Journal of Environmental Management 70 (2004) 275–282 www.elsevier.com/locate/jenvman
The history and evaluation of saltwater intrusion into a coastal aquifer in Mersin, Turkey Zeynel Demirel* Department of Geological Engineering, Faculty of Engineering, Mersin University, Turkey Received 26 February 2003; revised 10 October 2003; accepted 8 December 2003
Abstract The Mersin – Kazanlı region is a densely industrialized region. The factories and towns cover their water demand from groundwater. With the increased water demand, saltwater intrusion has occurred. The chloride concentration of the water samples from some wells has been analysed periodically since these wells were drilled. The results of these analyses and electrical conductivity measurements were used to show the history and development of saltwater intrusion up to the year 2000. The Cl2 concentration of the water within the alluvial aquifer increased to over 3000 mg/l in 1999 and the wells were closed completely. In 2001 new wells were drilled more than 1 km away from the sea and old well field. With the results of the analyses conducted in 2001, the current groundwater quality was determined. The ground water is of the magnesium – calcium-bicarbonate type and this composition is controlled by the interaction of the water with the sediments of alluvial deposits. q 2003 Elsevier Ltd. All rights reserved. Keywords: Coastal aquifer; Environmental monitoring; Mediterranean Sea; Turkey
1. Introduction The development of groundwater resources for water supply is a widespread practice in Turkey, favoured by the existence of basins with thick Quaternary deposits that form aquifers with good-quality water. Under steady-state conditions a state of equilibrium is established between seawater and freshwater in a coastal aquifer. However, owing to increasing demand for water, groundwater may be subjected to over-exploitation and the natural equilibrium is thus disturbed. This results in aggressive seawater intrusion, which may even reach the inland aquifer. The Mersin region is a dense industrialized region. There are many factories in this region, which cover their water demand from groundwater. Since the 1980’s, with the economic development of the coastal areas of Mersin, salt-water intrusion has been induced by overexploitation of groundwater. In the mid 1990’s, groundwater overexploitation exacerbated the salt-water intrusion, and the rate of intrusion continues to increase steadily. The salinization of the freshwater resources has seriously impeded the development of * Tel.: þ324-3610001; fax: þ324-361-00-32. E-mail address:
[email protected] (Z. Demirel). 0301-4797/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.jenvman.2003.12.007
industry, agriculture and the improvement of the people’s living standards in the region. The town of Kazanlı depends on the groundwater and today, because of the saltwater intrusion and pollution, the government plans to bring water from a dam which exists 25 km away from Kazanlı. This project is very expensive. Monitoring of contaminated aquifers is an integral part of the activity related to the modern human – environment interaction and seawater intrusion is a typical result of such interaction (Melloul and Goldenberg, 1997). A factory near Kazanlı has drilled 13 supply wells since the 1970’s at different dates and analysed the chloride concentration and measured the electrical conductivity periodically. After they detected high chloride concentrations they stopped pumping from this well. As a result, seawater intrusion reached the inland aquifers. Only adequate seawater intrusion monitoring can lead to better management of such coastal aquifer. The purpose of this study is to describe the history and current condition of salinization in the Mersin – Kazanlı region and to show that the seawater intrusion can be detected and managed with a few parameters which are important in the composition of seawater chemistry.
276
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
Fig. 1. The location and geological map of study area.
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
2. Site description Mersin is situated in the Mediterranean Sea region of Turkey. The study area is located between latitude 348 370 and 348 570 and longitude 368 450 and 378 000 . The area is a delta plain, which is formed by Berdan River at the east and Delic¸ay creek at the west. The morphology in the study area is characterized by a wide flatness and it has an elevated altitude northwards. Topographic structure in the north of the investigated area has a wavy character. The altitude reaches to 300– 1100 m in the north and decreases gradually southwards. The Berdan River flow regime is strongly dependent on the seasonal rains; average discharge at the Berdan Dam in Tarsus is 38 m3/s. The Delic¸ay creek flows only 3 –4 months in a year and its average discharge is 2– 3 m3/s. The study area is situated in a region with typical Mediterranean climate. Wet and mild winter combined with dry and hot summers are typical for the coastal zone around the Mediterranean Sea. Annual average temperature in the area is 18.6 8C. Yearly average of rainfall is about 600 – 650 mm.
