Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Dec 1.
Published in final edited form as: Sci Total Environ. 2021 Jul 22;798:149236. doi: 10.1016/j.scitotenv.2021.149236

Characterization of cooking-related ultrafine particles in a US residence and impacts of various intervention strategies

Jianbang Xiang a,*, Jiayuan Hao b, Elena Austin a, Jeff Shirai a, Edmund Seto a
PMCID: PMC8484057  NIHMSID: NIHMS1731551  PMID: 34340070

Abstract

Interventions that improve air exchange or filter the air have the potential to reduce particle exposures from residential cooking. In this study, we evaluated the effect of using a range hood, opening kitchen windows, and using portable air cleaners (PACs) in various home locations on the concentrations of ultrafine particles (UFPs) at different times and in different rooms during and after cooking. All experiments were conducted using a standardized cooking protocol in a real-world naturally-ventilated apartment located in the northwest United States. Real-time UFP measurements collected from the kitchen, living room, and bedroom locations were used to estimate parameters of a dynamic model, which included time-varying particle emission rates from cooking and particle decay. We found that 1-min mean UFP number concentrations in the kitchen and living room mostly peaked within 0–10 min after cooking ended at levels of 150000–500000 particles/cm3. In contrast, the bedroom UFP concentrations were consistently low except for the window-open scenario. While varying considerably with time, the 1-min UFP emission rates were comparable during and within 5-min after cooking, with means (standard deviations) of 0.8 (1.1) ×1012 and 1.1 (1.2) ×1012 particles/min, respectively. Compared with the no-intervention scenario, keeping the kitchen windows open and using a kitchen range hood reduced the mean indoor average UFP concentrations during and 1 h after cooking by ~70% and ~35%, respectively. Along with the range hood on, utilizing a PAC in the kitchen during and after cooking further reduced the mean indoor average UFP levels during and 1 h after cooking by an additional 53%. In contrast, placing the PAC in the living room or bedroom resulted in worse efficacy, with additional 2–13% reductions. These findings provide useful information on how to reduce cooking-related UFP exposure via readily accessible intervention strategies.

Keywords: cooking, ultrafine particle, emission, range hood, window opening, portable air cleaner

Graphical Abstract

graphic file with name nihms-1731551-f0001.jpg

1. Introduction

Residential cooking activities can produce a wide range of hazardous particles and gaseous organic compounds (Chen et al., 2018; Huang et al., 2021; Wallace et al., 2004; Xiang et al., 2017). Exposure to such cooking-related pollutants has been associated with adverse health effects (Gabdrashova et al., 2021; Ke et al., 2009; Neghab et al., 2017; Pan et al., 2008; Singh et al., 2016), even lung cancer risks (Gao et al., 1987; Seow et al., 2000; Wang et al., 1996; Wu-Williams et al., 1990; Zhong et al., 1999). Among the cooking-related pollutants, ultrafine particles (UFPs; particles with an aerodynamic diameter ≤ 100 nm) dominate the particle number concentrations (Wallace et al., 2004). Based on a review study by Zhao et al. (Zhao and Zhao, 2018), the residential UFP concentrations during some cooking events, e.g., pan-frying and stir-frying, can exceed 106 particles/cm3, up to 70–100 fold of typical urban levels (Presto et al., 2021) and 5–10 fold of on-road levels (Austin et al., 2021; Xiang et al., 2020).

Although cooking activities mainly occur in kitchens, occupants in other rooms may also be exposed to cooking-related UFPs due to cooking-fume dispersion. Some studies have demonstrated the remarkably high UFP levels and emission rates in kitchens in certain cooking scenarios (e.g., frying) (Chen et al., 2018; Lunden et al., 2015; Wan et al., 2011). However, few studies have investigated the dispersion of cooking-related UFPs from kitchens to other rooms in homes (Wan et al., 2011). In particular, it is unclear how cooking impacts UFP levels in other rooms with their doors open or closed, such as an open living room or a closed bedroom.

Several studies have measured the UFP or UFP emission rates from certain cooking scenarios by assuming a steady emission strength during cooking (Chen et al., 2018; Wallace et al., 2004). Specifically, the UFP concentration time-series curves were nonlinearly fitted for the entire increasing stage by assuming a constant emission rate. However, this assumption may lead to large biases because the UFP emission rates can be time-varying with oil and food temperature. A difference method with more discreet time steps should yield more accurate estimates.

Utilizing a kitchen range hood during cooking is a common way to mitigate the dispersion of cooking fumes. Brett et al. evaluated the pollutant (e.g., CO2) capture efficiency of various types of kitchen range hoods in California homes (Delp and Singer, 2012; Lunden et al., 2015; Singer et al., 2012; Singer et al., 2017), showing that the capture efficiency varied widely (from <15% to >98%). When the range hood is not very effective, people may open kitchen windows or use a portable air cleaner (PAC) during and after cooking to reduce indoor pollutant levels. However, it remains unclear how these strategies impact the time-varying profiles of indoor UFP levels in different locations in the residence. While using a PAC, it is not evident how the placement impacts the overall effectiveness. Moreover, no studies have compared the effectiveness of these strategies for reducing cooking-related UFPs in residences.

The present study is part of a larger study, which assesses cooking-related PM2.5, UFPs, and size-resolved particles ranging from 0.3 to 10 μm. In contrast to a previous analysis, which assessed the characteristics of cooking-related PM2.5 (Xiang et al., 2021a), the present study examines the profiles of cooking-related UFPs. Instead of examining the pollutant mixture from both fuel combustion and food fumes, as illustrated in the studies mentioned above (Chen et al., 2018; Lunden et al., 2015), the current study focuses on UFP emissions from food fumes using an electric range. Through multi-scenario field measurements in a US home, this study aims to 1) illustrate the time-varying profiles of cooking-related UFPs in different locations in the residence; 2) demonstrate the dynamic process of cooking-related UFP emission rates; and 3) evaluate the impact of different intervention strategies (i.e., utilizing a range hood, opening kitchen windows, or using a PAC in various indoor locations) on indoor UFP number concentrations.

