Establishing a novel combustion generator system to simulate and study emissions from diverse applications

Abstract

Inhalation of emissions from combustion events such as military burn pits and wildland-urban interface fires result in exposures to complex aerosols that may cause adverse health effects. A surrogate combustion generator was created to study these events. A pellet stove was modified to burn diverse fuels comprised of materials found in burn pits (plastic, rubber, and wood). Jet fuel (JF) was added during combustion. The purpose of this project was to: 1) operate the generator over diverse parameters; and 2) characterize the resultant emissions. Pellet combustion (8 g/minute; range: 1–34) ± JF (0.2 mL/minute; range: 0–1) was conducted at fixed rates. Real-time and off-line aerosol characterizations (size distributions, concentrations, morphology) and chemical measures (total and speciation of volatiles, organics, elementals) with subsequent analytic methodology were performed. Pellet combustion produced an average particle concentration of 7.0e+6±1.1e+6 #/cc, and a total volatile organic compound concentration (TVOC) of 0.97+0.41 parts per million (PPM). The addition of JF accelerated the combustion process and increased the total particle counts to 4.5e+7±8.3e+6 #/cc, but decreased the TVOC to 0.63±0.26 ppm. Ultrafine particles were predominantly generated with similar count median diameters (<100 nm) and polydisperse distributions (GSD>1.5). Chemical speciation revealed potential human carcinogens (formaldehyde, acetaldehyde) and the addition of JF resulted in a >40-fold drop in organic carbon constituents. Adding plastic and rubber into the pellet mixture resulted in emissions containing 4% methyl-methacrylate and 10% methyl-isobutyl-ketones. Future experiments will assess diverse wood varieties, synthetic building constituents, and asses the biological effects that follow these inhalation exposures.

Introduction

According to State of Global Air 2024 report, air pollution currently ranks as the second leading risk factor for mortality, accounting for 8.1 million deaths worldwide (Health Effects Institute 2024). Air pollution events due to combustion processes have been alarmingly increasing for a variety of reasons (World Health Organization Media Centre 2016). Inhalation exposure to the complex emissions generated by combustion is associated with increased global morbidity and mortality (WHO Report 2024). Open-air combustion, whether deliberate (e.g. waste incineration) or natural (e.g. forest fires, wildfires), leads to the generation and release of complex aerosols comprised of particles and chemicals (Aurell et al. 2012; Sharkey et al. 2014; American Public Health Association 2015; WHO Report 2024). Federal agencies have documented increasing health risks on exposures to particle and chemical mixtures in environmental, occupational, and domestic settings (NIEHS [date unknown]; US EPA 1997; USAPHC Environmental Health Risk Assessment Program 2013; US Department of Veterans Affairs 2016). An initiative from the National Institutes of Environmental Health Sciences (NIEHS) and the National Toxicology Program (NTP) for future research is focused to rigorously investigate the impact of mixed exposures on human health (https://www.niehs.nih.gov/health/topics/science/chemical-mixtures). These facts reflect the gravity of the current issue, and an urgent need exists to better understand the diverse and numerous factors involved in emission generation associated with combustion processes.

The combustion of wastes in dedicated areas, commonly known as military ‘burn pits’, has historically been a component of many US military bases. While considerable variation exists in their size, burn pits share commonality in their combustion of materials such as, but not limited to, plastics, rubber, paper, wood, electronic waste, and fuels such as diesel fuel, jet propellant-8 (JP8), or jet fuel with additives (JF) as accelerants. This waste disposal method produces very high airborne concentrations of mixed emissions to which not only personnel in the military base are exposed to, but also those in the local environment for many miles (US EPA [date unknown] “NAAQS Table”; US EPA [date unknown] “Criteria for the definition of solid waste and solid and hazardous waste exclusions”). Since 2001, ~3 million United States service members have been deployed to the Middle East for repeated tours of duty lasting months to years, during which they were exposed to military burn pit emissions (Steele 2000; Smith et al. 2013). The incidence and severity of adverse chronic health effects among these Veterans have recently been acknowledged with recent passage of the PACT Act (US Department of Veterans Affairs 2014). Approximately 28% of deployed Veterans from the 1991 Persian Gulf War suffer from a complex symptomatic condition termed ‘Gulf War Illness’ that includes: pulmonary morbidity, cardiovascular morbidity, cognitive impairments, behavioral disorders, fatigue/diminished energetics, compromised immune function and pain (Fukuda et al. 1998; Steele et al. 2011; Dursa et al. 2016; Graves et al. 2016; Miller et al. 2018; Kim et al. 2021). Currently, veterans presenting with at least two of these conditions are diagnosed with ‘Chronic Multisymptom Illness’ (CMI) (Porter et al. 2020). The longer time-lag between deployment and diagnosis in original Gulf War Veterans is predictive of an increased future impact of CMI on current era personnel. This incidence is estimated to be ~19% in current era personnel deployed in Iraq and Afghanistan which means ~570 000 to 840 000 veterans will present with CMI in the coming years (Brian et al. 2012). Given that burn pits are still used at active military bases, single or repeated exposure(s) could raise these estimated numbers (U.S. Department of Veterans Affairs, [date unknown]). Unfortunately, the full etiology of health impacts after burn pit exposures have not been thoroughly investigated (Brian et al. 2012; Porter et al. 2020; Kim et al. 2021). Open pit burning is a local as well as global method of waste disposal. Further, in addition to open pit burning, human exposures to wildfire emissions and wildland urban interface emissions continues to grow. Therefore a method to assess the health effects of these emission exposures in diverse populations is needed (Wiedinmyer et al. 2014; Airborne Hazards and Burn Pit Exposures 2025; Division of Environmental Health Solid Waste Program, [date unknown]). Therefore, recapitulating combustion process in controlled settings is critical to better study the health effects of inhalation exposures to complex combustion emissions in Veterans, first-responders and the general public. Therefore, recapitulating combustion processes in controlled settings is critical to better study the health effects of inhalation exposures to complex combustion emissions in Veterans, first-responders and the general public.

