14.2 EPA-developed estimates

Structure fires include fires from residential and commercial structures as well as mobile homes. Mobile vehicle fires cover all fires from onroad motor vehicles, including campers, recreational vehicles, and motorcycles, as well as nonroad equipment, excluding aircraft, marine vessels, and locomotives. To estimate emissions from both source categories, first the average amount of structure/vehicle and content material burned are estimated. These estimates are year-specific and based on fire incident reports from the annual National Fire Incident Reporting System (NFIRS) public release [ref 4]. NFIRS is released by U.S. Fire Administration (USFA), under the Federal Emergency Management Agency (FEMA). Next, fire department incidents for each fire type are assigned a valid state-county FIPS and to create a total county count of each incident type. Fuel loading and median emission factors for each fire type are based on a recent study [ref 1]. Annual county level estimates are computed separately for each fire type by multiplying the county total incidents, average fuel loading, and emission factors.

14.2.1 Activity Data

Activity data for both structure fires and motor vehicle fires are the number of fire incidents in each county, provided by the annual release of the NFIRS public release of fire incident information. Included in these data are uniform format basic incident reports by fire station identifier (FDID) for all types of fires, as reported by fire departments. The basic incidents report includes counts of specific types of structure fires and motor vehicle fires. The following Incident Types, as defined in Chapter 3, Section C of the “NFIRS_Complete_Reference_Guide_2015.pdf” are included in the incident counts for structure fires and motor vehicle fires, as defined here in Table 14.2.

Table 14.2: Incident types included in the methodology
Fire Type Incident Type Code Incident Type Code Description
Structure 111 Building fire. Excludes confined fires (113–118).
Structure 112 Fire in structure, other than in a building. Included are fires on or in piers, quays, or pilings: tunnels or underground connecting structures; bridges, trestles, or overhead elevated structures; transformers, power or utility vaults or equipment; fences; and tents.
Structure 120 Fire in mobile property used as a fixed structure, other.
Structure 121 Fire in mobile home used as a fixed residence. Includes mobile homes when not in transit and used as a structure for residential purposes; and manufactured homes built on a permanent chassis.
Structure 122 Fire in a motor home, camper, or recreational vehicle when used as a structure. Includes motor homes when not in transit and used as a structure for residential purposes.
Structure 123 Fire in a portable building, when used at a fixed location. Includes portable buildings used for commerce, industry, or education and trailers used for commercial purposes.
Motor Vehicle 130 Mobile property (vehicle) fire, other.
Motor Vehicle 131 Passenger vehicle fire. Includes any motorized passenger vehicle, other than a motor home (136) (e.g., pickup trucks, sport utility vehicles, buses).
Motor Vehicle 132 Road freight or transport vehicle fire. Includes commercial freight hauling vehicles and contractor vans or trucks. Examples are moving trucks, plumber vans, and delivery trucks.
Motor Vehicle 136 Self-propelled motor home or recreational vehicle. Includes only self-propelled motor homes or recreational vehicles when being used in transport mode. Excludes those used for normal residential use (122).
Motor Vehicle 137 Camper or recreational vehicle (RV) fire, not self-propelled. Includes trailers. Excludes RVs on blocks or used regularly as a fixed building (122) and the vehicle towing the camper or RV or the campers mounted on pickups (131).
Motor Vehicle 138 Off-road vehicle or heavy equipment fire. Includes dirt bikes, specialty off-road vehicles, earth-moving equipment (bulldozers), and farm equipment.

EPA does not include contained fires (incident codes 113 – 118, e.g., 118: Trash or rubbish fire in a structure, with no flame damage to structure or its contents). File formats are defined in the included in the NFIRS file “Public Data Release Format v1_8.pdf” included in every annual package download.

NFIRS does not provide valid state-county FIPS for many FDIDs in the basic incident report. There is a hierarchy used to assign valid state-county FIPS for these FDIDs. First, an associated FDID cross reference file (fdheader) provides some matches. Second, for FDIDs with a valid zip code, FDIDs are apportioned to associated FIPS based on proportion of population in different counties for each zip code. As a last resort, all remaining FDIDs without valid FIPS and zip codes are assigned FIPS based on a manual search by fire department name and search engine and mapping results.

14.2.2 Fuel Loading

For both structure fires and motor vehicle fires, fuel loading estimates are needed to associate the number of fires to units compatible with emission factors, which are based on the weight of the material burned.

14.2.2.1 Structure Fires

The fuel loading factor for structure fires is based on an estimate of the combustible structural content for an average-sized residence. Structural content includes both structural mass and building contents:

Fuel loading = structure loss + Content loss

In 1999, the average size of existing homes was assumed to be 1,649 square feet with a corresponding average of 11,000 board feet of lumber; an assumption of one ton of material per 1,000 board feet therefore yields approximately 11 tons of combustible material per structure fire in 1999 [ref 2].