2. Methods

2.1. Experimental site

The experimental methods were reported in greater detail elsewhere (Xiang et al., 2021a). Briefly, we conducted the experiments in a naturally ventilated apartment in Sand Point, Seattle, Washington State, United States (US), from August 6 to September 16, 2019. As shown in Appendix Fig. A1, the duplex apartment had an open kitchen (including a dining area) and living room on the first floor, and all three bedrooms on the second floor. There was an internal staircase, without a door or barrier, connecting the two floors. Openable windows were in the kitchen, living room, and bedrooms. Cooking was conducted on one of the front burners of the electric kitchen range (Hotpoint, GE Appliances, US) with 10 temperature options (i.e., OFF, 1–9 from low to high levels). In cooking scenarios involving the range hood (Broan BUEZ2, US), the exhaust hood with a nominal airflow of 90 liters/s over the electric range was used to vent the kitchen air outside.

2.2. Cooking scenarios

Pan-frying steak and asparagus were chosen as the primary cooking recipe since pan-frying is a common and high-emitting cooking method (Chen et al., 2018). We followed the same protocol to purchase, prepare, and cook the food for each experiment, as described in detail in the Appendix. Fig. 1 shows the timeline of the cooking protocol for each experiment. In brief, the same amount of food materials (mean ± standard deviation [SD]: 230 ± 17 g for steak and 227 ± 25 g for asparagus) were prepared about 30 min before cooking for each experiment. At the start of cooking (Minute 0), the electric range was turned on at the temperature level 9 with the pre-cleaned nonstick frying pan on one front burner. After 2 min, the steak was added to the pan and fried for 6 min. After removing the steak from the pan, we reheated the pan for 30 s and then fried the prepared asparagus for 8 min. That step was followed by turning off the range (Minute 17) and removing the asparagus from the pan. The uncovered pan was left on the same burner to cool for 1 h before being cleaned. No smoking, cleaning, candle burning, or other cooking activities but the designed cooking during Minutes 0–16 occurred during each experiment. Only the experimenters (1–2 persons) were present in the apartment during the experiments, when they mostly stayed in the kitchen during cooking and in the living room/bedroom after cooking.

We conducted seven experimental scenarios with various combinations of range hood, kitchen window, and PAC statuses as follows: 1) kitchen window closed + range hood off + PAC off, 2) kitchen window open + range hood off + PAC off, 3) kitchen window closed + range hood on + PAC off, 4) kitchen window closed + range hood on + kitchen PAC on, 5) kitchen window closed + range hood on + living-room PAC on, 6) kitchen window closed + range hood on + bedroom PAC on, and 7) kitchen window closed + range hood on + kitchen PAC on + living-room PAC on + bedroom PAC on. Because the measured indoor UFP levels in Scenario 1 were extremely high and decayed slowly, we opened the apartment’s main door and the kitchen window about 1 h after cooking to avoid acute exposure and closed them again after 5 min. Also, no repeated trials were conducted in Scenario 1 because of the extremely high exposure. In contrast, two trials were conducted for all the other scenarios (Scenarios 2–7). In Scenario 2, we opened the kitchen window at least 30 min before cooking until the end of all measurements. In Scenarios 3–7, we turned on the range hood at the start of cooking (Minute 0) and turned it off 1 min after cooking (Minute 18) due to the noise issue. In the scenarios involving PAC use, we turned on the PACs about 10 min before cooking and kept them on until the end of all measurements. One of the three bedrooms was chosen when the PAC was used in the bedroom, as illustrated in Appendix Fig. A1. We kept all doors and windows closed in the living room and bedrooms for all scenarios unless specified.

The PACs utilized in this study (Air Purifier 2000i, Philips, US) contain a high-efficiency particulate air (HEPA) filter, with a rated clean air delivery rate (CADR) of 179 m3/h for smoke. In Scenarios 4–7, we used the PACs’ auto operation mode, automatically adjusting its fan speed level based on an integrated particle sensor. This auto-mode feature, widely used in real-world settings due to its convenience, has been discussed in greater detail elsewhere (Huang et al., 2021; Xiang et al., 2021b; Xiang et al., 2021c).

2.3. UFP and CO2 measurement

We utilized P-Trak (Model 8525, TSI Inc., MN) condensation nuclei particle counters to measure the UFP number concentrations in the kitchen, living room, and bedroom at 1-min intervals from about 30 min before and 4 h after cooking. Note that the living-room UFPs in Scenario 1 were not measured due to the instrument unavailability. P-Trak monitors have been widely used for UFP measurement in previous studies (Austin et al., 2021; Liu et al., 2010; Xiang et al., 2020). Prior to the current study’s experiments, the P-Trak monitors were factory calibrated by TSI. Also, we conducted a relative calibration among the three monitors (choosing one of the monitors as the reference monitor) in a scenario similar to Scenario 1 with all three monitors co-located to ensure consistent measurements from the three monitors (see more details in Appendix Fig. A2). The normalized root mean squared errors (Xiang et al., 2020) of the post-calibrated monitors were 8–17%, indicating reasonably consistent measurements. Besides, we measured CO2 concentration in the kitchen with a factory-calibrated Q-Trak (Model 7575, TSI Inc., US) at 1-min intervals to assess air exchange rate (AER) as described below.

2.4. Data analysis

While investigating the UFP variations under various intervention scenarios, we evaluated the UFP number concentrations, decay-related parameters, and emission rates. To minimize the biased comparison attributed to window-closed air exchange rate (AER) variations among different scenarios, we selected one of the two trials in Scenarios 3–7, which had comparable window-closed AERs to that in Scenario 1. Also, we selected one of the two trials in Scenario 2 (window open) that had complete measurements in the three indoor locations. All calculations were made in R (Version 3.3.0) (R Core Team, 2013).

