Assessing and modelling drinking water quality at the railway stations of Tripura, India, with a possible strategic solution

Abstract

Availability of safe drinking water at the railway stations is a basic passenger amenity. To maintain the quality of drinking water, the Ministry of Railways, Government of India, follow the Uniform Drinking Water Quality Monitoring Protocol, 2013, with a benchmarking standard of Bureau of Indian Standards, 2012. The quality of drinking water varies from state to state, including station to station, due to the diversified regional structure of the country. The regional structure of Tripura, Northeastern state of India, is unique for its Tilla-Lunga topography, where 27 railway stations provide passenger services, including safe drinking water of heterogeneous quality. The present study aims to assess and develop a model of drinking water quality at the railway stations of Tripura, India, with a possible strategic solution. Multivariate statistical tools like multiple regression analysis, principal component analysis, and cluster analysis have been used. Weighted Arithmetic Water Quality Index (WAWQI) and Entropy Weighted Water Quality Index (EWWQI) have been applied to assess the drinking water quality of the railway stations of Tripura. It has been observed that several physical (r = 0.42, p-value = 0.064), chemical (r = 0.31, p-value = 0.047), and biological (r = 0.47, p-value = 0.077) parameters of drinking water are positively correlated with each other, whereas chloride (λ = 2.98), E. coli (λ = 2.13), turbidity (λ = 1.72), fluoride (λ = 1.49), and iron (λ = 1.08) are most significant water quality parameters in the railway stations of Tripura. The WAWQI and EWWQI classified the drinking water of 55.55% and 85.19% of railway stations as good for drinking, respectively. Drinking water quality can improve 55.55% to 81.48% by removing iron content from the water. Physiography and land use patterns of surrounding areas of railway stations influence drinking water quality. Elevation and urbanisation have significant negative correlation (− 0.438 and − 0.441 at 5% level) with drinking water quality. Policy-practice interference has been analysed, and passenger satisfaction with drinking water has been measured for the bottom-up planning approach. Through content analysis, a few technical suggestions have been made to improve the physio-chemical and biological properties of drinking water at the railway stations of Tripura.

Introduction

Water is the fundamental key to life (Chowdhury et al. 2022). The rapid growth of population, industrialisation, and anthropogenic activities have resulted in poor drinking water quality (Ramapriya and Elango 2022; Darko et al. 2021; Faroque and South 2021). The availability of safe drinking water is one of the minimum passenger amenities at the railway stations in India (Pravithra 2018). Indian railway is the fourth largest (65,000 km) railway network with 7083 railway stations after the USA (250,000 km), China (100,000 km), and Russia (85,500 km) (Mitra and Roy 2020). About 22 million passengers every day travelled through railways during the pre-pandemic era. India went into complete lockdown from March 25, 2020 due to COVID-19 outbreak (Pawar et al. 2020). In the fiscal year 2021, it has been reported that 1.3 billion passengers dropped across India (Sun 2022). Passengers’ perceptions of safety, security, comfortability, and higher speed compare to other modes of transportation; cost-effectiveness leads a considerable number of passengers to travel by train in India (Roy et al. 2019; Hagen et al. 2017; Akiyama and Okushima 2009). In 2018, during a public interest litigation (PIL) hearing, Delhi High Court was told to provide drinkable water to the railway passengers (Press Trust of India 2018). Even before that, it was reported that about 50% of the chlorination plants of Northern Railway (NR) had not been working for the past several years, including the major railway stations like Lucknow and Gaziabad (Kumar 2015). During that time, the judicial system expressed concern over the quality of water supplied to passengers at different railway stations and gave the verdict to collect water samples from different railway stations and sought a report on the quality of drinking water (Iqbal 2015). As per the Indian Constitution (Article 21), the right to intake clean water is one of the fundamental right of all Indian citizens protected under the Safe Drinking Water Act, 1996 (Kothari 2006). In response to the judicial system and welfare of the people of India, the Indian Railways (IR) ensure to provide safe drinking water to the passengers (Press Information Bureau 2019). However, the drinking water quality is mainly related to the water source (Carrard et al. 2020). The primary source of drinking water in the railway stations in India is ground water (Ministry of Railway 2017; Pandey 2015; Sahni et al. 2010). The ground water quality varies from region to region due to the geological structure, location, extension and quality of ground water table, the leaching of minerals, contamination of water due to agriculture, and other anthropogenic activities (Khatri and Tyagi 2014; Rango et al. 2012; Alexakis 2011). Rodvang et al. (2004) explain the impact of animal farming operations on ground water quality with spatio-temporal variations. Land use and land cover (LULC) of an area significantly control the water quality (Zeinalzadeh and Rezaei 2017; Wilson and Weng 2010). Physiography and climatic factors also affected the ground water quality. The physiography and climate of Tripura are unique, where 27 operational railway stations are situated. The state has mainly glass sands, limestone, plastic clay, and hard rock. The most significant resources in the state are oil and natural gas (Dikshit and Dikshit 2014). Due to leaching activities, ground water becomes non-potable because of contamination of heavy metals like iron (Fe), arsenic (As), cadmium (Cd), chromium, and fluoride (F⁻). Contamination of that toxic pollutant is increasing daily due to the rapid growth of population, urbanisation, industrial development, and infrastructural intervention. Recently, a number of research work has been done on water quality in Tripura (e.g. Barman et al. 2022; Karmakar and Singh 2021; Roy et al. 2021; Roy and Manna 2021; Paul et al. 2019; Singh and Kumar 2015; Lodh et al. 2014; Sarkar and Mishra 2014; Banerjee et al. 2011). Based on the bibliometric analysis, maximum research follows water quality parameters based on micro-spatial scale. However, no research was found on the macro-regional level of drinking water quality-driven policy intervention.

