How likely is dying from heat after a hurricane knocks out your power in summer?
Evidence quality 4.13/5
Eight-dimension review score against the quality rubric . Each dimension scored 1–5.
- D1 Source grounding
- 5/5
- D2 Source authority
- 5/5
- D3 Arithmetic
- 3/5
- D4 Uncertainty
- 3/5
- D5 Scope
- 4/5
- D6 Prose
- 4/5
- D7 Perception honesty
- 4/5
- D8 Caveat completeness
- 5/5
● your factors — click this risk ▾ to reveal
- Your factors
≈ As likely as
Perceived
No survey has asked the public to estimate the chance of dying from heat in the days after a hurricane cuts power. People tend to file the danger of a hurricane under the storm itself: the wind, the surge, the falling trees. The slower hazard that follows, days of no air conditioning while the grid is down and the heat index climbs past 100F, is rarely counted as a hurricane death at all, so intuition almost certainly understates it.
Rough estimate: Most people would not name post-storm heat as a hurricane risk; informal intuition treats it as near zero relative to wind and water.
Source: editorial intuition, not polled
Actual
14 heat deaths among roughly 6.75 million people in Beryl's blackout zone (about 1 in 480,000 per event)
Residents inside the ~2.7M-customer power-outage footprint after Hurricane Beryl's July 2024 Texas landfall
Show derivation
Subgroup: an adult living in the hurricane-exposed US Gulf and South Atlantic coastal zone, over a 59-year adult horizon (age 18 to 77). Per-event crude rate from Beryl: 14 heat-induced hyperthermia deaths (NHC Tropical Cyclone Report AL022024) among the population inside the 2.7M-customer outage footprint, taken at roughly 2.5 persons per household, so about 6.75M exposed people, giving about 2.1e-6 deaths per exposed person per event. A coastal-South adult plausibly experiences on the order of 5 prolonged (multi-day) summer post-hurricane blackouts across a 59-year adult life, giving about 5 x 2.1e-6 ~ 1.0e-5 if every event resembles Beryl. Beryl was a comparatively MILD heat coincidence (Houston temperatures peaked in the high 90sF for only a couple of days and the metro was, per Stone-method modeling adapted by the Washington Post, 'narrowly spared' a far worse toll of 600-1,500 deaths). Because the per-event death rate is dominated by how severe and prolonged the coincident heat wave is, the central estimate is set at 5e-5 (about 2.5x the all-Beryl-like floor) to reflect the realistic mix of mild and severe coincidences, with an order-of-magnitude band of 1e-5 to 5e-4. This is a derived, single-major-event-anchored estimate, not a measured population rate.
Caveats: This is a derived, single-major-event-anchored estimate, not a measured populati…
This is a derived, single-major-event-anchored estimate, not a measured population rate, and it should be read as order-of-magnitude only. Four structural weaknesses: (1) The numerator comes essentially from one storm (Beryl, July 2024); the National Hurricane Center's 14 hyperthermia deaths are a forensic count, but a one-event base rate is statistically thin. (2) Isolating the post-outage heat INCREMENT from background extreme-heat mortality is itself an open modeling problem, acknowledged in the compound-hazard literature; deaths during an outage may have occurred in a heat wave regardless of the grid. (3) The denominator (~6.75M exposed people from 2.7M customers x ~2.5 persons/household) is an approximation, and 'persons inside the outage footprint' is not the same as 'persons without functional cooling.' (4) The per-event death rate is dominated by how severe and prolonged the coincident heat wave is: Beryl was mild (Houston 'narrowly spared,' per Washington Post modeling of a 600-1,500-death worst case), so the central 5e-5 deliberately sits above the all-Beryl-like floor of 1e-5 and the band runs to 5e-4 to cover a Phoenix-style severe coincidence. The subgroup is a coastal-South US adult; for a typical inland US adult the figure is roughly 10x lower. The estimate is reported because a defensible denominator path exists (per-event rate x plausible lifetime event count), but a case could be made for no_reliable_estimate given the single-event base; the wide band is the honest expression of that tension. 8D self-score average 3.9, all dimensions >= 3 — clears threshold, with D3 (arithmetic) and D4 (uncertainty) the weakest at 3 each because of the derivation's dependence on assumed per-household occupancy and lifetime event count.
