An AP-42 emission factor relates the quantity of a pollutant released to an activity, such as pounds of NOx per million cubic feet burned [1]. EPA describes most of these factors as simply averages of all available data of acceptable quality for a source category [1]. Those averages are assumed to represent long-term averages for every facility in that category [1].
The tables below cover the combustion equipment an upstream or midstream emissions inventory actually contains, meaning boilers and heaters, turbines, reciprocating engines, and flares. Every factor carries a letter rating describing the strength of the data behind it, and the rating says nothing about your equipment [1].
Quick Answer: AP-42 gives natural gas combustion factors in pounds per million standard cubic feet [2]. Small boilers under 100 MMBtu/hr carry 100 lb NOx, and all natural gas combustion carries 5.5 lb VOC [2]. Turbines, engines, and flares are expressed per million Btu of fuel input instead [3]. Every factor describes equipment operating normally, because EPA keeps short-term fluctuation conditions out of its test program [1].
What an AP-42 Emission Factor Rating Actually Tells You
The A through E rating measures the reliability of the test data behind a factor, with A being the best [1]. An A-rated factor comes from A- and B-rated source tests taken at many randomly chosen facilities in the industry population [1]. An E-rated factor comes from C- and D-rated test data where there may be reason to suspect the tested facilities are not a random sample [1].
EPA is unusually direct about the limits of its own scale. The ratings are subjective, consider the scatter in the underlying data only indirectly, and should be seen as approximations [1]. They do not imply statistical error bounds or confidence intervals around any factor [1].
A Factor Is a Population Average, Not a Facility Estimate
A factor describes a category of equipment, so roughly half the sources in that category emit more than it says [1]. EPA makes the consequence explicit, noting that a permit limit set at an AP-42 factor would place half of the sources in noncompliance [1]. EPA does not recommend using the factors as source-specific permit limits or for compliance determinations [1].
That distinction survives every downstream use of the number. An inventory built from factors is an estimate of what a population of similar facilities emits on average, applied to yours.
The Rating Says Nothing About Your Equipment
Two factors with the same letter can behave very differently in practice. The large stationary diesel NOx factor is rated B [4]. EPA notes in the same table that actual emissions from a particular engine could vary considerably from the published level [4]. The factors are averaged across all manufacturers and duty cycles [4].
Sampling depth varies more than the letter suggests. The methane factor for large stationary diesel engines is rated E and rests on data from a single engine [4]. One engine standing in for a national equipment population is worth pausing on before that number enters a reported inventory.
AP-42 Emission Factors Change Units by Source Type
Combustion factors arrive in two different unit systems, and mixing them is the most common arithmetic error in a factor-built inventory. Natural gas external combustion is published per million standard cubic feet burned [2]. Turbines and reciprocating engines are published per million Btu of fuel input, and engines add a power-output basis in pounds per horsepower-hour [3].
Factors below are pounds per million standard cubic feet of natural gas fired, based on an average higher heating value of 1,020 Btu/scf. [2] Dividing by 1,020 converts them to pounds per million Btu [2].
| Combustor and control | NOx (lb/10⁶ scf) | Rating | CO (lb/10⁶ scf) | Rating |
|---|---|---|---|---|
| Large wall-fired boiler >100 MMBtu/hr, uncontrolled pre-NSPS | 280 | A | 84 | B |
| Large wall-fired boiler >100 MMBtu/hr, uncontrolled post-NSPS | 190 | A | 84 | B |
| Large wall-fired boiler >100 MMBtu/hr, low-NOx burners | 140 | A | 84 | B |
| Large wall-fired boiler >100 MMBtu/hr, flue gas recirculation | 100 | D | 84 | B |
| Small boiler <100 MMBtu/hr, uncontrolled | 100 | B | 84 | B |
| Small boiler <100 MMBtu/hr, low-NOx burners | 50 | D | 84 | B |
| Small boiler <100 MMBtu/hr, low-NOx burners and FGR | 32 | C | 84 | B |
| Tangential-fired boiler, uncontrolled | 170 | A | 24 | C |
| Residential furnace <0.3 MMBtu/hr | 94 | B | 40 | B |
AP-42 Table 1.4-1, expressed as NO2 [2].
Organic and greenhouse pollutants apply across all natural gas combustion sources rather than per combustor type [2]. VOC is 5.5 lb/10⁶ scf at a C rating, methane is 2.3 at B, and total organic compounds is 11 at B [2]. Carbon dioxide is 120,000 lb/10⁶ scf at an A rating, SO2 is 0.6 at A, and total PM is 7.6 at D [2].
