The Uptime report

Uptime Report Research · Grid Reliability

Power Outage Statistics: Outage Hours Rose 46% in a Decade, and Major Storms Caused 91% of the Increase

Power outage statistics from the federal government’s own utility reports show the average U.S. electricity customer lost 7.0 hours of power a year from 2021 to 2025, up from 4.8 hours in 2013 to 2017. In 2025 the figure was 4.5 hours (269 minutes) across 1.34 interruptions. Major event days, mostly storms, explain 91% of the decade’s increase.

4.5 hwithout power per customer in 2025 (269 minutes)
11.0 hin 2024, the worst year since reporting began in 2013
+10%everyday outage time, 2013 to 2017 vs 2021 to 2025
+70%outage time on major event days, same comparison

We downloaded every annual Form EIA-861 reliability file from the U.S. Energy Information Administration (2013 to 2025, the 2025 final data released October 7, 2026), which lists each utility’s SAIDI (minutes without power per customer) and SAIFI (interruptions per customer), with and without major event days. We combined 752 utility-state reports for 2025 into customer-weighted state and national figures. Our national numbers reproduce the EIA’s published series to within one minute. We then matched the result against 865 Department of Energy disturbance reports for large outages.

Power outage statistics by year: everyday outages vs. major events

0h2h4h6h8h10h12h2013: 3.8 hours total (227 min), 1.9 h everyday outages, 1.9 h on major event days3.820132014: 3.9 hours total (236 min), 1.9 h everyday outages, 2.0 h on major event days3.920142015: 3.5 hours total (209 min), 2.0 h everyday outages, 1.5 h on major event days3.520152016: 4.5 hours total (268 min), 2.0 h everyday outages, 2.5 h on major event days4.520162017: 8.4 hours total (506 min), 1.9 h everyday outages, 6.5 h on major event days8.420172018: 5.8 hours total (349 min), 2.0 h everyday outages, 3.8 h on major event days5.820182019: 4.9 hours total (296 min), 2.0 h everyday outages, 2.9 h on major event days4.920192020: 7.6 hours total (456 min), 1.9 h everyday outages, 5.7 h on major event days7.620202021: 7.9 hours total (476 min), 2.1 h everyday outages, 5.8 h on major event days7.920212022: 5.6 hours total (333 min), 2.2 h everyday outages, 3.4 h on major event days5.620222023: 6.1 hours total (367 min), 2.1 h everyday outages, 4.0 h on major event days6.120232024: 11.0 hours total (663 min), 2.2 h everyday outages, 8.8 h on major event days11.020242025: 4.5 hours total (269 min), 2.1 h everyday outages, 2.4 h on major event days4.52025Major event daysEveryday outages (without major event days)

Average hours without power per customer, 50 states and DC, IEEE 1366 method. Dark: interruptions on normal days. Red: interruptions on major event days, which each utility identifies under the IEEE 1366 standard (usually severe weather). Source: EIA Form EIA-861, The Uptime Report analysis.

The grid’s everyday performance barely moved. Without major event days, customers lost 1.9 hours a year in 2013 to 2017 and 2.1 hours in 2021 to 2025. The number of everyday interruptions was flat (1.04 vs 1.06 a year); the growth came from each one lasting longer, with the average everyday interruption up from 111 to 120 minutes.

Major event days are where the volatility lives. They added 2.9 hours a year on average in 2013 to 2017 and 4.9 hours in 2021 to 2025. In 2024, they accounted for 80% of all outage time. The EIA attributes most of 2024’s outage time to major events such as Hurricanes Beryl, Helene, and Milton; Helene alone left 5.9 million customers without power across 10 states. 2025 was calmer, at 4.5 hours, the lowest total since 2016.

Is it just more utilities reporting? No. The share of customers covered by IEEE reporting rose from 66% in 2013 to 79% in 2025, so we re-ran the numbers on a fixed panel of 366 utility-state units that reported every year from 2013 to 2025. Same pattern: everyday outage time rose from 1.9 to 2.1 hours, total outage time from 4.9 to 6.7 hours.

Power outages by state: the reliability scatter

70%+ of outage time on major event days 50% to 69% under 50% under half of customers covered

