Abstract
Groundwater has become an important source of water to meet the increasing requirement for domestic, industrial, and agricultural needs. The study was formulated to assess the availability of groundwater resources in Selangor State, Malaysia. An integrated MIKE SHE and MIKE 11-based surface water–groundwater interaction model was developed to fulfill the study objectives. Model results illustrated that groundwater flows from the eastern part of the hilly region to the western part and ultimately discharges into the sea. The study also revealed that Sungai Langat, Sungai Klang, and Sungai Selangor receive water from groundwater and do not feed the groundwater. Approximate potential groundwater resources based on water balance study varies from 860–960 million liter per day (MLD), 630–690 MLD, 810–870 MLD to 630–690 MLD for Langat basin, Klang basin, Selangor basin, and Bernam basin, respectively. Impacts due to extreme dry condition showed that groundwater table declined about 0.5–3 m in most of the area. However, groundwater level for the major portion of Selangor State will drop to about 0–1 m for increased abstraction due to changed land-use pattern. Study findings may be used for a holistic and comprehensive management of groundwater resources for sustainable development in Selangor State, Malaysia.
INTRODUCTION
Most major cities in Malaysia face problems of potable water supply due to the population boom, industrialization, pollution of surface water, and drought condition. The demand for clean water supply is already a major issue in certain places in Malaysia, particularly in Selangor, Kuala Lumpur, Johor Bahru, and Pulau Pinang (Saimy & Raji 2015). During previous dry spells in Malaysia, especially in Melaka, Selangor, and Sarawak, groundwater has proven to be a life saver (Mohamed et al. 2009).
Systematic basin studies have not been undertaken by the Jabatan Mineral dan Geosains Malaysia (JMG) in Selangor State, Malaysia (JPZ Sdn Bhd and IWM (M) Sdn Bhd 2015). In Selangor State, for the year 2008, there were more than 300 active tubewells recorded, most of which are in industrial areas, schools, mosques, and for rural area supply for drinking water from domestic wells. The extraction of fresh groundwater in Selangor in 2008 was estimated to be approximately 10.8 billion liters, which is a significant proportion of the total water supply in Selangor (Ranhill Consulting Sdn Bhd 2011). A number of factories in Selangor use groundwater as their water supply. This is based on the number of licenses issued to them by Lembaga Urus Air Selangor (LUAS). Examples of use include industries, hydroponics, fish farming, landscaping and, to a certain extent, for the cooling of industrial plants and for domestic supply.
Potable water is a major component for the survival of humans so it should not only be dependent on the surface water sources to meet this requirement but also for alternative water sources. Moreover, surface water resources are growing limited due to various factors, such as deteriorated water quality, exposure to pollution, and so on. Therefore, a detailed study on groundwater resources is essential to ensure that the raw water reserves of Selangor State, Malaysia are sufficient to meet the growing demand of potable water supply. The study was formulated with a view to conduct a comprehensive and detailed evaluation of the availability of groundwater resources for sustainable water management in the Selangor State. The study also provided information for better understanding of the groundwater flow system in the state using accurate estimation of hydrogeological parameters. The study was conceived with a view to assessing the water resources by an integrated way using the mathematical modeling techniques. Over the last two decades, mathematical modeling has become an increasingly important tool for analyzing and understanding groundwater systems (Saatsaz et al. 2011). For the present study, mathematical modeling techniques have been used to establish a scientific management tool for overcoming the present and future difficulties in water supply for Selangor State, Malaysia.
STUDY AREA
Selangor is the most developed state in Malaysia and is also the most populated and industrialized state compared to other states (JPZ Sdn Bhd and IWM (M) Sdn Bhd 2015). Selangor State together with the Federal territories of Kuala Lumpur and Putrajaya covers an area of 8,396 km2, as shown in Figure 1.
Selangor comprises nine administrative districts which are Klang, Kuala Langat, Gombak, Hulu Langat, Hulu Selangor, Petaling, Sabak Bernam, Kuala Selangor, and Sepang.
