Elsevier

Atmospheric Environment

Volume 42, Issue 4, February 2008, Pages 688-702
Atmospheric Environment

Factors affecting non-tailpipe aerosol particle emissions from paved roads: On-road measurements in Stockholm, Sweden

https://doi.org/10.1016/j.atmosenv.2007.09.064Get rights and content

Abstract

A large fraction of urban PM10 concentrations is due to non-exhaust traffic emissions. In this paper, a mobile measurement system has been used to quantify the relative importance of road particle emission and suspension of accumulated dust versus direct pavement wear, tire type (studded, friction, and summer), pavement type, and vehicle speed. Measurements were performed during May–September on selected roads with different pavements and traffic conditions in the Stockholm region. The highest particle mass concentrations were always observed behind the studded tire and the lowest were behind the summer tire; studded-to-summer ratios were 4.4–17.3 and studded-to-friction ratios were 2.0–6.4. This indicates that studded tires lead to higher emissions than friction and summer tires regardless to the asphalt type. By comparing with measurements in a road simulator, it could be estimated that the pavement wear due to the friction tires was 0.018–0.068 of the suspension of accumulated road dust. Likewise for studded tires road-wear was estimated to be 1.2–4.8 the suspension of accumulated dust. This indicates that wear due to friction tires is very small compared to the suspension of accumulated dust and that suspension due to studded tires may sometimes be as large as the wear of the road. But this will vary depending on, e.g. the amount of dust accumulated on the roads. An important dependence on vehicle speed was also observed. During May, the particle mass concentrations behind the studded tire at vehicle speed 100 km h−1 were about 10 times higher than that at 20 km h−1. The speed dependence was not so pronounced in September, which could be due to less accumulated dust on the roads. The particle number size distribution of the emissions due to road wear by studded tire was characterized by a clear increase in number concentrations of the coarse fraction of aerosol particles, with a geometric mean diameter between 3 and 5 μm. The size distribution of the emissions due to the summer tire was very similar with smaller concentrations. An important limitation with the measurements presented is that they were made by using a van, which is bigger than regular cars and has bigger tires. Thus, road wear and dust suspension due to cars are expected to be different.

Introduction

Combustion and non-combustion/non-tailpipe emissions due to traffic activities are the main sources of particulate matter (PM) in the urban atmosphere. Non-tailpipe emissions may include both road dust and wear of vehicular parts such as brakes, tires, etc. (e.g. Rogge et al., 1993; Garg et al., 2000; Kupiainen et al., 2005; Wåhlin et al., 2006; Almeida et al., 2006). While non-combustion is a more general term than non-tailpipe, it may additionally include emissions due to internal engine wear. Non-tailpipe emissions constitute a major fraction of the primary PM10 emissions in urban areas of Nordic countries. For example, road dust emission dominate PM10 emissions during the late-winter and early-spring seasons in Stockholm (Omstedt et al., 2005; Norman and Johansson, 2006; Johansson et al., 2007), Helsinki (Pohjola et al., 2002; Laakso et al., 2003), Oslo (Lützenkirchen and Lutnaes, 2005), Trondheim and Bergen (Laupsa et al., 2005), as well as Gothenburg and Lycksele (Areskoug et al., 2004). Road dust re-suspension is also a serious problem in other urban cities worldwide (Etyemezian et al., 2003a, Etyemezian et al., 2003b; Zhao et al., 2006), especially nearby dust sources such as quarries (Abu-Allaban et al., 2006).

Since the 1940s, studies have been focused on removal of particles from a surface due to near-by-the-surface airflow, turbulent induced bursts (e.g. due to traffic), and soil erosion (e.g. Bagnold, 1941; Chepil, 1945; Sehmel, 1973; Nicholson et al., 1989; Nicholson and Branson, 1990). Car speed and shape as well as car–body distance from the road surface play a significant rule in road dust re-suspension (Sehmel, 1973; Oude Weenink, 1976; Mollinger et al., 1993); mainly due to intensity of induced turbulences behind the car. Size of deposited dust particles and the topography of the road surface are also influencing factors (Mollinger et al., 1993). Large differences in PM10 emissions can be also found from paved and unpaved roads (e.g. Claiborn et al. 1995). In general, there may be a threshold value of the airflow speed that must be exceeded before a particle is re-suspended (Sehmel, 1973). Particles of a given size and density are more easily re-suspended from a smooth surface than from an irregular surface of an asphalt road. In contrast, if the particles are on dirt or wet road, they are less likely to become airborne again. The effect of wind speed on particle emissions in urban areas is small compared to vehicle-induced turbulence (Mårtensson et al., 2006). Road dust is also directly emitted as a result of tire–road surface interaction (Kupiainen, 2007).

