1. Home
  2. Data
  3. EPA GHG Emission Factors Hub
Last reviewed October 2026
Authored by Jeremiah Say

Founder and Lead Systems Architect of GreenCalculus. Translates GHG Protocol methodology into high-precision JavaScript calculation engines. Architect of the MasterBrain data layer covering 18,995 sourced emission factors, aligned with IPCC AR6 and the GHG Protocol Corporate Standard.

Full profile →

Verified by GreenCalculus Engineering

Automated verification pipeline that audits every page against its underlying calculation code, source documents, and MasterBrain data layer. Traces every figure cell-by-cell to its named source workbook, enforces cell-by-cell provenance attribution on every emission factor, and cross-checks methodology prose against the data layer to catch stated-vs-actual discrepancies before publication.

Governance & verification pipeline →

EPA Emission Factors: the GHG Emission Factors Hub, Table by Table

SourceEPA GHG Emission Factors Hub
Edition2025 (15 Jan 2025)
Tables12
GC ImplementationMasterBrain v2026.204
LicenceUS Government work, public domain

When someone says “the EPA emission factors”, they almost always mean one file: the GHG Emission Factors Hub, a five-page table that the EPA’s Center for Corporate Climate Leadership publishes for companies building a greenhouse gas inventory. It gives the default factors for burning fuel, buying electricity and steam, shipping goods, travelling for business and throwing things away. This page goes through all 12 of its tables, shows the headline values (served live from our database and traceable to the EPA workbook) and points out the three places where the Hub’s own notes change how its numbers should be used.

Quick answer

The EPA’s emission factors for company reporting are in the GHG Emission Factors Hub. The latest edition is 2025, published January 2025 and still current at 2 October 2026. Natural gas is 53.06 kg CO2 per mmBtu. Diesel is 10.21 kg CO2 per gallon. US-average grid electricity is 771.5 lb CO2 per MWh.

Which EPA emission factors do you need?

The EPA publishes several sets of numbers that all get called “EPA emission factors”, and they answer different questions. Picking the wrong one is the first mistake people make, so start here.

The five EPA factor sets and what each one is for
Publication What it gives you Use it for
GHG Emission Factors Hub (this page) Default greenhouse gas factors for fuels, vehicles, electricity, steam, freight, waste and travel A corporate Scope 1, 2 and 3 inventory: the general-purpose default
eGRID Emission rates for every US power plant, state and grid subregion Scope 2 electricity. The Hub’s Table 6 is copied from it
GHGRP, 40 CFR Part 98 The mandatory reporting rule and its factor tables (C-1, C-2) Facility reporting to the EPA. The Hub’s Table 1 is taken from it. See Subpart W for oil and gas
AP-42 Factors for air pollutants from stationary sources, by process Air permits and pollutant inventories, mostly not CO2e reporting
Supply-chain factors (USEEIO) kg CO2e per dollar for 1,016 commodities Spend-based Scope 3 Category 1 and 2 estimates

The EPA describes the Hub’s purpose in one sentence on its landing page:

EPA, GHG Emission Factors Hub (verbatim)

“EPA’s GHG Emission Factors Hub was designed to provide organizations with a regularly updated and easy-to-use set of default emission factors for organizational greenhouse gas reporting.”

The word that matters there is default. If your fuel supplier gives you a measured heat content or carbon content, the Hub itself says to use that instead. The Hub is for when you have nothing better.

What is in the Hub: all 12 tables

The 2025 edition runs to five pages and twelve tables. Most of its numbers are not original EPA research: they are taken from other EPA programmes, other federal agencies and, in a few places, from outside the US government. Knowing where each table comes from tells you how far to trust it and when it will change.