A second environmental risk where inhalation exposures have been impacting public health, nationally and globally, is WUI fires (Cohen 2008; National Academies of Sciences, Engineering, and Medicine 2022; Defossé GE 2023; Singh et al. 2023). Aerosols produced by WUI fires are complicated by the presence of synthetic building materials, and thereby their emission profiles as well as the toxicities of the mixtures have yet to be assessed. A large part of this challenge focuses on the potential synergistic toxicities of these emissions. Due to ongoing exposure of these mixed aerosols, short-term (acute) and/or long-term (chronic) health risks present significant health-care challenges and financial burdens to nearby populations as well as professionals (e.g. fire-fighters, first responders).

Generally, combustion emissions contain carbon-monoxide (CO; more with incomplete combustion), carbon dioxide (CO₂; more with complete combustion), and water along with complex combinations of particles and other chemicals. This mean combustion efficiency (MCE) is defined as: MCE (%) = (CO₂/(CO₂ + CO)) × 100. A higher percentage reflects more complete combustion. Whether complete or incomplete combustion, the resultant aerosols depend heavily on a variety of conditional factors. Therefore, it is important to consider the parameters such as temperature, relative humidity, fuels, accelerants, and air flow, all of which influence combustion, and its progression to emit a complex mixture of particles and chemicals. The existing literature has not fully addressed these parameters in terms of burn pits and WUI fires. This limits our understanding of the health effects and mechanisms of morbidity and mortality that follow inhalation exposures to modern combustion emissions (Kim et al. 2018; Sekimoto et al. 2018; Mouton et al. 2023). Current equipment used to study combustion emissions is predominantly focused on the use of quartz tubes and heating elements. While effective over a wide range of temperatures, these devices are limited by the amount and type of combustible materials to be studied. We sought to improve these limiting factors in our combustion system.

Therefore, we created a surrogate combustion generator capable of continuous operation with diverse bulk materials, that would serve multiple research aspects to: (1) generate reliable and repeatable combustion processes; (2) produce equally reliable and repeatable emissions; (3) offer multiple locations to accurately sample/characterize emissions; and (4) produce sufficient emissions to simultaneously create sample libraries, perform aerosol characterizations and rodent inhalation exposures. Such a generator (BP100, IEStechno, Morgantown, WV) has been developed at the West Virginia University Center for Inhalation Toxicology (iTOX). Our objectives in this study were to: (1) determine if the combustion generator could reproduce diverse combustion processes such as those seen in military burn pits, wildfires, and WUI fires; (2) identify the full range of operating parameters that could simulate these real-world inhalation exposures during combustion of diverse materials; and (3) to characterize the resultant emissions produced by these conditions.

A commercial pellet stove was modified to be capable of burning custom wood-based pellets of varying compositions and proportions that represent commonly burnt materials at military bases in the Middle East (e.g. plastics, rubber, and wood), wildfires (e.g. softwoods, hardwoods) and WUI fires (e.g. woods, insulation, vinyl, composites). Because accelerants were and still are commonly used in military burn pits, we also added the ability to drip accelerants such as JF at various rates during the combustion processes. Operating parameters for commercial pellets (with/without JF) were evaluated and coupled with extensive characterization of the resultant emissions. Achieving the above objectives will provide a foundation to permit evaluation of the biologic consequences that follow combustion emission inhalation exposures. Further, it will facilitate identification of the associated mechanisms of morbidity and mortality. Finally, it will demonstrate our capabilities with custom pellet formation to replicate current and future combustion events.

