Four real comparisons across the utilities this project tracks, built from the same closed-event records every other page here reads — nothing modeled or estimated. Coverage varies by utility (some report every individual incident; others only ever give one county-wide rolling total), so a smaller or newer utility having a short bar below usually means less historical data, not a better track record. Where a number is thin enough to not mean much yet, it's labeled limited rather than hidden.
Resolved outage count per utility, statewide, from every real closed event on file.
| Utility | Resolved outages |
|---|---|
| Duke Energy | 40,296 |
| TECO | 16,218 |
| FPL | 5,563 |
| Clay Electric Cooperative | 3,120 |
| Lee County Electric Cooperative | 927 |
| City of Tallahassee | 656 |
| Florida Keys Electric Cooperative | 442 |
| Lake Worth Beach Utilities | 416 |
| Orlando Utilities Commission | 415 |
| Florida Public Utilities Corporation | 319 |
| JEA | 62 |
| Peace River Electric Cooperative | 3 |
| Talquin Electric Cooperative | 3 |
Median outage duration per utility, longest first. A single outage lasting longer than 96 hours is excluded from this math — for a county-wide rolling-total source, that's usually several separate real outages blurred into one implausibly long reading, not one actual repair job (see the caption below).
| Utility | n | Median hours | p90 hours | Excluded as implausible |
|---|---|---|---|---|
| Talquin Electric Cooperative (limited) | 2 | 3.7 | 6.1 | 1 |
| FPL | 5486 | 2.6 | 16.7 | 77 |
| Lee County Electric Cooperative | 927 | 1.8 | 7.4 | 0 |
| JEA | 62 | 1.4 | 6.6 | 0 |
| Duke Energy | 40262 | 1.3 | 4.0 | 34 |
| Clay Electric Cooperative | 3120 | 1.3 | 3.9 | 0 |
| Peace River Electric Cooperative (limited) | 1 | 1.2 | 1.2 | 2 |
| City of Tallahassee | 656 | 1.1 | 5.0 | 0 |
| Orlando Utilities Commission | 415 | 0.8 | 2.9 | 0 |
| Florida Keys Electric Cooperative | 442 | 0.8 | 4.2 | 0 |
| Lake Worth Beach Utilities | 416 | 0.5 | 3.3 | 0 |
| TECO | 16218 | 0.5 | 4.0 | 0 |
| Florida Public Utilities Corporation | 319 | 0.5 | 1.9 | 0 |
"Limited" means fewer than 3 usable events on file — a real number, just too thin yet to call a trend.
When a utility tells you when your power will be back, how often is that true? On-time % per utility, most honest first — only utilities that report a real per-incident estimate (not just a county-wide rollup) can be checked this way, which is why only a few appear here.
| Utility | n | On-time % | Median error (hours) |
|---|---|---|---|
| TECO | 16434 | 90.9 | -3.4 |
| Clay Electric Cooperative | 8941 | 81.6 | -0.9 |
| Lake Worth Beach Utilities | 858 | 64.2 | -1.3 |
| FPL | 27917 | 56.3 | -0.4 |
"On-time" means power actually came back at or before the utility's own first stated estimate. A positive median error means the real median outage ran later than promised; negative means it beat its own estimate.
Share of closed incidents where the utility's own cause text actually resolved to a real category, per utility — then, among those, what the real mix looks like.
| Utility | Real cause % | Equipment | Planned work | Vegetation | Weather | Other / vehicle / animal |
|---|---|---|---|---|---|---|
| FPL | 36.6% | 9247 | 0 | 412 | 348 | 234 |
| Lake Worth Beach Utilities | 17.4% | 88 | 44 | 5 | 14 | 6 |
| Duke Energy | 8.3% | 0 | 3440 | 0 | 0 | 0 |
| TECO | 8.2% | 944 | 0 | 373 | 0 | 30 |
Known real quirk: Duke's own cause text almost never matches the keyword categories the other three utilities' phrasing does fall into — that reads here as a low "real cause %," which is honest about the comparison but isn't the same thing as Duke rarely knowing why the power went out. It's a wording mismatch with this project's shared categorizer, not a finding about Duke's own record-keeping.