Combustion vs Life-Cycle Emissions: Why Two Numbers Exist
A combustion number counts only the CO2 released when a fuel burns. A life-cycle number adds the other steps, such as extraction or farming, processing, transport and distribution, so it answers a wider question. Both are correct, and they are not interchangeable: one measures a moment, the other a chain.
Gasoline shows the gap. EPA's 2025 factor for burning it is 70.22 kg of CO2 per million Btu, while EPA's life-cycle baseline for petroleum gasoline is 98.2 kg of CO2e per million Btu. Ethanol shows something subtler: its burning number is close to gasoline's, but its life-cycle number is not. The sections below explain what each measure includes, how the two diverge for gasoline, ethanol and grid electricity, and how to read any number you meet. The sourced values for all three sit in the Comparator for gasoline, ethanol and grid electricity.
Two boundaries, two questions
EPA's guidance on direct emissions from stationary combustion covers what burning a fuel releases. It states that it does not address the upstream emissions from extraction, production and transport of the fuel, which an organization reports under a separate category. That is the combustion boundary: the stack or tailpipe.
The life-cycle boundary is wider. U.S. law, as quoted in an EPA fact sheet, defines lifecycle greenhouse gas emissions as the aggregate quantity of emissions, including direct emissions and significant indirect ones such as land use change, across the full fuel lifecycle. That runs from feedstock generation or extraction through distribution, delivery and use of the finished fuel, with each gas adjusted for its global warming potential. The unit shifts from CO2 to CO2e, "carbon dioxide equivalent," which converts other gases such as methane into CO2 terms.
| Stage | In a combustion number | In a life-cycle number |
|---|---|---|
| Extraction, mining or farming | No | Yes |
| Processing or refining | No | Yes |
| Transport and distribution | No | Yes |
| Land use change (for crop-based fuels) | No | Yes, in EPA's analysis |
| Burning the fuel | Yes | Yes |
| Gases counted | CO2 (other gases listed separately) | CO2, CH4 and N2O, as CO2e |
How researchers build a life-cycle number is the subject of how a life-cycle assessment of a fuel is built. Combustion factors for several fuels, on one energy basis, are laid out in CO2 emissions per unit of energy.
Gasoline and ethanol
| Fuel | Combustion CO2 (EPA 2025 factor, per unit of HHV energy) | Life-cycle (EPA, kg CO2e per million Btu) |
|---|---|---|
| Gasoline | 70.22 kg per million Btu (66.56 g/MJ) | 98.2, 2005 average petroleum baseline |
| Ethanol (E100) | 68.44 kg per million Btu (64.87 g/MJ), biogenic | 77.2, corn starch, natural-gas dry mill, 2022 average case |
EPA's life-cycle table splits the gasoline baseline into two stages: 19.2 for crude oil production, transport and refining, and 79.0 for distribution and use. Those two add to 98.2. Do not subtract the 70.22 combustion factor from 98.2 to find an "upstream share." The numbers come from separate EPA datasets with different base years, the life-cycle stage includes distribution and tailpipe CO2, CH4 and N2O for an average vehicle fleet, and the life-cycle table does not state a heating-value basis.
Ethanol is where the two measures part ways. On combustion alone, ethanol's factor is about 2.5% below gasoline's (68.44 ÷ 70.22 is about 0.975). EPA lists it as biogenic CO2, meaning carbon that plants took from the air. On EPA's life-cycle table, the corn ethanol case is 21% below the gasoline baseline (1 − 77.2 ÷ 98.2 is about 0.21). The reason is in the stage columns. For ethanol, distribution and use is 2.1, not 79.0. EPA's method treats carbon in a fuel from renewable biomass as carbon that came from the atmosphere. In a 2024 EPA determination letter, the agency says the uptake of that carbon by the biomass and the CO2 from burning it cancel each other out, so both are left out of the results. The same letter says land use change and other significant indirect emissions are still counted, and that EPA does not simply assume biofuels are carbon neutral. The burden moves to earlier steps: 16.5 for agriculture, 27.8 for land use change, 3.0 for feedstock transport and 27.9 for fuel production.
The plant matters too. In the same EPA table, corn ethanol from a coal-fired wet mill comes to 117.0, above the gasoline baseline, while a biomass-fired dry mill comes to 60.6. So "ethanol" has no single life-cycle number. The question of how biogenic carbon gets counted belongs to what carbon neutral fuel claims mean.
Electricity: zero at the outlet is not zero overall
An appliance burns nothing at the outlet, so it releases no CO2 itself. Many of the power plants that supplied it do release CO2. EIA's national profile for 2024, the latest on its page as of November 13, 2026, gives a CO2 factor of 785 lb per megawatt-hour for all U.S. electricity generation, which is 0.785 lb per kWh (about 356 g). The Comparator records that as a combustion-stage figure for the national average mix. It has no sourced life-cycle figure for grid electricity.