The main objective of the study was to assess the quality of drinking water of the railway stations in Tripura, India, and to provide a strategic technical solution for policy intervention.

Study area

The research has been conducted in Tripura, North-eastern India, which covers 10,486 sq. km geographical area. Bangladesh encircles the landlock state of Tripura to the north, south, and west (856 km) and the Indian states Assam (53 km) and Mizoram (109 km) to the northeast and east, respectively (Roy and Mitra 2020). Tripura is extravagantly a hilly state with valley and ridge (tilla and lunga) topography (Sen et al. 2015). About 60% of its land is hilly, while the remaining 40% is plain land. Even the plain land is not a dead-level land. Instead, many low hills and tillas break it by 30–60 m in elevation, covered with trees and shrubs. The state has six anticlinal hill ranges: Baramura, Atharamura, Longtharai, Shakhan, Jampui, and Deotamura. The railway track crosses the major hills Longtharai (515 m), Atharamura (481 m), and Baramura (249 m) from north to South (Roy and Mitra 2015). Many rivers, narrow streams, gullies, and ravines originated from those hills. The state has 11 prominent rivers named Longai (98 km), Juri, Deo (98 km), Manu (167 km), Dhalai (117 km), Khowai (70 km), Haora (53 km), Bijoy (26), Gomati, Muhuri (64 km), and Feni from north to south. The road network of Tripura is deplorable during monsoon; landslides are a common phenomenon in the hilly terrain of this region (Sen et al. 2013). In this circumstances, railway has become the prime mode of transportation in Tripura (Roy and Mitra 2016 a). According to the Census of India 2011, the state’s total population is about 3,673,917, where population density is 350. The net sown area of the state is only 2077.2 sq. km (Department of Agriculture, Cooperation & Farmers Welfare Ministry of Agriculture & Farmers Welfare 2018). Due to the high man-land ratio, forest cover and nature of undulating plateau enormous pressure promulgate the agricultural land. As a result, Tripura became a food-deficit state and the need to import food and other necessary materials from other states of India (De 2004). The cost of transport by road is much higher than by railway. The railway is always a more dependable mode of transportation, especially in hilly regions like Tripura, where during monsoon roads deteriorate due to landslides in the hill ranges of Tripura. In Tripura, since 2016 (after gauge conversion), railways have been playing a significant role in freight transport (Singh 2016). The state has about 264-km-long railway track from 27 railway stations. The railway line extended from Churaibari Railway Station [CBZ] (24°26ʹ N. and 92°14ʹ E.) in north to Sabroom Railway Station [SBRM] (23°00′ N. and 91°41′ E.) in south. There are twenty-five intermediate stations namely Nadiapur [NPU] (24°23ʹ N. and 92°12ʹ E.), Dharmanagar [DMR] (24°22ʹ N. and 92°10ʹ E.), Panisagar [PASG] (24°16ʹ N. and 92°09ʹ E.), Pencharthal [PEC] (24°11ʹ N. and 92°06ʹ E.), Kumarghat [KUGT] (24°09ʹ N. and 92°02ʹ E.), Nalkata [NLKT] (24°03ʹ N. and 92°00ʹ E.), Manu [MANU] (23°59ʹ N. and 91°59ʹ E.), S.K. Para [SKAP] (23°58ʹ N. and 91°8ʹ E.), Jawaharnagar [JWNR] (23°55ʹ N. and 91°54ʹ E.), Ambassa [ABSA] (23°55ʹ N. and 91°51ʹ E.), Mungiakami [MGKM] (23°53ʹ N. and 91°42ʹ E.), Teliamura [TLMR] (23°51ʹ N. and 91°37ʹ E.), Jirania [JRNA] (23°49ʹ N. and 91°25ʹ E.), Jogendranagar [JGNR] (23°48ʹN. and 91°18ʹE.), Agartala [AGTL] (23°47ʹN. and 91°16ʹE.), Sekerkote [SKKE] (23°44′ N. and 91°16′E.), Bishalgarh [BLGH] (23°40′ N. and 91°16′ E.), Bishramganj [BHRM] (23°35′ N. and 91°21′ E.), Udaipur [UDPU] (23°30′ N and 91°28′ E.), Garjee [JRJE] (23°25′ N. and 91°29′ E.), Santirbazar [STRB] (23°19′ N. and 91°31′ E.) Belonia railway station [BENA] (23°14′ N. and 91°29′ E.), Jolaibari [JLBRI] (23°11′ N. and 91°35′ E.), Thailik Twisa [THTW] (23°7′ N. and 91°36′ E.), and Manu Bazar [MUBR] (23° 3′ N. and 91°38′ E.) railway station (Fig. 1).

Fig. 1

Location map of the study area (source: prepared by the authors; data extracted from Handheld GPS receiver, Google Earth Pro and SAS Planet v. 210,805)

Methods and materials

Sampling and analytical procedure

The sample has been collected from all 27 railway stations of Tripura (Fig. 1). Water samples has been stored in 1-L HDPE containers [flush completely with distilled water and nitric acid (HNO₃)] from 27 reported railway stations of Tripura and 13 parameters were investigated as per guideline of UDWMQP to test its fitness for drinking. To obtain accurate results, physical parameters like pH, TDS, and turbidity were tested on-site immediately after sampling. Samples were conserved, adding 1:1 nitric acid (HNO₃) for analysis; chemical parameters include total hardness, alkalinity, chloride, sulphate, fluoride, nitrate, arsenic and iron. biological parameters like total coliform and E. coli were characterised in the laboratory using different methods (Table 1).

