Technical Report

CHARACTERIZATION OF INORGANIC POLLUTANTS IN URBAN DRAINAGE NETWORK Zaini, Z. 1

Undergraduate student, Bachelor (Hons,) in Civil Engineering, Faculty of Civil Engineering, Universiti Teknologi MARA, Malaysia

ABSTRACT The significant changes of hydrology and the land-use resulting from the rapid development in Malaysia, leads to the increase in the magnitude of stormwater events flow in downstream area and impact on land pollution thus for the water quality. Especially in urban area, the stormwater events will result to the potential of flooding, if the existing drainage systems fail to accommodate the discharge of the stormwater. The situation occurs mainly from a type of pollutants called sediments which accumulate into the drainage system. It is a need to study about the characterization of inorganic pollutants (bed sediment) in urban drainage network with consideration of land development and season variations. The study were conducted by using particle size distribution test to determine the type of sediment, composition, mean size and sorting characteristics in dry and wet season along small drainage area at hospital construction site, Section 7, Shah Alam. From the test, the results show that for both seasons, the bed sediment classified as very fine sand, with high composition of sand and silt, their mean size about 173.7294 µm to 176.8763 µm, and the sorting characteristics as very poorly sorted. Keyword: inorganic pollutants, urban drainage, season variations, bed sediment characteristics, particle size distribution

1

INTRODUCTION

Rapid urbanization in Malaysia lead to significant changes to the hydrology and land-use with the most obvious effect being the increase in the magnitude of stormwater events flow in downstream area and impact on flooding and water quality. With the climatic condition of uniform temperature, high humidity, copious rainfall and arise mainly from the maritime exposure, the effect of land and sea breezes on the general wind flow pattern determine the rainfall distribution patterns over the country. During monsoon season, some of the places will experiences heavy rain spells. Especially in urban area, this will result to the flash flood, because the inability of existing drainage system to accommodate the discharge of stormwater (Malaysian Meteorological Department, 2008). This situation occurs as the drain becomes shallower resulting from some impurity accumulated into the system. This is called as pollutant. Gross pollutants such as litter and sediments will cause pollution and therefore reduce water quality (Lambe Littlejohn, 1989). The sediment is affected from land-use which commonly accelerates soil erosion and can greatly affect into streams which is also receive surrounding runoff. The most visually striking impacts to streams are simply a consequence of the sediments presence in the water, such, greater turbidity and decreased transparency, and in extreme cases, zones of deposition and flow diversion (Edwards, 1969). Where sediment is a pollutant, the study is conducted to determine the characteristic of the inorganic pollutants in urban drainage network which deals with the effect of the land development and with respect to the season variations. This is deal to characterize the bed sediments in term of the type, composition, mean size and sorting characteristics.

2

RESEARCH METHODOLOGY

2.1

Samples Preparation

Samples of sediments were taken in a small drainage area at hospital’s construction site at Section 7, Shah Alam, Selangor. The samples include ten (10) samples of sediment and a sample of soil was taken during the dry season, which is at 22 July 2008 where there was no raining at that week and no water in the drain. For the wet season data, another ten samples of sediment were collected on 14 October 2008, where there was a raining season at that week and with drain containing water. The purpose of soil being taken was to determine whether the soil at the uppermost stream were eroded and transformed into sediment in the rain by comparing the type of that soil and the sediment samples in the drain. The sample were collected in every 10 metres interval along the drain beginning at the uppermost stream near the soil sample was taken. The point of sample to be collected were measured by the tape so that the same point of sediment samples can be identified for both dry and wet season sampling.

