Emissions estimation is the foundation of every air quality compliance obligation. Your permit application, annual emissions inventory, Title V compliance certification, greenhouse gas report, and emission fee calculation all depend on defensible emissions estimates. Use the wrong method or the wrong emission factor, and your entire compliance structure is built on sand. Underestimate and you risk enforcement action. Overestimate and you pay higher fees, trigger unnecessary permit requirements, and may be forced to install controls you do not actually need.
EPA provides a hierarchy of estimation methods, from most preferred to least: continuous emissions monitoring (CEMS), stack testing, material balance, and emission factors. In practice, most facilities use a combination. Large combustion sources may have CEMS for NOx and SO2 but use emission factors for CO2 and PM. Coating operations use material balance for VOC but emission factors for HAPs. Storage tanks use EPA software models. The key is matching the right method to each source based on accuracy requirements, regulatory specifications, and practical constraints. This guide walks through each method so you know when to use what.
Emission Factors: AP-42 and Beyond
EPA's AP-42, Compilation of Air Pollutant Emission Factors, is the most widely used source of emission factors for stationary sources. It has been published since 1972 and contains factors for hundreds of source categories organized into chapters: external combustion (Ch. 1), solid waste disposal (Ch. 2), stationary internal combustion (Ch. 3), evaporative losses (Ch. 4-7), mineral products (Ch. 8), food and agriculture (Ch. 9), wood products (Ch. 10), mineral products (Ch. 11), metallurgical (Ch. 12), and miscellaneous (Ch. 13). Each section provides emission factors in units of pollutant mass per unit of activity (lbs per MMBtu, lbs per ton of material processed, etc.).
AP-42 factors are assigned quality ratings from A (excellent) to E (poor) based on the number and quality of source tests used to develop them. An A-rated factor is based on many well-designed tests across multiple facilities and is considered representative. An E-rated factor may be based on a single test of questionable quality. Many commonly used factors carry C or D ratings, meaning they could be off by a factor of two or more for your specific equipment. Always check the rating and consider whether a source test or material balance would give you a more accurate and defensible number.
For greenhouse gases, the primary emission factor source is 40 CFR Part 98, Subpart C, Table C-1 and Table C-2. These factors are specific to the mandatory GHG reporting program and are organized by fuel type: natural gas at 53.06 kg CO2/MMBtu, distillate fuel oil at 73.16, residual fuel oil at 75.10, and dozens of other fuels. For combustion of common fuels, these factors are well-established and highly accurate because the chemistry is straightforward — the carbon content of the fuel determines the CO2 output.
Beyond AP-42, several other emission factor databases exist. The EPA WebFIRE database provides emission factors from source tests submitted to EPA and is searchable by source category and SCC code. The FIRE (Factor Information Retrieval) database is the older version and is being phased out. State agencies sometimes publish state-specific factors that may be required for use in their jurisdiction. Manufacturer data, equipment specifications, and engine certifications can also serve as emission factors for specific units, particularly for engines and turbines with certified emission rates.
1.4 — Natural gas combustion (boilers, heaters)
1.3 — Fuel oil combustion
3.2 — Heavy-duty diesel engines (generators)
3.4 — Large stationary gas turbines
4.3 — Waste/wastewater collection
7.1 — Organic liquid storage tanks
13.2.1 — Paved roads (fugitive dust)
13.2.2 — Unpaved roads (fugitive dust)
Always use the most recent version at EPA.gov/air-emissions-factors-and-quantification.
Material Balance: The Most Accurate Method You Already Have Data For
Material balance is based on the conservation of mass: what goes in must come out. For emissions estimation, this means tracking the mass of a pollutant-containing material into your process and subtracting whatever leaves in the product, is captured by controls, or goes out with waste. The difference is your emission. For many process sources, particularly coating, cleaning, printing, and adhesive operations, material balance is more accurate than emission factors because it reflects your actual usage rather than an industry average.