Table 1 Methods of analysis and BIS standard limits of physio-chemical and biological parameters

There are several standards available for drinking water monitoring like the American Public Health Association (APHA), World Health Organization (WHO), Bureau of Indian Standards (BSI), Central Pollution Control Board (CPCB), and Indian Council of Medical Research (ICMR). The Bureau of Indian Standards (BIS) has specified drinking water quality standards in India to provide safe drinking water to the people. BIS (IS_10500 and revised module IS 10500:2012) has Uniform Drinking Water Quality Monitoring Protocol specifications. This standard has been considered for benchmarking mainly because the Indian Railway follows BIS standards for drinking water monitoring.

Multivariate statistical analysis

Pearson product-moment correlation coefficient (PPMCC) (Pearson’s r) is calculated statistically and its value may be positive or negative and may range from − 1 to + 1. The r values of + 1 and − 1 indicate strong and inverse linear relationships between variables (Roy et al. 2021). Canonical correlation analysis (CCA) is used to identify and measure the associations among drinking water’s physical, chemical, and biological properties. CCA determines a set of canonical variates, orthogonal linear combinations of the variables which best explain the variability between the sets and within the sets (Hardoon et al. 2003). We have three variables, i.e. physical variables (pH, turbidity, TDS), chemical variables (chloride, sulphate, fluoride, nitrate), and biological variables (total coliform and E. coli). Principal component analysis (PCA) is a tool to decide on an independent variable by taking out exceptionally related parameters. Kaiser–Meyer–Olkin (KMO) and Bartlett’s sphericity test have been executed to confirm whether PCA analysis is reasonable or not for the present experimental dataset. Bartlett’s sphericity test has been tested to acquire the correlation matrix’s nature, whether this is a unique matrix or not (Hill 2011).

Cluster analysis (CA) has been interpreted in group monitoring sites to determine whether each sample had similar physio-chemical and biological characteristics. In CA, every sample from individual clusters and the pair of clusters combined relies on similarity (measured by squared Euclidean distance) and the linkage technique (Ward’s method). The dendrogram outline demonstrates the outcomes, providing a basic image of the cluster (groups).

Minitab v.18.1 and SPSS v. 26 were used for different statistical analyses. Origin Pro v.19 and Sigma Plot software have been used for graphical representation of the results.

Calculation of water quality index (WQI)

WQI values of 27 samples were estimated by taking 13 variables to obtain their fitness for drinking. To calculate WQI, two methods (weighted arithmetic water quality index and entropy weighted water quality index) have been employed.

Calculation of weighted arithmetic water quality index (WAWQI)

Horton (1965) developed the weighted arithmetic water quality index (WAWQI) model and Brown et al. (1970) modified it. Several researchers (Mahammad and Islam 2021; Menberu et al. 2021; Ayogu et al. 2020; Roy and Manna 2021; Adimalla and Qian 2019) used this method to calculate WAWQI, which has been computed using the following steps:

$${w}_{j}^{^{\prime}}=\frac{K}{{S}_{j}}$$

(1)

${S}_{j}$ is the corresponding parameter of jth variables.

K is the proportionality constant which can be calculated as follows:

$$K=\frac{1}{\sum_{j=1}^{n}\frac{1}{{S}_{j}}}$$

(2)

$${q}_{j}=\frac{{C}_{j}}{{S}_{j}}\times 100$$

(3)

where ${q}_{j}$ is the ratio of the analysed value of jth parameter (${C}_{j}$) to its Bureau of Indian Standards (BIS), 2012, recommended standard value (${S}_{j}$).

$$WAWQI=\frac{\sum_{j=1}^{n}{w}_{j}^{^{\prime}}\times {q}_{j}}{{\sum }_{j=1}^{n}{w}_{j}^{^{\prime}}}$$

(4)

Calculation of entropy weighted water quality index (EWWQI)

To accumulate these thirteen drinking water parameters in a linear combination, composite entropy weighted water quality index (EWWQI) has been proposed, and the respective coefficients (weights) of those indices are calculated by Shannon’s entropy (Shannon 1948). For determining the weightage value of each of the all thirteen drinking water variables, the following Shannon entropy formula has been used:

$${P}_{ij}=\frac{{x}_{ij}}{{\sum }_{i=1}^{m}{x}_{ij}}$$

(5)

where m represents the number of sampling areas (railway stations) and P_ij is the proportional occurrence of the value (x_ij) corresponds to jth evaluated drinking water variables (index) in ith sample (railway station), calculated by using the following equation:

$${E}_{j}=-k\sum_{i=1}^{m}{p}_{ij}\mathrm{ln}{p}_{ij}$$

(6)

where K is the proportionality constant and measured as

$$k={~}^{1}\!\left/ \!{~}_{\mathrm{ln}\left(m\right)}\right.$$

Corresponding weight of the jth variables (index), denoted by ${W}_{j}$, for j = 1,2,…,s, is evaluated as

$${W}_{j}=\frac{1-{E}_{j}}{{\sum }_{j=1}^{n}\left(1-{E}_{j}\right)}$$

(7)

Finally, EWWQI for each station “j” ($EWWQI$) has been calculated by using the following formula:

$${EWWQI}_{j}=\sum_{j=1}^{n}{w}_{j}\times {q}_{j}$$

(8)

Based on EWWQI, drinking water of the railway station is classified into five different categories (Adimalla 2021; Roy et al. 2021; Egbueri et al. 2020; Amiri et al. 2014).

