Exposure to ultrafine particles in different transport modes in the city of Rome

Environmental Pollution 228 (2017) 201e210

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Exposure to ultrafine particles in different transport modes in the city of Rome* Mario Grana a, *, Nicola Toschi a, b, c, Laura Vicentini a, Antonio Pietroiusti a, Andrea Magrini a a b c

Department of Biomedicine and Prevention, University of Rome “Tor Vergata” e Via Montpellier 1, 00133 Rome, Italy Department of Radiology, “Athinoula A. Martinos” Center for Biomedical Imaging, Boston, MA, USA Harvard Medical School, Boston, MA, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 5 October 2016 Received in revised form 8 May 2017 Accepted 9 May 2017

There is evidence of adverse health impacts from human exposure to particulate air pollution, including increased rates of respiratory and cardiovascular illness, hospitalizations, and pre-mature mortality. Most recent hypotheses assign an important role to ultrafine particles (UFP) (<0.1 mm) and to associated transition metals (in particular Fe). In a large city like Rome, where many active people spend more than one hour per day in private or public transportation, it may be important to evaluate the level of exposure to harmful pollutants which occurs during urban travelling. In this context, the aim of this work was to examine the relative contribution of different transport modes to total daily exposure. We performed experimental measurements during both morning and evening traffic peak hours throughout the winter season (December 2013eMarch 2014), for a total of 98 trips. Our results suggest that the lowest UFP exposures are experienced by underground train commuters, with an average number concentration of 14 134 cm
3, and are largely a reflection of the routes being at greater distance from vehicular traffic. Motorcyclists experienced significantly higher average concentrations (73 168 cm
3) than all other exposure classes, and this is most likely a result of the presence of highconcentration and short-duration peaks which do not occur when the same routes are traveled by car. UFP concentrations in subway train environments were found to be comparable to urban background levels. Still, in underground trains we found the highest values of PM10 mass concentration with a maximum value of 422 mg/m3. PM10 concentration in trains was found to be four and two times higher than what was measured in car and motorbike trips, respectively. Transport mode contribution to total integrated UFP daily exposure was found to be 16.3%e20.9% while travelling by car, 28.7% for motorbike trips, and 8.7% for subway trips. Due to lower exposure times, commuting by car and motorbike is comparable to other daily activities in terms of exposure. Our data can provide relevant information for transport decision-making and increase environmental awareness in the hope that the information about inhaled pollutants can translate into a more rational approach to urban travelling. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Commuter Exposure Transport mode Ultrafine particles PM10 SEM

1. Introduction Current knowledge of the health effects associated with air pollution from airborne particles date back to approximately half of the twentieth century. Since then, many countries have adopted air quality standards aimed at protecting human health and the environment. However, despite substantial investments in

*

This paper has been recommended for acceptance by Eddy Y. Zeng. * Corresponding author. E-mail address: [email protected] (M. Grana).

http://dx.doi.org/10.1016/j.envpol.2017.05.032 0269-7491/© 2017 Elsevier Ltd. All rights reserved.

pollution control, air quality in urban areas of industrialized countries remains alarming. This serious health impact mainly affects developed countries, for which it has been estimated that environmental pollution from airborne particles, measured in terms of PM10, is responsible for 6% of total mortality, of which about half is attributed to vehicular traffic (Künzli et al., 2000). Several epidemiological studies, mostly performed on the general population, have highlighted a number of adverse health effects associated with exposure to fine airborne particles with fine (<2.5 mm) and ultrafine (<0.1 mm) size characteristics. Specifically, there is experimental and epidemiological evidence that finer

202

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

airborne particles are responsible for a variety of diseases, not only of the respiratory system, but also of other organs, which appear to be especially affected by the ultrafine particles size class (WHO, 2004). People spend a largely variable but significant part of their time commuting. A WHO report indicates an average of 1e1.5 h per day spent travelling (Kryzanowski et al., 2005), and since air pollution levels in traffic are generally high, these relatively short durations can contribute substantially to the daily personal exposure and related health effects (Kaur et al., 2007). For example, three personal exposure studies carried out in the Netherlands documented that time spent in traffic contributed significantly to 24 h average personal exposure to PM10, PM2.5 and ultrafine particles (Wichmann et al., 2005; Van Roosbroeck et al., 2008; Boogaard et al., 2009). Similar results were reported in studies conducted in Belgium (IntPanis et al., 2010) and in Sidney (Knibbs and de Dear, 2010). Miao et al. (2015) suggest that, independently of the mode of commuting, all types of commuting during rush hours increase exposure to air pollution. Also, two studies performed in London (Kaur et al., 2005; Briggs et al., 2008) show that average exposures while walking are considerably greater than in-car exposure, and that exposure to PM2.5, ultrafine particles and CO are affected by transport mode, route and timing. Also, recent studies show that incar exposure could significantly contribute to total integrated exposure estimates (Knibbs et al., 2011), and air pollution exposures in traffic are generally higher than and poorly correlated with concentrations measured simultaneously at fixed monitoring locations, even if these are located in major streets (Kaur et al., 2007). Characterization of commuters’ exposure cannot therefore rely on fixed site monitoring. To date, three studies assessing exposure within cars within Italian urban settings exist (Cattaneo et al., 2009; Geiss et al., 2010 and Cattani et al., 2013), and only one study assessed PM exposure within underground trains (Cartenì et al., 2015). This study was conducted in Rome, Italy's most populous city with 2 885 272 residents (year 2011) in 1285.3 km2, with the aim of evaluating and characterizing commuter exposure to PM10 and UFP while travelling through different transport modes. 2. Methods 2.1. Study design and routes All experiments were carried out during working days in the period December 2013eMarch 2014 (winter season). Four different transport modes were selected: car, underground train, motorcycle and bus. All routes have a similar origin and the same destination. We employed two starting points in close proximity to each other and located in the West of the city, while the destination was located in the East-South East (Tor Vergata University) (Fig. 1). Measurements were performed at the same time of day for all transport modes (8.30e10.30 and 16.30e18.30). In car trips we employed two main types of routes: a) routes through the center of the city and b) routes through the Great Ring Junction (GRA), a dual-carriageway, three-lane, toll-free highway. During car trips, air conditioning (AC) was switched on in auto mode without air recirculation (AR). Additionally, in order to confirm the filtering capabilities of car AC equipment, a limited number of trips were performed with AC switched on in auto mode with AR. Car windows were kept closed throughout all trips. The route traveled by motorcycle was roughly the same as the car route through the city center, and route length varied from 22.6 km to 31.1 km across the city center and 34.5 kme39.4 km along the GRA. The morning subway route was along “line A”, starting at “Cornelia” station and ending at “Anagnina” station (and vice versa in the afternoon). This

