Do we aggregate the primary and secondary Aluminum industries and commodities?

[WesIngwersen]

BEA benchmark IOT has two Aluminum industries and commodities

Primary aluminum, 331313
Secondary aluminum, 33131B

USEEIO has kept these as is in v1 and 2.

CEDA has aggregated them into a single industry and commodity, 331313.

Part of project https://github.com/orgs/cornerstone-data/projects/6/

RESOLVED as of 12/11/2025 - see note below

[sarahcashmanz]

It would be nice to keep them separate given the GHG impact between the two varies so much. It would also help for any future hybridization.

[sangwonsuhvm]

In CEDA, we aggregated them as we weren't confident about the emissions data at that level of detail. I think that it makes sense to separate them if we are confident about the emissions data.

[MoLi7]

Agree with Sangwon - it's mainly a decision made due to data limitation.

I support separating them if we're confident about emission attribution to primary and secondary aluminum.

[WesIngwersen]

Awaiting #43 to provide insight into data quality.

[MoLi7]

In the 2023 EPA's GHG Inventory, emissions from aluminum production are not reported separately for primary and secondary aluminum products.

See CO2 emissions as an example:

@bl-young @catherinebirney could you share if FLOWSA uses additional data to attribute emissions to primary and secondary aluminum separately?

[bl-young]

We do not attribute to them separately but rather assign solely to 331313. I agree w/ your assessment.

[bl-young]

Need to clarify the original comment. I think we mean 331314, which is secondary aluminum not 33131B

[bl-young]

We need to confirm if we looking at the question correctly. @MoLi7 will confirm if CEDA aggregates all 3: 331313, 331314, and 33131B (which is 3 industries, and 2 commodities in the BEA tables)

[bl-young]

In our stewi datasets, specifically NEI, we can identify the assigned NAICS by facility according to the FRS registry (or that reported by the facility). We see more facilities at 331314 than 331313, but the largest emitters tend to be in 331313, which tracks.

[bl-young]

Other facility level datasets, like GHGRP, would provide some potential insight into this split as well for the GHG data.

[bl-young]

Dataset of select air emissions by facility is added in #61

Summary for 2020 dataset:

Flow	Unit	331313	331314	331315	331318	33131
Carbon Dioxide	1000 tonnes	2,292	1,041	584	590	245
Carbon Monoxide	tonnes	102,648	1,291	504	460	12,009
Nitrogen Oxides	tonnes	3,671	1,300	486	622	104
PM2.5 Primary (Filt + Cond)	tonnes	1,780	641	693	379	19
Sulfur Dioxide	tonnes	13,013	117	17	7	2,218
Volatile Organic Compounds	tonnes	917	1,025	4,035	872	0

[jvendries]

Resolution
Keep them separated given the availability of data for allocating emissions and the broader goal of having more detailed models.