Pile Burn Emissions Calculation Tool
Source-specific inventories of air pollutants are key inputs into air quality models and generating emissions trends. The US Environmental Protection agency produces a comprehensive set of emissions inventories for the National Emissions Inventory (NEI) every three years. In the 2020 NEI wildland fires were the single largest source of primary fine particulate matter (PM2.5) in the United States, emitting over 2.4 tons PM2.5 1. Accurate emissions estimates from wildland fires, including those from both planned and unplanned burns, are therefore critical to characterizing the PM2.5 in the US. The burning of piled woody biomass (e.g. pile burns) is conducted by state, local, federal, and private landowners to dispose of downed limbs, logging debris, and other sources. In the 2020 NEI pile burns (PB) were not differentiated from other burning detected in wildlands such as broadcast burns and wildfires due to lack of methodology and input data. Missing pile burns can cause underpredictions of emissions whereas the misclassification of pile burns may result in overpredictions of emissions as they tend to represent an overall lower total amount of biomass consumption than broadcast burns or wildfires. This guidance describes the use of a script package to estimate the pile burn emissions from disparate input activity data sources.
Across the US prescribed fire activity by area treated is primarily concentrated in the Southeast 2 in the form of broadcast burns. However, in Pacific Northwest states pile burns are often the most common fire-based method for the removal of logging debris and reducing hazardous fuel loads 3. States where pile burning is common, such as Washington, require the estimation of piled biomass as part of their permitting process 4. To aid in this estimation process the USDA Forest Service developed a pile estimate tool (https://depts.washington.edu/nwfire/piles/) based on the hand pile size and emissions estimate methods in Wright 2009 5 and the machine pile estimate methods in Hardy 1998 6. These same equations and methods for estimating piled mass and the resulting emissions from a burn are implemented in the CONSUME 7 fuel consumption and emissions tool found in wildland smoke models such as the BlueSky modeling framework 8. While the implementation of the pile burn equations in these tools are useful for meeting permit requirements, particularly when detailed data is available from the burner, they do not meet the operational needs of estimating pile burn emissions at the type of broad national level required in the NEI.
The development of the NEI is a collaborative process that utilizes both EPA generated data and domain expertise from federal, state, local, and tribal agencies (SLT). As most pile burns are missed by satellite instrumentation the EPA relies on operations and permit data from these agencies to create an emissions inventory for pile burns 9. The information recorded in the federal and SLT data varies based on the requirements of each agency, resulting in the need to use parameters not directly available in the USDA Forest Service pile estimate tool or CONSUME. The pile burn emissions tool described in the document addresses this gap by providing methods to estimate biomass consumed from multiple forms of pile burn activity including pile mass, pile geometry, pile volume, and treated area.
The method of estimating total pile mass from geometry follows those applied in the USDA Forest Service pile tool using the equations in Wright 2009 and Hardy 1998. At this time only the equations for half sphere, paraboloid, half-cylinder, and half-ellipsoid piles are implemented in this tool. Alternatively, the pile volume can be supplied in place of supplying a geometry type and the corresponding parameters. To estimate pile mass in hand piles the equations in Wright 2009 require biomass density. The process of retrieving the biomass density is operationalized using the Fuel Characteristic Classification System (FCCS) 10 of the location of the pile. The fractions of the downed species within the FCCS fuel bed are combined with the wood species density in Miles 2009 11 to estimate an effective pile biomass density for the Wright 2009 equations.
Many agencies do not require specific details of pile configuration or mass and therefore only supply area. This tool can accommodate area estimates as activity provided a piled mass per unit area loading value. However, loading values associated a pile burn operation may not be available for all burns or from all agencies. In these cases, a default loading value [Fig. 1] of 4.5 tons per acre was developed based on median piled mass per area treated reported in the 2020 and 2022 Washington permit database (WA CITE) and the 2017-2020 California PFIRS data (CITE).
Mass consumed is estimated from the piled mass available using a consumption fraction. Piled mass available can be supplied directly to the tool or calculated from one of the methods described in the paragraphs. The consumption fraction varies based on fuel and burn conditions. A default value of 0.9 is supplied in the tool based on reference values from Hardy 1996 and Wright 2019 12.
