This study aims to explore the influential sources of organic matter in first flush runoff from urban roadways by comparing organic carbon content and particle size distribution in road dust with those from discharge from vehicles during rainfall. Samples on first flush runoff and road dust were collected from urban roadways. In addition, vehicle drainage was assumed to flow from vehicles during rainfall events, so vehicle wash-off water was collected by spraying water onto the top and from the underside of vehicles to simulate accumulation during a vehicle run. In road dust, the organic carbon content in the <0.2 mm fraction was about twice that of the 0.2–2 mm fraction. The particle size distributions of both first flush runoff and vehicle wash-off water were similar, and particles <0.2 mm contributed to over 95% of the total volume. The dissolved organic carbon concentration in the vehicle wash-off water was considerably higher than that in the road dust/water mixture. The total organic carbon content in road dust was positively correlated with annual daily traffic. Therefore, vehicles were thought to strongly influence the nature of road dust.

INTRODUCTION

First flush runoff from urban roadways has increased as a result of increased population and traffic due to urban development and impervious land surfaces (Novotny 2003). Such runoff contains a number of pollutants, derived from vehicle exhaust emissions, tyres, asphalt, litter and so on, which are accumulated in road dust (US Environmental Protection Agency (US EPA) 1983). The runoff directly or indirectly acts as a source of water pollution when it eventually flows into rivers, lakes or enclosed water areas. These pollutants can enter and severely affect aquatic systems mainly via runoff. The water quality of public water bodies in Japan has shown a tendency for overall improvement in terms of organic matter content, but the improvement is still insufficient in enclosed water areas such as lakes and inland seas (Wada et al. 2006; Japanese Ministry of the Environment (Japanese MOE) 2014).

In many previous studies, organic matter pollution in roadway runoffs have been reported with a focus on toxic organic chemicals such as polycyclic aromatic hydrocarbons (PAHs), volatile organic compounds and pesticides (e.g., Björklund et al. 2009; Mahbub et al. 2011; Zgheib et al. 2012; Ozaki et al. 2015). The concentrations of these pollutants are in the range of μg–ng/L, and are significantly different from those of the organic matter present in runoffs. Wada et al. (2006, 2011) showed that organic compounds containing non-biodegradable substances were present in high concentrations in first flush runoff from urban roadways. In addition, it was confirmed that a certain amount of organic pollutants were discharged even after the first flush runoff by analysing the chronological variation in total organic carbon (TOC) concentration in roadway runoffs (Wada & Fujii 2006). Accordingly, part of organic pollutant loads by roadway runoffs might be derived from vehicle bodies, automobile fuel and engine oil (e.g., poly α-olefin and paraffin). However, since these loads comprise diverse organic compounds, it is very difficult to evaluate organic matter loadings by measuring all individual compounds.

Conversely, studies simulating non-point source pollution from roadways have often considered both the wash-off and deposition processes. Furthermore, other studies have adopted a total suspended solids (SS) concentration as one of the target parameters of pollutants (Sartor & Boyd 1972; Alley & Smith 1981; Hijioka & Furumai 2001). In addition, the SS contributes to water and soil contamination (Mangani et al. 2005; Murakami et al. 2006). It is especially suggested that traffic, rainfall condition and soil size are crucial for the contamination. However, a dominating process of organic loads via roadway runoff is not fully clarified.

The objective of this study is to explore the influential sources of organic matter in first flush runoff from urban roadways by comparing organic carbon content and particle size distribution in road dust with those in discharge from vehicles during rainfalls. The aim is to elucidate the contribution of vehicle wash-off as a potential source of organic matter in urban roadway runoff.