Methane and ethane are published separately, and reading them together is instructive. AP-42 Table 1.4-3 lists speciated organic compounds for all natural gas combustion sources [2].
| Compound | Factor (lb/10⁶ scf) | Rating |
|---|---|---|
| Methane | 2.3 | B |
| Ethane | 3.1 | E |
| Pentane | 2.6 | E |
| Butane | 2.1 | E |
| Hexane | 1.8 | E |
| Propane | 1.6 | E |
Methane from AP-42 Table 1.4-2, the C2 through C6 alkanes from Table 1.4-3, both per million standard cubic feet fired [2].
The ethane factor is larger than the methane factor [2]. That inverts what the fuel’s own composition would suggest, since natural gas is generally above 85 per cent methane [2]. Methane carries a B rating while ethane and every heavier alkane carry an E [2]. An E rating means the factor was developed from C- and D-rated test data that may not be a random sample [1].
Across the other source types the picture thins out quickly.
| Source | Methane | Rating | Ethane | Rating |
|---|---|---|---|---|
| Natural gas combustion (boilers, heaters) | 2.3 lb/10⁶ scf | B | 3.1 lb/10⁶ scf | E |
| Natural gas-fired turbine | 0.0086 lb/MMBtu | C | not published | n/a |
| Distillate oil-fired turbine | no data | n/a | not published | n/a |
| Landfill or digester gas turbine | not published | n/a | not published | n/a |
| Dual-fuel engine (95% gas, 5% diesel) | 0.6 lb/MMBtu | E | not published | n/a |
| Large stationary diesel engine | no value published | E | not published | n/a |
| Gasoline or diesel industrial engine | not published | n/a | not published | n/a |
| Two-stroke lean-burn gas engine (2SLB) | 1.45 lb/MMBtu | C | 0.0709 lb/MMBtu | A |
| Four-stroke lean-burn gas engine (4SLB) | 1.25 lb/MMBtu | C | 0.105 lb/MMBtu | C |
| Four-stroke rich-burn gas engine (4SRB) | 0.23 lb/MMBtu | C | 0.0704 lb/MMBtu | C |
| Elevated flare | THC only, 0.14 lb/10⁶ Btu | B | not published | n/a |
| Enclosed ground flare | THC only, 332 lb/10⁶ scf | Poorly representative | not published | n/a |
Compiled from AP-42 Tables 1.4-2, 1.4-3, 3.1-1, 3.1-2a, 3.1-2b, 3.3-1, 3.4-1, 13.5-1 and 13.5-3 [2].
Three entries in that table need reading carefully. The large stationary diesel row carries an E rating but no value [4]. The chapter offers only a note that total organics run 9 per cent methane by weight, from one engine [4]. The flare rows report total hydrocarbons rather than methane, measured as methane equivalent for elevated flares and as propane for enclosed ground flares [5].
Ethane is published in two places, and the second is the compressor fleet [6]. Natural gas combustion carries one value across every combustor variant, and each gas-fired reciprocating engine class carries its own [6]. It is absent from turbines, from the diesel and dual-fuel engines, and from flares [7].
Those chapters speciate organics into hazardous air pollutants such as benzene and formaldehyde rather than light alkanes [7]. An operator wanting an ethane figure for a gas compressor engine has one; for a diesel engine or a flare, none.
Turbine and Engine Factors Use Heat Input or Power Output
Turbine factors are pounds per million Btu of fuel input, derived only from units running at high load near 80 percent [3].
| Turbine or engine | Pollutant | Factor | Rating |
|---|---|---|---|
| Natural gas turbine, uncontrolled | NOx | 0.32 lb/MMBtu | A |
| Natural gas turbine, uncontrolled | CO | 0.082 lb/MMBtu | A |
| Natural gas turbine, water or steam injection | NOx | 0.13 lb/MMBtu | A |
| Natural gas turbine, lean premix | NOx | 0.099 lb/MMBtu | D |
| Natural gas turbine, uncontrolled | VOC | 0.0021 lb/MMBtu | D |
| Distillate oil turbine, uncontrolled | NOx | 0.88 lb/MMBtu | C |
| Large stationary diesel engine, uncontrolled | NOx | 0.024 lb/hp-hr (3.2 lb/MMBtu) | B |
| Large stationary diesel engine, controlled | NOx | 0.013 lb/hp-hr (1.9 lb/MMBtu) | B |
| Large stationary diesel engine | CO | 0.0055 lb/hp-hr (0.85 lb/MMBtu) | C |
| Industrial diesel engine, uncontrolled | NOx | 0.031 lb/hp-hr (4.41 lb/MMBtu) | D |
| Industrial diesel engine, uncontrolled | CO | 0.00668 lb/hp-hr (0.95 lb/MMBtu) | D |
| Industrial gasoline engine, uncontrolled | NOx | 0.011 lb/hp-hr (1.63 lb/MMBtu) | D |
AP-42 Tables 3.1-1, 3.1-2a, 3.3-1, and 3.4-1. Industrial engine conversions assume a brake-specific fuel consumption of 7,000 Btu/hp-hr [7].