2.4.1. Concentrations

Besides the location-specific concentrations, we calculated the indoor average concentrations by averaging the first-floor (kitchen and living room) and second-floor (bedroom) concentrations (see more details in the footnotes of Appendix Table A1). We then conducted the following analyses. First, we calculated the UFP level increases during and after cooking compared with the before-cooking levels to adjust for the daily variations in the indoor UFP background levels. As defined earlier, the time when we turned on the electric range was Minute 0. Minutes (−10)–(−1), 0–16, and 17–75 were defined as before-, during-, and after-cooking periods, respectively. The UFP levels after Minute 75 were not compared since the door and window statuses were altered at Minute 76 in Scenario 1. Second, the UFP level increases were compared among different indoor locations, i.e., the kitchen, living room, and bedroom. Finally, the UFP level increases were compared across different scenarios to determine the effectiveness of these intervention strategies. Scenarios 2 and 3 were compared with Scenario 1 to reveal the effects of keeping the kitchen window open and utilizing the range hood, respectively. Scenarios 4–7 were compared with Scenario 3 to demonstrate the efficacy of using a single PAC and multiple PACs in various locations.

2.4.2. Estimation of decay-related parameters

The UFP levels in the kitchen, living room, and bedroom after cooking (no emission source) can be described as Eq. (1) (Sun et al., 2019; Wallace et al., 2004; Xiang et al., 2021a), assuming the air was well mixed in each location.

Cin(t2)=Cin(bg)+(Cin(t1)−Cin(bg))⋅e−kt⋅(t2−t1) (1)

The model describes the dynamics of indoor UFP number concentration Cin(t) in units of particles/cm3 as a function of two times, t1 and t2. The total particle decay is modeled via the exponential function, with rate parameter kt in units of h−1. The background particle concentration indoors (i.e., before cooking) is represented by Cin(bg).

The total decay rate, kt, accounts for decay from ventilation, deposition, and PAC use. As shown in Eq. (1), we assume an exponential decay with time, and fit the parameter empirically to the monitored UFP concentrations as they decreased after each experiment’s cooking period, according to the following criteria: the decay curve showed a clear decreasing trend and was visually smooth over at least a 30 min period, at least 10 min had elapsed since the end of cooking, and the period over which the decay parameter was fit did not include any change in door or window openings or range hood use.

An approach based on CO2 tracer gas (Li et al., 2014) was used to estimate the AER for the first story (kitchen and living room) as described elsewhere in detail (Xiang et al., 2021a). We assumed that the AER applied to the open and connected kitchen and living areas, but not the upstairs bedroom behind the closed door. The UFP concentrations reported in the results for the different areas suggest that this assumption was valid. More details about the AER estimation were shown in the Appendix.

2.4.3. Estimation of emission rates

The dynamic model for indoor UFPs during cooking is described by Eq. (3):

dCin(t)dt=p⋅AER⋅Cout(t)+S(t)V−kt⋅Cin(t) (3)

In this model, the change in indoor UFP number concentrations dCin(t)/dt in units of particles/cm3 is a function of the current indoor concentration Cin(t), which is subject to decay as described in the previous section for kt, but also driven by the UFP emission rate at that time, S(t) from cooking, which is distributed over the indoor volume, V in units cm3. The model also assumes non-negligible infiltration of UFP from the outdoors, subject to the outdoor UFP concentration Cout(t), the AER, and the penetration factor p (unitless). Based on previous studies, we set p to 0.47 and 1 when windows were closed and open, respectively (Stephens and Siegel, 2012).

Eq. (3) can be solved using a difference method, assuming the AER, p, and kt remain constant over the time step Δt, as described in other studies (Bennett and Koutrakis, 2006; Sun et al., 2019):

Cin(t)=p⋅AER⋅Cout(t)kt+S(t)kt⋅V+(Cin(t−Δt)−(p⋅AER⋅Cout(t)kt+S(t)kt⋅V))⋅e−kt⋅Δt (4)

Thus, S(t) can be solved as Eq. (5):

S(t)=Cin(t)−p⋅AER⋅Cout(t)kt−(Cin(t−Δt)−p⋅AER⋅Cout(t)kt)⋅e−kt⋅Δt1−e−kt⋅Δt⋅kt⋅V (5)

The outdoor UFP levels during the experiments were not continuously measured due to instrument unavailability. We estimated the outdoor UFP concentrations using the measured before-cooking UFP concentrations in the kitchen in Scenario 2 and a steady-state model (see more details in the Appendix). In Scenario 2, the kitchen windows were opened at least 4 h before the UFP measurement, and no events generating amounts of particles (e.g., cooking, smoking, incense burning, and vacuuming) occurred within 4 h before the UFP measurement. Thus, the indoor sources’ impacts on indoor UFPs were assumed to be minimal. Based on the steady-state model, the average outdoor UFP concentrations were ~3000 particles/cm3. Austin et al. measured the ambient UFP levels in Sand Point of Seattle (~1.5 km from our experimental site) during May–July 2018 and found the mean levels of 5200 particles/cm3 (Austin et al., 2019; Austin et al., 2021), similar to our results. Given the two studies were both conducted in the warm season within two years, the UFP levels at this urban background site between these two periods should be similar. Considering that the measurements by Austin et al. involved more days, we took their measurements, i.e., 5200 particles/cm3, as the input of the outdoor UFP concentration in this study.

The increase in bedroom UFP levels during cooking (Minutes 0–16) was negligible compared to those in the kitchen and living room based on our measurements (see more details below). Thus, the cooking-related total UFP emission rates can be estimated using average UFP number concentrations and total decay rates in the kitchen and living room via Eq. (5). As the living-room UFPs were not measured in Scenario 1, the emission rates were calculated using the UFP concentrations and decay rates in the kitchen instead.

