Commuter exposure to particulate matter in urban public transportation of Xi’an, China

Abstract

Purpose

To investigate commuter exposures to particulate matter (PM) in urban public transportation buses and subways, PM concentrations were simultaneously monitored for these two modes, over the same routes, in Xi’an, China.

Methods

The microenvironment variabilities in each stage of the total trip were analyzed. Exposure doses for the different commute processes were estimated based on the heart rates of volunteers. Experimental measurements were taken during peak traffic hours in July and October (summer and autumn) on two typical commute routes, for a total of 36 trips. One-way ANOVA was used to analyze the effects of different variables on commuter exposures.

Results

On the same route, the average PM exposure concentration of bus commuters was higher than those of subway commuters. For example, on Route 1 in the case study, the average PM10, PM2.5, and PM1 exposure concentrations of bus commuters were 71.6%, 19%, and 10.4% higher, respectively, than those of subway commuters. In the ground transportation mode, the exposure concentration of bus commuters was affected by the type of vehicle. Particle concentrations were significantly higher inside compressed natural gas (CNG) buses, than in pure electric (PE) buses, and in summer, the PM10 concentration in a CNG bus was 4.3 times higher than that in a PE bus. In a CNG bus, commuters in the back door area suffered the highest PM10 exposure concentration (179.6 μg/m³), followed by those in the rear of the carriage (142.8 μg/m³), and then those in the front door area (105.4 μg/m³).

Conclusion

Commuters’ avoidance of ground traffic sources, effective ventilation systems in buses, and the use of screens in subway systems can all help to lower the PM exposure of commuters. For all the modes of transportation in our study, the hottest spots for PM exposure appeared in the period when commuters were waiting for transit vehicles to arrive.

Introduction

Particulate matter (PM)—one of the major traffic-related pollutants—is a major contaminant associated with a variety of adverse health outcomes, including respiratory diseases, cardiovascular events, hospitalization and even mortality [1]. Exposure to PM during commuting is the main contributor to total daily exposure for most urban residents. A previous study of Chinese behavior patterns showed that urban residents spend an average of 3.5 h day⁻¹ performing outdoor activities, and that adult residents spend an average of 87 min day ⁻¹ in traffic [2], contributing substantially to daily personal exposure and related health effects [3]. Another study, in Seoul, Korea [4], showed that the total contribution of transportation (i.e., walking, buses, subway, and other transportation methods) to PM 10 exposure was 12.4% in summer and 16.9% in winter; to PM 2.5 exposure, 7.8% in summer and 12.4% in winter. Still other studies have shown that short (1 h) and even very short (<1 h) exposures to traffic emissions exacerbate existing pulmonary and cardiovascular diseases [5–8].

Many metropolitan areas are developing policies that encourage the use of public transportation for commuting [9, 10]. Public transportation relieves the pressure of ground transportation to a certain extent, effectively reducing the emissions of motor vehicle pollutants and energy consumption. Common public transportation methods include buses, subway trains, light rail, public bicycles, taxis, etc. Among these, subway trains and buses make up the majority of public transit trips, because of their advantages of speed, convenience, and their ability to carry a large volume of passengers. Bus and subway passengers, however, are inevitably exposed to PM from vehicle exhaust emissions and piston effects, especially in subways. In public transportation, a large number of people are gathered together in a small space, increasing the severity of travelers’ air pollution exposure. Several studies have specifically investigated the health hazards caused by the transportation environment and the damages to human health caused by air pollution in the public transportation system [11, 12]. The results all indicate that there is an urgent need to address the pollution problem in the public transportation systems.

