Studies related to effluent produced by structure and vegetation fires often focus on gas phase or solid condensed phase, with limited treatment of liquid condensate generated as smoke cools to ambient. Recent post-fire human health concerns related to systemic human exposures to fire smoke and contamination of water distribution systems after wildland urban interface fires can be informed by understanding the chemical composition of liquid condensate resulting from large-scale fire experiments. In this pilot study, fire effluent (smoke) samples were continuously drawn from five different full-scale room-and-contents fire experiments, from which condensate was collected as the effluent cooled. Elevated concentrations of several volatile organic compounds (VOCs), including benzene, toluene, xylenes, styrene, naphthalene, and acetone along with several anions were detected in the acidic effluent. Many of these same VOCs have been identified in the air during firefighter safety experiments and in post-fire water distribution systems at levels that raise concern for human health. Benzene and naphthalene concentrations in the condensate were orders of magnitude above typical water quality standards and thus may directly contaminate large volumes of water. Peak benzene concentrations were similar to highest values reported from contaminated water distribution systems after wildfire events, though additional study is needed to understand the mechanisms by which this condensate may contribute to systemic contamination. Improved understanding of liquid condensate from fire effluent may be important to other areas of human and environmental health study, and some considerations are provided for future research.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
Fire effluent studies have largely focused on gas phase and solid phase products of combustion. While it is known that some harmful gases may condense onto solid carbonaceous soot particles, less is understood about the condensed liquid phase that may result from cooling fire effluent. Many of the compounds that are present in gas phase at high temperature, including high concentrations of water vapor, may become trapped in liquid phase as the effluent is cooled in the environment or when it comes in contact with cooler surfaces such as walls, building utilities, firefighting personal protective equipment (PPE), and skin. This phenomenon is of particular importance with volatile organic compounds (VOCs) that may pose human health and environmental concerns [1] and are nearly ubiquitous in smoke from residential fires [2‐4]. The composition of fire effluent will vary depending on materials burning and conditions in which they are burning, but several of the most common airborne VOCs measured during uncontrolled fires have been classified as known (e.g. benzene) and probable (e.g. styrene) carcinogens by the International Agency for Research on Cancer (IARC) [5‐7].
Research on the concentrations of airborne contamination produced by residential, wildland, and training fires have consistently identified elevated levels of VOCs such as benzene, toluene, ethylbenzene, xylenes, styrene (BTEXS) and polycyclic aromatic hydrocarbons (PAHs) such as naphthalene in the fire environment [2, 8‐18]. At the same time, these compounds have have been found to create an exposure risk for humans through inhalation and dermal absorption routes [18‐25]. While much of this research has focused on firefighter’s occupational exposure to smoke during firefighting operations, contaminants in the smoke can also create health concerns for general population located downwind from a fire [26, 27]. In addition to airborne risks, by-products of large fire events may create water contamination risks in water catchments [28, 29] and water distribution systems [30] after wildland and wildland urban interface (WUI) fires.