MATERIALS AND METHODS

First flush runoff

Samples of first flush runoff from rainfall events at five roads in an urban area of the watershed of Lake Biwa, which is the largest freshwater lake in Japan, were collected from 1999 to 2014. The survey station roadways, which were typically urban roadways, had average daily traffic (ADT) of approximately 15,000 to 36,000 vehicles (Japanese Ministry of Land Infrastructure Transport & Tourism (Japanese MLIT) 1999, 2005, 2010) and were surfaced with asphalt. The road runoff samples were the cumulative discharge of 2 to 7 mm from the onset of rain. The samples were collected by hand (St. 1, 2, 3 and 5) or by previously installed ‘First Flush Cleaner’ (Japanese Patent No. 3768186 (2006)) units (St. 4), which can collect only the dirty drainage portion at the beginning of a rain event. The specifications of the survey stations are detailed in Table 1. For improving the precision of the measurements of various roadway runoffs, new additional samples were collected and analyzed at St. 1 (2009–2010), St. 4 (2010–2014) and St. 5 (2013–2014). The surrounding area of St. 5 had undergone rapid urbanization with an increase in the traffic.

Table 1

Specifications of survey station for first flush runoff

St. Geographic locations Number of rainfall events Period of survey (year) ADTa (vehicles/day) Catchment area (m2References 
35.084 °N 136.067 °E 2004, 2005 24,457 75 Wada & Fujii (2010)  
2009 Wada et al. (2011)  
2009–2010 This study 
35.051 °N 135.920 °E 1999–2000 15,152 750 Wada & Fujii (2010)  
35.072 °N 136.028 °E 2003 17,032 285 Wada et al. (2006
35.038 °N 135.971 °E 2002–2005 22,288 72 Wada & Fujii (2007)  
2007 Wada et al. (2011)  
2010–2014 This study 
35.013 °N 135.937 °E 2013–2014 35,888 126 This study 
St. Geographic locations Number of rainfall events Period of survey (year) ADTa (vehicles/day) Catchment area (m2References 
35.084 °N 136.067 °E 2004, 2005 24,457 75 Wada & Fujii (2010)  
2009 Wada et al. (2011)  
2009–2010 This study 
35.051 °N 135.920 °E 1999–2000 15,152 750 Wada & Fujii (2010)  
35.072 °N 136.028 °E 2003 17,032 285 Wada et al. (2006
35.038 °N 135.971 °E 2002–2005 22,288 72 Wada & Fujii (2007)  
2007 Wada et al. (2011)  
2010–2014 This study 
35.013 °N 135.937 °E 2013–2014 35,888 126 This study 

Road dust

Road dust samples were collected from 14 roadside stations using a brush from road gutter surfaces in the Lake Biwa–Yodo River watershed area (8,240 km2), for a total of 23 data samples. The sampling station roadways, which were typically urban roadways or arterial roads, had ADT of 4,082 to 76,599 vehicles in 2010 (Japanese MLIT 1999, 2005, 2010) and were surfaced with asphalt. The sampling areas were 0.0625 m2 or 0.25 m2. The specifications of the sampling collection sites are detailed in Table 2. The road dust was air-dried at room temperature and homogenized after being sieved through a 2 mm mesh screen to remove coarse dust and pebbles. Then, seven dust samples were fractionated by stepwise filtration, using stainless screens of 212 μm and 2 mm mesh size, which screened two different particle size fractions: 0.2–2.0 mm and <0.2 mm. Road dust and distilled water were mixed at a solid/liquid (S/L) ratio of 10,000 mg-dry weight/L, and sufficiently shaken by a magnetic stirrer (ASONE, REXIM RSH-6DR) at a speed of 250 rpm for 1 h at room temperature to resuspend the settled particles. Particulate and dissolved fractions of TOC were then immediately separated with a 0.45 μm membrane filter. Dissolved organic carbon (DOC) and particulate organic carbon (POC) concentrations were transformed into those at the S/L ratio of 100 mg-dry weight/L, which was an average of total SS concentration (100 ± 65 mg/L; average ± standard deviation (SD), n = 23) observed in first flush runoffs. A series of mixing tests were performed using only seven road dust samples, because the remaining 16 samples could not provide sufficient quantity of dust for the test. The t-test showed that the difference in ADT between seven stations and the remaining 16 stations was not statistically significant at the P = 0.01 level.