Gas-fired reciprocating engines get their own chapter, and it is the one an upstream or midstream inventory leans on hardest [6]. AP-42 places these engines at pipeline compressor and storage stations and at gas processing plants, driving compressors and pumps [6].
| Engine class | CH₄ | C₂H₆ | TOC | VOC | NOx | CO |
|---|---|---|---|---|---|---|
| Two-stroke lean-burn (2SLB) | 1.45 (C) | 0.0709 (A) | 1.64 (A) | 0.12 (C) | 3.17 (A) | 0.386 (A) |
| Four-stroke lean-burn (4SLB) | 1.25 (C) | 0.105 (C) | 1.47 (A) | 0.118 (C) | 4.08 (B) | 0.317 (C) |
| Four-stroke rich-burn (4SRB) | 0.23 (C) | 0.0704 (C) | 0.358 (C) | 0.0296 (C) | 2.21 (A) | 3.72 (A) |
AP-42 Tables 3.2-1, 3.2-2 and 3.2-3, uncontrolled, lb/MMBtu of fuel input, rating in brackets. NOx and CO shown at 90 to 105 per cent load [6].
All three classes share a CO₂ factor of 110 lb/MMBtu at an A rating and an SO₂ factor of 0.000588 at an A rating [6]. NOx and CO are published split by load [6]. That matters, because a compressor engine below 90 per cent load is the normal case rather than the exception. For 4SLB the NOx factor falls from 4.08 to 0.847 below that line, and for 4SRB the CO factor moves from 3.72 to 3.51 [6].
One entry in the turbine table contradicts itself, and EPA says so in a footnote [3]. The uncontrolled CO factor is higher than the water-steam-injection and lean-premix factors, which runs contrary to expectation [3]. EPA could not identify the reason beyond the datasets being different [3]. A control technology that appears to raise emissions is a sign the factor is describing test populations rather than physics.
Flare Factors Report What Leaves the Flare, Not What Enters It
Flare factors describe emissions exiting the flare, because the flare is a control device rather than the originating source [5]. EPA warns they may not be appropriate for estimating the uncontrolled emissions routed to the flare [5]. The same caution applies to the potential to emit of the process feeding it [5].
| Flare type | Pollutant | Factor | Grade |
|---|---|---|---|
| Elevated flare | VOC | 0.66 lb/10⁶ Btu | Poorly representative |
| Elevated flare | CO | 0.31 lb/10⁶ Btu | Poorly representative |
| Elevated flare | THC | 0.14 lb/10⁶ Btu | B |
| Elevated flare | NOx | 0.068 lb/10⁶ Btu | B |
| Enclosed ground flare, gas production site | THC | 332 lb/10⁶ scf burned (0.335 lb/10⁶ Btu) | Poorly representative |
AP-42 Tables 13.5-1, 13.5-2, and 13.5-3. Newer flare factors carry a representativeness grade instead of a letter rating [5].
The conditions attached to these numbers matter more than the numbers. Elevated flare factors apply to well-operated flares achieving at least 98 percent destruction efficiency, and the underlying VOC dataset averaged 98.9 percent [5]. The enclosed ground flare dataset averaged 99.33 percent destruction efficiency on a gas volume basis [5]. Applying these factors to a flare that is unlit, or running outside its design envelope, assumes away the failure being asked about.
For a related look at where one piece of equipment dominates a facility total, see TEG Dehydrator Emissions: Why Glycol Pumps Are 90% of the Problem.
What Every Factor Assumes, and What Changes When You Model Failure
Every factor in the tables above describes equipment operating as designed, and EPA states the assumption plainly rather than hiding it [1]. EPA runs the underlying tests at normal operating conditions [1]. The conditions that drive short-term swings are kept out of the test program, and out of the test evaluation as well [1].