0h5h10h15h20h25h00.511.522.533.54U.S. average 1.42 interruptionsU.S. average 7.0 hoursInterruptions per customer per year (SAIFI)Hours without power per customer per year (SAIDI)Utah: 1.9 hours a year (116 min), 0.91 interruptions a year, 13% of minutes on major event days (2021 to 2025 average)Delaware: 1.6 hours a year (95 min), 0.89 interruptions a year, 30% of minutes on major event days (2021 to 2025 average)Wyoming: 2.7 hours a year (162 min), 1.17 interruptions a year, 31% of minutes on major event days (2021 to 2025 average)New Mexico: 3.2 hours a year (194 min), 1.19 interruptions a year, 32% of minutes on major event days (2021 to 2025 average)Arizona: 1.6 hours a year (99 min), 0.97 interruptions a year, 32% of minutes on major event days (2021 to 2025 average)District of Columbia: 0.8 hours a year (48 min), 0.34 interruptions a year, 33% of minutes on major event days (2021 to 2025 average)New Jersey: 2.2 hours a year (133 min), 1.00 interruptions a year, 33% of minutes on major event days (2021 to 2025 average)Hawaii: 5.4 hours a year (322 min), 3.92 interruptions a year, 35% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingIdaho: 4.1 hours a year (247 min), 1.30 interruptions a year, 43% of minutes on major event days (2021 to 2025 average)Montana: 4.1 hours a year (244 min), 1.41 interruptions a year, 44% of minutes on major event days (2021 to 2025 average)Alabama: 4.5 hours a year (267 min), 2.01 interruptions a year, 44% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingColorado: 3.2 hours a year (192 min), 1.16 interruptions a year, 45% of minutes on major event days (2021 to 2025 average)Iowa: 2.3 hours a year (140 min), 1.03 interruptions a year, 45% of minutes on major event days (2021 to 2025 average)Nevada: 2.4 hours a year (145 min), 1.00 interruptions a year, 46% of minutes on major event days (2021 to 2025 average)California: 5.1 hours a year (307 min), 1.33 interruptions a year, 50% of minutes on major event days (2021 to 2025 average)Alaska: 9.4 hours a year (563 min), 3.19 interruptions a year, 53% of minutes on major event days (2021 to 2025 average)West Virginia: 16.8 hours a year (1010 min), 2.68 interruptions a year, 53% of minutes on major event days (2021 to 2025 average)New York: 5.0 hours a year (299 min), 1.40 interruptions a year, 53% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingMaryland: 2.8 hours a year (170 min), 0.96 interruptions a year, 55% of minutes on major event days (2021 to 2025 average)Minnesota: 3.1 hours a year (186 min), 1.13 interruptions a year, 55% of minutes on major event days (2021 to 2025 average)Pennsylvania: 4.9 hours a year (292 min), 1.27 interruptions a year, 56% of minutes on major event days (2021 to 2025 average)Connecticut: 3.0 hours a year (178 min), 0.92 interruptions a year, 57% of minutes on major event days (2021 to 2025 average)South Dakota: 3.0 hours a year (178 min), 1.07 interruptions a year, 57% of minutes on major event days (2021 to 2025 average)Arkansas: 8.9 hours a year (533 min), 2.05 interruptions a year, 58% of minutes on major event days (2021 to 2025 average)Illinois: 2.4 hours a year (146 min), 0.84 interruptions a year, 59% of minutes on major event days (2021 to 2025 average)Kansas: 4.3 hours a year (259 min), 1.40 interruptions a year, 59% of minutes on major event days (2021 to 2025 average)Tennessee: 7.3 hours a year (438 min), 2.33 interruptions a year, 61% of minutes on major event days (2021 to 2025 average)Ohio: 5.8 hours a year (348 min), 1.33 interruptions a year, 61% of minutes on major event days (2021 to 2025 average)Indiana: 5.8 hours a year (347 min), 1.38 interruptions a year, 64% of minutes on major event days (2021 to 2025 average)Mississippi: 12.2 hours a year (731 min), 2.41 interruptions a year, 64% of minutes on major event days (2021 to 2025 average)Vermont: 9.2 hours a year (553 min), 2.40 interruptions a year, 65% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingMissouri: 4.9 hours a year (295 min), 1.24 interruptions a year, 65% of minutes on major event days (2021 to 2025 average)Massachusetts: 3.9 hours a year (237 min), 1.03 interruptions a year, 66% of minutes on major event days (2021 to 2025 average)Washington: 7.1 hours a year (427 min), 1.51 interruptions a year, 66% of minutes on major event days (2021 to 2025 average)Georgia: 7.2 hours a year (435 min), 1.75 interruptions a year, 66% of minutes on major event days (2021 to 2025 average)Virginia: 13.1 hours a year (785 min), 2.15 interruptions a year, 66% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingRhode Island: 3.1 hours a year (184 min), 1.01 interruptions a year, 67% of minutes on major event days (2021 to 2025 average)North Dakota: 4.5 hours a year (271 min), 1.14 interruptions a year, 67% of minutes on major event days (2021 to 2025 average)Wisconsin: 4.8 hours a year (287 min), 1.02 interruptions a year, 68% of minutes on major event days (2021 to 2025 average)Oklahoma: 7.3 hours a year (438 min), 1.67 interruptions a year, 72% of minutes on major event days (2021 to 2025 average); under half of customers covered by IEEE reportingNorth Carolina: 8.9 hours a year (535 min), 1.59 interruptions a year, 73% of minutes on major event days (2021 to 2025 average)Kentucky: 9.5 hours a year (568 min), 1.75 interruptions a year, 74% of minutes on major event days (2021 to 2025 average)Michigan: 12.1 hours a year (726 min), 1.46 interruptions a year, 77% of minutes on major event days (2021 to 2025 average)Maine: 17.5 hours a year (1053 min), 2.94 interruptions a year, 78% of minutes on major event days (2021 to 2025 average)New Hampshire: 7.2 hours a year (430 min), 1.35 interruptions a year, 80% of minutes on major event days (2021 to 2025 average)Oregon: 9.7 hours a year (584 min), 1.23 interruptions a year, 81% of minutes on major event days (2021 to 2025 average)Texas: 13.7 hours a year (822 min), 2.19 interruptions a year, 81% of minutes on major event days (2021 to 2025 average)Nebraska: 6.5 hours a year (388 min), 0.98 interruptions a year, 83% of minutes on major event days (2021 to 2025 average)Louisiana: 25.9 hours a year (1552 min), 2.36 interruptions a year, 85% of minutes on major event days (2021 to 2025 average)South Carolina: 13.0 hours a year (781 min), 1.63 interruptions a year, 86% of minutes on major event days (2021 to 2025 average)Florida: 9.4 hours a year (565 min), 1.19 interruptions a year, 88% of minutes on major event days (2021 to 2025 average)UTDEAZDCHIAKWVMSVANCKYMIMEORTXNELASCFL

