Impact of Kitchen Natural Gas Use on Indoor NO₂ Levels and Human Health: A Case Study in Two European Cities

Abstract

Natural gas (NG) is commonly used in kitchens, powering stoves, ovens, and other appliances. While it is known for its efficiency and convenience, NG contributes to the release of nitrogen dioxide (NO₂) and can have significant implications for human health. In this study, the importance of the use of NG in kitchens on human exposure to NO₂ was analyzed. An extensive literature review in the field was conducted, and the NO₂ levels were assessed in kitchens with NG cookers in Aveiro and electric cookers in Porto, both in Portugal. Higher levels of NO₂ were found in kitchens in Aveiro, where NO₂ levels outdoors are lower than in Porto. This pollutant can spread to other rooms, especially when ventilation is lacking, which is particularly concerning during colder seasons and at night. As around 70% of the time is spent at home, this can have a significant impact on human exposure to NO₂. Therefore, although Aveiro has low levels of NO₂ outdoors, its population may be exposed to much higher levels of this pollutant than the Porto population, a city with air quality issues, but predominantly using electric cookers. This finding emphasizes the need for the stricter regulation of NG use indoors to protect human health and also suggests a shift in human health protection policies from mere monitoring/control of outdoor air quality to a comprehensive assessment of human exposure, including exposure to indoor air quality.

1. Introduction

People spend nearly 90% of their lives indoors [1], where various pollutant emission sources like gas stoves, heaters, and tobacco smoke significantly influence indoor air quality. Among these sources, the stove is unique. Although adults spend less than an hour cooking, they stay between 1 and 3.3 h per day in the kitchen [2], increasing potential exposure to combustion constituents and compounds. Unlike other appliances, stove combustion byproducts are emitted directly into the home air without any requirement for venting the exhaust outdoors. As a result, the risk of respiratory symptoms related to gas stove exposure remains significant, even after the adjustment of the nitrogen dioxide (NO₂) levels [3].

To minimize pollution exposure from traditional fuels such as biomass, coal gas, and liquefied petroleum gas (LPG) burning, natural gas (NG) has been promoted by governments as an alternative fuel for cooking over the last few decades [4,5]. As a result, natural gas has become a popular fuel choice for home cooking. In both the US and Europe, over one-third of households cook with gas [6,7]. In some European countries, the proportion is substantially higher, with more than 60% of households using gas for cooking [8].

One of the main markers of combustion-related pollutants is NO₂. Several epidemiological studies have linked NO₂ exposure to various adverse health effects, including respiratory disease, cardiovascular disease, and diabetes [9]. Despite these significant health concerns and the increasing use of natural gas, research in this domain is scarce. According to the Web of Science database in May 2024, there were only 14 studies dedicated to examining the concentrations of NO₂ in kitchens resulting from the use of natural gas. The search was conducted directly on the topic, i.e., title, abstract, keyword plus, and author keywords, using the following query: (NO₂ & Kitchen & (“natural gas” or NG)). These studies compared different types of environments, fuels, ventilation conditions, and exposure periods.

Among the studies analyzed, some measure the concentrations in the short term, typically less than 1 h [2], while others focus on long-term exposure, generally several days or weeks [10]. Few studies were found in the literature monitoring periods of less than 1 day and more than 1 week. Table 1 and Table 2 summarize the main characteristics and findings of the studies in the field for short- and long-term periods, respectively.

Several authors compared the combustion products generated in kitchens from different fuel types. The impact of using biomass fuels [11,12,13,14], electricity [15,16,17,18,19], coal gas [20], liquefied petroleum gas [11,12], propane gas [19,21], and natural gas [10,12,14,15,16,19,20,22,23,24] was analyzed by several authors. Some of the studies compared indoor concentration in kitchens with outdoor levels [3,10,14,16,18,25,26,27,28].

High concentrations of NO₂ were found in kitchens using natural gas cooking burners, compared to homes with electric cooking [15,16,17]. Kornartit et al. [17] found that the average NO₂ levels in kitchens with a gas cooker were twice as high as those with an electric cooker, with no significant difference in the summer period, while Melia et al. [16] find that the average hourly concentration of NO₂ in gas kitchens was more than seven times greater than that in electric kitchens. NO₂ concentrations are even higher in the case of biomass burning. Kumie et al. [13] found NO₂ concentrations from biomass burning ranging from between 0 and 978 $\mathsf{\mu }$g·m⁻³.

Indoor levels are also influenced by infiltration through ventilation and building materials. Urban areas typically have higher NO₂ environmental concentrations, primarily attributable to traffic emissions [29]. Overall, the outdoor NO₂ source contributed 73–76% of the NO₂ in the kitchen [30]. These authors found that outdoor NO₂ sources were present indoors all the time; by contrast, indoor NO₂ sources were present sporadically but with a very high contribution.

Typically, higher NO₂ concentrations are found in rural kitchens than in urban ones. This happens as NO_x emissions are related to the power of indoor sources. Lebel et al. [31] found that NO₂ emissions are linearly related to the amount of natural gas burned ($r^2$ = 0.76; p < 0.01). Furthermore, in a recent literature review conducted by Hu and Zhao [9], it was concluded that the indoor-to-outdoor (I/O) NO₂ concentration ratio is primarily influenced by human activity patterns and ventilation. Typically, the I/O ratio of residential buildings exceeds 1 due to the substantial NO₂ emissions during cooking and smoking, combined with relatively low ventilation rates.

Near 80% of the NO₂ in indoor air is removed during the first hour after it is emitted (Traynor et al. (1989) in [2]). However, in some cases, NO₂ can react and form other compounds such as NO and nitrous acid. Cooking burner pollutants that are not exhausted directly by a range hood are removed from the air in the home by air exchange with the outdoors and, for NO₂, by deposits on interior surfaces. In the U.S., the reaction rate for NO₂ is estimated to be of 0.8 h⁻¹ [32].

