Exploring the Relative Effects of Natural Gas and Biogas Cooking on Indoor Air Quality in Residential Kitchens

Abstract

Indoor air pollution from gas stove combustion remains a public health concern, given links to adverse cardiorespiratory health effects, yet few studies have characterized or compared the air quality impacts of different gas-based cooking fuels. We investigated kitchen-level concentrations of nitrogen dioxide (NO₂) and fine particulate matter (PM_2.5) concentrations in four homes in Central Pennsylvania that used natural gas and/or biogas fueled stoves. We conducted time-resolved kitchen monitoring and assessed pollutant concentrations during cooking and non-cooking periods. We applied linear mixed-effect regression models with kitchen-level random effects and time-varying covariates to estimate the influence of fuel type on indoor air quality. During cooking, mean kitchen NO₂ concentrations during cooking were more than 160% higher in homes using natural gas compared with biogas (95% confidence interval [CI]: 109.4%, 211.1%), although both levels remained below the WHO guideline. PM_2.5 concentrations showed limited sensitivity to fuel type, with modest differences observed. Adjusted mixed-effect regression models revealed attenuated but consistent associations, with natural gas use increasing NO₂ exposure by 2.8 ppb, or 60.3% (95% CI: 1.7, 4.6 ppb). These findings suggest further research into understanding the exposure and health benefits of alternative fuels in residential kitchen settings is merited.

1. Introduction

Household air pollution is a well-recognized consequence of cooking with gas stoves in the United States, where natural gas remains a widely used fuel [1,2,3]. Among the pollutants generated, nitrogen dioxide (NO₂) is of particular concern, given the established links between exposure to elevated levels and adverse respiratory and cardiovascular outcomes [3,4,5,6,7]. Yet despite these concerns, NO₂ emissions from natural gas combustion have continued to receive limited public health attention. Similarly, the potential benefits of cleaner cooking fuel alternatives remain poorly characterized. Biogas has been proposed as a renewable energy substitute with potential environmental benefits [8,9]. In cooking applications, the presence of carbon dioxide (CO₂) in biogas acts as a thermal ballast, absorbing heat and slowing combustion, resulting in a lower flame temperature and potentially fewer high-thermal combustion byproducts [10,11]. However, the extent to which biogas reduces harmful indoor air pollutants remains underexplored. Most evidence comes from low-resource settings where biogas replaces solid fuels [12], with little known about its relative impact in higher-resource environments.

To address some of these gaps, this pilot study evaluated the association between biogas or natural gas fuel use and indoor air quality in four residential kitchens in Central Pennsylvania, using high-resolution monitoring and controlling for ambient and kitchen-level factors.

2. Materials and Methods

This study was informed by prior field research methods evaluating the contribution of household stoves to indoor air quality. We sampled four household kitchens using natural gas and/or biogas fueled stoves in Central Pennsylvania. Kitchens were selected non-randomly based on convenience and existing relationships with households participating in anaerobic biodigester projects in the region, given the relative rarity of biogas use in U.S. residential settings. Our selection included primarily suburban or small-town homes located in a mixed-use region with small- to medium-scale farms, moderate local traffic, and scattered industrial or warehouse facilities. Conditions during the study period were typically warm and humid, with occasional heat waves. Biogas for cooking fuel among participants was mainly derived from similar sources—farm animal manure and food waste.

We collected data during 72 h periods in the summer months (June–July) of 2024. Stoves were typically 2-burner (biogas) or 4-burner (natural gas) models, and all four households had kitchens in a room separate from other rooms in the house, with a study median kitchen size of 37.3 m². We instructed participants in the households to use the stoves as they usually would during the monitoring period.

Our primary outcome was indoor NO₂ exposure, assessed in a 72 h window using time-resolved and time-integrated sampling methods. Minute-resolved NO₂ concentrations in the kitchen area were measured using Home Health Box (HHB) aerosol samplers (Access Sensor Technologies, Fort Collins, CO, USA) placed on kitchen countertops, approximately 1–2 m from the stove combustion area. We conducted concurrent time-integrated passive NO₂ sampling using Ogawa passive samplers (Ogawa USA, Pompano Beach, FL, USA), analyzed by RTI International (Research Triangle Park, NC, USA). Outdoor air quality and meteorological conditions (temperature and humidity) were monitored using a uRADMonitor A4 station (MagnaSCI SRL, Dumbravita, Romania). Stove use was determined using electronic stove use monitors (SUMs; Lascar Electronics, Erie, PA, USA) that recorded temperature fluctuations over time. Cooking events were primarily defined based on highly positive temperature slopes over short periods (indicating temperature doubling) and when >40 °C [13], a threshold not otherwise observed in the absence of cooking. To supplement this approach, participants maintained activity logs documenting potential cooking and other indoor pollution events, which were used to cross-validate the temperature-based detections.

