Assessment of potability of minewater pumped out from Jharia Coalfield, India: an integrated approach using integrated water quality index, heavy metal pollution index, and multivariate statistics

The dense and industrialized populace in the mining areas of Jharia Coalfield (JCF) is plagued by a severe shortage of water supply. The unutilized pumped out coal minewater discharges may be utilized to cater to the increasing water demand of the region but it runs the risk of getting contaminated from domestic and industrial effluents. The current study aimed to assess the suitability of augmenting underground minewater for potable purposes. For this purpose, ninety underground minewater samples collected from 15 locations across JCF for the hydrological year 2019–2020 were analysed to gain an insight on the physicochemical characteristics of the minewater using an integrated approach of standard hydrochemical methods, integrated water quality index (IWQI), heavy metal pollution index (HPI), and multivariate statistical analysis. For the minewater quality to be deemed suitable for potable purposes, both IWQI (lower than 2) and HPI (lower than 30) values were considered. IWQI values of the minewater samples from the study area ranged from 1.97 to 5.08, while the HPI values ranged from 18.40 to 53.05. The pH of the samples were found to be mildly acidic to alkaline (6.5 to 8.3) with varying total hardness (149 to 719 mg L−1), total dissolved solids (341 to 953 mg L−1), and electrical conductivity (568 to 1389 µS cm−1), reflecting heterogeneity in underlying hydrosystems, variations in geological formations, and the influence of lithogenic and anthropogenic processes on the water chemistry of the region, which was corroborated by the principal component analysis (PCA) and hierarchical cluster analysis (HCA) of the minewater samples. Two major water types of the region were identified, viz., Ca–Mg–HCO3 and Ca–Mg–Cl–SO4. This multiparametric approach gives a holistically accurate assessment of the minewater quality, overcoming the limitations of traditional water quality indices and facilitating time-saving and effective water management practices, and sets the foundation for augmenting minewater for potable purposes to meet increasing demands.


Introduction
About one-sixth of the global population lack access to water of potable quality (WHO 2012), despite global advancements in technology and economy. Minewater discharged during mining operations is a valuable water resource, but it runs the risk of getting contaminated from domestic and industrial effluents, causing adverse effects on human health and stymieing the growth and development of the region (Neogi et al. 2018;Adimalla and Qian 2019). Water resources management is vital for sustainability in densely populated and industrialized regions. This becomes more relevant in water-stressed regions, prone to contamination with limited access to potable water (Badham et al. 2019;Xiang et al. 2021). Efficient water resource management requires a significant capital investment in smart infrastructure, efficient treatment technologies, quantifiable indicators, and monitoring tools (Behmel et al. 2016;Platikanov et al. 2019). Studies carried out in urban areas of different regions Responsible Editor: Xianliang Yi across the globe such as the states of Arizona and California in the USA, New South Wales and Victoria in Australia, Western Cape Provinces in South Africa, and the Mediterranean coast have revealed a significant reduction in quality and quantity of their potable water resources, influenced by environmental and anthropogenic pressures (Navarro-Ortega et al. 2015;Stefanidis et al. 2018;Best 2019;Bhurtun et al. 2019). The installation of specific infrastructures has been proposed and implemented to ameliorate potable water quality and to ensure its future availability. In this context, an integrated approach to assess the water quality of a region by using integrated water quality index (IWQI) and heavy metal pollution index (HPI) model reduces time and effort expended in the treatment processes. Based on the values of the integrated indices, corroborated by chemometric methods like principal component analysis (PCA) and hierarchical cluster analysis (HCA), appropriate treatment technologies can be set up, thus adopting a comprehensive water resource management strategy.
The Bureau of Indian Standards (BIS), World Health Organization (WHO), the European Union (EU), etc., define permissible limit as the maximum admissible concentration and acceptable limit as the minimum required concentration of ionic content in water. For maintenance of mineral homeostasis in the human body, the concentration of ionic content in the water needs to be within range of both the threshold limits. Treated water could be deemed unfit for consumption if it lacks a few essential minerals which are vital to the body. Studies on mineral deficiency have revealed its adverse effects on the human health (Shu 2015;Neogi et al. 2018;Mukate et al. 2019;Hembrom et al. 2020). While traditional water quality indices (WQI) consider the values of physicochemical parameters below the acceptable limits only to be good without accounting for the ion deficiency in the water, giving an erroneous assessment of water quality, IWQI uses both acceptable/desirable limits and permissible limits in its assessment of water quality from the aspect of ion concentration in the minewater.
In this paper, we seek to give an insight on the assessment of potability of minewater pumped out from Jharia Coalfield, using an integrated approach which constitutes application of the IWQI model, HPI model, and multivariate statistics to fill the research gap of comprehensive evaluation of water quality of the study area by overcoming the limitations of traditional WQI model by incorporating both the drinking water threshold levels in its calculation as well as accounting for the presence of heavy/trace metals present in the sampled water, giving a holistically accurate assessment of the minewater quality.

