Assessment of Heavy Metal Contamination in ...

1052

International Journal of Earth Sciences and Engineering ISSN 0974-5904, Vol. 04, No. 06, December 2011, pp. 1052-1058

Assessment of Heavy Metal Contamination in Sediments using Multivariate Statistical Techniques in an Abandoned Mining Site: A Case Study from Kolar Gold Fields Area, Karnataka, India A. KESHAV KRISHNA, K. RAMA MOHAN and N. N. MURTHY Environmental Geochemistry Group, National Geophysical Research Institute (Council of Scientific & Industrial Research), Hyderabad, INDIA Email: [email protected], [email protected], [email protected] Abstract: Multivariate statistical techniques such as Principal component analysis (PCA), Factor analysis (FA) were applied to understand the process of toxic metals in an abandoned mining area. Thus, the interpretation of a data set from abandoned Kolar Gold Fields area, generated during the monitoring of 23 elements at thirty two locations was studied. The metal concentration data for the sediment samples are reported in terms of basic distribution pattern of the elements in the form of bar diagrams, statistical parameters and metal to metal correlations. The varifactors obtained from analysis indicate that three factors were obtained expressing 85.85%, 78.98% and 61.73% of the total variance. Fe2O3, MnO, MgO, CaO, TiO2, Co, Cr, Cu, Ni and V in factor 1 are controlled by regional rock and soil weathering, whereas P2O5, S, As, Pb and Zn in factor 2 are controlled by industrial mining activity in the past and Al2 O3 in factor 3. Further, the correlation study along with linear regression supported the fact that various elevated metal concentrations emerged from the past mining activities, lead to contamination of the sediment, and also eventually to groundwater in their proximity. This study illustrates the use of multivariate statistical techniques for understanding and interpretation of large variable data sets in terms of evaluating pollution from mining activity. Keywords: Multivariate statistics, PCA, FA, heavy metals, kolar gold fields, abandoned mining. 1. Introduction: Mining is one of the most important sources of heavy metals and can be responsible for impacts in its surrounding environment. Metal pollution of the environment as a result of abandoned mining activities is an acute problem nowadays. Although mineral resources extraction has been carried out for centuries, until last few decades relatively little attention has been given to minimize the metal dispersal around these areas coming from the indiscriminately dumped mining wastes. As a consequence, one of the challenges facing society today is the identification, evaluation and remediation of these old abandoned areas to protect public health and environment. Mineral extraction and processing especially metals, rocks and tailings cause a potential risk to the environment; when these

materials are exposed to weathering [Riccardo etal., 2005]. Metals such as Cd, Cu, Ni, Pb, and Zn are often elevated over background levels in sediments due to mining and other human activities. Concentration of metals that cause toxicity can vary, however by one or more orders of magnitude among different sediments [DeMora etal., 2004]. The toxicity of chemicals in sediments is determined by the degree of which chemicals bind to the sediment. This modifies the chemical potentials of the metals and as a consequence, different sediments exhibit different degrees of toxicity. The present studies are highly dependent on observation and quantitative measurements of the distribution of metals around the abandoned mining activities. The main aim of the present study which is the first comprehensive investigation of distribution

#02041214 Copyright © 2011 CAFET-INNOVA TECHNICAL SOCIETY. All rights reserved.

A. KESHAV KRISHNA, K. RAMA MOHAN and N. N. MURTHY

of heavy metals in sediments in the Kolar Gold Fields is to study the distribution of different elements by applying multivariate statistics including correlation and factor analysis, and to identify possible sources of sediment bound heavy metal after post abandoned mining activity.

1053

80 x 2 to 4 km, volcanic dominated Archean schist belt in the eastern block of Dharwar craton. In the study area the three main mines of KGF are Nandidung Mine, Champion Reef Mine and Mysore Mine, respectively (Fig. 1).

2. Materials and Methods: 2.1 Study Area: The Kolar Gold Field is situated to the south of Bangarpet taluk in the state of Karnataka. It is located 100 km east of Bangalore and about 300 km west of Chennai.