There are four main river basins in the state, namely, Sungai Bernam, Sungai Selangor, Sungai Klang, and Sungai Langat. All the four river systems flow in the west direction and ultimately discharge into the Straits of Malacca.
The topography of Selangor State can broadly be divided into three distinctive regions, namely, mountainous zones, hilly areas, and lowland areas. The first zone, mountainous areas, is located in the north-eastern part of the state, and the highest peak has a height of 1,900 m from the mean sea level. The second zone, which is hilly in nature, is characterized by gentle slopes spreading widely from north to east in the middle part of the state. The area is below 100 m and the river flows gently along these areas. The third zone is the relatively flat alluvial plane located in the southwest of the state. This zone is bounded by the hilly area to the north and east and by the Straits of Malacca to the west.
The climatological data have been collected for the duration of 1982 to 2012. The mean annual rainfall of the study area ranges from 1,623 mm to 2,960 mm. The mean annual potential evapotranspiration varies from 1,431 mm to 1,610 mm. The temperature throughout the year is quite constant with a mean of 27 °C. The relative humidity ranges from 62% to 96% and averages around 82% (JPZ Sdn Bhd and IWM (M) Sdn Bhd 2015).
The most important aquifers in Selangor State are the Kuala Lumpur limestone, Kenny Hill Formation, Hawthornden Formation, Kajang Formation, and the unconsolidated alluvium comprising both river and coastal alluvium. The coastal alluvium aquifer is generally brackish or saline with very small pockets of freshwater of limited usage. The river alluvial aquifer is slightly better, but overall, yields are low except for areas around Banting. The Kenny Hill Formation and limestone aquifers are capable of sustaining moderate yields. Although there is some groundwater potential within the limestone, it is prone to pollution. Other than its carbonate incrustation potential it has high nitrate and coliform counts. Furthermore, large scale pumping from this formation can lead to land subsidence as a result of sand pumping and lowering of groundwater levels. The underlying Kenny Hill Formation is predominantly quartzite, sandstone, shale, and phyllite.
In the study area, crops are grown in rain-fed condition. The major area is covered by forest, which covers 55% of the total area, followed by water body (21%) and then homesteads (14%).
In Selangor State, the lowland soil is composed of strongly weathered ferralitic soils, whereas yellow podozolic soils, acid brown forest soils or podozols are found in the mountain areas. Quaternary deposition as clay, silt, and peat formed a low humic clay soil which becomes alluvial soil with the fluvial silt deposition.
Groundwater in Selangor State occurs in both alluvial and hardrock aquifers. The groundwater abstraction record for the years 2005 to 2013 indicates a rapid rise in groundwater abstraction from 1.8 MCM in the year 2005 to 21.6 MCM in the year 2009, mostly to meet the industrial demand. About 65% of the abstraction was taken by Megasteel for the steel mill operation located at Banting. Sungai Klang basin has the largest number of industries and commercial activities compared to the other river basins in Selangor State (JPZ Sdn Bhd and IWM (M) Sdn Bhd 2015).
MATERIALS AND METHODS
This study was conducted to assess groundwater availability of Selangor State, Malaysia for an efficient planning and management of water resources. Assessment and management of water resources can be done considering surface water and groundwater in isolation but this isolated approach fails to address the integrated behavior of the land and water ecosystem, interaction between the surface and groundwater within the water ecosystem. To ensure the integrated approach, a physically based distributed modeling system was applied under this study which is more realistic and appropriate than the traditional analytical approach.
Groundwater modeling is a powerful tool to optimize and predict the groundwater resources. Groundwater models apply the advantages of recent advances in computer technology, provide real-time modeling, visualization and analysis of two- and three-dimensional flow (Abdul Rahim & Abdul Ghani 2002; Welsh 2008). Simulation of groundwater modeling is an excellent tool to understand the behavior of an aquifer system subjected to different stresses (Rejani et al. 2008).