Study site
Jharia Coalfield is a sickle-shaped terrain that spreads across a 450-km 2 area between 23°37′ N -23°52′ N latitudes and 86° 09′E-86° 30′ E longitudes with an altitude averaging 220 m above mean sea level, situated in Dhanbad district, in the Indian state of Jharkhand. It is located at the centre of the Damodar valley, bounded by the Eastern Railway in the north and the Damodar River in the south (Singh et al. 2018a, b). The sampling sites selected for the current study comprised 15 operational minewater treatment plants (MWTPs) spread across Jharia Coalfield (Fig. 1)

Geomorphology and drainage
A major feature of the Jharia coal basin is its geology which belongs to the Gondwana group of sedimentary strata formed from the ancient rock types of Dharwar and post-Dharwar period and is characterised by coal seams and patches of sandstone (Singh et al. 2018a, b). The soil cover of the study region is of the alluvial type with a low nutrient and organic matter concentration lying above the coal-bearing Gondwana superstratum (Singh et al. 2019;Bharti et al. 2021). A large portion of the study area has fine-to-mixed loamy soil texture while the southern part of the study area comprises sandy soil and loose sand of alluvium sequence, favouring considerable infiltration (Rehman et al. 2020;Adhikari and Mal 2021). The entire region of the Jharia Coalfield has been of great importance due to its large reserves of lower Gondwana coal, the mining of which over the last century has affected the environment around it (Singh et al. 2012;Singh et al. 2018a, b).
The Damodar River forms the main component of natural drainage of the region, flowing west to east, capturing all the surface drainage from a few major streams (viz. Bansjora, Chatkari, Katri, Khudia, Kumari, Jamunia, Tisra, etc.), and the rest intermittent, which drain the Jharia Coalfield from the north to south, forming a dendritic drainage system (Panigrahy et al. 2015a, b;Singh et al. 2018a, b;Kumar et al. 2019). Mining activities on the surface and underground, below the piezometric level, change the stream gradient, influencing the flow regime of groundwater and surfacewater (Yuan et al. 2020). To understand the overall topographic expression and drainage network of the study area, a map ( Fig. 1)