3. Geological setting The basin is filled with Quaternary and Tertiary sediments (Fig. 1). These sediments are overlain Paleozoic metamorphic basement rocks, which crop out on the Taurus Mountain, along the northern border of the basin (Fig. 1). In the northwest of the study area ophiolitic rocks is settled. Tertiary sedimentary rocks consist of a succession of marine, lacustrine, and fluvial deposits. The Middle-Upper Miocene aged Kuzgun formation consists of sandstone, conglomerate, limestone, tuff-marlshale and sandstone alternation. Its thickness varies between 50 and 100 m in the studied area. The Upper Miocene aged Handere formation is composed of claystone – marl-siltstone, limestone, gypsum and sandstone – conglomerate series. The thickness of the formation ranges from 50 to 500 m (S¸enol, 1998). Handere formation is the widest formation in the investigation area. In the region, at the end of Pliocene, the environment changed to continental conditions. Early Quaternary aged fan deposits; delta deposits and shore deposits took place with paleosolic caliches of pedagogic origin, Mediterranean red soil and hard caliches (S¸enol, 1998).
4. Hydrogeology As the sandstone, conglomerate and limestone series of Kuzgun formation have karstik features they should form productive aquifers, but there are not any deep wells into these aquifers in the studied area. The conglomerate, sandstone and limestone are found mostly on the upper
277
Table 1 The hydrogeological character of supply wells in studied area Well number
Date
Pumping rate (l/s)
Depth (m)
Aquifer
S-1 S-2 S-3 S-4
1974 1974 1982 1981
100 150 90 150
60 90 60 120
Alluvial Alluvial Alluvial Alluvial
S-5
1985
130
120
Alluvial
S-6
1985
90
100
Alluvial
S-7
1981
60
90
Alluvial
S-9 S-10 S-11 S-12 S-13
1993 1995 2001 2001 2001
150 160 150 150 150
110 115 150 110 150
Current status
Pumping stopped Pumping stopped Pumping stopped Pumping stopped
Alluvial Alluvial Alluvial þ Handere Alluvial þ Handere Alluvial þ Handere
levels of Handere formation and form aquifers. There are some wells such as S-11, S-12 and S-13, which drain these series (Fig. 1). The recharge rate of Handere aquifer was estimated as 8.8 £ 106 m3/yr (Demirel and Tu¨rkmen, 2001). Alluvial deposits form the most productive aquifers in the study area. The wells drilled in the alluvial aquifer produce quite large amounts of water (Table 1). The thickness of alluvial aquifer varies between 30 and 120 m. Its water bearing formations are clayey gravels and gravel deposits. The transmissivity of these deposits varies between 3.5 £ 1022 and 3.5 £ 1024 m2/s (Demirel and Tu¨rkmen, 2001). The recharge rate of the alluvial aquifer is estimated as 7.4 £ 107 m3/yr. Static levels in the wells, generally, are in the range of 2 –3 m from surface. The direction of groundwater flow is from north to south (Demirel and Tu¨rkmen, 2001).
5. Methodology Saltwater can be detected directly in observation wells by electrical electrodes or by sampling, or indirectly by geoelectrical methods (Melloul and Goldenberg, 1997). In this study the direct method by sampling is used. Water samples for chloride analysis were obtained every 15 days since the wells were drilled and at the same time the electrical conductivity was measured at the wellhead. The wells were drilled from 1974 at different dates. After some wells were affected by seawater such as S-4, S-5, S-6, and S7, pumping was stopped and a new well was drilled. The periodical chloride analysis done by the factory laboratory is used to show the development of the salt-water intrusion history in time and space.