Grana et al. [13] investigated commuters’ exposure to ultrafine particles in different transportation modes (car, motorbike, subway, bus) in the city of Rome. Their results suggest that the lowest UFP exposures are experienced by underground train commuters, with an average particle number concentration of 14,134 cm⁻³, largely because the routes are at greater distance from vehicular traffic, than are other modes. Tsai et al. [14] compared commuters’ exposures to PM while they were using motorcycles, cars, buses, and mass rapid transit (MRT) on identical routes in Taipei, and the results showed that motorcycle commuters and car commuters were exposed to the highest and lowest concentrations of PM, respectively. Onat et al. [15] investigated personal exposure to PM2.5 under four public transportation modes: bus, metro–bus, car and walking. They found that the highest average PM2.5 concentration exposure was inside a bus during rush hours, and that commuters inside buses and those walking were exposed to higher coarse particle concentrations (>1 μm) than were commuters in metro–buses or cars. Knibbs et al. [16] investigated personal exposure to ultrafine particles and PM2.5 in four transportation modes (train, bus, ferry and automobile) in Sydney, Australia, and their results showed that the highest UFP exposure occurred during bus trips.

However, these studies focused on comparisons between different commuting modes, and ventilation changes are rarely accounted for in current commuting exposure studies. Using exposure instead of dose can cause measurement errors and can therefore bias exposure−response relationships [17]. Combining commuting mode, exposure time, and breathing rate is an important basis for accurately assessing the health risks of commuting exposure [18]. Therefore, a dose model was used in our study, because it can estimate ventilation and hence the inhaled pollutant load, thus improving the assessment of the effect of pollutants on health. In addition, the inhaled dose during each mode of transportation can be evaluated by the measured concentration, the time spent in various microenvironments, and the estimated volume of air exchanged with each breath.

Xi’an, the largest metropolis in northwestern China, has a total population of 961.67 million. Subway and bus are the main modes of travel for the people of Xi’an. We studied the particulate exposure of commuters in buses and subways, with the following objectives: 1) to assess the contribution of different microenvironments to overall commuting exposure; 2) to identify the greatest possible range of causes of variation in the air quality inhaled by bus and subway commuters; and 3) to combine commuting mode, travel time and inhaled dose, to comprehensively assess commuters’ exposure risk .

Methods

Study design and routes

All experiments were carried out on working days during July and October 2018 (summer and autumn). Measurements were performed at the same time of day for all transportation modes during the peak traffic hours (7:00–9:00 and 17:00–19:00). Two typical commuting routes were selected as the sampling routes, as they covered both commuting modes: bus and subway, as shown in Fig. 1. Route 1 is an arterial road that connects the two most prosperous business districts of Xiaozhai and North Street; these two areas are densely populated and experience a high flow of vehicles on their busy streets. Route 2 is a two-way ten-lane expressway with an hourly traffic volume of 8873 vehicles. The bus and subway trips had similar origins and destinations. For the bus trips, two types of vehicles were selected: CNG-powered and PE-powered. Xi’an City has generally promoted the modernization of public transportation, especially green and low-carbon modes, and the overall level of pure electrification in the Xi’an bus system is already ranked in the forefront over the entire country. By the beginning of 2018, the total number of PE buses had reached 3650, making Xi’an the largest city in the northwest region operating PE buses, and ranking among the best in the promotion of PE buses nationwide.

Fig. 1.

Map of field sampling routes

The three subway lines currently in operation have greatly eased the pressure for adequate ground transportation; in 2017, Xi’an Metro had a daily average passenger flow of more than 1.8 million. Two subway lines (Line 2 and Line 3) with the largest passenger volume, were selected as the subway research routes. The subway departure frequency is 1.5 min during the shifts studied. Detailed information about the experimental routes is shown in Table 1.

Table 1.

Description of the two sampling routes

Route	Length (km)	Lanes	Intersections	Bus stops	Subway stops	Bus speed (km/h)	Average Traffic volume (vehicles/h)
							car	bus	truck
1	5.8	12	5	8	5	12.34	4557	356	159
2	7.0	10	3	9	5	25	8665	50	158

Two experimenters carried instruments and set off at the same time, taking the bus and subway, respectively, to the same destination. Each trip included time spent waiting for the bus/subway, time spent riding inside the bus/subway, and time spent walking to and from the bus/subway, including transfer times between vehicles. A total of 40 trips were completed, and no measurements were taken during days with rain or smog pollution. After discarding measurements affected by technical problems with the instruments, 19 and 17 trips, for the bus and subway modes, respectively, were analyzed. The average (one-way) bus-trip duration was around 36 min, for subway, around 22 min. In order to account for day-to-day variability in outdoor particle concentrations, background measurements were carried out at the Xiaozhai Monitoring Station, and averaged, to obtain two background values for each day.

