A study on the soil characteristic and properties of riverbank soil samples from Sungai Perak, Kota Lama Kiri, Kuala Kangsar, Malaysia

Riverbank filtration (RBF) technology is new in Malaysia and only a few efforts have been made to understand the RBF mechanisms and processes. Soil characteristics and properties play important roles in determining the suitability of the site for the RBF application. A research has been carried out in Kota Lama Kiri, Kuala Kangsar, Perak, Malaysia to identify the characteristics of the riverbank soil for different layers of the pumping well (PW) and three adjacent monitoring wells namely MW2, MW3, and MW5. Particle size distributions and hydraulic conductivities of the soils were obtained from sieve analyses and constant head permeability tests. The subsurface soils of the study site consisted of medium sand, fine sand, coarse sand and gravel but them medium sand was highest in percentage over the other types of soil. The aquifer extended down to 8 m. The highest hydraulic conductivity value for the PW was 0.91 cm/s and obtained for sample taken from 6 m deep. The highest hydraulic conductivity value for the monitoring wells was 5.03 cm/s and obtained for sample taken from 2.20 to 3.20 m of MW5. The overall well production capacity determined from the pumping test was 112.10 m3/hr.


Introduction
RBF is a natural first-step process of water treatment. The technology is commonly employed in alluvial aquifers with complex hydrologic system that exhibits both physical and geochemical heterogeneity. Alluvial aquifers that are hydraulically connected to surface water bodies have been used for water production throughout the world. Sites have been selected because of the significant amount of groundwater that can be extracted at shallow depths [1]. River water flowing through the riverbed on its way to the underlying aquifer and subsequently to the production wells can undergo RBF. Riverbed and aquifer material act as filter media to the contaminants [2]. During the process, various contaminants such as trace organic pollutants, bacteria, viruses, and inorganic compounds can be removed [3].
The design of a RBF system usually requires detailed hydrogeological site investigation and knowledge about the hydrological characteristics of the catchment. The hydrological characteristics such as hydraulic conductivity, transmissivity, storage coefficient, and dispersivity affecting groundwater flow and contaminant transport can be characterized by various laboratory and field tests such as column tracer tests, grain size analyses, pumping tests, slug tests, and field tracer tests [4]. Alluvial aquifers near a river are typically multilayered due to the deposition processes resulting in vertical heterogeneities in hydraulic properties [5,6,7,8]. A suitable site for a RBF would involve sand and gravel aquifers with hydraulic conductivity, k f > 10 -4 m/s, a minimal thickness of 5 m and a good hydraulic connection to the adjacent surface water [9]. Studies on the characteristics of a feasible bank filtration site based on examples from Europe and North America would mention the relevance aquifer thickness, hydraulic conductivity, flow velocity gradient, river discharge, and clogging of the riverbed [10,11,12]. The thicknesses of some existing RBF sites in Germany range from 3-5 m at Böckingen, by the Necker River to 40-55 m at Torgau-Ost, by the Elbe River, with hydraulic conductivity ranging between 1 x 10 -2 and 7.5x10 -5 m/s [10].
The RBF technology is very new in Malaysia. The characteristics of the alluvial soils need to be identified in determining the suitability of a site. The aim of this study was to describe the characteristics of the alluvial soils of the selected site in Perak and evaluate the suitability of the site for RBF application. This paper focuses on soil classification, particle size distribution (PSD), and hydraulic conductivity values.

Study area
The study area is located in Kg Kota Lama Kiri, Kuala Kangsar, Perak. Kuala Kangsar site was selected for RBF application after considering the soil profile near the Sungai Perak. Figure1 shows the location of the study site at Kota Lama Kiri, Kuala Kangsar, Perak, Malaysia. As shown in Figure  1, the distance from the MW1 to the river was 10 meter. The distance between MW1 and MW2 was 8 meter, where the distances were applied similar to all of the monitoring wells, except for MW5 and MW6. The distance of MW5 from the PW was 25 meter, and 50 meter for the MW6, respectively. This paper will focus on the investigation of the soil samples from pumping well (PW) and the adjacent monitoring wells including MW2, MW3 and MW5.