Table 2

Specifications of sampling collection sites for road dust

Site Geographic locations Sampling date ADTa (vehicles/day) Previous rainfall amount** (mm) Dry weather periodb (day) Analysis of particle sizec 
RD1 34.842 °N 11/6/2010 17,974 134.5 17.8   
135.567 °E 
23/8/2010 17.0 9.2  
29/11/2010 7.5 6.5  
28/1/2011 11.0 37.8  
11/8/2011 31.5 5.8  
RD2 34.789 °N 28/1/2011 34,241 11.0 37.4 ✓ 
135.580 °E 
6/6/2012 3.0 3.6 ✓ 
2/8/2012 7.0 7.9 ✓ 
2/11/2012 8.5 4.8 ✓ 
RD3 34.996 °N 3/9/2010 57,356 135.5 22.1   
135.742 °E 
6/6/2012 11.0 7.7 ✓ 
RD4 35.039 °N 135.970 °E 6/6/2012 23,813 3.5 5.6 ✓ 
RD5 34.734 °N 135.337 °E 2/11/2012 76,599 8.5 4.6 ✓ 
RD6 34.991 °N 3/9/2010 43,186 135.5 22.1   
135.753 °E 
6/6/2012 11.0 7.7  
RD7 35.077 °N 136.048 ° 6/6/2012 23,401 3.5 5.6  
RD8 35.047 °N 135.912 °E 3/9/2010 15,535 7.0 4.8  
RD9 34.767 °N 136.131 °E 28/1/2011 9,110 5.5 27.8  
RD10 34.906 °N 135.410 °E 28/1/2011 4,082 13.0 37.7  
RD11 34.693 °N 135.812 °E 28/1/2011 49,189 6.0 28.5  
RD12 34.477 °N 135.963 °E 4/8/2011 5,439 93.0 15.7  
RD13 34.476 °N 135.933 °E 5/8/2011 9,785 93.0 16.7  
RD14 34.558 °N 135.461 °E 2/11/2012 26,525 3.0 4.8  
Site Geographic locations Sampling date ADTa (vehicles/day) Previous rainfall amount** (mm) Dry weather periodb (day) Analysis of particle sizec 
RD1 34.842 °N 11/6/2010 17,974 134.5 17.8   
135.567 °E 
23/8/2010 17.0 9.2  
29/11/2010 7.5 6.5  
28/1/2011 11.0 37.8  
11/8/2011 31.5 5.8  
RD2 34.789 °N 28/1/2011 34,241 11.0 37.4 ✓ 
135.580 °E 
6/6/2012 3.0 3.6 ✓ 
2/8/2012 7.0 7.9 ✓ 
2/11/2012 8.5 4.8 ✓ 
RD3 34.996 °N 3/9/2010 57,356 135.5 22.1   
135.742 °E 
6/6/2012 11.0 7.7 ✓ 
RD4 35.039 °N 135.970 °E 6/6/2012 23,813 3.5 5.6 ✓ 
RD5 34.734 °N 135.337 °E 2/11/2012 76,599 8.5 4.6 ✓ 
RD6 34.991 °N 3/9/2010 43,186 135.5 22.1   
135.753 °E 
6/6/2012 11.0 7.7  
RD7 35.077 °N 136.048 ° 6/6/2012 23,401 3.5 5.6  
RD8 35.047 °N 135.912 °E 3/9/2010 15,535 7.0 4.8  
RD9 34.767 °N 136.131 °E 28/1/2011 9,110 5.5 27.8  
RD10 34.906 °N 135.410 °E 28/1/2011 4,082 13.0 37.7  
RD11 34.693 °N 135.812 °E 28/1/2011 49,189 6.0 28.5  
RD12 34.477 °N 135.963 °E 4/8/2011 5,439 93.0 15.7  
RD13 34.476 °N 135.933 °E 5/8/2011 9,785 93.0 16.7  
RD14 34.558 °N 135.461 °E 2/11/2012 26,525 3.0 4.8  

aAverage daily traffic vehicles on weekdays, Road Traffic Census (Japanese MLIT 1999, 2005, 2010).

cThese seven samples were separated into two parts by size (0.2–2.0 mm and <0.2 mm) and were analysed as 0.01 w/v% road dust/water mixture.