route is comparable to the trip on the surface through the center and has a length of 17.2 km. All subway measurements were carried out inside train carriages. The bus ride covered a suburban area ran from “Anagnina” subway station to Tor Vergata University (distance: 6.8 km, most of the route covered on a dedicated bus-lane). Participants spent an average of 48 ± 6 min (one-way) commuting from house to work or back. In order to account of day-to-day variability in outdoor particle number concentration (PNC), background measurements were carried out at the origin and destination point of each trip, distant from traffic sources, and averaged to obtain a single background value for each day. Also, a second branch of the study was aimed to estimate and describe the contribution of commuting exposure to total daily exposure. To this end, indoor measurements were made in two private homes in the West area of the city (in the kitchen while cooking and eating dinner, in the living room and in the bedroom) as well as in one office building in the University campus. The average sampling time was set to 30 min, except for cooking, in which case sampling time was 1e1.5 h. 2.2. Vehicles under study Cars and motorcycle employed in this study belonged to colleagues who agreed to take part in this study. This resulted in a set of vehicles of different size and manufacturing year (1998e2010). 2.3. Instrumentation A P-Trak ultrafine particle counter (TSI Model 8525, TSI Incorporated, Shoreview, MN, USA), was used to measure total particle number concentration in the range from 20 nm to 1000 nm at a time resolution of 1 s. This unit is a condensation particle counter which employs isopropyl alcohol and is capable of detecting particle concentrations up to 5  105 particles per cm3 (cm
3). The PTrak has the advantage of being hand-held (small dimensions and weight, battery powered) and of allowing recordings with high time resolution. On the other hand the lower limit on particle size range (20 nm) leads to underestimation of the ultrafine component below this limit, which is especially relevant when detecting freshly emitted nucleation mode particles (Zhu et al., 2006). Still, Matson et al. (2004) found that the P-Trak is reliable and yields results comparable to another hand held instrument, TSI 3007, with a lower size limit (10 nm as compared to 20 nm). Additionally, the size ranges of particles coming from diesel exhaust and from petrol engines are 20e130 nm and 20e60 nm respectively (Morawska et al., 2008) and therefore fall within the P-Trak measurement range. The P-Trak was zero-checked before each use by means of an HEPA zero filter assembly, and the recording time interval was set to 1 s. For large particles (>300 nm) a TSI AeroTrak 9306 optical particle counter was used. Measurements with this instrument were performed simultaneously with a part of the measurements made with the P-Trak (6 car trips, 2 motorbike trips, 2 subway and 2 bus rides). In car cabins, CO and CO2 concentrations were measured in both air ventilation modes (3 car trips with AR and 10 car trips without AR), by means of TSI IAQCalc Indoor Air Quality Meter mod. 7545, which has a resolution of 0.1 ppm and 1 ppm for CO and for CO2, respectively. In cars, the devices were placed on the passenger seat, with inlets facing forwards, next to the breathing zone of the driver. On the train, motorbike and bus, units were carried in an especially adapted back-pack, with inlet facing upwards, and the P-Trak was held horizontally to avoid tilting. PM10 mass concentration (thoracic fraction) was measured through a personal sampler (SKC AirChek XR5000) connected to a

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

203

Fig. 1. Sampling route map (blue ¼ metro, green ¼ car and motorbike through the city center, light blue ¼ bus, red/orange ¼ car through the GRA). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

BGI GK2.69 cyclone set at a flow rate of 1.6 L/minute. This cyclone is one of the few that can perform personal sampling of thoracic fraction of particles. While high flow ambient sampling of PM10 can be more accurate, the only way to measure personal exposure to thoracic fraction is through the use of personal samplers (which are portable devices and consequently are based on low flow samplers) connected to cyclones. Particles were collected on 37 mm silver membrane filters (porosity 0.8 mm). These filters are conductive and non-sorptive, hence weight instability is minimized and not affected by changes in relative humidity and electrostatic charges. Other weight instability factors were taken into account through the use of at least three blank filters in each weighing session. Filters were weighted with a Sartorius ME5 microbalance located in a environmentally-controlled chamber with a temperature of 20 ± 2  C and relative humidity of 50± 5% and a filter conditioning time of two hours before weighing. The balance has a weighing resolution of 1 mg, the uncertainty of weighing and the LOD are 1.5 mg and 4.6 mg, determined according to ISO 15767:2009 (ISO, 2009).