Anzeige
WUI fires have become a common occurrence around the world including in the western United States. While acute damage to vegetation, structures, vehicles, and other property have long been documented, attention has recently turned to community resilience impacts related to damaged water distribution systems. In particular, widespread VOC contamination in drinking water systems in California after the Tubbs Fire (5,636 structures destroyed in 36,807 acres of Napa and Sonoma counties [31]) in 2017 and the Camp Fire (18,804 structures destroyed in 153,336 acres of Butte county [31]) in 2018 have initiated research into the source of these contaminants. After the Camp Fire, within the Paradise Irrigation District water system, benzene concentration ranged from nondetect (<0.5 \(\mu\)g/L) to >2,217 \(\mu\)g/L, while the peak value measured after the Tubbs Fire was 40,000 \(\mu\)g/L [30, 32, 33]. The US Environmental Protection Agency’s (EPA’s) 1-day health advisory for a 10-kg child drinking 1 L/day is 200 \(\mu\)g/L [34], while the Federal and California maximum contaminant levels (MCL) are 5 and 1 \(\mu\)g/L, respectively [34, 35]. The Tubbs and Camp Fire incidents are reportedly the first known wildfires where widespread contamination was found in the water distribution network but not in the source water after the fire [30]. In the years since, drinking water system contamination has been found after the Echo Mountain Fire, Almeda Fire and Lionshead Fire in Oregon [36], the Marshall Fire in Colorado [36, 37] and was a concern after the Hermits Peak/Calf Canyon wildfire in New Mexico [36]. Schulze and Fischer (2021) have linked local burn severity to the probability of water contamination exceeding maximum recommended levels in the City of Santa Rose and Town of Paradise after the Tubbs and Camp Fires, and found the most useful estimate of burn severity to be the density of damaged structures after these WUI fire events [38]. However, research is still needed to understand the mechanism by which fires contribute to this contamination. It has been hypothesized that this contamination could be the result of (1) the degradation of plastic materials present in the water systems, (2) back siphoning of contaminated water through damaged plumbing, and/or (3) contaminated air and/or ash being sucked into the distribution system as it depressurized [30, 38, 39]. Isaacson et al. (2021), Richter et al. (2022) and Metz et al. (2023) have begun to address certain aspects of this first concern [40‐42], but questions related to the impact of contaminated air being drawn into the water system and condensing have not been addressed. This concern is summarized by the Paradise Irrigation District in a description of the science of VOC contamination as a result of wildfire, where it stated that “During a wildfire, and especially in conjunction with depressurization, contaminants (especially VOCs) can get drawn into the water distribution system. It is suspected that these contaminants are drawn in as a gas.” [43].
In addition to VOCs and PAHs with known direct human health concerns, solvents, anions, and extreme pH levels in liquid condensate can impact the flowing and stationary bodies of water that can be impacted by fire effluent. The EPA recommends criteria for ambient water quality to protect aquatic organisms, including an acute limit for chloride of 860 mg/L and chronic limits for pH between 6.5 and 9 [44, 45]. Fire related contamination of water catchments can also be mobilized through suppression or rain water runoff from the debris [29] in a similar manner to contamination of water distribution system back siphoning contaminated water through damaged plumbing [30, 38, 39]. Thus it is valuable to understand relative compounds and concentrations collected in fire effluent condensate as well as runoff.
The purpose of this manuscript is to characterize the concentrations of VOCs, anions, and pH in liquid condensate and runoff water collected from residential room-and-contents fire experiments and describe how fire effluent condensate may help further understand impacts of fire on human health and the environment. This pilot study focuses on condensate and runoff from room-and-contents fires due to the high concentrations of airborne VOC contaminants noted from residential fires [2, 3, 14] and the link between water contamination and the density of damaged structures in WUI fires [38]. Suggestions for future studies are provided to expand on this initial effort into fires involving full structures and vegetative fuels.
2 Methods
2.1 Full Scale Compartments for Live Fire Experiments
Fire effluent condensate was collected from five full-scale, room-and-contents fire experiments in purpose-built residential-style structures (Table 1). The primary fuel source and ignition location in each experiment was an upholstered sofa comprised of polyester fabric, polyurethane foam, polyester fill, and an engineered wood frame (Fig. 1). The flooring consisted of plywood, polyurethane foam padding, and polyethylene terephthalate carpet. Experiments 1 and 2 included only the upholstered sofa, flooring, and a plywood box as fuels. Experiments 3 and 4 included additional fuels for a furnished kitchen and living room. Experiment 5 was conducted in a simulated furnished studio apartment including a bedroom, kitchenette, office, and living room with a fuel load that included 1.5 cm thick wood paneling on the walls and a wide variety of common household items to simulate a cluttered residential fire scene. In Experiments 1-4, fires were ignited with no exterior ventilation, and any changes in ventilation are described in Table 1. Experiment 5 was conducted with an exterior door throughout the fire test. Details of the experimental structures, including compartment, door, and window dimensions can be found in Section S.1 of the Supplemental Information. Additional information on the dimensions, weights, and materials of each item in the fuel load is detailed in Section S.2.