Methods

Wood pellets

Wood-based pellets were selected to deliver uniform, custom mixtures, under tightly controlled experimental conditions. Commercial grade pellets (Greene Team Pellet Fuel Company, Carmichaels, PA) were used herein for their ease of availability and uniformity. This was necessary for the development of the generator and comparison purposes for preliminary emission profiles for the range of operational parameters presented herein. Figure 1 shows a representative batch of pellets used in these procedures. Custom pellets were also manufactured with sawdust and materials generally combusted in military burn pits, such as plastic and rubber, in different proportions of total weight (e.g. 10%, 15%, and 25% of plastic and rubber). The percentage contributions were ranged based on minimum contribution, to maximum contribution that still forms a pellet. Some preliminary analysis of the emissions from these pellets is presented, but full-scale emission characterization of all the pellet mixtures available is beyond the scope of this manuscript. Therefore, we focused on emission profiles under various operating parameters using commercial pellets with or without JF.

Figure 1.

Picture of representative commercial pellets used for testing in the combustion generator.

Experimental set up, operating parameters and sampling strategies

A schematic of the WVU iTOX combustion generator is presented in Figure 2. Emissions were characterized with the array of instruments listed in Table 1. A pellet stove was modified to operate under a variety of computer-controlled pellet/accelerant feed rates. Pellets could be fed into the crucible burner over a range of 1–34 g/min. Liquid accelerants (JF) could be added at rates of 0–1 mL/min. Input airflow to the stove was set at 12 LPM and maintained for all trials described herein. Sampling was performed from three locations: (1) stove interior; (2) exhaust flue; and (3) exposure chamber. The sampling apparatus used in this study is presented in Table 1.

Figure 2.

(A) Schematic of the WVU iTOX combustion generator. (B) Maximum temperature (in degrees Celsius) in combustion generator at different pellet feed rate.

Table 1.

Instrumentation and sampling parameters for emission characterization with particles as well as chemicals.

Emission characterization	Sampling instrument	Manufacturer’s information	Metric (unit) of measurement	Sampling flow rate (LPM)	Particle size range (nm)	Data log period
Particles	1) SMPS – APS α	TSI Inc., Shoreview, MN	Particle number concentration (dN/dlogDp #/cc)	0.3 (SMPS) / 5 (APS)	14–660 (SMPS) / 370–20 000 (APS)	30 Min sample
	2) CPC	TSI Inc., Shoreview, MN	Particle concentration (#/cc)	0.3	100–1000	Sample / second
	3) ELPI®+ α	Dekati Ltd., Finland	Particle number concentration (dN/dlogDp #/cc)	10	6–10 000	30 Min sample
	4) MOUDIα	Thermo Electron Corporation, Franklin, MA	Particle mass concentration (dM/dlogDp mg/m3)	1	≥300	30 Min sample
	5) Gravimetric collection	37 mm PTFE filter,	Total particle mass (mg/m3)	1	–	Off-line*
	6) TEM	JEM 1400 (JEOL Ltd., Peabody, MA) on 200 mesh copper grid	Particle morphology	1	–	45 s sample
	6) FE-SEM with EDS	Hitachi S4800 with the Bruker 25 mm filter polycarbonate membrane	Particle morphology and qualitative heavy metal analysis	1	–	45 s sample
	7) TPS 100	Grid; RJ Lee Group	Particle morphology/composition	0.005		Off-line*
Chemicals	8) TVOC 2		Total volatile organic chemical concentration (PPM)	0.5	NA	Sample / second
	9) OC/EC	Filter; RJ Lee Group	Total organic and elemental carbon mass (mg/m3)	1	NA	Off-line*
	10) GC-MS	Agilent Technologies, Inc., Santa Clara, CA	Chemical speciation including VOCs, VCCs, and PAHs.	1	NA	Off-line*

Note: SMPS-APS: scanning mobility particle sizer – aerodynamic particle sizer; CPC: condensation particle counter; ELPI +: electric low pressure impactor; TEM: transmission electron microscopy; FE-SEM: field-emission scanning electron microscopy; EDS: energy dispersive X-ray spectroscopy; LPM: liters per minute; nm: nanometers

^αsampling device or instrument that provides particle size distribution

^*off-line measurements were sampled during entire exposure trial if not specified

OC: organic carbon; EC: elemental carbon; GC-MS: gas chromatography and mass spectrometry; TVOC 2: total volatile organic compounds (photoionization detector: PID, 10.6 eV lamp, TV2SXBMX-X, ION Science Ltd, TX); VCCs: volatile carbonyl compounds; PAHs: polyaromatic hydrocarbons; PPM: parts per million; NA: not applicable.

Emission profiles are dependent on multiple parameters such as pellet composition, degree of combustion (complete or incomplete), pellet feed rate, accelerant pump rate and airflow rate. These variables affect combustion temperatures in our system (ranges ~150–500°C) and can produce complex emission profiles (Sekimoto et al. 2018). When examining the association between maximum combustion temperature and pellet feed rate (Figure 2(B)), the relationship appeared linear up to a duty cycle of ~50% at which point the temperature plateaued (~500°C). For all the exposure assessment data presented in this study, the system was operated at an average target temperature of 200°C which was achieved with an auger duty cycle feed rate of 15% (range: 1–100%) which corresponded to ~8 g/min (range: 1–34 g/min) mass of pellets dropped.