278
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
Table 2 Results of the chemical analyses Parameter
S-1
S-10
S-11
S-12
Sea water
pH EC (mS/cm) Caþ2 (mg/l) Mgþ2 (mg/l) Fe(T) (mg/l) Naþ (mg/l) Kþ (mg/l) Cl2 (mg/l) SO22 4 (mg/l) HCO2 3 (mg/l) Chemical characters
8.2 807 56.1 50.1 ,0.1 32.4 2.6 45.1 68.7 299.5 Mg –Ca-HCO3
8.1 924 68.1 61.3 ,0.1 31.4 2.4 57.1 69.1 363.8 Mg– Ca-HCO3
8.2 814 66.1 50.5 ,0.1 22.6 2.4 43.6 60.5 314.9 Mg–Ca-HCO3
8.2 834 68.1 49.3 ,0.1 29.0 2.9 36.4 60.1 312.8 Mg–Ca-HCO3
8.2 58600 461 1480
In 2001, three new wells were drilled in the studied area, namely S-11, S-12 and S-13 (Fig. 1). These wells are located quite far from the Mediterranean Sea and collect water from both the alluvial aquifer and deeper Handere aquifer (Table 1). Today, the factory supplies its water demand mostly from these three new wells. To characterize the current groundwater quality, water samples were obtained from these and from S-1 and seawater. Samples were analysed by Inductively Coupled Plasma (ICP) in accordance with American Society for Testing and Materials (ASTM).
12600 603 22560 3098 Na-Cl
6. Results The results of the analyses from the samples taken in 2001 are shown in Table 2. The groundwater is of the magnesium – calcium-bicarbonate type (Table 2, Fig. 2) and this composition is controlled by the interaction of the water with the sediments of alluvial deposits. In all samples, the chloride ion concentration is also very high. Dissolved chloride is common in shallow groundwater, but concentrations resulting from natural sources are generally low (Hem, 1992). Chloride concentrations in
Fig. 2. Classification of ground waters in the study area (PIPER Diagram).
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
areas affected by the intrusion may exceed the U.S. Environmental Protection Agency (1986) Secondary Maximum Contaminant Level of 250 mg/l (Andreasen and Fleck, 1997). In the water from the wells S-1 and S10, the rates of the chloride are 17 and 18%mval, respectively. Both of these wells drain the alluvial aquifer, the others, namely S-11, S-12 and S-13 drain both aquifers together. This means, in the near future, if the production continues at this high rate, the chloride will also be a dominant ion in S-1 and S-10. The chemical character of seawater is Na-Cl, as expected (Table 2, Fig. 2). All the samples are in a narrow place in the Piper diagram but distant from the place of the seawater sample. With the periodical chloride analysis and electrical conductivity measurements it is possible to evaluate
279
the seawater intrusion to the wells near the Mediterranean Sea. Fig. 3 shows the change of the chloride concentration versus time for the wells of S-4, S-5, S-6, and S-7. The Cl2concentration of the water from the well S-4, which was drilled in 1981, increased slightly until 1989 and after this date the concentration was over the limit value of 250 mg/l and increased continuously. After 1992, the concentration was over 500 mg/l and in 1995 it was almost 2000 mg/l. After stopping the pumping, the concentration was decreased to the value of 250 mg/l. But, when the pumping was begun in 1996, the concentration increased again to the value of over 3000 mg/l and in 1999 this well was closed completely. The chloride concentration of the water of the well S-7, which was drilled in 1981, increased after 1987 quite rapidly
Fig. 3. The change of chloride concentrations versus time.