Instruments

A set of portable instruments was used for real-time measurements of exposure levels in the above-mentioned public transportation modes, as shown in Fig. 2. Two portable aerosol optical particle size spectrometers (Grimm 11-A) were used for counting total particle concentrations in the size range of 0.25–32 μm at 6-s intervals; the instruments were placed in a specially designed backpack. Each volunteer wore a TomTom Runner2 sports bracelet with GPS positioning to monitor and record his/her heart rate, which is an important indicator for calculating respiratory dose. Environmental variables (such as wind speed and direction) were obtained from a Kestrel 4500 Pocket Weather Tracker. All instruments were calibrated and operated simultaneously together, before each experiment. Temperature and relative humidity were also monitored, to ensure that all PM monitors were functioning under the range of preferred environmental conditions. Additional on-site personnel were responsible for recording traffic and other conditions, such as bus door opening time, passenger capacity, red light waiting time, other vehicles in the vicinity of the bus, and delays caused by traffic congestion, during each commute run. All the time data were read from the Grimm11-A to ensure consistency between the special case and the experimental data.

Fig. 2.

Instruments

Dose model

The mode of activity has tremendous effects on commuter PM exposure. The breathing rate varies with the activity mode, and directly affects the exposure dose per unit of time. Hence the combination of exposure concentration and breathing rate provides a direct measurement of the effects. In a recent assessment of current methods, Pablo et al. [17] concluded that the choice of method depends on the objectives, and on the scale and length of the study, and they suggested using a continuous method based on an exponential function, to better reflect changes in VE for different physical activities over a relatively long period of time. The dose model proposed by Campos et al. [19] was selected in our study.

$$\mathit{VE}=e^{0.58+\left(0.025\times \mathit{HR}\right)}$$ 1

where VE is the inhalation rate, defined as the amount of air entering the lungs per minute, calculated by the heart rate and related oxygen consumption; and HR is the heart rate of an individual volunteer. The inhaled dose was calculated on a 1-min time resolution as the product of the PM concentration and the per-minute ventilation and then linearly extrapolated to estimate the inhaled dose during the entire trip.

Data processing

All descriptive computations and statistical analyses were made using SPSS software (version 22.0). For each transportation mode, linear regression models were used to compare concentrations between different commuting modes. Real-time PM concentrations for different commuting modes were treated as dependent variables, and commuting microenvironments as fixed-effect covariates. The relative contributions to total daily exposure were calculated by multiplying median time spent for each activity by the related PM concentration, and dividing by the sum of all the activity doses. One-way ANOVA was used to analyze the effects of different variables (commuting mode, vehicle type, seating location, route) on commuter exposures. Pearson correlation analysis and scatterplots were used to analyze the relationship between the concentrations inside vehicles and ambient concentrations, and a correlation of p values <0.05 was considered statistically significant.

Results and discussion

Concentrations of PM exposure under different commuting modes

Ambient temperature and RH ranged from 8 °C to 33 °C (average was 20.3 °C) and 14%–88% (average was 57%), respectively. The prevailing wind was from the southwest, with an average speed of 1.5 m s ⁻¹ during the experiment. The total exposure for the entire trip and the spatial variation in each traffic microenvironment was studied. Descriptive data on the mass concentrations of PM in different microenvironments are shown in Table 2. There are significant differences in PM concentration, between different microenvironments, as can be seen in Fig. 3.

Table 2.

Summary of PM concentrations in different commuting microenvironments.

