5.2.4 Metal build-up on road surfaces
The metal build-up on road surfaces were determined based on the method described in Section 4.3.7. The metal build-up loads are represented in the form of mg per m2 of road surface and the concentrations of the metals are given in mg per g of solids build-up on the road surface.
Variability of metal build-up among different suburbs
A preliminary analysis was conducted to understand the behaviour of metal build-up on roads in different suburbs regardless of antecedent dry periods and the results are shown in Table 5-5. In Table 5-5, it can be seen that most dominant metals on road surfaces are Na, Mg, Al, K, Ca and Fe. It is commonly known that these metals are found in soil (Mostert, et al., 2010). Therefore, it is not unusual to find these metals in abundance as a major fraction of road build-up and could be originating from soil
from adjacent land (Gunawardana, 2011). However, as shown in Table 5-5, metal loading was different from one suburb to another. In particular, build-up in Surfers Paradise sites represents comparatively high average metal loading for Na, Mg, Al, K, Ca and Fe. As shown in Figure 4-1, Chapter 4, sites in Surfers Paradise are located close to (within 1.5 km distance of) the coast line. Therefore, there is a possibility of the presence of sea salts and associated metals such as Na, Ca and K in road build-up.
Table 5-5 Average metal loading (mg/m2) in build-up in different suburbs
Surfers Paradise Nerang Benowa Clearview Estate Average metal loading (± Standard deviation) (mg/m2)
Li 0.0315 (± 0.0527) 0.0227 (± 0.0256) 0.0217 (± 0.0369) 0.0032 (± 0.0016) Na 23.7302 (± 28.8214) 7.3834 (± 5.5097) 9.1824 (± 7.214) 2.8306 (± 1.592) Mg 21.8264 (± 26.115) 13.2217 (± 11.5752) 15.9998 (± 24.3208) 2.8147 (± 1.0645) Al 39.4816 (± 45.2724) 31.6925 (± 27.5795) 32.2154 (± 45.031) 7.2252 (± 2.3032) K 14.0185 (± 11.6247) 9.8949 (± 5.6674) 9.0631 (± 5.9479) 3.3337 (± 1.9585) Ca 118.0371 (± 144.8897) 69.8325 (± 46.4019) 59.9933 (± 71.0154) 11.4291 (± 4.4693) Ti 0.9805 (± 0.8089) 1.7165 (± 1.6249) 1.2843 (± 1.9494) 0.2242 (± 0.0736) V 0.0878 (± 0.1056) 0.0585 (± 0.0472) 0.0754 (± 0.1128) 0.0161 (± 0.0057) Cr 0.0861 (± 0.1136) 0.081 (± 0.07) 0.0906 (± 0.1482) 0.0157 (± 0.0046) Mn 1.6841 (± 1.9787) 1.3764 (± 1.2002) 1.2544 (± 1.8029) 0.3557 (± 0.1364) Fe 65.8006 (± 81.2706) 63.6777 (± 59.377) 63.972 (± 103.7979) 16.4375 (± 11.5503) Co 0.0537 (± 0.0761) 0.0306 (± 0.0272) 0.0328 (± 0.0464) 0.006 (± 0.0025) Ni 0.0852 (± 0.12) 0.0814 (± 0.0711) 0.0983 (± 0.1611) 0.0113 (± 0.0041) Cu 1.6882 (± 1.5818) 1.2029 (± 0.6565) 1.8295 (± 2.4982) 0.3208 (± 0.1443) Zn 4.6823 (± 5.5381) 4.2899 (± 3.1287) 5.446 (± 8.8373) 0.5498 (± 0.1513) Mo 0.0089 (± 0.0113) 0.009 (± 0.0028) 0.0099 (± 0.0154) 0.0022 (± 0.0019) Rh 0.0011 (± 0.0002) 0.0012 (± 0.0003) 0.0009 (± 0.0002) 0.0013 (± 0.0003) Pd 0.0054 (± 0.0009) 0.0061 (± 0.0013) 0.0047 (± 0.0009) 0.0063 (± 0.0013) Cd 0.0041 (± 0.0048) 0.0045 (± 0.0048) 0.0042 (± 0.0034) 0.0016 (± 0.0001) Sn 0.0249 (± 0.0129) 0.0186 (± 0.015) 0.0244 (± 0.029) 0.0078 (± 0.0051) Sb 0.0198 (± 0.0267) 0.0117 (± 0.0057) 0.0217 (± 0.0313) 0.0038 (± 0.001) Ba 1.0401 (± 1.2282) 0.9052 (± 0.7212) 1.2536 (± 2.1478) 0.4127 (± 0.7407) Pt 0.0011 (± 0.0002) 0.0012 (± 0.0003) 0.0009 (± 0.0002) 0.0013 (± 0.0003) Pb 0.689 (± 1.0989) 0.3761 (± 0.3069) 0.451 (± 0.4873) 0.132 (± 0.2189)
Moderate levels of loading can be seen for metals analysed across the sampling sites except for Na, Mg, Al, K, Ca and Fe (See Table 5-5). The highest average metal loading of Ti, V, Cr, Co, Ni, Cu, Zn, Mo, Cd, Sn, Sb, Ba and Pb were observed in Surfers Paradise, Benowa and Nerang compared to the Clearview Estate residential sites. It can be argued that the presence of comparatively high solids load on road surfaces in Surfers Paradise, Benowa and Nerang (Section 5.2.1) would result in high loading of metals. In addition, the amount of metal released into the environment can be higher in Surfers Paradise, Benowa and Nerang suburbs than Clearview Estate due to land use specific activities (high traffic, industrial and commercial activities). It could be reflected in the measured metal concentrations (mg/g) if the rate of release of metals is different between suburbs.