2.2

Testing the Sample

The samples taken from study area then were test in the laboratory. The test includes the particle size distribution test (sieve analysis and Hydrometer analysis). 2.2.1

Sieve Analysis

Sieve analysis is a method where sample of soil is separated on sieves of different sizes. Thus, sieve analysis of the particle size distribution is defined in terms of discrete size ranges such the percentage of sample between the sizes of the sieves used (Wikipedia, 2008). The apparatus consists of different sieve size ranging from 75 mm to 0.063 mm and receivers (pan), a balance, simple dividers, drying oven, sieve brushes and mechanical sieve shaker (BBB, 2005). An illustration of instrumental is as follows (Figure 2.1):

Figure 2.1: Sieve analysis instrument The objective of the test is to determine the amount and distribution of the particle size of the sediment greater than 0.063 mm in diameter. Where the sediment samples contain finegrained particles, a wet sieving procedure was first carried out to remove these and to determine the combined clay/silt fraction percentage. A suitable sized sub-sample was first oven dried and then sieve to separate the coarsest particles (>20 mm). The mass retained will be taken (Roy Whitlow, 2004). In a wet sieving process a vessel was used that has a sieving screen as a false bottom. The vessel was filled up to a predetermined level with water which also containing a dispersing agent (sodium hexametaphosphate, 2g per litre of water) about 4 hours to facilitate the dispersion of sediment particles. Then the sub-sample about 600g was immersed into the water. Fine solids are then washed through a 63 µm screen with the addition of further water until essentially clear water passes through the screen (Figure 2.2). The passed fraction will be used for the Hydrometer test. The retained fraction was again oven-dried and the dry sieving was made for that fraction (Roy Whitlow, 2004).

Figure 2.2: Wet sieving analysis In dry sieving method the sediment was being dried into the oven basically about a day. Normally, for dry sieve test, the sieve size between 0.063 mm to 5 mm was used. The sample then is being shaking about 15 minutes to cause the soil particles to fall through the sieves until they are retained on a particular sieve. After shaking, the mass of soil retained on each sample is determined (Roy Whitlow, 2004).

2.2.2

Hydrometer Analysis

A hydrometer analysis is the process by which fine-grained soils, silts and clays, are graded. Hydrometer analysis is performed if the grain sizes are too small for sieve analysis The objective of Hydrometer analysis is to quantitatively determine the particle size distribution in an essentially cohesionless sediment, which the size smaller than 0.063 mm diameter, hence determine the silt and clay fraction percentage by difference (Roy Whitlow, 2004). The apparatus for this test are BS test sieve, riffle box, mechanical shaker and other accessories which include evaporating dish, corrosion tray, sieve brushes and sodium hexametaphosphate as dispersing agent (BBB, 2005). The illustration of instrumental setup is shown as follows (Figure 2.3):

Figure 2.3: Hydrometer test apparatus (Google, 2008) For soil samples containing fine-grained particles, a wet sieving procedure is first carried out. The sample which passed through 63 µm will be used for the Hydrometer test. The passed material in the vessel will be keep to sedimentation occurs about 4 hours. Then the settle materials were collected at oven-dried about 4 hours). Then the dry sample (about 60g) will be used for the test (Roy Whitlow, 2004). It is important to make a pre-treatment for the sample before the Hydrometer test is done. The purpose of pre-treatment is to eliminate the organic matter which may contain in the sample. The Hydrogen Peroxide (H2O2) were added about 20% concentrations of the solution volume to remove the organic matter (J.T. Germaine, 2004). The samples of sediment were added into a cylinder and the water is filled up to 1000ml scale. The mixture was stir about a minute to allow the sedimentation process. The Hydrometer and thermometer then put into the mixture solution and each of the reading will be taken every 1, 2, 4, 8, 16, 30, 60, 120, 240, 480, 1440 and 2880 minutes. Note that for the absence of the specific gravity of sample, it can be assumed as 2.65 Mg/m3 (J.T. Germaine, 2004). 2.3

Characterize the Samples

2.3.1

Particle Size Distribution Graph

Particle size distribution (PSD) graph was used to plot the curve from sieve and Hydrometer analysis. The curve was representing the percentage cumulative passing of the sediment particles for certain sieve size diameter. The PSD curve is very significance in determining the type of sediment, the compositions, the mean value and also their sorting characteristics. The compositions of the sediments can be computed by determination the different of cumulative percentage passing which coincide with the lower boundary and upper boundary of mineral class in horizontal axis (Table 2.1)

Table 2.1: Method on determination of mineral compositions (Charles Thomas Haan and Billy J. Barfieldm, 2006)

Mineral Composition (Different in Percentage passing) (%) Gravel Sand Silt Clay

Mineral upper and lower boundary (size) (mm) More than 2 0.063 to 2 0.002 to 0.063 Lower than 0.002