EPA GHG Emission Factors Hub 2025: contents and upstream sources
Table Covers Unit Taken from
1 Stationary combustion: 58 fuels plus five biomass types (CO2, CH4, N2O) per mmBtu, and per short ton, scf or gallon 40 CFR Part 98, Tables C-1, C-2 and AA-1
2 Mobile combustion, CO2 only, 10 fuels per gallon or scf 40 CFR Part 98, Table C-1
3–5 Mobile CH4 and N2O for on-road petrol, on-road diesel and alternative-fuel, and non-road vehicles per mile or per gallon, by vehicle type and model year US GHG Inventory 1990–2022, annexes
6 Grid electricity for 27 eGRID subregions plus a US average, with grid gross loss % lb per MWh eGRID2023 (January 2025)
7 Purchased steam and heat per mmBtu EPA calculation, natural gas at 80% efficiency
8 Scope 3 Categories 4 and 9: road, rail, water and air freight per vehicle-mile or short ton-mile US GHG Inventory; Bureau of Transportation Statistics
9 Scope 3 Categories 5 and 12: 61 materials by disposal route metric tons CO2e per short ton EPA WARM documentation (December 2023), adjusted
10 Scope 3 Categories 6 and 7: car, rail, bus and air travel per vehicle-mile or passenger-mile US GHG Inventory; Amtrak; DEFRA 2022 for air
11 Global warming potentials of individual gases 100-year GWP IPCC AR5 (2013)
12 GWPs of blended refrigerants 100-year GWP IPCC AR5, with EPA SNAP blend compositions

Two of those rows deserve a second look. The air travel factors in Table 10 are not American at all: they are the UK government’s 2022 figures, which the Hub says are “held constant from 2022 release (2018 activity data)”, because later UK editions reflect low pandemic-era load factors. And Table 9’s waste factors are on a different GWP edition from the rest of the file, which gets its own section below.

Stationary combustion: Table 1

Table 1 covers fuel burned in boilers, furnaces, generators and kilns: the core of most Scope 1 inventories. The EPA lists each gas separately, CO2 in kilograms and CH4 and N2O in grams. Most fuels are given both per unit of energy (mmBtu, on a higher heating value) and per physical unit. The right-hand column below combines the three gases into CO2e using the Hub’s own AR5 values, which is the figure most people actually need.

Selected Table 1 factors: EPA per-gas values and the combined CO₂e
Fuel Unit CO₂ (kg) CH₄ (g) N₂O (g) kg CO₂e (AR5)
Natural gas mmBtu 53.06 1.0 0.10 53.11
Natural gas scf 0.05444 0.00103 0.00010 0.05450
Propane gallon 5.72 0.27 0.05 5.74
Distillate fuel oil No. 2 gallon 10.21 0.41 0.08 10.24
Motor gasoline gallon 8.78 0.38 0.08 8.81
Bituminous coal short ton 2,325 274 40 2343
53.06
kg CO₂ per mmBtu of natural gas: the most-used number in the Hub (Table 1, 2025)
99.9% of the CO₂e is CO₂; methane and nitrous oxide add 0.05 kg

That last line generalises. For fossil fuels burned in stationary equipment, CO2 is nearly the whole answer: the methane and nitrous oxide terms move natural gas by 0.1% and bituminous coal by under 1%. That is why many inventories report combustion as CO2 alone. The Hub still gives all three gases so that you can show the other two were counted.

Three notes under Table 1 are easy to miss:

  • Heating values are gross (HHV). If your energy figures are on a net basis (LHV), as European data usually is, convert before multiplying.
  • Combustion only. The EPA’s words: the factors “represent combustion emissions only and do not represent upstream emissions.” Extraction, refining and transport of the fuel are not in them.
  • Supplier data wins. “If heat content is available from the fuel supplier, it is preferable to use that value.”

Vehicles: Tables 2 to 5

Vehicle emissions are split in a way that surprises people. CO2 depends almost entirely on how much fuel is burned, so Table 2 gives it per gallon: 10.21 kg CO2 per gallon [GreenCalculus mobile_combustion.usa.diesel.gallon · EPA GHG 2025 Emission Factors Hub!D107 · v2026.248] of diesel and 8.78 per gallon [GreenCalculus mobile_combustion.usa.motor_gasoline.gallon · EPA GHG 2025 Emission Factors Hub!D112] of motor gasoline. Methane and nitrous oxide depend on the engine and its exhaust treatment, so Tables 3 to 5 give them per mile, by vehicle type and model year.

The result is that no single per-gallon CO2e figure exists for a US vehicle in the Hub. To report all three gases you need fuel used and miles driven, for each vehicle class. Our database stores Table 2 as CO2 only, labelled as such, for that reason: a “CO2e per gallon” figure built from Table 2 alone would quietly leave two gases out.

Electricity and steam: Tables 6 and 7

Table 6 is a copy of the eGRID2023 subregion summary table. It gives three gases for each of 27 subregions plus a US average, in pounds per megawatt-hour. The US average total output rate is 771.5 lb CO2 per MWh, or 0.350 kg per kWh. The subregions range from 241.0 lb in Upstate New York, where hydro and nuclear dominate, to 1,543.0 lb in Puerto Rico, which is double the national figure. Use your own subregion rather than the national average; our eGRID reference covers the subregion map and how to pick one.

