Next Article in Journal
Unveiling PM2.5 Transport Pathways: A Trajectory-Channel Model Framework for Spatiotemporally Quantitative Source Apportionment
Previous Article in Journal
Exploring the Explainability of a Machine Learning Tool to Improve Severe Thunderstorm Wind Reports
Previous Article in Special Issue
Children’s Allergic Sensitization to Pets: The Role of Air Pollution
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Occurrence and Mitigation of PM2.5, NO2, CO and CO2 in Homes Due to Cooking and Gas Stoves

1
School of Science, Technology, Engineering, and Mathematics, University of Washington, Bothell, WA 98011, USA
2
Department of Atmospheric Sciences, University of Washington, Seattle, WA 98195, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2025, 16(7), 882; https://doi.org/10.3390/atmos16070882
Submission received: 29 April 2025 / Revised: 7 July 2025 / Accepted: 14 July 2025 / Published: 18 July 2025

Abstract

We surveyed the air quality conditions in 18 homes with gas stoves for PM2.5, CO2, NO2 and CO using calibrated low-cost sensors. In each home, participants were asked to cook as usual, but to record their cooking activities and mitigation efforts (windows, ventilation fans, etc.). All homes showed enhanced pollutants during, and immediately after, times of cooking or stove use. For each home, we quantified the minutes per day and minutes per minute of cooking over known health thresholds for each pollutant. On average, homes exhibited 38 min per day over one or more of these thresholds, with PM2.5 and NO2 being the pollutants of greatest concern. Six homes had much higher occurrences over the health thresholds, averaging 73 min per day. We found an average of 1.0 min over one or more of the health thresholds per minute of cooking when no mitigation was used, whereas when mitigation was used (filtration or vent fan), this value was reduced by 34%. We further investigated several mitigation methods including natural diffusion, a commercial HEPA filter unit, a commercial O3 scrubber and a ventilation fan. We found that the HEPA unit was highly effective for PM2.5 but had no impact on any of the gaseous pollutants. The O3 scrubber was moderately effective for NO2 but had little impact on the other pollutants. The ventilation fan was highly effective for all pollutants and reduced the average pollutant lifetime significantly. Under controlled test conditions, the pollutant lifetime (or time to reach 37% of the original concentration), was reduced from an average of 45 min (with no ventilation) to 7 min. While no commercial filter showed efficacy for both PM2.5 and NO2, the fact that each could be removed individually suggests that a combined filter for both pollutants could be developed, which would significantly reduce health impacts in homes with gas stoves.

1. Introduction

Gas combustion and cooking emit large amounts of PM2.5, (particulate matter with a diameter of <2.5 µm), NO, NO2, CO, CO2 and volatile organic compounds (VOCs) into homes [1,2,3,4,5]. Gas combustion emits NO, NO2, CO and CO2 [2,3] while cooking, and is mainly responsible for PM2.5 emissions [5,6]. Gas is relied upon as a primary source of fuel for cooking by more than 60% of the global population [7], including over 47 million U.S. households [8]. There are numerous documented health effects from exposure to these pollutants including headaches, nausea, episodes of increased chest pain in people with chronic heart disease and increased risk of respiratory infection [9]. Gruenwald, et al. [10] showed that 12.7% of childhood asthma in the U.S can be attributed to gas stove use. Exposure to high levels of these pollutants have both short-term and long-term health effects. Long-term NO2 exposure is linked to respiratory health issues such as pediatric asthma [11], COPD [12], and lung cancer [13]. Decreased cognitive function and physiological changes are associated with increased CO2 exposure [14] and other pollutants [15]. Kashtan et al. [16] describe health impacts from NO2 and gas stove use and identify home size and occupant behavior as key variables controlling exposure.
Cooking is well known to generate aerosols. In one study by Huboyo et al. [6], the authors compared concentrations of PM2.5 produced via boiling and frying methods and compared concentrations of PM2.5 produced via cooking a low-fat food (tofu) and a high-fat food (chicken). They found that concentrations of PM2.5 increased from the background levels during all cooking events, but that the increase was greatest for frying high-fat foods.
Mitigation techniques, such as ventilation fans, windows or high efficiency particle air filters (HEPA) can reduce pollutant concentration in homes. Oanh et al. [5] identified cooking as an important source of PM2.5, but also found that ventilation played a key role in determining the concentrations. Vented exhaust fans are generally effective at removing pollutants but show large variations in performance [17,18,19]. Xiang et al. [19] found that pollutants reached a maximum concentration soon after cooking, but the return to healthy levels depended on mitigation and air exchange. In households where vented exhaust fans are not possible, air purifiers utilizing HEPA and/or activated carbon filters may be effective at removing PM2.5 and NO2 and O3, respectively [20,21,22,23]. Modifying the activated carbon filters with metal or metal oxides can increase NO2 adsorption [24,25,26]. As O3 and NO2 are chemically rather similar, both are polar oxidizing compounds, therefore we expect that their reactivity on these materials would be similar.
Complete evaluations of indoor air require measurements of a suite of pollutants, but this can be challenging due to the instrumentation. Most research-grade instrumentation take up a large amount space, are noisy and require frequent maintenance (e.g., [2]), thus making it difficult to document normal home behaviors. Low-cost sensors (LCSs) can help as they are smaller, less noisy and require less frequent maintenance. However, they do require careful and routine calibration against reference instruments.
In LCSs, CO and NO2 are usually measured with an electrochemical sensor, whereas non-dispersive IR absorption is used for CO2 and laser scattering for PM2.5. Various studies have evaluated the performance of gaseous LCSs by testing them against reference instruments [27,28,29,30]. Due to sensor aging and exposure to different temperature, humidity and pollutant history, frequent re-calibration is needed [30]. Whitehill et al. [31] found that many CO, CO2 and NO2 sensors performed reasonably without calibration, but there were large variations within one type and some units had poor response and/or high offsets. Thus, individual calibrations are necessary to use LCSs successfully.
In our previous work [3], we used a commercial LCS package which was calibrated on multiple occasions against reference methods, to measure PM2.5, CO, CO2 and NO2 in one home with a gas stove. We showed that during normal cooking activities, concentrations of all pollutants could reach extremely high levels, especially PM2.5 and NO2. In this study, we use a combination of research-grade instruments and LCSs to address the following questions:
(1)
How common is poor air quality in homes with gas stoves?
(2)
Which pollutants are of greatest concern?
(3)
Do normal mitigation methods improve air quality in homes?
(4)
Which mitigation methods provide the best tools to improve indoor air quality (IAQ)?
In the first part of our study, we address the first three questions above, using LCSs to look at IAQ during normal cooking and activities in 18 homes with gas stoves. In the second part of our study, we address the last question, using both the LCS and research-grade instruments to examine the efficacy of different mitigation methods. Our findings demonstrate the importance of mitigation to improve IAQ and highlight the most effective mitigation methods.

