Due for re-check: the scheduled review date of 2026-11-03 has passed. Every value below still carries the source it was taken from and the passage that states it — check it against the link.

1,519 records

When is this US generating unit scheduled to retire?

Across 1,519 records covering 256 plants, planned retirement year runs from 2026 (171 plants) to 2072 (3 plants), and 2027 is the most common — 222 of them. Every value is transcribed from the source that states it and carries the sentence it was taken from: 3 sources across 1 site, last checked 2026-09-24, due for re-check by 2026-11-03. Every record has a page of its own, quoting the line that states it.

About this data

Every utility-scale US electric generating unit its operator has told EIA it plans to retire, as printed in a single named monthly edition of EIA's Electric Power Monthly: the planned retirement month and year, the plant and unit it belongs to, the operating company, the state, the net summer capacity in megawatts, the generating technology, the fuel and the prime mover. Each record is one generating unit as one edition reported it, and every page names the edition it was read from. Answers 'when is the J.H. Campbell plant retiring', 'which generating units in Illinois are scheduled to retire in 2027' and 'how much coal capacity is scheduled to come off the grid next year' — for the edition named on the page. Three editions are held — EIA's data for April 2026, its data for May 2026 and its data for June 2026 — so where a unit appears in more than one, the pages side by side show whether its planned date moved; 469 units appear in all three and 33 of them carry a different planned date in June than in April, among them all three J.H. Campbell units in Michigan, which moved from May 2026 to August 2026. No single page states the movement, because a record is one unit as one edition reported it. Earlier editions are recoverable: EIA overwrites the previous month's figure on its rolling table, but it archives every monthly issue permanently and each archived issue carries the same table, so an edition can be read back from the issue that published it. These plans do move — EIA's own coverage notes that 'these plans may be more subject to change than usual, with policy shifts recently delaying retirements at several coal plants' — so the edition stamp is part of the fact, not a footnote.

The data

Showing the first 400 of 1,519 records. Every record has its own page, with its source and the quoted line that states it; these list the rest.

• Records 1–400ACCC Mays Landing · 60802 · PV1 · April 2026 … Fort Churchill · 2330 · 1 · June 2026
• Records 401–800Fort Churchill · 2330 · 2 · June 2026 … La Cygne · 1241 · 2 · April 2026
• Records 801–1,200La Cygne · 1241 · 1 · June 2026 … Rocky Ford · 6516 · IC1A · May 2026
• Records 1,201–1,519Rocky Ford · 6516 · IC2 · May 2026 … Zeeland · 1867 · 11 · May 2026

Where this came from

Every record above links the page it was taken from and quotes the sentence that states it. These are the 3 sources this dataset was assembled from.

• eia.govhttps://www.eia.gov/electricity/monthly/archive/june2026.pdf
• eia.govhttps://www.eia.gov/electricity/monthly/archive/august2026.zip
• eia.govhttps://www.eia.gov/electricity/monthly/archive/july2026.pdf

Machine-readable

• data.jsonThe whole dataset — every record with its source URL and source quote.
• Open Knowledge Format bundleOne JSON object per line — every record's frontmatter and quoted span exactly as it is held here, in one fetch.
• data.csvThe same records as one flat table, for a spreadsheet or a dataframe. The last four columns are the source URL, the quoted sentence it was read from, the date we last checked it, and which columns are our reading rather than the page's words.
• How this is made and checkedWhat "verified against source" does and does not mean.

From your own code

Same records, same quotes, without scraping the page: refsource is on PyPI and npm. Each value comes back carrying the URL it was read from and the sentence on that page that states it — .source and .quote sit on the value itself rather than in a side channel, so the checking step is available instead of skipped.

pip install refsource
refsource lookup power-plant-retirement-schedule plant_name="ACCC Mays Landing"

npx -y refsource lookup power-plant-retirement-schedule plant_name="ACCC Mays Landing"

Set your AI assistant up to use this

Two files and no account. Put this in .mcp.json at the root of your project — Claude Code, Cursor, Windsurf, VS Code and Codex all read that file — and your assistant can look this dataset up instead of recalling it. The server is remote, keyless and read-only.

{
  "mcpServers": {
    "referencesource": {
      "type": "http",
      "url": "https://referencesource.org/mcp"
    }
  }
}

Add to Cursor · or, on the command line: claude mcp add --transport http referencesource https://referencesource.org/mcp --scope project

Then one line in the project's CLAUDE.md or AGENTS.md, so the assistant knows when to reach for it:

When a question needs "Planned US generating-unit retirements, as EIA published them in one named month's Electric Power Monthly", call the referencesource MCP server at https://referencesource.org/mcp (tool `search_records`, dataset_slug `power-plant-retirement-schedule`) instead of answering from memory — every record it returns carries its source URL and a verbatim quote from that page.

What each tool does, and the servers built over single registers: Connect your AI assistant.

1,519 records. last verified against source 2026-09-24. due for re-check by 2026-11-03.

Licence. Facts extracted from the US Energy Information Administration's Electric Power Monthly, a US federal government publication. Plant names, identifiers, capacities and planned dates are facts and not copyrightable. Every record links back to the EPM table it was read from and quotes the row verbatim.