Each subregion has two rates, and the Hub is clear about which one belongs in an inventory:

EPA, Table 6 note (verbatim)

“Annual non-baseload output emission factors should not be used when developing a carbon footprint or emissions inventory, but can be used to estimate GHG emissions reductions on the grid from changes in electricity use.”

The non-baseload rate (1,393.7 lb CO2/MWh nationally) is the plant that ramps up or down when demand changes, so it is the right number for “how much did our efficiency project save?” and the wrong one for “what was our footprint?”. Mixing them up overstates Scope 2 by roughly 80% at the national level.

New in 2025: Table 6 also gives a grid gross loss percentage for each subregion (4.1–4.4%). The Hub says this is for the transmission-and-distribution part of Scope 3 Category 3, and should not be applied to an end user’s Scope 2.

Table 7 gives one figure for purchased steam and heat: 66.33 kg CO2 per mmBtu, or 66.40 kg CO2e [GreenCalculus district_heating.usa.steam_heat.mmbtu · EPA GHG 2025 Emission Factors Hub!D410:F410] with the other two gases added. The EPA notes that it assumes “natural gas fuel is used to generate steam or heat at 80 percent thermal efficiency”. The boiler loss is already inside the number, so do not add it again.

Scope 3: Tables 8, 9 and 10

The three Scope 3 tables follow the GHG Protocol’s Scope 3 calculation guidance, and each says which of its methods it supports. Tables 8 and 10 are for the distance-based method. If you have fuel data, the Hub tells you to use the fuel-based method with Tables 2 to 5 instead.

Business travel and commuting (Table 10)

One table serves both Category 6 (business travel) and Category 7 (employee commuting). Combined on AR5:

Passenger car (per vehicle-mile)

0.299

Air, short haul <300 mi

0.209

Air, long haul ≥2,300 mi

0.164

Commuter rail

0.134

Air, medium haul

0.130

Bus

0.067

Values are kg CO2e per passenger-mile, except the car, which is per vehicle-mile. Divide by the number of people in the car before comparing it with the others.

Freight (Table 8)

Table 8 has two kinds of factor. Per vehicle-mile is for a truck or car used only for your goods. Per short ton-mile is for shared transport. For shared freight the spread is large: 0.021 kg CO2 per ton-mile by rail, 0.186 by medium- and heavy-duty truck and 1.086 by air. Moving a ton by air emits about 50 times as much CO2 as moving it by rail.

Waste (Table 9)

Table 9 covers Category 5 and Category 12 for 61 materials and up to six routes: recycled, landfilled, combusted, composted and two kinds of anaerobic digestion. Selected values, in metric tons CO2e per short ton:

Selected Table 9 factors: metric tons CO₂e per short ton (AR4 basis)
Material Recycled Landfilled Combusted Composted
Corrugated containers 0.11 1.00 0.05 n/a
Food waste n/a 0.68 0.05 0.11
Mixed plastics 0.22 0.02 2.34 n/a
Aluminium cans 0.06 0.02 0.01 n/a
Mixed MSW n/a 0.58 0.43 n/a

Note the units: metric tons of CO2e per short ton of material (2,000 lb). Paper and food waste are worst in a landfill, where they rot into methane. Plastics are worst when burned, because they are made from fossil carbon. These figures will not match the EPA’s WARM tool, and the Hub explains why: they exclude “avoided emissions impact from any of the disposal methods”. Recycling here gets no credit for the virgin material it replaces. That is the correct treatment for a Scope 3 inventory, where avoided emissions are reported separately if at all.

Two GWP editions in one file

The Hub opens by telling you to convert methane and nitrous oxide with the IPCC’s Fifth Assessment Report values: 28 for CH4 and 265 for N2O. Tables 11 and 12 are also AR5. Table 9 is the exception, and its note says so in one line:

EPA, Table 9 note (verbatim)

“AR4 GWP values are used to convert all waste emission factors into CO2e.”

The waste factors are published already in CO2e, with no per-gas breakdown, so you cannot restate them onto AR5 yourself. The practical answer is to disclose it: “waste factors from EPA Hub 2025 Table 9, AR4 basis; all other factors AR5.” Note also that AR5 is not the newest IPCC set, and some frameworks now ask for AR6. If yours does, the per-gas tables (1, 6, 7, 8 and 10) can be recombined on AR6; Table 9 cannot.