The classic material balance for VOC from coating operations works like this: multiply the volume of each coating used (gallons) by its density (lbs/gallon) and its VOC weight fraction (from the Safety Data Sheet or technical data sheet). Sum across all coatings. If none of the VOC is captured by controls or incorporated into the product, the total is your VOC emission. For a shop using 200 gallons per month of a coating at 8.5 lbs/gallon with 45% VOC by weight: 200 × 8.5 × 0.45 = 765 lbs VOC per month, or 4.6 tons per year.
Material balance can also be applied to cleaning solvents, adhesives, inks, and any other material where the volatile component evaporates during use. Track purchases, subtract inventory changes (material on hand at end of period minus beginning of period), subtract any material sent to waste disposal with documentation, and the remainder is your emission. This method is preferred by most state agencies for VOC and HAP emissions from surface coating and solvent-using operations because it directly measures your facility's consumption.
The main limitation of material balance is that it requires good records. You need accurate purchasing data (invoices, delivery records), inventory tracking (monthly physical counts or automated systems), waste manifests for any VOC-containing waste sent off-site, and current SDS or technical data sheets with VOC content. If your purchasing records are sloppy or your SDS files are outdated, the material balance will be unreliable. Establishing a monthly material tracking system costs almost nothing and provides the data you need for both emissions reporting and cost control. Many facilities discover that simply tracking solvent usage monthly reduces consumption by 10-20% because it makes waste visible.
1. Beginning inventory of each VOC-containing material (gallons)
2. + Purchases received during month (gallons, from invoices)
3. − Ending inventory (gallons, from physical count)
4. = Total usage (gallons)
5. × Density (lbs/gal, from SDS)
6. × VOC weight fraction (from SDS or TDS)
7. = Monthly VOC emissions (lbs)
Keep this log for each material. Totals feed directly into your annual emissions inventory.
Facility Emissions Inventory Calculator
Aggregate emissions from multiple facility sources into totals per pollutant. Compare against Title V and PSD thresholds with dominant source identification.
Continuous Emissions Monitoring: When the Permit Demands Real Data
Continuous Emissions Monitoring Systems (CEMS) measure pollutant concentrations and flow rates in real time from a stack or vent. They provide the most accurate emissions data available and are the gold standard for compliance demonstration. CEMS are typically required for large combustion sources (boilers above 250 MMBtu/hr heat input), electric generating units, cement kilns, incinerators, and other major sources where ongoing measurement is needed to ensure continuous compliance with emission limits.
A typical CEMS installation measures SO2 (by UV fluorescence or NDIR), NOx (by chemiluminescence), CO (by NDIR or GFC), O2 or CO2 (by paramagnetic or NDIR), and stack flow (by ultrasonic or differential pressure). The raw data is converted to mass emission rates (lbs/hr) using the measured concentration and flow, and then rolled up to daily, monthly, and annual totals. Data availability requirements are strict — 40 CFR Part 75 requires 90% or higher data capture rates for affected units, with substitute data procedures for any gaps.
The cost of CEMS is substantial. A multi-pollutant system (NOx, SO2, CO, O2, flow) typically costs $150,000-$400,000 to purchase and install, with annual operating costs of $30,000-$80,000 for calibration gases, maintenance, and Relative Accuracy Test Audits (RATAs). RATAs, which involve parallel stack testing to verify the CEMS against reference methods, are required annually and cost $10,000-$25,000 each. For facilities with multiple stacks, the costs multiply. This is why CEMS are reserved for large sources where the emission rates justify the investment.
If your permit does not require CEMS but you want better data than emission factors provide, predictive emissions monitoring systems (PEMS) are an alternative. PEMS use process parameters (fuel flow, steam flow, combustion temperature, O2 levels) in a mathematical model to predict emissions without directly measuring them in the stack. PEMS cost a fraction of CEMS to install and maintain, but they require initial correlation testing against reference method stack tests and periodic recalibration. Some states accept PEMS as a compliance tool for certain source categories where CEMS would be disproportionately expensive.
AP-42 Emission Factor Lookup
Browse EPA AP-42 emission factors for combustion sources, engines, turbines, and crushers. Enter capacity to calculate annual emissions with full section references.