2. Methods

2.1. Part 1: Volunteer Home Study

We examined the IAQ in 18 volunteer homes with gas stoves using used calibrated LCSs. The multi-pollutant sensor we used was the TSI Air Assure (TSI Inc., Shoreview, MN, USA), which measures 6 gases (CO, CO2, NO2, SO2, VOCs and O3) plus PM2.5. For this study, we report LCS data only on CO, CO2, NO2 and PM2.5, as the SO2 and VOCs sensors were uncalibrated and the O3 sensor had poor performance in the high NOx kitchen environment. For the TSI LCS, CO and NO2 are measured using electrochemical sensors from SPEC Sensors LLC, models 110-1XX and 110-5XX, respectively (SPEC Sensors LLC, Irvine, CA, USA). The CO2 LCS measurement uses non-dispersive infrared absorption (model SCD30) and the PM2.5 module uses laser scattering (model SPS30), both from Sensiron Inc (Chicago, IL, USA). The sensor manufacturer does its own batch calibrations using standard reference materials. For the electrochemical sensors, the manufacturer also incorporates a temperature sensitivity for each batch. TSI incorporates the manufacturer’s batch calibration data and conducts further tests on each NO2, CO2 and PM2.5 sensor. For CO, the manufacturer’s calibration data are used without recalibration by TSI. (John Parrilli, TSI Inc. personal communication, November 2023). Data from each TSI unit are transmitted by Wi-Fi in real time and subsequently downloaded from the TSI Link website (https://tsilink.com/) (last accessed 13 July 2025) for analysis.
Before and after deploying sensors to the volunteer homes, we further tested and recalibrated each unit with research-grade instruments and reference calibration gases. We used three separate TSI units and each was calibrated independently. For NOx calibration we used a 2B Technologies model 405 NO/NO2 analyzer; for CO calibration we used a Teledyne-API (San Diego, CA, USA) model T300U; for CO2 calibration we used a Licor Inc. model 840A; and for PM2.5 calibration we used a Grimm/Durag Group (Hamburg, Germany) optical particle counter model EDM 180. The NOx, CO and CO2 instruments were calibrated against standard reference gases from commercial suppliers. Each LCS was calibrated at least 3 times, except for CO which was only calibrated once. To be acceptable, a calibration equation must have shown an R2 greater than 0.8. In general, we found that the calibration slopes were fairly stable, but the calibration intercepts were less stable. Because of these drifting intercepts, the data from each home were checked to ensure reasonable minimum values were recorded during clean periods (e.g., near zero for CO, NO2 and PM2.5 and near 450 ppm for CO2). Calibration results are shown in Table 1 below.
Volunteers were identified by an internet posting and word of mouth. To qualify for our study, participants must use a gas stove and/or gas oven for cooking, cook for at least 120 min per week, have Wi-Fi and be willing to log their cooking and mitigation activities. In total, 20 volunteer homes were identified and data were obtained on 18 of these homes. For each home, a calibrated sensor was sent with instructions on how and where to install the sensor, instructions on cooking logs and other information. We instructed volunteers to install the sensor in the same room as their stove, but at a distance of at least 1.7 m from it. We informed participants that an ideal location for the sensor would be at a table in the same room as the stove where family members frequently spend time, as this would give the best information on exposure. Each volunteer home was monitored with the TSI LCS for a minimum of two weeks. Volunteers were asked to ensure no mitigation was performed during the first week and then to use mitigation methods during the second week (e.g., ventilation fans, windows, etc.); however, not all volunteers followed this protocol exactly. Volunteer logs documented their cooking and mitigation actions each day. Volunteers were free to cook as normal and we did not attempt to identify the specific types of foods that were cooked. For each home, the data were retrieved as one-minute averages from the TSI Link website. The most recent sensor calibrations were applied, and the calibrated data were checked, especially minimum concentrations, which could indicate a problem with the intercept correction. From the cooking logs we quantified the amount of cooking and the minutes per day over key thresholds for each pollutant and whether mitigation was used or not. These thresholds (shown in Table 2) are based on U.S. EPA ambient (outdoor) air quality standards, except for CO2, where we use 2000 ppm, a level that is clearly associated with health impacts [32,33].