MIKE SHE and MIKE 11, a coupled numerical model-based modeling software was used under this study to fulfill the study objectives. MIKE SHE is a deterministic, fully distributed physically based modeling system which is capable of describing the entire hydrogeological cycle over the model area (Refsgaard & Storm 1995; Rahim et al. 2009). Six process-oriented components, interception/evaporation, overland/channel flow, unsaturated zone, saturated zone, precipitation/snow melt, and exchange between rivers and aquifer describe the principal physical processes of the land phase of the hydrological cycle. These six components are included as a modular structure in the MIKE SHE Water Movement module (Thompson et al. 2004).
MIKE SHE uses a finite difference approach to solve the partial differential equations, describing the process of overland flow (2D Saint-Venant equation), channel flow (1D Saint-Venant equation), unsaturated zone flow (1D Richards equation; Richards 1931), and saturated zone flow (3D Boussinesq equation). The methods proposed by Kristensen & Jensen (1975) were used to calculate the actual evapotranspiration based on potential evapotranspiration. The actual evapotranspiration was estimated by MIKE SHE on the basis of potential evapotranspiration rates, the root depths, and leaf area indices of different crops over the seasons.
MODEL SET-UP AND SIMULATION
Discretization
Due to the huge variation in topography, the model was developed for an area having an elevation within the range of 0 m to 200 m. The model area is shown in Figure 2. The model area was discretized into 1,000 m square grids. The model had 6,439 grid cells, where 478 grids are the boundary cells and the rest are computational cells.
Map showing the location of study area, model area, topography, hydro-stratigraphic cross-section.
Map showing the location of study area, model area, topography, hydro-stratigraphic cross-section.
Meteorological input
The spatial distribution of precipitation was generated by integrating Theissen polygons of rainfall stations. Time series data of the precipitation were obtained from secondary sources. Time series evaporation data for the station Loji Air Kuala Kubu Bahru were used in the model. Hydro-meteorological stations for Selangor State are shown in Figure 3. Land-use and vegetation information were used in the MIKE SHE model to calculate actual evapotranspiration. The crop database contains information on leaf area index, root depth, and other properties of each crop and was included in the model.
Unsaturated zone input
The most commonly occurring textures were found to be peat, loamy sand, and clay in unsaturated zone. Soil moisture retention curve and hydraulic conductivity function were given as inputs for each type of soil profile.
Surface water model
The four major rivers (Sungai Langat, Sungai Selangor, Sungai Klang, and Sungai Bernam) were included in the surface water model development. Uniform Manning's number (30 m1/3/s) was provided for different reaches of the river. The surface water model mainly comprised 12 upstream and three downstream boundaries. Upstream boundaries were provided with discharge and downstream boundaries were provided time series water levels.
Overland flow
Detention storage was taken as 50 mm for the initiation of the runoff flows. A constant value of Manning's number (M) as 10 m1/3/s was considered for the entire area.
Saturated zone input
The saturated zone component of MIKE SHE Water Movement module is the platform for exchanging water with the other components and estimates the saturated subsurface flow in the catchment. In the present study, 141 bore log data were collected from secondary sources to define the hydro-stratigraphic layers. Considering lithological variations and groundwater flow capacity, hydro-stratigraphic units of the study area were defined as top soil, aquitard, aquifer, and bedrock. A hydro-stratigraphic cross-section along A–A/ and B–B/ is shown in Figure 4. The hydraulic properties were obtained from secondary sources.
Drainage option was included into the model. Drain flow is simulated by a linear routing of water. In this study, drain water was routed to the respective rivers.
The Straits of Malacca are located along the west boundary line of the model area and thus the sea water level was considered as the west boundary. No monitoring well was available along the east and south boundary lines of the model area. Hence, along the east and south boundary lines, flux boundary was considered. Water level data from the MIKE 11 model result for Bernam River were used along the north boundary line.
A major water user within the Selangor State is water abstraction for industrial purposes from groundwater and surface water. Major water users for industries in the Selangor State were collected from LUAS and other related agencies. The numbers of production wells along with their daily abstraction and locations were used in the model set-up.