Certain indices reflect the state of emissions from combustion processes. For example, increased carbon monoxide (CO) indicates incomplete combustion, while higher levels of carbon dioxide (CO₂) reflect more complete combustion. Because water generation occurs in all combustion processes, it is important to consider the influence of relative humidity (RH) on the characterization of particle agglomerates, size distributions and chemicals. Therefore, our sampling strategy included multiple real-time measurements of RH and temperature (HMP60, Vaisala, Finland), and CO/CO₂ (AO2000, AAB, Zurich, Switzerland).

Modified combustion efficiency has been used to distinguish between smoldering and flaming phases during emission (Kim et al. 2018; Sekimoto et al. 2018; Mouton et al. 2023; Vance et al. 2023). Unlike some studies that focus on a particular combustion phase (smoldering or flaming), we aimed to be able to achieve both, as this approach would most closely resemble the uncontrolled smoldering and flaming that occurs during real-world fire events.

Finally, the capability to sample emissions through multiple ports was implemented. Complementary real-time and offline assessments were made to study the evolution of particles and chemicals in high resolution during combustion generation and animal exposures.

Particle and chemical characterization of emissions

Multiple aerosol assessments such as mass concentration, number concentration and size distribution were performed with a variety of instruments. These apparatuses and their operating parameters (e.g. sampling flow rates, manufacturer, size detection limits) are listed in Table 1. The purpose of these measurements was to provide complementary assessments of particle size distributions (PSD) from multiple perspectives (concentration with estimated mass and number metrics). Resultant PSDs are reported based on diameter (count median diameter, CMD; or mass median diameter, MMAD) and geometric standard deviation (GSD). The upper range of the particle characterization equipment has a detection limit of 10–20 μm. This particle size is beyond the respirable range, and as such was not considered in our assessments. Due to the greater mass of these larger particles, many did not exit the combustion chamber and fell into the ash. Therefore, an ‘Ash Library’ was established by collecting ash and stored for future studies.

The gravimetric mass of aerosols was determined for particles collected on a pre-weighed 37 mm diameter, 0.45 μm pore polytetrafluorethylene filter (SKC Inc., PA). Air was drawn from the flue/exposure chamber at 1 LPM. Particle morphology (i.e. shape, agglomeration, condensation etc.) was evaluated by electron microscopy (S-4800, Hitachi, Tokyo, Japan). Particles were collected onto 25 mm diameter, 0.2 μm pore size track-etched polycarbonate filters (Sterlitech Corporation, VA) by drawing air from the flue/exposure chamber at 1 LPM using a pre-calibrated sampling pump with a modified NIOSH 7300 method (Hasselriis and Licata 1996; NMAM 2003; Nzihou and Stanmore 2013). After collection, filters were prepared and analyzed with transmission and scanning electron microscopy (TEM and SEM, respectively) with energy dispersive X-ray spectroscopy (EDS, qualitative elemental analysis). A thermophoretic sampler (TPS100, RJ Lee Group, Pittsburgh, PA) provided a complementary off-line particle morphology assessment to TEM, and SEM. The TPS100 has a flow rate of 5mL/min and is ideally suited to collect particles less than 300 nm (Leith et al. 2014). The sampler collects particles directly onto a TEM grid with a carbon film and can be placed directly into the TEM or SEM with no sample preparation.

Chemical constituents in combustion emission were majorly segregated into: volatile organic compounds (VOCs), volatile carbonyl compounds (VCCs), polyaromatic hydrocarbons (PAHs), and heavy metals. An aggregate total level of organic vapors (e.g. VOCs, VCCs, and PAHs) were measured in real-time using a TVOC2 (Photoionization Detector: PID TV2SXBMX-X, ION Science Ltd, TX). This approach provides valuable information regarding emission patterns over a broad range (e.g. >900) of ‘volatiles’ during the course of a given combustion experiment. For additional speciation, emission mixtures were collected in a water trap as water has been documented as an inevitable product during combustion process (Sugaya et al. 2001; Eilhann and Castaldi 2012; Mojtaba et al. 2017; David and Niculescu 2021; Caron-Beaudoin et al. 2022). The combustion emissions were drawn at 1 LPM from the flue/exposure chamber and sampled in a flask containing 50 mL of ultra-pure water (≥17 mΩ) and evaluated separately for specific VOCs (>60), VCCs (>10), and PAHs (>20) of interest. After water trap collection, chemical speciation was performed off-line with a headspace sampler coupled to a gas chromatography/mass spectrometer (GC-MS, Model# 7697A, Model #8890, Model# 7000, respectively, Agilent Technologies Inc., Santa Clara, CA) with methods modified from previously documented literature (Sugaya et al. 2001; Eilhann and Castaldi 2012; Masiol et al. 2016; Mojtaba et al. 2017; David and Niculescu 2021). For efficiency, assessment of pellet combustion emissions, 3 of the highest cancer risk chemicals according to the International Agency for Research on Cancer (IARC) were used for screening purposes including: formaldehyde (CAS # 50–00–0; Group 1 carcinogenic to human), acetaldehyde (CAS # 75–07–0; Group 2B: possibly carcinogenic to human), and acrolein (CAS # 107–02–8; Group 2A: probably carcinogenic to human). A complete list of the chemicals and analytic methods is listed in the supplemental materials (PAHs in Table S1, VOCs in Table S2, and VCCs in Table S3). Qualitative assessments of certain chemicals resulting from combustion of customized pellets are also included. The GC-MS instruments used herein are capable of chemical characterization of specific compounds of interest (average ± standard deviation). Chemical speciation for these compounds was identified using the National Institute of Standards and Technology library databases (NIST, Gaithersburg, MD).

