Power outages are common enough to deserve routine preparation. Ready.gov says 25% of households may experience an outage in any year, while the 2023 American Housing Survey recorded about 33.9 million households reporting at least one outage in the previous 12 months. The disruption may last minutes, days, or weeks during a disaster, and its consequences include work loss, spoiled food, unsafe temperatures, and medical-device problems.
Contents
- How common are power outages?
- How long do outages last?
- Where outage exposure is highest
- Household consequences of a blackout
- Medical and energy-security risks
- How prepared are U.S. households?
How common are power outages?
The broadest household estimate comes from Ready.gov Power Outages: in any year, 25% of households may experience a power outage. The 2023 American Housing Survey provides a comparable national result: approximately 33.9 million households, or 1 in 4, reported at least one outage during the prior 12 months. These figures describe household exposure, not the number of separate events or the duration of each event.
The Energy Information Administration (EIA) measured interruption experience from the utility-customer perspective. In 2022, the average U.S. electricity customer experienced about 5.5 hours of power interruptions, or 5.6 hours when major events were included in the reported national average. The same EIA data recorded 1.4 outages per customer on average. In 2020, the average customer experienced just over 8 hours of interruptions.
The apparent differences are important. The household survey measures whether a home experienced an outage, while EIA statistics measure utility interruption duration and frequency. Neither single number describes every resident’s risk. Ready.gov also emphasizes that an outage can last a few minutes or extend for days or weeks when it occurs during a disaster.
The EIA’s outage-frequency and duration reporting covered distribution systems serving approximately 145 million customers. For the utilities reporting outage data in 2015, those utilities represented 28% of all utilities but about 72% of electricity sales. That coverage helps explain why utility datasets are valuable while still requiring attention to their reporting scope.
How long do outages last?
Duration changes sharply when major events are counted. EIA reported that average annual interruption duration was around 2 hours from 2013 through 2020 when major events were excluded. By contrast, the average U.S. customer experienced just over 8 hours of interruptions in 2020 when the year’s overall experience was considered. In 2022, the national average was about 5.5 hours, or 5.6 hours with major events included.
The 2022 state figures show why a national average can conceal substantial variation. Florida customers experienced 19.1 hours of interruptions on average, the highest state total mentioned in the EIA data. New Hampshire customers experienced 10.3 hours. At the low end, District of Columbia customers experienced 34 minutes, while Iowa customers experienced 85 minutes.
| EIA 2022 measure | Reported result |
|---|---|
| U.S. average interruption duration | 5.5 hours |
| U.S. average duration including major events | 5.6 hours |
| U.S. average interruptions per customer | 1.4 |
| Florida average interruption duration | 19.1 hours |
| New Hampshire average interruption duration | 10.3 hours |
| District of Columbia average interruption duration | 34 minutes |
| Iowa average interruption duration | 85 minutes |
| Alaska average interruptions per customer | 3.5 |
Alaska had 3.5 interruptions per customer on average in 2022, the highest frequency among the locations mentioned. The District of Columbia had the fewest interruptions per customer in that comparison. Frequency and duration are different preparedness signals: a community may experience several short interruptions or fewer, longer events.
Where outage exposure is highest
Weather-related events can produce unusually large interruptions. The EIA reported that the United States had 18 weather-related disasters exceeding $1 billion in damage each in 2022. In Florida, Hurricane Ian left more than 2.6 million customers without power in September 2022. Hurricane Nicole left more than 300,000 Florida customers without power in November 2022. Winter Storm Elliott left more than 1.5 million customers without power in December 2022.
The 2020 EIA state figures also show how event-heavy years can dominate local totals. Louisiana customers experienced 60 hours of interrupted power, the highest state total cited for that year. Alabama customers experienced almost 29 hours. At the shorter end of the cited 2020 totals, the District of Columbia experienced 44 minutes and South Dakota experienced 101 minutes.
The 2023 American Housing Survey adds household and community context. Rural households reported outages at a rate of 35.4%, compared with 22.8% of urban households. Homeowners reported a 28.3% outage rate, compared with 19.9% for renters. Among the 15 largest metropolitan areas, Detroit had the highest reported outage rate at 44.8%, followed by San Francisco at 36.3% and Seattle at 29.3%. New York’s rate was 11.2%, while Miami had the lowest rate in that group at 5.8%.
