Henry-WattTime
commented
1 year ago Hello @ciril-emaps I have some serious reservations about this proposal. Fundamentally, we are trying to create a specification that encourages action to reduce global emissions (what might be considered 'consequential', but used to inform action). This proposed PR does not maintain that intent.
I agree with Pieter Gagnon's letter that a short run marginal does not capture the full effect of load and renewable procurement activities. Pieter's letter calls for consideration of long run effects in decision making (a long run marginal signal), he does not advocate for any particular signal. We think that there could be many signals that might approximate that. I've tried to address this in PR #352 by making it clear that the marginal signal should include both short and long run effects.
I think there is consensus that a marginal emissions signal is the correct signal for making decisions to reduce emissions (Olivier's blog post agrees with this and correctly identifies that long run marginal is hard to identify and that average emissions in this context is just a proxy for long run marginal). Am I misunderstanding your point here? If not, a discussion of good signals for long run marginal might make sense in the guidance document.
Also, more granular data is not necessarily required for compliance with this standard. An organization can still make very informed siting decisions based on less granular data. For example, if an organization is making a decision on where to site computation load between Poland and France, annual data may be sufficient to make that decision, especially if load flexibility is not an option. Granular/hourly data alone should not be required if time-based load shifting decisions are being made. Global annual marginal emissions data (both operating and build margin) can be found from the UNFCCC, making this more accessible in a broader range of regions.
I don't particularly understand the bearing that other standards like the EU hydrogen standard has on this standard? I think this standard is attempting to measure something different and drive different action, especially because it includes market measures like renewable energy procurement, which the SCI explicitly excludes?
Does this make sense? Looking forward to discussing this in more depth.
ciril-emaps
atg-abhishek
navveenb
We are hereby submitting a formal proposal to amend the SCI standard.
From the evidence gathered, it seems like there is currently little consensus on which metric to use as a decision-making signal, as can be seen in these examples below (some of them are taken from this discussion in this forum):
Even though more evidence could be collected, the examples above highlight that mandating the use of a specific metric in the SCI seems to be an unnecessary restriction that might hinder the adoption of the SCI:
In the interest of a credible, robust and globally applicable SCI, we suggest making the SCI less restrictive, by not mandating marginal as the only decision-making signal. Instead, we suggest allowing multiple, equally relevant metrics in the SCI. Such a change facilitates the adoption of the SCI globally and provides companies with a more broadly applicable tool to decarbonise software.
We recognize the motivation of having a single signal in the standard, and would therefore like to express our intention in helping the group follow and stay aligned with the upcoming scientific consensus, with the ultimate goal to make the SCI more impact-driven and globally applicable.