Sampling and analysis
Two sampling campaigns were carried out in the premonsoon and post-monsoon seasons of March-April and October-November, respectively, for the hydrological year 2019-2020. Ninety underground coal minewater samples (3 composite samples from each of the 15 minewater treatment plants operating in JCF, Jharkhand) were collected from the storage tank outlets of the minewater treatment facilities in high-density polyethylene bottles (pre-washed) of 1-L capacity for the year 2019-2020. In each site, three field duplicates were collected and well mixed in situ subsequently. A total of 14 physicochemical parameters and 9 trace/heavy metals (iron, manganese, zinc, nickel, lead, copper, chromium, arsenic, and cadmium) determining water quality of the region were analysed. Type I (18.2 MΩcm) Milli-Q water was used to prepare all the solutions. EC, pH, and TDS values were gauged onsite using Multiparameter pH Tester 35 Ecotech instrument. Total alkalinity (titration with 0.02 N H 2 SO 4 using phenolphthalein and methyl orange indicator), total hardness (titration with EDTA), sulphate (turbidimetric method), nitrate (spectrophotometric method), chloride (argentometric method), and fluoride (SPADNS method) were analysed using a UV-visible spectrophotometer (UV-1800 Shimadzu), as per the methods prescribed in APHA 2017. Duplicate blanks and a laboratory water standard were analysed for quality control. EDTA titration method was used to analyse calcium while magnesium was determined by multiplying the magnesium hardness by a factor of 0.2431. Potassium and sodium were determined by the  (ESICO, M-1385). The trace/heavy metal concentration in the minewater was determined after digestion of the samples with 1 N HNO 3 prior to its analysis, using GBC Avanta PM Atomic Absorption Spectrophotometer in flame mode at the recommended wavelengths. The spatial distribution of the integrated water quality index was drawn using IDW technique in ArcGIS 10.2.2 software.

Quality control and quality assurance
The quality control and quality assurance were checked by corroborating analytical methods based on appropriate standards (calibration, blank reagents, detection limit, and accuracy) during all measurements. The calculation of correlation coefficient (R 2 = 0.99) of standard curves of known concentrations of anions and cations confirmed the accuracy of results. The accuracy of the chemical analysis of the minewater samples was determined by calculating the ionic balance error (IBE), as shown in the equation: where, A and C are the concentrations of total anions and cations in milliequivalent per litre (meq L −1 ) and B is the percentage of ionic balance error. To increase the accuracy of the analysis, values of B up to 10% were considered (Yadav et al., 2020).

Compilation of data and statistical methodology
Data compiled during the study followed the standard methods prescribed by APHA (2017). This data was processed using IWQI, HPI, and multivariate statistical analysis.

Calculation of IWQI
IWQI was applied to the coal minewater samples to evaluate the minewater quality and its suitability for consumption. Studies on potability of water have revealed its beneficial effects on human health to occur at acceptable concentrations (Li and Wu 2019). Therefore, acceptable limit (AL) and permissible limit (PL) values of the water quality parameters are assigned by the BIS based on the hazard it poses to human health (Table 1). The values between the two threshold limits fall under the range which can also be represented as the difference in deficit (15%) of the range of the specific water quality parameter from its permissible limit (PL) to give the modified permissible limit (MPL) given in Table 1.
The percentage (%) deficit can be changed as per the situation and is used to buy time to alleviate the existing contamination levels, preventing contamination from crossing the threshold level which otherwise could adversely affect the environment around it (S. Mukate et al. 2019). The values that fall within the AL and MPL are considered ideal for consumption. When the value of x th water quality parameter ( Q x ) is higher than AL but lower than MPL, i.e. AL ≤ Q x ≥ MPL, the value of the sub index (SI 1 ) is taken as 0. When the value of x th water quality parameter is lesser than the acceptable limit, i.e. Q x ≤ AL, then SI 2 = (AL−QX) AL . When the water quality parameter Q x has a higher value than the MPL, i.e. Q x≥ MPL, then SI 3 = (Qx−MPL) MPL . To compute the sub-indices, division of the difference in the specific water quality parameter (Q x ) with its acceptable limit (AL) or MPL is done by its respective AL or MPL. This is done to obtain a homogenized value for identification of increased or decreased concentration of a specific parameter with respect to its AL or MPL. These computed values are added to get the IWQI.
where SI xy = value of sub-index of x th sample and y th water quality parameter.