The inhalable fraction of particles was also collected on polycarbonate filter for qualitative analysis. Size, morphology and elemental composition of particles were determined by scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy. The observations and analysis were performed with a Zeiss LEO 420 scanning electron microscope operated at 20 kV acceleration voltage. Prior to analysis samples were gold coated by means of a Agar Auto Sputter Coater. The energy-dispersive X-ray detector used was an Oxford Inca Energy 250 (19 mm working distance). A summary of the type of pollutants measured along with the number of measurements for each transport mode or location are shown in Appendix A (Supplementary material - Table S1). 2.4. Statistical analysis In order to explore the independent effects of mode of transportation, time of day, and route on measured concentrations, we employed a general linear model (GLM) with concentrations as the dependent variable and transportation, time of day and route as

204

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

independent factors. Additionally, in order to account for possible inter-day variability, background concentrations were included as nuisance variables and the date of measurement was included as a random factor in the model. Wherever a significant effect of a factor was detected, pairwise comparisons between factor levels were performed post-hoc using Bonferroni correction for multiple comparison. p < 0.05 was considered statistically significant. Also, average time spent daily commuting in Rome was estimated to be 1.5 h, while median total time spent indoor at home and at office for Italian people was estimated by Schweizer et al. (2007) and set as 13 h and 8 h respectively. For each transport mode, the relative contribution to total daily exposure was calculated by multiplying median time spent for each activity to the related PNC concentration and dividing by the sum of all the activity doses. All analyses were performed using R software. 3. Results The GLM yielded statistically significant effects of all factors, and post-hoc comparisons revealed statistically significant differences between all pairs of levels of any single factor (p < 0.001, Bonferroni corrected). 3.1. Effects of mode of transport and route

Fig. 2. Distribution of ultrafine particle number concentration by transport mode. Solid lines indicate the median, boxes indicate the interquartile range, whiskers maximum value is Q3þ1.5*IQR, dots indicate outliers (y axis in log scale). In post hoc pairwise comparisons, all differences between modes of transport were statistically significant (p > 0.001 corrected).

The average UFP concentration (analyzed after aggregation in 1 min intervals) relative to different transport modes as well as indoor microenvironments are presented in Table 1. Fig. 2 provides a visual summary of measured data according to mode of transportation. The PNC highest median was found while commuting in motorbike through the city center (60 238 cm
3); car cabins PNC in AC mode without recirculation, resulted in median values of 29 173 and 39 518 cm
3 respectively for city center and GRA trips. These data sets showed remarkable variability: as can be seen in Fig. 2 interquartile range was 36 861e96 876 cm
3 (motorbike), 24 165e64 113 cm
3 (car GRA w/o AR), 21 440e42 328 cm
3 (car city center w/o AR). Car cabin PNC was strongly affected by route

type. Lower values (median ¼ 25 075 cm
3) were obtained while riding by bus in a suburban area. Subway trains PNC values are comparable to levels recorded in urban background environment and are the lowest among the commute modes, in accordance with Moreno et al. (2015a). Also, in accordance with previous literature (Li et al., 2017), in the case of PM2.5 and PM1 the motorcyclists inhaled significant high pollutants during commuting presumably as a result of high-concentration exposure and short-duration peaks. Table 2 shows that when travelling by car through the city center (with AR mode) the average CO2 concentration (2413 ppm)

Table 1 Ultrafine particle number concentration (cm
3).

Mode of transport Car (city center), w/o AR Car (city center), with AR Car (GRA), w/o AR Car (GRA), with AR Motorbike Subway Bus Outdoor background Timings Morning Car (city center) Car (GRA) Motorbike Subway Bus Afternoon Car (city center) Car (GRA) Motorbike Subway Bus Microenvironments Office Kitchen Living room Bedroom

N (average time per trip)

Mean (sd)

Median

Min

Max

23 (520 ) 3 (540 ) 29 (430 ) 2 (390 ) 20 (460 ) 24 (380 ) 23 (170 ) 152

35 768 (28 846) 5249 (3090) 48 428 (38 315) 11 958 (6957) 73 168 (50 279) 14 134 (6355) 29 299 (18 078) 13 466 (7055)

29 173 4213 39 518 8686 60 238 14 287 25 075 11 890

4963 2921 2335 4658 5921 4784 2963 1408

422 283 17 873 384 950 31 280 299 867 47 675 107 358 51 915

13 12 10 11 11

41 459 (33 474) 55 850 (50 001) 88 285 (55 185) 16 863 (5620) 36 506 (20 735)

32 44 74 17 32

828 653 527 107 919

9111 2400 5921 4784 2963

422 283 381 600 299 867 30 580 107 358

10 17 10 13 12

26 499 (14 987) 43 443 (26 181) 59 491 (40 873) 11 725 (5985) 20 951 (10 136)

22 812 38 494 47 382 9601 18 918

4963 2566 6801 5124 8808

126 893 221 118 228 667 47 675 63 300

18 11 5 7

5676 (2862) 38 996 (28 808) 12 155 (2190) 8243 (1850)

5696 31 410 12 348 8833

1559 8400 9412 5241

15 467 199 583 15 880 10 010

M. Grana et al. / Environmental Pollution 228 (2017) 201e210 Table 2 Summary statistics for average carbon monoxide and carbon dioxide concentration (ppm). N