2.2. Part 2: Home Characterization Study

For one home that had generally very high pollutant levels, we performed further tests to evaluate several mitigation measures, using the reference instruments mentioned above, plus the TSI LCS (following calibration) to measure CO, CO2, NO, NO2 and PM2.5. For these tests, we enclosed the main kitchen/dining room area, including the gas stove, with plastic sheets to minimize air flow and exchange and obtain more reproducible results. We calculated the area and volume of this space to be 17.2 m2 and 46 m3, respectively. Reference instruments and three TSI LCSs were placed on a table near the middle of this space and a ventilation fan or a filter unit, depending on the test, was used for mixing and/or filtration during each test. In early tests, we confirmed that this arrangement gave similar pollutant accumulation and emission factors as in our earlier work in the same space with the same stove [3].
We evaluated four different mitigation measures:
  • Natural ventilation, diffusion and/or reaction on the room materials.
  • A stove exhaust fan.
  • A commercial catalytic O3 scrubber unit (CDA250) to test for NO2 removal.
  • A commercial HEPA air purifier (Coway Airmega AP-1512HH Mighty).
We expected that the natural ventilation in this space to be minimal, given the attempt to minimize air flow. However, reactions on the walls and kitchen materials could also contribute to pollutant reductions [34,35,36]. The stove exhaust fan was located directly above the gas stove/oven and vented outside. Its rated exhaust flow was 8.5 m3 min−1, however we measured its flow to be 7.9 m3 min−1. The O3 scrubber (CDA250) is a commercial unit manufactured by Oxidation Technologies LLC (Inwood, IA, USA). It uses an aluminum honeycomb filter coated with a MnO catalyst and is specified as 90% efficient for O3 removal at a flow of 250 cubic feet per minute or 7.1 m3 min−1. While this unit is not intended for residential use (it is rather noisy), we wanted to examine whether this type of unit could reduce NO2 concentrations from gas stoves. We measured the removal efficiency for NO2 directly by sampling from the intake and exhaust ports during periods of high NO2 and found an average NO2 removal of 65%. Finally, we tested a commercial HEPA air purifier (Airmega AP-1512HH Mighty) from Coway Inc (Seoul, Republic of Korea). The unit flow is 6.8 m3 min−1 and we measured a PM2.5 removal of 92% during periods of high concentrations. The Coway Airmega unit has an “activated carbon” pre-filter in front of the HEPA filter; however, the manufacturer provides no specific information on the efficacy of this component. Potentially, this activated carbon pre-filter could reduce NO2 and other gases.
For each mitigation test, the room was filled with a high level of pollutants from the stove (for gases) or cooking (for PM2.5) and the rate of change (dX/dt) was used to obtain the first order removal rate constant (k) in min−1 for each of the pollutants measured. During these tests, the pollutant time series generally followed an exponential decay given by the following equation:
X ( t )   =   X ( 0 )     e k t
where X(0) is the initial concentration, X(t) is the concentration at time t and k is the first order decay rate. An example of this decay is shown in Figure 1.
The k values were determined two ways:
  • By quantifying the time for 50% removal of a given pollutant.
  • By averaging the decay (ΔX/X(t)) for each time step over the measurement period.
For the first method, we can calculate k via the following equation:
k = 0.693/time1/2
And for the second method we can use the following:
k = ln [ΔX/X(t)]/time
These differences between the two calculations were small, between 0–7%, depending on the tests. The pollutant lifetime is given by 1/k. The k values can also be related to the Clean Air Delivery Rate (CADR, in m3 min−1). The CADR is a measure of the clean air flow for an air purifier or other mitigation tool. The CADR can be calculated by either of two methods:
CADR = scrubber flow rate * removal efficiency
CADR = k * room volume