In practice, quantification of road dust emissions is complicated because of the many different factors that might need to be controlled—vehicle type and speed, tire type, and pavement type and conditions, as well as the use of road salt and sand. Sanding has been found to increase the PM10 emissions due to wear of sand granules and enhanced wear of the road surface (e.g. Kupiainen and Tervahattu, 2004; Kupiainen et al., 2005; Tervahattu et al., 2006). The use of studded tires also enhances the road surface wear which increases the PM10 concentrations, especially during dry road conditions (e.g. Kupiainen et al., 2003, Kupiainen et al., 2005; Norman and Johansson, 2006). De-icers such as calcium magnesium acetate and calcium chloride have been seen to reduce the amount of re-suspended road dust due to the wetting of the road, whereas road sweeping may significantly increase the PM10 concentrations during and right after the sweeping (e.g. Gertler et al., 2006; Norman and Johansson, 2006).

Recently, intensive laboratory measurements with road test facilities (e.g. Dahl et al., 2006; Kupiainen et al., 2003, Kupiainen et al., 2005) and on-road measurements with mobile laboratories (e.g. Kuhns et al., 2001) have been used to quantify aerosol particle emissions from road surfaces. Both techniques indicated that aerosol particle emissions from tire–road interface are not only coarse but also ultrafine (UF, diameter <100 nm). Mobile laboratories, e.g. TRAKER, have been very useful to quantify the effect in real-world conditions with respect to different road surfaces and dust loading, weather condition, location, seasons, vehicle speed, traffic volumes, etc. (Etyemezian et al., 2003a, Etyemezian et al., 2003b; Kuhns et al., 2003).

It is interesting to note that total road wear due to studded tires has been studied in a road simulator at VTI (the Swedish National Road and Transport Research Institute, Linköping, Sweden) since many years and this has been summarized in a model of road wear (Jacobson and Wågberg, 2004). From these measurements, it has been concluded that the most important factors for the wear are the quality of the stone material in the asphalt (wear can differ by a factor 8 between different asphalt types), type of studs (light weight studs (maximum 1.1 g) may cause ca. 50% less wear than heavier steel studs of 1.8 g), vehicle speed (increasing from 50  to 100 km h−1 increase wear by a factor 2), and road surface wetness (higher wear on a wet road compared to dry). One might think that PM10 emissions are in some way directly related to the total road wear when studded tires are used, but this has not yet been investigated.

In order to find the most efficient actions, a better understanding about the processes involved in non-tailpipe emissions is needed. In this paper, we present results from on-road measurements of non-tailpipe emissions. The purpose is to assess quantitative relationships between different key factors that cause the generation of road particle emissions: tire type (summer/friction/studded), asphalt type, role of re-suspension versus direct emission, and seasonal effects. This study is limited to only one brand of each tire type but it covers a wide area of the Stockholm region to include several types of road pavements. The winter season was not included in this study because of the limited time and the road surface conditions (wet/frozen road surfaces).

Section snippets

Mobile laboratory and instrumentation

Our mobile laboratory setup was based on a modified version of the second version of the TRAKER (Testing Re-entrained Aerosol Kinetic Emissions from Roads) as described by Etyemezian et al. (2003b). Our vehicle was a Volkswagen van (LT 35, 2002) equipped with three metal tube inlets: two were mounted behind the front tires and one was mounted underneath the van that extended to sample background air bellow the front bumper (Fig. 1). The three inlet lines entered the van compartment through the

Road survey measurements: influence of tire type and asphalt type

The concentrations measured behind the tires and in front of the bumper during driving and standing are shown in Fig. 4. The concentrations observed during standing periods represent background concentrations and they were quite similar to the concentration in front of the van, which shows that the impact of emissions from other vehicles on the roads was small during the measurements. The highest particle mass concentrations were always observed behind the studded tire and the lowest

Conclusions

In this study, we have successfully applied an on-road mobile measurement system to show that studded tires (Gislaved Nord Frost C) increase road dust emissions by 2.0–6.4 times as compared to friction tires (Nokian Hakkapeliitta CQ) and 4.4–17 times as compared to summer tires (Dunlop SP LT8). For the friction tire most emission was due to suspension, whereas for the studded tire road wear was the main cause of the emissions during the studied period (May–September). The emission depends on

Acknowledgments

The authors would like to thank Kai Rosman at the Department of Applied Environmental Science, Stockholm University, for programming the data-logging and Leif Bäcklin at the Department of Meteorology, Stockholm University, for skillful technical assistance.

The Swedish Road Administration is acknowledged for the financial support to the EMMA mobile laboratory. Dr. Tareq Hussein acknowledges the financial support from the European Science Foundation (ESF) within the Interdisciplinary Troposphere

References (41)

Cited by (132)

View all citing articles on Scopus
View full text