Portable engines span small nonroad spark-ignition equipment, handheld tools, and generators. The available evidence combines EPA modeled emission factors and regulatory estimates with U.S. Consumer Product Safety Commission (CPSC) reported fatal carbon-monoxide incidents from 2011–2021, so the figures describe benchmarks and reported outcomes rather than a single market-size measure.
Key Portable Engine Statistics
The most broadly useful figures are:
- In the United States, EPA estimated that Phase 1 standards for small spark-ignition engines would reduce hydrocarbon plus nitrogen-oxide emissions by approximately 33% versus unregulated engines in the 1995 rule estimate.
- In the United States, EPA estimated that Phase 2 nonhandheld standards would reduce HC+NOx emissions by approximately 60% beyond Phase 1 levels in the 1999 rule estimate.
- EPA projected that Phase 2 handheld standards would reduce U.S. handheld-engine HC+NOx emissions by 70% beyond the Phase 1 reduction.
- EPA projected 500,000 tons per year of U.S. handheld-engine exhaust HC+NOx reductions by 2027.
- EPA estimated that Phase 2 compliance would lower fuel consumption by approximately 30% for U.S. handheld engines.
- EPA’s Phase 1 small-engine rule covered engines at or below 19 kW, or approximately 25 horsepower, in the United States.
- EPA’s Class I baseline gas two-stroke factor was 207.92 g/hp-hr of HC for nonhandheld engines under 225 cc.
- EPA’s Class I Phase 3 overhead-valve four-stroke factor was 0.04 g/hp-hr of particulate matter for nonhandheld engines under 225 cc.
- EPA’s Class II Phase 3 overhead-valve four-stroke factor was 1.01 g/hp-hr of NOx for nonhandheld engines over 225 cc.
- EPA’s Class IV Phase 2 catalyst gas two-stroke factor was 26.87 g/hp-hr of HC for handheld engines over 20 and under 50 cc.
- A peer-reviewed study tested 23 new and used handheld two-stroke engines from model years 1981–2003.
- In that study, 10% ethanol reformulated gasoline reduced HC by 15% on average on the tested handheld two-stroke engines.
- CPSC recorded 710 fatal non-fire CO incidents involving engine-driven tools in the United States from 2011–2021.
- Those reported U.S. engine-driven-tool incidents involved 900 deaths from 2011–2021.
- Portable generators accounted for 581 of the 710 reported incidents and 758 of the 900 reported deaths from 2011–2021.
- CPSC reported 39 engine-driven-tool CO incidents and 52 deaths in 2021 as of May 10, 2022.
- From 2011–2021, people aged 25 and older represented 88% of engine-driven-tool CO decedents in the United States.
- From 2011–2021, 743 of 900 reported deaths, or 83%, occurred at residential locations.
- From 2011–2021, cold months accounted for 391 of 900 reported deaths, or 43%.
- Among deaths with known alarm status, 194 deaths, or 83%, were in incidents with no alarm.
Contents
- Key Portable Engine Statistics
- Portable Engine Emissions and Regulatory Benchmarks
- Emission Factors by Engine Class and Technology
- Fuel Use and Brake-Specific Performance Data
- Portable Generator CO Deaths and Annual Trends
- Where Portable-Engine CO Fatalities Occur
- Seasonal Patterns and CO-Alarm Data
Portable Engine Emissions and Regulatory Benchmarks
EPA’s small spark-ignition rules provide modeled policy benchmarks, not measurements of every portable engine in service. Phase 1 covered U.S. engines at or below 19 kW, equivalent to approximately 25 horsepower.
- EPA estimated a 33% reduction in HC+NOx versus unregulated engines under the 1995 Phase 1 estimate for U.S. small SI engines.
- EPA estimated a 60% reduction beyond Phase 1 under the 1999 Phase 2 estimate for U.S. nonhandheld small SI engines, including generators.
- EPA projected a 70% reduction beyond the Phase 1 reduction for U.S. handheld small SI engines under Phase 2.
