Flexible operations, on-site generation, and AI-driven loads make annual-average Scope 2 factors unreliable. This whitepaper shows how hourly and marginal emissions metrics sharpen accuracy by up to ~20 percent and provide a more truthful view of site performance.
Key insights:
- Annual-average factors mask meaningful intra-day and seasonal variability.
- Hourly emissions factors materially improve Scope 2 accuracy for flexible loads.
- Marginal emissions clarify avoided-emissions claims from dispatch, DR, and DERs.
- Time-aligned clean-energy procurement (including hourly CFE%) ensures PPAs actually match site consumption.
The paper includes ERCOT and CAISO modelling plus clear recommendations for operators and sustainability teams adopting time-aware accounting.
Jan. 13, 2026, 11:27 a.m. - Jose M. Moya
This is a strong step toward aligning accounting with how flexible data centres actually operate. Time-averaged Scope 2 metrics clearly fail to reflect the real impact of load shifting and grid-aware operation. Moving to hourly and marginal factors is a necessary correction. That said, as several lifecycle studies have shown, Scope 3 (particularly IT and infrastructure embodied carbon) will dominate in many data centre contexts. The next challenge is connecting time-aware Scope 2 signals with Scope 3 trade-offs: utilisation, refresh cycles, capacity planning, and the decision not to build. Without that link, we risk optimising carbon reporting faster than we optimise the systems themselves.