grgmiller
commented
2 years ago The language describing residual mixes in the GHG Protocol Scope 2 Guidance is generally pretty vague, and describes it as:
- "untracked or unclaimed energy and emissions"
- "the emissions rate left after the three other contractual information items [EACs, PPAs, supplier specific rates] are removed from the system"
- "all unclaimed energy emissions, which is formulated by removing contractual claims data from energy production data (often the same as grid average data)"
- "represents the regional emissions data that consumers should use if they operate in a market with choice for consumers, differentiated products, and supplier specific data, but did not purchase certificates or a specified product, do not have a contract with a specified source, or do not have supplier-specific information"
However, in two cases, a more specific definition is provided, which seems to suggest that compliance RECs would remain in the residual mix calculation (although it is still not explicit):
- "uses energy production data and factors out voluntary purchases"
- "characterizing the GHG intensity of unclaimed or publicly shared electricity"
ewezerek
In market-based carbon accounting, the residual mix describes the carbon intensity of the system after other contractual claims (generally tracked using RECS) have been netted out.
I need to look into the official definition, but my understanding is that a residual mix only nets out voluntary REC retirements, and not retirements as part of a state RPS. In a state with a minimum renewable portfolio standard, all utilities are required to retire the same minimum amount of RECs, so no matter which utility you have, the residual mix will include all state-mandated renewables. However, utilities and other energy buyers may voluntarily procure additional RECs, which would not be counted as part of the residual mix. Thus, the challenge of calculating a residual mix is identifying which renewable generation is being claimed as part of a state RPS mandate, and which is being retired as part of voluntary procurement.
however, another way to potentially think about residual mix is to remove all supplier-specific claims, regardless of whether it was mandated or voluntary.
Data sources
Calculating a residual mix would require information on all voluntary REC retirements that occur.
Green-e publishes annual-average residual mix information that is aggregated at the eGRID subregion. It seems that this data only nets out voluntary RECS. https://www.green-e.org/residual-mix
PJM-GATS also publishes data about the residual fuel mix of their system on an annual basis, and seems to net out all REC claims (RPS and voluntary): https://gats.pjm-eis.com/GATS2/PublicReports/PJMResidualMix/Filter
Potential Approaches
Approximation approach
If we assume that all (REC-eligible) renewable energy generation in a region is claimed by someone else, we could approximate the residual mix using existing data as the average emissions intensity of all non-renewable sources. This is likely a reasonable approximation of residual mix if there is no RPS (or if residual mix excludes RPS renewables as well), although more in depth research would be needed to validate this as a robust approach to approximating residual mix.
However, by quickly comparing PJM-GATS data on residual mix to their system mix, it appears that the residual mix only contains non-renewable resources.
More robust approach
If we could get information on REC purchases (or even better, REC retirements), we could more accurately calculate residual mix. In order to perform this calculation, we would need a way to match each REC with a specific generator and specific month. At the very least, we would need aggregated REC data for each fuel type at a BA level (e.g. how many wind RECs were retired in ERCOT in May 2020) in order to reasonably estimate the residual mix based on REC data. We would also potentially need information on whether each REC is a compliance REC or voluntary REC.
Other questions