Besides emission sources, other factors may influence pollutant concentrations. Hu et al. [15] examined eight conditions involving fuel type (electricity or natural gas), cooking type (boiling water or cooking), and ventilation (with or without), indicating that both the period and style of cooking affect pollutant levels. Willer et al. [23] developed a model that combined the presence of sources of combustion products and the ventilation characteristics in the kitchen to determine the ‘Chance of Accumulation of Combustion Products’ in the kitchen.

Ventilation is one of the main factors affecting indoor concentrations. Lebel et al. [31] discovered that families who do not use their range hoods or have poor ventilation can exceed the 1-h standard of NO₂ (100 ppb) within minutes of stove usage, especially in smaller kitchens.

Some kitchens have “ductless” hoods that recirculate fumes through activated charcoal filters, which are generally less effective at cleaning the air. However, since exhaust hoods are separate from the stove and must be operated manually, they are used, in practice, only 25–40% of the time [27,33]. Vented hoods vary in effectiveness and work best when positioned directly over the stove [26]. Singer et al. [34] observed that NO₂ decay rates were much faster with a forced air unit operating, suggesting the effective removal of NO₂ by the air handler, and Tian et al. [20] found that mechanical exhaust systems are more efficient than natural ventilation in removing pollutants.

The literature review has enabled us to identify that studies on NO₂ exposure, considering both indoor and outdoor levels, have been conducted since at least the 1980s for different cities. These studies typically compare different fuel types under various environmental, ventilation, or cooking conditions. However, the focus has often been narrowly limited to measuring indoor and outdoor NO₂ concentration levels, neglecting the critical study of human exposure. This oversight limits the development of more comprehensive and robust government policies for human health protection, which currently tend to focus primarily on assessing and controlling outdoor air quality levels.

In this study, the main objective is to assess the effect of the use of NG in kitchens on human exposure to NO₂ under different outdoor air quality levels. To this end, we measured NO₂ levels both indoors and outdoors in kitchens using natural gas and electric stoves in two distinct urban areas with different levels of outdoor air quality. One area complies with outdoor European Union NO₂ standards, while the other typically exceeds these standards.

The article begins with a literature review and an overview of NO₂ air quality standards across various countries and time periods. Following this, the methodology is detailed, and the presented results are discussed. The article concludes by summarizing the main findings.

Table 1. Studies of short-term NO₂ exposure (less than 1 week) monitoring in kitchens using natural gas (number of samples).

Ref.	Location	Environment	Fuels	NO2 Concentration		Exposure Period
				Indoors	Outdoors
[22]	Adelaide	Urban and Suburban (193)	NG	33.28 ± 3.76  $\mathsf{\mu }$g·m−3	-	-
			Without NG	15.61 ± 4.70  $\mathsf{\mu }$g·m−3
[15]	Wuhan, China	Urban	NG	Winter:		30 min
				167.0 ± 215.1 $\mathsf{\mu }$g·m−3	31.5 ± 23.0 $\mathsf{\mu }$g·m−3
				Without ventilation:
				294.5 ± 372.0 $\mathsf{\mu }$g·m−3	-
				With ventilation:
				97.9 ± 57.1 $\mathsf{\mu }$g·m−3	-
			Electric	Winter:
				32.9 ± 15.1 $\mathsf{\mu }$g·m−3	29.0 ± 20.0 $\mathsf{\mu }$g·m−3
				Without ventilation:
				28.7 ± 11.6 $\mathsf{\mu }$g·m−3	-
				With ventilation:
				38.9 ± 17.8 $\mathsf{\mu }$g·m−3	-
[20]	6 Chinese cities			Summer:
	(different climatic areas)	Urban	NG (6)	20 ± 0 $\mathsf{\mu }$g·m−3	75 $\mathsf{\mu }$g·m−3	3 h
			Coal gas (20)	100 ± 80 $\mathsf{\mu }$g·m−3
[11]	Tanzania	Rural	Biomass	59.80 ± 66.19 $\mathsf{\mu }$g·m−3	12.79 ± 23.88 $\mathsf{\mu }$g·m−3	24 h
				(48)	(18)
[12]	Pakistan	Urban (3) Rural (3)	NG, LPG	0  $\mathsf{\mu }$g·m−3	-	24 h
			Biomass	14,200 $\mathsf{\mu }$g·m−3
			Animal dung	20,300 $\mathsf{\mu }$g·m−3
[13]	Ethiopia		Biomass		-	24 h
		Rural (17,215)		97 ± 91.41 $\mathsf{\mu }$g·m−3	-
		RuralHighland (9500)		116 ± 101 $\mathsf{\mu }$g·m−3	-
		RuralLowland (8015)		176 ± 74 $\mathsf{\mu }$g·m−3	-
[23]	Netherlands	-	NG (49)	26.59  ± 2.04 $\mathsf{\mu }$g·m−3	-	48 h
			Electric (19)	22.41 ± 2.48 $\mathsf{\mu }$g·m−3
[16]	London	Suburban (4)	NG	135.96  $\mathsf{\mu }$g·m−3	-	96 h
			Electric	17.86  $\mathsf{\mu }$g·m−3
[24]	California	-	NG (343)	Cold seasons:
				31.78 ± 4.33  $\mathsf{\mu }$g·m−3	27.23 ± 3.39  $\mathsf{\mu }$g·m−3	6 days

Fuels: NG: natural gas; LPG: liquefied petroleum gas.

Table 1. Studies of short-term NO₂ exposure (less than 1 week) monitoring in kitchens using natural gas (number of samples).