Coupling of MIKE SHE and MIKE 11
All of the branches within the study area were specified as coupled reaches and so exchanged water with adjacent MIKE SHE grid squares. The leakage coefficient of the riverbed materials is used for the evaluation of the river–aquifer exchange.
RESULTS AND DISCUSSION
Calibration and validation of coupled model
The developed coupled model was calibrated against observed groundwater level for the hydrological years 2007–2012 and validated for the years 2013–2014. During calibration, overland leakage coefficient, drain level, hydraulic conductivity, and storage coefficient were adjusted. The final calibration parameters are given in Table 1. A sample of calibration plots are given in Figure 5. Overall calibration results show a good match between simulated and observed values of groundwater levels.
Final calibration parameters
Parameters . | Statistics . | Top soil . | Aquitard . | Aquifer . | Bedrock . |
---|---|---|---|---|---|
Horizontal hydraulic conductivity, Kx (m/day) | Maximum | 0.086 | 0.047 | 14.4 | 0.047 |
Minimum | 0.044 | 0.024 | 1.75 | 0.024 | |
Average | 0.069 | 0.038 | 4.92 | 0.038 | |
Vertical hydraulic conductivity, Kv (m/day) | Maximum | 0.0086 | 0.0047 | 1.44 | 0.0047 |
Minimum | 0.0044 | 0.0024 | 0.175 | 0.0024 | |
Average | 0.0069 | 0.0038 | 0.492 | 0.0038 | |
Specific storage, Ss (m−1) | Maximum | 1.00 × 10−06 | 6.60 × 10−09 | 4.57 × 10−09 | 6.00 × 10−06 |
Minimum | |||||
Average | |||||
Specific yield, Sy | Maximum | 0.03 | 0.02 | 0.11 | 0.04 |
Minimum | |||||
Average |
Parameters . | Statistics . | Top soil . | Aquitard . | Aquifer . | Bedrock . |
---|---|---|---|---|---|
Horizontal hydraulic conductivity, Kx (m/day) | Maximum | 0.086 | 0.047 | 14.4 | 0.047 |
Minimum | 0.044 | 0.024 | 1.75 | 0.024 | |
Average | 0.069 | 0.038 | 4.92 | 0.038 | |
Vertical hydraulic conductivity, Kv (m/day) | Maximum | 0.0086 | 0.0047 | 1.44 | 0.0047 |
Minimum | 0.0044 | 0.0024 | 0.175 | 0.0024 | |
Average | 0.0069 | 0.0038 | 0.492 | 0.0038 | |
Specific storage, Ss (m−1) | Maximum | 1.00 × 10−06 | 6.60 × 10−09 | 4.57 × 10−09 | 6.00 × 10−06 |
Minimum | |||||
Average | |||||
Specific yield, Sy | Maximum | 0.03 | 0.02 | 0.11 | 0.04 |
Minimum | |||||
Average |
Examples of validation results against groundwater levels for the period 2013–2014 are shown in Figure 6. Overall, validation results show a similar trend of groundwater fluctuation and good matching of groundwater levels between observed and simulated values also for the validation periods. From the results of the model validation, it could be concluded that the parameters used in the calibrated model are acceptable, thus the model can be used for prediction purposes.
Uncertainty/sensitivity analysis of coupled model
Model uncertainty/sensitivity analysis is a procedure for quantifying the impact on an aquifer's simulated response due to variation of model parameter or model stress. The purpose of sensitivity analysis is to identify those parameters which are most important in determining aquifer behavior. Sensitivity analysis was carried out for the horizontal hydraulic conductivity, specific storage, and specific yield. The hydraulic conductivities and specific yield values were multiplied by 2 and 0.5 times to their base condition. The impacts on specific storage have been evaluated by taking a value of 1 × 10−4/m and 1 × 10−6/m whereas it was considered as 4.57 × 10−9/m in base condition. The sensitivity of horizontal hydraulic conductivity and specific yield is shown in Figures 7 and 8, respectively.