3. Results and Discussion

3.1. Overview

Fig. 2 shows the profile of 1-min indoor average and location-specific UFP levels for each scenario. In spite of the varying magnitudes, the UFP levels across all scenarios mostly exhibited a similar time-varying pattern. Throughout all the experiments, the mean (SD) background indoor UFP levels (before cooking) were 6100 (4300) particles/cm3 in this apartment, with relative differences of less than 20% among the kitchen, living room, and bedroom levels. The average indoor levels were slightly higher than ambient concentrations (~5200 particles/cm3 on average), possibly due to the secondary organic aerosols generated from indoor ozone-initiated chemical reactions (Weschler, 2016; Xiang et al., 2016). Previous studies have shown that ozone can react with unsaturated indoor organic compounds (e.g., terpenes from building materials and cleaning products; squalene and unsaturated fatty acids from occupants’ skin oil), resulting in secondary organic aerosol (SOA) formation (Weschler, 2016; Xiang et al., 2016). The UFP levels in the kitchen and living room started to rise 0–2 min after the steak was added (Minutes 2–4). The peak concentrations mostly occurred 0–10 min after cooking ended, at levels of 150000–500000 particles/cm3, comparable to those reported in previous studies for pan-frying cooking events (Zhao and Zhao, 2018). The kitchen and living-room UFP levels then gradually declined to the background levels after at least 4 h after cooking. An after-cooking spike occurred in Scenario 3, potentially resulting from the fluctuations of AERs and outdoor ozone concentrations. Similarly, a spike of indoor UFP concentrations dominated by the increasing outdoor ozone concentrations was previously reported by Xiang et al. (2016). Compared with the kitchen and living room, the bedroom UFP levels during and after cooking were consistently low except for Scenario 2, in which the kitchen window was open, and the bedroom levels were even higher than the two first-floor rooms during Minutes 40–150.

Fig. 2.

Fig. 2.

Open in a new tab

Time-series plots of 1-min indoor ultrafine particle number concentrations for each scenario. “S1–7” represents Scenarios 1–7. “Average” refers to indoor mean concentration. Note that the living-room UFPs in Scenario 1 were not measured due to the instrument unavailability.

3.2. Concentrations

Fig. 3 shows the means (SDs) of 1-min indoor UFP level increases during and 1-h after cooking compared with before-cooking levels for each scenario (see Appendix Fig. A3 for the boxplots and Table A1 for the statistical summary). In Scenario 1 (no intervention strategies), the indoor average mean (SD) UFP levels during and 1-h after cooking increased by 82000 (55400) and 123300 (54200) particles/cm3, respectively, compared with the before-cooking levels. Comparing indoor locations, the pooled during-and-after (D&A) UFP concentration in the kitchen increased by 211300 (113300) particles/cm3, which was approximately 12 times higher than the mean bedroom-level increase (16800 particles/cm3). The living-room levels were not measured in this scenario due to instrument unavailability.

Fig. 3.

Fig. 3.

Open in a new tab

Means (standard deviations) of 1-min indoor ultrafine particle concentration increases during and 1-h after cooking compared with before-cooking levels in each scenario. “S1–7” represents Scenarios 1–7. “During”, “After”, and “D&A” refer to during cooking, 1-h after cooking, and pooled period, respectively. “Average” refers to indoor mean concentration. Note that the living-room UFPs in Scenario 1 were not measured due to the instrument unavailability.

In Scenario 2 (kitchen window open), the mean (SD) indoor average UFP level increase in the pooled D&A period was 34400 (17100) particles/cm3, about 80000 particles/cm3 (70%) lower than that in Scenario 1. The mean kitchen level increase in the D&A period was 169000 particles/cm3 (80%) lower, while the bedroom increase was 15000 particles/cm3 (87%) higher than seen in Scenario 1. Comparing indoor locations, the kitchen level increases in the D&A period were 10000 particles/cm3 (30%) higher than those for the living room, and 11000 particles/cm3 (35%) higher than those for the bedroom. The kitchen-bedroom concentration differences in the D&A period in Scenario 2 were minor compared to those in Scenario 1. The substantial increase in bedroom UFP levels in Scenario 2 suggests that keeping the kitchen window open significantly increased the diffusion rates of the cooking-emitted UFPs because of the large AERs in this scenario (see more results of AERs in Section 3.3).

In Scenario 3 (range hood on during cooking), the mean (SD) indoor average UFP level increase in the pooled D&A period was 73600 (29200) particles/cm3, about 40000 particles/cm3 (35%) lower than that in Scenario 1. The mean kitchen and bedroom level increases in the D&A period were about 69000 particles/cm3 (33%) and 7000 particles/cm3 (42%) lower than those in Scenario 1, respectively. Comparing the indoor locations, the kitchen level increases in the D&A period were 10000 particles/cm3 (8%) higher than those for the living room, and 133000 particles/cm3 (14 times) higher than those for the bedroom. The kitchen-bedroom concentration differences in the D&A period in Scenario 3 were larger than those in Scenario 1. While less effective than Scenario 2, the range hood utilization consistently reduced the cooking-related UFP levels for multiple indoor locations compared with Scenario 1. Larger concentration reductions for the bedroom reveal that the range hood captured a substantial amount of cooking-related UFPs before they migrated to the upstairs bedroom.

In Scenario 4 (PAC in the kitchen), the mean (SD) indoor average UFP level increase in the pooled D&A period was 34700 (30600) particles/cm3, about 39000 particles/cm3 (53%) lower than that in Scenario 3. The mean kitchen, living room, and bedroom level increases in the D&A period were about 78000 particles/cm3 (54%), 72000 particles/cm3 (54%), and 3000 particles/cm3 (32%) lower than those in Scenario 3, respectively. Comparing the indoor locations, the kitchen level increases in the D&A period were 4000 particles/cm3 (7%) higher than those for the living room, and 58000 particles/cm3 (9 times) higher than those for the bedroom. In contrast, using the PAC in the living room (Scenario 5) and bedroom (Scenario 6), the mean indoor average UFP level increases in the pooled D&A period were 2% and 17% lower than those in Scenario 3, respectively, but 107% and 76% higher than those in Scenario 4. The sheer difference of the mean indoor average UFP level increases in the pooled D&A period among Scenarios 4–6 highlights the importance of the PAC placement location (see more results and discussion below). When using PACs in all three rooms (Scenario 7), the mean indoor average UFP level increases in the pooled D&A period were 49% lower than those in Scenario 3 but 9% higher than those in Scenario 4. Theoretically, Scenario 7 should yield better efficacy in mitigating indoor UFP levels than Scenario 4 if all experimental conditions except PAC usage were the same. However, there were likely variations in cooking-related UFP emission rates across the different scenarios, impacting UFP decay rates and explaining the higher UFP levels seen in Scenario 7 compared to Scenario 4. In light of this, we further evaluated the decay-related parameters and UFP emission rates below.