Ref.	Location	Environment	Fuels	NO2 Concentration		Exposure Period
				Indoors	Outdoors
[22]	Adelaide	Urban and Suburban (193)	NG	33.28 ± 3.76  $\mathsf{\mu }$g·m−3	-	-
			Without NG	15.61 ± 4.70  $\mathsf{\mu }$g·m−3
[15]	Wuhan, China	Urban	NG	Winter:		30 min
				167.0 ± 215.1 $\mathsf{\mu }$g·m−3	31.5 ± 23.0 $\mathsf{\mu }$g·m−3
				Without ventilation:
				294.5 ± 372.0 $\mathsf{\mu }$g·m−3	-
				With ventilation:
				97.9 ± 57.1 $\mathsf{\mu }$g·m−3	-
			Electric	Winter:
				32.9 ± 15.1 $\mathsf{\mu }$g·m−3	29.0 ± 20.0 $\mathsf{\mu }$g·m−3
				Without ventilation:
				28.7 ± 11.6 $\mathsf{\mu }$g·m−3	-
				With ventilation:
				38.9 ± 17.8 $\mathsf{\mu }$g·m−3	-
[20]	6 Chinese cities			Summer:
	(different climatic areas)	Urban	NG (6)	20 ± 0 $\mathsf{\mu }$g·m−3	75 $\mathsf{\mu }$g·m−3	3 h
			Coal gas (20)	100 ± 80 $\mathsf{\mu }$g·m−3
[11]	Tanzania	Rural	Biomass	59.80 ± 66.19 $\mathsf{\mu }$g·m−3	12.79 ± 23.88 $\mathsf{\mu }$g·m−3	24 h
				(48)	(18)
[12]	Pakistan	Urban (3) Rural (3)	NG, LPG	0  $\mathsf{\mu }$g·m−3	-	24 h
			Biomass	14,200 $\mathsf{\mu }$g·m−3
			Animal dung	20,300 $\mathsf{\mu }$g·m−3
[13]	Ethiopia		Biomass		-	24 h
		Rural (17,215)		97 ± 91.41 $\mathsf{\mu }$g·m−3	-
		RuralHighland (9500)		116 ± 101 $\mathsf{\mu }$g·m−3	-
		RuralLowland (8015)		176 ± 74 $\mathsf{\mu }$g·m−3	-
[23]	Netherlands	-	NG (49)	26.59  ± 2.04 $\mathsf{\mu }$g·m−3	-	48 h
			Electric (19)	22.41 ± 2.48 $\mathsf{\mu }$g·m−3
[16]	London	Suburban (4)	NG	135.96  $\mathsf{\mu }$g·m−3	-	96 h
			Electric	17.86  $\mathsf{\mu }$g·m−3
[24]	California	-	NG (343)	Cold seasons:
				31.78 ± 4.33  $\mathsf{\mu }$g·m−3	27.23 ± 3.39  $\mathsf{\mu }$g·m−3	6 days

Fuels: NG: natural gas; LPG: liquefied petroleum gas.

Table 2. Studies of long-term NO₂ exposure (1 week or more) monitoring in kitchens using natural gas (number of samples).

Ref.	Location	Environment	Fuels	NO2 Concentration		Exposure Period
				Indoors	Outdoors
[14]	Pakistan	Urban	NG	Winter (16):		1 week
				218 ± 12 $\mathsf{\mu }$g·m−3	64 $\mathsf{\mu }$g·m−3
				Summer (30):
				234 ± 196 $\mathsf{\mu }$g·m−3	138 $\mathsf{\mu }$g·m−3
		Rural	NG	Winter (20):
				242 ± 57 $\mathsf{\mu }$g·m−3	22 $\mathsf{\mu }$g·m−3
				Summer (16):
				81 ± 65 $\mathsf{\mu }$g·m−3	26 $\mathsf{\mu }$g·m−3
		Rural	Biomass	Winter (20):
				256 ± 40 $\mathsf{\mu }$g·m−3	22 $\mathsf{\mu }$g·m−3
				Summer (24):
				51 ± 27 $\mathsf{\mu }$g·m−3	40 $\mathsf{\mu }$g·m−3
[17]	Hertfordshire (UK)	Urban	Gas	Winter:		1 week
				38.74 ± 12.98 $\mathsf{\mu }$g·m−3	-
				- Non-smokers:	-
				18.9 $\mathsf{\mu }$g·m−3	-
				- Passive smokers:	-
				21.7 $\mathsf{\mu }$g·m−3	-
				- Smokers:	-
				24.8 $\mathsf{\mu }$g·m−3	-
				Summer:
				26.70 ± 2.45 $\mathsf{\mu }$g·m−3	-
				- Non-smokers:	-
				14.0 $\mathsf{\mu }$g·m−3	-
				- Passive smokers:	-
				14.0 $\mathsf{\mu }$g·m−3	-
				- Smokers:	-
				16.2 $\mathsf{\mu }$g·m−3	-
			Electric	Winter:
				13.35 ± 5.27 $\mathsf{\mu }$g·m−3	-
				Summer:
				20.69 ± 3.20 $\mathsf{\mu }$g·m−3	-
[10]	Taiwan			Winter:
		Rural (11)	NG (5), LPG (4)	64.69 ± 17.11 $\mathsf{\mu }$g·m−3	75.41 ± 18.05 $\mathsf{\mu }$g·m−3	1 week
		Urban (12)	LPG	46.01 ± 12.41 $\mathsf{\mu }$g·m−3	44.19 ± 11.85 $\mathsf{\mu }$g·m−3
[28]	Hong Kong	Urban (12)	Town gas	Summer:		1 week
				61.0 ± 22.4 $\mathsf{\mu }$g·m−3	71.8 ± 15.1 $\mathsf{\mu }$g·m−3
[18]	Netherlands		NG, Electricity (5%)	Winter:		1 week
		Urban (299)Rural (149) Rural (164)		96 ± 59 $\mathsf{\mu }$g·m−3   64 ± 50 $\mathsf{\mu }$g·m−3   75 ± 62 $\mathsf{\mu }$g·m−3	62 ± 12 $\mathsf{\mu }$g·m−3 35 ± 19 $\mathsf{\mu }$g·m−3 22 ± 9 $\mathsf{\mu }$g·m−3
		Urban		Unvented geiser (181):
				120 ± 55 $\mathsf{\mu }$g·m−3
				Vented geiser (48):
				71 ± 53 $\mathsf{\mu }$g·m−3
				No geiser (44):
				49 ± 30 $\mathsf{\mu }$g·m−3
		Rural		Unvented geiser (181):
				143 ± 60 $\mathsf{\mu }$g·m−3
				Vented geiser (105):
				60 ± 38 $\mathsf{\mu }$g·m−3
				No geiser (34):
				32 ± 24 $\mathsf{\mu }$g·m−3
[19]	Portage	Rural	NG (I:237, O:240)	65.5 ± 30.7 $\mathsf{\mu }$g·m−3	15.8 ± 6.3 $\mathsf{\mu }$g·m−3	1-week
			LPG (I:568, O: 555)	65.6 ± 38.4 $\mathsf{\mu }$g·m−3	11.8 ± 5.6 $\mathsf{\mu }$g·m−3
			Electric (I:174, O:173)	8.4 ± 4.7 $\mathsf{\mu }$g·m−3	12.8 ± 5.6 $\mathsf{\mu }$g·m−3
[35]	Boston	Urban	Gas	Heating season (57):		2-weeks
				94  ± 36 $\mathsf{\mu }$g·m−3	39 ± 11 $\mathsf{\mu }$g·m−3
				Non heating season (43):
				62 ± 34 $\mathsf{\mu }$g·m−3	32 ± 13 $\mathsf{\mu }$g·m−3
				Total (100):
				81 ± 38 $\mathsf{\mu }$g·m−3	36 ± 12 $\mathsf{\mu }$g·m−3

