59 Electricity Resilience Statistics on Outages, Costs and Grid Readiness

Electricity resilience can be measured through customer outage duration, interruption frequency, severe-weather exposure, household self-sufficiency, economic losses and grid investment. In the United States, these measures range from 2024 customer reliability data to modeled transmission outcomes through 2050, and they describe different populations and time periods.

Key Electricity Resilience Statistics

The most broadly useful figures include:

  • In 2024, U.S. electricity customers experienced nearly 11 hours of interruptions on average.
  • In 2024, major events caused 80% of U.S. hours without electricity.
  • In 2024, major-event interruptions averaged nearly nine hours per U.S. customer.
  • During 2014–2023, major-event interruptions averaged nearly four hours per U.S. customer per year.
  • In 2024, the average U.S. customer experienced 1.5 non-momentary interruptions.
  • In 2024, Maine recorded 1,748.6 minutes of SAIDI per customer-year including major events.
  • In 2024, New York recorded 0.910 SAIFI interruptions per customer including major events.
  • In September 2024, Hurricane Helene left 5.9 million customers without power across 10 U.S. states.
  • In October 2024, Hurricane Milton left 3.4 million Florida customers without power.
  • In 2024, South Carolina customers experienced nearly 53 hours of interruptions on average.
  • In 2024, Hawaii customers experienced 4.4 interruptions on average.
  • In FEMA’s 2024 survey, 63% of U.S. household respondents said they could live at home more than three days without power in warm temperatures.
  • In the same 2024 survey, 59% said they could live at home more than three days without power in cold temperatures.
  • From 2003–2012, weather-related outages cost the U.S. economy an inflation-adjusted $18 billion to $33 billion per year on average.
  • Since 2002, severe weather accounted for 87% of U.S. outages affecting at least 50,000 customers.
  • In the DOE/EOP 2013 baseline, about 70% of transmission lines and power transformers were more than 25 years old.
  • Between June 2011 and February 2013, Recovery Act funds installed 6.2 million smart meters in the United States.
  • Through 2050, DOE’s enhanced-transmission scenario projects $320 billion in present-value power-sector cost savings.

Contents

National Electricity Outage Duration and Frequency

The Energy Information Administration (EIA) reported that U.S. electricity customers experienced an average of nearly 11 hours of interruptions in 2024. Major events, including utility-classified hurricanes and other severe weather, accounted for 80% of those hours, while major-event interruptions averaged nearly nine hours per customer.

EIA’s comparison with the preceding decade provides a useful measure of weather-related volatility: major-event interruptions averaged nearly four hours per customer per year during 2014–2023. The 2024 result is an observed national average, and the treatment of major events follows utility reporting practices.

The national interruption-frequency measure was 1.5 instances per customer in 2024. SAIFI counts non-momentary interruptions, and EIA cautions that state reporting coverage can vary.

Source: EIA, “Hurricanes in 2024 led to the most hours without power in the United States in 10 years”.

State Electricity Resilience Data

EIA’s 2024 any-method state series includes major events. SAIDI measures interruption duration in minutes per customer-year, while SAIFI measures interruptions per customer-year; the figures below should therefore be read as paired duration and frequency indicators, not as a single resilience score.

State SAIDI, 2024 SAIFI, 2024
Maine 1,748.6 minutes 3.636 interruptions
New Hampshire 922.7 minutes 2.021 interruptions
Vermont 696.7 minutes 2.516 interruptions
Pennsylvania 336.1 minutes 1.353 interruptions
New York 219.9 minutes 0.910 interruptions
New Jersey 174.2 minutes 1.119 interruptions

Source: EIA, “Electric Power Annual: Table 11.3 Reliability Metrics Using Any Method of U.S. Distribution System by State, 2024 and 2023”.

The state figures show why geography matters in electricity resilience statistics. Maine’s 2024 duration and frequency were both higher than New York’s, while Vermont recorded fewer minutes than New Hampshire but more interruptions per customer, illustrating that outage length and outage frequency can move differently.

Major weather events can affect millions of customer accounts even when annual averages are reported per customer. EIA recorded several large 2024 events:

  • Hurricane Beryl left 2.6 million Texas customers without power in July 2024.
  • Hurricane Helene left 5.9 million customers without power across 10 states in September 2024; at least 1.2 million of those affected customers were in South Carolina.
  • Hurricane Milton left 3.4 million Florida customers without power in October 2024.

These event totals are not equivalent to unique annual customers, because an event count can include repeat or overlapping service accounts. EIA also reported that South Carolina customers experienced nearly 53 hours of interruptions on average in 2024, the highest state value cited in that analysis.