An average loss percentage of 7.3 percent for each fire incident is assumed from the 1999 CARB report [ref 2]. Note that this is far less than the 80% loss assumption from the 2022 WUI study [ref 1] as those fires are based on wildland fires impacting residential and commercial buildings. This loss rate percentage should be adjusted locally based on more information if available. As shown in equation (14.1), applying this 7.3 percent loss rate leads to an estimated structure loss component of fuel load of 0.80 tons per fire.

\[\begin{equation} \text{Structure Loss (tons per fire)} = \text{Combustible Material (tons)} \times \frac{\text{Loss Rate}}{100} \tag{14.1} \end{equation}\]

Next, weighted average combustible contents within the structure are also estimated as most fires start in the kitchen or a bedroom. Applying the same 7.3 percent loss rate to weighted average of 7.91 pounds per square foot of combustible contents yields an additional 0.48 tons of building contents per fire as shown in equation (14.2).

\[\begin{equation} \text{Content Loss (tons per fire)} = SF \times \frac{WACC}{2,000} \times \frac {\text{Loss Rate}}{100} \tag{14.2} \end{equation}\]

Where:
\(SF\) = Average housing square footage = 1,649 in 1999
\(WACC\) = Weighted average combustible contents = 7.91 lbs per square foot
\(\text{Loss Rate}\) = Average loss rate for structure fires in percent = 7.3

Cumulatively, the fuel load is therefore assumed to be 1.28 tons (0.80 tons from structure loss and 0.48 tons from content loss) per fire based on 1999 assumptions. However, based on the 2023 study [ref 1], we have scaled up fuel loads for both structure and content losses based on the average sized house increasing from 1,649 to 2,150, yielding 1.67 tons per structure fire:

\[\begin{equation} FL_{y} = \text{(Structure Loss + Content Loss)} \times \frac{SF_{y}}{SF_{1999}} \tag{14.3} \end{equation}\]

Where:
\(FL_{y}\) = Fuel load in inventory year y in tons per fire
\(\text{Structure Loss}\) = Average Structure Loss in 1999 = 0.80 tons
\(\text{Content Loss}\) = Average Content Loss in 1999 = 0.48 tons
\(SF_{y}\) = Average housing square footage in current inventory year y (2,150 square feet in 2023)
\(SF_{1999}\) = Average housing square footage in 1999 = 1,649 square feet

14.2.2.2 Motor Vehicle Fires

The fuel loading for motor vehicle fires is based on the 2023 study of wildland urban interface fires [ref 1]. Unlike structure fires, there is no assumption of reduced loss rate for motor vehicle fires that typically spread rapidly to consume much of the combustible composition of the vehicle. Appendix Table S6 aggregates the sum of combustible materials from plastics, coatings, textiles, fluids and lubricants, rubber (tire), fuel, and other non-metal combustible components from a typical year 2017 mid-sized car (4,032 lbs). The total combustible materials, and thus fuel loading was computed as 1,016 lbs, or 0.508 tons.

14.2.3 Allocation Procedure

Emissions are estimated for each fire station identifier (FDID) which, after preprocessing discussed above, are associated with specific state-county(s) FIPS.

14.2.4 Emission Factors

Emission factors for both structure fires and motor vehicle fires are based on the 2023 study [ref 1], which is based on a compilation from a comprehensive literature search of fire simulations of structures, vehicles, and their components. Numerous caveats were used to filter out outliers and those studies that only provided concentration data.

From the comprehensive set of emission factors, a summary of emission factor dataset was developed by taking the median of all observations for all structure materials, regardless of scale of combustion conditions was used. For motor vehicle fires, the median emission factors from the full-scale tests were used and where full-scale vehicle tests were not available, the median emission factor from the entire dataset for any vehicle material was used.

The EPA Wagon Wheel estimate emissions for all CAPs, acid gases, and HAP-Metals; all HAP VOCs and HAP-VOC/PAH pollutants are estimated in EIS via HAP Augmentation multiplication factors that are simply the emission factor for the HAP-VOC divided by VOC emission factor. All PM species (EC, OC, NO3, SO4, and PMFINE) are estimated in EIS via PM speciation. Structure fire emission factors are provided in Table 14.3.