The elemental carbon (EC) and organic carbon (OC) content of the combustion samples collected on a pre-fired quartz fiber filter was analyzed using the NIOSH Method 5040. The EC/OC analysis involves placing a portion of the filter in a thermal-optical analyzer (Sunset Laboratories, Inc. OCEC Dual Optics Lab Instrument, Tigard, OR) where it is heated in a controlled manner, causing the organic and elemental carbon to oxidize at different temperatures, allowing for their separate quantification. Results were reported in mg/m³.

While the combustion chamber was thoroughly vacuumed after each experiment and purged with HEPA filtered air for at least 12 h, the potential for residual particles and chemicals could not be ruled out. Therefore, particle and chemical evaluations were conducted with real-time monitors and GC-MS analysis to identify any carry over in subsequent trials. This was achieved by conducting measurements in the absence of combustion. Trials with ‘cold’ and ‘hot’ air were conducted over the range of temperatures associated with pellet combustion to assess background aerosol profiles and determine if volatile chemicals were released from the oven or stack due to re-volatilization of any residue from previous experiment(s). The ‘cold test’ was conducted in the absence of combustion with room air being pulled into the combustion chamber with the input fan operating (moving parts were activated). The ‘hot test’ was conducted with 2 high wattage cylindrical electrical heaters (CSS-508885/120V, Omega Engineering Inc., Norwalk, CT) placed in the generator that could produce a similar amount of heat, in the absence of combustion (~78°C Flue Temperature). As with the ‘cold test’, the fan was active for the ‘hot test’.

Series of TEM and/or SEM samples collected during a trial were analyzed to track the evolution of particle morphology over time and correlated with particle size distribution and concentration measurements.

Statistical analyses for data evaluation

To establish consistency of assessments with the described instrumentation, each combustion experiment was sampled/characterized during a 1-h period. Sampling began when the pellet material ignited, with visible fire in the combustion generator, and with a crucible temperature approximately 200°C or higher. Three replicates were performed for each analysis with and without JF.

Data and concentration analyses were performed as: average + standard deviation (SD) and were analyzed using JMP^® Pro 16.0.0 (SAS Institute, Inc., Cary, NC). Particle size distributions are reported as CMD (measured with SMPS-APS and ELPI) or MMAD (measured with MOUDI) in micrometers (μm) ± GSD.

Combustion is a dynamic process that produces complex particles and chemicals throughout a given trial. In our experience, a 1-h trial was a sufficient time period to collect enough data (for a given piece of equipment) to generate/calculate the relative standard deviation (SD).

Results

Background measurements with ‘cold tests’ and ‘hot tests’ reflected negligible residual particles or volatile chemicals (e.g. below the level of detection). Real-time analysis of total VOCs resulted in to 0.047 and 0.031 PPM for the ‘cold’ and ‘hot’ test, respectively. Chemical speciation with GC-MS analysis for both tests did not reveal any volatile chemicals above detection limits. The ‘cold test’ resulted in particle counts ~4–6 orders of magnitude lower than during the combustion process. The ‘hot test’ produced higher counts of number of nanometer-sized particle, but PSD was considerably different than aerosols emitted from combustion. The ‘hot test’ revealed the presence of nano-sized particles (TEM and SEM analysis, data not shown). Subsequent tests confirmed that the heaters, themselves, released these particles when a voltage was applied to the heaters. Therefore, based on these observations, we concluded that any particle and/or VOC production caused by other sources or due to residue from previous runs was close to negligible.

Evaluating combustion generator system

The feed rate of the fuel pellets and the JF drip rates were controlled via the computerized control system. The rate of the pellets was controlled by adjusting the percentage (duty cycle) of time the auger was activated during each 15 s period. Auger duty cycle rates (%) of: 12.5, 25, 50, 75 and 100 delivered commercial pellets at a mass rate (g/min) of: 7, 12, 19, 31, 34. Similarly, JF was delivered linearly at rates (mL/min) of: 0, 0.2, 0.4, 0.6, 0.8, and 1.0.