Hawaii had the highest state interruption frequency cited by EIA, at 4.4 interruptions per customer in 2024. EIA identified weather, volcanic activity and plant outages among contributing factors.

At the other end of the reported thresholds, Arizona, South Dakota, North Dakota and Massachusetts each averaged less than two hours of interruptions in 2024. South Dakota, Maryland, Illinois and Massachusetts each averaged less than one interruption per customer in that year; EIA provided threshold statements rather than exact values for these observations.

Source: EIA, “Hurricanes in 2024 led to the most hours without power in the United States in 10 years”.

Household Electricity Resilience During Extended Outages

The Federal Emergency Management Agency’s 2024 National Household Survey measured perceived household self-sufficiency, not observed outage endurance or generator ownership. Respondents were asked how long they believed they could live at home without power under warm and cold conditions.

Reported duration at home without power Warm temperatures Cold temperatures
More than three days 63% 59%
More than one week 43% 40%
More than two weeks 26% 24%
More than one month 16% 15%
More than three months 11% 9%

Source: FEMA, “2024 National Household Survey on Disaster Preparedness Findings”.

The survey shows a steady decline in reported self-sufficiency as the outage duration lengthens. The warm-versus-cold percentages are perceptions from U.S. household survey respondents, so they should not be interpreted as measured survival rates during a grid interruption.

Economic Costs of Weak Electricity Resilience

The Department of Energy and Executive Office of the President reported an inflation-adjusted average annual cost of $18 billion to $33 billion for U.S. weather-related outages during 2003–2012. The modeled range included lost output and wages, spoilage, delays, inconvenience and grid damage.

The same report gave wider event-year ranges: weather-related outage costs were estimated at $40 billion to $75 billion in 2008, the year of Hurricane Ike, and $27 billion to $52 billion in 2012, the year of Superstorm Sandy. A Congressional Research Service estimate cited by DOE put annual weather-related outage costs at $25 billion to $70 billion in a 2012 estimate.

These estimates are not directly interchangeable because they use different assumptions and methods. Primen’s 2001 estimate for all outage costs was $132 billion to $209 billion in 2012 dollars for U.S. businesses, based on a survey-based extrapolation from 985 industrial and digital-economy firms and excluding residential customers.

The DOE/EOP report counted 679 widespread severe-weather outages from 2003–2012, defining each event as affecting at least 50,000 customers. Severe weather accounted for 58% of observed outages since 2002 and 87% of outages affecting at least 50,000 customers.

Source: DOE/EOP, “Economic Benefits of Increasing Electric Grid Resilience to Weather Outages”.

Grid Infrastructure and Resilience Investment Statistics

The DOE/EOP report described a historical U.S. grid baseline serving more than 144 million end-use customers, with about 5,800 major power plants and more than 450,000 miles of high-voltage transmission lines. About 70% of transmission lines and power transformers were more than 25 years old, and the average age of U.S. power plants was more than 30 years in that cited baseline.

The 2009 Recovery Act allocated $4.5 billion to DOE investments in modern grid technology. Between June 2011 and February 2013, those funds were reported to have deployed 343 advanced grid sensors, upgraded 3,000 distribution circuits with digital technology, installed 6.2 million smart meters and supported 16 energy-storage projects.

These infrastructure figures describe a historical system scale and a historical investment program, not current inventories or current spending. The energy-storage figure is a project count rather than an installed-capacity measure.

Source: DOE/EOP, “Economic Benefits of Increasing Electric Grid Resilience to Weather Outages”.

Transmission Reliability and 2050 Electricity Resilience Forecasts

DOE’s enhanced-transmission scenario projects $320 billion in present-value power-sector cost savings through 2050. The same scenario projects 3,420 million metric tons of cumulative power-sector emissions reductions, $730 billion in climate benefits and $50 billion in health benefits through 2050.

These are modeled scenario results, not observed resilience outcomes or guaranteed savings. The climate and health figures are benefit valuations, and the emissions result is not itself a customer-outage measure.

NERC’s M-16 transmission metric covers AC circuits and transformers operated at 200 kV and above in the North American bulk transmission system. On the metric page, NERC reports the historical availability values shown for 2010–2011 as above 90%; this bulk-transmission element measure should not be presented as the same thing as customer distribution reliability.

Sources: DOE, “Transmission Impact Assessment”; NERC, “M-16 Element Availability Percentage and Unavailability Percentage”.

Table of contents