The sensitivity of specific storage is shown in Figure 9.
Figure 7 shows that horizontal hydraulic conductivity has an influence in the model calibration. The sensitivity plots also indicate that specific storage and specific yield do not have significant influence on model calibration.
Groundwater flow direction
Based on the potentiometric surface, a contour map along with the groundwater flow direction was produced and is shown in Figure 10. The minimum groundwater level occurs in Megasteel area whereas maximum levels occur in the eastern part of the hilly region. Due to the higher elevation in groundwater level in the eastern part of the model area, the groundwater flows from east to west and finally discharges into the sea.
Assessment of surface water–groundwater interaction
Exchange between surface water and groundwater for Sungai Langat, Sungai Klang, and Sungai Selangor River were calculated. Sungai Bernam was discarded as it was used as boundary condition for model set-up. The approximate exchange volume estimated from the calibrated model is presented in Table 2. Table 2 reveals that all the three rivers only receive water from groundwater and do not feed the groundwater. This may be due to the higher in ground elevation; the groundwater table always remains higher than the surface water level. Sungai Selangor receives the greatest flow from the aquifer.
Year-wise surface water–groundwater exchange volume
River name . | Aquifer to river exchange volume (Mm3) . | ||||
---|---|---|---|---|---|
2008 . | 2009 . | 2010 . | 2011 . | 2012 . | |
Langat | 74 | 59 | 61 | 52 | 72 |
Klang | 41 | 23 | 38 | 33 | 45 |
Selangor | 109 | 96 | 96 | 97 | 99 |
River name . | Aquifer to river exchange volume (Mm3) . | ||||
---|---|---|---|---|---|
2008 . | 2009 . | 2010 . | 2011 . | 2012 . | |
Langat | 74 | 59 | 61 | 52 | 72 |
Klang | 41 | 23 | 38 | 33 | 45 |
Selangor | 109 | 96 | 96 | 97 | 99 |
Basin-wise sustainable groundwater resources assessment
Sustainable groundwater resources were computed based on potential resources estimation. To evaluate the potential recharge for the specific zone, a model simulation was carried out considering average year hydrological condition. Model simulation was performed by lowering the groundwater level and a trial and error procedure was conducted to visualize whether the groundwater table regains its original position.
Water balance techniques have been extensively used to quantitative estimates of water resources. The groundwater resources were estimated from model results simulated for the period of average year hydrological condition considering the corresponding rainfall events of each rainfall station. Water balance for Langat basin area considering average year hydrological condition is presented in Figure 11. Using the water balance obtained from the model simulation, potential recharge for the Langat basin area was estimated as 221 mm.
The available sustainable groundwater resources, as calculated from potential recharge based on the water balance study for different basins, are summarized in Table 3. The approximate potential groundwater resources based on water balance study varies in the ranges 860–960 million liter per day (MLD), 630–690 MLD, 810–870 MLD, and 630–690 MLD for Langat basin, Klang basin, Selangor basin, and Bernam basin, respectively. Table 3 shows that the sustainable yield (221 mm) for Langat basin area is higher than other basins. This may be due to the fact that the major portion of the basin area is being formed by alluvium formations. The results presented here may vary with other studies due to the variation in approaches and parameters used.
Available groundwater resources for different basins
Basin name . | Potential recharge (mm) . | Basin area (km2) . | Basin area in model domain (km2) . | Available resources (MLD) . |
---|---|---|---|---|
Langat | 221 | 2,423 | 1,580 | 860–960 |
Klang | 190 | 1,288 | 1,288 | 630–690 |
Selangor | 179 | 2,200 | 1,723 | 810–870 |
Bernam | 221 | 1,131 | 1,131 | 630–690 |
Basin name . | Potential recharge (mm) . | Basin area (km2) . | Basin area in model domain (km2) . | Available resources (MLD) . |
---|---|---|---|---|
Langat | 221 | 2,423 | 1,580 | 860–960 |
Klang | 190 | 1,288 | 1,288 | 630–690 |
Selangor | 179 | 2,200 | 1,723 | 810–870 |
Bernam | 221 | 1,131 | 1,131 | 630–690 |
Spatial distribution of groundwater resources
The availability of groundwater resources was determined for the model area by subdividing it into a number of zones. The zones were subdivided based on topography, geological formation, and rainfall distribution. The geological map as produced by JMG was used to define the geology of the model area. Finally, the rainfall distribution was done by the Theissen polygon method. A total number of 70 zones was identified. The distribution of groundwater resources is presented in Figures 12–14.