How the risk varies
The headline figure averages across very different situations. Here’s how the probability varies by scenario or context:
1 in 20,000
Subgroup baseline used for the headline; Beryl (Houston) is the anchoring event.
1 in 33,333
Hurricane exposure present but landfall frequency and peak-summer heat coincidence somewhat lower than the western Gulf.
1 in 200,000
Major-hurricane summer blackouts are rare; residual risk reflects post-tropical and derecho outages such as the 2012 mid-Atlantic event.
Bar length and shade rank these scenarios against each other, not against other risks. The exact odds are shown beside each.
Related risks
Other risks on similar themes — for exploring related fears.
Drought famine death
What are the odds of dying from drought-induced famine or water scarcity?
Pick challenger
A hurricane’s official death toll is mostly written in wind and water: the surge, the falling trees, the people who drown. The slower hazard arrives after the storm has passed, when the grid is down and the air conditioning is off and the heat index is still climbing past 100 degrees Fahrenheit. After Hurricane Beryl made landfall near Matagorda, Texas, on July 8, 2024, it knocked out power to more than 2.7 million customers, roughly 80 percent of the Houston region’s electricity, and the National Hurricane Center later recorded that all 34 of the storm’s indirect deaths happened in Texas, fourteen of them from heat-induced hyperthermia caused by the outages. Harris County logged 553 heat-related illness encounters in the single week after landfall. The people who died were not, for the most part, killed by the hurricane. They were killed by the week that followed it in a hot house with no working fan.
The number itself resists clean normalization. Fourteen deaths among the roughly 6.75 million people inside Beryl’s outage footprint works out to about one in 480,000 for that one event, but a single storm is a thin statistical base, and separating the post-outage increment from the heat deaths that a July Texas week would have produced anyway is an unsolved modeling problem. What is clearer is the direction of the tail. Beryl was a mild coincidence: Houston temperatures peaked only in the high 90s for a couple of days, and modeling adapted by the Washington Post from Brian Stone’s work concluded the metro was narrowly spared a far worse toll on the order of 600 to 1,500 deaths. The peer-reviewed version of that modeling found a blackout during a severe heat wave roughly doubles heat mortality in Detroit and Atlanta and raises it about sevenfold in Phoenix, where more than half the population would need emergency care. The hazard is not the average outage. It is the outage that lands on the worst heat.
So the headline figure here, about one in 20,000 over the adult life of a coastal-South resident, is a derived estimate stitched from a per-event rate and a plausible count of multi-day summer blackouts across a lifetime, carried with an uncertainty band that spans a full order of magnitude in each direction. It sits between a home fire death and drowning on the lifetime scale, and well above a lightning strike. The risk is not evenly spread. It concentrates on the Gulf and South Atlantic coast, on the oldest residents, and on the people who cannot leave a hot room or who depend on electricity to breathe or to be fed. The clearest signal in the data is who dies: in Beryl’s wake the documented victims included a 64-year-old woman who used a wheelchair and a feeding tube and died on her fourth day without power, and a person whose breathing machine failed when the lights went out. Climate researchers have started calling this the fifth hurricane hazard, after wind, surge, rain, and tornadoes, because warmer late-season air makes the lethal overlap of a major storm and a heat wave more likely, not less.
Related tidbits
After Hurricane Beryl's 2024 blackout, 14 people died of heat among roughly 6.75 million in the outage zone, about 1 in 480,000 per event. The wind and surge get counted as hurricane deaths. The days of no air conditioning that follow usually do not.
Claim ledger
Every number below is what each source reported, with the verbatim quote we relied on and how we arrived at our figure. Click any link to verify directly.