Causality analysis

The heterogeneous drinking water quality has been accessed through the Remote Sensing and Geographical Information System (GIS) technique. Geological information of the study area (railway stations) has been extracted from the Bhukosh portal of the Ministry of Geology, Government of India. Alos pulsar radiometric corrected digital elevation model (DEM) data with 12.5 m resolution has been collected from NASA Earth data source. The hydrological model of the study area has been prepared from DEM data. A soil map has been prepared from Food and Agriculture Organisation Soil Portal. Infrastructural facilities and passengers’ movement of the station has been analysed to understand the causal interference of developmental factors. Based on infrastructural data, calculate composite infrastructure index (CII). CII_j for station “j” has been calculated by using the following formula:

$${CII}_{j}=\sum_{k=1}^{v}\frac{{\partial }_{kj}}{v}, for j=\mathrm{1,2}\dots \dots \dots s$$

where s = number of stations and v = number of infrastructural parameters,

$${\partial }_{kj}=\frac{{X}_{kj}-{X}_{k\left(1\right)}}{{X}_{k\left(n\right)}-{X}_{k\left(1\right)}},$$

X_kj=jth observational value of corresponding infrastructural parameter X_k,

X_k(1)=minimum value of corresponding infrastructural parameter X_k,

X_k(n)=maximum value of corresponding infrastructural parameter X_k.

Remarks: Range of ${CII}_{j}$ is [0,1]

Proof: Let us consider a state “M,” which consists of the highest value of observation for each of the corresponding “v” infrastructural parameters, i.e. ${X}_{kM}={X}_{k\left(n\right)}$ , for k = 1,2,…,v. So, ${\partial }_{kM}=\frac{{X}_{k\left(n\right)}-{X}_{k\left(1\right)}}{{X}_{k\left(n\right)}-{X}_{k\left(1\right)}}$ = 1 for k = 1, 2,…,v, hence ${CII}_{M}$= $\sum_{k=1}^{v}\frac{1}{v}=\frac{v}{v}=1$. Similarly, we may consider another state say “m,” which always consists of only the lowest value of observation for each of the corresponding “v” infrastructural parameters (${X}_{km}={X}_{k\left(1\right)}$, for all k). So ${\partial }_{km}=0$, hence ${CII}_{j}$= $\sum_{k=1}^{v}\frac{0}{v}=0$.

Pearson correlation coefficient among CII, WAWQI, and EWWQI has been calculated. Correlation between average monthly passengers’ mobility from the railway station also assessed with WAWQI and EWWQI.

Landuse map of the state has been prepared through Sentinal-2 satellite image with five land-use classes, i.e. agricultural land, built-up area, fallow land, vegetation cover, and water bodies. We combined the land-use classes into a unit called weighted percentage index (WPI).

$$WPI=\left({p}_{1}\times w\right)+\left({p}_{2}\times w\right)+\left({p}_{3}\times w\right)+\left({p}_{4}\times w\right)+\left({p}_{5}\times w\right)$$

where

p ₁ :

percentage of agricultural land

p ₂ :

percentege of built up area

p ₃ :

percentege of fallow land

p ₄ :

percentege of vegetation cover

p ₅ :

weigth of LULC classes

It has been considered total weight (TW = 1), i.e. the equal weightage (w = 0.20) has been given to all land-use classes. Kappa coefficient has been calibrated for accuracy assessment (Tallón-Ballesteros and Riquelme 2014). To understand the impact of urbanisation and forest cover, normal difference built-up index (NDBI) and normal difference vegetation index (NDVI) have been calibrated (Verhoeven and Dedoussi 2022; Gandhi et al. 2015), and final assess the associations of those factors with WAWQI and EWWQI.

Cartographic representation has been made in ArcMap v. 10.8, Global Mapper v.23, and Map Info v.17.

The strategic suggestion has been drafted based on content analysis and policy review. Long-term, medium-term, and short-term strategies explain the policy gap and alternative solutions for better drinking water.

Results and discussion

Analysis of physio-chemical and biological parameters

The most common drinking water variable which has been measured in nature (acidic or basic) of water is pH. The pH scale varies between 0 to 14. pH value 7 represents the neutral water, lower than 7 signifies that water is acidic and pH value of more than 7 indicates alkaline water. According to Amendment No. 1 in June 2015 to IS 10500:2012 drinking water specifications (Second Revision), acceptable range of pH is 6.50 to 8.50. It has been observed that the pH value of Churaibari (6.06), Nadiapur (6.47), Manu (6.35), S.K. Para (6.27), Ambassa (6.25), Mungiakami (6.33), and Thailik Twisa (6.09) railway stations are less than 6.50, which reveals that the water of those stations is acidic in nature (Fig. 2a). For the remaining 20 stations, the pH value of water is within an acceptable limit.

Fig. 2

a pH and b turbidity of drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

It has been found that only 18.51% of railway stations (4 railway stations out of 27 railway stations) have the acceptable limit of turbidity (0NTU-1NTU) in the drinking water as per the standard Bureau of Indian Standards 2012 (IS 10500:2012). However, the permissible water limit in stations has been found to be 59.25% (Fig. 2b). The remaining 22.23% of railway stations have very high turbidity in drinking water. Manu, Santirbazar, S.K. Para, and Nalkata railway stations have more than 15 NTU turbidity in drinking water, which is not fit for drinking (Fig. 2b). A high level of turbidity in drinking water might associate with endemic gastrointestinal illness (Mann et al. 2007; Schwartz et al. 2000).

