Discussion: Updates to the "Remaining emissions" success criteria in section 4.1

[ryansholin]

In section 4.1, "Choose a Sustainable Hosting Provider," the current "Remaining emissions" success criteria includes mentions of "carbon credits" and "offsets" in the text:

Remaining emissions are compensated, keeping in mind that the priority should be to avoid then reduce them and only compensate for them if they cannot be avoided. Carbon credits may not be sustainable, therefore the effectiveness of an offset solution must be verified, shown to be both environmentally viable and sustainable, and part of a longer-term Strategy to eliminate emissions entirely from a chain, benefitting the wider ecosystem.

For context, the immediately preceding success criteria is "Renewable electricity", which may need its own edit and discussion here. I believe the "Remaining emissions" criteria is intended to address the emissions that are not covered by location-based renewable electricity, though there are currently mentions of market-based electricity purchases such as RECs, as well.

As the current "Remaining emissions" text implies, the voluntary carbon market is a controversial space, and while terms like "high-quality verifiable offsets" are helpful, they're not backed by a particular standard, and I don't think we should attempt to create one here. We could be more specific in discussing location-based and market-based emissions, and we could cite the existing and relevant GHG Protocol guidance.

Approaches to updating these criteria could include:

1. Merge the "Remaining emissions" criteria with the preceding "Renewable electricity" criteria, removing mentions of "credits" or "offsets" while adding specificity about measuring location-based and market-based emissions as per the GHG Protocol. "Market-based" is a broader way to talk about the dynamic credit/offset/REC/EAC/PPA/offtake/drawdown/capture space without being too prescriptive about the "how."
2. Keep both criteria; rewrite the first to address location-based emissions (i.e. maximize the use of renewable energy from the local grid) and rewrite the second to address market-based emissions (i.e. where 100% renewables aren't possible, use market-based instruments to account for emissions).

[ryansholin]

Separately (or perhaps nested inside this issue?) we should decide whether to stick with "renewable" here, or "carbon-free" which would include nuclear.

[mrchrisadams]

There's a few things to refer to here:

"Renewable" is used in legislation in a few places, not to really refer to the grid being free of nuclear power, but to emphasise various market based interventions as @ryansholin says, to invest in renewables.

As I understand it, one reason that nuclear had not been included previously is that is was assumed that commercial actors like corporates don't buy nuclear power themselves (the state almost always some signficant chunk of the costs, either directly or indirectly), so there was little point in trying to incentivise it.

It's also worth bearing in mind that there is now a early stages of a fossil free standard for digital services being developed by CENELEC. This also seems to exclude nuclear being permitted in the mix, which I think may restrict the grids it could apply to. Here's the link:

https://www.cencenelec.eu/media/CEN-CENELEC/CWAs/RI/2025/cwa18203_2025.pdf

Energy Tag also have some updates about how cleaner energy forms look like they'll be recognised, and this is likely to be worked on throughout this year.

https://energytag.org/ghg-protocol-scope-2-update-7-key-messages/

Finally if you're looking for something to refer to do, there's some info about the likely minimum standards being adopted in Europe below, and our org has written a bit about the emerging practice for using market based approaches:
https://www.thegreenwebfoundation.org/news/a-may-update-on-the-first-e-e-d-day-in-europe/

[mrchrisadams]

This is in the pubic domain now as well. it gives an idea of how discussions are going about introducing a new regulatiuon requiring these minimum standards, and what might be reasonable to expect

Nikolaos Kontinakis An EU energy label for data centres.pdf

[AlexDawsonUK]

Personally I think renewables is clearer language.

Is any form of energy "carbon-free"? Sure the generation itself doesn't produce carbon but the equipment used to produce it (extraction, production, etc) will produce some kind of traceable emissions (as there will be human activity involved).

[ryansholin]

"Carbon-free energy" is a relatively standard term (though I don't think it's defined in a capital-S Standard) for energy that is generated without directly releasing carbon emissions.

tl;dr: If we want to use the broadest possible and widely accepted definition of "green" energy, I would recommend the term "carbon-free," mostly to include nuclear, rather than the term "renewable" which specifically excludes nuclear.

Renewable energy comes from natural sources that replenish themselves over time. This includes solar, wind, hydro, geothermal, and biomass. (I include biomass here for comprehensiveness but I am not well-read on the debate about the pros and cons of biomass -- usually burning wood -- as an energy source.)

Carbon-free energy refers specifically to energy that produces no carbon dioxide (CO₂) emissions at the point of generation. This includes solar, wind, hydro, nuclear, and sometimes geothermal. Biomass is typically not considered carbon-free, even though it's renewable, because burning it releases CO₂.

To boil that down a bit more (no steam-generation puns intended):

• All carbon-free energy sources produce no direct emissions.
• All renewable energy sources are naturally replenished.
• An energy source can be renewable but not carbon-free (e.g., biomass).
• An energy source can be carbon-free but not renewable (e.g., nuclear).

So, carbon-free is a more inclusive term than renewable. Terms like "green" or "clean" mean nothing particular in the context of electricity procurement, unfortunately!