Calculation of HPI
The trace/heavy metal contamination affecting minewater quality was evaluated by computing the analysed samples with HPI, an indexing model which incorporates weighted arithmetic quality in assigning weights ( w x ) ranging from 0 to 1 to individual metals (Singh and Kamal 2017). The value for critical pollution index is set at 100 for drinking water (Bhardwaj et al. 2017). The standard for the analysed minewater quality parameter is designated as S x which is in inverse proportion to the unit weightage ( W x ) of the respective minewater quality parameter.
The formula proposed by Mohan et al. (1996) was utilized in calculation of HPI: where W x denotes unit weightage of x th minewater quality parameter, sub-index of the x th parameter is denoted by Q x and n is the number of minewater quality parameters analysed. W x is computed by applying the formula: where the permissible limit (BIS 2012) of x th minewater quality parameter is denoted by S x and p is assigned to be the proportionality constant. Q x denotes sub-index of x th minewater quality parameter and was calculated using the formula: where S x is the permissible limit (BIS 2012) for the x th parameter and M x which is expressed in microgram per litre denotes monitored value of x th minewater quality parameter. In this study, HPI values were grouped into three categories which have been demarcated as high (> 30), medium (15-30), and low (< 15) depending upon trace/heavy metal concentration in the analysed minewater (Giri and Singh 2014;Panigrahy et al. 2015a, b;Tiwari et al. 2015).

Multivariate statistics and geostatistical modelling
PCA and HCA were used in identifying the minewater contamination source and the similarities among the sampling locations, respectively. PCA and HCA of the analysed data were done using XLSTAT 2019.1 software to reduce the large dataset for interpreting patterns within the data. The original dataset was reduced after alteration into a newer group of variables known as principal components, which were obtained from linear combinations of variables of the original dataset and categorized in a manner that the first principal components are accountable for variations in the original dataset (Ogwueleka 2015;Kamtchueng et al. 2016).
The minewater quality of the study area was represented spatially using the inverse distance weightage (IDW) method to interpolate IWQI and HPI data in Arc GIS 10.2.2. This method weighs the sampling points during interpolation in a manner that the impact of one point over the other increases with decrease in relative distance from the unknown point created for the study.

Results and discussion
A descriptive statistic of the minewater quality parameters of the minewater samples collected from the study area across two campaigns (pre-Monsoon and post-Monsoon) for the year hydrological year 2019-2020, with its respective acceptable limit (AL) and permissible limit (PL)  Table 3. A detailed description of the hydrological compositions of the major cations, anions, and the charge balance errors (%) of the study area for the hydrological year 2019-2020 is given in Supplementary  Table 2. The cationic concentration of major ions in the coal minewater of JCF in order of their relative abundance was calcium > magnesium > sodium > potassium, while the anionic concentrations were in the order: sulphate > bicarbonate > chloride > nitrate > fluoride. The value for the total cationic charge TZ + ranged between 3.95 and 11.9 meq L −1 (mean 8.85 meq L −1 ). The total anionic charge (TZ − ) ranged between 4.25 and 12.83 meq L −1 , having an arithmetic mean of 9.38 meq L −1 . Charge balance error up to N ± 5% was considered acceptable for all the minewater samples (N). In total, 37% of the samples showed a charge balance error exceeding 5%, exhibiting a deficit in total cations due to an anionic excess because of accumulation of contaminated load due to mining operations near the area of study.
The physicochemical characteristics of minewater of the study area for the hydrological year 2019-2020 are summarized by box plots (Fig. 2) which were plotted using Origin Pro 9.0 software.
Considering the most important trace/heavy metals from previous investigations in JCF, whose concentration levels affect human lives, nine trace heavy metals were chosen for this study. Their order of relative abundance was found to be Fe > Mn > Zn > Ni > Pb > Cu > Cr > As > Cd. The trace/heavy metal concentrations in most of the analysed samples, barring a few exceptions, were lower than the permissible limits set by BIS 2012. Fe, Mn, Pb, Cu, and Ni were observed to occur in concentrations exceeding the acceptable limits at some sites. Fe concentration ranged from 226 to 626 µg L −1 (mean 421 µg L −1 ) across the study sites. A total of 74% of the samples had Fe concentration exceeding the maximum acceptable limit of Fe (300 µg L −1 ). The Mn concentration ranged from 4.5 to 697.4 µg L −1 (mean 118.4 µg L −1 ). Of the Table 3 Statistical summary of physico-chemical characteristics of the sampled minewater † No guideline values have been established for some of the parameters as its concentration levels do not concern health. SD, standard deviation.  analysed samples, 27% had Mn concentration exceeding the permissible limit of 300 µg L −1 , with Bera (697.4 µg L −1 ) and Kuya (594 µg L −1 ) having the highest Mn concentration. The concentrations of Pb varied between 7.6 and 24.8 µg L −1 (mean 12.3 µg L −1 ), with 60% of the analysed samples crossing the maximum acceptable limit of 10 µg L −1 . Cr, Cd, and As, which are highly toxic trace/heavy metals, were within the permissible limits set by BIS 2012. Ni concentration exceeded the maximum acceptable limit of 20 µg L −1 in 27% of the analysed minewater samples.
To identify the impact of mining on the water quality of the region, chemometric methods like PCA (Fig. 3A, B) and HCA were used to differentiate and verify the potential sources of major contaminants in the study region. The 15 locations from where the minewater samples were collected were grouped into clusters based on their physicochemical characteristics using HCA and expressed by a dendrogram (Fig. 4) plotted using XLSTAT 2019.1 software.