Vehicle wash-off water

The water pollutant load from first flush runoff depends on the previous dry weather period and the wash-off process during the actual rainfall event. Road dust is a deposit of dry fallout during dry weather periods and is washed away by rainfall. However, in the same rainfall event, the TOC concentration in the first flush runoff was high compared with that in the road dust solution prepared with the same level of SS concentration. This suggested that first flush runoff is also influenced by contaminants other than road dust, such as pollution from vehicles. Therefore, vehicle drainage was assumed to flow from the vehicles during rainfall events. The intended experiment was conducted using two vehicles which were driven almost daily, and before the vehicles were washed (Table 3).

Table 3

Specifications of vehicle using experiment for vehicle wash-off water

Car type Sampling date Previous rainfall amounta (mm) Dry weather periodb (day) Driving area after previous rainfall 
Type A: Toyota Hiace Van S-GL (2490-cc diesel engine) 15/4/2013 16.5 Shiga pref. 
1/11/2013 21 Shiga pref. 
15/4/2014 11 Shiga pref. 
7/8/2014 11 Shiga and Kanagawa pref. 
3/10/2014 Shiga pref. 
Type B: Toyota Regius Ace (2700-cc gas engine) 15/4/2013 16.5 Shiga and Kyoto pref. 
1/11/2013 21 Aichi and Shiga pref. 
15/4/2014 11 Fukui, Kyoto and Shiga pref. 
7/8/2014 11 Aichi and Shiga pref. 
3/10/2014 Aichi, Fukui and Shiga pref. 
Car type Sampling date Previous rainfall amounta (mm) Dry weather periodb (day) Driving area after previous rainfall 
Type A: Toyota Hiace Van S-GL (2490-cc diesel engine) 15/4/2013 16.5 Shiga pref. 
1/11/2013 21 Shiga pref. 
15/4/2014 11 Shiga pref. 
7/8/2014 11 Shiga and Kanagawa pref. 
3/10/2014 Shiga pref. 
Type B: Toyota Regius Ace (2700-cc gas engine) 15/4/2013 16.5 Shiga and Kyoto pref. 
1/11/2013 21 Aichi and Shiga pref. 
15/4/2014 11 Fukui, Kyoto and Shiga pref. 
7/8/2014 11 Aichi and Shiga pref. 
3/10/2014 Aichi, Fukui and Shiga pref. 

The spray amount corresponding to the first flush runoff from the surface area (37–39 m2) of the car body was calculated as the cumulative discharge amount of 5 mm per car. The vehicle wash-off water experiment simulated the rain on the vehicle during its run. The experimental method comprised the following steps. First, 2 mm distilled water (pH = 7.4) was sprayed on the top and side of the car body in a manner that emulates rainfall. Thereafter, 3 mm distilled water was sprayed underneath the car to simulate the effect of water and mud splashed by the wheels. This spray intensity corresponded to a rainfall intensity of approximately 2.4 mm/h. All discharge samples corresponding to the first flush runoff were caught by a polyethylene sheet (washed and protected from contamination prior to the experiment), which lay under the car, and was collected in a container. This experiment was conducted for ten samples (two vehicles per experiment and five times). A schematic of the vehicle wash-off water collection process is shown in Figure 1.

Figure 1

Schematic and actual photo of vehicle wash-off water collection.

Figure 1

Schematic and actual photo of vehicle wash-off water collection.

Chemical and particle size distribution analyses

Total organic carbon was used as an index of organic compounds. Particulate and dissolved fractions of TOC were separated with a 0.45 μm membrane filter. Then, DOC was determined by JIS K0102-22.1 (Japanese Industrial Standards Committee 1998) and measured with a TOC analyzer (Toray Engineering TOC-650). Particulate organic carbon was measured with an NC analyzer (Sumigraph NC-22F), and TOC was calculated by summing DOC and POC.

Particle size distributions of six first flush runoffs and all vehicle wash-off water collections were measured with a laser diffraction particle size analyzer (Shimazu SALD-3100).

RESULTS AND DISCUSSION

Organic carbon concentrations in samples

First flush runoff

The concentration of TOC, POC, DOC and SS in the first flush runoff was 32.2 ± 19.3 mgC/L, 17.0 ± 11.6 mgC/L, 15.3 ± 13.9 mgC/L, and 100 ± 65 mg/L (average ± SD), respectively (Table 4). The average concentrations of POC and DOC were found to be similar in the first flush runoff. However, each maximum value was much higher than the average and median values. This suggests that the concentrations of contaminants contained in first flush runoff vary greatly from one rainfall event to another.