Table 14.3: Structure fire emission factors (lb/ton)
Pollutant Code Pollutant Name Pollutant Group Emission Factor Estimated in WW? Reference
CO Carbon Monoxide 138.00000 Y Ref. 1; Table S2
NH3 Ammonia 1.64400 Y Ref. 1; Table S2
NOX Nitrogen Oxides 0.66200 Y Ref. 1; Table S2
PM10-PRI Primary PM10 78.60000 Y Ref. 1; Table S2
PM25-PRI Primary PM25 78.60000 Y Ref. 1; Table S2
SO2 Sulfur dioxide 0.12500 Y Ref. 1; Table S2
VOC Volatile Organic Compounds 116.40000 Y Ref. 1; Table S2
74908 Hydrogen Cyanide Acid-Gases 3.00000 Y Ref. 1; Table S2
7647010 Hydrochloric Acid Acid-Gases 22.00000 Y Ref. 1; Table S2
7782505 Chlorine Acid-Gases 7.12000 Y Ref. 1; Table S2
7439921 Lead HAP-Metal 0.02200 Y Ref. 1; Table S2
7440360 Antimony HAP-Metal 0.00080 Y Ref. 1; Table S2
7783064 Hydrogen Sulfide 4.38000 Y Ref. 1; Table S2
50000 Formaldehyde HAP-VOC 0.49800 N Ref. 1; Table S2
71432 Benzene HAP-VOC 1.13600 N Ref. 1; Table S2
75070 Acetaldehyde HAP-VOC 0.38000 N Ref. 1; Table S2
91203 Naphthalene HAP-VOC 0.22000 N Ref. 1; Table S2
92524 Biphenyl HAP-VOC 0.06840 N Ref. 1; Table S2
100414 Ethylbenzene HAP-VOC 0.11200 N Ref. 1; Table S2
100425 Styrene HAP-VOC 0.18620 N Ref. 1; Table S2
107028 Acrolein HAP-VOC 0.02700 N Ref. 1; Table S2
108883 Toluene HAP-VOC 0.23000 N Ref. 1; Table S2
108907 Chlorobenzene HAP-VOC 0.59000 N Ref. 1; Table S2
108952 Phenol HAP-VOC 0.07420 N Ref. 1; Table S2
50328 Benzo(a)pyrene HAP-VOC/PAH 0.01432 N Ref. 1; Table S2
53703 Dibenzo(a,h)anthracene HAP-VOC/PAH 0.00180 N Ref. 1; Table S2
56553 Benz(a)anthracene HAP-VOC/PAH 0.01500 N Ref. 1; Table S2
83329 Acenaphthene HAP-VOC/PAH 0.00210 N Ref. 1; Table S2
85018 Phenanthrene HAP-VOC/PAH 0.18000 N Ref. 1; Table S2
86737 Fluorene HAP-VOC/PAH 0.03000 N Ref. 1; Table S2
120127 Anthracene HAP-VOC/PAH 0.02180 N Ref. 1; Table S2
129000 Pyrene HAP-VOC/PAH 0.07400 N Ref. 1; Table S2
191242 Benzo(g,h,i,)perylene HAP-VOC/PAH 0.01900 N Ref. 1; Table S2
192972 Benzo(e)pyrene HAP-VOC/PAH 0.01280 N Ref. 1; Table S2
193395 Indeno(1,2,3-c,d)pyrene HAP-VOC/PAH 0.02000 N Ref. 1; Table S2
198550 Perylene HAP-VOC/PAH 0.00280 N Ref. 1; Table S2
205992 Benzo(b)fluoranthene HAP-VOC/PAH 0.02600 N Ref. 1; Table S2
206440 Fluoranthene HAP-VOC/PAH 0.07400 N Ref. 1; Table S2
207089 Benzo(k)fluoranthene HAP-VOC/PAH 0.00780 N Ref. 1; Table S2
208968 Acenaphthylene HAP-VOC/PAH 0.17520 N Ref. 1; Table S2
218019 Chrysene HAP-VOC/PAH 0.02000 N Ref. 1; Table S2
130498292 PAH, total HAP-VOC/PAH 0.13540 N Ref. 1; Table S2
EC Elemental Carbon 15.36000 N Ref. 1; Table S2
OC Organic Carbon 3.52000 N Ref. 1; Table S2

Several errors for motor vehicle emission factors were found in Supplementary data Table S-2 [ref 1], and these corrections [ref 5] are reflected in the table of emission factors included here in Table 14.4. Note, Chromium VI (18540299) and Chromium III (16065831) are computed as 34 percent and 66 percent, respectively, of total Chromium.