Carbon monoxide

Carbon dioxide

Mean ± sd (median) Min-Max Mode of transport Car (city center), w/o AR

7

Car (city center), with AR

3

Car (GRA), w/o AR

4

Car (GRA), with AR

2

Outdoor

13

1.6 ± 1.2 (1.5) <0.1e5.6 0.6 ± 0.3 (0.8) <0.1e1.0 0.7 ± 0.9 (0.6) <0.1e7.1 0.5 ± 0.4 (0.6) <0.1e1.4 <0.1 ± 0.2 (<0.1) <0.1e1.0

683 ± 92 (675) 402e1070 2413 ± 624 (2568) 638-3020 654 ± 96 (622) 514e933 1335 ± 221 (1369) 592-1801 438 ± 35 (438) 370e523

205

is higher than what is measured along the route through the GRA (1335 ppm). In this context, Knibbs et al. (2009) showed that air exchange increased with vehicle speed. Fig. 3 shows UFP and CO2 concentrations in car cabins (with only the driver present) while travelling at different speeds with AR on. Increasing the speed determines a reduction in CO2 rising and a concomitant peaking of UFP concentration, possibly due to a higher rate of air leakage from outdoor. There is a direct correlation between CO concentrations and UFP counts: peaks in CO and UFP were often experienced almost simultaneously by car drivers (see Fig. S1 in Supplementary material). Table 3 compares the PM10 concentrations found in car and motorbike trips with those found in the underground trains. Particle concentration in trains is four and two times higher (respectively), with a maximum value of 422 mg/m3. Previous studies have reported high concentrations of PM in underground trains/platforms and have highlighted resulting health concerns (Cartenì

Fig. 3. Example of UFP and CO2 real time concentrations during car trips in AR mode throughout the GRA (data collected on January the 27th and 28th 2014).

206

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

Table 3 Summary statistics for average PM10 personal exposure (mg/m3).

Mode of transport Car Motorbike Subway

N

Mean (sd)

Median

Min

Max

14 7 6

61 (17) 138 (33) 268 (79)

64 130 251

38 104 207

93 197 422

et al., 2015; Cusack et al., 2015; Martins et al., 2015; Moreno et al., 2014). UFP concentrations are uniform, comparable to outdoor background levels and do not show peaking phenomena recorded in surface transport modes, which means these are not influenced by train movements and air exchange between tunnels and trains. This is also confirmed by particle number concentrations in the range 0.3e20 mm: Fig. 4 shows that in motorbike and car finer particles prevail (range 0.3e0.5 mm) as opposed to measurements in subway trains, while conversely in train carriages particles above 1 mm diameter (coarser fraction) prevail. This could possibly mean that particles generated in the subway system are in the coarse

fraction range, while finer particles present underground are conveyed from outdoor through the tunnel ventilation systems and hence outdoor morning/afternoon variability in UFP concentration is reflected also in the subway system. Mechanical abrasion and wear and tear of the brake system and friction between rail and wheels can also generate and resuspend particulate matter in electricity-powered subway/metro systems; these are more likely to elevate levels of particle mass rather than UFP number count (Nieuwenhuijsen et al., 2007). 3.2. Estimation of exposure to peaks The percentages of measurements above an absolute threshold value (chosen as 100 000 cm
3) were: motorbike ¼ 22.6%, car (GRA) ¼ 7.2%, car (city center) ¼ 2.3%, subway ¼ 0%. Additionally, the percentage of measurements which exceeded the threshold of two standard deviation above the mean value for any given day gave the following results: motorbike ¼ 7.0%, car (city center) ¼ 5.1%, car (GRA) ¼ 3.2%, subway ¼ 4.3%. In summary, as expected motorcyclists are exposed to absolute peaks more often that all

Fig. 4. Particle number concentration in different size range (mean þ - SEM over the number of trips after computing trip median). (Car: 6 trips through GRA; Motorbike: 2 trips through city center; Subway: 2 trips; Bus: 2 tests; Background: 9 tests).