Fuels: NG: natural gas; LPG: liquefied petroleum gas.

Table 2. Studies of long-term NO₂ exposure (1 week or more) monitoring in kitchens using natural gas (number of samples).

Ref.	Location	Environment	Fuels	NO2 Concentration		Exposure Period
				Indoors	Outdoors
[14]	Pakistan	Urban	NG	Winter (16):		1 week
				218 ± 12 $\mathsf{\mu }$g·m−3	64 $\mathsf{\mu }$g·m−3
				Summer (30):
				234 ± 196 $\mathsf{\mu }$g·m−3	138 $\mathsf{\mu }$g·m−3
		Rural	NG	Winter (20):
				242 ± 57 $\mathsf{\mu }$g·m−3	22 $\mathsf{\mu }$g·m−3
				Summer (16):
				81 ± 65 $\mathsf{\mu }$g·m−3	26 $\mathsf{\mu }$g·m−3
		Rural	Biomass	Winter (20):
				256 ± 40 $\mathsf{\mu }$g·m−3	22 $\mathsf{\mu }$g·m−3
				Summer (24):
				51 ± 27 $\mathsf{\mu }$g·m−3	40 $\mathsf{\mu }$g·m−3
[17]	Hertfordshire (UK)	Urban	Gas	Winter:		1 week
				38.74 ± 12.98 $\mathsf{\mu }$g·m−3	-
				- Non-smokers:	-
				18.9 $\mathsf{\mu }$g·m−3	-
				- Passive smokers:	-
				21.7 $\mathsf{\mu }$g·m−3	-
				- Smokers:	-
				24.8 $\mathsf{\mu }$g·m−3	-
				Summer:
				26.70 ± 2.45 $\mathsf{\mu }$g·m−3	-
				- Non-smokers:	-
				14.0 $\mathsf{\mu }$g·m−3	-
				- Passive smokers:	-
				14.0 $\mathsf{\mu }$g·m−3	-
				- Smokers:	-
				16.2 $\mathsf{\mu }$g·m−3	-
			Electric	Winter:
				13.35 ± 5.27 $\mathsf{\mu }$g·m−3	-
				Summer:
				20.69 ± 3.20 $\mathsf{\mu }$g·m−3	-
[10]	Taiwan			Winter:
		Rural (11)	NG (5), LPG (4)	64.69 ± 17.11 $\mathsf{\mu }$g·m−3	75.41 ± 18.05 $\mathsf{\mu }$g·m−3	1 week
		Urban (12)	LPG	46.01 ± 12.41 $\mathsf{\mu }$g·m−3	44.19 ± 11.85 $\mathsf{\mu }$g·m−3
[28]	Hong Kong	Urban (12)	Town gas	Summer:		1 week
				61.0 ± 22.4 $\mathsf{\mu }$g·m−3	71.8 ± 15.1 $\mathsf{\mu }$g·m−3
[18]	Netherlands		NG, Electricity (5%)	Winter:		1 week
		Urban (299)Rural (149) Rural (164)		96 ± 59 $\mathsf{\mu }$g·m−3   64 ± 50 $\mathsf{\mu }$g·m−3   75 ± 62 $\mathsf{\mu }$g·m−3	62 ± 12 $\mathsf{\mu }$g·m−3 35 ± 19 $\mathsf{\mu }$g·m−3 22 ± 9 $\mathsf{\mu }$g·m−3
		Urban		Unvented geiser (181):
				120 ± 55 $\mathsf{\mu }$g·m−3
				Vented geiser (48):
				71 ± 53 $\mathsf{\mu }$g·m−3
				No geiser (44):
				49 ± 30 $\mathsf{\mu }$g·m−3
		Rural		Unvented geiser (181):
				143 ± 60 $\mathsf{\mu }$g·m−3
				Vented geiser (105):
				60 ± 38 $\mathsf{\mu }$g·m−3
				No geiser (34):
				32 ± 24 $\mathsf{\mu }$g·m−3
[19]	Portage	Rural	NG (I:237, O:240)	65.5 ± 30.7 $\mathsf{\mu }$g·m−3	15.8 ± 6.3 $\mathsf{\mu }$g·m−3	1-week
			LPG (I:568, O: 555)	65.6 ± 38.4 $\mathsf{\mu }$g·m−3	11.8 ± 5.6 $\mathsf{\mu }$g·m−3
			Electric (I:174, O:173)	8.4 ± 4.7 $\mathsf{\mu }$g·m−3	12.8 ± 5.6 $\mathsf{\mu }$g·m−3
[35]	Boston	Urban	Gas	Heating season (57):		2-weeks
				94  ± 36 $\mathsf{\mu }$g·m−3	39 ± 11 $\mathsf{\mu }$g·m−3
				Non heating season (43):
				62 ± 34 $\mathsf{\mu }$g·m−3	32 ± 13 $\mathsf{\mu }$g·m−3
				Total (100):
				81 ± 38 $\mathsf{\mu }$g·m−3	36 ± 12 $\mathsf{\mu }$g·m−3

Fuels: NG: natural gas; LPG: liquefied petroleum gas.