Figure 2: Sample Location Points in the Study Area 2.2 Geological Setting of the Area:

Figure 1: Map of the Study Area The Kolar schist belt of South India is well known for its gold deposits and has been mined systematically for gold over 100 years [Siddaiah et al., 1989]. KGF mines occur in the semi – arid region in the southern plains of Karnataka. The region is characterized by annual rainfall of 740mm.The warm dry temperature spread over most part of the year is around 27350C. The gold loads of Kolar Gold Fields lie in a narrow bond of rocks of Dharwar series. Kolar schist belt is a north- south trending ~

In Kolar schist belt, gold occurs as 2 types; gold quartz sulfide lodes and gold quartz calcite vein systems. Gold sulfide loads include significant quantities of magnetite, ilmenite and graphite. Total sulfide content and the amounts of pyrrohotite, chalcopyrite, spharelite, graphite and ilmenite increase from east to west, whereas arsenopyrite and magnetite contents decrease east to west. The host rock for gold is hornblende schist of lower Dharwar age accompanied with pegmatites and champion gneisses. The auriferous quartz veins strike nearly N-S, parallel to the general geological trend of the schist belt. The KGF mining region is traversed by three geologic fault system with the prominent one striking NW-SE right through the centre of the region and is known as Mysore North Fault (MNF). The other two namely Tennant Fault and Gifford Fault, are minor faults sub parallel to MNF and lie on either across the latter [Arora et al., 2001].


1054

Assessment of Heavy Metal Contamination in Sediments using Multivariate Statistical Techniques in an Abandoned Mining Site: A Case Study from Kolar Gold Fields Area, Karnataka, India

3. Sampling, Preparation and Analysis: Thirty two surface sediment samples were collected from (0-25 cm) depth. Sampling locations were chosen to provide good area coverage (Fig. 2). After sampling, sediment samples were dried and kept in an oven for 48 h at 600C temperature. The dried sediments were then minced with mortar and pestle and sieved through 2 mm sieve and further these samples were ground in agate swing grinding mill to make the sample homogeneous and to get accurate analytical data, as it is essential that the surface layer should be representative of the bulk specimen. Pressed pellet for X-ray fluorescence analysis were prepared by using collapsible aluminium cups [Krishna et al., 2007]. Elemental composition was determined using an X-ray fluorescence spectrometer, (Philips MagiX PRO model PW2440) with Rh 4KW tube. Its high level performance enables, therefore, a very sensitive and accurate determination of major and trace elements (Si, Al, Na, Mg, Ca, Fe, P, S, As, Ba, Co, Cd, Cu Mo, Ni, Pb, Rb, Sr, V, Zn & Zr) from few ppm to % level measurement. International soil and sediment reference materials from the US Geological Survey (USGS), Canadian Certified reference materials (CRMP), (SO-1, 2, 3, 4; LKSD-1, 2, 3, and 4), were used to prepare the calibration curves for major and trace elements, and to check the accuracy of analytical data.

samples followed by varimax rotation in order to make the components more interpretable [Jolliffe., 1986]. 5. Results and Discussion: 5.1. Characterization Deposits:

of

Mine

Waste

Mine waste deposits from ore treatment processes are the major source of heavy metal contamination in mining environment worldwide [Mlayah et al., 2008, Dinelli et al., 2001, Letlermoser et al., 2005, Bobos et al., 2006]. To obtain the overall distribution pattern of the elements, bar diagrams were used to compare the condition of each value of particular element at a sampling site. Table 1: Statistical Summary of Metal Concentrations in Sediment Samples in Kolar Gold Fields (mg/kg)

4. Analysis using Computer Software: Raw sediment data for basic statistical parameters were calculated to acquire the overall feature of the datasets. The tests for normality of the raw and transformed data were performed using SPSS® software prior to multivariate analysis. To assess and classify the interrelationships among the major and trace elements in the sediments into groups that show similar geochemical features, multivariate analysis of Principal component analysis (PCA) and correlation analysis were carried out using SPSS® software [Alvin., 2002, Mlayah et al., 2008, Anazawa et al., 2004, Danielsson et al., 1999]. In this study variables are concentrations of elements in sediment

A brief statistical summary of concentrations of all elements are shown in Table 1. The spatial distribution in the form of bar graphs for twenty two elements analyzed is represented. Sediments were found to contain a wide range and complicated distributions of trace metal concentrations as shown in Fig 3 and Fig 4. In general, high content of As, Ba, Cr, Co, Pb, Sr, Zn and Zr mainly distributed in the mining area near the three main mines of KGF which are Nandidung Mine, Champion Reef Mine and Mysore Mine, respectively.