Each dot is a state, averaged over 2021 to 2025, all events included. Up means more hours without power; right means more separate interruptions. Hover for values.

The scatter separates two different problems. The storm states sit high but not far right: South Carolina, Michigan, Oregon, and Florida have interruption counts near the national average but lost 9.4 to 13.0 hours a year, with 70% or more of that time on major event days. In South Carolina, the 2024 total reached 52.6 hours. Further right sit the frequent-outage states, where customers see many separate interruptions: Alaska (3.2 a year), Maine (2.9), and West Virginia (2.7), against a national average of 1.4.

West Virginia is the outlier on both axes. Even with major event days removed, its customers lost 7.9 hours a year, the most of any state and 3.7 times the national everyday figure. Across the decade, 38 of 51 states lost more time per customer in 2021 to 2025 than in 2013 to 2017. Because severe weather drives most of this, our analysis of FEMA disaster declarations by state is a useful second input when scoring a client site’s continuity risk.

Puerto Rico, which the national figures exclude, is in a different category. Its utility reported 59.1 hours without power per customer in 2025, 35.9 of them outside major event days. Its everyday figure averaged 27.3 hours a year from 2021 to 2024, in line with the 27 hours the EIA has reported for that period.

Most hours without power

  1. Louisiana25.9 h
  2. Maine17.5 h
  3. West Virginia16.8 h
  4. Texas13.7 h
  5. Virginia†13.1 h

Fewest hours without power

  1. District of Columbia0.8 h
  2. Delaware1.6 h
  3. Arizona1.6 h
  4. Utah1.9 h
  5. New Jersey2.2 h

Five-year averages, 2021 to 2025, all events. † Under half of customers covered by IEEE reporting (see the note under the full table).

State ranking: all 50 states and DC

Click a column header to sort. Single years swing hard: one hurricane can multiply a state’s total, so the ranking uses the five-year average.