2. Nitrogen Dioxide Air Quality Standards

Upon inhalation, 70–90% of NO₂ can be absorbed through the respiratory tract ([36] in [14]). Short-term NO₂ exposure can exacerbate respiratory diseases, particularly asthma, resulting in symptoms such as coughing, wheezing, and difficulty breathing. It can also lead to hospital admissions and emergency room visits [37]. Prolonged exposure may contribute to the development of asthma and potentially increase vulnerability to respiratory infections [9,38]. For instance, Ciuk et al. [22] found a positive association between NO₂ exposure from gas appliances and the prevalence of respiratory symptoms in a study conducted on 1335 preschool children, and, recently Kashtan et al. [21] found that gas and propane stoves increase long-term NO₂ exposure by 4.0 ppbv on average across the United States, which is 75% of the World Health Organization’s exposure guideline.

Reflecting the findings of various studies, the World Health Organization (WHO) updated its Air Quality Guidelines (AQGs) in 2021, and significantly reduced the annual nitrogen dioxide AQG level from 40 $\mathsf{\mu }$g·m⁻³ (as established in the Global update in 2005) to 10 $\mathsf{\mu }$g·m⁻³ [39,40]. Regarding the short-term guideline for nitrogen dioxide (200 $\mathsf{\mu }$g·m⁻³ in 1 h), it was not re-evaluated in the 2021 edition of the WHO guidelines [39] and, therefore, until now, remains valid. The WHO also recommends a short-term (24-hour) nitrogen dioxide AQG level of 25 $\mathsf{\mu }$g·m⁻³, defined as the 99th percentile (equivalent to three to four days of exceedance per year) of the annual distribution of 24-hour average concentrations. These adjustments align with emerging scientific evidence on the health impacts of air pollution. The revised Air Quality Guidelines apply to both outdoor and indoor environments.

Both the United States (US) [38] and the European Union (EU) [41] are adapting their air quality standards to the new WHO guidelines, albeit more quickly for outdoor air than for indoor air, where there is a notable lack of specific legislation in the US and in most EU member states. However, to reflect the most recent health evidence on the impacts of NO₂, some countries and organizations have defined stricter limits than those set by the WHO. For instance, Australia significantly strengthened NO₂ reporting standards for the 1-h and annual average to 150 $\mathsf{\mu }$g·m⁻³ and 28.2 $\mathsf{\mu }$g·m⁻³, respectively, bringing forward standards initially proposed for 2025 [42,43]. Similarly, the National Institute of Public Health from Norway set guidelines for the 1-hour and annual average NO₂ limits of 100 $\mathsf{\mu }$g·m⁻³ and 30 $\mathsf{\mu }$g·m⁻³, respectively [44].

As mentioned above, although indoor sources play an important role in NO₂ exposure, the EU as many countries, including the U.S., do not have indoor standards for protecting the general population from NO₂ exposure [2]. Exceptions include countries like Canada and China. Canada, for instance, has defined a value of 170 $\mathsf{\mu }$g·m−³ for short-term exposures and 20 $\mathsf{\mu }$g·m−³ for long-term exposures, which are stricter than those set by the WHO levels.

Table 3. Air quality NO₂ ($\mathsf{\mu }$g·m⁻³) guidelines and standards for several countries and organizations.

Authority, Ref.	I / O a	Year to Be Met	Max. 15 min Average	Max. 1 h Average	Max. 24 h Average	Annual Concentration
WHO [45]	Both	1992	-	400	150	-
WHO [46]	Both	2000	-	200	-	40
WHO [39]	Both	2021	-	200	25	10
China  [47]	I	2022	-	200	-	-
Canada b [48]	I	2015	-	170	-	20
Canada b [48]	I	1987	-	480	-	100
Australia [42]	O	2021	-	182	-	38.5
Australia [43]	O	2025	-	150	-	28.2
EU  [41]	O	2010	-	200 (18)	-	40
Norway [44]	O	2013	300	100	-	30
US EPA [49]	O	1971	-	-	-	100
US EPA [49]	O	2010	-	188	-	-
South Africa  [50]	O	2002	-	200 (88)	-	40 (0)

^a I: indoors, O: outdoors. ^b Values defined for the short term and long term.

provides an overview of the air quality NO₂ guidelines and standards for various countries and organizations over time.

3. Materials and Methods

A field campaign was conducted to measure the concentrations of NO₂ in dwellings with natural gas cookers and electric cookers in two different cities. Both indoor and outdoor levels of NO₂ concentrations were measured and the exposure levels of the population living in each city were estimated.

Section 3.1 outlines the site selection, Section 3.2 provides an overview of the sampling process, Section 3.3 presents the data analysis methodology and, in Section 3.4, exposure scenarios are evaluated.

3.1. Site Selection

Sampling was conducted in dwellings using natural gas stoves and dwellings using electrical stoves from two Portuguese cities, Porto and Aveiro. Porto is the largest city in the northern region with a population of approximately 237,000, while Aveiro, is a smaller city with around 78,000 residents. The reason why two cities were used instead of just one is to emphasize that the population of a city with no outdoor air quality problems, such as Aveiro, which mostly uses NG for cooking, is more exposed to NO₂ than the population of the Porto region, a city with outdoor air quality problems but that mostly uses electric cookers for cooking.