1055

Figure 3: Distribution Bar Graph Pattern of Trace Elements (Cr, Co, Ba, As, Rb, Pb, Ni, Cu, Zr, Zn, V and Sr).

Figure 4: Distribution Bar Graph Pattern of Major Elements (SiO2, Al2O3, Fe2O3, MnO, MgO, CaO, Na2O, K2O, TiO2, P2O5 and S).

In detail, high concentrations were observed at mine dumps whereby the elements might have leached into the sediments particularly near the sampling points SED-24 to SED- 28 and some at sampling points SED-3 to SED6. The increase of some of the concentration values either reflects mining activity (tailings, mine waste), and could be associated with complex sulphides intimately associated with barite or weathering of naturally mineralized profiles such as arsenopyrites.

PCA was applied to obtain information on the geochemical association of elements, mobility of metals and in order to know the sources and sinks of these elements [Bermejs et al., 2003]. In this study varimax method was chosen as the rotation method in PCA as shown in Table 2. The dimensions of the parameters showed three principal factors by using PCA. Factor 1 represented Fe2O3, MnO, CaO, TiO2, Co, Cr, Ni and V. Petrogenic elements Fe, together with Ti was the major component of silica minerals that are the products of rock and soil weathering on land. Generally Co, Ni and Cr belong to siderophile element, and are main rock-forming elements. It is easy for them entering into iron–magnesium silicate mineral for isomorphism with Fe and


1056


Mg; in natural environment. Factor 2 represented P2O5, S, As, Pb and Zn and Factor 3 represented Al2O3. These three principal factors explained 85.85% of the total variance. The past mining activity and agricultural activities in the surroundings of KGF have a large contribution to the Factor 2 as these types of components due to run off and atmospheric deposition indicating anthropogenic sources. The metal correlations in the samples are listed in Table 3. The Pearson’s correlation coefficients were selected as the measurement between groups, and the correlation between individual elements produced similar results to those of principal component analysis (PCA). Also, it suggests that the distribution of TiO2 might be related to ilmenite present in the area because ilmenite always combines with V, a high correlation (1.00) was also observed and V. There was good between TiO2 correlation between As and Pb (1.00) and was greater than 0.75 between Cu and Zn. Arsenic is mainly present in as arsenopyrites in the gold ore and arsenic can characteristically be found in the mine dumps.

Table 2: Component Loadings of Three Factors, Eigen Values, and Explained Variance for Measured Variables

Table 3: Pearson’s Correlation Coefficient Matrix for Metal Concentrations in Sediments in Kolar Gold Fields (mg/kg)

Correlation in Bold are Significant at p< 0.01



6. Conclusions: Data analysis takes an important place often, when objects of interest are described by different types of parameters. At this stage of data analysis, it is important to achieve a good visual representation of the data. In the present study the multivariate statistical approaches like PCA served as a tool to study sources of heavy metal elements in sediments of Kolar Gold Fields. Furthermore, linear regression and Pearson correlation coefficients were calculated to examine the relationships between the elements based on which natural and anthropogenic input of heavy metals in sediment samples were distinguished. Present research study suggests that firstly, distribution of Fe2O3, MnO, MgO, CaO, TiO2, Cr, Ni and V in sediment is controlled by natural factor of the lithogenic process to be originated from parent material and sediment. Secondly, As, Cu, Pb, Zn , P2O5 and S have a majority of contribution from the past exploration and excavations of gold from the abandoned mining area which can be concluded as the local industrial factor and past mining activity. Based on the results the element clustering behaviour in sediment samples reveal the metal contents are controlled by both natural and anthropogenic sources, based on the individual factors observed. The results also reconfirmed that multivariate statistical method (principal component analysis) of a small set of data can provide a valuable information for a study area especially the present status of an abandoned mining area for the decades. Acknowledgements: The present work was carried out as part of the CSIR Network Project COR-008-28. The authors are thankful to Dr. Y.J.Bhaskar Rao, Acting Director of the National Geophysical Research Institute in Hyderabad for his continuous support, encouragement and his permission to publish this paper. References: [1] Riccardo Biddau, Rosa Cidu (2005) Hydrogeochemical baseline studies prior