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

other routes/modes of transport, both in absolute and in relative terms. The slightly higher relative incidence of peaks in the subway may have little significance in view of the overall low exposures/ concentrations in this mode of transport. 3.3. Effects of time of day Exposure to UFP shows differences between morning and afternoon trips (the morning median concentrations were between 14% and 44% higher in comparison to the afternoon concentration levels). As reported by Aalto et al. (2005), in Rome, the winters are more polluted compared with summers by as much as a factor of 4e5 and during the weekdays the maximum concentration is detected during the morning hours between 7 a.m. and 10 a.m. This may be due to morning rush hour nucleation events related to higher emissions of condensable substances and to lower temperatures, especially during winter season. 3.4. Effects of air recirculation (AR) When air was recirculated in cars, UFP concentrations were reduced in comparison to cars without AR (86% reduction through the city center and 75% reduction through GRA) and became similar to concentrations typical of an office building. While this indicates the shielding efficacy of AR, the accompanying increase in CO2 concentration limits the use of this mode to short driving periods (a few minutes). In this context, it is important to note that the ASHRAE standard 62.1e2013 (ASHRAE, 2013) suggests that in confined spaces CO2 concentration should be maintained to levels no greater than about 700 ppm above outdoor air concentration. In view of an outdoor average CO2 concentration of 438 ppm, in the case of AR ventilation mode, the limit is exceeded after 4e5 min of driving (Fig. 3). While this trend was evident in our data, we did not include this factor (AR on vs AR off) in our GLM due to the limited number of data collected (3 trips) using AR. 3.5. SEM and energy-dispersive X-ray spectroscopy analysis of airborne particles In terms of particle number, submicron organic/soot particles were the most abundant type. SEM analysis of samples of airborne particles taken from subway trains showed that PM is represented mostly by the coarse fraction, and it is Fe-rich (Cusack et al., 2015; Moreno et al., 2015b; Ripanucci et al., 2006); the most common morphology occurs in the form of irregular, leaves and splinters (Fig. 5-a) of at most a few microns in size. Samples included mineral particles, such as silicates (Fig. 5-b), calcium carbonate and sulphates (from soil and building materials); we also observed diesel soot particles or aggregates coming from outside air. Fe-rich particles from the underground railway systems could be related to mechanical abrasion and wear at the rail-wheel-brake interfaces (Jung et al., 2010). Airborne samples taken from a motorbike trip show that the traffic emission consists mostly of soot, which occurs as agglomerates (Fig. 5-c) or single spherical carbon particles. Other particles in the motorbike sample are characterized by the presence of elements probably related to brake and tire abrasion: Fe, Cu, Ba (Fig. 5-d). There are also Fe-Cr particles (Fig. 5-e) with irregular (non-spherical) shapes, probably coming from wearing and corrosion of vehicle parts. Emissions from motorized vehicles are the primary source of metallic particles to the atmosphere, probably originating from engine combustion material, brake abrasion (Gietl et al., 2010; Thorpe and Harrison, 2008), and tire wear (Sanderson et al., 2014). Among metals we also identified Zn, which in addition to coming from brakes abrasion, and diesel engine emission sources, is also used as activator of vulcanisation in the tire industry.

207

Other elements recognized in the samples, from the resuspension component include Si, Ca, Mg, Al. In summary, qualitative SEM analysis showed that, in the underground transport mode, the coarse fraction (with high Fe content) is most abundant, while at the surface the highest proportion of particles is constituted by submicrometric particles of organic nature and carbonaceous soot agglomerates formed directly by combustion. The presence of heavy metals coming from different parts of vehicles (tyres, brakes) is also non-negligible. While the toxicological relevance of metallic aerosols is not known in detail, it is becoming increasingly clear that even small amounts of metal emitted may affect human health (Fairbrother et al., 2007). 3.6. Exposure distribution Table 4 shows the influence of measured UFP concentrations on total daily exposure. Car travel along GRA and city center between home and workplace (1.5 h on average) contributed respectively 20.9% and 16.3% to total daily UFP exposure; motorbike trips contributed for 28.7% on total exposure. The lowest contribution comes from subway trips with 8.7%. 4. Discussion and conclusions In this paper, we examined the relative contribution of different transport modes, routes, and time of day to total daily exposure. The highest UFP concentrations were found while travelling in surface transport modes (motorbike, car, bus) along high density traffic routes. Except for subway trips, these values are greater than those measured during other daily activities. Higher concentrations were detected in motorbike and car trips (as compared to other modes of transport). These findings also exhibited the spatial and temporal variations of UFP concentrations typical of the roadway environment (Morawska et al., 2008). Among others, train commuters are less exposed to UFPs because of the distance from vehicular traffic sources. On the other hand, motorcyclists are most exposed, with UFP concentrations also showing high frequency peak events. The main factors affecting differences in UFP concentrations measured in this paper can be hypothesized to be related to a) varying proximity to sources: line A metro system consists of shallow underground stations, on average
10 m deep with respect to the road surface, and deep stations, on average between
19 m and
54 m with respect to the road surface. These greater distances from the traffic sources determines lower UFP concentrations as compared to other transport modes (14 134 cm
3 in subway carriages as compared to 35 768 cm
3 in car cabins following the same route on the surface), b) shielding effects of vehicle cabins, as highlighted by the difference in concentrations found between car cabins and motorcycle travelling through the city center (35 768 cm
3 in cars as compared to 73 168 cm
3 in motorbikes), it therefore appears that vehicle barrier has an important role in shielding commuters from the highest UFP concentrations, reducing mean and maximum values, c) traffic density: the higher traffic density is found along the GRA, where car journeys show an average value of 48 428 cm
3 compared with a value of 35 768 cm
3 through the city center, d) average speed: higher speed imply a higher rate of air intake from outdoor and thus a parallel increase of UFP concentration (Fig. 3), e) cabin ventilation mode: AR mode drastically reduces UFP concentrations. In this context, a way to reduce UFP exposure for car drivers could be a better cabin isolation and a greater control over highemitting vehicles. Also, subway commuters are most exposed to coarser fraction of particles which is Fe-rich. The priority in subway air quality should therefore be to reduce high mass concentrations

208

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

Fig. 5. SEM images; a) typical morphological aspect of FePM in subway samples; b) geological origin particles; c) C-rich agglomerate (soot) in motorbike airborne sample; d) particle containing Fe-Cu-Ba; e) Fe-Cr particle.

