63 Generator Safety Statistics: Fatalities, Locations, Alarms and CO Outcomes

Generator safety statistics show a persistent carbon-monoxide hazard during backup-power use. The figures below cover reported U.S. fatal incidents, broader unintentional CO surveillance, and hurricane-related hospital cases; the populations, periods and definitions differ by source.

Key Generator Safety Statistics

The most broadly useful figures are:

  • 814 deaths in 624 reported incidents associated with engine-driven generators and other tools occurred in the United States during 2011–2021.
  • 758 deaths in 581 incidents involved portable generators in the U.S. CPSC dataset during 2011–2021.
  • 85 deaths per year from 68 incidents per year was the reported generator average in the United States during 2017–2019.
  • 327 deaths, or 40% of generator-related deaths, involved U.S. victims ages 45–64 during 2011–2021.
  • 624 male deaths, or 77% of reported generator deaths, were recorded in the United States during 2011–2021.
  • 355 deaths, or 44% of reported generator deaths, occurred in cold months from November through February during 2011–2021.
  • 581 deaths, or 71% of reported generator deaths, occurred in fixed-structure homes in the United States during 2011–2021.
  • 111 of 464 deaths with known victim position occurred in the same room or space as the generator in U.S. fixed-structure homes during 2011–2021.
  • 186 deaths, or 84% of known alarm-status deaths, involved no CO alarm in the United States during 2011–2021.
  • 18 deaths, or 8% of known alarm-status deaths, involved an alarm that failed because batteries were removed, incorrectly inserted or drained during 2011–2021.
  • 216 deaths in 140 home incidents involved no reported ventilation in U.S. fixed-structure homes during 2011–2021.
  • 358 of 422 deaths with known home size, or 87%, occurred in homes under 2,000 square feet in the United States during 2011–2021.
  • 157 deaths in 114 incidents were linked to weather or power-distribution outages in the United States during 2011–2021.
  • 58 deaths in 37 incidents were linked to hurricanes or tropical storms in the United States during 2011–2021.
  • 56.5 emergency-department visits per million people per year was the CDC’s estimated all-source unintentional CO rate in the United States during 2005–2018.
  • Six deaths and 167 nonfatal patients in 51 incidents were identified at Florida hospitals after hurricanes during August 13–October 15, 2004.

Contents

The Consumer Product Safety Commission recorded 624 generator-associated fatal non-fire CO incidents and 814 deaths in the United States during 2011–2021. These are reported cases, a minimum count rather than a complete census, and CPSC says reporting for 2020 and 2021 was incomplete.

Portable generators accounted for 581 incidents and 758 deaths in that dataset. The portable-generator category excludes fixed-location generators and some products outside CPSC jurisdiction. CPSC, Fatal Incidents Associated with Non-Fire Carbon Monoxide Poisoning from Engine-Driven Generators and Other Engine-Driven Tools, 2011–2021

Year or period Reported incidents Reported deaths
2011 69 96
2012 43 49
2013 46 59
2014 38 49
2015 52 77
2016 57 75
2017–2019 annual average 68 85
2020 77 103
2021 39 52

The annual CPSC counts were 67 incidents and 83 deaths in 2017, 57 and 76 in 2018, and 79 and 95 in 2019. The 2020 and 2021 values may rise as delayed reports are added, so they should not be treated as final totals.

Portable Generator Demographics and Seasonal Data

Age patterns in the 814 reported generator-related deaths describe the fatal cases, not population-level risk. The largest supplied age group was 45–64, with 327 deaths, or 40%; victims ages 25–44 accounted for 244 deaths, or 30%, while victims 65 or older accounted for 137 deaths, or 17%, and those ages 15–24 accounted for 62 deaths, or 8%.

Sex and race categories also describe reported victims rather than exposure-adjusted rates:

  • Female victims accounted for 190 deaths, or 23%, while male victims accounted for 624 deaths, or 77%, in the United States during 2011–2021.
  • White victims represented 458 deaths, or 56%, during 2011–2021.
  • Black or African American victims represented 186 deaths, or 23%, during 2011–2021.
  • Hispanic victims of any race represented 103 deaths, or 13%, during 2011–2021; ethnicity may overlap race categories.

Seasonal shares were highest in the cold-month grouping, defined as November through February. Cold months produced 285 incidents and 355 deaths, or 46% of incidents and 44% of deaths; transitional months—March, April, September and October—produced 198 incidents and 264 deaths, both 32%; warm months from May through August produced 141 incidents and 195 deaths, or 23% and 24%.

These seasonal percentages describe reported fatal incidents, not exposure-adjusted risk. The demographic and seasonal figures come from CPSC’s 2011–2021 fatal-incident report.

Fixed-structure homes accounted for 430 incidents and 581 deaths, or 69% of incidents and 71% of deaths, in the CPSC’s U.S. 2011–2021 data. This category includes houses, mobile homes, apartments and attached structures; it excludes RVs and detached structures.