To evaluate this hypothesis, the metal concentrations (mg/g) were examined. The data is given in Table C1, Appendix C. It can be seen that among Ti, V, Cr, Co, Ni, Cu, Zn, Mo, Cd, Sn, Sb, Ba and Pb, the average concentrations are higher in Surfers Paradise, Benowa and Nerang compared to Clearview Estate except for Pb and Mo. These metals are mainly released to the environment via traffic related sources (Adachi and Tainosho, 2004) and industrial activities (in Nerang). Therefore, high concentrations of these metals can be expected as there are high vehicular traffic activities and heavy duty vehicular traffic activities in sites within Surfers Paradise, Benowa and Nerang.
As evident in Table 5-5, the lowest metal loading was reported for Rh, Pd and Pt in all suburbs. Rh, Pd and Pt are generally products released from catalytic converters of vehicles (Palacios et al., 2000). Palacios, et al. (2000) have noted that the release of Rh, Pd and Pt depends highly on the type and age of the catalytic converters. They further noted that the variability in the concentrations of Rh and Pd released from fresh catalytic converters is higher than that from aged products. The presence of these metals is justifiable as vehicular activities are common in all suburbs.
Statistical significance of the above observations needed to be investigated to test the influential role of land use specific activities such as industrial and commercial activities on metal accumulation. In this regard, One-way ANOVA was conducted for the metal data set (mg/m2) same as in Section 5.2.1.The One-way ANOVA was
carried out for individual metal elements separately. This was to test the statistical significance of the data set observed in different suburbs.
Similar to the analysis undertaken for the total solids data set, key assumptions relating to ANOVA were tested prior to the analysis. Outliers were identified in each of the data sets. However, the sensitivity to the outliers to the final one-way ANOVA results was negligible. The normality of the data sets was disregarded as the sample size in different groups (suburbs) are similar (Lix, et al., 1996). The data sets were also examined for homogeneity of variances. The elements such as Al, Cr, Mn, Fe, Cu, Cd, Sn and Ba satisfied the assumption of homogeneity of variances out of 24 metals. For the cases where the homogeneity of variances were violated, Welch ANOVA test results were used instead of standard One-way ANOVA test results (Laerd, 2013). Welch ANOVA is the modified version of the standard One-way ANOVA. Similar to the standard One-way ANOVA, p<0.05 in Welch ANOVA is considered as statistically significant and hence there is a difference in groups means (Laerd, 2013).
Test results of the standard one-way ANOVA and Welch ANOVA are given in Table C2, Appendix C. As evident in Table C2, Appendix C, significance level, p is greater than 0.05 for metals Li, Al, V, Cr, Mn, Fe, Cu, Cd, Sn, Ba and Pb indicating that there were no significant differences in metal loads among different suburbs. In contrast, loading of Na, Mg, K, Ca, Ti, Co, Ni, Zn, Mo, Rh, Pd, Sb and Pt indicates significant difference among suburbs. As noted above, there is a possibility of the influence of sea salt on metals build-up for Na, Mg, K and Ca as these metals were significantly different among suburbs. Therefore, distance from the coastline could be a factor that influences the metal loadings on roads. Similarly, the differences in concentrations of other metals among different suburbs could be due to the influence of contributing sources of pollutants. However, it is difficult to derive firm conclusions prior to identifying exact sources of metals.
Moreover, this study’s results were compared with two recent studies conducted in the same region (Gold Coast, Australia) as shown in Table 5-6. As evident in Table 5-6, loading of Cr, Ni and Cd on road surfaces have similar loading ranges as Gunawardena (2012)and Gunawardana (2011), whereas elevated loading of Mn, Cu,
Zn and Pb was reported in this study. Dissimilarities can be attributed to environmental conditions at the sampling time and differences in antecedent dry days prior to sample collection in different studies. The other elements analysed in this study were not reported in the previous studies.