As per Amendment No. 1 June 2015 to IS 10500:2012 drinking water specifications (Second Revision), acceptable range of total dissolve solid (TDS) is 500 mg/L and the permissible limit is 2000 mg/L. It has been found that the drinking water of all railway stations does not exceed the acceptable limit (Fig. 3a).

Fig. 3

a TDS and b total hardness of drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

Total hardness is generally expressed in milligrams of calcium carbonate equivalent per litre water. According to Amendment No. 1 June 2015 to IS 10500:2012 drinking water specifications (Second Revision), acceptable range of total hardness is 200 mg/L, and the permissible limit is 600 mg/L. The drinking water of all the railway stations of Tripura is below the accepted limit of BIS standard. Water containing calcium carbonate at concentrations below 60 mg/L is usually considered soft water; 60–120 mg/L, moderately hard water; 120–180 mg/L, hard; and more than 180 mg/L, very hard water (McGowen et al. 2001). Churaibari (32.67), Nadiapur (28.00 mg/L), Dharmanagar (53.33 mg/L), Panisagar (54.00 mg/L), Kumarghat (23.33 mg/L), Manu (44.67 mg/L), S.K. Para (46.67 mg/L), Teliamura (55.33 mg/L), Jirania (45.33 mg/L), Jogendranagar (21.33 mg/L), Agartala (36.00 mg/L), Sekerkote (52.67 mg/L), Bishramganj (42.00 mg/L), Udaipur (46.00 mg/L), Belonia (25.33 mg/L), Jolaibari (37.33 mg/L), Thailik Twisa (18.67 mg/L), and Sabroom (36.67 mg/L) railway station have soft drinking water. It has been observed that only the Pencharthal railway station has hard drinking water (Fig. 4).

Fig. 4

Heavy materials in the drinking water of Pencharthal railway station

Field emission scanning electron microscopy (FESEM) and energy dispersive X-ray (EDx) are the most advanced technology to detect and analyse element morphology and properties, which has been used for all the railway stations of Tripura to identify heavy materials in drinking water provided to the passengers, and it has been found that magnesium (Mg), calcium (Ca), sodium (Na) fluorine (F), silicon (Si), chlorine (Cl), potassium (K), and iron (Fe) are the common elements in the drinking water of railway stations of Tripura. The drinking water of Pencharthal railway station has relatively higher hardness (Fig. 5), and less hardness has been found in the drinking water of Thailik Twisa railway station (Fig. 5).

Fig. 5

Heavy materials on the drinking water of Thailik Twisa railway station

Hard water positively impacts human health because it contains calcium and magnesium, which helps protect against cardiovascular disease (Akram and Rehman 2018). There is also some evidence that calcium and magnesium in drinking water may help protect against gastric, colon, rectal cancer, and pancreatic cancer, and that magnesium may help protect against oesophageal and ovarian cancer (McVean 2019). Sometimes very hard water may be hazardous to human bodies (Sengupta 2013). Koufman and Johnston (2012) found that drinking of natural carbonated alkaline water with a pH of 8.8 may help deactivate pepsin, the primary enzyme that causes acid reflux. Drinking alkaline ionised water may benefit people with high blood pressure, diabetes, and high cholesterol (Yu Lian 2006).

The alkalinity illustrates the determination of water’s capability to nullify the acids. The alkalinity of water may be due to the presence of one or more ions. These include hydroxides, carbonates, and bicarbonates. As per IS 10500:2012, 3025 (Part 23) accepted limit of alkalinity is 200 mg/L and the permissible limit is 600 mg/L. It has been observed that the average alkalinity of drinking water of railway stations in Tripura is 33.457 mg/L, which is far below than the Bureau of Indian Standards (BIS), 2012. Maximum alkalinity (66.67 mg/L) was found in the Pencharthal railway station (Fig. 6a). At the same time, minimum alkalinity (12.67 mg/L) was observed in the S. K. Para railway station, followed by Jogendranagar (14.00 mg/L) and Thailik Twisa (14.67 mg/L), respectively (Table 1). The range of alkalinity is 54 mg/L in the drinking water of railway stations of Tripura. Comparatively, more alkaline drinking water was found in the Pencharthal (66.67 mg/L), Jirania (56.67 mg/L), and Jawaharnagar (56.00 mg/L), Santirbazar (51.33 mg/L), and Ambassa (50.00 mg/L). Kurtosis value (− 0.77) with 0.872% standard error depicts that the degree of the tailedness thickens in the alkalinity probability distribution.

Fig. 6

a Alkalinity and b chloride of drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

Sulphate is one of the primary dissolved components of water. Different natural and anthropogenic factors are responsible for sulphate influx in drinking water like sewage treatment plants, industrial discharge, and agricultural fertiliser (Wang and Zhang 2019; Miao et al. 2012). This study estimates the risk level for sulphate in the drinking water of railway stations in Tripura. According to BIS Standard, 2012, the acceptable limit of sulphate in drinking water is 200 mg/L and the permissible limit is 400 mg/L, respectively. The average value of sulphate in the drinking water of the railway station in Tripura is 63.44 mg/L (Table 2). It has been observed that the sulphate level in the drinking water of different railway stations in Tripura is far below than the Bureau of Indian Standard defined acceptable limit for consumption (Fig. 7a). The maximum concentration of sulphate is found in the Jogendranagar (94.00 mg/L), Nadiapur (89.00 mg/L), and Garjee (85.00 mg/L) railway stations. The result of sulphate depicts that the water of railway stations of Tripura is fit for drinking.