Soil sampling and soil investigation
Soil samples were collected during the construction and development of the PW and MW's in Kuala Kangsar, Perak. The coordinate of the PW is 04°48'08.5''N and 100°57'06.9''E. The samples were collected for each meter depth as the drill penetrates deep into the ground. The soils sample in each layer were collected and labeled. The maximum depth was 8.20 meters. Figure 2 shows the borelog and schematic well design of PW. The soil samples were transported to the Geotechnical Engineering Laboratory in School of Civil Engineering, Universiti Sains Malaysia (USM) for the soil characteristics investigation.

Sieve analyses
The sieve analysis was performed with reference to British Standard 1377: Part 2:1990. Sieve analyses were conducted by using a mesh wire woven sieve size of 14.00 mm at the top and 0.063 mm at the bottom. Soil samples were oven dried at 105 °C for 24 hours. 100g of soil samples were washed through the sieve size 0.063 mm. The soil sample must be washed carefully to avoid losing the soil material. Washed sample will be used for the mechanical sieving and the washed samples were oven dried at 105 °C for 24 hours. The dry samples were weighed again and were put on the top sieve having the largest opening which is 14.00 mm. The mechanical shaker was set to operate for 10 to 15 minutes. The soils retained on top of each sieve were weighed using the balance readable to 0.01g.

Permeability (Constant Head Test)
The constant head test was done in accordance with British Standard 1377: Part 1-9:1990. Permeability of a soil is a measure of its capacity to allow the flow of the water through the pore spaces between solid particles. The degree of permeability is determined by applying a hydraulic pressure gradient in a sample of saturated soil and measuring the flow rate of the water. The ancillary apparatus were prepared in the Geotechnical Laboratory. The internal diameter the of the permeameter cell were measured and recorded. The length and the weight of the specimen were also measured. The permeameter were placed at the test equipment and the water valves were opened. The water was allowed to flow for several minutes to let all of air out from the permeameter. A measuring cylinder was placed under the outlet of permeameter and the timer was started simultaneously. The levels of the water in the manometer tubes which is H 1 and H 2 were recorded. The quantity of water collected in the cylinder was measured. The temperature of the water was recorded.

Pumping test
The pumping test was carried using a DT 95-10 Dynatech Submersible Pump. A Submersible Pump was installed using a crane and 125 mm diameter GI riser pipe was connected to the submersible pump. A 125 mm diameter gate valve was connected to the riser pipe to regulate the flow rate. The riser pipe was then directed to a 90 V Notch tank to measure the flow rate of pumping. After the installation of the pump and all other necessary setup was completed, a calibration test was carried out for 2 hours to determine the capacity of the pumping wells, and also to determine the pumping rates for the Step Drawdown Test. The pumping test program consists of step drawdown and a 72 hours constant discharge test and recovery test.