Rank State Hours a year, 2021 to 2025 Everyday hours (no major events) Interruptions a year Share on major event days Hours, 2025 Hours, 2024 Change vs 2013 to 2017 Customers covered, 2025
1 Louisiana 25.9 3.9 2.36 85% 4.5 10.5 +232% 73%
2 Maine 17.5 3.8 2.94 78% 6.1 29.1 0% 100%
3 West Virginia 16.8 7.9 2.68 53% 16.9 19.4 +46% 98%
4 Texas 13.7 2.6 2.19 81% 3.5 26.9 +209% 66%
5 Virginia† 13.1 4.5 2.15 66% 12.7 16.0 +95% 30%
6 South Carolina 13.0 1.9 1.63 86% 2.5 52.6 +21% 95%
7 Mississippi 12.2 4.4 2.41 64% 8.7 10.1 +115% 69%
8 Michigan 12.1 2.8 1.46 77% 10.1 8.0 +28% 94%
9 Oregon 9.7 1.9 1.23 81% 5.0 9.7 +140% 87%
10 Kentucky 9.5 2.5 1.75 74% 6.6 9.0 +168% 94%
11 Florida 9.4 1.1 1.19 88% 1.2 24.8 -15% 84%
12 Alaska 9.4 4.4 3.19 53% 9.4 5.4 +66% 80%
13 Vermont† 9.2 3.3 2.40 65% 4.2 14.9 +4% 11%
14 North Carolina 8.9 2.4 1.59 73% 3.2 25.3 +26% 90%
15 Arkansas 8.9 3.7 2.05 58% 7.9 8.8 +64% 82%
16 Tennessee 7.3 2.9 2.33 61% 3.8 4.8 +41% 69%
17 Oklahoma† 7.3 2.1 1.67 72% 5.0 3.7 +26% 46%
18 Georgia 7.2 2.5 1.75 66% 4.5 20.1 +4% 88%
19 New Hampshire 7.2 1.5 1.35 80% 2.5 12.3 -4% 88%
20 Washington 7.1 2.5 1.51 66% 6.3 10.7 +20% 89%
21 Nebraska 6.5 1.1 0.98 83% 9.0 11.6 +292% 66%
22 Ohio 5.8 2.2 1.33 61% 3.4 8.5 +77% 97%
23 Indiana 5.8 2.1 1.38 64% 6.3 5.2 +42% 90%
24 Hawaii† 5.4 3.5 3.92 35% 5.6 6.4 +87% 7%
25 California 5.1 2.6 1.33 50% 6.0 4.8 +120% 93%
26 New York† 5.0 2.3 1.40 53% 4.1 8.0 +28% 24%
27 Missouri 4.9 1.7 1.24 65% 7.8 3.5 +33% 83%
28 Pennsylvania 4.9 2.1 1.27 56% 6.9 5.9 +37% 84%
29 Wisconsin 4.8 1.5 1.02 68% 4.4 4.9 +81% 74%
30 North Dakota 4.5 1.5 1.14 67% 3.5 1.7 +143% 84%
31 Alabama† 4.5 2.5 2.01 44% 4.5 3.3 +7% 15%
32 Kansas 4.3 1.7 1.40 59% 3.2 4.8 0% 80%
33 Idaho 4.1 2.3 1.30 43% 4.9 3.4 +13% 93%
34 Montana 4.1 2.3 1.41 44% 4.8 5.7 +23% 71%
35 Massachusetts 3.9 1.4 1.03 66% 2.0 1.7 +21% 92%
36 New Mexico 3.2 2.2 1.19 32% 2.7 4.5 +70% 87%
37 Colorado 3.2 1.8 1.16 45% 4.5 4.3 +54% 93%
38 Minnesota 3.1 1.4 1.13 55% 3.0 3.7 -16% 86%
39 Rhode Island 3.1 1.0 1.01 67% 3.0 1.7 -56% 97%
40 Connecticut 3.0 1.3 0.92 57% 3.1 2.5 -5% 99%
41 South Dakota 3.0 1.3 1.07 57% 4.7 1.4 -45% 72%
42 Maryland 2.8 1.3 0.96 55% 3.0 2.0 +19% 99%
43 Wyoming 2.7 1.9 1.17 31% 3.7 2.3 -30% 72%
44 Illinois 2.4 1.0 0.84 59% 2.4 2.6 -9% 96%
45 Nevada 2.4 1.3 1.00 46% 2.1 2.6 +57% 100%
46 Iowa 2.3 1.3 1.03 45% 3.1 3.3 +7% 53%
47 New Jersey 2.2 1.5 1.00 33% 2.6 2.9 -13% 99%
48 Utah 1.9 1.7 0.91 13% 1.9 1.9 -38% 86%
49 Arizona 1.6 1.1 0.97 32% 1.1 1.4 +16% 95%
50 Delaware 1.6 1.1 0.89 30% 1.7 2.0 -42% 87%
51 District of Columbia 0.8 0.5 0.34 33% 0.8 0.5 -52% 99%

† Under half of the state’s customers are served by utilities reporting under the IEEE standard in 2025 (Alabama, Hawaii, New York, Oklahoma, Virginia, Vermont), so the figure reflects a minority of customers. Including utilities that report under another standard gives these five-year averages: Alabama 4.6 hours; Hawaii 5.0 hours; New York 2.7 hours; Oklahoma 6.1 hours; Virginia 7.0 hours; Vermont 9.9 hours. Customer-weighted averages of utility SAIDI and SAIFI; coverage is capped at 100% where utility counts slightly exceed the EIA sales-file total.