3.3. Decay-related parameters

Table 1 shows the location-specific UFP total decay rate (kt) and first-floor AERs for each scenario. As described in the Methods, the AERs were measured based on a CO2-based approach. The decay rates (kt) for the kitchen and living room were comparable across all the scenarios, with a mean relative difference of 15% (range: 0–30%). Thus, kt for these two locations were aggregated into the first-floor kt, ranging from 1.10 (0.02) h−1 (Scenario 1) to 7.16 (0.24) h−1 (Scenario 7). In contrast, kt for the first floor was 256% larger than that for the bedroom on average. The large difference in kt between the bedroom and first floor suggests that the bedroom door, or equivalent barrier, can substantially reduce the UFP decay rates between the spaces. The first-floor kt (range: 1.10–1.26 h−1) and AERs (range: 0.23–028 h−1) in the scenarios with windows closed and no PACs in use (i.e., Scenarios 1 and 3) were consistent with those reported in Wallace et al., where the median kt measured in 74 Canadian homes was 1.26 h−1 with a median AER of 0.23 h−1 (Wallace et al., 2013). In the window-open scenario (Scenario 2), kt for the first floor increased markedly due to the increased AER, consistent with a previous study that evaluated the importance of window opening conditions on indoor UFPs in a full-scale test building (Rim et al., 2013). Scenario 7 with three PACs yielded the largest kt for the first floor among all the intervention scenarios, about 6 times larger than that in Scenario 1 and 63%–497% larger than those with a single PAC in use (Scenarios 4–6). In the scenarios using a single PAC, placing it in the kitchen (Scenario 4) led to a much larger kt for the first floor. Notably, using the PAC on the first floor had a minimal effect on kt for the bedroom, and vice versa. Ventilation contributed to 18–26% on average of total UFP decay rates for the first floor when the windows were closed and no PACs were in use. In contrast, the AER/kt ratio increased substantially with the kitchen window open, and decreased to below 10% with a PAC used in the kitchen. The varying ratios further reveal the relative importance of each UFP decay pathways in each scenario.

Table 1.

The total decay rate of indoor UFP number concentrations and air exchange rate in each scenario.

Scenario kt (h−1)
 AER (h−1) AERkt(F1)(%)
Kitchen Living room Bedroom F1 a
S1 1.10 (0.02) NM b 1.64 (0.10) 1.10 (0.02) 0.28 (0.17) 26 (15)
S2 5.09 (0.12) 4.75 (0.15) 0.71 (0.03) 4.92 (0.24) 5.12 (3.18) c 104 (65)
S3 1.42 (0.02) 1.10 (0.04) 0.69 (0.01) 1.26 (0.23) 0.23 (0.22) 18 (18)
S4 4.41 (0.04) 4.40 (0.03) 0.80 (0.07) 4.40 (0.01) 0.36 (0.20) 8 (5)
S5 2.80 (0.02) 2.13 (0.06) 0.42 (0.02) 2.46 (0.48) 0.41 (0.25) 17 (11)
S6 1.29 (0.02) 1.12 (0.03) 2.26 (0.08) 1.20 (0.11) 0.22 (0.18) 19 (15)
S7 7.32 (0.08) 6.99 (0.11) NA d 7.16 (0.24) 0.31 (0.18) 4 (3)
Open in a new tab
a

Floor 1, which includes the kitchen and living room.

b

Not measured.

c

Because of the relatively large measurement error of window-open AER, the AER was even larger than kt.

d

Not applicable because no eligible periods were found for the fitting.

3.4. Emission rates

Fig. 4 displays the time-varying UFP emission rates during and after cooking for each scenario (see Appendix Table A2 for the statistical summary). Remaining negligible during Minutes 0–3, the emission rates began to increase from Minute 4 (about 2 min after the steak was added), peaked at Minutes 6–8 (Dish 1) and 15–18 (Dish 2), and then declined to 0 gradually about 5–10 min after cooking. The time-varying pattern confirms our hypothesis that the cooking-related UFP emission rates varied considerably with time. Hence, the emission rates cannot be estimated via a statistical fitting approach by assuming a constant value during cooking. The method of using a more discreet time step (i.e., 1 min), as in the present study, will likely yield more accurate estimates.

Fig. 4.

Fig. 4.

Open in a new tab

Time-series plots of 1-min cooking-related ultrafine particle emission rates for each scenario. “S1–7” represents Scenarios 1–7. Dishes 1 and 2 refer to the steak and asparagus, respectively.

Results also revealed significant UFP emissions within several minutes after cooking. The mean (SD) emission rates during (Minutes 0–16) and 5-min after cooking (Minutes 17–21) in Scenarios 1–7 were 0.8 (1.1) ×1012 and 1.1 (1.2) ×1012 particles/min, respectively. The after-cooking emissions possibly came from food residue in the hot pan. Thus, it would likely be beneficial to take some mitigating measures to reduce emissions not only during but after cooking. In this study, the range hood was turned off shortly after cooking due to the noise issue, so it could not reduce the after-cooking UFP emissions. Some measures, such as keeping the range hood on, covering the pan, removing the pan from the burner, or cleaning the pan immediately after cooking, may reduce the indoor UFP levels after cooking.

The mean (SD) UFP emission rates during (Minutes 0–16) and 5-min after cooking (Minutes 17–21) with the range hood off and kitchen window closed (Scenario 1) were 1.0 (1.3) ×1012 and 1.7 (1.7) ×1012 particles/min, respectively. The during-cooking emission rates in Scenario 2 were about 70% lower than those in Scenario 1, which suggests that larger ventilation rates or airflow velocities can reduce cooking-related emissions. Further investigation is needed to understand the underlying reasons for this observation, which may in turn help setting up control strategies to reduce cooking-related UFP emissions at the source.