Table 4 Estimated relative contributions of each considered microenvironment to UFPs exposure. Microenvironment

Median time spent (h/d)

Commuting Office Home - Bedroom Home e Living room Home e Kitchen Outdoor

1.5 8 8 3.5 1.5 1.5

Median PNC (cm
3)

Relative contribution to total integrated exposure (%)

Car GRA

Car city center

Motorbike

Subway

Car GRA

Car city center

Motorbike

Subway

39 518

29 173

60 238

14 287

20.9 16.1 24.9 15.2 16.6 6.3

16.3 17.0 26.4 16.1 17.6 6.7

28.7 14.5 22.5 13.7 15.0 5.7

8.7 18.5 28.7 17.6 19.2 7.3

5 696 8 833 12 348 31 410 11 890

of aerosol. This study shows that commuting in the subway system increase the PM10 24-h average exposure posing a potential health risk to commuters. The highest CO concentrations recorded in car cabins while commuting through the GRA and through the city center (without AR) show maximum values of 7.1 ppm and 5.6 ppm respectively. These values are below the 8-h (10 ppm) guideline value proposed by World Health Organization (WHO, 2000). An important aspect which should be taken into account in future studies is the influence of meteorological factors, like wind speed, which potentially affect UFP concentration as well as its temporal variability. Additionally, particle mapping could be

coupled with the GPS positioning in order to analyze pollutant concentrations also in relationship to the intra-trip distribution of time (e.g. isolating the time spent in high pollutant concentrations areas such as crossroads and traffic lights). Transport is a ubiquitous component of life, and initial evidence suggests that UFP exposures incurred during this time can contribute substantially to daily exposure and be associated with adverse health effects in susceptible as well as otherwise healthy individuals. Further research to better define this link is therefore well-justified, and will be of considerable benefit to urban planning, policy development and public health. It is important to transform urban cities into more

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

sustainable and livable environments. This could be achieved by providing incentives for people to take public transportation, e.g. through congestion charging schemes for central districts. It is also important to change people's perception of public transport (hence leading to changes in behaviour) by informing both the public and policymakers about the environment and health implications of transport choices. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.envpol.2017.05.032. References €meri, K., Paatero, P., Kulmala, M., Bellander, T., Berglind, N., Bouso, L., Aalto, P., Ha ~ o-Vinyals, G., Sunyer, J., Cattani, G., Marconi, A., Cyrys, J., von Klot, S., Castan €vall, B., Forastiere, F., Peters, A., Zetzsche, K., Lanki, T., Pekkanen, J., Nyberg, F., Sjo 2005. Aerosol particle number concentration measurements in five european cities using TSI-3022 condensation particle counter over a three-year period during health effects of air pollution on susceptible subpopulations. J. Air& Waste Manag. Association55 1064e1076. http://dx.doi.org/10.1080/ 10473289.2005.10464702. http://dx.doi.org/10.1080/10473289.2005.10464702. ANSI/ASHRAE Standard 62.1-2013. Ventilation for acceptable indoor air quality. ASHRAE, Atlanta GA. https://www.ashrae.org/resourcesepublications/ bookstore/standards-62-1e62-2. Boogaard, H., Borgman, F., Kamminga, J., Hoek, G., 2009. Exposure to ultrafine and fine particles and noise during cycling and driving in 11 Dutch cities. Atmos. Environ. 43, 4234e4242. http://dx.doi.org/10.1016/j.atmosenv.2009.05.035. http://www.sciencedirect.com/science/article/pii/S1352231009004506. Briggs, D.J., de Hoogh, K., Morris, C., Gulliver, J., 2008. Effects of travel mode on exposures to particulate air pollution. Environ. Int. 34, 12e22. http://dx.doi.org/ 10.1016/j.envint.2007.06.011. http://www.sciencedirect.com/science/article/pii/ S0160412007001249. Cartenì, A., Cascetta, F., Campana, S., 2015. Underground and ground-level particulate matter concentrations in an Italian metro system. Atmos. Environ. 101, 328e337. http://dx.doi.org/10.1016/j.atmosenv.2014.11.030. http://www. sciencedirect.com/science/article/pii/S1352231014008954. Cattaneo, A., Garramone, G., Taronna, M., Peruzzo, C., Cavallo, D.M., 2009. Personal exposure to airborne ultrafine particles in the urban area of Milan. Inhaled Particles X, (23e25 September 2008, Manchester). J. Phys. Conf. Ser. 151, 012039. http://dx.doi.org/10.1088/1742-6596/151/1/012039, 2009. http:// iopscience.iop.org/article/10.1088/1742-6596/151/1/012039/meta; jsessionid¼7184A77DC123EA41C086CFD139F51C6A.c1. Cattani, G., Di Menno di Bucchianico, A., Inglessis, M., 2013. Car commuters' personal exposure to ultrafine particles. Environ. Eng. Manag. J. 12 (S11), 217e220. http://omicron.ch.tuiasi.ro/EEMJ/pdfs/vol12/no11suppl/56_Cattani_13.pdf. Cusack, M., Talbot, N., Ondracek, J., Minguillon, M.C., Martins, V., Klouda, K., Schwarz, J., Zdímal, V., 2015. Variability of aerosols and chemical composition of PM10, PM2.5 and PM1 on a platform of the Prague underground metro. Atmos. Environ. 118, 176e183. http://dx.doi.org/10.1016/j.atmosenv.2015.08.013. http:// www.sciencedirect.com/science/article/pii/S1352231015302582. Fairbrother, A., Wenstel, R., Sappington, K., Wood, W., 2007. Framework for metal risk assessment. Ecotoxicol. Environ. Saf. 68, 145e227. http://dx.doi.org/ 10.1016/j.ecoenv.2007.03.015. http://www.pebblescience.org/Pebble-Mine/ acid-drainage-pdfs/MetalsRiskAssessmentEPA07.pdf. Geiss, O., Barrero-Moreno, J., Tirendi, S., Kotzias, D., 2010. Exposure to particulate matter in vehicle cabins of private cars. Aerosol Air Qual. Res. 10, 581e588. http://dx.doi.org/10.4209/aaqr.2010.07.0054. http://aaqr.org/VOL10_No6_ December2010/7_AAQR-10-07-OA-0054_581-588.pdf. Gietl, J.K., Lawrence, R., Thorpe, A.J., Harrison, R.M., 2010. Identification of brake wear particles and derivation of a quantitative tracer for brake dust at a major road. Atmos. Environ. 44, 141e146. http://dx.doi.org/10.1016/j.atmosenv.2009.10.016. http://www.sciencedirect.com/science/article/pii/ S1352231009008735. IntPanis, L., de Geus, B., Vandenbulcke, G., Willems, H., Degraeuwe, B., Bleux, N., Mishra, V., Thomas, I., Meeusen, R., 2010. Exposure to particulate matter in traffic: a comparison of cyclists and car passengers. Atmos. Environ. 44, 2263e2270. http://dx.doi.org/10.1016/j.atmosenv.2010.04.028. http://www. sciencedirect.com/science/article/pii/S1352231010003225. ISO 15767, 2009. Workplace Atmospheres - Controlling and Characterizing Uncertainty in Weighing Collected Aerosols (Geneva). Jung, H.-J., Kim, B., Ryu, J., Maskey, S., Kim, J.-C., Sohn, J., Ro, C.-U., 2010. Source identification of particulate matter collected at underground subway stations in Seoul, Korea using quantitative single-particle analysis. Atmos. Environ. 44, 2287e2293. http://dx.doi.org/10.1016/j.atmosenv.2010.04.003. http://www. sciencedirect.com/science/article/pii/S1352231010002852. Kaur, S., Nieuwenhuijsen, M., Colvile, R., 2005. Personal exposure of street canyon intersection users to PM2.5, ultrafine particle counts and carbon monoxide in Central London, UK. Atmos. Environ. 39, 3629e3641. http://dx.doi.org/10.1016/