The introduction goes further and names the sampling bias directly [1]. Units get tested more often when they are new, and when the operator already believes they are running properly [1]. EPA notes that either circumstance can bias the resulting factor [1]. That sentence has been in AP-42 since 1995, and it is the most useful sentence in the document.
The Gap Shows Up When Someone Measures
Failure conditions are not exotic, and inventories that model them land above inventories that do not. In Colorado’s Denver-Julesburg Basin, a measurement-informed inventory came in at 1.16 times the total operators reported to the state [8]. Across Colorado statewide, the MAES model in the Colorado Ongoing Basin Emissions study returned 1.47 times the reported inventory on 2024 data [9]. A second statistical model in the same study returned a higher ratio, so the two approaches did not converge [9].
Neither gap comes from a disputed factor value. Both come from including operating states that the factor’s test population never contained. Flaring is the source where that gap is widest, walked through in Oil and Gas Flaring: What Flares Actually Destroy.
What a Mechanistic Model Holds That a Factor Cannot
A mechanistic model simulates the equipment and the process instead of multiplying an activity by a constant [10]. The Mechanistic Air Emissions Simulator is a physics-based simulator of that kind, developed at CSU and UT Austin [10]. It generates an expected emission range for a facility rather than a single value [10]. TetraSoft uses MAES through a partnership with CSU [11].
The practical difference is which questions stay answerable. A stuck dump valve, an open thief hatch, and a flare outside its operating envelope have no factor to look up. No factor was built from equipment in that state. Factors and mechanistic models are not competing answers to one question.
For the measurement side of the same question, see Why Satellites Show About 2x More Methane Than Inventories.
What AP-42 Does Not Give You
The chapters above cover combustion. An upstream or midstream inventory contains a good deal that burns nothing, and for most of it AP-42 has no factor at all.
Storage tanks are the exception that proves the shape of the problem. AP-42 Chapter 7.1 handles them with equations rather than emission factors, and it declines to give an equation for flashing at all [12]. Our Storage Tank Emissions Calculation: Formulas and Failure Modes walks through the two equations it does publish and what the missing third one costs.
Dehydrators get no AP-42 treatment whatsoever, and the working method is GRI-GLYCalc rather than a lookup. TEG Dehydrator Emissions: Why Glycol Pumps Are 90% of the Problem covers where those emissions actually concentrate.
Three more sit outside the compilation entirely. Equipment leaks are estimated with EPA’s Protocol for Equipment Leak Emission Estimates. Pneumatic controllers and pumps are counted under the Greenhouse Gas Reporting Program’s Subpart W. Well completions and workovers are handled the same way.
Loading and transportation losses do have an AP-42 chapter, 5.2. They are a working loss rather than a combustion source, so they sit outside this post.
The practical consequence is that a factor-built inventory assembled only from this post is a combustion inventory, not a facility inventory.
Frequently Asked Questions
Why is the ethane factor rated E when methane is rated B?
The rating reflects the test data behind each factor, not the compound [1]. An E-rated factor is developed from C- and D-rated test data, where the tested facilities may not be a random sample [1]. A B-rated factor comes from A- or B-rated data at a reasonable number of facilities [1]. Ethane and the heavier alkanes sit in the speciated table, which is measured less often than the headline pollutants.
Can I use an AP-42 factor as a permit limit?
EPA does not recommend it [1]. Because a factor is an average, roughly half of the sources in a category emit more than the factor and half emit less [1]. EPA notes in the AP-42 introduction that a permit limit set at the factor would put half the sources in noncompliance [1]. Reviewing agencies set the applicable method for any given submission.
Why do AP-42 factors use different units for the same pollutant?
Units follow whatever activity the factor was measured against. Natural gas combustion factors are pounds per million standard cubic feet burned [2]. Turbine factors are pounds per million Btu of fuel input [3].
That heat-input denominator is only as good as the gas behind it, which varies more than the factor assumes, as set out in Natural Gas BTU: What Wellhead Gas Actually Measures. Reciprocating engine factors carry both a power-output basis in pounds per horsepower-hour and a fuel-input basis [7]. Converting between them requires a heating value or a brake-specific fuel consumption [7].
Are enclosed ground flares treated as flares under OOOO and OOOOa?
No [5]. Under 40 CFR part 60 subparts OOOO and OOOOa, these units are not considered flares [5]. The same exclusion applies under 40 CFR part 63 subparts HH and HHH [5].
The definition of flare in those subparts specifically excludes them [5]. The AP-42 factor still applies to the unit even though the subpart definition does not.
How do I apply a control efficiency to an AP-42 factor?