Research microenvironment	PM concentration(μg/m3)
	N(trips)	PM 10			PM 2.5			PM 1
		mean	SD	range	mean	SD	range	mean	SD	range
Roadside walking	32	149.4	53.5	36.4 ~ 383	69.8	26.4	8 ~ 127	53.7	23.8	6.5 ~ 105
Bus stop	17	172.0	55.0	86.2 ~ 393.5	72.4	19.6	25.6 ~ 128.7	51.4	21.8	12.6 ~ 107
Platform (subway)	24	102.3	35.0	23.9 ~ 201	54.0	14.6	19 ~ 97.5	39.3	10.6	11.9 ~ 55
CNG bus (OW)	10	130.0	61.0	23.2 ~ 364	58.0	28.0	10.7 ~ 136.5	39.3	22.0	8.3 ~ 88.3
PE bus (AC)	27	24.3	12.5	6.9 ~ 98	16.7	5.5	6 ~ 55.2	15.0	4.7	4.9 ~ 40
Subway	21	68.0	17.7	16.3 ~ 127	46.3	11.3	16.3 ~ 74.7	35.6	8.0	14 ~ 55.5

Fig. 3.

PM concentrations in major environments during the sampling periods (The upper and lower whiskers represent the ranges of 5% to 95%, the small box represents the average, and the upper and lower lines of the box represent the 25th and 75th quartiles)

OW: open windows; AC: air conditioner turned on.

Bus mode

For bus commuters, a single trip was split into four sections (walking along the roadside to and from the stop, waiting at the stop, and actual travel on the bus). Two types of buses were selected: CNG-powered and PE-powered, as they have the different ventilation modes (CNG buses usually have open windows to facilitate air exchange, while PE buses generally keep the windows closed and the air conditioners turned on). The measurements and statistical results showed that the overall average particle concentrations of PM10 varied from 24.3 to 172 μg/m³, and the average personal exposure concentrations of PM10 were 24.3, 130, 149.4 and 172μg/m³, when commuting by PE bus, commuting by CNG bus, walking alongside the road, and waiting at the stop, respectively (the stop-waiting phase is a hot spot for PM exposure). Average personal concentrations of PM2.5 ranged from 16.7 to 72.4μg/m³. Similarly, the mean personal exposure concentrations of PM2.5 tended to be highest when waiting at the bus stop (72.4μg/m³), compared with riding on a CNG bus (58μg/m³), walking (69.8μg/m³), and riding on a PE bus (16.7μg/m³). PM1 concentrations when walking alongside the road were the highest, followed by at the bus stop, riding on a PE bus and riding on a CNG bus.

Differences in pollutant concentrations inside buses may be affected by the time of day, with higher concentrations typical in the morning compared with the afternoon, especially for PM10. This may be due to morning rush hour nucleation events related to higher emissions of condensable substances, and to lower temperatures [13]. Figure 4 shows the changes in PM concentration inside CNG buses during the morning and evening peak periods.

Fig. 4.

Typical pollutant time series for PM 10, PM 2.5 and PM 1 concentrations inside CNG buses during the morning and evening peak periods

Our measurements showed that the bus-stop waiting population is exposed to high concentrations of PM. Bus stops in urban areas generally serve several or even as many as a dozen bus lines, making bus stops at the site very frequent. After entering the bus stop area, a bus will generally experience three stages: inbound (decelerating), idling, and outbound (accelerating, emitting the largest amount of pollutants). When the bus stops at a site near the roadside, the pollutants, especially PM, are not very diffused during this short period of time, and basically linger near the source. The health of the commuters waiting at the stop consequently suffers.

Subway mode

The first Xi’an Metro line was opened on September 16, 2011, making Xi’an the first city in northwestern China to open a subway. By September 2018, three subway lines were in operation, with a total length of 91.35 km and a total of 66 platforms; the highest daily passenger count reached 2.648 million. Similar to a bus itinerary, a subway trip also includes walking to and from the platform, waiting, and actual travel on the subway, in addition to negotiating a section of underground access to the waiting point. Real-time monitoring data indicated that PM concentrations on the subway platforms (PM10: 119 μg/m³, PM2.5: 55.5 μg/m³, PM1: 39.6 μg/m³) were relatively higher than inside the carriage (PM10: 75 μg/m³, PM2.5: 47.5 μg/m³, PM1: 36 μg/m³). Figure 5 is a comparison of the concentrations of PM inside and outside the subway carriage.