Year by year: the national series

Year Minutes, all events Hours, all events Minutes, without major events Minutes on major event days Major event share Interruptions, all Interruptions, without major events Utility reports Customers covered
2013 227 3.8 112 115 51% 1.19 0.99 541 66%
2014 236 3.9 114 122 52% 1.26 1.04 598 74%
2015 209 3.5 117 92 44% 1.28 1.07 643 77%
2016 268 4.5 120 149 55% 1.33 1.08 691 78%
2017 506 8.4 117 389 77% 1.42 1.02 731 78%
2018 349 5.8 121 228 65% 1.34 1.05 771 76%
2019 296 4.9 122 173 59% 1.33 1.04 809 78%
2020 456 7.6 116 340 75% 1.39 1.01 689 77%
2021 476 7.9 126 351 74% 1.44 1.04 702 78%
2022 333 5.6 131 202 61% 1.43 1.09 712 78%
2023 367 6.1 124 243 66% 1.35 1.02 732 78%
2024 663 11.0 132 531 80% 1.53 1.06 744 79%
2025 269 4.5 125 144 53% 1.34 1.08 752 79%

Major disturbances timeline: what the Department of Energy reports show

Utilities and grid operators must file Form OE-417 with the Department of Energy when an incident cuts power to more than 50,000 customers for an hour or more. From 2015 (when the summaries began recording the alert criterion) through 2025, 865 reports cited that criterion, and 97% gave weather or natural disaster as the cause. A single storm can produce several reports, one per affected utility, so the chart counts reports, not storms.

weather or natural disaster other causes

01020302015-02: 8 large outage reports (8 weather)2015-03: 1 large outage reports (1 weather)2015-04: 6 large outage reports (5 weather)2015-05: 5 large outage reports (5 weather)2015-06: 8 large outage reports (7 weather)2015-07: 5 large outage reports (5 weather)2015-08: 9 large outage reports (8 weather)2015-09: 1 large outage reports (1 weather)2015-10: 3 large outage reports (2 weather)2015-11: 6 large outage reports (6 weather)2015-12: 7 large outage reports (6 weather)2016-01: 3 large outage reports (3 weather)2016-02: 5 large outage reports (5 weather)2016-03: 2 large outage reports (2 weather)2016-04: 2 large outage reports (2 weather)2016-05: 4 large outage reports (1 weather)2016-06: 1 large outage reports (1 weather)2016-07: 12 large outage reports (12 weather)2016-08: 2 large outage reports (2 weather)2016-09: 5 large outage reports (5 weather)2016-10: 7 large outage reports (7 weather)2016-12: 1 large outage reports (1 weather)2017-01: 6 large outage reports (5 weather)2017-02: 4 large outage reports (4 weather)2017-03: 12 large outage reports (12 weather)2017-04: 7 large outage reports (6 weather)2017-05: 7 large outage reports (7 weather)2017-06: 7 large outage reports (7 weather)2017-07: 5 large outage reports (5 weather)2017-08: 5 large outage reports (5 weather)2017-09: 8 large outage reports (8 weather)2017-10: 5 large outage reports (5 weather)2017-11: 3 large outage reports (3 weather)2017-12: 5 large outage reports (5 weather)2018-03: 20 large outage reports (20 weather)2018-04: 5 large outage reports (5 weather)2018-05: 15 large outage reports (14 weather)2018-06: 4 large outage reports (4 weather)2018-07: 5 large outage reports (5 weather)2018-08: 3 large outage reports (3 weather)2018-09: 4 large outage reports (4 weather)2018-10: 11 large outage reports (11 weather)2018-11: 8 large outage reports (8 weather)2018-12: 6 large outage reports (6 weather)2019-01: 7 large outage reports (7 weather)2019-02: 10 large outage reports (10 weather)2019-03: 4 large outage reports (4 weather)2019-04: 5 large outage reports (5 weather)2019-05: 6 large outage reports (6 weather)2019-06: 10 large outage reports (10 weather)2019-07: 10 large outage reports (10 weather)2019-08: 4 large outage reports (2 weather)2019-09: 7 large outage reports (6 weather)2019-10: 9 large outage reports (9 weather)2019-11: 6 large outage reports (6 weather)2020-01: 3 large outage reports (3 weather)2020-02: 8 large outage reports (8 weather)2020-03: 4 large outage reports (4 weather)2020-04: 23 large outage reports (23 weather)2020-05: 6 large outage reports (6 weather)2020-06: 7 large outage reports (7 weather)2020-07: 6 large outage reports (6 weather)2020-08: 28 large outage reports (28 weather)2020-09: 8 large outage reports (7 weather)2020-10: 17 large outage reports (17 weather)2020-11: 8 large outage reports (8 weather)2020-12: 7 large outage reports (7 weather)2021-01: 9 large outage reports (9 weather)2021-02: 10 large outage reports (10 weather)2021-03: 3 large outage reports (3 weather)2021-04: 1 large outage reports (1 weather)2021-05: 3 large outage reports (3 weather)2021-06: 4 large outage reports (3 weather)2021-07: 8 large outage reports (8 weather)2021-08: 13 large outage reports (13 weather)2021-09: 10 large outage reports (10 weather)2021-10: 7 large outage reports (7 weather)2021-11: 2 large outage reports (2 weather)2021-12: 12 large outage reports (12 weather)2022-01: 7 large outage reports (7 weather)2022-02: 5 large outage reports (5 weather)2022-03: 5 large outage reports (5 weather)2022-04: 4 large outage reports (4 weather)2022-05: 3 large outage reports (3 weather)2022-06: 11 large outage reports (11 weather)2022-07: 4 large outage reports (4 weather)2022-08: 8 large outage reports (7 weather)2022-09: 9 large outage reports (9 weather)2022-10: 2 large outage reports (1 weather)2022-11: 8 large outage reports (7 weather)2022-12: 11 large outage reports (11 weather)2023-01: 6 large outage reports (6 weather)2023-02: 7 large outage reports (7 weather)2023-03: 15 large outage reports (15 weather)2023-04: 2 large outage reports (2 weather)2023-05: 1 large outage reports (1 weather)2023-06: 6 large outage reports (6 weather)2023-07: 18 large outage reports (18 weather)2023-08: 1 large outage reports (1 weather)2023-09: 4 large outage reports (4 weather)2023-10: 1 large outage reports (1 weather)2023-11: 2 large outage reports (2 weather)2023-12: 1 large outage reports (1 weather)2024-01: 15 large outage reports (15 weather)2024-02: 6 large outage reports (6 weather)2024-03: 8 large outage reports (7 weather)2024-04: 8 large outage reports (8 weather)2024-05: 19 large outage reports (19 weather)2024-06: 9 large outage reports (9 weather)2024-07: 6 large outage reports (6 weather)2024-08: 10 large outage reports (9 weather)2024-09: 14 large outage reports (13 weather)2024-10: 5 large outage reports (5 weather)2024-11: 7 large outage reports (7 weather)2024-12: 7 large outage reports (7 weather)2025-01: 7 large outage reports (7 weather)2025-02: 7 large outage reports (6 weather)2025-03: 10 large outage reports (10 weather)2025-04: 5 large outage reports (5 weather)2025-05: 6 large outage reports (6 weather)2025-06: 5 large outage reports (5 weather)2025-07: 12 large outage reports (12 weather)2025-08: 1 large outage reports (1 weather)2025-10: 2 large outage reports (2 weather)2025-11: 2 large outage reports (1 weather)2025-12: 7 large outage reports (7 weather)1. 2018-03: Winter storms12. 2020-04: Severe storms23. 2020-08: Storms Isaias and Laura34. 2020-10: Storms Zeta and Delta45. 2023-03: Severe storms56. 2023-07: Severe storms67. 2024-01: Severe storms78. 2024-05: Severe storms820152016201720182019202020212022202320242025