The information from the curve of PSD graph also can be used to determine the type of sediment presence. This was representing by computing the median diameter, d50 from the PSD curve. The median diameter of sediment, d50 is the minimum diameter of the sediment passed through 50% of finer materials. It is the particle diameter most widely adopted to represent the sediment materials. The median size is a suitable choice for lognormals distributed grain sizes, an assumption commonly employed for natural sediments, because it coincides with the geometric mean and mode sizes of such a distribution (Roy Whitlow, 2004). The d50 coefficient which coincides to the grain diameter size in horizontal axis will characterize the type of sediment has. The classification is based on table 2.2 below: Table 2.2: Grain size characteristics (Charles Thomas Haan and Billy J. Barfield, 2006)

Grain size (mm) 1.0 < d50 ≤ 2.0 0.5 < d50 ≤ 1.0 0.25 < d50 ≤ 0.5 0.10 < d50 ≤ 0.25 0.05 < d50 ≤ 0.10 0.002 < d50 ≤ 0.05 d50 ≤ 0.002 2.3.2

Mean Size of Sediment

Type of sediment Very Coarse Sand Coarse Sand Medium Sand Fine Sand Very Fine Sand Silt Clay

Mean size of sediment, Dm represent the index of grain size measurement due to its weight. The determination of mean size of sediment can be obtained directly from the PSD curve which represented by graphical method below (Figure 2.4).

Figure 2.4: Determination of mean size (Charles Thomas Haan and Billy J. Barfield, 2006)

The mean size was calculated using this formula the equation:

Where: Dm

= Mean size of sediment (mm)

∆i

= Portion of the percentages shown on the y-axis of Figure 2.4

di

= The mean value of the sizes established by the extreme values of the interval ∆i

(Charles Thomas Haan and Billy J. Barfield, 2006). 2.3.3 Sorting Characteristics of Sediment Sorting is also refered as standard deviation and indicates therange of forces which determined the sediment size distribution Sorting indicates the distribution of grain size of sediments, either in unconsolidated deposits or in sedimentary rocks. Poorly sorted indicates that the sediment sizes are mixed whereas well sorted indicates that the sediment sizes are similar. The sorting of sediment were mainly influence by the degree of mineral compositions and grain size of sediment (Wikipedia, 2008). Sorting, σ was determined by this equation:

Where dm = Mean size (Charles Thomas Haan and Billy J. Barfield, 2006). Table 2.3 below show the sorting classes for sediment: Table 2.3: Sorting classes (Charles Thomas Haan and Billy J. Barfield, 2006)

Sorting, σ (mm) σ > 0.78 0.71 < σ ≤ 0.78 0.61 < σ ≤ 0.71 0.50 < σ ≤ 0.61 0.25 < σ ≤ 0.50 0.0625 < σ ≤ 0.25 σ < 0.0625

Classification Very well sorted Well sorted Moderately well sorted Moderately sorted Poor sorted Very poorly sorted Extremely poorly sorted

3

RESULTS, ANALYSIS AND DISCUSSION

3.1

Types of Sediment Sample

The median size, d50 found from particle size distribution (PSD) cumulative curve can describes the classification of the sediment type. By referring to the Table 2.2 above, for the sediment median size between 0.05 mm to 0.10 mm, the sediment is classified as very fine sand. This condition was valid to all samples of sediment in dry season. The median sizes for the samples were ranging from 0.065 mm to 0.078 mm. It can be shown that the distribution of the sediment was same along the drain because there is no effect of water in transporting the sediment. For the wet season, the median size of sediment was ranging from 0.053 mm to 0.076 mm. It can be seen that, the median size of sediment become smaller downstream to the drain. Because of the range is in between 0.05 mm to 0.10 mm, the type of sediment also classified as very fine sand. This condition occurs because of the presence of water will transported and deposited lighter sediments relatively longer distance from upstream point. So, the sediment becomes finer and the median size becomes smaller. The results are shown in table below: Table 3.1: Type of sediment along the drain (Dry and Wet Season)