Location category Incidents Deaths
Fixed-structure homes 430 (69%) 581 (71%)
Detached home structures 52 (8%) 53 (7%)
Temporary shelters 41 (7%) 59 (7%)
Boats or vehicles 43 (7%) 50 (6%)

Within fixed-structure homes, 161 incidents and 226 deaths involved a generator in a non-basement living space. A basement or crawlspace was involved in 87 incidents and 118 deaths, while a garage, enclosed carport or attached barn was involved in 77 incidents and 105 deaths.

Victim position was known for 464 of the 581 home deaths. At least 111 of those deaths, or 24% of deaths with known position, occurred in the same room or space as the generator; the denominator is not all home deaths. CPSC’s report notes that location categories are not interchangeable with generator product categories.

Generator Alarms, Ventilation and Home-Size Statistics

Alarm status was known for 222 of the 814 generator deaths. Among those known-status deaths, 186, or 84%, involved no CO alarm, 34, or 15%, involved an alarm being present, and alarms sounded in 9 deaths, or 4%.

The same CPSC table records 18 deaths, or 8% of known-status deaths, in which the alarm did not work because batteries were removed, incorrectly inserted or drained. These figures describe reported fatal cases and do not measure the effectiveness of alarms across all generator users.

Ventilation was recorded for fixed-structure-home incidents, but it was unknown in 212 incidents. No ventilation was reported in 140 incidents involving 216 deaths, or 37% of all home deaths; some ventilation was attempted in 78 incidents involving 103 deaths, or 18% of all home deaths.

Attempted ventilation is not a measure of adequate ventilation. Generators placed outside but near a window, door or vent were associated with 14 incidents and 22 deaths, although proximity and airflow were not standardized measurements.

Home size was known for 310 of 430 incidents and 422 of 581 deaths. Among known-size incidents, 201, or 65%, occurred in homes under 1,500 square feet; among known-size deaths, 358, or 87%, occurred in homes under 2,000 square feet.

The median home size in known-size fatal cases was approximately 1,250 square feet, not a causal threshold. CPSC’s fatal-incident report supplies the denominators and limitations.

Outages, Storms and Temporary-Use Generator Data

Reported reasons for generator use connect many fatal cases to disrupted electricity, but the categories are not mutually presented as exposure rates. Weather or power-distribution outages were linked to 114 incidents and 157 deaths in the United States during 2011–2021, while electricity shut off by a power company was linked to 100 incidents and 139 deaths.

Reported context Incidents Deaths
Hurricanes or tropical storms 37 58
Ice or snow storms 37 47
Wind storms 6 13
Thunderstorms, rainstorms or flooding 6 8
Temporary locations such as trailers, boats, campers, cabins or campsites 108 Not used

The temporary-location incident count is retained without a death total because the cited table excerpt did not expose that continuation line. Hurricane, snow, wind and rain categories are subsets of generator cases tied to a reported power-outage reason, so they should not be added to the broader outage total without a complete cross-tabulation. CPSC’s report classifies the reason for use from incident reports.

CDC surveillance provides context for CO poisoning beyond generator-specific fatalities. For 2005–2018, estimated annual U.S. rates for unintentional CO poisoning from all sources were 39.5 poison-center exposure calls, 56.5 emergency-department visits, 7.3 hospitalizations and 3.3 deaths per million people per year; the ED estimate covered 17 states and the hospitalization estimate covered 26 states.

Non-fire CO cases comprised 74.0% of ED visits, 60.1% of hospitalizations and 40.9% of deaths in those surveillance datasets. These are all-source figures, not generator-specific rates. CDC, Morbidity and Mortality of Unintentional Carbon Monoxide Poisoning: United States 2005 to 2018 also reported that children ages 0–9 had the highest poison-center exposure rate, 54.1 per million, and ED visit rate, 70.5 per million; adults older than 80 had the highest hospitalization rate, 20.2 per million, and death rate, 9.9 per million.

Among poison-center exposure cases, 45.9% were treated at a health-care facility. Poison-center callers are not a census of all exposures, so this share should not be generalized to every CO event.

A Florida hospital surveillance study after hurricanes in 2004 identified six deaths and 167 nonfatal patients in 51 exposure incidents from August 13 through October 15, 2004. The 10-hospital design likely overrepresented more severe cases.

Among the 167 nonfatal Florida patients, 52.1% were female and the median age was 29. Patients younger than 16 represented 31.1%, while those older than 65 represented 6.6%; the most common symptoms were headache at 80.0%, nausea at 51.5% and dizziness at 50.9%.

COHb was measured for 162 patients, with a mean of 19.8%, a median of 21.1% and a range of 0.2%–45.1%. Treatment outcomes included 48.5% treated and released from the ED without hyperbaric oxygen, 43.7% released after hyperbaric treatment, 7.8% hospitalized and 46.1% receiving hyperbaric oxygen. CDC, Carbon Monoxide Poisoning from Hurricane-Associated Use of Portable Generators — Florida, 2004

A separate 2005 CDC report cited an estimated 15,000–40,000 annual U.S. ED visits and 500 annual deaths from unintentional non-fire CO poisoning, but those were historical all-source estimates, not generator-specific counts. The same report stated that generator-related deaths reported to CPSC rose from 18 in 2001 to 36 in 2003. CDC, Carbon Monoxide Poisonings After Two Major Hurricanes — Alabama and Texas, August–October 2005

Table of contents