For individual technologies, the IPCC publishes both measures. Its Fifth Assessment Report lists medians in grams of CO2e per kWh of electricity, with direct emissions at the plant shown beside the life-cycle total.
| Technology | Direct emissions | Life-cycle emissions |
|---|---|---|
| Coal, pulverized | 760 | 820 |
| Gas, combined cycle | 370 | 490 |
| Utility-scale solar PV | 0 | 48 |
| Wind, onshore | 0 | 11 |
Wind and solar have no direct emissions, yet their life-cycle values are above zero because of equipment manufacture and supply chains. Coal's life-cycle value (740 to 910 across studies) is higher than its direct value because the IPCC adds 9.6 g per kWh for infrastructure and supply chain and 47 g for methane. As a check, 760 + 9.6 + 47 is about 817, close to the 820 median, though medians do not add exactly. How a grid's mix adds up is covered in carbon intensity of electricity.
Reading a number you meet
Before using any emissions figure, ask four questions:
- Where does the boundary start and stop: the stack, or the whole chain?
- What is counted, CO2 alone or CO2e?
- What is the unit: per million Btu, per MJ, per kWh of electricity, per kg?
- Which year, plant type and region does it describe?
Two figures can both be right and still not be comparable if any answer differs.
Common mistakes
- Treating the two numbers as rivals. They answer different questions, and neither replaces the other.
- Subtracting one from the other. A combustion factor and a life-cycle value from different datasets do not subtract into an upstream share.
- Calling electricity emission-free because the outlet emits nothing. Many power plants do emit, and even wind and solar show life-cycle values above zero.
- Assuming a fuel has one life-cycle value. EPA's corn ethanol cases range from 60.6 to 117.0 depending on the plant.
- Reading CO2e as CO2. CO2e adds methane and nitrous oxide, converted with global warming potentials.
Frequently asked questions
What is the difference between combustion and life-cycle emissions?
Combustion emissions are what leaves the tailpipe or stack when the fuel burns. Life-cycle emissions add the steps before that, such as extraction or farming, processing and transport, and sometimes use and land change. The unit moves from CO2 to CO2e because other greenhouse gases are included.
Why is ethanol's life-cycle number lower than gasoline's if it burns about the same?
EPA's life-cycle method treats the CO2 from burning ethanol as offset by CO2 the crops absorbed, so the use stage is small, and the farming, land use and production steps carry the total. For a corn natural-gas dry mill, EPA's table gives 77.2 against 98.2 for gasoline, about 21% lower.
Is electricity from the grid a zero-emission energy source?
It depends on the boundary. At the outlet, nothing burns. At the power plant, EIA's 2024 national factor is 785 lb of CO2 per MWh. Over the full life cycle, the IPCC lists medians from 11 g CO2e per kWh for onshore wind to 820 for pulverized coal.
Which of the two numbers should a reader use?
Use the one that matches your question. For emissions from burning a fuel in your own equipment, use a combustion factor. For the total effect of choosing one fuel over another, use a life-cycle value from a study whose boundary and year you have checked.
Sources
- GHG Emission Factors Hub: U.S. Environmental Protection Agency. Accessed November 13, 2026.
- Greenhouse Gas Inventory Guidance: Direct Emissions from Stationary Combustion Sources: U.S. EPA Center for Corporate Climate Leadership. Accessed November 13, 2026.
- Lifecycle Greenhouse Gas Emissions for Select Pathways (kg CO2e per mmBtu): U.S. EPA Office of Transportation and Air Quality. Accessed November 13, 2026.
- EPA Lifecycle Analysis of Greenhouse Gas Emissions from Renewable Fuels, EPA-420-F-10-006: U.S. Environmental Protection Agency. Accessed November 13, 2026.
- Lifecycle Analysis of Greenhouse Gas Emissions under the Renewable Fuel Standard: U.S. Environmental Protection Agency. Accessed November 13, 2026.
- Chevron El Segundo Soy FCC D5 Fuel Pathway Determination under the RFS Program, September 11, 2024: U.S. Environmental Protection Agency. Accessed November 13, 2026.
- United States Electricity Profile 2024: U.S. Energy Information Administration. Accessed November 13, 2026.
- Annex III: Technology-specific cost and performance parameters, Climate Change 2014: Mitigation of Climate Change: Intergovernmental Panel on Climate Change. Accessed November 13, 2026.
Spotted something wrong or out of date? Tell us through the corrections page. We answer within 10 business days.