Aveiro has an extensive household natural gas network, while Porto predominantly relies on electricity. To the best of the authors’ knowledge, there are no statistical data on the number of natural gas cookers per capita in either city. Despite this, there are figures for domestic natural gas consumption. In fact, in 2022, the city of Aveiro had a 363% higher per capita domestic consumption of natural gas when compared to the city of Porto (https://www.pordata.pt/municipios/consumo+de+gas+natural+por+habitante-485, accessed on 9 January 2024). Therefore, to represent the most common scenario, in our field study in Porto, all the kitchens were equipped with electric cookers, whereas, in Aveiro, all the kitchens had natural gas cookers.

In this work, we define a “cooktop” as a flat surface with individual gas burners/electric coils, typically four or five per stove, and an occasional “griddle” element. The “oven” is an enclosed heated space, usually containing two burners: a “bake” burner below the bottom steel plate and a “broil” burner, typically exposed and affixed to the oven’s roof. A “stove” is a freestanding unit that contains both a cooktop and an oven, typically both using gas, although some combine a gas cooktop with an electric oven. All cookers had an cooker hood.

Aveiro and Porto, both coastal cities, are separated by approximately 60 km. Porto has a temperate maritime climate, with rainy and relatively mild winters, and moderately warm summers. Aveiro experiences a similar climate, but its closer proximity to the ocean leads to constant winds and higher humidity. Both cities see increased precipitation during autumn and winter. Although the two cities are relatively close and in the same climate zone, the meteorological variables of temperature, humidity, and wind speed were recorded during the campaigns.

3.2. General Sampling Scheme

Sampling was conducted in 9 kitchens using natural gas stoves (P1-9) and 9 kitchens using electrical stoves (P10-18). As mention before, in Porto, all the kitchens were equipped with electric cookers, whereas, in Aveiro, all the kitchens had natural gas cookers.

Our sample set included both houses and flats. All selected dwellings were non-smoking and relied solely on natural ventilation.

Preferably, as in previous studies, passive samplers were placed in the middle of the kitchen, away from the cooker, obstacles, windows, and doors [15]. The sampling height was set at the breathing zone (2 m), suspended from the ceiling.

The passive samplers were exposed at the same period in all the kitchens. The experiment was conducted over a period of three weeks, from 22 February 2021 to 15 March 2021, ensuring that factors that might affect the use of the cookers, such as weather and day of the week, would be comparable.

In all the dwellings, a passive sampler was also placed outdoors, nearby, but in such a way as to be representative of the air quality around the dwellings. Indoor and outdoor levels were measured in each dwelling simultaneously.

The average floor area of the dwellings used in the field campaigns in both towns was around 150 m². Almost all of these dwellings were located in the city center and faced the main road or the car park, except for P1, P10, and P11, which were houses located in suburban residential areas.

Table 4. Main characteristics of samples collected.

ID	City	House Characteristics
		Type	Typology a	Total n.° of Inhab.	Cooking Fuel
				(n.º of Children b)
P1	Aveiro	House	T5	3 (0)	NG
P2	Aveiro	Flat (2nd floor)	T4	4 (0)	NG
P3	Aveiro	Flat (3rd floor)	T3	3 (0)	NG
P4	Aveiro	Flat (5th floor)	T3	4 (0)	NG
P5	Aveiro	Flat (2nd floor)	T3	4 (0)	NG
P6	Aveiro	Flat (G floor)	T2	3 (0)	NG
P7	Aveiro	Flat (1st floor)	T3	2 (0)	NG
P8	Aveiro	Flat (3rd floor)	T3	2 (0)	NG
P9	Aveiro	Flat (2nd floor)	T4	2 (0)	NG
P10	Porto	House	T4	3 (0)	Electric
P11	Porto	House	T4	3 (0)	Electric
P12	Porto	Flat (3rd floor)	T4	5 (0)	Electric
P13	Porto	Flat (1st floor)	T4	4 (0)	Electric
P14	Porto	Flat (2nd floor)	T3	4 (0)	Electric
P15	Porto	Flat (5th floor)	T4	3 (0)	Electric
P16	Porto	Flat (9th floor)	T4	4 (0)	Electric
P17	Porto	Flat (4th floor)	T3	4 (0)	Electric
P18	Porto	Flat (2nd floor)	T3	4 (2)	Electric

^a T1: one bedroom, dining room, kitchen, and bathroom (at least one). ^b Number of inhabitants under 18 years old.

outlines the main characteristics of the samples collected.

NO₂ concentrations were measured using passive samplers from Passam AG. This type of sampler is based on the principle of NO₂ diffusion in the triethanolamine absorbent. All the information on the specifications of the sampler used in the field campaigns can be found on the NO₂ data sheet [51]. At the end of the exposure period (3 weeks), the NO₂ collected was determined spectrophotometrically using the Saltzmann method in the Passam AG accredited testing laboratory for air quality. For outdoor sampling, the polypropylene tube of the passive sampler was placed in a special shelter to protect it from all kinds of meteorological disturbances. No shelter of any kind was used for indoor sampling. It should be noted that the P12 indoor sample was removed due to sampling problems.

3.3. Data Analysis

To assess the impact of gas appliances on indoor NO₂ levels, we compared measurements from homes in Aveiro, where gas appliances are commonly used, with those from homes in Porto, where electrical appliances are predominantly used. The initial analysis was carried out using simple descriptive statistics, followed by an inference analysis of the two series. As the samples follow a normal distribution, as confirmed by the Shapiro–Wilk test (N = 9), we applied the parametric t-test to compare the two groups. Following this, the indoor/outdoor (I/O) ratios of the concentrations observed in both cities are also presented and analyzed. Finally, given their potential relevance, a comparison was made of the meteorological parameters observed during the campaigns in both cities.