3. Results

3.1. Part 1: Volunteer Home Study

LCSs were installed in 18 volunteer homes for 2 weeks or more, while participants carried out normal cooking. All volunteers used some form of mitigation at some point during the experiment, while 14 homes also carried out unmitigated cooking for a part of the study period. All homes showed elevated pollutant concentrations, especially during the late afternoon/early evening time periods, when cooking was most pronounced. Figure 2 shows a typical pattern for 2 days in one home. Concentrations were frequently elevated, especially for PM2.5 and NO2, especially around mealtimes due to cooking. In this case, PM2.5 and NO2 were elevated in the afternoons of November 1st and 2nd. However, while these two pollutants were generally elevated at similar times of day, closer examination reveals that the pollutants are uncorrelated in time, suggesting different sources. Jaffe and Creekmore [3] found that gas stove combustion emits large amounts of CO, CO2 and NO2, but no detectable PM2.5. In contrast, cooking is a significant source of PM2.5 aerosols. This is further supported by the observations, which showed a very good correlation using the one-minute data between the gaseous species CO, CO2, and NO2 (R2 = 0.5−0.8), but very poor correlation between PM2.5 and any of the gaseous species (R2 < 0.05).
To quantify the occurrence of high pollutants, we quantified the minutes per day over known health thresholds for each pollutant. These thresholds, along with the average minutes per day of cooking in each home and the minutes per day over the health thresholds for each pollutant are shown in Table 2, with PM2.5 and NO2 being the pollutants of greatest concern. The results were at first surprising since the minutes over the standards in certain homes (and the average) for some pollutants went up when mitigation was used. However, these data need to be adjusted for the amount of cooking by the volunteer participants, which was greater when mitigation was used. Table 3 shows the minutes over each health threshold per minute of cooking, averaged across all participants. By doing this adjustment, we can now see the true impact of mitigation efforts. Overall, mitigation reduced the time over the standards per minute of cooking by 34%.
On average, we found that these homes averaged 38 min per day over one or more of the health standards (for either mitigated or the unmitigated cases), but a few homes had much higher occurrences. In six homes, the mean (unmitigated) value over one or more of the standards was 73 min per day, with the highest home having 157 min per day over one or more thresholds. These homes do a lot of cooking and likely have poor ventilation. Therefore, adequate ventilation and mitigation measures are essential to reduce occupants’ exposure to harmful pollutants, especially for these homes.

3.2. Results Part 2: Home Characterization Study

In the worst home, with poor air quality (157 min per day over one or more thresholds for the unmitigated days and 53 min per day for mitigated days), we examined four types of mitigations.
  • Natural ventilation, diffusion and/or reaction on the room materials.
  • A stove exhaust fan.
  • A commercial catalytic O3 scrubber unit (CDA250).
  • A commercial HEPA air purifier (Coway Airmega AP-1512HH Mighty).
Figure 1 shows an example of one of these tests and Table 4 shows a summary of the calculated k values. During this test (Figure 2), only natural ventilation occurred at the start and so the pollutant concentrations declined slowly. At 14:33 local time, the Coway HEPA filtration unit was turned on and concentrations of PM2.5 declined much more rapidly. However, the concentrations of CO, CO2 and NO2 showed no additional reduction. For these tests, with the Coway HEPA unit, we calculated first order decay rates of 0.13 min−1 for PM2.5 and 0.01–0.03 min−1 for the gases. These k values for the gases are essentially indistinguishable from the natural ventilation tests and demonstrate that while the HEPA is excellent at reducing PM2.5, the HEPA and activated carbon pre-filter have little or no impact on these gases. The exhaust fan gave much faster removal rates for all pollutants, with k values of 0.13–0.16 min−1. The O3 scrubber (CDA250) showed moderate removal for NO2 (k value of 0.09 min−1), but little efficacy for the other pollutants. The stove exhaust fan showed nearly identical behavior for PM2.5, NO and CO2, with k values of 0.13 min−1 for each. NO2 showed a higher k value, likely reflecting additional removal due to reactivity on surfaces.
Using the measured k values from Table 4, we can calculate the efficacy of each mitigation method, as indicated by the CADR. Table 5 shows the CADR values calculated by two different methods, either the flow rate and scrubber efficiency (Equation (4)) or the measured k values and room volume (Equation (5)). For the Coway HEPA and the CDA250, the CADR values are only reported for PM2.5 and NO2, respectively, since these are the only pollutants that each unit removes to any significant degree. While ventilation does not directly “filter” the air, it does provide a flow of relatively clean air by pulling in outside air. For this calculation, we assume that the concentrations in the outside air are 10% or less of those in the indoor air, during periods of poor IAQ. In all cases, the two CADR calculation methods show generally good agreement. These results demonstrate that each of these methodologies (HEPA, catalytic scrubber or ventilation fan) can result in significant improvements in IAQ, but that only the ventilation fan impacts all pollutants. In the case of the catalytic scrubbers, new developments would be needed to incorporate these into commercial air filtration products. Our data show that the “activated carbon” filter in one commercial product (Coway) has little impact on NO2. This is in contrast to results from Matthaios et al. [23], who show that some activated carbon filters are effective for NO2 in vehicles. Future work should resolve these differences in filter types, as this could potentially lead to simple filtration methods for NO2 in homes with gas stoves.