Emissions estimates are calculated from the product of the mass consumed and a pollutant-specific emission factor [Show equation]. This tool includes two sets of emission factors, one from the USDA Forest Service Smoke Emissions Reference Application (SERA) database 13 and a second set consistent with those found in the USDA Forest Service pile burn tool and CONSUME from Hardy 1998. The SERA emission factors are averaged across pile burn emissions measurement studies by combustion phase (flaming and smoldering) where sufficient data points are available or as composite values across studies (Table S1). The Hardy 1998 emission factors are more limited than SERA in the available pollutants, notably lacking an emission factor for nitrogen oxides (NO¬x), and require additional parameterization on the quality of the pile that is typically not available in bulk state and federal datasets.
https://depts.washington.edu/nwfire/sera/ Download timestamp (UTC): 2023-11-27 15:10:20Filters applied: Burn Types = Pile burn (Field); Pile burn (Lab) | Slash included
The pile burn emissions calculator (pilecalc) is a Python script utilizing the Numpy and Pandas packages. While the calculator may be compatible with multiple versions of Python and the related packages, it was tested with Python 3.9.16, Numpy 1.26.4, and Pandas 2.2.2. The pile_to_ff10 utility script aggregates pilecalc outputs into a set of SMOKE-compatible FF10 inventories (https://www.cmascenter.org/smoke/documentation/5.0/html/ch06s02s08.html).
A JSON formatted file is used to configure the pilecalc scenario (Fig. 1). This file defines the set of emission factors to apply the consumption and default values used to fill required variables. Valid emission factor options are “sera” and “hardy” for the SERA and Hardy emission factors, respectively.
Figure 1. Example configuration file entries for pilecalc
Default values are used when a variable required for the consumption and emissions calculations is not available from the input activity. A description of the default variables is shown in Table 2. Additional variables from Table 3 can be added to the default section of the JSON file to override pilecalc defaults.
Table 2. Default value descriptions for the pilecalc configuration.
Variable Description Valid Options ceff Consumption fraction of the pile burn 0-1 [0.9 Recommended] tons_acre Woody piled material loading (tons) per area (acre) Value > 0 method Method used to construct the pile, used for estimating loading based on pile geometry hand or machine fccs FCCS Number for regional default land cover type, used for estimating loading from geometry Any valid FCCS number
Activity is supplied to pilecalc in the form of a comma-delimited text file. Each activity record requires the location of the pile burns, the burn date, and information on the fuel loading or consumption. The pilecalc tool can estimate consumption from five distinct sets of loading or activity parameters. At a minimum one form of activity is required for each record. If multiple forms of activity are specified for a single record, then the loading and consumption are determined based on a hierarchy of run types (from highest and lowest rank): Piled Biomass Consumed, Piled Biomass Available, Pile Volume, Pile Geometry, and Treated Area. The required columns for each run type are specified in Table 3. All input columns are passed through to the pilecalc output regardless of their use in any loading or consumption calculations.
Table 3. Columns for pilecalc activity input file Column Name Description Run Type Requirement Valid Values Default Value Latitude Latitude of piles All -90. to 90. N/A Longitude Longitude of piles All -180. to 180. N/A Date Date of burn All Any date as YYYYMMDD N/A consumed_tons Total biomass consumed for all piles at location on burn date (tons) Piled Biomass Consumed Number > 0 N/A piletons Total available biomass for all piles at location on burn date (tons) Piled Biomass Available Number > 0 N/A pileheight Height of average pile (ft) Pile Geometry Number > 0 N/A Pilewidth Width of average pile (ft) Certain Pile Geometries Number > 0 N/A pilelength Length of average pile (ft) Certain Pile Geometries Number > 0 N/A Pilevol Average pile volume (cubic ft) Pile Volume Number > 0 N/A geompiles Number of piles at burned at the location on the burn date Pile Geometry and Volume Integer > 0 1 fccs FCCS number associated with the piled material Pile Geometry and Volume Any valid FCCS 0 geometry Geometry of piles Pile Geometry spheroid, paraboloid, ellipsoid, rectangle, cylinder paraboloid pilemethod Method of piling Pile Geometry and Volume machine, hand machine packing Pile packing ratio Pile Geometry and Volume 0 - 1 0.2 dirt Fraction of pile volume that is dirt Pile Geometry and Volume 0 - 1 0.1 acres Total area from where the piled material was gathered (acres) Treated Area Number > 0 N/A tonsacre Mass of piled material per acre (tons/acre) Treated Area Number > 0 4.5 ceff Consumption fraction for the pile burns All 0 - 1 0.9 flaming Fraction of fuel consumed in the flaming phase All 0 - 1 0.7 smoldering Fraction of fuel consumed in the smoldering phase All 0 - 1 0.15 residual Fraction of fuel consumed in the residual smoldering phase All 0 - 1 0.15
Sample input file snippets for each of the five run types are provided in the appendix (Fig. A1 - A5).