Participants also completed household surveys to provide information on kitchen configuration, cooking fuel/stove type and usage, ventilation, and cooking patterns.

Our primary analysis involved assessing the effect of cooking fuel type (natural gas vs. biogas) on minute-level kitchen NO₂ concentrations. The primary outcome was natural log-transformed NO₂ concentration, derived from time-resolved measurements. We assessed the association between cooking fuel type (natural gas vs. biogas) and log-transformed NO₂ concentrations during cooking events using linear fixed-effects regressions estimated via ordinary least squares, incorporating kitchen fixed-effects to account for all time-invariant household characteristics. Guided by the literature and a conceptual directed acyclic graph, final models included covariates for range hood or window use (ventilation), minute-level temperature, relative humidity and PM_2.5 from the indoor HHB sensor, and concurrent outdoor PM_2.5, temperature and relative humidity from the uRADMonitor A4 station.

Separately, we estimated the association between cooking fuel type and log-transformed PM_2.5 concentrations during cooking events using the same fixed-effects modeling approach described above. As part of our sensitivity analyses, we examined fuel–exposure associations within the one household that used both fuels and compared kitchen NO₂ concentrations during cooking and non-cooking periods (both overall and stratified by fuel type) to better isolate the contribution of combustion-related cooking activities to indoor air quality.

For our primary models, robust standard errors were clustered at the kitchen level to account for repeated measurements. All statistical analyses were conducted using R statistical software (version 4.2.3; R Core Team, Vienna, Austria, 2024), and results with p < 0.05 were deemed statistically significant.

Participants provided informed consent verbally, and the study received approval from the Dickinson College Institutional Review Board.

3. Results

We assessed up to 559 h of air quality data from four households over eight independent monitoring periods.

Table 1. Descriptive statistics of study environmental variables in households cooking with natural gas and/or biogas.

Environmental Variables	Overall	Natural Gas	Biogas	p-Value 2
	N = 559 1	N = 146 (25.7%) 1	N = 413 (74.3%) 1
Indoor NO2 (ppb)	4.4 (4.6)	6.9 (6.7)	3.5 (3.1)	<0.001
Indoor PM2.5 (µg/m3)	9.3 (12.0)	8.5 (15.3)	9.5 (10.6)	0.14
Indoor Temperature (F)	80.8 (3.9)	82.5 (2.7)	80.2 (4.1)	<0.001
Indoor Relative Humidity (%)	55.5 (8.7)	54.2 (9.1)	55.9 (8.5)	0.11
Outdoor PM2.5 (µg/m3)	15.5 (8.2)	12.8 (6.2)	16.4 (8.6)	<0.001
Outdoor Temperature	82.7 (8.5)	82.8 (7.6)	82.7 (8.9)	0.7
Outdoor Relative Humidity (%)	56.1 (14.8)	61.7 (12.4)	54.2 (15.1)	<0.001
Observation Hours with Self-Reported Cooking Activities	73 (13.1%)	20 (13.7%)	53 (12.8%)	0.8

¹ n (%); Mean (standard deviation). ² Pearson’s Chi-squared test; Wilcoxon rank sum test. Abbreviations: NO₂—Nitrogen dioxide; PM_2.5—Particulate matter with aerodynamic diameter ≤ 2.5 µg/m³; ppb—parts per billion; F—Fahrenheit.

summarizes air pollution measurements and contextual characteristics across these hours, stratified by fuel type. The majority of our observations (74.3%) occurred during periods with biogas use.

Across all observations, mean indoor NO₂ concentrations during natural gas use were almost double those during biogas use, representing about a 98.3% increase over the biogas household mean of 3.48 ppb (95% CI: 74.7% to 121.6%, Table 1). In contrast, the mean PM_2.5 concentrations did not differ significantly by fuel type, with higher average values observed during periods of biogas use (9.5 µg/m³) than natural gas (8.5 µg/m³), though with less variability. Cooking (at least one event defined by temperature slope threshold or self-report) was detected during 13.1% of all monitored hours, with a greater share of cooking time occurring in households using natural gas.