Table 1
Summary of Fire Experiments
Exp. #
Fire Compartment Size [m\(^2\)]
Fuel Load
Summary of Timeline
1
13
Upholstered sofa and
Fire becomes ventilation limited
carpeted flooring
at 5 min., suppression at 20 min
2
13
Upholstered sofa and
Fire becomes ventilation limited
carpeted flooring
at 3 min., window failure at 33 min.
flashover at 35 min., suppression at
40 min
3
65
Furnished kitchen and
Fire becomes ventilation limited at
living room
5 min., open exterior door at 18 min.
flashover at 22 min., window failure
at 23 min, suppression at 26 min
4
65
Furnished kitchen and
Fire becomes ventilation limited at
living room
5 min., open exterior door at 12 min.
flashover at 18 min., window failure
at 20 min, suppression at 24 min
5
42
Furnished studio
Door to exterior open, flashover
apartment
at 8 min., window failure at 5.5 min
and 9.5 min, suppression at 18 min
Anzeige
Gas temperatures were measured in the fire compartment with 1.3 mm bare-bead, chromel-alumel (type K) thermocouples arranged in a vertical array of eight thermocouples with a spacing of 0.3 m, starting 0.03 m below the ceiling. Thermocouple arrays were located near the center of each room. (Additional detail is provided in Section S.3 of the Supplemental Information.)
Figure 1
Pretest pictures of the fuel load for each experiment
×
2.2 Fire Effluent Condensate Collection
2.2.1 Condensation Trap Design
Fire effluent condensate was collected by pumping fire gases from ports inside the fire compartment through stainless steel condensation traps designed for high temperature fire gas sampling. Four to six gas sampling ports were located throughout the fire compartment in each fire at elevations ranging between 10 cm above the floor and 2.5 cm below the ceiling (see Section S.3 in Supplemental Information for detailed locations). Each port consisted of 0.95 cm tubing which was routed out of the fire compartment to a condensation trap. The condensation traps were coils of 0.95 cm tubing, approximately 680 cm in length, which drained into 90 mL reservoirs (Fig. 2). The coils and traps were submerged in an ice/water bath throughout each experiment. Fire gases were pulled from each sampling port by a 21 lpm vacuum pump.
The total volume of combustion gases sampled was estimated by assuming a constant volumetric flow rate based on the pump specifications over a fixed duration. The pumps were activated prior to ignition, but it was assumed that condensate does not form until combustion gases are passing through the condensation traps. Therefore, the duration of condensate collection was defined as the time between when the temperature measured nearest to each sampling location increased more than 10% above the ambient temperature (i.e. about 1–3 min after ignition) and when the pumps were turned off at the end of the test.
Figure 2
(left) Example stainless condensation trap for sample collection. (right) Collection of condensate from reservoir by direct transfer into volatile organics analysis (VOA) vial
×
2.2.2 Condensation Trap Cleaning
New condensation traps were constructed for these experiments such that no prior contamination from combustion gases was present. One set of traps was used for Experiments 1, 3, and 5, and a separate set was used for Experiments 2 and 4. The condensation traps were cleaned after each experiment as follows: fill with water, clear with compressed air, rinse with 91% isopropyl alcohol, force a foam plug through the tubing sections with compressed air, fill with water, and clear with compressed air. An oil-free air compressor was used when clearing the condensation traps. No other solvents or cleaning agents (e.g. acetone) were used.
2.2.3 Condensation Trap Baseline Sampling
Baseline measurements were collected prior to each experiment to characterize the presence of any residual contamination in the condensation traps. 25 mL of distilled water was poured into both the inlet and outlet of each condensation trap. After 30 min (to simulate the time between extinguishment and sample collection of the condensate), the rinse water was collected from the traps and analyzed using the same test methods as the condensate samples. Baseline sampling was also conducted after the last experiment for both sets of condensation traps. These samples will be referred to as “baseline rinse” samples.