Table 4

Organic carbon and SS concentration in first flush runoff from road

 Average ± SD Median Minimum Maximum Number of samples 
TOC (mgC/L) 32.2 ± 19.3 28.7 3.0 88.3 36 
POC (mgC/L) 17.0 ± 11.6 15.8 0.5 50.8 36 
DOC (mgC/L) 15.3 ± 13.9 11.4 1.9 68.7 36 
SS (mg/L) 100 ± 65 90 15 320 23 
 Average ± SD Median Minimum Maximum Number of samples 
TOC (mgC/L) 32.2 ± 19.3 28.7 3.0 88.3 36 
POC (mgC/L) 17.0 ± 11.6 15.8 0.5 50.8 36 
DOC (mgC/L) 15.3 ± 13.9 11.4 1.9 68.7 36 
SS (mg/L) 100 ± 65 90 15 320 23 

SD: standard deviation.

Road dust

Organic carbon content in the road dust was 46 ± 40 mgC/g (n = 23), of which seven road dust samples applied to the mixing tests and the remaining 16 samples were 77 ± 47 mgC/g and 33 ± 30 mgC/g, respectively. Here, the concentration of the selected seven dust samples were evaluated as follows. Arnarson & Keil (2000) reported that the adsorption mass of the natural organic matter (NOM) onto montmorillonite was proportional to its equilibrium concentration in the range 0–25 mgC/L (Equation (1)): 
formula
1
where A is the specific adsorbed mass [mgC/g], CL is the equilibrium concentration of NOM [mgC/L] and Kd is the partition coefficient for the NOM. Kd is determined as the slope of the linear regression of a Cs vs. CL plot.
In the mixing test in this study the following equation was established according to the mass balance of organic carbon: 
formula
2
where Ai is the initial adsorbed mass (the organic contents of road dust) [mgC/g], A is the adsorbed mass after mixing [mgC/g] and M is the road dust dose [g/L]. The obtained data of 1 w/v% (e.g., RD2-2 (0.2–2 mm) were: Ai = 40 mgC/g, M = 10 g/L, and C = 1.7 mgC/L) can be substituted into Equation (2) and solved for Kd (=23.5). Assuming that Equation (1) is established in the mixing test in this study, DOC concentration at a discretionary M can be estimated by simultaneously solving Equations (1) and (2). For instance, when M is 0.1 g/L, the DOC concentration is calculated to be 1.2 mgC/L using Kd of 23.5.

The evaluated DOC concentrations for 0.2–2 mm and <0.2 mm particle size fractions at M = 0.1 g/L were 1.7 ± 0.8 mgC/L and 3.2 ± 1.9 mgC/L, respectively. The POC concentrations calculated from these results and the organic contents of road dust were 3.5 ± 1.7 mgC/L and 7.0 ± 4.2 mgC/L, respectively. These results are shown in Figure 2.

Figure 2

Concentration of TOC, POC and DOC in 0.01 w/v% road dust/water mixture (n = 7).

Figure 2

Concentration of TOC, POC and DOC in 0.01 w/v% road dust/water mixture (n = 7).

All organic carbon concentrations of the <0.2 mm samples were higher than the corresponding concentrations of the 0.2–2 mm samples except RD4, whose organic carbon content was extremely low. The average content of organic carbon in the 0.2–2 mm and <0.2 mm particle sizes in the seven road dust samples were 52 ± 22 mgC/g and 102 ± 57 mgC/g, respectively. Thus, the organic carbon content in the <0.2 mm fraction was about twice that of the 0.2–2 mm fraction. The average weight ratios of the 0.2–2 mm and <0.2 mm particle sizes in the seven road dust samples accounted for about 48% and 52%, respectively. Therefore, the calculated pollution load of the <0.2 mm particles was estimated to be more influential compared to the 0.2–2 mm particles.