209

j.atmosenv.2005.02.046. http://www.sciencedirect.com/science/article/pii/ S1352231005002268. Kaur, S., Nieuwenhuijsen, M.J., Colvile, R.N., 2007. Fine particulate matter and carbon monoxide exposure concentrations in urban street transport microenvironments. Atmos. Environ. 41, 4781e4810. http://dx.doi.org/10.1016/ j.atmosenv.2007.02.002. http://www.sciencedirect.com/science/article/pii/ S1352231007001343. Knibbs, L.D., de Dear, R.J., Atkinson, S.E., 2009. Field study of air change and flow rate in six automobiles. Indoor Air 19, 303e313. http://dx.doi.org/10.1111/ j.1600-0668.2009.00593.x. http://onlinelibrary.wiley.com/doi/10.1111/j.16000668.2009.00593.x/pdf. Knibbs, L.D., Cole-Hunter, T., Morawska, L., 2011. A review of commuter exposure to ultrafine particles and its health effects, 45 Atmos. Environ. 16, 2611e2622. http://dx.doi.org/10.1016/j.atmosenv.2011.02.065. http://www.sciencedirect. com/science/article/pii/S1352231011002299. Knibbs, L.D., de Dear, R.J., 2010. Exposure to ultrafine particles and PM2.5 in four Sydney transport modes. Atmos. Environ. 44, 3224e3227. http://dx.doi.org/ 10.1016/j.atmosenv.2010.05.026. http://www.sciencedirect.com/science/article/ pii/S1352231010003997. Kryzanowski, M., Kuna-Dibbert, B., Schneider, J., 2005. Health Effects of Transportrelated Air Pollution: Summary for Policy-makers. WHO Regional Office for Europe, Copenhagen. http://www.euro.who.int/__data/assets/pdf_file/0006/ 74715/E86650.pdf. Künzli, N., Kaiser, R., Medina, S., Studnika, M., Chanel, O., Filliger, P., Hery, M., Horak Jr., F., Puybonnieux-Texier, V., Quenel, P., Schneider, J., Seethaler, R., Vergnaud, J.-C., Sommer, H., 2000. Public health impact of outdoor and trafficrelated air pollution: a European assessment. Lancet 356, 795e801. http:// dx.doi.org/10.1016/S0140-6736(00)02653-2. http://www.sciencedirect.com/ science/article/pii/S0140673600026532. Li, H.C., Chiueh, P.T., Liu, S.P., Huang, Y.Y., 2017. Assessment of different route choice on commuters' exposure to air pollution in Taipei, Taiwan. Environ. Sci. Pollut. Res. 24 (3), 3163e3171. n, M.C., Amato, F., de Miguel, E., Capdevila, M., Martins, V., Moreno, T., Minguillo Querol, X., 2015. Exposure to airborne particulate matter in the subway system. Sci. Total Environ. 511, 711e722. http://dx.doi.org/10.1016/j.scitotenv.2014.12.013. http://www.sciencedirect.com/science/article/pii/ S004896971401715X. Matson, U., Ekberg, L.E., Afshari, A., 2004. Measurement of ultrafine particles: a comparison of two handheld condensation particle counters. Aerosol Sci. Technol. 38, 487e495. http://dx.doi.org/10.1080/02786820490462200. http:// dx.doi.org/10.1080/02786820490462200. Miao, Q., Bouchard, M., Chen, D., Rosenberg, M.W., Aronson, K.J., 2015. Commuting behaviors and exposure to air pollution in Montreal, Canada. Sci. Total Environ. 508, 193e198. http://dx.doi.org/10.1016/j.scitotenv.2014.11.078. http://www. sciencedirect.com/science/article/pii/S0048969714016726. Morawska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U., Ling, X., 2008. Ambient nano and ultrafine particles from motor vehicle emissions: characteristics, ambient processing and implications on human exposure. Atmos. Environ. 42, 8113e8138. http://dx.doi.org/10.1016/ j.atmosenv.2008.07.050. http://www.sciencedirect.com/science/article/pii/ S1352231008006961. rez, N., Reche, C., Martins, V., de Miguel, E., Capdevila, M., Centelles, S., Moreno, T., Pe  n, M.C., Amato, F., Alastuey, A., Querol, X., Gibbons, W., 2014. Subway Minguillo platform air quality: assessing the influences of tunnel ventilation, train piston effect and station design. Atmos. Environ. 92, 461e468. http://dx.doi.org/ 10.1016/j.atmosenv.2014.04.043. http://www.sciencedirect.com/science/article/ pii/S135223101400315X.  n, M., Martins, V., Vargas, C., Moreno, T., Reche, C., Rivas, I., Cruz Minguillo Buonanno, G., Parga, J., Pandolfi, M., Brines, M., Ealo, M., Sofia Fonseca, A., Amato, F., Sosa, G., Capdevila, M., de Miguel, E., Querol, X., Gibbons, W., 2015a. Urban air quality comparison for bus, tram, subway and pedestrian commutes in Barcelona. Environ. Res. 142, 495e510. rube , K., Minguillo n, M.C., Amato, F., Moreno, T., Martins, V., Querol, X., Jones, T., Be Capdevila, M., de Miguel, E., Centelles, S., Gibbons, W., 2015b. A new look at inhalable metalliferous airborne particles on rail subway platforms. Sci. Total Environ. 505, 367e375. http://dx.doi.org/10.1016/j.scitotenv.2014.10.013. http:// www.sciencedirect.com/science/article/pii/S004896971401451X. mez-Perales, E., Colvile, R.N., 2007. Levels of particulate air Nieuwenhuijsen, M.J., Go pollution, its elemental composition, determinants and health effects in metro systems. Atmos. Environ. 41, 7995e8006. http://dx.doi.org/10.1016/j.atmosenv.2007.08.002. http://www.sciencedirect.com/science/article/pii/ S135223100700698X. Ripanucci, G., Grana, M., Vicentini, L., Magrini, A., Bergamaschi, A., 2006. Dust in the underground railway tunnels of an Italian town. J. Occup. Environ. Hyg. 3, 16e25. http://dx.doi.org/10.1080/15459620500444004. Sanderson, P., Delgado-Saborit, J.M., Harrison, R.M., 2014. A review of chemical and physical characterisation of atmospheric metallic nanoparticles. Atmos. Environ. 94, 353e365. http://dx.doi.org/10.1016/j.atmosenv.2014.05.023. http:// www.sciencedirect.com/science/article/pii/S1352231014003707. Schweizer, C., Edwards, R.D., Bayer-Oglesby, L., Gauderman, W.J., Ilacqua, V., Jantunen, M.J., Kan Lai, H., Nieuwenhuijsen, M., Kunzli, N., 2007. Indoor timeemicroenvironmenteactivity patterns in seven regions of Europe. J. Expo. Sci. Environ. Epidemiol. 17, 170e181. Thorpe, A., Harrison, R.M., 2008. Sources and properties of non-exhaust particulate matter from road traffic: a review. Sci. Total Environ. 400, 270e282. http://

