is another argument for consistent and transparent reporting standards.