Assumptions and Decisions

Here are the assumptions we made on key aspects on this project. Some we made at the beginning, some emerged as the project progressed. We used these to guide various decisions as the project progressed.

Using publicly available data

One of core values at GWF is to work in the open, in part because we believe this is an important way to foster faster action on the climate crisis. Therefore we decided early on that we’d only rely on data that had already been openly published. This meant we would not rely on tools, methodologies or datasets behind paywalls.

Estimating for impact or reporting

Whilst the TCS broadly aligns with the scoping of the GHG Protocol, its purpose is not to produce accounting-level compliant reporting. Instead the focus of this project is to attempt to estimate emissions with a view to looking for areas of impact, and consequently what can be done to improve.

Estimates vs measurements: what models are we using?

It’s important to be aware of the kind of model you’re working with, as it will impact estimates results. Top down models, tend to look at entire systems and as such are more likely to overestimate emissions. Bottom up models look to estimate smaller components of larger systems, and are likely to underestimate emissions.

That is to say, all models have flaws and it is important to understand their flaws and limitations. We’ve done our best to keep this in mind throughout this project, and clearly state the known flaws and limitiations for any approaches we’ve used.

Fershad, from GWF, has spoken quite a bit about this difference in the past - his slides on The Nuance of Quantifying Digital Carbon Emissions.

High-level estimates

We decided that for each category in the TCS we would attempt to produce at least a high-level estimate.

We assumed a high-level estimate would be characterised as:

  • Imperfect but relatively quick to do
  • Containing broad assumptions and averages
  • Favouring publicly available data rather than specific measurements unique to GWF
  • Intended to give a feel for the magnitude of emissions on particular category vs another, allowing us to identify areas of most potential concern
  • Better than nothing

Detailed estimates

We decided that after producing our initial set of high-level estimates, we would deep dive into particular areas of concern or interest as remaining capacity allowed.

We assumed that detailed-level estimates would be characterised as:

  • Still imperfect, aligned to GHG accounting standards when it makes sense to do so.
  • Time-consuming, complex and required detailed knowledge to produce.
  • Using specific measurements unique to GWF’s set-up as much as possible.
  • Helpful for understanding what can be changed to reduce impact.

Our approaches to producing these high-level, and when relevant detailed estimates, are documented in this project write-up.

Units

For this project of mapping the GWF technology estate using TCS, we calculate the final figures to kgCO2e/year.

We typically use kW and kWhs when quoting energy use and consumption.

Carbon intensity is typically quoted in gCO2e/kWh.

We’ve chosen to round calcuation outputs when appropriate, which may lead to some rounding errors and very minor variation in the results.

Regional Carbon Intensity

The GWF team are a remote team, positioned around the world. We used the regional electricity grid’s carbon intensity to calculate the operational carbon emissions.

The following regional carbon intensities are used:

  • Germany: 363.982gCO2e/kWh
  • Netherlands: 386.189gCO2e/kWh
  • Taiwan: 565.629gCO2e/kWh
  • UK: 268.255gCO2e/kWh
  • Global: 490gCO2e/kWh
  • Europe: 328gCO2e/kWh

These values were obtained from the GWF Grid Intensity CLI tool and the Tech Carbon Standard (based on Ember Data Explorer).

Digital sustainability vs carbon

Digital sustainability and decarbonising digital are not the same thing. Sustainability, or understanding ones wider impact, involves a much broader set of considerations covering other planetary boundaries (water, rare raw materials, land use) and social implications.

This project narrowed in on carbon emissions as a starting lens to begin understanding these considerations.

Hannah, from GWF, has spoken a lot about this issue in the past. Her fellowship project, doingthedoughnut.tech, goes into this more detail.

Next - Estimating the Upstream Emissions