On the other hand, the results of this study suggest that it is challenging to control the emissions from pan-frying cooking events given the significant emission variations across Scenarios 3–7 (e.g., emission rates in Scenario 7 were twice those seen in Scenario 6), even though a standard cooking operating procedure was followed. This finding is also supported by a previous study with three repeated experiments for each cooking scenario (Chen et al., 2018). Even when the pan temperature and food weight were kept consistent, the underlying factors specific to a food item, such as the fat content and shape of the food materials, are difficult to control. The emission variations due to such food-specific variability also explain why the mean indoor average UFP level increases in the pooled D&A period in Scenario 7 (PACs used in three locations) were higher than those in Scenario 4 (PAC only used in the kitchen). As shown in Appendix Table A2, the mean (SD) UFP emission rates during and 5-min after cooking in Scenario 7 were 1.3 (1.3) ×1012 particles/min, approximately 60% higher than those in Scenario 4 on average. Thus, the concentration increases in Scenario 7 were 9% higher than those in Scenario 4, although the UFP decay rates for the first floor in the former scenario were ~60% larger (Table 1). Still, similar experiments utilizing homogenous foods (e.g., processed foods) may provide useful insight into the effectiveness of different mitigation strategies.

3.5. Comparison with PM2.5 results

In a companion paper (Xiang et al., 2021a), we assessed the characteristics of cooking-related PM2.5 in the apartment and the impacts of various interventions. Overall, the varying patterns and the impacts of these interventions were consistent for PM2.5 and UFPs. In the no-intervention scenario (Scenario 1), the bedroom concentrations were > 90% lower than the first-floor concentrations for both pollutants. Both studies revealed large variations in pollutant emission rates during cooking and substantial pollutant emissions within 5 min after cooking. Also, both studies found that opening kitchen windows during and after cooking reduced the first-floor pollutant concentrations substantially but increased the bedroom levels after cooking. As for scenarios involving using a PAC, placing it in the kitchen led to better efficacy in reducing indoor PM2.5 or UFP concentrations. Nonetheless, PM2.5 is larger in size and mass than UFPs and thus has the potential to behave differently than the UFPs described in this paper. In the comparison of the two studies, larger decay rates were found for UFPs than PM2.5 in all window-closed scenarios (Appendix Fig. A4), consistent with a previous study that measured PM2.5 and UFP decay rates in 74 Canadian homes and found the total decay rates for PM2.5 were ~17% greater than those for UFPs (Wallace et al., 2013). As a result of the larger decay rates, the lingering effects of cooking on indoor UFPs lasted for a shorter period than PM2.5. Taking Scenario 3 as an example, the lingering effects on UFPs and PM2.5 lasted for ~4 h and ~7 h, respectively.

3.6. Selection of cooking-related UFP mitigation strategies

This study illustrated the impacts of several commonly-used cooking fume mitigation strategies on cooking-related indoor UFP concentrations. Overall, opening kitchen windows can substantially and cost-effectively reduce indoor UFPs attributed to cooking. However, the efficacies vary with the actual window-open AERs, and thus, can be less significant when the AERs are low (see more discussion on window-open AERs elsewhere (Nazaroff, 2021; Xiang et al., 2021a)). Also, the kitchen windows may be physically unopenable in some dwellings or not opened when ambient PM levels are high (Xiang et al., 2021d; Xiang et al., 2019). On the other hand, the efficacies of kitchen range hoods vary widely with brands and styles, e.g., from <15% to >98% based on previous studies (Delp and Singer, 2012; Lunden et al., 2015; Singer et al., 2012; Singer et al., 2017). The relatively low efficiency and large noise (as high as 70 dB) of some range hoods may prevent people from running them for a long time. All the above issues may lead to a relatively low frequency of range hood use and window opening during cooking in real-world settings. This is consistent with a recent study, which investigated residential cooking and kitchen ventilation behaviors in 132 Canadian households and found that only 27% of the cooking activities were conducted with added ventilation (range hood use 10%, window opening 15%, and both 2%) (Sun and Wallace, 2021).

By contrast, the use of PACs is more flexible despite the availability of effective range hoods and openable windows, or the condition of meteorology. This study revealed that the single PAC use resulted in an additional 2–53% reduction of the mean indoor average UFP levels in the pooled D&A period. Remarkably, the placement location impacted the overall efficacy of PAC use. Specifically, placing the PAC in the kitchen, i.e., closer to the source, reduced indoor average UFP concentrations more significantly. However, further investigation on how the placement of a PAC impacts occupants’ time-weighted exposure is needed as exposure is also dependent on people’s time-activity patterns.

3.7. Limitations

This study has several potential limitations. First, the second floor was not included when we estimated the total cooking-related UFP emission rates since we did not measure the AERs in the bedroom. However, the impacts on the emission rate estimates during cooking (Minutes 0–16) should be negligible because the bedroom level increase during cooking was minimum. Although the after-cooking emission rates (Minutes 17–21) could be underestimated, especially in Scenario 2 (open kitchen window), where UFP levels in the bedroom clearly occurred, such underestimates do not change our conclusion that it is meaningful to take some measures to reduce such emissions not only during but after cooking. Second, we did not measure ambient UFP levels, which could result in some biases in estimating the UFP emission rates. Nonetheless, as explained in the Methods, our estimates of outdoor UFP concentrations were similar to those measured ~1.5 km away from the experimental site. Third, the variations in UFP emission rates from pan-frying cooking scenarios across different trials were not fully controlled, although the same standard operating procedure was followed. Future studies will benefit from a more controlled emission source. Fourth, in the window-open scenario (Scenario 2), we did not examine similar situations where the windows elsewhere in the apartment (e.g., bedroom) or main entry door were open (Scenario 2), both of which could significantly alter the indoor airflow and, thus, the spatial distribution of indoor air pollutants. Lastly, the quantitative results obtained in the present study are specific to the pan-frying cooking scenario and the apartment where the experiments were conducted. Future studies with more cooking scenarios and housing types are warranted.

In spite of these limitations, this study provides useful information on the characteristics of cooking-related UFP levels with high-emitting cooking activities (e.g., pan-frying and stir-frying) in a residence and a sense of the magnitude of the reduction in indoor UFP levels that may be achieved by utilizing one or more strategies.