Here's a specific decision-making example linked to the guidelines:
If we recommend using a data center that runs on "renewable" energy in Europe, that would exclude most of France, even though France has an extremely low carbon intensity thanks to a large amount of nuclear power.

[mrchrisadams]

I agree with @ryansholin on this one.

I think something like carbon-free is likely the option that would make more sense if you want wider acceptance, particularly in places where nuclear is on the grid, for the reasons that Ryan mentions. It would exclude the France, parts of the USA, and the UK straight away.

It's also consistent with the Climate Group's recently published technical guidelines on their new 24 / 7 carbon-free criteria. See page 8-10 in the doc linked below:

https://www2.theclimategroup.org/247-technical-criteria-May-2025

Why link to those technical standards? Generally because the work has already been done to get a degree of buy-in from people who aren't in the meetings where the WSG's are being discussed - they've already got corporates on board with what I understand to be a strict, but deliverable definition, of what we might call "clean" or "green" energy.

If you don't like either, there's always Fossil Free…

I'm obviously ridiculously biased, because it's literally our organisation's mission, but amazingly, I think the term it might actually be compatible with the 24 / 7 carbon-free definition above, because that definition from the Climate Group excludes the use of fossil-gas with carbon capture and storage anyway.

I also think it's consistent with the Fossil Free product definition I linked above, and where I had misread it. The text in that document doesn't exclude nuclear, it just has it as a should not, rather than shall not. Here's the text here:

5.1 Energy sources
5.1.1 Oil, fossil gas, coal and peat shall not be used. All products and their components, including both
permanent and temporary components, shall be produced without burning oil, fossil gas, coal, or peat
throughout their supply chain.
5.1.2 Renewable energy sources should be utilised throughout the supply chain.
5.1.3 Offsets shall not be accepted as a substitute for using actual renewable energy.
5.1.4 Renewable energy sources include:
• Solar power (photovoltaic and solar thermal),
• Wind power (onshore and offshore),
• Geothermal energy,
• Ocean energy (tidal, wave, and ocean thermal energy conversion),
• Biomass should not be used in installations exceeding 5 MW capacity.
5.1.5 Hydropower should follow the guidelines of the World Commission on Dams.
5.1.6 Nuclear power should not be used.
5.1.7 Renewable energy used in production shall be either generated on-site or traceable through
Energy Attribute Certificates (EACs) such as Guarantees of Origin (GOs) that are procured from the
same generation source as the underlying energy. Such procurement should provide a traceable
financial flow for both the power and its attributes, such as a Power Purchase Agreement (PPA). PPAs
allow companies to directly purchase renewable energy from a generator and may incentivize long-
term investments in renewable energy infrastructure. Certificates should follow the “three pillars”
principles adopted in both the European Union and the United States for renewable and “clean”
hydrogen projects.

That also includes the some of the same "Three pillars" stuff that both Energy tag and Electricity maps have spoken about, and is written into European and (at the time of this comment, at least) US law for "green hydrogen" projects.

I think this means that if we don't want to use carbon free, and we don't want to use renewable, we could always defer to the definition in the Energy sources section of that CENELEC Fossil Free Products - Guidelines and Requirements for Fossil Fuel Free Supply Chains document that I've pasted above.

Like the Climate Group definition, that's already gone through some fairly lively discussions in workshops, and seems to allow for nuclear, which on a number of grids, it's pretty much impossible to say you're not partially running on it.

[ryansholin]

Thanks to @mrchrisadams and @AlexDawsonUK for the discussion here.

Coming back to the tactical and practical bits, I am leaning toward this approach, with revisions:

Keep both criteria; rewrite the first to address location-based emissions (i.e. maximize the use of renewable energy from the local grid) and rewrite the second to address market-based emissions (i.e. where 100% renewables aren't possible, use market-based instruments to account for emissions).

I want to make sure we don't (re)write a whole recommended approach to emissions accounting in the WSGs, but rather suggest alignment with the GHG Protocol (the going standard for global compliance regulation) and provide more context, examples, and resources in the Additional Information, Resources, and Machine Testable sections.

The current "Renewable Electricity" under 4.1 could be re-titled to "Minimize Carbon Intensity" and focus on location-based emissions of web hosting providers, something like:

Select a hosting provider that uses electricity with the lowest possible carbon intensity. Location-based emissions factors should be used to calculate the carbon intensity of available electricity from the regional grid where the data centers in use are located. Onsite electricity generation, including backup generators, should also be taken into account.

Note: Data about behind-the-meter "onsite electricity generation" is not easily accessible, but it should be disclosed in sustainability reports from relevant vendors.

Then, the current "Remaining Emissions" under 4.1 could be revised to focus on existing (and evolving) standards for market-based emissions accounting, something like:

Select a hosting provider that follows the GHG Protocol Scope 2 guidance on market-based carbon emissions accounting, seeking to match its location-based emissions from electricity consumption with purchases of carbon-free electricity.

Note: I am leaving out many possible additional modifiers about CFE there in deference to the GHGP Scope 2 guidance. If the W3C wanted to go further, we could call it "...purchases of 24/7 hourly-matched carbon-free electricity."