4. Conclusions

We surveyed the air quality conditions in 18 homes with gas stoves for PM2.5, CO2, NO2 and CO using calibrated low-cost sensors. In each home, participants were asked to cook as usual, but to record their cooking activities and mitigation efforts (windows, ventilation fans, etc.). All homes showed enhanced pollutants during, and immediately after, times of cooking or stove use. We quantified the minutes per day that each home was above known health thresholds for each pollutant. We found an average of 38 min per day in each home over one or more of these thresholds on days when no mitigation was used, with PM2.5 and NO2 being the pollutants of greatest concern. However, in six of the homes we surveyed, the time with concentrations over the health thresholds was much higher, averaging 73 min per day when no mitigation was used and one home averaged 157 min per day over these thresholds. To compare unmitigated and mitigated periods, we quantified the minutes over the standards per minute of cooking. For unmitigated periods, we found an average of 1.0 min over one or more standards per minute of cooking, whereas for mitigated periods, this value was reduced by 34%.
In one of our test homes, we further investigated several mitigation methods including natural diffusion, a commercial HEPA filter unit, a commercial O3 scrubber and a ventilation fan. We found that the HEPA unit was highly effective for PM2.5 but had no impact on any of the gaseous pollutants. The O3 scrubber (CDA250) was moderately effective for NO2 but had little impact on the other pollutants. The ventilation fan was highly effective for all pollutants and reduced the average pollutant lifetime significantly. Under controlled tests, the pollutant lifetime (or time to reach 37% of the original concentration), was reduced from an average of 45 min (with no ventilation) to 7 min. Thus, for all homes, but especially those with gas stoves, we recommend the use of both a ventilation fan and a HEPA filter unit during cooking and for at least 20 min after cooking is completed.
We note that our results did not consider the impacts of building layout (e.g., Jiao et al. 2025 [37]) or population demographics (e.g., Schachter et al. 2020 [38]).
These results demonstrate that each of these methodologies (HEPA, catalytic scrubber or ventilation fan) can result in significant improvements in IAQ, but that only the ventilation fan impacts all pollutants. In the case of the catalytic scrubbers, new developments would be needed to incorporate these into commercial air filtration products. Our data show that the “activated carbon” filter in one commercial product (Coway) has little impact on NO2, but a unit with a MnO catalyst had significant NO2 removal. Further evaluation and development of commercial units for homes that scrub both PM2.5 and NO2 could provide significant health benefits for homes with gas stoves.