Furumai et al. (2002) reported that the finest fraction (<45 μm) was mainly washed off from the road surface, although most of the road dust was coarser than 125 μm. Also, Goonetilleke et al. (2009) compared particle size distribution data for different sites and suggested that stormwater quality improvement strategies should target the 0.75–150 μm range for removal. Our results agreed with the previous ones. However, the TOC concentration in the road dust/water mixture (<0.2 mm particles) was 10.2 ± 5.7 mgC/L (n = 7), which was three times lower than that of the first flush runoff shown in Table 4 (32.2 mgC/L). This significant difference in results under a similar SS concentration is very interesting.

Vehicle wash-off water

Vehicle wash-off water which was discharged during a rainfall event was investigated for pollutants. Figure 3 shows that the resulting concentration of TOC, POC and DOC in vehicle wash-off water was 14.0 ± 5.2 mgC/L, 8.2 ± 3.0 mgC/L and 5.8 ± 2.6 mgC/L (n = 10), respectively. Thus, POC concentration was a little higher than the DOC concentration. However, the DOC concentration in the vehicle wash-off water was considerably higher than that in the road dust/water mixture. It is known that non-point sources in urban areas are related to the road, surface road dust, and vehicles. Oil and PAH pollution, especially, are caused by inadequate automotive maintenance and automobile emissions (Campbell et al. 2004). This result is worthy of special mention because the car is a self-evident source of dissolved organic substances like these.

Figure 3

Concentration of TOC, POC, DOC and SS in vehicle wash-off water (n = 10).

Figure 3

Concentration of TOC, POC, DOC and SS in vehicle wash-off water (n = 10).

Particle size distribution in first flush runoff and vehicle wash-off water

Particle size distribution is shown in Table 5, and Figure 4 is an example of a particle size distribution in first flush runoff and vehicle wash-off water. The particle size distributions on volumetric basis of both first flush runoff and vehicle wash-off water have a higher proportion of <0.2 mm particles, with each ratio accounting for 98.8 ± 1.0% (n = 6) and 97.5 ± 2.6% (n = 10), respectively. Both samples accounted for approximately 90% in the range of 0.002–0.2 mm. It was interesting that the two different pollution sources had similar particle size distribution. Moreover, the average total SS concentration in vehicle wash-off water was 88 ± 41 mg/L (n = 10). As discussed in the section ‘Road dust’, the water pollution from road dust was different for the two sizes, and the organic carbon concentration for the <0.2 mm fraction was significantly higher than that for the 0.2–2 mm one. This result approximately accorded with a particle size range related to the PAH pollutant source reported by some researchers (e.g., Pengchai et al. 2005). This suggests the possibility that organic matter could have been adsorbed onto the surface of fine road dust particles.

Table 5

Percentages of different particle size distributions in first flush runoff and vehicle wash-off water

Particle size First flush runoff Vehicle wash-off water  
<0.002 mm (Clay) 8.9 ± 1.0 5.6 ± 2.9 
0.002–0.02 mm (Silt) 41.5 ± 8.7 29.2 ± 8.3 
0.02–0.2 mm (Fine sand) 48.4 ± 9.8 62.8 ± 9.9 
0.2 mm–2 mm (Coarse sand) 1.2 ± 0.9 2.4 ± 2.5 
Particle size First flush runoff Vehicle wash-off water  
<0.002 mm (Clay) 8.9 ± 1.0 5.6 ± 2.9 
0.002–0.02 mm (Silt) 41.5 ± 8.7 29.2 ± 8.3 
0.02–0.2 mm (Fine sand) 48.4 ± 9.8 62.8 ± 9.9 
0.2 mm–2 mm (Coarse sand) 1.2 ± 0.9 2.4 ± 2.5 

Particle amount of volumetric basis %: average ± SD.

Figure 4

Particle size distributions of volumetric basis for first flush runoff and vehicle wash-off water.

Figure 4

Particle size distributions of volumetric basis for first flush runoff and vehicle wash-off water.