Fig. 5.

Comparison of PM concentrations, between subway carriage and platforms

The main sources of PM on the platforms include the dirty air entering the platform from the tunnel (especially coarse particles), re-suspension of these accumulated PM caused by air turbulence, and particles carried by passengers from the external environment. Particles enter the platform from the outside through the ventilation system, the elevator system and wayside gates, and from mechanical abrasion while trains are running or braking. The completely enclosed structure of the underground metro system makes the particles difficult to exhaust, and they accumulate over time and are re-suspended under the movements of commuters and piston winds, resulting in still higher PM concentrations. Upon the departure of a train from the platform, PM air quality deteriorates as the in-draught dilution effect declines and the train produces PM resuspension after passing through the station [20]. PM levels measured in the system were in the range of those measured in Los Angeles [21] and Taipei [22] but were lower than those from Beijing [12], Seoul [23] and Barcelona [24].

Factors affecting the concentration of PM

Seating location

A study in Stuttgart (Germany) [25] explored the impact of seating location on black carbon (BC) exposure in public transit buses, for vulnerable groups. The results of this study suggest that although the average BC exposure of passengers is not affected by seating location, the priority seating area in the middle of the bus is associated with a higher likelihood of experiencing a peak in concentration, compared to seating at the back of the bus. In our study, the interior of the CNG bus was divided into three parts. The measurement locations used within the CNG bus are shown in Fig. 6: front door, back door and rear of the carriage.

Fig. 6.

Measurement locations inside the CNG bus

Table 3 is the description of PM concentrations in different seating locations inside the CNG bus. The concentration of PM at the back door was highest, where the average concentration of PM10 reached 179.6μg/m³, followed by the rear of the carriage (142.8μg/m³), and finally the front door (105.4μg/m³). At a bus stop, the back door is frequently opened and closed in order to pick up and drop off passengers, and since the area of the bus near the back door is relatively large, this action facilitates the entry of pollutants in the external environment into the vehicle when the door is open. For bus commuters, PM 10, PM 2.5 and PM1 concentrations were significantly increased, by 33% (from 160.6 to 213μg/m³), 6% (from 45.9 to 48.6μg/m³) and 3% (from 15.8 to 16.3μg/m³), respectively, when the doors were opened compared to when they were closed. These figures are consistent with those in Taipei [14]: that study showed that higher in-vehicle PM10 concentrations occurred when the bus doors were opened. Not only do ambient air pollutants enter the cabin during this time, but the particles may also be re-suspended. When commuters get in the car, they usually move to the back door, and the airflow generated by this movement drives more particles to accumulate at the back door area. Figure 7 shows a comparison of particle concentrations at different positions inside the bus. However, we did not find this difference in the subway system, probably because the subway system is equipped with a screen door that opens simultaneously with the train doors and can better prevent the invasion of external pollutants. The study in Korean indicated that the mean PM 10 concentration (after the introduction of screen doors) was significantly reduced, by 16%, relative to the period before the doors were installed [26].

Table 3.

PM concentrations at different seating locations inside a CNG bus

	PM10(μg/m3)				PM2.5(μg/m3)				PM1(μg/m3)
	mean	max	min	SD	mean	max	min	SD	mean	max	min	SD
front door	105.4	160.7	51.1	21.6	39.6	54.1	24.3	6.3	17.15	19.7	13.7	1.3
back door	179.6	335.0	61.5	91.2	53.2	78.7	24.3	8.7	17.26	19.6	12.9	1.2
rear of the carriage	142.8	248.5	86.2	39.0	50.7	77.4	34.3	9.5	17.18	20.3	14.3	1.4

Fig. 7.

Comparison of PM concentrations at different locations inside a CNG bus

(Logarithmic conversion of concentration data; FD indicates front door; BD back door; and RC rear of carriage)

Vehicle type

Over the same route, the concentrations of PM in CNG and PE buses were measured in real time. Figure 8 shows the comparison of PM concentrations in the two kinds of vehicles, in different seasons.

Fig. 8.