Monthly count of OE-417 reports for outages affecting more than 50,000 customers. Numbered markers flag the eight busiest months, listed in the table below with the matching FEMA declaration titles for the states involved (OpenFEMA). Source: DOE OE-417 annual summaries, The Uptime Report analysis.

Months with 13 or more weather-related large outage reports:

Chart marker Month Weather-related large outage reports Most-reported states Matching FEMA declaration
1 Mar 2018 20 NY, RI, MA Severe Winter Storm and Snowstorm; Severe Storm and Flooding
May 2018 14 NY, MI, NH Severe Storms, Tornadoes, and Straight-Line Winds; Severe Storm and Flooding
2 Apr 2020 23 MS, TX, KS Severe Storms, Tornadoes, Straight-Line Winds, and Flooding; Severe Storms, Tornadoes, and Straight-Line Winds
3 Aug 2020 28 NY, LA, NJ Tropical Storm Isaias; Hurricane Laura
4 Oct 2020 17 LA, MS, TX Hurricane Zeta; Hurricane Delta
Aug 2021 13 MI, LA, IL Hurricane Henri; Tropical Storm Ida
5 Mar 2023 15 WA, KY, VA Severe Storms, Straight-Line Winds, and Tornadoes; Severe Storms, Straight-Line Winds, Tornadoes, Flooding, Landslides, and Mudslides
6 Jul 2023 18 MI, MD, MO Severe Storms, Straight-Line Winds, Tornadoes, and Flooding; Severe Storms and Flooding
7 Jan 2024 15 WA, NC, MI Severe Storms and Flooding; Severe Winter Storms, Straight-Line Winds, Landslides, and Mudslides
8 May 2024 19 TX, FL, KS Severe Storms, Straight-Line Winds, Tornadoes, and Flooding; Severe Storms, Straight-Line Winds, Tornadoes, Landslides, and Mudslides
Sep 2024 13 SC, FL, NC Hurricane Helene; Hurricane Milton

Every OE-417 report by cause

The full set of OE-417 reports is broader than outages: it also covers attacks on equipment, cyber incidents, fuel shortages, and grid emergencies that did not black anyone out. Physical security reports have outnumbered weather reports in every year since 2022, but none of the 865 large outage reports was classified as physical security. The form, its alert criteria, and the filing portal have changed over time, so year-to-year totals are not strictly comparable.