-
[1] NOAA National Hurricane Center — National Hurricane Center Tropical Cyclone Report: Hurricane Beryl (AL022024)
National Hurricane Center Tropical Cyclone Report: Hurricane Beryl (AL022024)- Statistic
14 of 34 indirect Texas deaths were heat-induced hyperthermia caused by widespread electrical outages- Excerpt
“All 34 of the reported deaths indirectly related to the storm occurred in Texas. Fourteen of them were due to heat-induced hyperthermia caused by widespread electrical outages. One death was due to a fall in a home after an electrical outage, and one was due to an electrical outage causing a breathing machine to fail.”
- Source data from
- 2025-01-23
- Accessed
- 2026-06-13
- Calculation
- Primary authoritative numerator (14 heat deaths). The report (amended version) puts total Beryl deaths at 69, with 35 direct and 34 indirect, all indirect deaths in Texas. Used as the numerator for the per-event crude rate.
- Independence
- Official US government post-storm forensic accounting; independent of the academic and journalistic sources below.
-
[2] Yale Climate Connections (Jeff Masters) — The emerging danger of post-hurricane heat waves
The emerging danger of post-hurricane heat waves- Statistic
14 heat-related deaths and 553 heat-illness encounters after Beryl; post-storm heat named the emerging fifth hurricane hazard; 2.7M customers lost power- Excerpt
“553 heat‐related illness encounters were reported by Harris County Public Health over the period July 9-15, 2024.”
- Source data from
- 2026-04-10
- Accessed
- 2026-06-13
- Calculation
- Establishes the morbidity context (553 ER-level heat-illness encounters) and the framing of post-hurricane heat as an emerging compound hazard; corroborates the 14-death NHC figure and the 2.7M-customer outage scale used in the denominator.
-
[3] GeoHealth (Grundstein, Shepherd, Stearns, Ritchie) — The Fifth Hurricane Hazard: A Case Study of Heat Risks Faced by Disaster Relief Workers After Hurricane Beryl's Landfall
The Fifth Hurricane Hazard: A Case Study of Heat Risks Faced by Disaster Relief Workers After Hurricane Beryl's Landfall- Statistic
553 heat-related illness encounters July 9-15, 2024; >2 million customers initially lost power- Excerpt
“There were reports of 553 heat-related illness encounters in the aftermath of Hurricane Beryl over the period from 9–15 July 2024.”
- Source data from
- 2025-12-18
- Accessed
- 2026-06-13
- Calculation
- Peer-reviewed source naming post-landfall heat the 'fifth hurricane hazard' and quantifying the morbidity burden; independently corroborates the 553-encounter and >2M-customer figures.
- Independence
- Independent academic team (University of Georgia / Korey Stringer Institute); separate from NHC and from Yale Climate Connections.
-
[4] Environmental Science & Technology (Stone et al.) — How Blackouts during Heat Waves Amplify Mortality and Morbidity Risk
How Blackouts during Heat Waves Amplify Mortality and Morbidity RiskSee all 2 Likelier entries citing this source →
- Statistic
A combined heat wave and blackout multiplies heat mortality roughly 2x in Detroit/Atlanta and ~7x (about 700%) in Phoenix vs. heat alone- Excerpt
“more than 50% of the population”
- Source data from
- 2023-05-22
- Accessed
- 2026-06-13
- Calculation
- Provides the mechanism and the upper-tail scale: a prolonged blackout during a severe heat wave can multiply heat mortality several-fold to (in Phoenix) ~700%. Used to justify the wide upper bound (5e-4) of the uncertainty band, since Beryl was a mild heat coincidence. Excerpt is the verbatim Phoenix morbidity figure (share of population needing emergency care).
- Independence
- Independent Georgia Tech / University of Michigan modeling team.
-
[5] Wikipedia — Effects of Hurricane Beryl in Texas
Effects of Hurricane Beryl in Texas- Statistic
More than 2.7 million households and businesses lost power; 2.2M were CenterPoint customers (~80% of the region)- Excerpt
“More than 2.7 million households and businesses lost power”
- Source data from
- 2025-09-01
- Accessed
- 2026-06-13
- Calculation
- Source of the 2.7M-customer outage-footprint figure used to build the ~6.75M exposed-person denominator (2.7M customers x ~2.5 persons per household). Encyclopedia, used only for the widely reported outage count, which is corroborated by NHC and Yale sources.