These figures are not forecasts for a particular home. They are measured results for specified populations and periods. Local infrastructure, weather exposure, housing type, and the definition of an outage can all affect comparisons.
Household consequences of a blackout
The 2023 American Housing Survey measured consequences beyond the initial loss of electricity. About 23.6 million outage-reporting households, roughly 70%, had at least one outage lasting 6 hours or more. About 2.4 million housing units had someone miss work because of an outage. That equals 7.9% of households that both experienced an outage and had at least one worker.
Food and medicine losses were also reported. About 4.9 million housing units, or approximately 14% of outage-affected households, reported spoiled food. Among households with refrigerated medicine, 705,000 households, or 2.9%, reported spoiled medicine. About 2.9 million households, or 8.4%, said they had to stay away from home overnight because of an outage.
Cold-weather and water-system effects appeared in the same survey. Approximately 673,000 housing units had frozen pipes in the prior 12 months because of outage-related conditions. Another 343,000 households reported water in a basement or crawl space because a sump pump stopped working during an outage. A blackout plan therefore needs to account for work, food, medicine, temperature, and water—not only lighting.
Medical and energy-security risks
Electricity dependence is especially consequential for households using medical equipment. The 2023 American Housing Survey counted 14.5 million households reporting medical devices that require electricity. Of those households, 31.6% were affected by outages. The survey also found that 28.2% of households with a person with a disability reported an outage, equal to 8.6 million households.
The EIA’s 2020 Residential Energy Consumption Survey measured a related vulnerability. It found that 27% of U.S. households, or 34 million households, reported difficulty meeting energy needs in 2020. In that year, 1.4 million households, or 1%, reported that someone needed medical attention because the home was kept at an unsafe temperature.
The 2020 RECS covered 123.53 million primary occupied housing units and its survey frame covered 99.6% of eligible U.S. addresses. It sampled 18,496 households, the largest responding sample in the program’s history. These are survey-coverage details, not estimates of blackout frequency, but they identify the population context for the energy-security figures.
Household equipment can influence what an outage affects. Near 90% of households used air conditioning in 2020, and 31% reported main heating equipment that was at least 15 years old. Separate freezers were reported by 51% of households in the West North Central division and 40% in the East South Central division. More than three-quarters of electric-vehicle owners reported charging at home in the 2020 RECS.
How prepared are U.S. households?
Preparedness measures have changed across FEMA surveys and should not be treated as one continuous series without noting the year. In the FEMA 2019 National Household Survey, 80% of respondents said they had gathered enough supplies to last at least 3 days, and 48% said they had an emergency action plan. In the FEMA 2024 National Household Survey, 69% said they had assembled supplies for an emergency; among those with supplies, 88% said the supplies would last more than 3 days.
The 2024 FEMA survey reported that 78% had read, seen, or heard at least some preparedness information during the previous year. Specific information topics reached different shares: 42% received information about emergency supply kits, 34% about emergency plans, 32% about alerts and warnings, and 27% about evacuation plans or routes. Information about protecting a home reached 26%, saving money for disasters 25%, protecting important documents 24%, and staying in contact with family 23%.
Smaller shares reported receiving information about practicing emergency drills and property insurance, each at 21%. Nineteen percent received information about getting involved in the community, and 18% received information about helping neighbors during a disaster. In the same FEMA 2024 survey, 83% reported taking at least three preparedness actions, up from 57% in 2023.
The FEMA 2024 National Preparedness Report recorded several specific actions. Forty-eight percent assembled or updated emergency supplies, 32% saved for a rainy day, 30% documented or insured private property, and 14% got involved with their communities to take preparedness actions. Two-thirds of respondents expressed high concern about lacking necessities in a disaster. A separate FEMA survey discussed in 2024 NPR found that 77% of urban respondents said they likely would be impacted by disasters in the coming year.
The practical implication of these measurements is specific: outage readiness should cover at least three days of essential needs, while accounting for the possibility that a disaster-related outage lasts longer. The statistics identify exposure and reported preparation levels; they do not establish that any particular kit, generator, device, or plan will prevent harm.