2.2.4 Condensate Collection
Condensate samples were collected after each live-fire experiment was completed and fires were suppressed by pouring liquid directly from each of the the 90 mL reservoirs into the appropriate sampling containers for each method employed (Fig. 2). To minimize the impact of the contaminants evaporating or aging, all samples were collected within 30 min after fire extinguishment. It is possible that loss of some compounds of interest may have occurred during this time or through retention on the tubing and/or reservoir surface, which should be evaluated in future work. All samples were stored in a refrigerator immediately after sample collection and prior to shipment (on ice) to an analytical laboratory under chain of custody. Samples were analyzed for VOCs via method EPA 8260B [46], pH via American Public Health Association (APHA) method SM 4500-H\(^+\)B [47], and anions via method EPA 300.0 [48]. Laboratory method detection and reporting limits are provided in section S.4 of the Supplemental Information.
2.3 Suppression Runoff Water Collection
In order to compare concentrations of the same compounds that may be present in water used during firefighting activities, a sample of runoff water flowing out of the test compartments from fire suppression activities was collected during Experiments 2–5 in plastic buckets and transferred to appropriate sample collection containers within 30 min of fire extinguishment. Runoff water samples were analyzed using the same methods as the condensate. Samples were also collected directly from the suppression hosestream as baseline measurements for the runoff water samples.
3 Results
3.1 Fire Effluent Condensate Results
The total volume of condensate and combustion gases sampled along with peak and average compartment gas temperatures at the sampling locations were calculated over the duration of condensate collection (Table 2). Time resolved plots of gas temperature and oxygen/carbon dioxide/carbon monoxide concentrations are included in section S.5 of the Supplemental Information to provide further detail regarding the conditions within the fire compartment during condensate collection.
Table 2
Fire Effluent Condensate Volumes and Sampling Conditions.
Exp. #
Condensate Volume (mL)
Volume of Combustion Gases Sampled (L)
Peak Inlet Temperature (\(^\circ\)C)
Average Inlet Temperature (\(^\circ\)C)
1
130
2,930
190 ± 60
110 ± 20
2
475
5,810
1,100 ± 10
230 ± 20
3
185
4,440
750 ± 270
140 ± 30
4
255
3,890
830 ± 300
180 ± 30
5
150
1,260
1,470 ± 430
430 ± 60
Volumes are calculated as summations of all (4-6) sampling locations. Temperatures are presented as the mean and standard deviation among every sampling location
Fire effluent condensate from all five structure fire experiments was collected and analyzed for VOCs (Table 3) as well as pH and anion concentrations (Table 4). These tables only include compounds that were detected in at least three of the five experiments, while all compounds detected in each experiment are included in section S.6 (Tables S8 and S9) of the Supplementary Information. All BTEXS compounds were detected in condensate from at least three of the five experiments except for ethyl benzene (only detected in Experiment 4 at 54 \(\mu\)g/L). 1,2,4-Trimethylbenzene, 2-hexanone, and chloromethane were also detected in at least one condensate sample.
Table 3
Volatile Organic Compound (VOC) Concentrations (\(\mu\)g/L) in Condensate from 5 Room-and-contents Fires.
Exp. #
1
2
3
4
5
Benzene
1,100
6,400
2,600
3,600
33,000
Styrene
<400
1,200
470*
1,400
1,800
Toluene
180*
1,000
<340
660
3,900
Xylenes
<290
110*
<740
153
910*
Naphthalene
2,700*
8,100
7,400*
8,100
10,000
2-Butanone
2,100*
3,600*
7,300*
13,000
31,000
Acetone
57,000
31,000
74,000
110,000
250,000
Ethanol
<40,000
<40,000
67,000*
49,000
61,000*
* Result is less than the reporting limit but greater than or equal to the minimum detection limit and the concentration is an approximate value
< Below minimum detection limit (MDL is reported)
Limited to compounds with concentration above detection limit in at least 3 experiments
Table 4
pH and anion Concentrations (mg/L) in Condensate from 5 Room-and-contents Fires.