Median size, d50 (mm) Sample 1 2 3 4 5 6 7 8

Dry Season Size (mm) Type 0.066 Very Fine Sand 0.066 Very Fine Sand 0.066 Very Fine Sand 0.066 Very Fine Sand 0.068 Very Fine Sand 0.065 Very Fine Sand 0.065 Very Fine Sand 0.066 Very Fine Sand

Sample 1 2 3 4 5 6 7 8

Wet Season Size (mm) Name 0.076 Very Fine Sand 0.076 Very Fine Sand 0.068 Very Fine Sand 0.065 Very Fine Sand 0.062 Very Fine Sand 0.060 Very Fine Sand 0.059 Very Fine Sand 0.058 Very Fine Sand

9 10

3.2

0.066 0.066

Very Fine Sand Very Fine Sand

9 10

0.055 0.053

Very Fine Sand Very Fine Sand

Compositions of Sediment Sample

Figure 3.1: Mineral compositions along the drain (dry season)

The majority of mineral compositions in dry season were represented by half of the sand constituents in the samples. The range is about 50.51% to 50.91% of sediment composition. It was follows by silt constituents (34.41% to 34.80%) and clay constituent (14.49% to 14.77%). The percentage of fine minerals was almost same with coarse materials in a range 49.09% to 49.46%. The percentage of coarse minerals was shown about 50.54% to 50.91% (Figure 3.1). The mineral compositions of the sediment samples were likely same along the drain. This is because of there are no effect of water in transporting the mineral constituents of sediment to the downstream area. As the result, there was no change in pattern to the percentage of mineral compositions along the drain.

Figure 3.2: Mineral compositions along the drain (wet season)

In wet season, the majority of mineral compositions in sediment were represented by high percentage of sand about 44.02% to 53.93%. It was follows by silt constituents (38.16% to 42.55%) and clay constituents (7.92% to 13.43%). The percentage of fine minerals were varies from 46.08% to 55.98% and coarse minerals about 44.02% to 53.93% (Figure 3.2). The finer mineral compositions (clay and silt) of sediment samples were increase with respect to distance. This is due to the effect of water presence in the drain was transported lighter minerals (clay and silt) and deposited them in downstream area. Lighter minerals will transporting in longer distance and hardly to deposited due to high retention time to settle. Heavier minerals such sand will transported in shorter distance and deposited easily because of low retention time to settle. However, the sand compositions were decrease with respect to the distance.

3.3

Mean Size of Sediment Samples

It can be shown that the mean diameter for dry season sediment samples were in the range of 169.8917 µm to 185.7565 µm (Table 3.2) whereas for the wet season, the mean size is in between 162.7314 µm to 189.6941 µm (Table 3.3). All of the mean size is very small and the average size of sediment lies in the range of fine sand type. The mean size of sediment along the drain in both seasons was fluctuated increase and decrease due to the compositions size of the mineral in the sediments. It is essential to compute the average mean size of the sediment in the drain. For dry season, the average mean size was about 176.8763 µm but for the wet season was 173.7294 µm, both are fine sand type.

Table 3.2: Mean size of sediment samples (dry season)

Sample

Distance 0

Dm (mm) 0.1699

Dm (µm) 169.8917

1 2

10

0.1826

182.6454

3

20

0.1722

172.1893

4

30

0.1738

173.7905

5

40

0.1858

185.7565

6

50

0.1737

173.7116

7

60

0.1735

173.5487

8

70

0.1776

177.6420

9

80

0.1831

183.1240

10

90

0.1765

176.4637

0.1769

176.8763

Average

Table 3.3: Mean size of sediment samples (wet season)

Sample 1

Distance 0

Dm (mm) 0.1897

Dm (µm) 189.6941

2

10

0.1722

172.1936

3

20

0.1816

181.5808

4

30

0.1742

174.1735

5

40

0.1730

173.0147

6

50

0.1627

162.7314

7

60

0.1701

170.0503

8

70

0.1733

173.3158

9

80

0.1708

170.7760

10

90

0.1698

169.7638

0.1737

173.7294

Average 3.4

Standard Deviation of Sediment Samples

The sorting classification can be found by referring to the Table 2.3. For dry seasons, the samples were very poorly sorted because the range of sorting values lies between 0.0637 mm to

0.0975 mm (Table 3.4), whereas for wet season all the samples except 9th sample were classified as very poorly sorted poor because it lies in range of 0.0666 mm to 0.0972 mm (Table 3.5). The 9th sample was classified as extremely poor sorted because the sorting value was zero. The class of very poorly sorted range in between 0.0625 mm to 0.25 mm. The very poorly sorted and extremely poor sorted indicates that the sediment sizes were mixed which is influence by the degree of mineral compositions and grain size in the sediment samples. Only little grain were transported and deposited by water each time to the downstream part of the drain.