- EPA projected 500,000 tons per year of exhaust HC+NOx reductions by 2027 from the Phase 2 handheld standards in the United States.
- EPA estimated that Phase 2 compliance would reduce fuel consumption by approximately 30% for U.S. handheld engines.
These figures are estimates or forecasts tied to new-engine compliance, implementation, and fleet-turnover assumptions. They should not be read as direct market-share, fleet-use, or annual-emissions statistics.
Sources: EPA Small Business Advocacy Review Panel report and EPA Phase 2 handheld-engine regulatory announcement.
Portable Engine Emission Factors by Class and Technology
EPA’s NONROAD2008a factors show how modeled steady-state emissions vary across engine classes and technologies. The units are brake-specific grams per horsepower-hour, so they are not annual totals and should not be compared as identical real-world duty cycles.
| EPA class and configuration | HC (g/hp-hr) | CO (g/hp-hr) | NOx or PM (g/hp-hr) |
|---|---|---|---|
| Class I baseline gas two-stroke, under 225 cc | 207.92 HC | 485.81 CO | — |
| Class I Phase 3 overhead-valve four-stroke, under 225 cc | 3.80 HC | — | 0.04 PM |
| Class I Phase 3 side-valve four-stroke, under 225 cc | — | 238.45 CO | 1.04 NOx |
| Class II baseline gas two-stroke, over 225 cc | 207.92 HC | 485.81 CO | — |
| Class II Phase 3 overhead-valve four-stroke, over 225 cc | 3.17 HC | 321.94 CO | 1.01 NOx; 0.06 PM |
The same EPA report assigns Class II engines over 225 cc to a 6–25 horsepower range, but the class is broader than portable generators. Class I and Class II entries are model factors rather than field tests of named products.
Handheld classes provide related portable-engine benchmarks:
- Class IV baseline gas two-stroke HC was 261.00 g/hp-hr, compared with 26.87 g/hp-hr for the Class IV Phase 2 catalyst gas two-stroke factor.
- Class IV baseline gas two-stroke CO was 718.87 g/hp-hr, compared with 141.69 g/hp-hr for the Phase 2 catalyst factor.
- Class V baseline gas two-stroke CO was 519.02 g/hp-hr, compared with 283.37 g/hp-hr for the Phase 2 factor.
The Class V Phase 2 HC factor was 47.98 g/hp-hr for handheld engines over 50 cc. These are steady-state model factors; the handheld categories are related portable-engine benchmarks, not a complete generator emissions inventory.
Source: EPA Exhaust Emission Factors for Nonroad Engine Modeling—Spark-Ignition.
Fuel Use and Brake-Specific Performance Data
Brake-specific fuel consumption (BSFC) expresses modeled fuel use per horsepower-hour. It is useful for comparing the listed engine configurations, but it does not establish the fuel use of a particular generator under a particular load.
| Configuration | BSFC (lb/hp-hr) | Period or factor |
|---|---|---|
| Class I baseline gas two-stroke | 0.870 | NONROAD2008a baseline factor |
| Class I Phase 3 overhead-valve four-stroke | 0.781 | NONROAD2008a Phase 3 factor |
| Class II Phase 3 overhead-valve four-stroke | 0.735 | NONROAD2008a Phase 3 factor |
| Class IV baseline gas two-stroke | 1.365 | NONROAD2008a baseline factor |
| Class IV Phase 2 catalyst gas two-stroke | 0.822 | NONROAD2008a Phase 2 factor |
| Class V baseline gas two-stroke | 0.921 | NONROAD2008a baseline factor |
| Class V Phase 2 | 0.608 | NONROAD2008a Phase 2 factor |
The EPA model assigns Class IV handheld engines over 20 and under 50 cc to 1–3 horsepower, and Class V handheld engines over 50 cc to 3–6 horsepower. Those are assigned modeling ranges, not measured distributions of engines sold or operated.
A peer-reviewed Atmospheric Environment study tested 23 new and used handheld two-stroke engines from model years 1981–2003. In the laboratory trial, 10% ethanol reformulated gasoline reduced HC by 15% on average and CO by 29% on average on the studied engines.