Comparison of PM concentration in two kinds of vehicles in different seasons

A vehicle’s ventilation conditions exert a significant influence on cabin pollutants. Lower values of PM were found in AC buses, compared to non-AC buses, in our study, a result that agrees with those reported in Bangkok [27], and PE buses are indeed more airtight. Airtightness plays an important role in shielding commuters from the highest PM concentrations, reducing both mean and maximum values. Not surprisingly, only in summer, when the air conditioning is running in PE buses, can an obvious difference in the concentration of PM in the two types of buses be observed. Open windows increase the risk of commuter exposure, while an air conditioner can filter out some of the coarse particles to effectively reduce the exposure concentration. This study demonstrated that bus drivers may be able to reduce their commute exposure by being cognizant of their pollutant environment and applying dynamic behavior modification to adapt to changing scenarios.

SOW: summer open window; AOW: autumn open window (both for CNG bus).

SAC: summer air conditioner turned on; ACW: autumn closed window (both for PE bus).

Route characteristics

Real-time monitored data demonstrated that in-cabin PM concentrations were strongly affected by route type, and lower exposure values were obtained for commuters riding in a CNG bus on an arterial road. Figure 9 shows a comparison of PM2.5 concentrations inside CNG buses, between the two routes. The driving speed of buses varied from 12.34 to 25 km/h between the arterial road and the expressway. Several studies have suggested that the pollutant levels inside buses decrease with increased driving speed [28, 29], but the opposite conclusion was drawn in our research, perhaps because of the high volume of traffic on Route 2, and the large amount of exhaust emissions exacerbated the pollution level of the road. The measured concentration of pollutants in a vehicle is greatly affected by road emissions. Studies have shown that higher driving speeds (from 20 mph to 60 mph) could result in higher ventilation rates (up to 3.4 times as high) inside the bus. Increasing the speed results in an increase in particle concentration and a concomitant peaking of PM concentration, possibly due to a higher rate of air leakage from outdoors. Higher external concentrations of pollutants enter the car, worsening the level of pollutants in the car. Although the concentrations of PM measured in the two subway lines we selected differed, the difference is not obvious in general. Passengers who changed from Line 2 (corresponding to Route 1) to Line 3 (corresponding to Route 2) had to go downstairs, as Line 3 runs deeper underground, and the accumulated pollutants there are not easily diffused. These differences were also the cause of the difference in concentration between the two lines.

Fig. 9.

Comparison of PM2.5 concentrations inside CNG buses, between the two routes. Route 1 is an arterial road, Route 2 an expressway

Ambient concentration

Many studies have shown that transportation systems can be strongly influenced by PM from ambient air. In Nantes (France), Muresan et al. [29] found high PM correlations in the public transportation system, between platform concentrations and direct ambient street concentrations. Leavey et al. [30] simultaneously measured selected pollutants outside and inside an on-road car, and found that outdoor concentrations had a strong effect on cabin concentrations. And a previous study in Los Angeles [21] found that PM levels of underground rail systems were positively correlated with outdoor concentrations.

To investigate the influence of ambient PM concentrations on bus and subway commuting, correlation analysis was performed on environmental conditions and exposure concentrations for 16 groups of bus trips and 14 groups of subway trips. The ambient concentrations during the experiment and the commuter exposure concentrations of the buses and subways are shown in Table 4. During bus trips, the Pearson correlation coefficient of PM10 was 0.719, and of PM2.5, 0.839, both of which were significant at 0.01. For subway commuting, the Pearson correlation coefficient of PM 10 was 0.681—significant at 0.01, and of PM 2.5, 0.556—significant at 0.05. Ground-level commuting (bus trips) is more influenced by ambient PM levels than is underground commuting (subway trips). The positive correlation between PM levels in the ambient concentration and inside transportation vehicles emphasizes the influence of surrounding pollution (Fig. 10).

Table 4.