Table 5-6 Comparison of metal loading with past studies
(2011) This study
Min Max Min Max Min Max
(mg/m2) (mg/m2) (mg/m2) (mg/m2) (mg/m2) (mg/m2) Cr <0.001 0.195 0.394 0.478 0.008 0.447 Mn 0.035 0.482 0.037 3.661 0.142 6.144 Ni <0.001 0.491 0.001 0.32 0.005 0.470 Cu 0.75 2.448 0.024 1.719 0.169 7.871 Zn 1.066 3.513 0.049 8.436 0.355 26.765 Cd <0.001 0.003 <0.001 0.021 0.001 0.015 Pb <0.001 0.195 <0.001 1.912 0.033 3.231
Variability of metal build-up with antecedent dry periods
Preliminary analysis was extended to understand the behaviour of metal build-up with respect to antecedent dry periods. Particularly, this was to investigate the difference between metal build-up and solids build-up with varying antecedent dry periods. For this purpose, the total metal loading in each road site for the two different dry periods were calculated separately. The ratio between the total metal loading relating to the antecedent dry period more than seven days and the fewer than seven days were then determined as shown in Eq. 5.1. The ratio presented in Eq. 5.1 is biased towards the metal elements present in abundance. However, it provides an overall assessment of the variability of metal loading between sites without undertaking extensive comparison based on individual metal elements.
𝑀𝑒𝑡𝑎𝑙 𝑟𝑎𝑡𝑖𝑜 = ∑ 𝐿𝑜𝑎𝑑𝑖𝑛𝑔 𝑜𝑓 𝑒𝑎𝑐ℎ 𝑚𝑒𝑡𝑎𝑙 𝑎𝑡 𝐴𝐷𝐷𝑠>7𝑑𝑎𝑦𝑠(
∑ 𝐿𝑜𝑎𝑑𝑖𝑛𝑔 𝑜𝑓 𝑒𝑎𝑐ℎ 𝑚𝑒𝑡𝑎𝑙 𝑎𝑡 𝐴𝐷𝐷𝑠 <7𝑑𝑎𝑦𝑠 (𝑚𝑔𝑚2) Eq. 5.1
The results are shown in Figure 5-3 with the solids build-up ratio related to antecedent dry period more than seven days and fewer than seven days. As evident in Figure 5-3, total metal loading ratio follows the solids build-up ratio except in I3
(Patrick Road) and R2 (Merloo Drive). Ball, et al. (1998) mathematically represented solids build-up on road surfaces by using a power function of antecedent dry days. Metal build-up process was also represented by the power function of antecedent dry days by Egodawatta, et al. (2013). This indicates that solids build-up and metal build-up follows the same behaviour with antecedent dry days, indicating consistency with this study results.
In C2 and C3 roads in Surfers Paradise there was a slight increment of metal loading ratio than the solids build-up ratio, which is attributed to the increase in concentrations (mg of metal in a g of build-up) with the antecedent dry days. The increase in concentration could be attributed to the increase in the release of metals to the environment via the different sources. In contrast, there are some sites, specially, I3, R1, R2 and R3, which show high solids build-up ratio with comparatively low metal loading ratio. Even though the solids were accumulated, the associated metals were less in quantity on these roads. This could be possible due to the deposition of solids of organic nature particularly at residential sites at Clearview Estate.
Figure 5-3 Comparison of metal and solids build-up on road surfaces based on antecedent dry periods
Variability of metal build-up in different particle size ranges
The analysis was further extended to observe the behaviour of metal concentrations in different particle size fractions. This was to investigate the relative importance of different particle size fractions in accumulating metals and to examine the consistency of study results with past studies. The laboratory test results obtained for the metal concentrations (mg/g) in five different particle size fractions, namely >425
0 1 2 3 4 5 6 C1 C2 C3 C4 I1 I2 I3 I4 M1 M2 M3 M4 R1 R2 R3 R4 Lo ad in g ra tio Road site Metal ratio Solid ratio
µm, 300-425 µm, 150-300 µm, 75-150 µm and <75 µm, are given in Table C3, Appendix C. The results revealed that the metal concentrations (except Ca and Pb) in the finest particle size fraction (<75 µm) were high compared to the other fractions, irrespective to the land use type. This observation is justifiable as the particles in this size fraction have high adsorption capacity for metals due to high surface to mass ratio and metals released from land use related activities can also be in small size fractions. The results were also in good agreement with past studies conducted in the same area of research confirming the consistency of the results of this study (Bi, et al., 2013; Gunawardana, 2011; Herngren, et al., 2006).
The percentage of total average metal concentrations related to the different particle size fractions in each land use are shown in Figure 5-4. As seen in Figure 5-4, 35- 50%, 14-20%, 13-19%, 6-11% and 12-25% of total metals were observed in <75 µm, 75-150 µm, 150-300 µm, 300-425 µm and >425 µm particle fractions respectively. Figure 5-4 again confirms that the highest percentage of metals is adhered to the finest fraction. However, it should be noted that the coarser fraction also bears a significant percentage of metals (11-24%) in the road build-up, especially, in Surfers Paradise and Benowa sites. There was a great influence from the presence of a high level of Ca in these two suburbs in the coarser fraction (Table C3, Appendix C) to have a significantly high percentage of total metal loading. As noted before, the presence of high level of Ca in Surfers Paradise and Benowa sites was possible because the sites are close to the coast line. In past studies, the importance of the coarser particle size fraction has not been generally highlighted as the trace metal elements are commonly the point of interest.
Figure 5-4 Percentage of total average metal concentrations in different particle size fractions in different suburbs