Physicochemical characteristics of minewater
The analysis of physicochemical parameters of the pumped-out coal minewater samples from 15 minewater treatment facilities (sampling sites) operating across Jharia Coalfield during the hydrological year 2019-2020 revealed variations in pH ranging from 6.5 to 8.3 with an arithmetic mean of 7.7, suggesting the nature of the coal minewater to be between mildly acidic and alkaline. The electrical conductivity (EC) of the coal minewater samples ranged between 568 and 1389 µS cm −1 (mean 1016 µS cm −1 ) and the TDS concentration ranged between 341 and 953 mg L −1 (mean 674 mg L −1 ). The fluctuation in concentrations of pH, TDS, and EC could be attributed to existing mining conditions in the region, variations in geological formations, and underlying hydrosystems (Singh et al. 2012;Rakotondrabe et al. 2018). Analysis of the ionic concentration of the minewater samples revealed calcium, magnesium (major cations), bicarbonate, and sulphate (major anions) to be the dominant ions responsible for the TDS in the coal minewater of JCF, contributing 30.38, 36.93, 10.50, and 7.70% of the total TDS, respectively. Chloride (6.23%) and sodium (5.31%) were the secondary contributors, while nitrate, potassium, and fluoride collectively accounted for just 1.95% of the TDS. Bicarbonates with concentrations varying between 33 and 442 mg L −1 (mean 220 mg L −1 ) were most dominant in the minewater of Muraidih, Shatabdi, and Pootkee-Balihari Area of JCF, which could be attributed to the sequestration of CO 2 from the soil zone and reaction of silicates in the underlying soil layer with the carbonic acid, causing carbonate dissolution. The decaying of organic matter in the subsurface of the soil zone along with root respiration results in a rise in CO 2 pressure in the subsurface environment, which on contact with percolated rainwater forms bicarbonates (Singh et al. 2012). The sulphate concentration, ranging between 52 and 576 mg L −1 (mean 293 mg L −1 ), showed higher concentration in the minewater of Jamunia, Mudidih, West Mudidh, Ramkanali, Victoria, Bera, and Kuya.  Tiwari et al. 2017). The spatial distribution of chloride concentration was below the maximum acceptable limit of 250 mg L −1 in all the sampling locations, while the fluoride concentration contributed to less than 1% of total anionic balance. Calcium concentration in the coal minewater of JCF varied between 36 and 123 mg L −1 (mean 82 mg L −1 ), contributing 44% of the total cationic concentration overall. Ramkanali (123 mg L −1 ), Victoria (121 mg L −1 ), Bera (113 mg L −1 ), and Kuya (120 mg L −1 ) showed relatively higher concentrations of calcium in comparison to the other sampling locations. This could be the result of weathering and dissolution of underlying rocks comprising limestone (CaCO 3 ) and other calc-silicate minerals such as biotite, pyroxene, and olivine (Tiwari et al. 2016). Magnesium concentration in the analysed minewater samples fluctuated between 19 and 102 mg L −1 (mean 58 mg L −1 ), accounting for 32% of total cations (TZ + ) overall, in equivalent units. West Mudidih (99 mg L −1 ) and Jamunia (102 mg L −1 ) showed the highest magnesium concentration among the sampled locations. The weathering of ferromagnesian minerals such as hornblende, olivine, etc., associated with metamorphic and igneous rocks and dolomite in sedimentary rocks forms the primary magnesium source in the coal minewater of JCF ). Among the cations, potassium was the least dominant in the coal minewater of the study area, accounting for less than 3% of the total cations. The results from analysis of the physico-chemical characteristics of the minewater samples also revealed the major water types of the Jharia Coalfield by plotting a Piper diagram (Fig. 5) using AqQA 1.1.1 software.
Two main water types were identified, viz. Ca-Mg-Cl-SO 4 (red circle) and Ca-Mg-HCO 3 (purple circle), which represent 60% and 33.33% of the coal minewater samples of the study site, respectively. The plotted points in the Piper diagram revealed 93.33% of the coal minewater samples of the study area fell in regions 1 and 2 which indicated the alkaline earth metals (Mg and Ca) to be dominant over the alkali metals (K and Na). The other water type observed was NaCl-SO 4 (orange circle) group, representing the remaining (6.66%) minewater samples of the study site. The predominance of Ca-Mg-Cl-SO 4 and CaMg-HCO 3 agrees with earlier investigations in similar geological environment (Saini et al. 2016;Singh et al. 2016).