210

M. Grana et al. / Environmental Pollution 228 (2017) 201e210

dx.doi.org/10.1016/j.scitotenv.2008.06.007. http://www.sciencedirect.com/ science/article/pii/S004896970800658X. Van Roosbroeck, S., Hoek, G., Meliefste, K., Janssen, N.A., Brunekreef, B., 2008. Validity of residential traffic intensity as an estimate of long-term personal exposure to traffic-related air pollution among adults. Environ. Sci. Technol. 42, 1337e1344. http://dx.doi.org/10.1021/es0712827. http://pubs.acs.org/doi/pdf/ 10.1021/es0712827. Wichmann, J., Janssen, N.A.H., van der Zee, S.C., Brunekreef, B., 2005. Traffic-related differences in indoor and personal absorption coefficient measurements in Amsterdam, The Netherlands. Atmos. Environ. 39, 7384e7392. http:// dx.doi.org/10.1016/j.atmosenv.2005.09.015. http://www.sciencedirect.com/ science/article/pii/S1352231005008113.

World Health Organization, 2000. Air Quality Guidelines for Europe. European series n. 91. World Health Organization Regional Office for Europe, second ed. World Health Organization Regional Publications, Copenhagen. http://www. euro.who.int/__data/assets/pdf_file/0005/74732/E71922.pdf. World Health Organization (WHO), 2004. Health Aspects of Air Pollution Results from the WHO Project “Systematic Review of Health Aspects of Air Pollution in Europe”. WHO Regional Office for Europe, Copenhagen, Denmark. WHO 2004. http://www.euro.who.int/__data/assets/pdf_file/0003/74730/E83080.pdf? ua¼1. Zhu, Y., Yu, N., Kuhn, T., Hinds, W., 2006. Field comparison of P-Trak and condensation particle counters. Aerosol Sci. Technol. 40, 422e430. http://dx.doi.org/ 10.1080/02786820600643321. http://dx.doi.org/10.1080/02786820600643321.