3.4. Exposure Estimate

Various daily exposure scenarios to NO₂ concentrations were evaluated, considering the number of hours an individual spends in different environments, both indoors and outdoors. Seven scenarios were considered. The first scenario assumes that an individual spends 10 h indoors, specifically in their home, with the remaining time spent outdoors. The second scenario considers 12 h indoors and the remaining time outdoors, and so on, until the last scenario, which assumes 24 h spent entirely indoors.

4. Results and Discussion

Descriptive statistics of NO₂ concentrations, categorized by fuel type, house typology, and house type, are outlined in

Table 5. Statistics of NO₂ concentrations obtained during field campaigns (µg·m⁻³).

		Indoors			Outdoors
City/Fuel Type	N	Min.	Max.	Avg. ± SEM	Min.	Max.	Avg. ± SEM	I/O
Fuel type
- NG (Aveiro)	9	23.8	59.9	38.6  ± 11.3	13.8	29.6	20.4 ± 5.3	1.90
- Electricity (Porto)	9	10.8	26.1	16.7 ± 5.5	23.1	34.2	28.8 ± 3.6	0.58
House typology (flats)
- NG (Aveiro)
- T2	1	-	-	30.0	-	-	17.7	1.69
- T3	5	29.2	46.5	38.4 ± 8.0	17.1	29.6	23.2 ± 6.1	1.64
- T4	2	43.8	59.9	51.9 ± 11.4	13.8	10.8	16.8 ± 4.2	3.09
- Electricity (Porto)
- T3	3	13.4	23.8	18.0 ± 5.3	28.1	31.0	29.7 ± 1.5	0.61
- T4	4	10.8	26.1	17.6 ± 7.8	23.1	34.2	29.5 ± 5.0	0.60
House type (flats)
- NG (Aveiro)
- 1st floor	1	-	-	46.5	-	-	29.6	1.57
- 2nd floor	3	43.8	59.9	49.8 + 8.8	13.8	28.8	20.8 ± 7.5	2.39
- 3rd floor	2	29.2	31.1	30.2 + 1.3	17.1	23.3	20.2 ± 4.4	1.49
- 5th floor	1	-	-	37.8	-	-	17.1	2.21
- Electricity (Porto)
- 1st floor	1	-	-	10.8	-	-	23.1	0.47
- 2nd floor	2	16.7	23.8	20.3 ± 5.0	28.1	29.9	29.0	0.70
- 3rd floor	1	-	-	-	-	-	34.2	-
- 4th floor	1	-	-	13.4	-	-	31.0	0.43
- 5th floor	1	-	-	26.1	-	-	32.7	0.80
- 9th floor	1	-	-	16.0	-	-	28.0	0.57

N: n.° of samples, Min: minimum, Max: maximum, Avg: average, SEM: Standard Error of the Mean.

. Additionally, the concentrations obtained at each point are shown in Figure 1, and indoor-to-outdoor ratio levels are presented in Figure 2.

Different patterns of NO₂ levels are found in the two cities. In Aveiro, the highest levels are observed indoors (59.9 µg·m⁻³) and the lowest outdoors (13.8 µg·m⁻³). In the Porto region, the pattern is the opposite, with lower values indoors (10.8 µg·m⁻³) and higher values outdoors (34.2 µg·m⁻³). The average NO₂ concentration observed in kitchens in the city of Aveiro is 38.6 µg·m⁻³, as opposed to the average value outdoors, which is 20.4 µg·m⁻³, corresponding to an average I/O ratio value of 1.90. On the other hand, in the kitchens of homes in the Porto region, the average value of NO₂ concentrations observed is 16.7 µg·m⁻³, which is 56.7% lower than the concentration observed in Aveiro. Furthermore, in the Porto region, the average value observed outdoors is 28.8 µg·m⁻³, which is 41.1% more than the concentration observed in Aveiro. These concentrations corresponded to an average I/O ratio of 0.58 in Porto, which is 69.3% lower than the I/O observed in Aveiro. Regarding the standard error of indoor NO₂ data, we observed a higher value for Aveiro (11.3 µg·m⁻³) compared to Porto (5.3 µg·m⁻³). This discrepancy is due to the presence of indoor NO₂ sources, particularly the use of natural gas appliances in Aveiro kitchens.

By applying the Shapiro–Wilk test, we observed that the indoor NO₂ concentration levels follow the normal distribution in both cities, Aveiro (p = 0.843) and Porto (p = 0.216). Therefore, despite the small sample size, values obtained are compatible with a normal distribution.

Significantly higher levels of NO₂ were found in homes with gas appliances (Aveiro) compared with homes without gas appliances (Porto) (p < 0.001, t-test). If we look at the results presented in detail in Figure 1, we can see that the results are consistent in both cities, both indoors and outdoors, following the described pattern. In Aveiro, indoor NO₂ concentrations are consistently higher than outdoor levels, whereas in Porto, the opposite trend is observed.

Byproducts from gas stove emissions are not distributed evenly throughout the house. Therefore, it is expected that, during peak NO₂ exposures from using the gas stove, the values may exceed the 3-week averages measured in this study by as much as an order of magnitude, especially near the source. The few studies that have measured both short-term and long-term NO₂ exposures confirm this relationship [2,8,10].

It is also important to note that larger dwellings in Aveiro city, with gas stoves, tend to have higher NO₂ values, when compared to smaller dwellings. In Porto, with electric stoves, this pattern, as expected, does not occur. In both cities, the NO₂ indoor and outdoor concentration values tend to decrease from the lowest to the highest floors, further away from road emission sources. This is particularly observed in the houses located in the city of Aveiro.

Concerning the I/O ratio, the pattern is identical to that observed previously, with both cities behaving in a clearly different way. Aveiro has I/O values that are always greater than 1 and the Porto region has I/O values that are always less than 1 (Figure 2).

