How reliable is your power? Minutes without electricity, by state and utility

Minutes without power by state, 2024

Customer-weighted SAIDI. Darker is more minutes. Grey states have no figure: Alabama, Hawaii, New York, Oklahoma, Vermont, Virginia. Utilities there mostly report under another method; each state page says how many.

All states in one table

Download state data as CSV

Click a header to sort. SAIDI is minutes without power per customer in the year; SAIFI is interruptions per customer. Not-ranked states are explained on their pages.

Ranking: 45 states, fewest minutes first

A normal year against the year as lived

Each dot is a state. States far above the line lost most of their minutes to major events. The biggest drops in rank from adding major events: Florida from 6th to 41st; Nebraska from 9th to 37th; New Hampshire from 14th to 38th. The smallest gap between the two figures is in District of Columbia (6 minutes); the largest in South Carolina (2939).

Why six states have no number

EIA's form asks each utility to report SAIDI and SAIFI under the IEEE 1366 standard or, if it uses another method, in a separate block. In 2024, 747 utility-state records are in the first block and 224 in the second; none are in both. Another method can count outages differently (for example, by what size of interruption is included), so we do not add the two together. A state is ranked when the IEEE group covers at least 60% of its customers. Not ranked: Alabama (14%); Hawaii (7%); New York (24%); Oklahoma (49%); Vermont (12%); Virginia (30%). EIA's State Totals sheet is used only as a check.

SAIDI, SAIFI and CAIDI in plain words

These are the three numbers every utility files. An interruption counts when it lasts more than five minutes. All three are averages over every customer, including the many who lost power for no time at all, so one household's year can be far from the average in either direction.

What leaving out major events hides

Utilities may remove "major event days" from their reliability figures. Under IEEE 1366 a utility decides which days qualify with a statistical rule applied to its own daily history, so what counts as a major event is relative to what is normal for that utility, and EIA publishes the result, not the list of days removed. The without-events figure covers ordinary days; the with-events figure also includes the large events.

In 2024 the gap was large: 532 of 662 minutes (80%) sat on major event days across these 45 states. The share is not even: from 7% in Utah to 96% in South Carolina. The first number depends on how many days a utility removes; compare utilities on both, and compare only within one reporting method. The outage reports behind many big events are on the Department of Energy outage pages.

Is my utility better or worse than the state median?

Pick your state and the company on your bill. You get its 2024 figures, where it sits among the utilities in your state and in the country, and how it compares with the median utility in your state.

What this means for you

• Compare your utility with the median utility in your state, first, not with the U.S. figure; the tool above does that.
• A figure near or above twice the state median is worth raising: ask the utility or the state commission for the reliability report it files.
• If you depend on power for medical equipment, home work or a freezer, a high SAIDI with major events is the number to look at; for how often it will flicker, look at SAIFI.
• Small utilities (under 10,000 customers) can move a lot from year to year. They are listed but not ranked.
• Smart meters, time-based pricing and efficiency programs are separate questions: see the utility report card.

Browse by state

• Alabama
• Alaska
• Arizona
• Arkansas
• California
• Colorado
• Connecticut
• Delaware
• District of Columbia
• Florida
• Georgia
• Hawaii
• Idaho
• Illinois
• Indiana
• Iowa
• Kansas
• Kentucky
• Louisiana
• Maine
• Maryland
• Massachusetts
• Michigan
• Minnesota
• Mississippi
• Missouri
• Montana
• Nebraska
• Nevada
• New Hampshire
• New Jersey
• New Mexico
• New York
• North Carolina
• North Dakota
• Ohio
• Oklahoma
• Oregon
• Pennsylvania
• Rhode Island
• South Carolina
• South Dakota
• Tennessee
• Texas
• Utah
• Vermont
• Virginia
• Washington
• West Virginia
• Wisconsin
• Wyoming

Best and worst utilities · median-utility ranking · states with the most major outage reports

Method, sources and dates

Source. EIA, Form EIA-861 annual electric power industry report, data year 2024, file eia861_2024.zip, retrieved 2026-10-01. Sheets used: Reliability, Sales to Ultimate Customers, Advanced Meters, Dynamic Pricing, Demand Response, Energy Efficiency, Net Metering, Operational Data. EIA publishes the utilities' own filings; we add nothing to a utility's figure.

State figures. A state figure is the average of the utilities' published numbers, weighted by the customers each utility reported next to them. Nothing is estimated. Only utilities that follow the IEEE 1366 standard are counted: EIA's form has a second block for utilities that use another method, and the two do not measure the same thing, so they are never added together. No utility reports in both blocks.

When a state gets a figure. The utilities that report under IEEE 1366 must serve at least 60% of the customers EIA lists for the state and at least 100,000 customers. Otherwise the state page shows the utilities and the reason, and the state is left out of rankings. Utility ranking floor: 10,000 customers; below that one bad day moves the average too far.

With and without major events. The comparison uses only utilities that published both numbers. One utility, Nashville Electric Service (Tennessee), published an all-events SAIDI of 0.0 against 86.6 minutes without major events, which cannot both be right; its all-events figure is dropped and its without-events figure kept. 82 of 102 state figures match EIA's own State Totals sheet within 1%. The rest differ, the largest being Tennessee and Mississippi; we have not traced which utilities EIA counted there, and we do not use that sheet.

State CSV, utility CSV.