A planned U.S. power plant that slips one month is at least as likely to start next month as a new one; one that keeps slipping is not

Description and use

Planned U.S. generators with a one‑month slip are as likely to start on time as new ones, while those with three or more slips have a lower on‑time probability.

Problem

Determining whether a generator’s earlier missed start dates predict its likelihood of starting on time in the following month.

Outcome

Generators with one prior slip are as likely to start on time as those with no prior slips; generators with three or more prior slips have a markedly lower on‑time share (0.306 vs 0.362). One‑month slips yield a higher on‑time share (0.390) than 2‑3‑month slips (0.260).

Workflow

• 1. Collect EIA‑860M monthly generator inventory data for 2022‑2025 and 2016‑2019.
• 2. Count prior missed start months for each generator before its current planned start month.
• 3. Group generators by number of earlier misses (0, 1, 2, 3+).
• 4. Calculate the share of generators that start on time (appear as operating in the next monthly file) for each group.
• 5. Compare shares across groups and time periods to assess impact of slip count.
• 6. Repeat analysis using slip length (1‑month slip vs 2‑3‑month slip) to evaluate effect of slip duration.
• 7. Publish predictions and code to GitHub for reproducibility.

Evidence

Two held. Three or more earlier misses is a bad sign, and that is new.

In 2022-2025 the share on time was 0.362 with no earlier next-month listing, 0.365 with one, 0.357 with two and 0.306 with three or more.

A generator planned for next month again right after a one-month slip was on time 0.390 of the time in 2022-2025, a little more often than a first listing.

The code, the raw output and a test that recomputes every number are here: https://github.com/munimtechnologies/eia860m-slip

Source