Author Contributions

Conceptualization, D.J.; Validation, D.N. and S.B.; Formal analysis, D.J., D.N. and S.B.; Writing—original draft, D.J. and D.N. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by a grant from the UW Bothell Scholarship, Research, and Creative Practice Seed Grant program.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors acknowledge funding from the UW Bothell Scholarship, Research, and Creative Practice Seed Grant program. We want to thank the Washington State Department of Ecology for the loan of the reference CO instrument and the University of Montana for loan of the reference NOx instrument.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Singer, B.C.; Pass, R.Z.; Delp, W.W.; Lorenzetti, D.M.; Maddalena, R.L. Pollutant concentrations and emission rates from natural gas cooking burners without and with range hood exhaust in nine California homes. Build. Environ. 2017, 122, 215–229. [Google Scholar] [CrossRef]
  2. Lebel, E.D.; Finnegan, C.J.; Ouyang, Z.; Jackson, R.B. Methane and NOx Emissions from Natural Gas Stoves, Cooktops, and Ovens in Residential Homes. Environ. Sci. Technol. 2022, 56, 2529–2539. [Google Scholar] [CrossRef] [PubMed]
  3. Jaffe, D.A.; Creekmore, A. Emissions and exposure to NOx, CO, CO2 and PM2.5 from a gas stove using reference and low-cost sensors. Atmos. Environ. 2024, 331, 120564. [Google Scholar] [CrossRef]
  4. Kashtan, Y.S.; Nicholson, M.; Finnegan, C.; Ouyang, Z.; Lebel, E.D.; Michanowicz, D.R.; Shonkoff, S.B.C.; Jackson, R.B. Gas and Propane Combustion from Stoves Emits Benzene and Increases Indoor Air Pollution. Environ. Sci. Technol. 2023, 57, 9653–9663. [Google Scholar] [CrossRef] [PubMed]
  5. Kim Oanh, N.T.; Huy, L.N.; Maneepatra, W.; Winijkul, E.; Giandomenico, A.; Tantrakarnapa, K.; Co, H.X.; Cuong, D.M.; Tsou, M.-C.M.; Hien, T.T.; et al. Comparative analysis of PM2.5 levels in various microenvironments associated with common cooking practices in selected Asian countries. Air Qual. Atmos. Health 2024, 17, 2967–2984. [Google Scholar] [CrossRef]
  6. Huboyo, H.S.; Tohno, S.; Cao, R. Indoor PM2.5 Characteristics and CO Concentration Related to Water-Based and Oil-Based Cooking Emissions Using a Gas Stove. Aerosol Air Qual. Res. 2011, 11, 401–411. [Google Scholar] [CrossRef]
  7. World Health Organization (WHO). Database: Cooking Fuels and Technologies. Available online: https://www.who.int/data/gho/data/themes/air-pollution/cooking-fuel-and-technology-database-by-fuel-category (accessed on 28 April 2025).
  8. U.S. Energy Information Administration, U.S.E.I.A. Residential Energy Consumption Survey (RECS). 2020 REVS Survey Data. Available online: https://www.eia.gov/consumption/residential/data/2020/#fueluses (accessed on 13 July 2025).
  9. Environmental Protection Agency (EPA). Sources of Combustion Products. Available online: https://www.epa.gov/indoor-air-quality-iaq/sources-combustion-products (accessed on 28 April 2025).
  10. Gruenwald, T.; Seals, B.A.; Knibbs, L.D.; Hosgood, H.D. Population Attributable Fraction of Gas Stoves and Childhood Asthma in the United States. Int. J. Environ. Res. Public Health 2023, 20, 75. [Google Scholar] [CrossRef] [PubMed]
  11. Achakulwisut, P.; Brauer, M.; Hystad, P.; Anenberg, S.C. Global, national, and urban burdens of paediatric asthma incidence attributable to ambient NO2 pollution: Estimates from global datasets. Lancet Planet. Health 2019, 3, e166–e178. [Google Scholar] [CrossRef] [PubMed]
  12. Zhang, Z.; Wang, J.; Lu, W. Exposure to nitrogen dioxide and chronic obstructive pulmonary disease (COPD) in adults: A systematic review and meta-analysis. Environ. Sci. Pollut. Res. Int. 2018, 25, 15133–15145. [Google Scholar] [CrossRef] [PubMed]
  13. Chen, Z.; Liu, N.; Tang, H.; Gao, X.; Zhang, Y.; Kan, H.; Deng, F.; Zhao, B.; Zeng, X.; Sun, Y.; et al. Health effects of exposure to sulfur dioxide, nitrogen dioxide, ozone, and carbon monoxide between 1980 and 2019: A systematic review and meta-analysis. Indoor Air 2022, 32, e13170. [Google Scholar] [CrossRef] [PubMed]
  14. Allen, J.G.; MacNaughton, P.; Satish, U.; Santanam, S.; Vallarino, J.; Spengler, J.D. Associations of Cognitive Function Scores with Carbon Dioxide, Ventilation, and Volatile Organic Compound Exposures in Office Workers: A Controlled Exposure Study of Green and Conventional Office Environments. Environ. Health Perspect. 2016, 124, 805–812. [Google Scholar] [CrossRef] [PubMed]
  15. MacNaughton, P.; Spengler, J.; Vallarino, J.; Santanam, S.; Satish, U.; Allen, J. Environmental perceptions and health before and after relocation to a green building. Build. Environ. 2016, 104, 138–144. [Google Scholar] [CrossRef] [PubMed]
  16. Kashtan, Y.; Nicholson, M.; Finnegan, C.J.; Ouyang, Z.; Garg, A.; Lebel, E.D.; Rowland, S.T.; Michanowicz, D.R.; Herrera, J.; Nadeau, K.C.; et al. Nitrogen dioxide exposure, health outcomes, and associated demographic disparities due to gas and propane combustion by U.S. stoves. Sci. Adv. 2024, 10, eadm8680. [Google Scholar] [CrossRef] [PubMed]
  17. Singer, B.C.; Delp, W.W.; Price, P.N.; Apte, M.G. Performance of installed cooking exhaust devices. Indoor Air 2012, 22, 224–234. [Google Scholar] [CrossRef] [PubMed]
  18. Sun, L.; Wallace, L.A.; Dobbin, N.A.; You, H.; Kulka, R.; Shin, T.; St-Jean, M.; Aubin, D.; Singer, B.C. Effect of venting range hood flow rate on size-resolved ultrafine particle concentrations from gas stove cooking. Aerosol Sci. Technol. 2018, 52, 1370–1381. [Google Scholar] [CrossRef]