Ambient concentration and exposure concentration during the experimental period

Date		PM 10 (μg/m3)			PM 2.5 (μg/m3)
		Ambient	Bus (OW)	Subway	Ambient	Bus (OW)	Subway
July14	Morning	101	248		35	77
	Afternoon	131	166		43	45
July16	Morning	20	95	48.6	13	40	28.5
	Afternoon	88	92	49	38	47	31
July17	Morning	16	60	46	9	25	27.8
	Afternoon	83	62	52	35	32	31
October 10	Morning	84	138	64.5	28	42	34.6
	Afternoon	28	71	59.6	9	22	29.6
October 12	Morning	172	194	85.7	84	85	51.8
	Afternoon	101	144	72.7	68	68	44.8
October 15	Morning	96	215	88.5	61	110	59
	Afternoon	109	162	71	74	80	47
October 16	Morning	68	85	82	33	49	48
	Afternoon	164	203	77	99	99	56
October18	Morning	72	44	61	143	64	39.5
	Afternoon	63	36	58.5	97	46	38.6

Fig. 10.

Correlation analysis between environmental concentration and exposure concentration during the experiment

The concentration of PM during CNG bus trips under OW mode is positively correlated with the ambient concentration, because of the proximity to road traffic emissions and because a negative pressure layer is formed on the outer surface of the cabin under the state of motion; these conditions allow air to be exchanged between the interior and exterior environments of a vehicle. Subway systems may be influenced by PM from ambient air that enters the platforms through the ventilation system and from particles carried by passengers from the external environment. However, compared with a bus, a subway is far away from the road and hence relatively less affected by the external environment.

Inhaled dose

On the same route, a CNG bus, a PE bus and a subway were ridden to the same destination from the same starting point. The full travel time for each commute was recorded, and the PM exposure of each commute was monitored in real time. In addition, changes in the heart rate of volunteers under different modes of transportation were monitored. The inhaled dose was calculated using the dose model proposed by Campos et al. [19], and the total VE from 16 different volunteers (some male and some female) was calculated, producing a wide range of values reflecting the health of each individual. Finally, only the average was used in the analysis, for the purpose of obtaining reliable estimates to compare inhaled doses for different modes of transportation. Table 5 shows the inhaled doses for CNG, PE and subway commuters at different stages of commuting, throughout the commute process.

Table 5.

Inhaled doses for CNG, PE and subway commuters at different stages of commuting and throughout the commute process

	Section	Duration (min)	Time ratio (%)	VE (Lmin−1)	PM10 (μg)	PM2.5 (μg)	PM1 (μg)
CNG bus	Walking	5.4	15	26.58	21.44	10.0	7.7
	Bus stop	4	11	16.04	11.04	4.65	1.4
	In-vehicle	27	74	14.59	51.2	22.85	15.48
	Whole-trip	36.4		19.07	83.68	37.5	24.58
PE bus	Walking	5.4	15	26.58	21.44	10.0	7.7
	Bus stop	5.0	14	16.04	13.79	5.8	1.75
	In-vehicle	25.0	71	13.87	8.43	5.79	5.2
	Whole-trip	35.4		18.83	43.66	21.59	14.65
subway	Walking	5.0	23	26.58	19.86	9.28	7.14
	underground	4.5	20	23.45	9.18	2.95	1.58
	platform	1.5	7	14.22	2.18	1.15	0.84
	In-carriage	11.0	50	13.87	10.37	7.02	5.43
	Whole-trip	22.0		19.53	41.5	20.4	14.99

Inhalation dose varied greatly for different transportation modes and at different stages of travel. The highest whole-trip inhaled dose was obtained for the CNG mode, followed by the PE and subway modes. In the CNG commuting process, the proportion of inhaled dose in the carriage was the largest, and the whole-trip exposure ratios for PM10, PM2.5 and PM1 were 0.61, 0.61 and 0.63, respectively. During the PE commute, however, the highest inhaled dose appeared in the walking stage, while the whole-trip exposure ratios of PM10, PM2.5 and PM1 accounted for 0.49, 0.46 and 0.53, respectively. The highest inhaled dose for subway commuting also occurred during the walking stage, with PM10, PM2.5 and PM1 accounting for 0.48, 0.45 and 0.48, respectively, of the whole-trip exposure.