EPA’s general equation is emissions equal activity rate times emission factor times one minus ER over 100 [1]. ER is the overall emission reduction efficiency, expressed as a percentage [1]. EPA defines ER as the control device destruction or removal efficiency multiplied by the capture efficiency of the control system [1]. For a period as long as a year, EPA states both terms should account for upset periods as well as routine operations [1].
Interested in building a Measurement-Informed Inventory for your operations? Contact us to learn about our MAES-based estimation services.
References
- EPA AP-42, Introduction to AP-42 Volume I, Fifth Edition, January 1995. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-09/documents/c00s00.pdf
- EPA AP-42 Chapter 1.4, Natural Gas Combustion, April 2026. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-09/documents/1.4_natural_gas_combustion.pdf
- EPA AP-42 Chapter 3.1, Stationary Gas Turbines, April 2000. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/c03s01.pdf
- EPA AP-42 Chapter 3.4, Large Stationary Diesel and All Stationary Dual-Fuel Engines, October 1996. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/c03s04.pdf
- EPA AP-42 Chapter 13.5, Industrial Flares, February 2018. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/13.5_industrial_flares.pdf
- EPA AP-42 Chapter 3.2, Natural Gas-fired Reciprocating Engines, July 2000. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/c03s02.pdf
- EPA AP-42 Chapter 3.3, Gasoline and Diesel Industrial Engines, October 1996. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/c03s03.pdf
- Santos et al., ACS ES&T Air 2025, 2, 1598-1611, DOI 10.1021/acsestair.5c00089. Using Measurement-Informed Inventory to Assess Emissions in the Denver-Julesburg Basin. https://doi.org/10.1021/acsestair.5c00089
- Brown et al., Colorado Ongoing Basin Emissions (COBE) Updated Final Report, November 20, 2025. Colorado State University METEC and Colorado School of Mines, for the Colorado Department of Public Health and Environment. https://metec.colostate.edu/colorado-ongoing-basin-emissions-cobe/
- Mollel et al., ACS ES&T Air 2025, 2, 723-735, DOI 10.1021/acsestair.4c00168. Using Prototypical Oil and Gas Sites to Model Methane Emissions in Colorado’s Denver-Julesburg Basin Using a Mechanistic Emission Estimation Tool. https://doi.org/10.1021/acsestair.4c00168
- TetraSoft, MAES Platform. https://tetrasoftco.com/maes-platform/maes-landing.html
- EPA AP-42 Chapter 7.1, Organic Liquid Storage Tanks, October 2024. United States Environmental Protection Agency, Compilation of Air Pollutant Emission Factors. https://www.epa.gov/sites/default/files/2020-10/documents/ch07s01.pdf
This post is for informational purposes only and does not constitute legal or compliance advice. Consult qualified legal counsel or a compliance professional for guidance specific to your operations and jurisdiction.
Interactive tool
What does AP-42 actually publish for your source?
Pick a source and a pollutant. Where AP-42 publishes a factor this returns it with its rating and does the arithmetic. Where it publishes nothing, it says so and says which kind of nothing, because for methane and ethane that is usually the answer.
Published
230 lb/yr
Where ethane exists at all published factors per group
- Natural gas combustion 9 of 9
- Gas-fired reciprocating engines 3 of 3
- Turbines 0 of 3
- Diesel and dual-fuel engines 0 of 3
- Flares 0 of 2
Ethane is published in two places: natural gas combustion, where all nine combustor variants share one value, and the gas-fired reciprocating engines of Chapter 3.2, where each class carries its own. It is absent from turbines, from the diesel and dual-fuel engines, and from flares, because those chapters speciate organics into hazardous air pollutants such as benzene and formaldehyde rather than into light alkanes.
- A factor describes a category of equipment, so roughly half the sources in that category emit more than it says.
- EPA states a factor is generally not an appropriate basis for a permit limit on a specific facility.
- The rating measures the test data behind the factor, not your equipment. EPA calls the scale subjective.
Hourly figures annualise at 8,760 hours, the same basis the other tools on this site use. Factors from EPA AP-42 Tables 1.4-1, 1.4-2, 1.4-3, 3.1-1, 3.1-2a, 3.1-2b, 3.2-1, 3.2-2, 3.2-3, 3.3-1, 3.4-1, 13.5-1, 13.5-2 and 13.5-3. Conversions between the two unit systems use the chapter's own 1,020 Btu/scf. Nothing leaves your browser.
Comments
0Share your thoughts. All comments are moderated before appearing.
No comments yet. Be the first to share your thoughts!