The study’s sample was small and specific to handheld two-stroke engines. Its results should therefore be treated as trial findings, not fleet-wide estimates for portable engines or generators.
Sources: EPA emission-factor report and Emissions profile from new and in-use handheld, 2-stroke engines.
Portable Generator CO Deaths and Annual Trends
CPSC’s U.S. data cover reported fatal non-fire carbon-monoxide incidents involving engine-driven tools from 2011–2021. The agency warns that incidents may be underestimated and that recent counts can change as delayed reports arrive.
| Measure | Reported count | Scope |
|---|---|---|
| Fatal engine-driven-tool incidents | 710 | United States, 2011–2021 |
| Deaths in those incidents | 900 | United States, 2011–2021 |
| Portable-generator incidents | 581 | United States, 2011–2021 |
| Portable-generator deaths | 758 | United States, 2011–2021 |
| Engine-driven-tool incidents in 2021 | 39 | As of May 10, 2022 |
| Engine-driven-tool deaths in 2021 | 52 | As of May 10, 2022 |
CPSC identified 2017–2019 as the most recent substantially complete years in its report. Across that period, the annual average was 78 engine-driven-tool incidents and 95 deaths, while the generator category averaged 68 incidents and 85 deaths per year.
Annual generator-related deaths were 96 in 2011, 49 in 2012, and 59 in 2013. They were 77 in 2015, 75 in 2016, and 83 in 2017.
These are reported counts, not exposure-adjusted rates per generator, household, user, operating hour, or fuel type. The generator category also reflects CPSC’s product classification and is not identical to every possible portable-engine application.
Where Portable-Engine CO Fatalities Occur
The CPSC fatality distribution describes who was represented among reported decedents and where incidents occurred. It does not provide age-adjusted or use-adjusted risk.
- People aged 25 and older represented 88% of all engine-driven-tool CO decedents from 2011–2021 in the United States.
- People aged 25 and older represented 87% of generator-related CO decedents, based on 814 decedents in the CPSC report.
- People aged 45 and older represented 83% of other engine-driven-tool decedents, based on 86 decedents.
- Victims over age 44 represented 60% of EDT-related CO fatalities, compared with 41% of the U.S. population in the report’s 2014–2018 comparison.
- Victims under age 25 represented 12% of EDT-related CO fatalities, compared with 32% of the U.S. population in that comparison.
Residential locations dominated the reported fatality data. 743 of 900 deaths, or 83%, occurred at residential locations, and fixed-structure homes accounted for 627 deaths, or 70% of the total.
For generator deaths, fixed-structure homes accounted for 581 deaths, or 71%, while temporary shelters accounted for 59 deaths, or 7%. CPSC defines temporary shelters to include trailers, RVs, cabins, tents, and campers; built-in generators may be excluded.
Source: CPSC fatal incident report.
Seasonal Patterns and CO-Alarm Data
CPSC’s seasonal categories show when reported fatal incidents occurred, while alarm figures are limited by missing information. “Unknown or unreported” alarm status must not be treated as proof that no alarm was present.
- Cold months, November–February, accounted for 321 of 710 fatal incidents, or 45%, from 2011–2021.
- Cold months accounted for 391 of 900 deaths, or 43%, during the same U.S. period.
- Transitional months, March–April and September–October, accounted for 226 incidents, or 32%.
CO-alarm reporting was incomplete in most cases. Status was unknown or unreported for 549 of 710 fatal incidents, or 77%, and for 664 of 900 deaths, or 74%.
Among the 159 fatal incidents with known alarm status, an alarm was present in 28 incidents, or 18%. Among deaths with known status, 194 deaths, or 83%, were in incidents with no alarm, and a CO alarm was known to have sounded in 11 fatal incidents involving 13 deaths.
The known-status subset is much smaller than the full incident and death totals. CPSC also notes that alarm response and victim location can affect outcomes, so these figures describe reported circumstances rather than a complete measure of alarm effectiveness.
Source: CPSC fatal incident report.