Year Weather Physical security Grid operations and equipment Cyber Fuel supply Other or unknown All reports Large outage reports Weather share of large
2013 56 79 30 2 6 1 174 n/a n/a
2014 87 73 31 3 17 3 214 n/a n/a
2015 66 42 29 0 0 6 143 59 92%
2016 42 44 38 4 7 6 141 44 93%
2017 80 45 18 0 7 0 150 74 97%
2018 102 58 55 0 5 0 220 81 99%
2019 92 81 98 0 7 0 278 78 96%
2020 166 94 113 3 7 0 383 125 99%
2021 156 92 121 7 11 0 387 82 99%
2022 95 163 118 9 5 0 390 77 96%
2023 81 175 50 4 1 37 348 67 100%
2024 133 180 69 62 43 135 622 114 97%
2025 72 131 12 42 61 108 426 64 97%

What this means for MSPs

Size UPS runtime for shutdown, not ride-through. In 2025, the average interruption outside major event days lasted 116 minutes. A small UPS sized for a few minutes of runtime will not carry a server room through that; its job is to keep network gear up for brief flickers and to trigger a clean shutdown of servers and storage. Test that the shutdown agent actually fires, because that is the outcome most outages will produce.

Set a generator threshold in hours, by state. Interruptions attributable to major event days averaged roughly 9.2 hours each in 2025 and 19.0 hours in 2024. For clients in the top of the table above, where five-year averages run from 9 to 26 hours a year, a standby generator or an agreed relocation plan belongs in the continuity conversation. For clients in the District of Columbia, Delaware, or Arizona, under two hours a year, a UPS plus cloud failover usually covers the risk.

Compare the grid to your SLA. The average U.S. customer lost 422 minutes a year in 2021 to 2025, which works out to 99.92% grid availability, close to the 525.6 minutes a year allowed by a 99.9% SLA. In Louisiana it was 99.70%. An on-premises system with no generator cannot promise a higher number than the utility feeding it.

Do not forget the phones and the internet edge. Business phone service is now mostly IP-based (see our VoIP statistics), so a power cut takes down the phones along with the firewall and switches unless they sit on the UPS. Cloud-hosted workloads keep running during a local outage, but staff only reach them if the office edge or their homes have power and connectivity.

Treat it as a year-round risk. Large weather outages are not just a hurricane-season problem. From 2015 to 2025, every calendar month logged at least 50 weather-related large outage reports to the DOE (November was the quietest), and the busiest month, July, logged 91. June to September accounted for 37% of them, but December to February accounted for 23% and March to May for 26%, so generator load tests and UPS battery checks belong on a fixed schedule, not just before late summer.

Frequently asked questions

How many hours of power outages does the average American have per year?

In 2025, the average U.S. electricity customer was without power for 269 minutes (4.5 hours) across 1.34 interruptions, based on utility reports to the EIA under the IEEE standard. The 2021 to 2025 average was 7.0 hours a year, and 2024 was the worst year in the series at 11.0 hours.

What states have the most power outages?

Averaged over 2021 to 2025, customers in Louisiana lost the most time (25.9 hours a year), followed by Maine (17.5) and West Virginia (16.8). West Virginia has the most everyday outage time once major event days are removed (7.9 hours a year). The District of Columbia has the least (0.8 hours).

Are power outages getting worse in the US?

Yes, mostly because of major events. Average outage time rose from 4.8 hours a year in 2013 to 2017 to 7.0 hours in 2021 to 2025. Time lost on major event days grew 70%, while everyday outage time grew 10%. Major events account for 91% of the increase.

What causes most large power outages?

Weather. Of the 865 reports utilities filed with the Department of Energy from 2015 to 2025 for outages affecting more than 50,000 customers, 841 (97%) gave weather or natural disaster as the cause.

How long does a typical power outage last?

In 2025, an average interruption outside major event days lasted 116 minutes (about 2 hours). Interruptions attributable to major event days averaged roughly 9.2 hours in 2025 and 19.0 hours in 2024.

Download the data. power-outage-statistics-by-state-2013-2025.csv (CSV). Licensed under CC BY 4.0: free to reuse with attribution to The Uptime Report.

Methodology

Sources. U.S. Energy Information Administration, Form EIA-861 annual files 2013 to 2025 (Reliability and Sales to Ultimate Customers tables; 2025 final data released October 7, 2026). U.S. Department of Energy, Form OE-417 Electric Emergency Incident and Disturbance annual summaries, 2013 to 2025. FEMA OpenFEMA Disaster Declarations Summaries (v2). All retrieved October 2026.