The Hub is a reliable file with one internal inconsistency: its waste table is on a different warming-potential edition from everything else, and it says so in a single line.

EPA against DEFRA

The UK government’s DEFRA conversion factors are the other widely used national set, and many multinational inventories use both. On the same fuel, they agree closely. Per kWh of natural gas on a gross basis, the EPA figure restated is 0.18123 kg CO2e [GreenCalculus fuels.usa.natural_gas.kwh_gcv · EPA GHG 2025] against DEFRA 2026’s 0.18231 [GreenCalculus fuels.gbr.natural_gas.kwh_gcv · DEFRA 2026 'Fuels'!D42]. That is a 0.6% gap, and natural gas is chemically much the same everywhere.

The bigger difference is coverage, not values:

EPA GHG Emission Factors Hub 2025 and DEFRA 2026, side by side
EPA Hub 2025 DEFRA 2026
Units US customary: gallon, scf, mmBtu, short ton, mile Metric: litre, m³, kWh, tonne, km
Gases Per gas (CO₂, CH₄, N₂O), you combine them Per gas and pre-combined CO₂e
Upstream (well-to-tank) Not published: “combustion emissions only” Published for every fuel. Natural gas adds 0.03021 kg/kWh, about 17%
Electricity 27 US subregions (eGRID) One UK-wide grid factor
Update rhythm Irregular: 2023, 2024 and 2025 editions came out in different months, and no 2026 edition yet Annual, every June
Licence US Government work, public domain Open Government Licence v3.0

The upstream row matters most. A US inventory built only on the Hub has no Scope 3 Category 3 fuel figure, and for natural gas that missing piece is about a sixth of the combustion figure. US upstream factors come from other sources, such as Argonne’s GREET model or NETL’s electricity baseline, not from the Hub.

Worked example

Worked example: a US office and its small fleet, one year
Activity Quantity Hub factor t
Natural gas, heating (Table 1) 1,200 mmBtu (12,000 therms) 53.1145 kg CO₂e/mmBtu 63.74
Diesel, fleet (Table 2) 4,000 gallons 10.21 kg CO₂/gallon 40.84
Electricity, US average (Table 6) 500 MWh 771.5 lb CO₂/MWh 174.97
Air travel, medium haul (Table 10) 150,000 passenger-miles 0.130103 kg CO₂e/passenger-mile 19.52
Total 299.07

Each line is quantity times factor. The electricity line converts pounds to kilograms (× 0.45359237). Two lines are CO2 only: the diesel needs miles by vehicle type from Tables 3–4 for its methane and nitrous oxide, and the electricity needs the CH4 and N2O rates from Table 6. Electricity is 59% of the total on the national average. Using the office’s own eGRID subregion could cut that line to about a third (Upstate New York) or double it (Puerto Rico).

Common errors

Errors that recur with the EPA Hub
Error Why it is wrong Do this instead
Using the non-baseload electricity rate for Scope 2 The Hub says it “should not be used” for a footprint; it overstates the US average by about 80% Use the total output rate for your eGRID subregion
Adding CH₄ and N₂O grams as if they were kilograms Table 1 gives CO₂ in kg but CH₄ and N₂O in g Divide the gram columns by 1,000 before applying the GWP
Applying HHV factors to LHV energy data The Hub is on higher heating value; net-basis energy is about 10% lower for gas Convert the energy figure to HHV first
Treating Table 2 as the whole vehicle footprint It is CO₂ only; CH₄ and N₂O are per mile in Tables 3–5 Add them from miles driven, or state that they are excluded
Restating the waste factors on AR5 They are published as CO₂e on AR4 with no per-gas split, so there is nothing to restate Use them as published and disclose the AR4 basis
Expecting the waste factors to match WARM The Hub removes WARM’s avoided-emissions credits on purpose Report avoided emissions separately, if at all
Using an archived edition without saying so The EPA page keeps every edition back to 2011, marked “ARCHIVED” Cite the edition and its date, e.g. “EPA GHG Emission Factors Hub, January 2025”

Is there a 2026 edition?

Not yet. On 2 October 2026, the EPA’s Hub page (last updated 12 January 2026) still lists the 2025 edition, dated January 2025, as the latest. Every earlier edition is marked “ARCHIVED”. Despite the word “annual”, release timing has varied: the 2022 PDF is dated April, 2023 September, 2024 June and 2025 January. For a 2025 or 2026 reporting year, the January 2025 file is the current EPA default. Name the edition in your methodology note, so that a later edition does not make your figures look wrong.

