CCUS.jl's update_ccus_gens! splits every carbon capture generator into a "saline" and an "eor" row sharing the same physical capacity (matched via match_capacity! requiring sum(pcap) <= pcap_max). For most gens is okay, but for ccs build_types "real" or "exog" with build_status "unbuilt" (which are planned gens with a build year inside the model horizon), add_build_constraints! independently fixes each row to its own full pcap_max in the build year. This leads to an infeasible model.
There are very instances of planned CCS builds but could be more common with time. The 2024 EIA860 form has a planned ngcccs in Louisiana which is causing this issue. However, this project was canceled in summer 2026.
CCUS.jl's update_ccus_gens! splits every carbon capture generator into a "saline" and an "eor" row sharing the same physical capacity (matched via match_capacity! requiring sum(pcap) <= pcap_max). For most gens is okay, but for ccs build_types "real" or "exog" with build_status "unbuilt" (which are planned gens with a build year inside the model horizon), add_build_constraints! independently fixes each row to its own full pcap_max in the build year. This leads to an infeasible model.
There are very instances of planned CCS builds but could be more common with time. The 2024 EIA860 form has a planned ngcccs in Louisiana which is causing this issue. However, this project was canceled in summer 2026.