As shown in Figure 3, combustion with JF produced similar CO (PPM: 62.8 ± 6.6 vs 62.7 ± 5.9) and slightly higher values for CO₂ (PPM: 4198.5 ± 164.5 vs 4035.1 ± 245.1) compared to without JF. Additionally, exposure chamber environmental conditions were maintained at levels that would be acceptable during animal exposures for all tests (temperature 19–22°C, RH 30–70%, data not shown).

Figure 3.

Real time measurements (PPM: parts per millions) of carbon dioxide (CO₂, A) and carbon monoxide (CO, B) during the combustion of commercial pellets with and without JF.

Particle characterization

Real-time total particle measurements made with the Condensation particle counter (CPC: TSI Inc., Shoreview, MN) are presented in Figure 4(A). Particle concentration was relatively lower in the presence of JF (Figure 4(A), #/cc: 7.0e+ 6 ± 1.1e+ 6 vs 4.5e+ 7 ± 8.3e + 6). In Figure 5, microscopy images are presented: Panel B) representative TEM morphology of agglomerates; Panel C) representative SEM (EDS); Panel D) EDS detection of elements in emission samples.

Figure 4.

Real time measurement of total particle concentration (# per cc) using CPC (A) and total volatiles (PPM: parts per millions) using TVOCs (B) on combustion of commercial pellets with and without JF.

Figure 5.

Off-line measurment of chemical speciation with GC-MS (A) of aldehydes, particle morphology using TEM (B), and SEM (C) with EDS (D) on combustion of commercial pellets with JF.

PSD, as determined with MOUDI measurements had a MMAD of 0.162 μm and GSD = 1.62, at a concentration of 2.84 mg/m³ in the absence of JF. This corresponded with a gravimetric filter measurement of 2.76 mg/m³ (Figure 6(C)). Combustion in the presence of JF produced a slightly larger PSD measurement (as measured with MOUDI): MMAD of 0.183 μm; and GSD = 1.93; at a concentration of 3.64 mg/m³. This also corresponded with a gravimetric filter measurement of 3.83 mg/m³ (Figure 7(C)).

Figure 6.

PSD and concentration using SMPS/APS (A); ELPI (B); and MOUDI (C) on combusting commercial pellet with No JF. MOUDI measurement MMAD = 0.162 μm GSD = 1.62; MOUDI concentration = 2.84 mg/m³. Corresponding gravimetric filter measurement = 2.76 mg/m³.

Figure 7.

PSD and concentration using SMPS/APS (A); ELPI (B); and MOUDI (C) on combusting commercial pellet with JF. MOUDI measurement MMAD = 0.183 μm; GSD = 1.93; MOUDI concentration = 3.64 mg/m³; corresponding gravimetric filter measurement = 3.83 mg/m³.

Considered together, this reflects increased small particle generation that creates less mass in the absence of JF; whereas, in the presence of JF, more large particles are present that tend to create a greater mass.

Chemical characterization

Figure 4(B) shows the real-time measurements of total volatile organic compounds with TVOC 2 (Photoionization Detector: PID, 10.6 eV lamp, TV2SXBMX-X, ION Science Ltd, Stafford, TX). With JF, commercial pellet pellets emit slightly lesser volatiles compared to without JF (PPM: 0.63 ± 0.26 vs 0.97 ± 0.41). In general, there is an overlap of total volatiles ~0.5–1.2 PPM for both conditions. To further speciate VCCs, we analyzed formaldehyde at 20–30 μg/mL, acetaldehyde at 2.5–3.5 μg/mL, and acrolein at 1.2–2.2 μg/mL during combusting with JF over 1-h trial (Figure 5(A)). Other chemicals within groups of VCCs, VOCs and PAHs were below detection limits (0.02 μg/mL) with commercial pellet combustion.

In Figure 8, pellets were formulated with 100% red oak wood (Panel A). In panels B, C and D the wood was mixed with different proportions of plastic, and rubber. Qualitative assessment of emissions was performed by combusting the pellets with plastic and rubber and sampling for subsequent GC-MS characterization (Figure 9(A) for VOCs [e.g. methyl methacrylate and methyl isobutyl ketones] and VCCs [e.g. Hexanal], Panel B for PAHs [e.g. naphthalene and other derivatives], and Panel C for furfural derivatives).

Figure 8.

Customized: (A) 100% commercial pellet, (B) 90% wood (W), 5% plastic (P), 5% rubber (R), (C) 80% W, 10% P, 10% R, and (D) 70% W, 15% P, 15% R. As the commercial pellet content decreased, the pellet weights decreased, and the pellets became less stable. Picture of each type of pellet fabricated (mixture percentages in terms of mass).

Figure 9.

GC-MS spectra for qualitative assessment of customized pellet composed on plastic, rubber, and wood – (A) for VOCs and VCCs, (B) for PAHs, and (C) for furfural derivatives.