Exp. #
1
2
3
4
5
pH
2.56
1.10
1.93
1.96
1.59
Bromide
<3.0
5.5
6.6
9.8
13
Chloride
270
39,000
3,000
2,400
4,700
Nitrate
13
2.4
5.7
<1.0
6.4
Sulfate
330
9,200
2,700
2,100
2,300
< Below reporting limit (RL is reported)
Limited to compounds with concentration above detection limit in at least 3 experiments
Baseline rinse water collected prior to each experiment and after the final two experiments had relatively low concentrations of most VOCs and inconsistent detection other than for acetone (detected in 4 of 7 samples; median 15 \(\mu\)g/L) and toluene (detected in 4 of 7 samples; median 0.8 \(\mu\)g/L). Chloride (median 3.8 \(\mu\)g/L) and sulfate (median 2.3 \(\mu\)g/L) were detected in all baseline rinse water samples, which had a median pH of 5.87. A sample of the distilled water used for the baseline rinse had a pH of 7.32 with no detectable VOC or anion contamination. All contaminants detected in any baseline rinse water and the pH of each sample are included in section S.6 (Tables S10 and S11) of the Supplemental Information.
3.2 Suppression Runoff Water Results
The concentration of contaminants in the runoff was notably lower than in the condensate and was basic (pH>7.00) in all experiments (Table 5 and 6). Samples of water from the suppression hose stream had a pH of 6.80 with detectable chloride, sulfate and nitrate concentrations (79, 13, and 0.7 mg/L, respectively). None of the VOCs reported in Table 5 were found in the baseline hose stream water. All contaminants detected in any runoff water and the baseline hose stream sample and the pH of each sample are included in section S.6 (Table S12 and S13) of the Supplemental Information.
Table 5
Volatile Organic Compound (VOC) Concentrations (\(\mu\)g/L) in Runoff Water from 4 Room-and-contents Fires.
Exp. #
2
3
4
5
Benzene
<79
11
10*
68
Styrene
<150
8.0*
<15
98
Acetone
<4,000
1,100
1,900
1,900
* Result is less than the reporting limit but greater than or equal to the minimum detection limit and the concentration is an approximate value
< Below minimum detection limit (MDL is reported)
Limited to compounds with concentration above detection limit in at least 2 experiments
Table 6
pH and anion Concentrations (mg/L) in Runoff Water from 4 Room-and-contents Fires.
Exp. #
2
3
4
5
pH
12.04
9.38
10.39
11.61
Bromide
<7.5
3.4
<3.0
4.6
Chloride
280
230
660
230
Nitrate
<2.5
1.3
1.9
3.2
Sulfate
1,500
7,700
3,000
2,200
< Below reporting limit (RL is reported)
Limited to compounds with concentration above detection limit in at least 2 experiments
4 Discussion and Future Work
Benzene concentrations in the effluent condensate were the highest of the measured BTEXS compounds, similar to typical airborne smoke samples collected from fireground studies [14, 18] and water contamination in samples collected after the Tubbs Fire (benzene samples reported in the highest average concentration) and Camp Fire (benzene samples represented largest number exceeding regulatory limits) [30, 32]. Benzene concentrations from the condensate were orders of magnitude higher than the concentration allowable in drinking water standards [34, 35] and similar to the peak concentration detected in water distribution systems after the Camp Fire, with the highest concentrations from Experiment 5 nearing the peak concentrations from the Tubbs Fire [30]. Toluene and styrene were detected at the next highest concentrations (among the BTEXS) in the condensate, which is again consistent with airborne smoke samples collected from the fireground [14] and the reported contamination in water distribution systems after Tubbs Fire (toluene samples reported with highest number of detection among BTEXS) and Camp Fire (styrene samples reported at the highest maximum concentration of all BTEXS) [30, 32].
In addition to BTEXS, high concentrations of naphthalene were found in the condensate, a compound which is the most abundant PAH that is typically measured in airborne smoke samples collected from structure fires [14, 21] and noted in water samples from contaminated water distribution systems after wildfires [30, 32]. These levels also exceed EPA’s 1-day health advisory for a 10-kg child drinking 1 L/day (500 \(\mu\)g/L) as well as the California MCL of 100 \(\mu\)g/L [34, 35]. In each experiment, acetone was present in the highest concentration of any of the VOCs measured, typically at least an order of magnitude higher than benzene. High concentrations of ethanol and 2-butanone were also detected. Acetone and 2-butanone were reported in the City of Santa Rosa water distribution network monitoring data set after the Tubbs Fire (including concentrations exceeding 100 \(\mu\)g/L in at least one water sample in which benzene was not detected) [30]. In fact, all of the compounds listed in Table S8 other than 2-hexanone and ethanol were reported in at least one sample collected from the water distribution network after the Tubbs Fire [30]. It is possible that other VOCs are also present in the condensate, though dilution of the sample was necessary due to the high concentrations of the reported compounds, such that other compounds present in lower concentrations may then be below detection limits. The condensate was strongly acidic with anion and sulfate concentrations that were two to three orders of magnitude above concentrations found in the baseline rinse water.