  19. Xiang, J.; Hao, J.; Austin, E.; Shirai, J.; Seto, E. Residential cooking-related PM2.5: Spatial-temporal variations under various intervention scenarios. Build. Environ. 2021, 201, 108002. [Google Scholar] [CrossRef] [PubMed]
  20. Chen, C.-F.; Hsu, C.-H.; Chang, Y.-J.; Lee, C.-H.; Lee, D.L. Efficacy of HEPA Air Cleaner on Improving Indoor Particulate Matter 2.5 Concentration. Int. J. Environ. Res. Public Health 2022, 19, 11517. [Google Scholar] [CrossRef] [PubMed]
  21. Paulin, L.M.; Diette, G.B.; Scott, M.; McCormack, M.C.; Matsui, E.C.; Curtin-Brosnan, J.; Williams, D.L.; Kidd-Taylor, A.; Shea, M.; Breysse, P.N.; et al. Home interventions are effective at decreasing indoor nitrogen dioxide concentrations. Indoor Air 2014, 24, 416–424. [Google Scholar] [CrossRef] [PubMed]
  22. Riederer, A.M.; Krenz, J.E.; Tchong-French, M.I.; Torres, E.; Perez, A.; Younglove, L.R.; Jansen, K.L.; Hardie, D.C.; Farquhar, S.A.; Sampson, P.D.; et al. Effectiveness of portable HEPA air cleaners on reducing indoor PM2.5 and NH3 in an agricultural cohort of children with asthma: A randomized intervention trial. Indoor Air 2021, 31, 454–466. [Google Scholar] [CrossRef] [PubMed]
  23. Matthaios, V.N.; Rooney, D.; Harrison, R.M.; Koutrakis, P.; Bloss, W.J. NO2 levels inside vehicle cabins with pollen and activated carbon filters: A real world targeted intervention to estimate NO2 exposure reduction potential. Sci. Total Environ. 2023, 860, 160395. [Google Scholar] [CrossRef] [PubMed]
  24. Kante, K.; Deliyanni, E.; Bandosz, T.J. Interactions of NO2 with activated carbons modified with cerium, lanthanum and sodium chlorides. J. Hazard. Mater. 2009, 165, 704–713. [Google Scholar] [CrossRef] [PubMed]
  25. Levasseur, B.; Gonzalez-Lopez, E.; Rossin, J.A.; Bandosz, T.J. Effect of Reduction Treatment on Copper Modified Activated Carbons on NOx Adsorption at Room Temperature. Langmuir 2011, 27, 5354–5365. [Google Scholar] [CrossRef] [PubMed]
  26. Sager, U.; Däuber, E.; Bathen, D.; Asbach, C.; Schmidt, F.; Tseng, J.-C.; Pommerin, A.; Weidenthaler, C.; Schmidt, W. Influence of the degree of infiltration of modified activated carbons with CuO/ZnO on the separation of NO2 at ambient temperatures. Adsorpt. Sci. Technol. 2016, 34, 307–319. [Google Scholar] [CrossRef]
  27. Malings, C.; Tanzer, R.; Hauryliuk, A.; Kumar, S.P.N.; Zimmerman, N.; Kara, L.B.; Presto, A.A.; Subramanian, R. Development of a general calibration model and long-term performance evaluation of low-cost sensors for air pollutant gas monitoring. Atmos. Meas. Tech. 2019, 12, 903–920. [Google Scholar] [CrossRef]
  28. Zimmerman, N.; Presto, A.A.; Kumar, S.P.N.; Gu, J.; Hauryliuk, A.; Robinson, E.S.; Robinson, A.L.; Subramanian, R. A machine learning calibration model using random forests to improve sensor performance for lower-cost air quality monitoring. Atmos. Meas. Tech. 2018, 11, 291–313. [Google Scholar] [CrossRef]
  29. Tryner, J.; Phillips, M.; Quinn, C.; Neymark, G.; Wilson, A.; Jathar, S.H.; Carter, E.; Volckens, J. Design and testing of a low-cost sensor and sampling platform for indoor air quality. Build. Environ. 2021, 206, 108398. [Google Scholar] [CrossRef] [PubMed]
  30. Miech, J.A.; Stanton, L.; Gao, M.; Micalizzi, P.; Uebelherr, J.; Herckes, P.; Fraser, M.P. In situ drift correction for a low-cost NO2 sensor network. Environ. Sci. Atmos. 2023, 3, 894–904. [Google Scholar] [CrossRef]
  31. Whitehill, A.R.; Long, R.W.; Urbanski, S.P.; Colón, M.; Habel, A.; Landis, M.S. Evaluation of Cairpol and Aeroqual Air Sensors in Biomass Burning Plumes. Atmosphere 2022, 13, 877. [Google Scholar] [CrossRef] [PubMed]
  32. Jacobson, T.A.; Kler, J.S.; Hernke, M.T.; Braun, R.K.; Meyer, K.C.; Funk, W.E. Direct human health risks of increased atmospheric carbon dioxide. Nat. Sustain. 2019, 2, 691–701. [Google Scholar] [CrossRef]
  33. Satish, U.; Mendell Mark, J.; Shekhar, K.; Hotchi, T.; Sullivan, D.; Streufert, S.; Fisk William, J. Is CO2 an Indoor Pollutant? Direct Effects of Low-to-Moderate CO2 Concentrations on Human Decision-Making Performance. Environ. Health Perspect. 2012, 120, 1671–1677. [Google Scholar] [CrossRef] [PubMed]
  34. Spicer, C.W.; Coutant, R.W.; Ward, G.F.; Joseph, D.W.; Gaynor, A.J.; Billick, I.H. Rates and mechanisms of NO2 removal from indoor air by residential materials. Environ. Int. 1989, 15, 643–654. [Google Scholar] [CrossRef]
  35. Ault, A.P.; Grassian, V.H.; Carslaw, N.; Collins, D.B.; Destaillats, H.; Donaldson, D.J.; Farmer, D.K.; Jimenez, J.L.; McNeill, V.F.; Morrison, G.C.; et al. Indoor Surface Chemistry: Developing a Molecular Picture of Reactions on Indoor Interfaces. Chem 2020, 6, 3203–3218. [Google Scholar] [CrossRef] [PubMed]
  36. Liu, J.; Deng, H.; Lakey, P.S.J.; Jiang, H.; Mekic, M.; Wang, X.; Shiraiwa, M.; Gligorovski, S. Unexpectedly High Indoor HONO Concentrations Associated with Photochemical NO2 Transformation on Glass Windows. Environ. Sci. Technol. 2020, 54, 15680–15688. [Google Scholar] [CrossRef] [PubMed]
  37. Jiao, X.; Zhou, L.; Zhao, W.; Yuan, W.; Yang, B.; Zhang, L.; Huang, W.; Long, S.; Xu, J.; Shen, H.; et al. Significant Cross-Contamination Caused by Cooking Fume Transport between Dwelling Units in Multilayer Buildings. Environ. Sci. Technol. 2025, 59, 9665–9675. [Google Scholar] [CrossRef] [PubMed]