Accordingly with chemical analysis, the EC/OC chemical speciation results from samples collected during combustion of commercial pellets had more than 40-fold higher emissions without JF vs with JF for organic (with JF: 4 mg/m³ ± 2 without JF: 170 mg/m³ ± 15) and elemental carbon (with JF: 1.7 mg/m³ ± 1; without JF: 77 mg/m³ ± 6).

Discussion

The results provided herein represent our initial identification of the optimal operating conditions for combustion processes created a surrogate generator that is associated with military burn pits (with JF), and wildland/WUI fires (without JF). This is a fundamental step toward establishing the combustion generator in the WVU iTOX for future inhalation exposure research. Proper aerosol characterization and control are critical in the identification of the underlying mechanisms of acute and/chronic illnesses that result from inhalation exposures to these mixed emissions. The greater goal of these collective endeavors in terms of public health is to prevent and/or mitigate morbidity and mortality associated with inhalation exposures to combustion emissions.

One of our goals in the design of the generator was the ability to add solid and liquid fuels, continuously, for extended time periods. We have presented our capability to generate complex combustion mixtures under modifiable operating parameters with real-time data logging for accurate and repeatable measurements. Further, sampling and exposure assessments can be performed at multiple locations from the exhaust flue. When combined with variable dilution ratios, the system can be used to model a variety of exposure conditions ranging from the immediate source of a fire to distal locations. This will provide an opportunity to study how time and/or distance influence aerosol characteristics. Considering recent wildfires and WUI fires in Canada (US EPA Report 2025), Maui (US Fire Administration 2024) and Los Angles (MacCarthy and Richter 2025), this is a relevant consideration as emissions from these types of events change over distance and differentially impact large populations. Added value is present in that the surrogate generator is capable of combusting virtually any substance including synthetic and biomass sources specific to geographic locations across North America (as might be experienced in wildfires or WUI fires). Therefore, applications with the generation system herein may be extended not only to Veterans, and co-deployed personnel; but also, domestic service personnel suffering from idiopathic conditions such as 9/11 first-responders, firefighters, police officers and civilians exposed to equally diverse combustion conditions. The major challenge in characterizing combustion emissions is effective sampling of the complex mixture of particles and chemicals relative to a specific fire. In certain aspects, queries may be focused on the conditions of combustion. Whereas, in others, the focus may be on the actual inhalation exposure. Further, mixtures are very dynamic as their constituents experience different conditions throughout the course of the combustion system. These conditions include RH, temperatures, feed rate of combusting materials, and air flow. All of these are capable of influences the degree of complete or incomplete combustion during the entire sampling period. Because of this dynamic nature, we developed a methodology incorporating real-time measurements in addition to offline analyses of these exposure conditions. This is a necessary step in the pursuit of identifying causative emission exposure parameters associated with adverse health outcomes.

The addition of JF to burning pellets enhanced combustion compared to without JF (respectively), and this was reflected by fewer total particles (#/cc:7.0e+ 6 ± 1.1e + 6 vs 4.5e+ 7 ± 8.3e + 6) and volatiles (PPM: 0.63 ± 0.26 vs 0.97 ± 0.41). JF did not appear to greatly influence particle size distribution as the medians were < 200 nm with a polydisperse distribution (GSD >1.5). The CMD measured with the SMPS-APS and MMAD determined with the MOUDI resulted in a smaller particle size distribution with JF (SMPS-APS: 80.6 nm, GSD:1.66 and MMAD: 162 nm, GSD: 1.62); compared to those without JF (SMPS-APS: 84.3 nm, GSD:1.62 and MMAD: 183 nm, GSD: 1.93). The CMD determined with the ELPI device resulted in smaller sized particle distribution emissions without JF (ELPI: 62.3 nm, GSD:1.74) compared to with JF (ELPI: 74.5 nm, GSD:1.81). Electron microscopy confirmed a 10 nm to 1 μm size range of particles (Figure 5, Panel B) that contained elemental particles (Figure 5(C,D)). Analysis of a TEM grid collected outside of the chamber during combustion with JF using the TPS100 confirms the presence of particles on the order of 10 nm to 1 micrometer (as shown in Figure 10). TEM, SEM-EDS, and TPS100 image confirm diverse morphological features of fine and/or ultrafine size particles would favor deposition in the respiratory tract. Metal containing samples could include local damages and potentially migrate from the lungs to systemic compartments. It is also noteworthy that these particles simultaneously exist with harmful chemicals that may further exacerbate biologic impact(s) of inhalation exposures to combustion emissions.

Figure 10.

Image using thermophoretic sampler (TPS 100) for commercial pellet with JF.

Chemical speciation using EC/OC resulted in a 40+ fold increase in emissions without JF vs with JF for organic (170 ± 15 mg/m³ vs 4 ± 2 mg/m³, respectively) and elemental carbon (77 ± 6 mg/m³ vs 1.7 ± 1 mg/m³, respectively). This potentially contained human carcinogens such as formaldehyde (IARC Group 1: carcinogenic to human), acetaldehyde (IARC Group 2B: possible carcinogenic to human), and acrolein (IARC Group 2A: probable carcinogenic to human). The addition of synthetics such as plastic and rubber to the pellets resulted in 4 ± 0.17% methyl methacrylate and 10 ± 1.26% methyl isobutyl ketones in the total chemical yield, respectively. These potentially harmful chemicals are mixed with nanosized particles in the aerosols and may alter dose-responses and/or exert synergistic biologic effects.