In comparison to suppression water runoff, the liquid condensate was produced in much smaller volumes but had markedly higher concentration of VOCs and anions. Compounds measured in the highest concentration in both condensate and runoff (e.g. benzene, acetone, chloride and sulfate compounds) were consistent. Carbon disulfide and trichloroethene were found in low concentrations in the runoff water from one experiment each, but were not detected in the condensate. Carbon disulfide was reported in the City of Santa Rosa water distribution network after the Tubbs Fire, also in low concentrations [30]. The pH of the condensate was strongly acidic (pH 1.10\(-\)2.56) while the runoff water was basic (pH 9.38\(-\)12.04). The cause of this variation in pH should be studied in future work and the impact of highly corrosive condensate on resilience and reconstruction of communities due to corrosion of utilities and infrastructure should be considered.
4.1 From Condensate to Human Health Risks
Further understanding of human health risks from exposure to fire effluent may be informed by expanded study of fire effluent condensate. These initial experiments reinforce concerns for the risk of a number of contaminants condensing out of the fire effluent and being available for dermal absorption. For example, dermal absorption of liquid condensed phase contaminants may help explain why studies that have ensured firefighters wear self contained breathing apparatus (SCBA) throughout their exposure (thus eliminating the inhalation pathway) have still identified systemic exposure to VOCs (particularly BTEXS) and PAHs [20, 24, 25]. Mayer et. al. (2022) [49] have shown that high concentrations of benzene may be present in the air under firefighting PPE. While small concentrations of benzene vapor can be absorbed dermally [50, 51], it is possible that systemic exposure could be enhanced through condensation of benzene to liquid phase on the skin. Furthermore, dermal uptake of contaminants may be influenced by pH and/or presence of compounds such as solvents (e.g. acetone, ethanol) on the skin [52]. The complex mixture of compounds identified in the fire effluent condensate may result in synergistic or antagonistic impacts related to dermal absorption and human health risk. Condensation of VOCs and other contaminants (including PAHs and other semi-volatile organic compounds (sVOCs)) onto human skin may be an important route for human exposure during fire events. Future studies may focus on a more broad analysis of biologically relevant dermal exposure scenarios including the effect of different concentrations of condensed contaminants and complex mixtures with solvents such as acetone on transdermal absorption.
A more detailed understanding of the composition and magnitude of compounds in this highly acidic liquid condensed phase may be useful for further understanding damage to the respiratory tract for those who inhale fire effluent, including individuals who have been rescued from inside burning structures, firefighters who have had their respiratory protection compromised, or even those who have been exposed to wildland/WUI fire smoke over long duration. Fire effluent can impact different depths of the respiratory tract based on properties of the contaminant [53]. Water soluble compounds are likely to deposit in the upper airway and result in upper airway irritation. Compounds that are less soluble may travel further in the airway. For example, aromatic hydrocarbons such as benzene can cause mucus membrane irritation as well as systemic toxicity [54]. Further characterization of the condensed liquid phase including impacts of different effluent compounds and acidity may provide additional insights into risk. Future studies should also characterize condensate from different phases of fire (e.g. pre-flashover, post-flashover, decay, smoldering) to understand how the presence and concentration of contaminants may change throughout the duration of a fire.