Table 2 Descriptive statistics of drinking water variables of different railway stations in Tripura

Fig. 7

a Sulphate and b nitrate of drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

Nitrate is a polyatomic ion generally found in surface water (Howe 2020). It is a joint compound of fertiliser, and from the agricultural field, nitrate has contaminated ground water, which is the primary source of drinking water in the railway stations of Tripura. It has been observed that the average nitrate value in the drinking water of the railway stations of Tripura is about 24.489 mg/L The acceptable limit of nitrate in the drinking water is 45 mg/L (Fig. 7b). No railway stations in the state were found with drinking water above the acceptable nitrate limit. The result indicates that the water of railway stations of Tripura is drinkable with respect to nitrate concentration. Nevertheless, agricultural practice and close proximity to urban centres influence the concentration of nitrate. Nitrate in about 48.15% of railway stations like Jogendranagar, Jirania, Churaibari, Bishramganj, Manu, Teliamura, Kumarghat, Nadiapur Dharmanagar, Garjee, Manu Bazar, Panisagar, and Jawaharnagar have been found with above mediocre value (Fig. 7b). Those railway stations located in a forested area or far from urban centres like Thailik Twisa, S. K. Para, Nalkata, Mungiakami, Udaipur, Sabroom, and Belonia have a comparatively low concentration of nitrate in drinking water (Fig. 7b).

Fluoride is the world’s 13th most abundant element and constitutes 0.08% of the earth surface (Peckham and Awofeso 2014; Hardwick et al. 2000). Fluoride is an inorganic, monatomic anion of fluorine that generally exists at low concentrations in most fresh and saltwater sources and might be available in rainwater. As per IS 10500:2012, 3025 (Part 60) accepted limit of fluoride is 1 mg/L and the permissible limit is 1.5 mg/L (Bureau of Indian Standards 2012). A high level of fluoride in drinking water may cause dental, skeletal, and neuro problems (Balasubramanian and Umarani 2012). It has been observed that the average level of fluoride in the drinking water of the railway stations of Tripura is 0.682 mg/L, with a standard deviation of 0.106. The maximum concentration of fluoride was found in the Panisagar railway station (0.960 mg/L), followed by Churaibari railway station (0.890 mg/L) and Pencharthal railway station (0.870 mg/L). The lowest fluoride concentration was found in the Nadiapur railway station (0.500 mg/L). Water of about 51.85% (14) railway stations of the state, i.e. Nadiapur, Bishramganj, Udaipur, Santirbazar, Jolaibari, Garjee, Ambassa, Belonia, Sabroom, Bishalgarh, Manu, S. K. Para, Nalkata, and Mungiakami, have been observed with a low (below average) concentration of fluoride (Fig. 8). About 64.28% (8) railway stations of this category are surrounded by forest areas where the level of fluoride is relatively low, indicating the antagonistic relation between fluoride and forest cover.

Fig. 8

Fluoride of drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

The World Health Organisation (WHO) has classified a group of 10 chemicals of primary public health concern and each of them has conveyed risk controlling endorsements; those are air pollution, arsenic (H₃AsO₄), asbestos (Mg₃Si₂O₅(OH)₄), benzene (C₆H₆), cadmium (Cd⁺²), dioxin (C₄H₄O₂), and dioxin-like substances, water with inadequate or excess fluoride (F −), lead (Pb⁺²), mercury (Hg), and very hazardous pesticides (WHO 2011). Arsenic is a metalloid element that forms several poisonous compounds. It is widely distributed throughout the earth’s crust and can be released into the atmosphere and water through natural and human activities (James et al. 2017; Bhatia et al. 2014; Kapp 2005). Many health effects, like skin lesions, peripheral neuropathy, gastrointestinal symptoms, diabetes, renal disorders, cardiovascular disease, and cancer, have been linked to arsenic contamination (Kumar et al. 2021). However, the signs and symptoms can take years to develop depending on exposure (Handmarsh et al. 2002; WHO 2010).

It has been observed that arsenic at a microscopic level is present in the drinking water of railway stations of Tripura (Fig. 9a). The level of arsenic in the drinking water of railway stations does not violate the BIS standard, 2012, where the acceptable limit of drinking water is 0.01 mg/L and the permissible limit is 0.05 mg/L (Fig. 9a). The average value of arsenic-contaminated water is 0.000097592592593 mg/L, far lesser than the BIS standard. Maximum arsenic-contaminated water (0.00021 mg/L) was found in Ambassa railway station and a minimum concentration (0.000015 mg/L) of arsenic was found in Sabroom railway station (Fig. 9a).

Fig. 9

a Arsenic and b iron on drinking water in different railway stations of Tripura (source: prepared by the authors, 2022)

Iron is the second most abundant metal in the earth’s crust, which accounts for about 5%. Elemental iron is rarely found in nature, as the iron ions Fe²⁺ and Fe³⁺ readily combine with oxygen- and sulphur-containing compounds to form oxides, hydroxides, carbonates, and sulphides. Iron is most commonly found in nature in the form of its oxides (WHO 1996). Iron is not hazardous to health, but it is considered a secondary or aesthetic contaminant. Essential for good health, iron helps transport oxygen in the blood. According to Amendment No. 1 June 2015 to IS 10500:2012 drinking water specifications (Second Revision), acceptable value of iron is 0.3 mg/L.