Comparison of concentrations in first flush runoff, vehicle wash-off water and road dust/water mixture

Figure 5 shows the relationship between the three respective concentrations of organic carbon. Roughly the same value ranges are noted, although the road dust/water mixture (dust size: <0.2 mm, 100 mg/L, n = 7) has a slightly lower organic carbon concentration than the first flush runoff and the vehicle wash-off water. This suggests that the three factors are interrelated and important. As Figure 6(a) suggests, the organic carbon content in road dust has a strong correlation with ADT (R2 = 0.68). Next, examining the two different particle sizes of the seven selected samples, as is shown in Figure 6(b), the TOC contents of the <0.2 mm fraction in the road dust samples were higher than those of the 0.2‒2 mm fraction, which further increased with the increase in the ADT. These results indicate a possibility that fine road dust particles (<0.2 mm) led to a discharge of a significant amount of organic matter pollution in the first flush runoff. Furthermore, they indicate that the organic content pollution in stormwater runoff from urban roadways is caused not only by fine particle discharge in road dust but also by vehicles running in the early stages of rainfall events. Egodawatta et al. (2007) reported that stormwater pollutants undergo build-up, wash-off and transport before accumulating in receiving waters. According to our previous study, organic matter (chemical oxygen demand and TOC) rapidly decreased at the beginning of rain but attained a constant value after some time (Wada et al. 2006). Our result indicates that the wash-off load process expressed mathematically in stormwater quality models should depend not only on roadway surface deposit and wet fallout in the rainfall but also on the number of running vehicles.

Figure 5

Total organic carbon in first flush runoff, vehicle wash-off water and 0.01 w/v% road dust/water mixture (dust size: <0.2 mm).

Figure 5

Total organic carbon in first flush runoff, vehicle wash-off water and 0.01 w/v% road dust/water mixture (dust size: <0.2 mm).

Figure 6

Correlation between TOC content (all samples (a) and two different particle size fractions: 0.2–2.0 mm and <0.2 mm (b)) in road dust and average daily traffic.

Figure 6

Correlation between TOC content (all samples (a) and two different particle size fractions: 0.2–2.0 mm and <0.2 mm (b)) in road dust and average daily traffic.

CONCLUSIONS

In this study, the influential source of organic matter in first flush runoffs from urban roadways was explored by comparing organic carbon content and particle size distribution in road dust samples with those in discharge from vehicles during rainfall. The obtained results are summarized as follows:

  1. The road dust contained a higher amount of finer particles (<0.2 mm) than larger ones (0.2–2 mm) in numbers, although the weight of both size fractions were almost the same. Since the carbon contents of fine dust was higher than those of larger ones, the fine dust was thought to contribute to organic loadings from the roadway runoffs more significantly than the larger ones. However, the TOC concentration in the road dust/water mixture (<0.2 mm) was three times lower than that of the first flush runoff under a similar SS concentration.

  2. The particle size distributions of both first flush runoff and vehicle wash-off water were similar, and particles less than 0.2 mm contributed to over 95% of total volume.

  3. The DOC concentration in the vehicle wash-off water was considerably higher than that for the road dust/water mixture. This was suggested to be caused by dissolved organic substances such as automobile fuel and engine oil.

  4. The TOC content in road dust positively correlated with annual daily traffic regardless of the particle size. Therefore, vehicles were thought to strongly influence the nature of road dust.

ACKNOWLEDGEMENTS

This research was partially supported by the Ministry of Land, Infrastructure, Transport and Tourism and Shiga prefectural government. The authors wish to thank Professor Kishimoto at Ryukoku University for his precise advice.