Principal component analysis
The distribution of variables of PCA among the minewater quality parameters in the factorial spread F1-F2 revealed three main groups (Fig. 3A). The pH, total alkalinity, and bicarbonates (HCO 3 − ) comprised the first group (encircled in green), representing a strong correlation among the three parameters: pH and total alkalinity (0.654), pH and . This group indicated that changes in pH were a result of the variations in concentration of total alkalinity of the minewater samples which correspond to bicarbonate concentration, predominant in the region. The second group (encircled in blue) consisted of K + , Na + , and Cl − . The leaching of inorganic fertilizers which is composed of potassium chloride (KCl) from nearby croplands serves as evidence for the association of chlorides (Cl − ) with potassium (K + ). Chlorides and sodium are leached from rock formations in the region, along with industrial effluents and sewage contributing to its concentration levels. The third group (encircled in purple) was composed of EC, TDS, TH, Ca 2+ , Mg 2+ , and SO 4 2− . The strong correlation between Ca 2+ and SO 4 2− ions provides evidence on gypsum's role in mineralisation of the minewater, which is consistent with the geology of the region consisting of limestone interbedded with layers of dolomite (Tiwari et al. 2016). The contribution of magnesium ion to the TDS and total hardness in the minewater samples indicates residence time of the underground minewater in the underlying carbonate aquifers due to its slow dissolution kinetics (Martins-Campina et al. 2008).
PCA of the trace/heavy metal concentrations revealed three main groups (Fig. 3 B). The first group (encircled in green) comprising Fe and Ni showed a strong correlation (0.766) which is attributable to the geological formations of the region, indicating a lithogenic source. High Fe concentration could also be due to organic material present in the minewater, leaching from underlying soil/rocks or seasonal variations affecting dissolved oxygen content (e.g. recharge) and in turn Fe concentrations (Fashola et al. 2016). Zn, Cd, Cu, and As formed the second group (encircled in blue). Zn, which is used as a minor additive in gasoline as well as automobile lubricants, is released into the environment during combustion or spillage. Coal combustion caused by mine fires deposits ash rich in these trace/heavy metals on the soil surface which subsequently leach into the groundwater (Masto et al. 2011;Zheng et al. 2013). Occurrence of mine fires in Jharia Coalfield, where the sampling sites are situated, corroborates the trace/heavy metals of the second group belonging to the same source, most likely coal mining wastes which are of anthropogenic origin (Siddiqui et al. 2020). The third group (encircled in purple) comprised Pb and Cr. The incineration of industrial wastes and coal, with its use in fuel tanks, bearings, solder, seals, and wheel weights, contributes to Pb accumulation in the region. The brake linings of vehicles containing asbestos wear down with usage, which along with the usage of catalytic converters represent the source of Cr to be of vehicular origin (Mahato et al. 2017). Mn showed no conspicuous correlation with other trace/heavy metals which could be due to anthropogenic activities.