It has been observed that the average iron contamination in drinking water of the railway stations in Tripura is 0.376 mg/L which is beyond the acceptable limit of drinking water as per BIS recommendation (Table 1). The maximum concentration of iron found in the drinking water of the Nadiapur railway station (1.45 mg/L), followed by Jirania (1.01 mg/L) and Sabroom (0.99 mg/L) railway stations. About 44.44% railway stations, namely Nadiapur, Jirania, Sabroom, Panisagar, Manu, Jawaharnagar, Ambassa, Udaipur, Bishalgarh, S. K. Para, Manu Bazar, and Kumarghat, have been identified with a higher concentration of iron (Fig. 9b). Minimum concentration of iron was found in Thailik Twisa railway station (0.05 mg/L). The result depicts that about 66.66% of railway stations have drinkable water in terms of iron content.

A smelly biofilm can form when iron exists and certain kinds of bacteria. The bacteria use the iron to survive, leaving behind a reddish-brown or yellow slime. Two bacteria, namely total coliforms and E. coli, have been tested.

Total coliforms are an assemblage form of different types of gram-negative anaerobic bacteria with different attributes (Payus and Nandini 2014). Generally, total coliforms are found in the environment, such as in soil and the intestines of animals, including humans (Olstadt et al. 2007). The primary sources of total coliforms in water are infected human and animal waste. Another reason for total coliforms in drinking water is the failure to maintain a drinking water supply system like wells, tanks, chlorinators, and equipment malfunction (Elmund et al. 1999). Coliform bacteria are rod-shaped gram-negative non-endospore-forming bacteria that ferment lactose with the production of acid and gas at 35–37 °C. A laboratory experiment has been conducted to test the presence of coliform in water and it has been found that total coliform is higher in the water of S.K. Para, Jawaharnagar, Nalkata, Nadiapur, and Jolaibari railway stations.

As per IS 10500:2012, clause 4.1.1, coliform shall not be detectable in any 100-mL sample. It has been observed that the drinking water of all railway stations of Tripura has a significant number of coliform bacteria (Fig. 10).

Fig. 10

Total coliform growth in the drinking water in different railway stations of Tripura (source: experimented by the authors, 2022)

It has been observed that only the drinking water of the Sabroom railway station (0 CFU/L) has followed the BIS recommendation on total coliform. The maximum concentration of total coliform was found in Jawaharnagar (2613 CFU/mL) railway station, followed by Bishramganj (1500 CFU/mL) and Teliamura (1345 CFU/mL) railway stations. Comparatively low concentrations of total coliform were found in Churaibari, Panisagar, Manu, Dharmanagar, Mungiakami, Kumarghat, Agartala, and Bishalgarh (Fig. 10). The result shows that a clean source of drinking water does not guarantee safe and drinkable water at the point of consumption. Probably, to remove the coliform drinking water in the railway station, another extra treatment is required before sending to tap for drinking water.

E. coli is generally contaminated in stored drinking water (Kundu et al. 2018). As per BIS standard, 2012 shall not be detectable in any 100-mL sample. Some of the E. coli are not harmful, but in some cases, they can cause diarrhoea, gastrointestinal symptoms, urinary tract infections, respiratory illness, pneumonia, and other illnesses (Torres et al. 2018; Kockerling et al. 2017; Bourgeois et al. 2016).

A high concentration of E. coli was found in Udaipur, Garjee, Belonia, Jolaibari, and Kumarghat railway stations (Fig. 11). Drinking water of Sabroom, Thailik Twisa, and Panisagar, respectively, is better in terms of concentration of total coliform. Source, storage, and supply system of drinking water influence the bacterial concentration. S.K. Para, Nalkata, and Nadiapur railway stations have a significantly higher concentration of E. coli, mainly due to the source of drinking water (Fig. 11). The waters containing higher E. coli required extra treatment before sending to the tap for drinking.

Fig. 11

E. coli experiment during 0 h (a) and 24 h of incubation in nutrient broth media from the collected water sample from different railway stations

Multivariate statistical analysis

It has been observed that a correlation between TDS and total hardness (r = 0.86, the corresponding p-value is > 0.05), TDS and alkalinity (r = 0.79, corresponding p-value is > 0.05), and alkalinity and total hardness (r = 0.78, corresponding p-value is > 0.05) are significantly very high. pH has negative relation with turbidity, TDS, iron, and total coliform (Fig. 12). Turbidity has a significant positive relationship with total coliform (r = 0.43, the corresponding p-value is 0.012). Iron, sulphate, and iron have low degrees of association with turbidity (Fig. 12). Nitrate has a statistically significant negative relationship with chloride (r =  − 0.40, the corresponding p-value is 0.019). A significant positive relationship was also observed in total coliform’s relationship with TDS (Fig. 12). E. coli has a positive correlation with total coliform as E. coli is proportional to total coliform.

Fig. 12

Correlation among the variables (source: prepared by the authors, 2022 using Minitab v.18.1)

Properties of the variables are classified into three categories, i.e., physical (set variable 1), chemical (set variable 2), and biological (set variable 3). Canonical correlation has been calibrated to understand the association’s maturity among the different sets of variables.

It has been observed that physical (r = 0.42, p-value = 0.064), chemical (r = 0.31, p-value = 0.047), and biological (r = 0.47, p-value = 0.077) variables of drinking water have a positive correlation (Table 3). However, it has been found that physical variables are significantly associated with biological variables.

Table 3 Canonical correlations among physical, chemical, and biological variables of drinking water

Principal component analysis (PCA) has been applied to identify individual loadings of 13 variables in water quality. Eigenvalues are utilised to achieve principal components (PCs). An eigenvalue determines significant variables with maximum value. It has been observed that chloride (λ = 2.98) is the principal component in the drinking water of different railway stations of Tripura (Table 4). According to Muangthong and Shrestha (2015), eigenvalues ≥ 1 are the most significant variables. Here, it has been found that chloride (λ = 2.98), E. coli (λ = 2.13), turbidity (λ = 1.72), fluoride (λ = 1.49), and iron (λ = 1.08) have ≥ 1 eigenvalues which depict their level of significance (Table 4). Chloride, E. coli, turbidity, and fluoride are significant at 1% level, respectively.