Available groundwater resources for (a) Kula Langat and (b) Sepang district.
Available groundwater resources for (a) Hulu Langat, (b) Port Klang, (c) Petaling, (d) FT Kuala Lumpur, (e) Kula Selangor, and (f) Gombak district.
Available groundwater resources for (a) Hulu Langat, (b) Port Klang, (c) Petaling, (d) FT Kuala Lumpur, (e) Kula Selangor, and (f) Gombak district.
Available groundwater resources for (a) Hulu Selangor and (b) Sabak Bernam district.
Available groundwater resources for (a) Hulu Selangor and (b) Sabak Bernam district.
Impact on groundwater level due to extreme dry condition
The calibrated and validated model was simulated to identify the impacts of extreme dry condition on groundwater level. Technicalities of the options are briefly described below:
Hydrological condition for the respective extreme dry year for each rainfall station.
Land use was considered for the present condition.
Present abstraction was considered.
All the parameters obtained through calibration were kept the same in this simulation.
The changes of the groundwater level for extreme dry condition are presented in Figure 15. The model simulation reveals that there is a significant impact on groundwater level due to the extreme dry rainfall event. Figure 15 shows that most of the area during the dry period of the groundwater table will decline to a depth of 0.5 m. Some parts of the Selangor State groundwater table will decline to a depth of 1.5 m.
Impact on groundwater level due to land-use change
Study of the land-use changes and their effects on runoff for the watershed level are essential in water resource planning and management. This study provided an approach to identify the effects of land-use changes on groundwater level for Selangor State. The proposed land-use map for the year 2035 of Selangor State was obtained from Jabatan Perancangan Bandar dan Desa (JPBD). A total number of 41 abstraction wells were proposed near the newly developed area to abstract the groundwater for fulfilling the domestic water demand. The capacity of each well was 150 m3/hr. Hydraulic parameters obtained through calibration were kept the same in this option simulation. The number of geological and computational layers with their top and bottom elevations, soil properties and soil moisture retention curves, DEM of the study area, crop database were also kept the same.
The changes of the groundwater level for future land-use pattern are presented in Figure 16. Figure 16 reveals that due to changed land-use pattern and increased abstraction, the groundwater level for the major portion of Selangor State will decline down to about 0.2 m.
CONCLUSION
The aim of groundwater resource studies of Selangor State is to provide information and data for a holistic and comprehensive management of groundwater resources for sustainable development.
Integrated, physically and distributed models were set up and successfully coupled using the MIKE 11, MIKE SHE modeling system. The simulation of both surface water level and groundwater level provided a quite good visual description of the hydrodynamic interaction for Selangor State.
The study is a reference for the proposed water resources development of Selangor State, Malaysia by using the groundwater resources, particularly during periods of water crisis in the future. Considering the urgency of groundwater usage, the study is important to gain an initial assessment of groundwater resources that are available and suitable to be developed in Selangor State.
ACKNOWLEDGEMENTS
This paper represents the partial findings of the study which was conducted by Jurutera Perunding Zaaba Sdn Bhd (JPZ) in association with Institute of Water Modelling (M) Sdn Bhd for the Lembaga Urus Air Selangor (LUAS). The authors are grateful to the Lembaga Urus Air Selangor (LUAS) for funding the study. Special thanks also to the Jurutera Perunding Zaaba Sdn Bhd (JPZ) for their support to complete the project.