4.2 From Condensate to Water Distribution System Contamination
This pilot study provides evidence for the generation of liquid condensate containing a high concentration of VOC contaminants when fire effluent is drawn into plumbing and cooled. Many of these compounds have been found at similar concentrations in contaminated water distribution systems after WUI fires. However, further study is needed to understand how this condensate may result in systemic contamination. For example, concentrations of benzene measured in this study are all above 1,000 \(\mu\)g/L. In cases where the liquid condensate may pool undiluted, high concentrations such as the peak values noted in the Tubbs and Camp Fires may be partially attributed to this source. These peak values and generally the highest levels of contamination were found in the months immediately after the fires [30]. Four months after the Camp Fire (March 2019), the average benzene level sampled from Paradise’s water system was 31-33 \(\mu\)g/L [33], which decreased as the infrastructure was replaced and flushing occurred. The relatively small volume of liquid condensate could be diluted by flowing water from the main during sampling, particularly if larger flushing volumes are used. It is also possible that the high concentration of contaminants from condensate may pool in stagnant locations and begin to adsorb into components of a water distribution system. These compounds may then subsequently desorb back into the water once service is restored in the system [43]. The adsorption of such compounds from condensate and subsequent re-release should be further studied.
Additional research is needed to contextualize the magnitude of condensate produced by fires compared to the levels that may be important for contaminating individual or community water systems. The volumes of condensate collected here depend on the sampling locations and fire conditions (fuel load, ventilation (timing and area), compartment volume, etc.) and do not reflect the total quantity of condensate produced by the fires. As expected, longer duration of post-flashover burning produced greater volumes of condensate when comparing similar fire configurations (Experiment 1 versus 2 and Experiment 3 versus 4). As for comparisons to wildfires, samples collected in this study were from single structure room-and-contents fires, while the Tubbs and Camp Fires involved 5,636 and 18,804 structures respectively along with adjacent vehicles, infrastructure and vegetation [31]. The magnitude of condensate produced from different fire conditions (fuel load, ventilation (timing and area), compartment volume, etc.), larger structure fires, multiple dwelling fires, wildfires, WUI fires, community fires, and complex fires should be studied. Future research may also focus on drawing in fire effluent through configurations that are more commonly encountered in damaged water distribution systems.
Finally, fire effluent condensate and water runoff may have an extended impact on the environment, including source water bodies and catchments. Elevated anion and VOC concentrations as well as acute acidity can be harmful to freshwater aquatic organisms and other wildlife. The contribution of liquid condensate from fire effluent and water runoff to these environmental risks should be further studied. Four of the five condensate samples contained chloride concentrations well above the EPA criteria of 860 mg/L and all resulted in low pH values well outside the chronic limits (pH 6.5\(-\)9.0), though the relatively small volume may be quickly diluted in large bodies of water. Suppression water runoff was collected in much larger volumes and resulted in pH values more basic (pH 9.38\(-\)12.04) than the EPA limits, while chloride concentrations were elevated above baseline values. Other important contaminants such as sVOCs (including many PAHs) have been quantified in the air during residential fires and are likely to be present in fire effluent condensate and runoff water that may result in contamination of these bodies.
5 Conclusion
This pilot study provides an initial characterization of liquid condensate from smoke produced by full-scale residential structure fire scenarios, establishing groundwork for future studies to further understand human health and environmental impacts of these fire events. High concentrations of benzene and naphthalene detected in fire effluent condensate are similar to peak concentrations detected in contaminated water distribution systems after recent wildland fires and may be partially responsible for contaminants that adsorb into pipes and subsequently desorb into drinking water. Elevated concentrations of benzene and other VOCs in the condensed effluent provides insight into an important mechanism for human exposure risks associated with fire smoke. The presence of solvents such as acetone and ethanol in the fire effluent condensate may impact absorption of other contaminants known to be present in fire effluent. The condensates are strongly acidic which may cause human health risk and impact corrosion of infrastructure.
Acknowledgements
The authors thank Keith Stakes, Brad Morrissey, Rebekah Schrader, Daniel Madrzykowski, Craig Weinschenk, and Adam Barowy for their support of this study. Additionally, valuable insights were gained from conversations with several colleagues in medical and industrial hygiene research who helped frame the discussion of future study related to fire effluent condensate.
Declarations
Conflict of interest
None of the authors have any commitments or funding that would cause a conflict of interest for the publication of this manuscript.
Ethical approval
Not applicable.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