4. Conclusions

This study illustrates the large variations in indoor UFP levels and emission rates during and after pan-frying cooking events. The 1-min mean UFP number concentrations in the kitchen and living room peaked at levels of 150000–500000 particles/cm3. In contrast, the bedroom UFP concentrations were consistently low except for the window-open scenario. Related, the lingering effects of cooking on indoor UFPs lasted for up to 4 h, despite cooking time being short (< 20 min in this study). Large variations in the 1-min UFP emission rates were found from the cooking events, with a mean (SD) of 0.8 (1.1) ×1012 and 1.1 (1.2) ×1012 particles/min during and with 5 min after cooking. Given the substantial UFP emissions in both periods, it would likely be beneficial to take some mitigating measures to reduce emissions not only during but after cooking. Compared with the no-intervention scenario, keeping the kitchen windows open and using a kitchen range hood reduced the mean indoor average UFP concentrations during and 1 h after cooking by ~70% and ~35%, respectively. Along with the range hood on, utilizing a PAC in the kitchen during and after cooking further reduced the mean indoor average UFP levels during and 1 h after cooking by an additional 53%. In contrast, placing the PAC in the living room or bedroom resulted in worse efficacy, with additional 2–13% reductions. These findings provide useful information on how to reduce cooking-related UFP exposure via readily accessible intervention strategies.

Supplementary Material

1

Fig. 1.

Fig. 1.

Open in a new tab

Timeline of the cooking protocol.

Highlights.

  • The time-varying profile of cooking-related UFPs in a residence was examined.

  • The effect of various cooking-fume intervention ways on indoor UFPs was evaluated.

  • Large variations were found in 1-min UFP emission rates during and after cooking.

  • Opening kitchen windows can reduce indoor average UFP concentrations by 70%.

  • Placing an air cleaner in the kitchen resulted in better efficacy than other locations.

Acknowledgments

The study was funded by the National Institute of Environmental Health Sciences (5R33ES024715-05).