280
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
to the value of 700 mg/l and this well has also been closed completely. The Cl2 concentration of the water of the well S-6, which was drilled in 1985 is also increased very slightly until mid 1995, but after this time, Cl2 concentration increased very rapidly. In the mid of 2000 the pumping from this well was also stopped. Same situation is detected for the well S-5, which was also drilled in 1985. The Cl2 concentration went in an allowable range until mid 1993 and after this time the increase was very rapid and the pumping was stopped in 2000. Fig. 4 shows the change in electrical conductivity of the water from some wells. The views of the charts have the same character as the chloride charts.
The chloride content and electrical conductivity measurements given in Figs. 3 and 4 indicate that the origin of groundwater salinity in Kazanlı region is due to seawater intrusion which resulted from intensive pumping. The extent of saltwater intrusion horizontally is reflected by the distribution of chloride concentrations as an equiconcentration map (Fig. 5). The periodical chemical analyses are used to show the development of saltwater intrusion in space for four selected times. In 1984, under steady-state conditions a state of equilibrium was established between seawater and freshwater in Kazanlı coastal aquifer. At that time, the maximum chloride concentration was in the range of 50 –55 mg/l. However, owing to
Fig. 4. The change of electrical conductivity (mS/cm) versus time.
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
281
Fig. 5. Equi-chloride concentration (mg/l) contours in the alluvial aquifer in Kanzanli.
increasing demand for water, groundwater was pumped to over-exploitation and natural equilibrium was thus disturbed. This resulted in seawater intrusion reaching inland aquifers. In 1988, the chloride concentration was measured up to 250 mg/l around the wells S-5 and S-6, which were close to the Sea. In 1992, the seawater intrusion reached more to the inland side and in the year 2000, the concentration was more than 1000 mg/l around wells S-5, S-4, S-6, and S-7 and the 100 mg/l contour reached to the wells S-2 and S-3, which are located at the northern part of the studied area.
7. Conclusions Groundwater is an indispensable resource for water supply for the town Kazanlı, for the factories in Kazanlı field and for the agriculture. Between 1970 and 1980, under steady-state conditions a state of equilibrium was established between seawater and freshwater in this coastal aquifer. A factory in this region, which has 13 supply wells,
overexploited the groundwater. As a result, the groundwater was polluted by seawater. The periodical chemical analysis and electrical conductivity measurements are used to show the history of saltwater intrusion in time and space. This study showed that saltwater intrusion due to excessive pumping started in mid 1980. Between 1990 and 2000 the wells drilled before 1985 were polluted by seawater and this well field could not be used anymore for groundwater supply. With the equi-concentration mapping of chloride ion the extent of the seawater pollution in the aquifer was determined. The seawater intrusion reached more than 500 m inland in the north direction. More data will be obtained to provide more accurate mapping of seawater encroachment into the aquifer, and will be useful for management of the coastal system. In 2001 the new wells were drilled more than 1 km away from the sea and old well field. The chemical composition of the groundwater from the new wells is suitable to use for industrial and agricultural purposes. However, the new well field needs to be well managed by considering the hydrogeological parameters such as aquifer transmissivity and groundwater recharge.
282
Z. Demirel / Journal of Environmental Management 70 (2004) 275–282
References Andreasen, D.C., Fleck, W.B., 1997. Use of bromide–chloride ratios to differentiate potential sources of chloride in a shallow, unconfined aquifer affected by brackish-water intrusion. Hydrogeology Journal, V.5. Demirel, Z., Tu¨rkmen, S., 2001. The hydrogeological investigation of Kazanlı region. Soda Factory AG, Kazanlı-Mersin, unpublished report. Hem, J.D., 1992. Study and interpretation of the chemical characteristics of
natural water, third ed., US Geological Survey Water Supply Paper 2254,. Melloul, A.J., Goldenberg, L.C., 1997. Monitoring of seawater intrusion in coastal aquifers: basic and local concerns. Journal of Environmental Management 51, 73 –86. S¸enol, M., 1998. The geological investigation of Mersin region. General Directorate of Mineral Research and Exploration of Turkey, Ankara, unpublished report.