The pilecalc tool can be run from the command line by specifying 2 – 3 arguments: path to the activity, path to write the output, and the JSON configuration file (defaults to pile_config.json). The script contains checks to verify that all necessary columns are specified in the input file. A successful run of the tool will result in a comma-delimited output file containing the fields from the activity input file, the parameters applied in the calculation, and calculated values. Input columns may vary depending on the source. All parameters and calculated values are specified in Table 4.
Table 4. Output file parameter and calculated value columns
Column Name Description latitude Latitude from activity longitude Longitude from activity date Date from activity method Name of applied run type/activity method plumepiles Number of piles assigned for plume rise calculation flaming Flaming phase combustion fraction residual Residual smoldering phase combustion fraction smoldering Smoldering phase combustion fraction ceff Consumption fraction dirt Dirt fraction tonsacre Tons per acre of loading piletons Total tons of piled biomass consumed_tons Total tons consumed flaming_tons Total tons consumed in flaming phase smoldering_tons Total tons consumed in smoldering phase residual_tons Total tons consumed in residual phase Pollutants (e.g. CO, NOX, etc.) Columns of emissions (tons) for individual pollutants (currently CAPs only) heat Heat output (btu) for plume rise calculation plumeacres Average pile surface area (acres) for plume rise calculation
The method, parameters, and emissions estimates should be spot-checked for quality assurance. The number of records in should match the number the number of records out. Selection of an incorrect method may be the result of improperly defined column names and data types. Individual pile burns tend to result in relatively low emissions by wildland fire standards (< 1 ton PM2.5).
Runtime errors commonly occur when required input columns are missing, data is incorrectly formatted, or command line arguments are missing.
A utility script, pile_to_ff10, is used to generate a flat file (FF10) formatted inventory for SMOKE from one or more pilecalc output files. A comma-delimited list file of pilecalc outputs contains the path, data source, and fire name field associated with each respective output (Fig. 2). The data source field is used as metadata within the FF10 to differentiate attribution of activity. The fire name field is a field from the input data that defines the name of individual pile burn operations. Those burns without a fire name field are assigned an automatically generated default name.
Figure 2. Example pile_to_ff10 list file format
The name of the list file and the label of the output annual and daily FF10 inventories are defined at the top of the pile_to_ff10 utility as the lst and label variables. The TIGER county boundaries shapefile (https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html) is used with the OSGeo OGR Python package to assign county FIPS codes to each pile burn location. The county boundary shapefile should be updated as necessary to include any updates to county designations. Burns that fall outside the boundaries of any county are written to a file with the defined label followed by the suffix “_oob”. This out-of-bounds file should be scanned for properly formatted latitudes and longitude. The output annual and daily FF10 inventories should be checked against the input data to confirm that the number of unique fires matches and that the total emissions match between the inputs and outputs.
Figure A1. Sample input for the Piled Biomass Consumed run type. The latitude, longitude, date, and consumed_tons columns are used as inputs to the pile emissions calculation. Consumption fraction was set in the ceff column but not used by pilecalc because the consumed mass was provided in this dataset.
Figure A2. Sample input for the Piled Biomass Available run type. The latitude, longitude, date, and piletons columns are used as inputs to the pile emissions calculation. The consumption fraction, ceff, used the default value because it is not specified in this dataset.
Figure A3. Sample input for Pile Volume run type. The latitude, longitude, date, pilevol, pilemethod, and fccs columns are used as inputs to the pile emissions calculation. The number of piles (geompiles) is not specified and filled with the default value. Columns indicating pile geometry parameters (pilelength, pilewidth, pileheight) are supplied but not used because the volume is higher in the run type hierarchy.
Figure A4. Sample input for Pile Geometry run type. The latitude, longitude, date, geompiles, pilewidth, pileheight, and fccs columns are used as inputs to the pile emissions calculation. The geometry type, geometry, was not provided and filled with the default configuration.
Pollutant Combustion Phase EF (lb/ton) Source CO2 flaming 3,485.856 SERA Flaming Pile-Field CO flaming 81.73 SERA Flaming Pile-Field PM2.5 flaming 10.988 SERA Flaming Pile-Field PM10 flaming 12.967 SERA PM2.5 Scaled by Hardy ratio of PM10:PM2.5 VOC flaming 1.426 SERA Flaming Pile-Field NMHC CO2 smoldering 3,095.082 SERA Smoldering Pile-Field CO smoldering 249.378 SERA Smoldering Pile-Field PM2.5 smoldering 31.446 SERA Smoldering Pile-Field PM10 smoldering 37.106 SERA PM2.5 Scaled by Hardy ratio of PM10:PM2.5 VOC smoldering 3.402 SERA Smoldering Pile-Field NMHC NOX 4.522 SERA All Pile-Field and Lab SO2 0.224 SERA All Pile-Field and Lab NH3 0.901 SERA Broadcast + Piles
Figure A5. Sample input for Treated Area run type. The latitude, longitude, date, and acres columns are used as inputs to the pile emissions calculation. The loading, tonsacre, and consumption fraction, ceff, columns were not provided and therefore filled with the default configuration.