Hierarchical cluster analysis
The results revealed grouping of the sampling sites into two main clusters, viz. cluster 1 and cluster 2. Two sub-clusters, 1 A and 1 B, constituted cluster 1. Sub-cluster 1A comprised 3 sampling locations, viz. Jamunia, Mudidih, and West Mudidih which showed similar water quality characteristics with relatively higher concentrations of Mg 2+ , SO 4 2− , Fe, Pb, and total hardness. Sub-cluster 1B included 5 sampling locations, viz. Khas Kusunda, Bera, Kuya, Ramkanali, and Victoria. These locations were characterized by higher Fe and SO 4 2− concentrations compared to the other clusters in addition to high levels of Ca 2+ , Mn, and total hardness (Sahu et al. 2017;Adhikari and Mal 2019). Higher Ca 2+ and SO 4 2− ion concentration suggested that the presence of gypsum in the underlying rocks significantly affected the hydrogeological conditions of the locations under sub-cluster 1B. While Ca 2+ and SO 4 2− contributed significantly to the minewater hardness of the locations under sub-cluster 1A, Mg 2+ and SO 4 2− concentrations influenced the minewater hardness of locations under sub-cluster 1B. Cluster 1 represented natural mineralization of the coal minewater and was characterised by higher concentrations of SO 4 2− , Ca 2+ , and Mg 2+ which contribute to its hardness, in addition to higher concentrations of Fe and Mn (Singh et al. 2012;Kumar and Singh 2016;Tiwari et al. 2016). Cluster 2 comprised the sub-clusters 2A and 2B. Three sampling locations, viz. Bastacolla, Kharkharee, and Sinidih, constituted sub-cluster 2A. The minewaters in these locations are characterised by high concentrations of total alkalinity (Singh et al. 2018a, b;Yadav et al. 2021). Sub-cluster 2B comprised 4 sampling locations, viz. Shatabdi, South Tisra, Muraidih, and P.B. Area. This sub-cluster was characterised by high concentrations of total alkalinity, HCO 3 − , NO 3 − , and F − along with Pb Pal and Maiti 2018;Yadav et al. 2021). The increased NO 3 − , Pb, and Ni levels in this group (cluster 2) indicate the influence of anthropogenic