Table 14.4: Motor vehicle fire emission factors (lb/ton)
Pollutant Code Pollutant Name Pollutant Group Emission Factor Estimated in WW? Reference
CO Carbon Monoxide 96.000000 Y Ref. 1; Table S2
NH3 Ammonia 0.800000 Y Ref. 1; Table S2
NOX Nitrogen Oxides 8.600000 Y Ref. 5
PM10-PRI Primary PM10 114.400000 Y Ref. 1; Table S2
PM25-PRI Primary PM25 114.400000 Y Ref. 1; Table S2
SO2 Sulfur dioxide 3.800000 Y Ref. 1; Table S2
VOC Volatile Organic Compounds 21.700000 Y Ref. 5
74908 Hydrogen Cyanide Acid-Gases 1.800000 Y Ref. 1; Table S2
7647010 Hydrochloric Acid Acid-Gases 12.800000 Y Ref. 1; Table S2
7664393 Hydrogen Fluoride Acid-Gases 3.000000 Y Ref. 1; Table S2
7782505 Chlorine Acid-Gases 5.000000 Y Ref. 1; Table S2
7439921 Lead HAP-Metal 0.065000 Y Ref. 1; Table S2
7439965 Manganese HAP-Metal 0.011400 Y Ref. 1; Table S2
7440020 Nickel HAP-Metal 0.005600 Y Ref. 1; Table S2
7440360 Antimony HAP-Metal 0.206000 Y Ref. 1; Table S2
7440382 Arsenic HAP-Metal 0.000520 Y Ref. 1; Table S2
7440439 Cadmium HAP-Metal 0.000990 Y Ref. 1; Table S2
18540299 Chromium VI HAP-Metal 0.003196 Y Ref. 1; Table S2*
16065831 Chromium III HAP-Metal 0.006204 Y Ref. 1; Table S2*
7440484 Cobalt HAP-Metal 0.001180 Y Ref. 1; Table S2
7723140 Phosphorus 0.109600 Y Ref. 1; Table S2
50000 Formaldehyde HAP-VOC 0.703600 N Ref. 1; Table S2
71432 Benzene HAP-VOC 3.024300 N Ref. 5
75070 Acetaldehyde HAP-VOC 0.592600 N Ref. 1; Table S2
91203 Naphthalene HAP-VOC 0.591959 N Ref. 5
92524 Biphenyl HAP-VOC 0.200000 N Ref. 1; Table S2
100414 Ethylbenzene HAP-VOC 0.000400 N Ref. 1; Table S2
100425 Styrene HAP-VOC 1.000000 N Ref. 1; Table S2
108883 Toluene HAP-VOC 2.000000 N Ref. 1; Table S2
108952 Phenol HAP-VOC 0.720000 N Ref. 1; Table S2
50328 Benzo(a)pyrene HAP-VOC/PAH 0.005766 N Ref. 5
53703 Dibenzo(a,h)anthracene HAP-VOC/PAH 0.002080 N Ref. 5
56553 Benz(a)anthracene HAP-VOC/PAH 0.017018 N Ref. 5
83329 Acenaphthene HAP-VOC/PAH 0.000903 N Ref. 5
85018 Phenanthrene HAP-VOC/PAH 0.210513 N Ref. 5
86737 Fluorene HAP-VOC/PAH 0.019788 N Ref. 5
120127 Anthracene HAP-VOC/PAH 0.020048 N Ref. 5
129000 Pyrene HAP-VOC/PAH 0.058225 N Ref. 5
191242 Benzo(g,h,i,)perylene HAP-VOC/PAH 0.014468 N Ref. 5
192972 Benzo(e)pyrene HAP-VOC/PAH 0.030000 N Ref. 1; Table S2
193395 Indeno(1,2,3-c,d)pyrene HAP-VOC/PAH 0.019801 N Ref. 5
198550 Perylene HAP-VOC/PAH 0.002000 N Ref. 1; Table S2
205992 Benzo(b)fluoranthene HAP-VOC/PAH 0.035324 N Ref. 5
206440 Fluoranthene HAP-VOC/PAH 0.095584 N Ref. 5
207089 Benzo(k)fluoranthene HAP-VOC/PAH 0.009217 N Ref. 5
208968 Acenaphthylene HAP-VOC/PAH 0.089928 N Ref. 5
218019 Chrysene HAP-VOC/PAH 0.024340 N Ref. 5
130498292 PAH, total HAP-VOC/PAH 2.200000 N Ref. 1; Table S2

14.2.5 Controls

There are no controls assumed for this category.