Challenges, limitations, and assumptions

Our first challenge was to understand the behavior of emitted particles and chemicals produced by combustion. To address this, all materials were combusted at relatively constant temperatures (with <5% deviation); however, the conditions and fuels used resulted in different temperatures. Once repeatability was achieved, emission profiles were then evaluated. We have also calibrated a uniform feed rate to deliver every pellet type on equal mass bases. This uniformity of mass combustion allows for comparison between respective emission profiles.

Although we evaluated test conditions that simulated real-world scenarios, we relied heavily on stability and repeatability. We acknowledge that there is some degree of variability with environmental combustion. As such, the observations and conclusions reached herein must be considered under this context. Perhaps more importantly, it should be acknowledged that the conditions considered in these studies are commonly observed in fires at one point or another, and emissions are generated that humans are exposed to are implicitly identified. Similarly, the pellet constituents and amounts may not be representative of all fires discussed. For our pellet mixtures, we began with the most commonly reported materials observed in burn pits, and WUI fires, as well as geographic biomass variation. Perhaps more importantly, the percent contributions had to be in proportions that could be pressed into a usable pellet form. However, this may not fully reflect the diversity of real-world combustion scenarios.

It is known that ‘early’ chemical mixtures are generated while heating (prior to formal burning/combustion) at temperatures <200°C. These initial toxicants may be of concern to actual inhalation exposures or simply contribute to error in our assessments. Because the focus of this manuscript was to evaluate our combustion generator mechanics and resultant emissions, only assessments during the established combustion process itself are provided. Future characterizations will be performed that assess emission of toxicants generated during this brief period.

Conclusions and future directions

Pellet materials, JF feed rates, temperature, and air flow influence the combustion process and resultant emissions. We have evaluated these parameters over a range of operational conditions. Pellet composition, qualitatively and quantitatively, influenced the particle size distribution, aerosol concentration, chemical concentration, and speciation of volatile chemicals including Proton Transfer Reaction – Mass Spectrometry (PTRMS). The combustion mixtures created, were dominated by a large number of respirable, nanometer-sized particles and potentially harmful chemicals that may be associated to acute or chronic health impacts (including CMI). The successful evaluation of our combustion generator demonstrated the scope and depth of our capabilities to produce and characterize complex combustion aerosols and chemicals that in future studies will mimic real-world exposures in rodent models.

Future inhalation exposure studies will rely heavily on the ability of our novel surrogate generator to recapitulate combustion processes reliably and repeatedly for diverse materials and conditions that humans are increasingly exposed to. Given the increasing scope and frequency of extreme temperature events, the influence of temperature and RH on the likelihood and intensity of wildfires and WUI fires should be explored. As the reach of wildfire emissions appears to be growing, studies should also evaluate the ‘aging of aerosols’ to characterize the evolution of particles and ‘volatiles’ during movement from source to target. This is necessary as combustion emissions travel great distances and the influence of the surrounding ecosystem (e.g. environmentally relevant factors such as ultraviolet light, temperature, humidity) alter the traveling emissions. We understand that combustion of wood from different geographic locations could influence emission constituents and differentially impact toxicity (Climate Change Indicators: Wildfires 2025). Therefore, variations in regional wood sources/biomasses and synthetic buildig materials should be explored to better understand emissions from WUI and wildfires.

Finally, a surrogate ‘house pellet’ is in development with representative composition that will include materials specific to modern construction (e.g. shingles, siding, pipes, insulation, plastic, rubber, and electric wiring, etc.). Full scale particle and chemical characterizations of emissions from combusting these custom formulated pellets will be conducted in the next phase of our studies. Measurements in the current study were conducted during steady state combustion (e.g. mid-trial). Because different ratios of smoldering and flaming occur during the beginning, middle and end of a trial, and novel materials are increasingly being consumed in WUI fires, we will be able to adapt our system to explore the health effects of inhalation exposures to these emissions. Therefore, all these periods may be explored in future studies with yet to be created materials.

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/08958378.2025.2531115.

HIGHLIGHTS.

A novel surrogate combustion emission generation system is presented. We demonstrate multiple methodologies to characterize the particles and chemicals present in the emissions produced during the combustion process from pellets made with combinations of wood products and synthetic materials with and without liquid accelerants. By establishing these detailed exposure assessments for the combustion processes described herein, a foundation is established for future experiments with rodents to determine the health impacts of inhalation exposures to these complex toxicants.

Data availability statement

The data that support the findings of this study are available from the corresponding author, [TRN], upon reasonable request.