Stack Testing: Source-Specific Data That Holds Up
Stack testing (also called source testing or performance testing) involves physically sampling the flue gas from a stack or vent using EPA reference methods and analyzing the samples in a laboratory. It provides emission data specific to your equipment, your fuel, your operating conditions, and your control devices. Stack test results are considered more accurate than emission factors for the specific unit tested and are often used to develop site-specific emission factors that replace generic AP-42 values.
EPA has published over 30 reference test methods (40 CFR Part 60, Appendix A) for measuring different pollutants. Method 5 measures particulate matter by drawing a stack gas sample through a heated filter and weighing the collected material. Method 6C measures SO2 by instrumental analysis. Method 7E measures NOx. Method 10 measures CO. Method 25A measures total gaseous organic compounds. Method 201A measures PM10 and PM2.5. Each method has specific sampling procedures, equipment requirements, and quality assurance criteria that must be followed exactly for the results to be valid.
A typical stack test costs $5,000-$20,000 per source per test event, depending on the pollutants measured, the number of test runs (usually three), and the complexity of the source. Tests on large stacks with multiple pollutants or difficult access can cost more. Add travel costs if the testing company is not local. Despite the cost, stack testing can be a smart investment when a source is close to a regulatory threshold. If AP-42 says your boiler emits 10.5 tons per year of NOx and the major source threshold is 10 tons, a $8,000 stack test that shows actual emissions are 7.2 tons could save you the entire cost and hassle of Title V permitting.
Stack tests must be conducted under representative operating conditions, which is defined in your permit or by agreement with the regulatory agency. Testing at maximum capacity establishes a maximum emission rate for permit compliance. Testing at normal operating load establishes a rate that better represents annual emissions for inventory purposes. Some permits require testing at multiple loads. Always coordinate with your state agency before conducting a test to agree on the test plan, operating conditions, and test methods. Test results that do not follow the approved protocol may be rejected.
• Your AP-42 PTE is near a major source threshold (test may prove you are below)
• You installed controls and need to demonstrate control efficiency for a permit limit
• Your emission fees are based on emission factors that overestimate your actual rates
• A permit limit is based on an old test and you have since improved the process
• You need to resolve a compliance dispute with measured data
Request proposals from at least two testing firms. Check their state certifications and EPA method experience.
Building the Annual Emissions Inventory
The annual emissions inventory brings together all of your estimation methods into a single comprehensive report of facility-wide emissions. Most states require an annual emissions inventory from permitted sources, submitted within 90 to 180 days after the end of the reporting year. The inventory lists every emission unit, the pollutants emitted, the estimation method used, the key operating parameters (fuel usage, production throughput, hours of operation), and the resulting emissions in tons per year for each pollutant and each unit.
Start with your permit. It lists every emission unit at the facility and the applicable emission limits. For each unit, gather the operating data for the year: fuel consumption from utility records or tank gauging, production throughput from process logs, operating hours from runtime meters or production records, and material usage from purchasing records. Apply the appropriate emission factor, material balance, CEMS data, or stack test result to each unit for each pollutant. If a unit has CEMS data for some pollutants and emission factors for others, use the best available data for each.
Quality-check the inventory by looking for internal consistency. Do the total emissions make sense relative to last year? If emissions jumped 30%, is there a corresponding increase in production or fuel use to explain it? If a unit shows zero emissions, did it really not operate? Cross-check fuel usage against utility bills. Verify that VOC emissions from coatings match the material balance from purchasing records. These internal checks catch data entry errors, unit conversion mistakes, and missing sources before the inventory goes to the regulator.
Keep detailed working files behind the inventory. State agencies and EPA auditors will ask how you calculated each number. Maintain the spreadsheets, fuel records, production logs, SDS files, stack test reports, and CEMS summary data that support every line in the inventory. Organize files by emission unit and year. If you use a software tool (many states offer online reporting systems like SLEIS, EQUIS, or state-specific platforms), keep a local backup of the underlying data. An emissions inventory is only as good as the records behind it, and you may need to defend the numbers years after they were submitted.
January: Begin gathering fuel, production, and material usage records for the previous year
February: Calculate emissions for each unit using applicable methods
March: Cross-check calculations, review for consistency with prior years
April: Compile the inventory in the state's required format and submit
Do not wait until the deadline. Chasing down records and resolving discrepancies always takes longer than expected.