14.2.6 Emissions

Emissions are quantified using county-level estimates on the number of incident reports, assumed fuel loading per incident, and the emission factors provided above as follows:

\[\begin{equation} E_{t,p,c} = I_{t,c} \times FL_{t} \times \frac{EF_{t,p}}{2,000} \tag{14.4} \end{equation}\]

Where:
\(E_{t,p,c}\) = Estimated annual emissions for fire type (i.e., structure, motor vehicle) f, pollutant p, and county c, in tons
\(I_{t,c}\) = Number of total fire incidents by fire type t in county c, in tons
\(FL_{t}\) = Average fuel load for each fire type incident t, in tons
\(EF_{t,p}\) = Emissions factor for pollutant p and fire type t, in lbs/ton

State and Local data submitters have the option to provide, separately for structure fires vs motor vehicle fires, more local information on the number of fires in each county (incidents), fuel loadings and emission factors.

Please note that there are no point source SCCs that overlap with the processes described here. Therefore, point source subtraction is not performed.

14.2.7 Sample Calculations

Table 14.5 lists sample calculations to determine the benzene emissions for nonpoint source POTWs. The values in these equations are demonstrating program logic and are not representative of any specific NEI year or county.

Table 14.5: Sample Calculations
Eq. # Equation Values Result
1 \(\text{Combustible Material} \times \frac{\text{Loss Rate}}{100}\) \(\text{11 tons} \times \frac{\text{7.3}}{100}\) 0.80 tons Structure Loss per average fire
2 \(SF \times \frac{WACC}{2,000} \times \frac {\text{Loss Rate}}{100}\) \(\text{1,649 sq feet} \times 7.91 \frac{lbs}{\text{sq feet}} \times \frac {1}{2000} \times \frac{\text{7.3}}{100}\) 0.48 tons Content Loss per average fire
3 \(\text{(Structure Loss + Content Loss)} \times \frac{SF_{y}}{SF_{1999}}\) \(\text{(0.80 tons + 0.48 tons)} \times \frac{\text{2,150 square feet (2023)}}{\text{1,649 square feet (1999)}}\) 1.67 tons per structure fire
4 \(I_{t,c} \times FL_{t} \times \frac{EF_{t,p}}{2,000}\) \(\text{61.67 structure fires} \times \text{1.67 ton/fire} \times \frac{78.6}{2000}\) 4.05 tons PM2.5 in county “c”

14.2.8 Improvements/Changes in the 2023 NEI

These sources had not previously been estimated by the EPA; therefore, these are new sources of emissions for the 2023 NEI. Any NEI prior to 2017 reflected sparse emissions from the few States that submitted; however, for the 2020 NEI, we removed any State-submitted estimates. SLTS could again submit their own estimates for the 2023 NEI and subject to quality assurance checks; these submissions are present in the 2023 NEI in lieu of EPA estimates discussed in this section.

14.2.9 Puerto Rico and U.S. Virgin Islands

Since insufficient data exists to calculate emissions for the counties in Puerto Rico and the U.S. Virgin Islands, we based emissions for those domains on two proxy counties in Florida: 12011, Broward County for Puerto Rico and 12087, Monroe County for the U.S. Virgin Islands. The total emissions in pounds for these two Florida counties are divided by their respective populations creating a pound per capita emission factor. For each Puerto Rico and U.S. Virgin Island county, the pound per capita emission factor is multiplied by the county population (from the same year as the inventory’s activity data) which serves as the activity data. In these cases, the throughput (activity data) unit and the emissions denominator unit are “EACH”.

14.2.10 Lahaina Fire of August 2023

We account for the Lahaina PM fire, the most devastating of several fires on Maui Hawaii on August 8-9 2023, via data compiled for the Lahaina Fire Incident Analysis Report, particularly, Table 4.2.1.1 highlighting the number of destroyed and damaged structures based on Fire Safety Research Institute (FSRI) and FEMA independent review.

For structure fires fuel loading, the default 7.3 percent loss rate was replaced with 100 percent for these fires, which translates to 22.87 tons per structure fire for Maui County Hawaii in 2023. No change was needed for motor vehicle fuel loading as the default loss rate is already 100 percent.

For number of incidents of structure fires, we use the 2,117 destroyed structures from FSRI and a scaled portion of the FEMA estimated 283 damaged structures because the FSRI damaged structures include a significant amount of structures that suffered wind rather than fire damage. Because we are assigning 100 percent structure loss in the fuel loading computation, we then scaled these 283 fires back down to the 7.3 percent loss rate used elsewhere yielding an equivalent of 20.7 structures (100 percent) consumed. In total, this yields 2,137.7 incidents for Maui in 2023 vs 45 in the default NFIRS dataset for 2023.

For number of incidents of motor vehicle fires, the default 158 incidents in NFIRS is replaced with an estimate of 3,643 based on a 2024 report where, according to U.S. Army Corps of Engineers, 3,502 burned vehicles were accounted for along with 141 boats.