Assuming that the values observed in both cities can be indicatively compared with the NO₂ air quality guideline level of the WHO annual average (10 µg·m⁻³), it can be observed that, for both cities, both indoors and outdoors, this guideline is being exceeded. Looking at the NO₂ annual average limit value of EU directive 2008/50/EC (40 µg·m⁻³), four of the nine dwellings in Aveiro exceed this value, unlike the observations made in the Porto region where there is no exceedance indoors. Finally, if we take the annual average limit value to be taken by the new air quality directive (20 µg·m⁻³), all the observations made indoors in Aveiro exceed this value, while, in Porto, only two observations out of eight exceed this standard.

As far as meteorology is concerned, no significant differences were found in the values of the meteorological variables observed in the two cities during the campaigns. The average absolute difference in temperature was around 0.6 °C, in relative humidity less than 4%, and in wind intensity less than 0.7 m.s^$-1$.

We spend around 70% of our time in our homes, which makes residential air quality of crucial importance [52]. Based on this, it is important to consider for the assessment the outdoor air levels, but also the indoor ones. Figure 3 outlines the daily exposure of the Aveiro and Porto population to NO₂ concentrations for different scenarios.

The exposure results indicate that the average daily exposure to NO₂ levels is higher in homes located in the city of Aveiro, a less polluted city where people typically use natural gas for cooking. In these homes, exposure tends to increase as the number of hours spent indoors increases. Conversely, in Porto, the NO₂ concentration levels are approximately 35.8% lower, and the trend shows a decrease in daily exposure as the number of hours spent indoors increases.

In fact, the population of small-/medium-sized European cities with no problems in terms of outdoor air quality, such as Aveiro, may be exposed to high levels of NO₂ due to the consumption of NG in their kitchens. On the other hand, the population of medium-sized cities, such as Porto, with air quality problems, particularly in terms of NO₂, may be less exposed to this pollutant if they do not use NG in their cooking.

This is not to suggest that outdoor air quality is unimportant, quite the opposite. Outdoor air quality is crucial to ensuring good ventilation of indoor spaces. However, the findings indicate that the good air quality observed in Aveiro was insufficient to ensure acceptable NO₂ levels in its kitchens, where natural gas is used. Conversely, the poor outdoor air quality in Porto did not significantly degrade the NO₂ levels observed in its kitchens, which are generally powered by electricity.

The results highlight that stricter air quality monitoring and enforcement are needed. Air is a fundamental resource for human health and well-being. The quality of the outdoor air, where we spend part of our time and which we use to ventilate our buildings, is important, but, above all, the quality of the indoor air, where we spend around 90% of our time (70% at home), is crucial for our health and well-being. Therefore, besides the restrictions on outdoor levels, policies should focus on improving indoor air quality through better ventilation systems and stricter regulations on indoor pollution sources to align with WHO guidelines, aiming for long-term health benefits.

In Aveiro, the high NO₂ levels found indoors indicate a pressing need for public health interventions, including raising awareness about the health impacts of NO₂, and providing information on mitigation strategies such as the improving ventilation. Given the differences between Aveiro and Porto, policies should be tailored to regional conditions. Aveiro may require more aggressive measures to tackle indoor air quality issues compared to Porto, which currently shows better compliance with air quality standards. Therefore, in regions where air quality is poor, financial support or incentives may be needed for households to improve their ventilation systems and/or invest in new cookers with cleaner technologies.

5. Conclusions

This study investigated the effect of the use of NG in kitchens on human exposure to NO₂ under different outdoor air quality levels. To achieve this goal, NO₂ concentrations associated with different cooking fuels were measured in kitchens placed in two different medium-sized European cities, Porto and Aveiro. In Porto, where electric cookers predominated, and Aveiro, where natural gas cookers were prevalent, significant differences in indoor NO₂ levels were observed. Homes with natural gas cookers consistently exhibited higher NO₂ concentrations and exposure compared to those with electric cookers, aligning with previous research highlighting the emissions from gas combustion as a substantial indoor pollutant source.

The exposure results reveal that average daily NO₂ exposure is higher in homes in Aveiro, a less polluted city where natural gas is commonly used for cooking. In these homes, exposure tends to rise with the number of hours spent indoors. In contrast, Porto exhibits NO₂ levels that are approximately 35.8% lower, with a trend of decreasing daily exposure as indoor time increases.

The results have significant policy implications, indicating that cities with better outdoor air quality, like Aveiro, may still have residents exposed to higher levels of pollution compared to those in more polluted cities, like Porto. However, it is important to note that these findings are preliminary due to the limited sample size of nine kitchens per stove type in each city. Expanding the sample size and including kitchens from both cities for each stove type could enhance the reliability of the comparison.

In addition, several factors, such as cooking behavior—particularly the duration and frequency of gas stove use—as well as the type and level of ventilation, can significantly impact indoor NO₂ concentrations. These factors should be carefully monitored in future studies, and guidelines should be developed to improve indoor air quality. On the other hand, Passam samplers are designed to operate accurately over a wide range of temperatures (5 to 40 °C) and humidity (20 to 80%); we did not anticipate that these factors would significantly influence NO₂ concentration measurements made in this work. However, in order to maintain experimental control, in future sampling work, the indoor temperature and relative humidity of the study domain should be recorded.

The findings underscore the importance of considering cooking fuel type in indoor air quality management. Despite the absence of indoor NO₂ standards in many regions, including the European Union and the U.S., countries like Canada have implemented stricter guidelines. This disparity underscores the need for broader adoption of indoor air quality regulations to mitigate health risks associated with indoor NO₂ exposure, particularly in homes relying on natural gas for cooking. Given the amount of time we spend in our homes (around 70% of the time), the results also suggest the need for a change in current human health protection policies, moving from a partial view focused only on outdoor air quality to a holistic view where indoor air quality must also urgently be considered.