To give a more intuitive comparison of the PM exposure and inhaled doses for CNG, PE and subway commuters, we considered the three commuting processes as a whole and calculated the proportion of each commute separately. Figure 11 reflects the ratio of whole-trip exposure and inhaled dose for CNG, PE and subway commuters. The ratio of inhaled dose in our study was calculated using Eq. (2). Among the three commuting processes, the average exposure of the PE commuters was the lowest, and the effective ventilation conditions in the PE bus contributed to a reduction of the whole-trip exposure. However, when VE and commute time were considered, subway commuters had the lowest inhaled dose ratio, and inhaled PM concentrations were 45% and 2% lower than for CNG and PE commuters, respectively.

Fig. 11.

Ratio of whole-trip PM exposure and inhaled dose for CNG, PE and subway commuters

Inhaled dose ratio=$\frac{\mathit{VE}•\mathit{\text{Time}}•\mathit{\text{Exposure concentration}}/1000}{\mathit{\text{transit}}\mathit{bus}+\mathit{\text{Subway}}}$ (2).

Conclusions

Commuters’ PM exposures are significantly influenced by their choice of commuting modes. Over the same route, the average exposure concentration of bus commuters is higher than that of subway commuters. For example, on Route 1, the average PM10, PM2.5, and PM1 exposure concentrations of bus commuters were 71.6%, 19%, and 10.4% higher, respectively, than those of subway commuters. Compared with CNG buses, taking a PE bus can reduce a passenger’s commuting exposure risk by 4.3 times in summer. Seat location, vehicle type, ventilation mode, route characteristics, and the use of screen doors all have an impact on commuter exposure. For bus mode, the priority seat at the front door is a good choice for respiratory patients, the elderly and children.

The segment exposure concentration ranking of bus commuters is Bus Stop > Walking > CNG Bus > PE Bus. Bus commuters are potentially exposed to the highest PM emissions of their entire trip while they are waiting at roadside bus platforms—due to emissions from vehicles passing by. Re-suspended dust on the road, containing particles from brake linings and steel, is another source of their PM exposures, especially PM 10 and PM 2.5 [31]. Concentrations of PM 10 and PM 2.5 reached 172μg/m³ and 72.4μg/m³, respectively, at the bus stop, when the buses were parked and just after they had left the stop.

For passengers commuting by subway, the segment exposure concentration ranking was platform > walking > underground > in-carriage. The exposure concentrations of PM10, PM2.5 and PM1 on the platform were 1.6, 1.2 and 1.1 times those in the carriage, respectively. However, due to the high frequency of subway departures, the waiting time on a platform is generally less than 1.5 min, and PM exposure at the platform makes up only 7% of the whole-trip PM exposure. The relatively large distance from road traffic emissions, and the use of screen doors, were together responsible for the lower exposure of subway commuters.

Regardless of whether a commuter is a bus or a subway traveler, exposure during the roadside walking period contributed significantly to the whole-trip exposure, for CNG, PE and subway commuters, accounting for 17%, 34.7% and 23%, respectively. Traffic-related emission sources, such as tailpipe emissions and re-suspended road dust, may explain the high exposures during walking.

The results of the study showed that commuters should consider the different particle exposure risks and their own physical condition, when choosing a mode of travel. Respiratory patients, the elderly and children should determine the environmental conditions of the day before going out, take protective measures, and travel as far as possible on a subway or a PE bus. The urban traffic management agency should ensure that the vehicle has an effective ventilation system, and the urban public transportation system should use clean energy as much as possible; these two considerations have important practical significance for reducing vehicle emissions and passenger exposure. In order to reduce the exposure of passengers at the stop, the bus operation schedules and stop locations should be designed so as to minimize passenger waiting time at stops, and bus progress inquiry software should be available to passengers, so that they can minimize their walking and waiting times.

Compliance with ethical standards

Conflict of interest

The study was carried according to a protocol approved by Chang’an University, the approval number is 20180517. The objectives of the study were explained to the study participants and all participants signed volunteer consent form.

Conflict of interest

The authors declare that they have no conflict of interest.