Footnotes

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix

The Appendix is provided.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Austin E, Xiang J, Gould T, Shirai J, Yun S, Yost M, et al. Mobile ObserVations of Ultrafine Particles: The MOV-UP study report. University of Washington, Seattle, WA, 2019. [Google Scholar]
  2. Austin E, Xiang J, Gould TR, Shirai JH, Yun S, Yost MG, et al. , 2021. Distinct ultrafine particle profiles associated with aircraft and roadway traffic. Environmental Science & Technology 55, 2847–2858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bennett DH, Koutrakis P, 2006. Determining the infiltration of outdoor particles in the indoor environment using a dynamic model. Journal of Aerosol Science 37, 766–785. [Google Scholar]
  4. Chen C, Zhao Y, Zhao B, 2018. Emission rates of multiple air pollutants generated from Chinese residential cooking. Environmental science & technology 52, 1081–1087. [DOI] [PubMed] [Google Scholar]
  5. Delp WW, Singer BC, 2012. Performance assessment of US residential cooking exhaust hoods. Environmental science & technology 46, 6167–6173. [DOI] [PubMed] [Google Scholar]
  6. Gabdrashova R, Nurzhan S, Naseri M, Bekezhankyzy Z, Gimnkhan A, Malekipirbazari M, et al. , 2021. The impact on heart rate and blood pressure following exposure to ultrafine particles from cooking using an electric stove. Science of The Total Environment 750, 141334. [DOI] [PubMed] [Google Scholar]
  7. Gao YT, Blot WJ, Zheng W, Ersnow AG, Hsu CW, Levin LI, et al. , 1987. Lung cancer among Chinese women. International journal of cancer 40, 604–609. [DOI] [PubMed] [Google Scholar]
  8. Huang C-H, Xiang J, Austin E, Shirai J, Liu Y, Simpson C, et al. , 2021. Impacts of using auto-mode portable air cleaner on indoor PM2.5 levels: An intervention study. Building and Environment 188, 107444. [Google Scholar]
  9. Ke Y, Cheng J, Zhang Z, Zhang R, Zhang Z, Shuai Z, et al. , 2009. Increased levels of oxidative DNA damage attributable to cooking-oil fumes exposure among cooks. Inhalation toxicology 21, 682–687. [DOI] [PubMed] [Google Scholar]
  10. Li H, Li X, Qi M, 2014. Field testing of natural ventilation in college student dormitories (Beijing, China). Building and Environment 78, 36–43. [Google Scholar]
  11. Liu L-JS, Phuleria HC, Webber W, Davey M, Lawson DR, Ireson RG, et al. , 2010. Quantification of self pollution from two diesel school buses using three independent methods. Atmospheric environment 44, 3422–3431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Lunden MM, Delp WW, Singer BC, 2015. Capture efficiency of cooking‐related fine and ultrafine particles by residential exhaust hoods. Indoor Air 25, 45–58. [DOI] [PubMed] [Google Scholar]
  13. Nazaroff WW, 2021. Residential air-change rates: A critical review. Indoor Air 31, 282–313. [DOI] [PubMed] [Google Scholar]
  14. Neghab M, Delikhoon M, Baghani AN, Hassanzadeh J, 2017. Exposure to cooking fumes and acute reversible decrement in lung functional capacity. Int. J. Occup. Environ. Med. 8, 207–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Pan C-H, Chan C-C, Wu K-Y, 2008. Effects on Chinese restaurant workers of exposure to cooking oil fumes: a cautionary note on urinary 8-hydroxy-2′-deoxyguanosine. Cancer Epidemiology, Biomarkers & Prevention 17, 3351–3357. [DOI] [PubMed] [Google Scholar]
  16. Presto AA, Saha PK, Robinson AL, 2021. Past, present, and future of ultrafine particle exposures in North America. Atmospheric Environment: X 10, 100109. [Google Scholar]
  17. R Core Team, 2013. R: A language and environment for statistical computing. URL http://cran.univ-paris1.fr/web/packages/dplR/vignettes/intro-dplR.pdf (accessed on December 7, 2020).
  18. Rim D, Wallace LA, Persily AK, 2013. Indoor ultrafine particles of outdoor origin: Importance of window opening area and fan operation condition. Environmental Science & Technology 47, 1922–1929. [DOI] [PubMed] [Google Scholar]
  19. Seow A, Poh W-T, Teh M, Eng P, Wang Y-T, Tan W-C, et al. , 2000. Fumes from meat cooking and lung cancer risk in Chinese women. Cancer Epidemiology, Biomarkers & Prevention 9, 1215–1221. [PubMed] [Google Scholar]
  20. Singer BC, Delp WW, Price P, Apte M, 2012. Performance of installed cooking exhaust devices. Indoor Air 22, 224–234. [DOI] [PubMed] [Google Scholar]
  21. Singer BC, Pass RZ, Delp WW, Lorenzetti DM, Maddalena RL, 2017. Pollutant concentrations and emission rates from natural gas cooking burners without and with range hood exhaust in nine California homes. Building and Environment 122, 215–229. [Google Scholar]
  22. Singh A, Nair KC, Kamal R, Bihari V, Gupta MK, Mudiam MKR, et al. , 2016. Assessing hazardous risks of indoor airborne polycyclic aromatic hydrocarbons in the kitchen and its association with lung functions and urinary PAH metabolites in kitchen workers. Clinica Chimica Acta 452, 204–213. [DOI] [PubMed] [Google Scholar]
  23. Stephens B, Siegel JA, 2012. Penetration of ambient submicron particles into single-family residences and associations with building characteristics. Indoor Air 22, 501–513. [DOI] [PubMed] [Google Scholar]
  24. Sun L, Wallace LA, 2021. Residential cooking and use of kitchen ventilation: The impact on exposure. Journal of the Air & Waste Management Association 1–14. [DOI] [PubMed] [Google Scholar]
  25. Sun Z, Liu C, Zhang Y, 2019. Evaluation of a steady-state method to estimate indoor PM2.5 concentration of outdoor origin. Building and Environment 161, 106243. [Google Scholar]
  26. Wallace L, Kindzierski W, Kearney J, MacNeill M, Heroux ME, Wheeler AJ, 2013. Fine and ultrafine particle decay rates in multiple homes. Environmental Science & Technology 47, 12929–12937. [DOI] [PubMed] [Google Scholar]
  27. Wallace LA, Emmerich SJ, Howard-Reed C, 2004. Source strengths of ultrafine and fine particles due to cooking with a gas stove. Environmental Science & Technology 38, 2304–2311. [DOI] [PubMed] [Google Scholar]
  28. Wan M-P, Wu C-L, To G-NS, Chan T-C, Chao CY, 2011. Ultrafine particles, and PM2. 5 generated from cooking in homes. Atmospheric Environment 45, 6141–6148. [Google Scholar]
  29. Wang T-j, Zhou B-s, Shi J-p, 1996. Lung cancer in nonsmoking Chinese women: a case-control study. Lung cancer 14, S93–S98. [DOI] [PubMed] [Google Scholar]
  30. Weschler CJ, 2016. Roles of the human occupant in indoor chemistry. Indoor air 26, 6–24. [DOI] [PubMed] [Google Scholar]
  31. Wu-Williams A, Dai X, Blot W, Xu Z, Sun X, Xiao H, et al. , 1990. Lung cancer among women in north-east China. British journal of cancer 62, 982. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Xiang J, Austin E, Gould T, Larson TV, Yost M, Shirai JH, et al. , 2020. Using vehicles’ rendezvous for in-situ calibration of instruments in fleet vehicle-based air pollution mobile monitoring. Environmental Science & Technology 54, 4286–4294. [DOI] [PubMed] [Google Scholar]
  33. Xiang J, Hao J, Austin E, Shirai J, Seto E, 2021a. Residential cooking-related PM2.5: Spatial-temporal variations under various intervention scenarios. Building and Environment 201, 108002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Xiang J, Huang C-H, Austin E, Shirai J, Liu Y, Seto E, 2021b. Energy consumption of using HEPA-based portable air cleaner in residences: A monitoring study in Seattle, US. Energy and Buildings 236, 110773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Xiang J, Huang C-H, Shirai J, Liu Y, Carmona N, Zuidema C, et al. , 2021c. Field measurements of PM2.5 infiltration factor and portable air cleaner effectiveness during wildfire episodes in US residences. Science of The Total Environment 773, 145642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Xiang J, Seto E, Mo J, Zhang J, Zhang Y, 2021d. Impacts of implementing Healthy Building guidelines for daily PM2.5 limit on premature deaths and economic losses in urban China: A population-based modeling study. Environment International 147, 106342. [DOI] [PubMed] [Google Scholar]
  37. Xiang J, Weschler CJ, Mo J, Day D, Zhang J, Zhang Y, 2016. Ozone, electrostatic precipitators, and particle number concentrations: Correlations observed in a real office during working hours. Environmental Science & Technology 50, 10236–44. [DOI] [PubMed] [Google Scholar]
  38. Xiang J, Weschler CJ, Wang Q, Zhang L, Mo J, Ma R, et al. , 2019. Reducing indoor levels of “outdoor PM2.5” in urban China: impact on mortalities. Environmental Science & Technology 53, 3119–3127. [DOI] [PubMed] [Google Scholar]
  39. Xiang ZY, Wang HL, Stevanovic S, Jing SG, Lou S, Tao S, et al. , 2017. Assessing impacts of factors on carbonyl compounds emissions produced from several typical Chinese cooking. Building and Environment 125, 348–355. [Google Scholar]
  40. Zhao Y, Zhao B. Emissions of air pollutants from Chinese cooking: A literature review. Building Simulation. Springer, 2018, pp. 1–19. [Google Scholar]
  41. Zhong L, Goldberg MS, Gao Y-T, Jin F, 1999. Lung cancer and indoor air pollution arising from Chinese-style cooking among nonsmoking women living in Shanghai, China. Epidemiology 10, 488–494. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1731551-supplement-1.docx (1.1MB, docx)

RESOURCES