Sieve analyses for soil samples
Grain size analyses were carried out in the Geotechnical Engineering Laboratory, School of Civil Engineering, Universiti Sains Malaysia (USM). From particle size distribution (PSD), the types and class of the soil can be determined. Gravel, coarse sand, medium sand, fine sand, silt and clay can be found within soil sample. D 10 represents the 10% of the particles are finer and 90% of the particles are coarser than that particular size of D 10 . D 30 means 30% of the particles are finer and 70% of the particles are coarser than that particular size of D 30 . The uniformity coefficient, termed as C u , is the ratio of D 60 to D 10 . C c is another measurement of gradation, which is coefficient of gradation. Table 1 shows the classification of the soil samples within PW, MW2, MW3 and MW5 at different depth. Table 2 shows the result of sieve analyses for PW, MW2, MW3 and MW5.  Table 2 indicates that the soils from PW had D 10 ranging from 0.15 to 0.38 mm, D 30 ranging from 0.40 to 0.75 mm, and D 60 ranging from 0.60 to 1.50 mm. For MW2, MW3, and MW5, the D 10, D 30, D 60 had ranged from 0.30 to 4.00 mm, 0.45 to 11.00 mm, and 0.60 to 15.00 mm, respectively. The coefficient of uniformity, C u , for PW, MW2, MW3 and MW5 ranged from 2.0 to 14.0 while the coefficient of gradation, C c , ranged from 0.05 to 2.75. The grain size distributions of soils from PW, MW2, MW3, and MW5 are given in Figures 3, 4, 5, and 6.
The grain size distributions for PW indicate that down to a depth of 8 m, the soils have a high percentage of medium sand, fine sand, coarse sand and gravel; and more gravel beyond the 8 m depth, thus making it suitable for the purpose of a pumping well. Nevertheless, the main aquifer was considered to reside above the 8 m mark due to the general presence of bedrocks at 8 m. The soils of MW2, MW3 and MW5 were predominantly medium and fine sands from surface down to 4.8 m. There was no silt or clay found in the soil samples of all boreholes.     Table 3 shows the results of permeability of the soil samples.  The table indicates that the hydraulic conductivity, k, of soil samples from the PW to range between 0.10 and 0.91 cm/s. The values of k for MW2, MW3 and MW5 ranged between 0.38 and 5.03 cm/s or between 3.8×10 −3 and 5.03×10 −2 m/s. The literature indicates that an alluvial aquifer near a river such as the Rhine River often have a hydraulic conductivity of 10 −3 to 10 −2 m/s [13,14]. In previous RBF application, the hydraulic conductivity, k, of Illinois River, USA had ranged between 2×10 -3 and 3×10 -3 m/s. The Ohio River in Louisville, Kentucky, USA had a k value of 6×10 -4 m/s, while the value for Great Miami River in Cincinnati, USA ranged between 8.8×10 -4 and 1.5×10 -3 m/s [15].
These result shows that hydraulic conductivity at field site and k obtained from constant head permeability test were quite similar with the natural materials from Sungai Perak, Kota Lama Kiri, Kuala Kangsar, Perak. Based on the screen depth between 5.00 to 8.00 metre witihin PW, soil samples hydraulic conductivity of 0.26 cm/s, 0.91 cm/s, 0.10 cm/s for depth 5.00 to 6.00 metre, 6.00 to 7.00 metre, 7.00 to 8.00 metre, respectively, and the result of pumping test 112.10 m 3 /h, it can be inferred that the average value of hydraulic conductivity of 0.42 cm/s, has the possibility to yield the abstraction of 112.10 m 3 /h.

Pumping test
Pumping test was conducted from 27 September 2013 to 30 September 2013 for 72 hours. Constant discharge rate for PW was 112.10 m 3 /h. The recovery percentage for PW was 84.67% for duration time of 180 minutes. The static water level within PW before conducting the pumping test was 2.60 metre. The measuring drawdown during pumping test was 0.79 metre.

Conclusion
A typical aquifer of a riverbank alluvial deposit considered for RBF application was investigated for the characteristics of its soil samples. The grain size analysis of PW, MW2, MW3 and MW5 showed that soils consisted of medium sand, fine sand, coarse sand and gravel with the medium sand having the highest percentage. The percentages of medium sands at various depths of PW, MW2, MW3, and MW5 were 45.12 to 77.90, 30.16 to 59.35, 0.70 to 54.51, and 0.32 to 81.34 respectively. The aquifer extended down to 8 meter deep. The maximum hydraulic conductivity for the PW was 0.91 cm/s at 6 m deep. The highest permeability, k was 5.03 cm/s, which is for the depth of 2.20 to 3.20 m of MW5. The constant discharge pumping rate of 112.10 m 3 /h verified the suitability of the site for RBF application.