Cooking events were associated with increases in both kitchen NO₂ and PM_2.5 levels, as illustrated by the time-resolved spikes in Figure 1. Multiple NO₂ measurements during cooking exceeded the WHO 24 h guideline of 25 µg/m³ (13.3 ppb) [14], particularly during natural gas use. On average, the kitchen NO₂ concentrations during cooking events were 7.5 ppb higher with natural gas compared to biogas use (Figure 1C), representing a 160.9% increase over the biogas stove mean of 4.7 ppb (95% CI: 109.4% to 211.1%). Consistent with these findings, exploratory passive NO₂ monitoring over 72 h showed that natural gas use was associated with a 1.6 ppb higher mean kitchen NO₂ concentration than biogas, reflecting a crude increase of approximately 38%. In contrast, PM_2.5 concentrations showed less discrepancy, with natural gas use producing a mean increase of 1.0 µg/m³ compared to biogas (Figure 1D). However, biogas cooking episodes displayed greater variability and more frequent exceedances of the WHO 24 h PM_2.5 guideline of 15 µg/m³ [14].

The adjusted linear mixed-effects models showed that during cooking periods, natural gas use was associated with a 2.8 ppb higher mean kitchen NO₂ concentration compared to biogas use (95% CI: 1.7, 4.6 ppb), corresponding to a 60.3% increase over the biogas mean of 4.7 ppb. Similarly, natural gas use was associated with a 0.8 µg/m³ higher mean kitchen PM_2.5 concentration relative to biogas (95% CI: 0.6, 1.2 ppb), or approximately a 6.2% increase over the biogas cooking period mean of 12.8 µg/m³. These results were robust to sensitivity analyses restricted to the household that used both biogas and natural gas fueled stoves.

In analyses comparing NO₂ concentrations during cooking versus non-cooking periods, cooking events were associated with a 2.0 ppb higher mean kitchen NO₂ concentration (95% CI: 1.6, 2.5 ppb) compared to non-cooking periods, a 57.8% increase in mean concentration over the average NO₂ levels (3.5 ppb) during non-cooking times. Importantly, this cooking-related increase in NO₂ levels was similar (~2.0 ppb) in both natural gas and biogas households.

4. Discussion

We hypothesized that the benefits of cleaner cooking fuel alternatives like biogas may extend to a relative reduction in cooking-related indoor air pollution compared to more conventional fuels like natural gas. Findings from this small pilot study support that possibility, within the study limitations: natural gas fueled stove use was consistently associated with higher kitchen NO₂ concentrations during cooking periods, while PM_2.5 differences were more modest. The small sample size and limited variability in fuel use and kitchen or cooking behaviors, along with possible misclassification of cooking periods, constrain the strength of our conclusions. However, the use of calibrated high-resolution monitoring and fixed-effects models that account for time-invariant kitchen characteristics strengthens the internal validity of the findings.

The relatively modest differences in PM_2.5 by fuel type may reflect similarities in emission rates during the combustion process [12] but could also stem from uncontrolled factors such as cooking intensity, food type, and ventilation practices [15], raising the possibility that our findings reflect study limitations rather than true fuel-related effects.

Across fuel types, NO₂ levels rose during cooking periods, reinforcing the role of combustion processes as a key driver of indoor air quality. Still, NO₂ concentrations were notably lower with biogas use, suggesting that the relatively lower calorific value associated with biogas may result in reduced production of high-temperature byproducts like NO₂ [11] and a cleaner combustion profile.

Our findings build on evidence that biogas performs similarly to natural gas in terms of PM_2.5 emissions [9,16], but extend this to NO₂ exposure. To our knowledge, this is the first study to directly compare kitchen-level NO₂ concentrations during biogas and natural gas use in a residential setting. Epidemiological data on NO₂ emissions from biogas cooking are limited, which makes meaningful comparisons difficult. However, we can lean on the public health significance of reducing cooking and ambient NO₂ emissions as reported in the literature [17,18,19], given the well-established associated respiratory health risks.

Given growing evidence that even low-level NO₂ exposure can impact respiratory health [1,2,20], and emerging policy interest in regulating gas stove emissions and enabling alternatives [1,21,22], our study, though preliminary and exploratory, contributes by characterizing and quantifying potential differences between cooking fuel types. Biogas, an important alternative fuel source in rural low-resource contexts [9,23], may offer modest indoor air quality benefits in higher-income homes where natural gas remains prevalent. Our future research plans to further assess these possible benefits with larger samples, personal exposure monitoring, and richer behavioral data on ventilation and stove use, within the context of key trade-offs in terms of performance parameters and overall economic and environmental costs [24,25].