Assessment of potability of minewater
Minewater quality of the study site for the year 2019-2020 was assessed by applying IWQI and HPI (Fig. 6) to the analyses from the aspect of major ion and trace/heavy metal concentration, respectively. The quality of the analysed minewater samples from their respective sampling locations was classified as excellent (< 1), good (1-2), poor (2-3), very poor (3-5), and unsuitable (> 5); and low (< 15), medium (15-30), and high (> 30) based on the mean values of IWQI and HPI, respectively, for the entire year (Table 4).
In this study, Jamunia showed high trace/heavy metal contamination (HPI-50.30), due to high concentrations of Fe (553 µg L −1 ) and Ni (31 µg L −1 ), while the rest of the physiological parameters including all the major ions were within the permissible limits of drinking water standards. From this, it is apparent that installing a water softening unit or iron filters to treat the pumped-out minewater discharges to utilize it for potable purposes is warranted. Based on the results obtained by application of IWQI and HPI, respectively, to the minewater analyses, it was revealed that barring the minewater of Shatabdi (1.97; 25.30), all the other locations discharged minewaters that were unsuitable for direct consumption and required prior treatment (Singh et al. 2011(Singh et al. , 2012Saini et al. 2016  a IWQI value of 1.97, it requires treatment prior to consumption due to high trace/heavy metal concentration (HPI 50.30). The mean HPI values of trace/heavy metal concentrations were found to be 35.62 for the pre-monsoon month of March-April and 31.09 for the post-monsoon month of October-November, which fall below the critical pollution index limit of 100. Minewater quality during post-monsoon was found to be better in terms of trace/heavy metal concentration Singh and Kamal 2017;Chaturvedi et al. 2018). The spatial variation in minewater quality of the entire study area for the year 2019-2020 is represented by interpolation of the sampled data using IDW technique in ArcGIS 10.2.2. The interpolation of IWQI (Fig. 7) and HPI (Fig. 8) values of analysed minewater could potentially enable efficient water resource management by assessing the water quality of JCF.
The integrated approach of combining both IWQI and HPI for qualitative studies of minewater could prove to be efficient in characterization and classification of water for its utilization in various purposes. This method helps in making quick decisions about the destination of pumped-out minewater discharges and its subsequent treatment (if any) to improve the quality of the water. A preliminary conceptual model demonstrating the water quality assessment process by using the integrated approach of IWQI, HPI, and multivariate statistics is depicted in Fig. 9.

Conclusion
The integrated approach for water quality assessment checks the limitations of eclipsing, aggregation, and ambiguity posed by traditional water quality indices, considering the maintenance of ionic balance in potable water, and thus representing optimal water quality as per the standard guidelines with respect to human health. The integrated model incorporates percent deficit in its calculations which makes the users aware of the pollution levels reaching the threshold limit, giving enough time to address the situation, thus, rendering it flexible in its usage as the percent deficit can be adjusted as per the nature of the environment. This would facilitate quicker operational responses to different types of pollution events and protect the water plant installations. This attribute would also allow for a wide applicability across different regions around the world. Water management strategies could be greatly helped by the integrated approach as it would lead to quick decision making about the destination of the pumped-out coal minewater or the type of treatment to be employed to improve the quality of the minewater. PCA of minewater quality is an efficient chemometric tool to explore the relationship between physicochemical water quality parameters and the geological domain. In the current study, PCA revealed the trace/heavy metal concentration in pumpedout underground minewater to be from a mixed source of anthropogenic and lithogenic origin, highlighting the impact of mining activities on groundwater. The results of the study suggest that multivariate statistical processes using PCA and HCA could be used to improve water monitoring strategies and management of water resources. The multivariate statistical approach holistically identifies areas of contamination, giving an assessment of the quality of minewater and its suitability for potable purposes and laying the framework for future investigations in augmenting minewater for potable purposes.
Acknowledgements The authors are grateful to Indian Institute Technology (Indian School of Mines) Dhanbad for providing all the necessary laboratory facilities during the research work. The authors also express their gratitude towards anonymous reviewers for their constructive comments which helped in strengthening the quality of the manuscript.
Author contribution Gurdeep Singh supervised the work. Pritam Mazinder Baruah performed the analysis and calculations and wrote the manuscript in consultation with Gurdeep Singh.

Data availability
The data that support the findings of this study are available from the corresponding author upon request. Fig. 9 Conceptual model demonstrating the water quality assessment by using the integrated approach of IWQI, HPI, and multivariate statistics Declarations Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare no competing interests.