1057

to gold mining; a case study in Sardinia (Italy). Journal of Geochemical Exploration, 86: 61-85. [2] DeMora S, Sheikholeslami MR, Wyse E, Azemard S, Cassi R (2004) An assessment of metal contamination in coastal sediments of the Caspian Sea. Mar Pollut Bull 48:61–77. [3] Krishna B R, Gejji F H (2001) The mill tailings of Kolar gold mines. Current Science, 80: 1475 – 1476. [4] Sudhakar M Rao, Venkatramana Reddy B V (2006) Characterization of Kolar gold mine tailings for cyanide and acid drainage. Geotechnical and Geological Engineering, 24: 1545 – 1559. [5] Hernandez L, Probst A, Probst J L, Ulrich E (2003): Heavy metal distribution in some French forest soils: evidence for atmospheric contamination. The Science of the Total Environment, 312: 195 – 219. [6] Siddaiah N S, Rajamani V (1989) The Geologic Setting, Minerology, Geochemistry, and Genesis of Gold Deposits of the Archean Kolar Schist Belt, India. Economic Geology, 84: 2155 – 2172. [7] Arora S K, Willy Y A, Srinivasan C, Benady S (2001) Local Seismicity due to rock bursts and near-field attenuation of ground motion in the Kolar Gold mining region, India. International Journal of Rock Mechanics and Mining Sciences 38: 711-719. [8] Krishna, A.K., Murthy, N.N., Govil, P.K. (2007). Multielement analysis of soils by Wavelength-Dispersive X-ray Fluorescence Spectrometry. Atomic Spectroscopy.2007, 28(6), Nov/Dec. [9] Alvin CR. (2002): Methods of multivariate analysis, John Wiley & Sons, Inc., USA. [10] Mlayah A, Ferreira E, de Silva, Rocha F, Ben Hamza Ch, Charef A, Noronta F. (2008) The Qued Mellegue: Mining activity, stream sediments and dispersion of base metals in natural environments; North-Western Tunisia, Journal of Geochemical Exploration, online. [11] Anazawa K, Kaida Y, Shinomura Y, Tomiyasu T, Sakomoto H, (2004).


1058


Heavy metal distribution in river waters and sediments around a ‘fivefly village’ Shikoku, Japan: application of multivariate analysis. Analytical Sciences 20, 79-84. [12] Danielsson A, Cato I, Carman R et al (1999): Spatial clustering of metals in the sediments of the Skagerrak/Kattegat. Appl Geochem 14: 689-706. [13] Jolliffe IT (1986): Principal component analysis. Springer, New York. [14] Dinelli E. Lucchini F. Fabbri M. Corteeci G. (2001) Metal distribution and environmental problems related to sulfide oxidation in the Libiora copper mine area (Ligurian Apennines, Italy). Journal of Geochemical Exploration 74(1); 141-152. [15] Letlermoser B.G, Ashley PM. (2005): Tailings dam seepage at the rehabilited Mary Kathleen Uranium mine, Australia. Journal of Geochemical Exploration 85(3). 119- 137.

[16] Bobos I, Duraes N, Noronha F (2006). Mineralogy and geochemistry of mill tailings impoundments from Algares (Aljustrel), Portugal: implications for acid sulfate mine water formation. Journal of Geochemical Exploration 88, 1-5. [17] Bermejs Santos JC, Beltron R, Gomez Araiza JL. (2003) : Spatial variations of heavy metals contamination in sediments from Odiel River (southwest Spain) Environ. Int. 29(1): 69-77. [18] Shanshan Wang, Zhimin Cao , Dongzhao Lan, Zhichang Zheng, Guihai Li (2008)Concentration distribution and assessment of several heavy metals in sediments of west-four Pearl River Estuary. Environ Geol (2008) 55:963– 975. [19] Manta DS, Angelone M, Bellanca A (2002): Heavy metals in urban soils: a case study from the city of Palermo (Sicily), Italy. Sci Total Environ 300:229-243.
