Update 2025 Interconnection Queue

[Yunzhi-Chen]

Summary

Update interconnection queues to LBNL's 2025 data vintage (LBNL_Ix_Queue_Data_File_thru2025.xlsx), producing interconnection_queues_2025.csv over a 2027–2031 window. Also repoints the county mapping, which was broken against current ReEDS main, and carries two processing bug fixes (isolated and verified in #16).

Prior-vintage outputs are untouched: interconnection_queues_2024.csv, _2023.csv, queue_versions_2024.html, queue_versions_2023.html.

Technical details

Implementation notes

• Params: version=2026, t_1=2028, t_2=2031 — columns 2027–2031, continuing the +1/vintage pattern (2024 file was 2026–2030).
• LBNL renamed columns this vintage (IA_status_clean → IA_phase_clean, type1/mw1 → type_1/mw_1). Added a third branch to the existing version check and rename them back, so the rest of the script is untouched.
• mw_* now read as object because of the header offset — added a pd.to_numeric cast.
• ReEDS/inputs/county2zone.csv no longer exists on ReEDS main; the mapping moved to ReEDS/inputs/zones/county_state.csv (same FIPS,county_name,state, 3109 rows). The script did not run without this.
• reeds_path now reads a REEDS_PATH env var, defaulting to the previous hardcoded path.
• The version-1 figure is no longer regenerated if it already exists, so a run can't overwrite a prior vintage's figure. Fixed chart 1's x-axis sort to use year_range_version_1.

Bug fixes carried on this branch

Both are detailed and verified against the 2024 input in #16:

1. Counties matched on name instead of FIPS. LBNL doesn't always use the ReEDS spelling (Acadia Parish vs acadia), so unmatched rows were dropped at the merge — Louisiana lost essentially its entire queue. Now FIPS first with county-name fallback, since LBNL's FIPS are sometimes stale (Oglala Lakota SD) or concatenated for multi-county projects. Unmatched active capacity drops 4.4% → 1.9%, the remainder out of ReEDS scope.
2. t_2 double-counted the IA Executed queue — built as max(cap_t1, cap_t2) + cap_t1 instead of cap_t1 + cap_t2. t_1 was never affected.

Recovering biomass and pumped-hydro

LBNL folded Pumped Storage into Other Storage and Biofuel/Biomass into Other this vintage. Both are filtered out by the script, so those two tech groups would have disappeared from the output — and ReEDS treats a missing (tg,r) as a 0 limit, which would have blocked all new biomass and PSH builds through 2031.

LBNL sent a supplement with the detailed types behind those aggregated categories for active requests (inputs/queues_other_forNLR_2025.xlsx); add_detailed_types() merges it back in before any other processing:

• Matched on q_id + entity — a q_id is only unique within an entity.
• Hybrid requests list their types in a different order than the public file, so each detailed type is matched to the aggregated category it was folded into rather than by position.
• Capacities always come from the public file; the supplement only relabels.

Recovers 6,841 MW pumped-hydro and 6,320 MW biomass, in line with the 2024 vintage (6,786 / 2,602). All other tech groups are byte-identical to the pre-merge output. LBNL confirmed the supplement can be posted publicly.

Not recovered: compressed air (1,737 MW), waste heat (72 MW) and wave (20 MW) have no ReEDS tech group and stay aggregated. Five supplement records have no match in the public file (~900 MW PSH); the script prints them at runtime.

Figures

2025 interconnection queue (outputs/figures/queue_versions_2025.html):

What ReEDS uses today vs. what it will use (outputs/figures/compare_queue_versions.html):

Read the 2026 and 2031 bars as constraint presence, not a limit of zero: eq_interconnection_queues is guarded by $(sum{(tgg,rr), cap_limit(tgg,rr,t)}), so a year absent from the csv is unconstrained rather than blocked.

Additional changes

• README updated: new filename, REEDS_PATH note, supplement merge documented.
• .gitignore: atb/ (unrelated, carried on this branch).

Relevant sources or documentation

• LBNL Queued Up: https://emp.lbl.gov/queues

Additional details

What this changes for ReEDS. In every overlapping year the limit tightens by 35–55% (2027 −39%, 2028 −55%, 2029 −41%, 2030 −35%) — a combination of the window shifting a year, the data update, and the two bug fixes. This is the main consequence of the PR.

Needs a paired ReEDS change: interconnection_start in inputs/scalars.csv should go 2025 → 2026 to match the new window. Without it, 2025 and 2026 builds count against the 2027 limit — which matters more than usual given how much that limit tightens.

Output verification: 23,201 rows (1,925 csp + 12 tech groups × 1,773 counties), 2,132 counties, 13 tech groups; no NaNs, no negatives, no duplicate (r,tg), values monotonic across years; all r are valid ReEDS counties; interconnection_queues_2025.csv matches interconnection_queues.csv.

Documentation: README updated. No ReEDS documentation changes needed.

LLM usage: Claude Code — adapting the script to the new column schema, locating the moved county mapping file, the two bug fixes, and the supplement merge for biomass/pumped-hydro.

[atpham88]

Can we get rid of version 2024 and its associated processing script? I think we can just keep two most recent versions.

[atpham88]

Can remove this if we remove 2024 version

[atpham88]

the queue data has FIPS now and we don't care about its name or state so can remove this and also any merging with this data, right? I believe I only originally did this because the first version of queue data I worked with did not have FIPS column.