Definitions. SAIDI is the total minutes of sustained interruption per customer served in a year; SAIFI is the number of sustained interruptions per customer. Major event days follow IEEE Standard 1366 (2003 or 2012 edition), under which each utility identifies days beyond its normal operating conditions; the EIA notes these are usually severe weather but can include vegetation interference and atypical utility operations. We follow the EIA-861 instructions: an interruption counts if it lasts more than 5 minutes, and momentary interruptions are excluded. “Everyday outages” means SAIDI without major event days.

Aggregation. State and national figures are customer-weighted averages of each utility’s reported SAIDI and SAIFI, using only utilities reporting under the IEEE standard, as the EIA does. Multi-state utilities report each state separately. National figures cover the 50 states and DC (territories excluded). Five-year averages are simple means of the five annual values. Average interruption length is SAIDI divided by SAIFI; the length attributable to major event days is the difference in SAIDI with and without major event days divided by the difference in SAIFI, an approximation because each figure is a separate customer-weighted average. Coverage is IEEE-reporting customers divided by total customers in the EIA sales file (bundled, delivery, and adjustment rows).

Checks. Our national series matches EIA Electric Power Annual Table 11.1 exactly for every year from 2014 to 2024 except 2021 (0.6 minutes higher, likely a later data revision), including 662.6 minutes (11.0 hours) for 2024, which the EIA reported as 11 hours with 80% from major events, and 333.0 minutes for 2022, which the EIA reported as about five and one-half hours. Our 2024 state figures match the EIA’s published state values (Table 11.4 and the EIA-861 State Totals sheet) within one minute in 51 of 51 states (for example, South Carolina 3155.0 minutes, which the EIA described as nearly 53 hours). For 2025, the EIA’s own “State Totals” sheet omits several large utilities that appear in its utility-level file (for example PacifiCorp in Utah, Oregon, and Wyoming, and Delmarva Power in Delaware), so our 2025 figures differ from that sheet by more than one minute in 16 states; ours include every IEEE utility report.

OE-417. Large outage reports are rows whose alert criterion is loss of electric service to more than 50,000 customers for one hour or more (criterion recorded from 2015). Cause categories are our grouping of the free-text event type: anything mentioning weather, natural disaster, storms, or wildfire counts as weather. Customer counts in OE-417 are initial estimates and often double-count across reports, so we do not sum them. Thirty-two 2023 rows with a blank start date are counted in 2023 and dated by restoration date. FEMA titles are matched to the busiest months by state and incident start date within the month.

Limitations. About 21% of U.S. customers are served by utilities that report under another standard or not at all (many small cooperatives and municipal utilities file a short form without reliability data). SAIDI is an average: individual buildings can see far more or less. Major event day thresholds are utility-specific, so the split between everyday and major event outages is not identical across utilities.

Not advice. This is an analysis of public reliability data, not an engineering assessment of any site’s power protection needs.

Cite as:The Uptime Report. "Power Outage Statistics: Outage Hours Rose 46% in a Decade, and Major Storms Caused 91% of the Increase." October 2026. https://theuptimereport.com/power-outage-statistics/

References

  1. U.S. EIA, Form EIA-861 detailed data files (2013 to 2025). Accessed October 2026.
  2. U.S. EIA, Electric Power Annual, Table 11.1: Reliability metrics of U.S. distribution system. Accessed October 2026.
  3. U.S. EIA, Electric Power Annual, Table 11.4: SAIDI values of U.S. distribution system by state. Accessed October 2026.
  4. U.S. EIA, “Hurricanes in 2024 led to the most hours without power in the United States in 10 years” (December 1, 2025). Accessed October 2026.
  5. U.S. EIA, “Electricity customers in Puerto Rico lose about 27 hours of power per year”. Accessed October 2026.
  6. U.S. EIA, “U.S. electricity customers averaged five and one-half hours of power interruptions in 2022” (January 25, 2024). Accessed October 2026.
  7. U.S. EIA, Form EIA-861 instructions, Schedule 3 Part B (distribution system reliability, IEEE). Accessed October 2026.
  8. U.S. Department of Energy, OE-417 Electric Emergency Incident and Disturbance Reports, annual summaries. Accessed October 2026.
  9. FEMA, OpenFEMA Disaster Declarations Summaries v2. Accessed October 2026.

Research library

More original data reports from The Uptime Report. Every report includes its methodology and a free CSV download. Browse all research

Threats and breaches

Infrastructure and uptime

Regulation and disclosure

Workforce and pay

MSP market and spending

Comments

One response to “Power Outage Statistics: Outage Hours Rose 46% in a Decade, and Major Storms Caused 91% of the Increase”

  1. CVE Statistics 2026: Vulnerabilities Per Year (NVD)

    […] Power Outage Statistics by StateOutage hours by state, EIA data […]

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

More posts