Pollutant Clean Dirty Really Dirty PM 21.9 27 36 PM10 15.5 20 28 PM2.5 13.5 17 23.6
Pollutant Flaming Smoldering Residual CO 52.66 130.37 130.37 CO2 3429.24 3089.88 3089.88 CH4 3.28 11.03 11.03 NMHC 3.56 6.78 6.78
(1) US EPA. 2020 National Emissions Inventory Technical Support Document: Fires – Wild, Prescribed, and Agricultural Field Burning. 2023. https://www.epa.gov/system/files/documents/2023-03/NEI2020_TSD_Section7_Fires.pdf (accessed Apr. 9 2025). (2) Melvin, M. A. 2021 National Prescribed Fire Use Survey Report; Coalition of Prescribed Fire Councils, Inc., 2021. https://www.prescribedfire.net/pdf/2021-National-Rx-Fire-Use-Report_FINAL.pdf (accessed 22 July 2024). (3) Oregon Department of Forestry. Oregon Smoke Management Annual Report 2022. https://www.oregon.gov/odf/Documents/fire/smr2022.pdf (accessed Apr. 9 2025). (4) Washington Department of Natural Resources. Burn Permits. 2025. https://www.dnr.wa.gov/programs-and-services/wildfire/outdoor-burning/burn-permits (accessed Apr. 9 2025). (5) Wright, C. S.; Balog, C. S.; Kelly, J. W. Estimating volume, biomass, and potential emissions of hand-piled fuels; U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, 2009. http://dx.doi.org/10.2737/PNW-GTR-805DOI: 10.2737/pnw-gtr-805. (6) Hardy, C. C. Guidelines for estimating volume, biomass, and smoke production for piled slash; U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, 1998. http://dx.doi.org/10.2737/PNW-GTR-364DOI: 10.2737/pnw-gtr-364. (7) Prichard, S. J.; Ottmar, R. D.; Anderson, G. K. CONSUME user's guide and scientific documentation; 2007. https://www.frames.gov/documents/usfs/fera/consume30_users_guide.pdf. (8) Larkin, N. K.; O'Neill, S. M.; Solomon, R.; Raffuse, S.; Strand, T.; Sullivan, D. C.; Krull, C.; Rorig, M.; Peterson, J.; Ferguson, S. A. The BlueSky smoke modeling framework. Int J Wildland Fire 2009, 18 (8), 906-920. DOI: https://doi.org/10.1071/WF07086. (9) Beidler, J.; Vukovich, J.; Pouliot, G.; Rao, V. Development of Emissions from the Burning of Piled Woody Biomass in the 2022 Emissions Modeling Platform. CMAS, Chapel Hill, NC, 2024. https://gaftp.epa.gov/Air/nei/2023/doc/supporting_data/events/draft/Pile%20Burn%20Draft%20Method%202022v1%20CMAS.pdf (10) Prichard, S. J.; Sandberg, D. V.; Ottmar, R. D.; Eberhardt, E.; Andreu, A.; Eagle, P.; Swedin, K. Fuel Characteristic Classification System version 3.0: technical documentation; U.S. Department of Agriculture, Forest Service, Pacific Northwest Research Station, 2013. http://dx.doi.org/10.2737/PNW-GTR-887DOI: 10.2737/pnw-gtr-887. (11) Miles, P. D. Specific gravity and other properties of wood and bark for 156 tree species found in North America; US Department of Agriculture, Forest Service, Northern Research Station, 2009. (12) Wright, C. S.; Evans, A. M.; Grove, S.; Haubensak, K. A. Pile age and burn season influence fuelbed properties, combustion dynamics, fuel consumption, and charcoal formation when burning hand piles. Forest Ecol Manag 2019, 439, 146-158. (13) Prichard, S. J.; O'Neill, S. M.; Eagle, P.; Andreu, A. G.; Drye, B.; Dubowy, J.; Urbanski, S.; Strand, T. M. Wildland fire emission factors in North America: synthesis of existing data, measurement needs and management applications. Int J Wildland Fire 2020, 29 (2), 132-147. DOI: https://doi.org/10.1071/WF19066.