REFERENCES

REFERENCES
Alley
W. M.
Smith
P. E.
1981
Estimation of accumulation parameters for urban runoff quality
.
Water Resour. Res.
17
(
6
),
1657
1664
.
Björklund
K.
Cousins
A. P.
Strömvall
A.-M.
Malmqvist
P.-A.
2009
Phthalates and nonylphenols in urban runoff: occurrence, distribution and area emission factors
.
Sci. Total Environ.
407
,
4665
4672
.
Campbell
N.
D'Arcy
B.
Frost
A.
Novotny
V.
Sansom
A.
2004
Diffuse pollution, an introduction to the problems and solutions
. In:
D'Arcy
B. J.
Ellis
J. B.
Ferrier
R. C.
Jenkins
A.
Dils
R.
(eds),
Diffuse Pollution Impacts: The Environmental and Economic Impacts of Diffuse Pollution in the UK
.
Terence Dalton Publishers
,
Sudbury
,
UK
, pp.
8
12
.
Furumai
H.
Balmer
H.
Boller
M.
2002
Dynamic behavior of suspended pollutants and particle size distribution in highway runoff
.
Water Sci. Technol.
46
(
11–12
),
413
418
.
Hijioka
Y.
Furumai
H.
2001
Modified washoff models of roof and road for urban non-point pollution analysis
.
Doboku Gakkai Ronbunshu
685/VII-20
,
123
134
(in Japanese)
.
Japan Meteorological Agency Database
,
Precipitation data
. .
Japanese Industrial Standards Committee
1998
Japanese Industrial Standard Testing Methods for Industrial Wastewater
.
JIS K 0102
.
Japanese Standards Association
,
Tokyo
,
Japan
.
Japanese Ministry of Land, Infrastructure, Transport, Tourism (Japanese MLIT)
1999, 2005, 2010
Roadways traffic census
. .
Japanese Ministry of the Environment (Japanese MOE)
2014
.
Japanese Patent No. 3768186
2006
Shiga prefecture, Lake Biwa-Yodo River Water Quality Preservation Organization and Toray Engineering Co., Ltd
,
Japan platform for Patent Information, National Center for Industrial Property Information and Training, https://www.j-platpat.inpit.go.jp/web/tokujitsu/tkbs/TKBS_GM401_ToPDF.action (accessed 23 March 2015) (in Japanese)
.
Mahbub
P.
Goonetilleke
A.
Ayoko
G. A.
Egodawatta
P.
2011
Effects of climate change on the wash-off of volatile organic compounds from urban roads
.
Sci. Total Environ.
409
,
3934
3942
.
Mangani
G.
Berloni
A.
Bellucci
F.
Tatano
F.
Maione
M.
2005
Evaluation of the pollutant content in road runoff first flush waters
.
Water Air Soil Pollut.
160
,
213
228
.
Murakami
M.
Nakajima
F.
Furumai
H.
Kato
Y.
2006
Leaching characteristics of heavy metals in road dust in Tokyo
.
Jpn Soc. Water Environ.
29
(
11
),
731
735
(in Japanese)
.
Novotny
V.
2003
Water Quality: Diffuse Pollution and Watershed Management
.
2nd edn
.
Urban and Highway Diffuse Pollution
.
John Wiley & Sons
,
New York
,
Chapter 8
, pp.
374
429
.
Pengchai
P.
Nakajima
F.
Furumai
H.
2005
Estimation of origins of polycyclic aromatic hydrocarbons in size-fractionated road dust in Tokyo with multivariate analysis
.
Water Sci. Technol.
51
(
3–4
),
169
175
.
Sartor
J. D.
Boyd
G. B.
1972
Water pollution aspects of street surface contaminants
,
US Environment Protection Agency Report, EPA-R2-72-081
.
US EPA
,
Washington, DC
,
USA
.
US Environmental Protection Agency (US EPA)
1983
Results of the Nationwide Urban Runoff Program
,
Vol. I, Final report
,
Water Planning Division, US EPA
,
Washington, DC
,
USA
.
Wada
K.
Fujii
S.
2006
Characterization of stormwater runoff processes from urban roadways
.
Jpn Soc. Water Environ.
29
(
11
),
699
704
(in Japanese)
.
Wada
K.
Fujii
S.
2007
Pollution control measures of runoff from urban roadways
. In:
The 5th Seminar on Water and Wastewater Management and Technologies
,
Kyoto
,
Japan
, pp.
47
53
.
Wada
K.
Fujii
S.
2010
Estimation of pollutant loads from urban roadway runoff
.
Water Sci. Technol.
61
(
2
),
345
354
.
Wada
K.
Nishikawa
M.
Kishimoto
N.
2011
Control approach to non-biodegradable organic matter in roadway runoff
.
Water Pract. Technol.
6
(
1
),
doi:10.2166/WPT.2011.007
.