Table 3.4: Sorting characteristics of sediment samples (dry season)

Distance 0 10 20 30 40 50 60 70 80 90

Sample 1 2 3 4 5 6 7 8 9 10

Sorting (mm) 0.0864 0.0928 0.0637 0.0894 0.0923 0.0799 0.0921 0.0937 0.0975 0.0911

Classification Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted

Table 3.5: Sorting characteristics for sediment samples (wet season)

Distance 0 10 20 30 40 50 60 70

Sample 1 2 3 4 5 6 7 8

Sorting (mm) 0.0972 0.0666 0.0940 0.0906 0.0768 0.0754 0.0891 0.0877

80

9

0.0000

90

10

0.0964

Classification Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Very poorly sorted Extremely poorly sorted Very poorly sorted

4.0

RECOMMENDATION AND CONCLUSION

4.1

Recommendation

The particle size distribution only characterizes the physical characteristic of inorganic pollutants. However, this does not determine the chemical characteristic such whether the samples contain metals such as zinc, plumbum, phosphorus and so on. However, the physical characteristic of the sediment can be used as the indicator for the determination of the composition of materials which contributing to the turbidity of the water, as well as, factors which degrade the quality of the water. It is recommended that for the future a test will be carried out to determine the chemical characteristics of inorganic pollutants in a drainage area. The significant of the study is that, the presence of some chemical constituents can contribute to the formation of eutrophication in the water. The eutrophication will result to the increase in the ecosystem's primary productivity (excessive plant growth and decay), and further effects including lack of oxygen and severe reductions in water quality, fish, and other animal populations.

4.2

Conclusion

From the test, the results show that for both seasons, the bed sediment classified as very fine sand, with high composition of sand and silt, their mean size about 173.7294 µm to 176.8763 µm, and the sorting characteristics as very poorly sorted. So, all the objectives were achieved.

REFERENCES Malaysian Meteorological Department (2007), Climate of Malaysia. http://www.met.gov.my/english/education/climate/climate01-02.html., Ministry of Science, Technology and Innovation (MOSTI), 17th February 2008. Littlejohn, L. (1989), Earth's water: Rivers and sediment, http://ga.water.usgs.gov/edu/earthriverssed.html, USGS SCIENCE, 16th March 2008. Edwards (1969), Introduction to Environmental Quality, Soils and Environmental Quality, 3rd edition. Wikipedia, 2008, Particles size distribution, http://en.wikipedia.org/wiki/Particles_size_distribution.htm, Wikimedia Foundation, Inc, 13th March 2008. BBB (2005), Manual Laboratory, Soil Mechanics Laboratory, Universiti Teknologi MARA. Wikipedia, 2008, Hydrometer analysis, http://en.wikipedia.org/wiki/Hydrometer_analysis.htm, Wikimedia Foundation, Inc., 11th March 2008. Hydrometer, http://images.google.com, Google Inc., 2nd April, 2008.

Whitlow, R. (2004), Basic Soil Mechanics, Pearson Prentice Hall South Asia Pte Ltd., 4th edition. Germaine, J.T. (2004), Hydrometer test, http://www.nexternal.com/shared/affiliates/?CS=hbm&Affiliate=1&Target=http%3A%2F%2Fw ww.nexternal.com%2Fhbm%2FProduct29, 02nd February 2008. Wikipedia (2008), Sorting_sediment, http://en.wikipedia.org/wiki/sorting_sediment.htm, Wikimedia Foundation Inc., 02st November 2008 Haan, C.T., and Barfield, B.J. (2006), Design Hydrology and Sedimentology for Small Catchments, Elsevier.