  38. Schachter, E.N.; Rohr, A.; Habre, R.; Koutrakis, P.; Moshier, E.; Nath, A.; Coull, B.; Grunin, A.; Kattan, M. Indoor air pollution and respiratory health effects in inner city children with moderate to severe asthma. Air Qual. Atmos. Health 2020, 13, 247–257. [Google Scholar] [CrossRef]
Figure 1. Removal of PM2.5 and CO2, (top) and CO and NO2 (bottom) by ventilation and HEPA filter. The HEPA filtered was initiated at 14:33 (LT), as shown by the vertical line. The data is from volunteer home #20 on 17 November 2024.
Figure 1. Removal of PM2.5 and CO2, (top) and CO and NO2 (bottom) by ventilation and HEPA filter. The HEPA filtered was initiated at 14:33 (LT), as shown by the vertical line. The data is from volunteer home #20 on 17 November 2024.
Atmosphere 16 00882 g001
Figure 2. Typical pattern for NO2 and PM2.5 inside one home for 2 days (volunteer home #20). The health thresholds we used are 100 ppb for NO2, shown as blue dashed line, and 35 µg m−3 for PM2.5, shown as the red dashed line.
Figure 2. Typical pattern for NO2 and PM2.5 inside one home for 2 days (volunteer home #20). The health thresholds we used are 100 ppb for NO2, shown as blue dashed line, and 35 µg m−3 for PM2.5, shown as the red dashed line.
Atmosphere 16 00882 g002
Table 1. Calibration results for the TSI LCSs. Calibration equations are given as LCS values = Reference value × slope + intercept.
Table 1. Calibration results for the TSI LCSs. Calibration equations are given as LCS values = Reference value × slope + intercept.
PM2.5
(n = 3 per Sensor)
CO2
(n = 3 per Sensor)
CO
(n = 1 per Sensor)
NO2
(n = 3 per Sensor)
Calibration (min, max)0.58, 0.770.80, 0.841.3, 1.50.55, 0.98
Intercepts (min, max)−7.97, 0.97 µg m−3 −55, 201 ppm0.1, 0.6 ppm−11, 63 ppb
Table 2. Average cooking and minutes per day over each of the health thresholds, with and without mitigation. These averages exclude days with no reported cooking. Note that not all homes included observations with no mitigation. The health threshold we applied is shown in parentheses for each pollutant.
Table 2. Average cooking and minutes per day over each of the health thresholds, with and without mitigation. These averages exclude days with no reported cooking. Note that not all homes included observations with no mitigation. The health threshold we applied is shown in parentheses for each pollutant.
Minutes per Day CookingPM2.5
(35 µg m−3)
CO2
(2000 ppm)
CO
(9.4 ppm)
NO2
(100 ppb)
Any
Threshold
No mitigation (n = 14)47.724.98.10.615.037.6
With mitigation (n = 17)64.331.02.70.27.338.4
Table 3. Average minutes over health threshold per minute of cooking with and without mitigation. Note that not all homes included observations with no mitigation. These averages exclude days with no reported cooking.
Table 3. Average minutes over health threshold per minute of cooking with and without mitigation. Note that not all homes included observations with no mitigation. These averages exclude days with no reported cooking.
PM2.5 CO2 CO NO2Any Threshold
No mitigation (n = 14)0.820.120.0070.261.03
With mitigation (n = 17)0.550.040.0030.110.68
% reduction3268535634
Table 4. First order removal coefficients (k, min−1) measured using various mitigation methods.
Table 4. First order removal coefficients (k, min−1) measured using various mitigation methods.
PM2.5NO2NOCO2
Natural ventilation, diffusion and/or reaction on room materials (n = 3)0.030.02not tested0.02
Exhaust fan (n = 3)0.130.160.130.13
CDA250 (ozone scrubber) (n = 3)not tested0.090.030.04
Coway HEPA (n = 3)0.130.03not tested0.01
Table 5. Measured removal efficiencies and calculated CADR values for PM2.5 and NO2. 1 CADR = scrubber flow rate × removal efficiency’; 2 CADR = k × room vol. 3 For the ventilation fan, we use a scrubber efficiency of 0.9, which assumes that the incoming clean air has >90% lower pollutant concentrations than the indoor air. All k values are from Table 4.
Table 5. Measured removal efficiencies and calculated CADR values for PM2.5 and NO2. 1 CADR = scrubber flow rate × removal efficiency’; 2 CADR = k × room vol. 3 For the ventilation fan, we use a scrubber efficiency of 0.9, which assumes that the incoming clean air has >90% lower pollutant concentrations than the indoor air. All k values are from Table 4.
Mitigation MethodCoway HEPACDA250Ventilation Fan 3
PollutantsPM2.5NO2all
Room volume (m3)464646
Measured scrubber efficiency0.920.650.93
Flow rate (m3/min)6.87.17.9
CADR from flow and scrubber efficiency 1 (m3/min)6.24.67.1
CADR from decay tests 2 (m3/min)5.94.16.7
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Jaffe, D.; Nirschl, D.; Birman, S. Occurrence and Mitigation of PM2.5, NO2, CO and CO2 in Homes Due to Cooking and Gas Stoves. Atmosphere 2025, 16, 882. https://doi.org/10.3390/atmos16070882

AMA Style

Jaffe D, Nirschl D, Birman S. Occurrence and Mitigation of PM2.5, NO2, CO and CO2 in Homes Due to Cooking and Gas Stoves. Atmosphere. 2025; 16(7):882. https://doi.org/10.3390/atmos16070882

Chicago/Turabian Style

Jaffe, Daniel, Devon Nirschl, and Stephanie Birman. 2025. "Occurrence and Mitigation of PM2.5, NO2, CO and CO2 in Homes Due to Cooking and Gas Stoves" Atmosphere 16, no. 7: 882. https://doi.org/10.3390/atmos16070882

APA Style

Jaffe, D., Nirschl, D., & Birman, S. (2025). Occurrence and Mitigation of PM2.5, NO2, CO and CO2 in Homes Due to Cooking and Gas Stoves. Atmosphere, 16(7), 882. https://doi.org/10.3390/atmos16070882

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop