Driver exposure to particulate matter in Bangkok

Abstract

The aims of this study were to determine the particulate matter with aerodynamic diameters ≥2.5 μm (PM_2.5) and 2.5–10 μm (PM_10–2.5) exposure levels of drivers and to analyze the proportion of elemental carbon (EC) and organic carbon (OC) in PM_2.5 in Bangkok, Thailand. Four bus routes were selected. Measurements were conducted over 10 days in August (rainy season) 2008 and 8 days in January (dry season) 2009. The mean PM_2.5 exposure level of the Tuk-tuk drivers was 86 μg/m³ in August and 198 μg/m³ in January. The mean for the non-air-conditioned bus drivers was 63 μg/m³ in August and 125 μg/m³ in January. The PM_2.5 and PM_10–2.5 exposure levels of the drivers in January were approximately twice as high as those in August. The proportion of total carbon (TC) in PM_2.5 to the PM_2.5 level in August (0.97 ± 0.28 μg/m³) was higher than in January (0.65 ± 0.13 μg/m³). The proportion of OC in the TC of the PM_2.5 in August (0.51 ± 0.08 μg/m³) was similar to that in January (0.65 ± 0.07 μg/m³). The TC exposure by PM_2.5 in January (81 ± 30 μg/m³) remained higher than in August (56 ± 21 μg/m³). The mean level of OC in the PM_2.5 was 29 ± 13 μg/m³ in August and 50 ± 24 μg/m³ in January. In conclusion, the PM exposure level in Bangkok drivers was higher than that in the general environment, which was already high, and it varied with the seasons and vehicle type. This study also demonstrated that the major component of the PM was carbon, likely derived from vehicles.

Implications:

Exposure to fine particulate matter (PM_2.5) in public transportation vehicles could have harmful health effects on both drivers and commuters in Bangkok, Thailand, where air pollution from vehicle exhaust is a serious problem. Exposure to fine particulate matter within moving vehicles has not been sufficiently investigated. Thus, the authors measured the levels of PM_2.5 within various types of transportation vehicles in Bangkok. The results are the first to demonstrate that drivers and commuters in Bangkok are exposed to a high level of PM_2.5, which cannot be detected by current roadside monitoring, and suggest the need for further pollution control measures.

Introduction

Studies of airborne particulate matter (PM) and its impact on public health have demonstrated adverse health effects at exposure levels that are currently experienced by urban populations. (WHO, 2005). Many studies have demonstrated that fine particulate matter, with an aerodynamic diameter of less than 2.5 μm (PM_2.5), is more hazardous than PM₁₀ (i.e., fine particulate matter with an aerodynamic diameter of less than 10 μm) and PM_10–2.5 (i.e., fine particulate matter with an aerodynamic diameter of larger than 2.5 μm and less than 10 μm). (Barnett et al., 2005; Barnett et al., 2006; Bell et al., 2008; Burnett and Goldberg, 2003; Burnett et al., 2004; Dockery et al., 1989; Dockery et al., 1993; Dominici et al., 2006; Dominici et al., 2007; Franklin et al., 2007; Holloman et al., 2004; Host et al., 2008; Klemm and Mason 2003; Miller et al., 2007; Ostro et al., 2006; Peng et al., 2008; Schwart et al., 1996) The guidelines set forth by the World Health Organization (WHO) for PM_2.5 are 10 μg/m³ for the annual mean and μg/m³ for the 24-hr mean; the guidelines for PM₁₀ are 20 and 50 μg/m³, respectively (WHO, 2005). Several studies have evaluated the importance of chemical composition, especially carbonaceous fractions and organic carbon (OC), for adverse health effects (Pedersen et al., 2005; Schnelle-Kreis et al., 2009). OC includes polycyclic aromatic hydrocarbons, some of which are carcinogens (Kume et al., 2007; Pedersen et al., 2005; Schnelle-Kreis et al., 2009; US Environmental Protection Agency, 2010). The measurement of elemental carbon (EC) to identify the source of carbonaceous pollutants is also required (Bae et al., 2009; Mkoma et al., 2010).

Bangkok is a typical large Asian city with a population of approximately 9.1 million and heavy traffic volume. In 1998 and 1999, almost all PM₁₀ concentrations in the four highly polluted districts, which are located less than 4 km from the city center of Bangkok, exceeded the Thailand National Ambient Air Quality Standard of 120 μg/m³ for the daily average concentration (Jinsart et al., 2002; Ostro et al., 2009; Vichit-Vadakan 2008). Urban traffic air pollution has been linked to an increase in the prevalence of respiratory function impairment among traffic police in Bangkok and policemen's wives (Karita et al., 2001; Karita et al., 2004). Furthermore, higher prevalence of respiratory symptoms and impaired lung function has been observed among children in areas with high pollution compared with those in areas in Bangkok with lower pollution (Tamura et al., 2003; Langkulsen et al., 2006). Not only the traffic policemen, but also the general population in Bangkok may be exposed to high levels of air pollution, especially when they are moving along the roadside. Information relating to this is scarce. The measurement of PM exposure levels in drivers on the streets in Bangkok is important in determining countermeasures for preventing their high exposure to PM and other air pollutants.

The aims of this study were to determine the PM_2.5 and PM_10–2.5 exposure levels of drivers, and to analyze the proportion of EC and OC in PM_2.5. We also investigated the differences in PM_2.5, PM_10–2.5, and PM₁₀ exposure levels among drivers of different types of vehicles (air-conditioned bus, non-air-conditioned bus, Tuk-tuk [motorized three-wheelers], and taxis), and examined seasonal effects in Bangkok.

Subjects and Methods

Subjects

In Bangkok, we searched for bus routes where both air-conditioned buses and non-air-conditioned buses drove the same routes, and we selected four routes: No. 4, No. 12, No. 29, and No. 93 of the Bangkok Mass Transit Authority. For all transportation modes, the air samplings were conducted close to the drivers (behind the driver's seat in the bus and Tuk-tuk and next to the driver in the taxi). The pollutants were sampled at the same time and on the same routes for all four vehicle types.

Methods

Sampling PM

Figure 1 shows the four bus routes and the seven Pollution Control Department Stations. The starting point was a large bus station in Bangkok. The measurements were conducted for 10 days in August (rainy season) 2008 and 8 days in January (dry season) 2009, with five collections for each day (total, 90 samples). Each driver drove each route for an average 5.7 hr (range, 3.3–9.8). The time for one round trip was 1.4–6.6 hr. Each driver completed a roundtrip once in the morning and another in the afternoon.

Measurements of the drivers' PM_2.5 and PM_10–2.5 exposure levels were conducted using personal samplers (ATPS-20H, filterholder with cascade impactors; Sibata Scientific Technology Ltd., Tokyo, Japan), which were set near the drivers. The PM_2.5 and PM_10–2.5 levels outside the windows of the vehicles were also measured. The filter holder was connected to a portable pump (MP-Σ3; Sibata) providing a constant airflow of 1.5 L/min. PM, which was larger than PM₁₀, was originally collected using a metal impaction plate coated with silicon grease immediately downstream of the inlet. PM_10–2.5 that was not captured by the impaction plate was collected on a 10-mm Teflon filter, which was made of borosilicate glass microfibers reinforced with woven glass cloth and bonded with Polytetrafluoroethylene (PTFE) (TX40HI20; Pall Corp., Port Washington, NY), with a 50% cut point of 10 μm. Next, PM_2.5 was collected on a similar 19-mm Teflon filter with a 50% cut point of 2.5 μm. Two samplers were always used for collecting and measuring PM, OC, and EC concurrently.

Weighing PM and the analysis of OC and EC levels

The PM_10–2.5 and PM_2.5 mass concentrations were obtained by weighing the filters before and after the sampling, always after a storage period (24 hr) in a temperature- and humidity-controlled room (ambient temperature, 23 ± 0.2 °C; relative humidity, 50% ± 1%), using an ultra-microbalance with a sensitivity of 0.1 μg (UMX-2; Mettler-Toledo, Inc., Columbus, OH). The PM₁₀ mass concentration was obtained as the sum of PM_10–2.5 and PM_2.5 mass concentrations. PM concentration levels below the detection limit were recorded as 0.05 μg/m³, which was calculated as a tenth of the analytical limit, from the combination of the 0.1 μg weight detection limit and approximately 0.2 m³ of the lowest sampling air volumes.

The concentrations of OC and EC in the quartz fiber filter samples were measured using a thermal/optical carbon analyzer (DRI Model 2001 Carbon Analyzer; Desert Research Institute, Las Vegas, NV) (Chow et al., 1993). Punched filter samples (8 mm ϕ) were analyzed using the IMPROVE protocol (OC1 = 120 °C, OC2 = 250 °C, OC3 = 450 °C, OC4 = 550 °C [in a 100% He atmosphere], EC1 = 550 °C, EC2 = 700 °C, EC3 = 800 °C [in a 2% O₂/98% He atmosphere]) and the detection limits of the sample were 0.5 μg (Chow et al., 2001). Because the PM_10–2.5 masses from the bus and taxi samples were small, and most of the OC and EC concentrations in PM_10–2.5 were equivalent to the negative controls, we excluded the OC and EC in PM_10–2.5 from the analysis. For the PM_2.5 samples, a pyrolysis correction of OC was made based on the filter reflectance.

Statistical analyses

All data analyses were performed using the SPSS Statistics 18 (SPSS Japan Inc., Tokyo, Japan) package. The means were compared using the Mann-Whitney U test or Kruskal-Wallis rank-sum test. P values < 0.05 were deemed statistically significant.

Results

PM exposure levels of drivers in Bangkok

From the total of 90 samples collected, due to trouble with the sampler, seven PM_10–2.5 samples and an additional two samples with damaged filters were excluded. The mean, median, and range of the PM_2.5 exposure levels of air-conditioned bus, non-air-conditioned bus, Tuk-tuk, and taxi drivers in August 2008 and January 2009 are listed in Table 1. All but one of the PM_2.5 exposure levels in the four vehicle types were greater than 50 μg/m³. The mean PM_2.5 exposure level for the Tuk-tuk drivers was 86 μg/m³ in August and 198 μg/m³ in January. The mean for the non-air-conditioned bus drivers was 63 μg/m³ in August and 125 μg/m³ in January. Seventy percent of the PM₁₀ exposure levels of the four types of vehicle drivers in August 2008 (mean of 75 μg/m³; median of 64 μg/m³; range of 23–190 μg/m³) and all of the PM₁₀ exposure levels of the drivers in January 2009 (mean of 164 μg/m³; median of 150 μg/m³; range of 64–403 μg/m³) were greater than 50 μg/m³. Table 2 presents the mean, median, and range of the PM_10–2.5 exposure levels of the four types of drivers and four routes in August 2008 and January 2009. The PM_10–2.5 levels in January were significantly higher than those in August for each type of vehicle (P < 0.01). PM_10–2.5 levels in January were significantly higher than those in August along routes Nos. 12, 29, and 93 (P < 0.01). The proportion of PM_2.5 to PM₁₀ was 76% in August 2008 and 80% in January 2009.

Table 1. Exposure levels to PM_2.5 * (μg/m³) of four driver types and outside the taxi (a) and four routes (b) in August 2008 and January 2009

	August 2008				January 2009
	Mean	Median	Range	n	Mean	Median	Range	n
(a)
Air-con bus	36	32	23–68	7	84	90	54–113	8
Non-air-con bus	63	58	39–105	7	125	123	104–143	8
Tuk-tuk	86	80	50–131	9	198	179	150–336	8
Taxi (inside)	52	52	34–97	10	110	110	78–155	8
Taxi (outside)	63	55	48–116	8	134	127	99–196	8
(b)
Route No. 4	64	51	31–131	11	108	104	59–201	8
Route No. 12	58	58	23–84	6	153	132	93–336	8
Route No. 29	64	70	32–104	7	129	123	54–222	8
Route No. 93	54	52	27–99	9	127	120	89–195	8
Notes: *PM2.5 levels in January were significantly higher than those in August on each of the driving routes by Mann-Whitney U test; (a) P< 0.01, (b) P < 0.05.

Table 2. Exposure levels to PM_10–2.5 * (μg/m³) of four driver types and outside the taxi (a) and four routes (b) in August 2008 and January 2009

	August 2008				January 2009
	Mean	Median	Range ^#	n	Mean	Median	Range	n
(a)
Air-con bus	6.9	0.05	0.05–23	7	26.3	26	5–50	8
Non-air-con bus	29.4	21	16–60	7	56.7	61	19–73	8
Tuk-tuk	25.1	23	0.05–60	9	46.9	47	23–68	8
Taxi (inside)	2.5	0.05	0.05–15	10	12.6	12	6–20	8
Taxi (outside)	22.1	18	6–50	8	44.5	43	20–77	8
(b)
Route No. 4	21.9	18	0.05–60	11	18.4	15	5–40	8
Route No. 12	7.2	3	0.05–21	6	40.8	40	14–73	8
Route No. 29	9.7	14	0.05–24	7	37.9	35	8–63	8
Route No. 93	16.0	10	0.05–47	9	43.8	47	10–72	8
Notes: *PM10–2.5 levels in January were significantly higher than those in August driving on routes Nos. 12, 29, or 93 by Mann-Whitney U test; (a) P < 0.01, (b) P < 0.01.
^#PM10–2.5 levels below the detection limit were set at 0.05.

The PM_2.5, PM_10–2.5, and PM₁₀ levels in January 2009 were significantly higher than those in August 2008 (P < 0.01). In August 2008, the PM_2.5, PM_10–2.5, and PM₁₀ levels on the No. 4 route were the highest among the four routes. In January 2009, the PM_2.5 and PM₁₀ levels on the No. 12 route were the highest among the four routes, and the PM_10–2.5 on the No. 93 route was the highest among the four routes.

The Tuk-tuk drivers were exposed to the highest levels of PM_2.5 and PM₁₀, and the non-air-conditioned bus drivers were exposed to the highest PM_10–2.5 levels. In terms of PM_2.5 exposure, the Tuk-tuk drivers showed significantly higher levels than the other drivers, in both August 2008 and January 2009 (P < 0.01). In terms of PM_10–2.5 exposure, the Tuk-tuk drivers and the non-air-conditioned bus drivers showed significantly higher levels than the air-conditioned bus drivers and taxi drivers, in both August 2008 and January 2009; however, the PM_10–2.5 exposure level of the Tuk-tuk drivers was not significantly higher than that of the non-air-conditioned bus drivers in August 2008 or January 2009 (P < 0.05). In terms of PM₁₀ exposure, the Tuk-tuk drivers and the non-air-conditioned bus drivers showed significantly higher levels than the air-conditioned bus drivers and the taxi drivers, in both August 2008 and January 2009 (P < 0.01). The PM₁₀ exposure level of the Tuk-tuk drivers was higher than that of the non-air-conditioned bus drivers in January 2009 (P < 0.01), but not in August. Table 3 lists the proportion of PM_2.5 to PM₁₀ for the four types of drivers and four routes in August 2008 and January 2009. The majority of the particulates collected were fine ones; on average, PM_2.5 comprised more than 70% of PM₁₀.

Table 3. The proportions of PM_2.5 levels to PM₁₀ levels of four driver types and outside the taxi (a) and on four routes (b) in August 2008 and January 2009

	August 2008				January 2009
	Mean	Median	Range ^#	n	Mean	Median	Range	n
(a)
Air-con bus	0.75	0.75	0.73–0.75	3	0.78	0.77	0.65–0.92	8
Non-air-con bus	0.70	0.73	0.52–0.81	7	0.70	0.67	0.65–0.85	8
Tuk-tuk	0.76	0.76	0.60–0.88	8	0.81	0.79	0.76–0.87	8
Taxi (inside)	0.91	0.91	0.87–0.94	3	0.90	0.90	0.85–0.93	8
Taxi (outside)	0.76	0.77	0.65–0.92	5	0.76	0.77	0.64–0.89	8
(b)
Route No. 4	0.72	0.72	0.60–0.87	8	0.86	0.86	0.72–0.93	8
Route No. 12	0.82	0.82	0.74–0.91	3	0.79	0.80	0.66–0.90	8
Route No. 29	0.82	0.73	0.75–0.88	7	0.77	0.77	0.65–0.93	8
Route No. 93	0.73	0.73	0.52–0.94	6	0.75	0.76	0.65–0.90	8
Note: ^#PM10–2.5 levels below the detection limit were excluded.

The proportion of PM_10-2.5 exposure in the non-air-conditioned bus drivers and the Tuk-tuk drivers compared with that of the taxi drivers was 11.5 and 9.8 in August 2008 and 4.5 and 3.7 in January 2009, respectively. Furthermore, the proportion of PM_2.5 exposure of the non-air-conditioned bus drivers and the Tuk-tuk drivers compared with that of the taxi drivers was 1.1 and 1.6 (August 2008) and 1.1 and 1.8 (January 2009), respectively. The PM_10–2.5, PM_2.5, and PM₁₀ levels outside taxis were significantly higher than those inside taxis, both in August 2008 and January 2009.

OC, EC, and total carbon (TC) concentration

Table 4 lists the TC levels in the PM_2.5 and three relevant proportions (TC to PM, OC to PM, and OC to TC) in August 2008 and January 2009. All of the exposure levels to TC and OC/TC differed significantly among the four driver types according to the Kruskal-Wallis rank-sum test (P < 0.01). In the PM_2.5 samples, the mean OC levels were 38-74 μg/m³ in August 2008 and 54–123 μg/m³ in January 2009. The mean proportions of the TC in the PM_2.5 were 0.89–1.07 and 0.57–0.71 in August 2008 and January 2009, respectively. The proportions of TC in PM_2.5 to the PM_2.5 levels in August (0.97 ± 0.28) were higher than those in January (0.65 ± 0.13), and those of the OC in TC in August (0.51 ± 0.08) were similar to those in January (0.65 ± 0.07), but the proportion of OC to TC remained higher in January compared to August.

Table 4. Total carbon (TC) exposure levels in the PM_2.5 and three relevant proportions (TC levels to PM_2.5 levels, organic carbon [OC] levles to PM_2.5 levels, and OC levels to TC levels [mean ± SD]) in August 2008 (a) and January 2009 (b)

	n	TC*	TC/PM2.5	OC/PM2.5	OC/TC*
(a) August 2008
Air-con bus	7	31 ± 7	0.95 ± 0.33	0.46 ± 0.16	0.50 ± 0.7
Non-air-con bus	7	57 ± 26	0.89 ± 0.20	0.42 ± 0.11	0.48 ± 0.04
Tuk-tuk	9	75 ± 14	0.93 ± 0.27	0.54 ± 0.17	0.59 ± 0.06
Taxi (inside)	10	52 ± 10	1.07 ± 0.31 ^#	0.50 ± 0.19	0.46 ± 0.08
Total	33	55 ± 22	0.97 ± 0.28	0.49 ± 0.16	0.51 ± 0.08
(b) January 2009
Air-con bus	8	54 ± 8	0.67 ± 0.13	0.44 ± 0.10	0.65 ± 0.04
Non-air-con bus	8	71 ± 14	0.57 ± 0.12	0.34 ± 0.08	0.60 ± 0.07
Tuk-tuk	8	123 ± 22	0.65 ± 0.15	0.47 ± 0.13	0.72 ± 0.06
Taxi (inside)	8	77 ± 18	0.71 ± 0.08	0.43 ± 0.07	0.62 ± 0.05
Total	32	81 ± 30	0.65 ± 0.13	0.42 ± 0.11	0.65 ± 0.07
Notes: *All of the exposure levles to TC and OC/TC differed significantly among the four driver types by Kruskal-Wallis rank-sum test (P < 0.01).
^# The reason for TC > PM2.5 is described in the Discussion.

The proportions of OC to TC in the PM_2.5 in August 2008 were significantly lower than those in January 2009. The proportions of OC in the PM_2.5 were 0.42–0.54 and 0.34–0.47 in August 2008 and January 2009, respectively. The proportions of OC to TC in the PM_2.5 were 0.46–0.59 and 0.60–0.72 in August 2008 and January 2009, respectively. Table 5 presents the TC, OC, and EC exposure levels in the PM_2.5 of drivers along the four bus routes in Bangkok during August 2008 and January 2009. The TC exposures by PM_2.5 in January (81 ± 30 μg/m³) were still higher than those in August (56 ± 21 μg/m³). The mean ± SD of OC in the PM_2.5 was 29 ± 13 μg/m³ in August and 50 ± 24 μg/m³ in January. The difference between the OC and EC levels was found to be significant by the Mann-Whitney U test.

Table 5. Exposure levels (mean ± SD) to total carbon (TC), organic carbon (OC), and elemental carbon (EC) in the PM_2.5 of the drivers on the four driving routes in Bangkok in August 2008 (a) and January 2009 (b)

	n	TC	OC	EC	P Value*
(a) August 2008
Route No. 4	12	58 ± 24	33 ± 14	25 ± 10	0.003
Route No. 12	7	56 ± 21	26 ± 14	29 ± 10	0.499
Route No. 29	8	64 ± 19	29 ± 11	35 ± 12	0.123
Route No. 93	10	48 ± 19	24 ± 14	24 ± 7	0.386
Total	37	56 ± 21	29 ± 13	28 ± 10	0.690
(b) January 2009
Route No. 4	8	82 ± 40	56 ± 33	26 ± 8	0.012
Route No. 12	8	91 ± 32	56 ± 25	36 ± 11	0.012
Route No. 29	8	79 ± 23	44 ± 15	35 ± 11	0.017
Route No. 93	8	73 ± 27	44 ± 21	29 ± 7	0.012
Total	32	81 ± 30	50 ± 24	31 ± 10	<0.001
Notes: *Difference between OC and EC levels by Mann-Whitney U test.

Discussion

In our study, drivers in Bangkok were exposed to high PM levels. The PM_2.5 exposure levels of the drivers of Tuk-tuk and non-air-conditioned buses, who were directly exposed to air pollution in the street, were similar to the levels (82–143 μg/m³) in highly polluted areas of Bangkok in 1998 and 1999 (Karita et al., 2001). These levels were above the WHO guideline for PM_2.5 (25 μg/m³), although the guideline is based on a 24-hr exposure, whereas our values were for approximately 6 hr. PM_2.5 exposure levels of the drivers of Tuk-tuk and non-air-conditioned buses were also higher than the environmental PM_2.5 levels of other cities, such as New York in 2001 and 2002 (0.47–53.7 μg/m³), Grenoble, France, in 1996 and 1997 (4.3–49.1 μg/m³), Shizuoka, Japan, in 2001 and 2002 (31.2 [19.3–41.7] μg/m³), and London (43.4 μg/m³ inside taxis), and similar to the levels in Beijing, China in 2006 (76 μ/m³ in the winter, 94 μ/m³ in the summer) (Hopke et al., 2003; Kaur et al., 2009; Kume et al., 2007; Wang et al., 2008; Zmirou et al., 2000). Additionally, 84% of the PM₁₀ exposure levels of the drivers were higher than the WHO guidelines, which allow a 24-hr exposure of 50 μg/m³. (WHO, 2005).

Our results revealed that the PM exposure to the drivers of Tuk-tuk and the non-air-conditioned bus was significantly higher than that of the air-conditioned bus and taxi drivers. The windows of the air-conditioned buses and the taxies were usually closed during driving. Tuk-tuk drivers and the non-air-conditioned bus drivers were directly exposed to the air in the street. The relative proportions of PM_2.5 exposure to the drivers of Tuk-tuks and non-air-conditioned buses compared with that of the taxi drivers were higher than those for PM_10–2.5 exposure. These results suggested that the filtering devices in the air-conditioned buses and the taxis may decrease the levels of PM_10-2.5 more effectively than the levels of PM_2.5, and that more efficient filtering devices may be required to decrease the PM_2.5 levels inside the vehicles.

In this study, the PM levels in January were significantly higher than those in August. The average PM_2.5 level in January 2009 was higher in air-conditioned buses by 48 μg/m³, in non-air-conditioned buses by 62 μg/m³, in Tuk-tuks by 112 μg/m³, inside taxis by 58 μg/m³, and outside taxis by 71 μg/m³, compared to the average PM_2.5 levels in August 2008 (Table 2). Likewise, the Pollution Control Department stations (PCD stations) in Bangkok (Figure 1) reported that the PM₁₀ level in January 2009 was higher than that in August 2008, suggesting a higher level of PM₁₀ during the dry season compared with the rainy season in Bangkok. The daily average PM₁₀ level measured at the PCD on the same days of our measurements at the general stations was 37.4 μg/m³ for August 2008 and 42.3 μg/m³ for January 2009. The PCD data for PM₁₀ from the roadside stations during the same periods showed 61.5 μg/m³ for August 2008 and 79.3 μg/m³ for January 2009. The average PM₁₀ levels measured at both of the general stations and the roadside stations were higher in January 2009 than in August 2008 (P < 0.05). The PM₁₀ levels at Din Daeng, one of the roadside PCD stations on 6 of 8 days in January 2009 were greater than 120 μg/m³, the National Standard of Thailand. The background levels in Bangkok were greater than 37.4 μg/m³; thus, our results suggest that the exposure level of drivers exceeded the WHO guidelines due to additional PM₁₀ exhausted from vehicles. According to the explanation of the WHO, the standard level of PM_2.5 is 50% of the standard level of PM₁₀. The proportion of PM_2.5 to PM₁₀ for the drivers in our study was greater than 70%. Our results suggest that the average PM_2.5 exposure levels of drivers in Bangkok are probably higher than the levels that the WHO guidelines allow.

Figure 1. Bangkok bus routes and Pollution Control Department stations. PCD = Pollution Control Department, Ministry of Natural Resources and Environment.

Other studies have shown that PM levels in Asian countries are strongly influenced by seasonal variations ( Begum et al., 2006; Gatari et al., 2006; Ho et al., 2006; Wang et al., 2008). During our survey in the rainy season (autumn), the temperature, dew-point temperature, and precipitation were 29.0 °C, 23.4 °C, and 8.5 mm, respectively, on average. Rainfall was observed 8 out of 10 days. During our survey in the dry season (winter), the average temperature was 25.3 °C, the average dew-point temperature was 16.5 °C, and the average precipitation was 0.0 mm. No precipitation was observed on these study days. We believe that rain or forest fires may influence the PM levels seasonally, but further studies are needed to confirm this assumption.

The level of OC in the PM samples and the proportion of TC to the PM_2.5 levels in Bangkok were higher than those of other cities (Japan Central Environment Council, 2009). It was unusual that the proportion of TC to PM_2.5 ranged from 0.89 to 1.07 in August 2008; this may be attributable to the evaporation of organic carbons and volatile inorganic ions due to high temperatures during sampling or storage at room temperature after sampling.

From another experiment, the decrease in weight during storage at room temperature for a month was 10% ± 11% (data not shown). Some volatile and semivolatile organic carbons may have evaporated. Thus, the proportion of TC to PM_2.5 levels may have been overestimated. The true PM_2.5 value may have been higher had we analyzed the filters immediately after they were collected. Another study also reported that the proportion of TC to PM_2.5 levels was much higher during the wet season than in the dry season, but the reason was unclear (Mkoma et al., 2010). Nevertheless, the proportions of TC levels to the PM_2.5 levels were higher than those of other cities (0.06–0.17 in Japan in 2008) (Japan Central Environment Council, 2009). PM_2.5 has been reported to be produced primarily by human activities, including driving cars and manufacturing (Fushimi et al., 2008; Japan Central Environment Council, 2009; WHO, 2005). To date, measurements of the exposure concentrations of EC and OC in PM_2.5 while driving vehicles have been scarce. Although the concentrations of EC and OC in PM_2.5 along the roadside in Bangkok were not measured simultaneously in our study, our results indicated that the concentrations OC in PM_2.5 to which drivers of vehicles in Bangkok were exposed were high compared to those in other cities.

Because the current study collected data in only two series, in August and January, the conclusions reached may require further confirmation by repeated measurements. Our results, however, seem to be typical in Bangkok, Thailand, and may be used to facilitate considerations of making changes to the traffic system in Bangkok and improving the work environment for drivers.

Conclusions

This study demonstrated that drivers in Bangkok were exposed to high concentrations of fine particulate matter (PM_2.5) as well as organic and elemental carbon, and that there were different PM exposure levels among drivers of different vehicle types. Also, the PM_2.5 exposure level of drivers in Bangkok varied by season. The carbonaceous component of PM_2.5 while driving vehicles in Bangkok was high and differed from that of environmental PM_2.5. Because the PM_2.5 in Bangkok consisted largely of OC, which usually contains various toxic substances, only weighing PM_2.5 may not be sufficient to evaluate the health effects of air pollution in Bangkok.

Acknowledgments

Partial funding of this study was received from The National Research University Project of CHE and the Ratchadaphiseksomphot Endowment Fund (CC307A). Mr. Chanin Kaewmanee is most appreciative of the post-graduate scholarship from the 90th Anniversary of Chulalongkorn University Fund.