Put the Hub’s fuel and travel factors to work, or look up any factor with its source attached.

Provenance and data access

From the EPA publication to this page
Stage Detail
Primary source US EPA Center for Corporate Climate Leadership, GHG Emission Factors Hub, 2025 edition (PDF “Last Modified: January 15, 2025”, and the matching xlsx).
In MasterBrain Tables 1, 2, 7, 9 and 10, source ID EPA_GHG_HUB_2025, each row citing its workbook cell. Table 6 electricity is carried from eGRID directly (source EPA_EGRID_2023). Tables 3–5 and 8 are not carried yet.
CO₂e Recombined from the Hub’s per-gas values on AR5 (CH₄ 28, N₂O 265), the Hub’s own stated basis. Table 2 is kept as CO₂ only. Per-kWh, per-litre and per-tonne rows are exact unit restatements, labelled as such.
Licence US Government work, public domain in the United States. Free to use and redistribute; citing the EPA is good practice rather than a condition.

Every Hub row is served from the open factor API, grouped by key prefix:

  • /wp-json/greencalculus/v1/factors?key_prefix=fuels.usa: Table 1, 39 rows
  • /wp-json/greencalculus/v1/factors?key_prefix=mobile_combustion.usa: Table 2, 19 rows
  • /wp-json/greencalculus/v1/factors?key_prefix=business_travel.usa: Table 10, 12 rows
  • /wp-json/greencalculus/v1/factors?key_prefix=waste_disposal.usa: Table 9, 183 rows
  • /wp-json/greencalculus/v1/factors?key_prefix=district_heating.usa: Table 7, 2 rows

Each response carries the source registry entry and the workbook cell reference for every value.

Citation guidance

U.S. Environmental Protection Agency (2025). GHG Emission Factors Hub, January 2025. Center for Corporate Climate Leadership. https://www.epa.gov/climateleadership/ghg-emission-factors-hub

Primary citation

Frequently asked questions

53.06 kg CO2 per mmBtu, plus 1.0 g of methane and 0.10 g of nitrous oxide, from Table 1 of the 2025 Hub. Combined on the Hub’s AR5 values that is 53.11 kg CO2e per mmBtu, or 0.0545 kg CO2e per standard cubic foot. One therm is 0.1 mmBtu, so a therm of gas is about 5.31 kg CO2e.

From the EPA’s GHG Emission Factors Hub page, which offers the current edition as an Excel workbook and a PDF, with archived editions back to 2011. The values on this page come from the 2025 workbook, and the same rows are available as JSON from the open factor API listed under Data access.

AR5: 28 for methane and 265 for nitrous oxide, from the IPCC’s Fifth Assessment Report. The one exception is the waste table (Table 9), which the EPA converts with AR4 values and publishes only as CO2e. If you combine the two in one inventory, say so in your methodology note.

They give the same answer, because the Hub’s Table 6 is copied from eGRID’s subregion summary table. Use the Hub if you want everything in one file. Go to eGRID itself if you need state or plant-level rates, or the most recent eGRID release if one comes out between Hub editions. Either way, use the total output rate, not the non-baseload rate.

For fuel combustion, largely yes: a mmBtu of natural gas or a gallon of diesel emits about the same anywhere, and the EPA and DEFRA gas figures are within 1% of each other. For electricity, no: Table 6 describes US grids only, so use a factor for the country where the power is consumed. For freight and travel, the factors reflect US vehicle fleets and load factors, so a local source is better when one exists.

Default values, published by the US Environmental Protection Agency, for how much greenhouse gas an activity releases per unit: per gallon of fuel, per megawatt-hour of electricity, per mile travelled or per ton of waste. For company reporting they are collected in the GHG Emission Factors Hub, whose latest edition is January 2025. You multiply your activity data by the factor to get emissions.

Version history

Substantive revisions to this dataset reference page
Version Date MasterBrain Summary
1.0 2026-10-02 v2026.204 Initial publication: EPA GHG Emission Factors Hub 2025, all 12 tables described, live Table 1, 2, 7 and 10 values, Table 6 and 9 figures from the PDF, the AR4/AR5 split, EPA vs DEFRA comparison, worked example, 6-item FAQ.

Scroll to Top