Table 4 Principal component analysis of different variables of drinking water

Principal components with eigenvalues of less than 1 were discarded on account of their low essentiality (Jombart et al. 2010; Jolliffe 2002). It has been found that eight variables, namely sulphate (λ = 0.81), arsenic (λ = 0.74), total hardness (λ = 0.62), total coliform (λ = 0.52), nitrate (0.37), TDS (λ = 0.36), pH (λ = 0.09), and alkalinity (λ = 0.04), have less than 1 eigenvalue (Table 4). Except pH and alkalinity, all the variables are significant at 5% level.

Condition number (K) and condition index (CI) have been calculated to diagnose multicollinearity among all drinking water quality.

$$\begin{array}{c}k=\frac{\mathrm{Maximum}\;\mathrm{eigenvalue}}{\mathrm{Minimum}\;\mathrm{eigenvalue}}\\CI=\sqrt{\frac{\mathrm{Maximum}\;\mathrm{eigenvalue}}{\mathrm{Minimum}\;\mathrm{eigenvalue}}}=\sqrt k\end{array}$$

A scree plot diagram (Fig. 13a) has been identified where the first five factors, namely chloride, E. coli, turbidity, fluoride, and iron, with more than 1 eigenvalue. As a result, the first five components contained with a total variance of 72.509%. It has been observed that after five eigenvalues, a minor fall in slope took place (Fig. 13a). The first nine variables hold more than 90% of the total variance. The last four variables, i.e., alkalinity, pH, TDS. and nitrate, have been found that the last four variables contribute less than 5% of the total variance.

Fig. 13

Principal component analysis by a scree plot and b 3D loading plot (source: prepared by the authors, 2022)

It has been depicted that a strong association exists between TDS, alkalinity, and total hardness (Fig. 13b). Another group of associations found among iron, total coliform, and E. coli signify that some metallic ions help bacterial growth. The proximity between iron and turbidity has been observed, which indicates that some metallic ions are suspended in water and increase the turbidity level (Fig. 13b).

A varimax rotation was performed on the drinking water quality data through the rotated factor. Turbidity (0.948), total hardness (0.891), and alkalinity (0.835) have large positive loadings on factor 1, so this factor describes physical factors that influence the drinking water quality of the railway station (Table 5).

Table 5 Factor analysis on different variables of drinking water

Turbidity (0.665) has significant positive and fluoride (− 0.699) has large negative loadings on factor 2, so this factor describes the relation between physio-chemical properties of water.

Sulphate (0.524) and nitrate (0.646) have large positive loadings on factor 3, so this factor describes influential factors of the chemical properties (Table 5).

E. coli (− 0.713) have large negative loadings on factor 4, so this factor describes the negative impact of E. coli on drinking water.

Iron (0.695) has large positive loadings on factor 4, so this factor describes the impact of heavy metals on drinking water (Table 5).

Together, all five factors explain 84% of the variation in the data.

The Kaiser–Meyer–Olkin (KMO) test shows sampling adequacy is 0.471 and Bartlett’s test measure of aphericity (approx. chi-square 128.140) which is statistically significant at 1% level.

The dendrogram method was also applied to understand the association of drinking water parameters (Roy et al. 2021). It has been observed in cluster 1, pH has relation to turbidity and metallic elements of the water have associated with alkalinity and TDS. In cluster 2, sulphate, total hardness, chloride, and E. coli are associated. Cluster 3 comprises total coliform and all variables of cluster 2 (Fig. 14).

Fig. 14

Cluster dendrogram (Ward’s method) showing the linkage between the variables (source: prepared by the authors, 2022)

Water quality index (WQI) analysis

Weighted arithmetic water quality index

One of the most popular and widely used methods to determine water quality is WAWQI (Zait et al. 2022; Ahmed et al. 2021; Pradhan et al. 2021; Roy and Manna 2021). Railway stations have been classified based on estimated WAWQI. It has been observed that good quality drinking water is found in 55.55% of railway stations i.e., Thailik Twisa, Belonia, Teliamura, Mungiakami, Sekerkote, Bishramganj, Jogendranagar, Churaibari, Agartala, Kumarghat, Jolaibari, Garjee, Dharmanagar, Pencharthal, and Ambassa (Table 6). About 37.04% of railway station has poor drinking water quality, namely Udaipur, Manu Bazar, Bishalgarh, Jawaharnagar, Nalkata, Panisagar, Jirania Sabroom, Santirbazar, and S. K. Para. According to WAWQI, only 7.41% of railway stations (2 out of 27) have been observed with very poor drinking water, namely Manu and Nadiapur railway stations.

Table 6 Classification of drinking water quality of the railway station through WAWQI with iron

Drinking water quality can be improved by removing iron from the water. Iron is one of the principal components of the drinking water of Tripura. Now as per WAWQI, 55.55% of railway stations have good quality drinking water, but removing iron it will be able to provide good quality water in 81.48% of the railway stations of Tripura (Table 7). It will ensure by eradicating iron from drinking water that no railway station will be left with very poor water quality (Table 7).

Table 7 Classification of drinking water quality of the railway station through WAWQI without iron

Entropy weighted water quality index

Another widely used method called EWWQI has been applied to estimate water quality analysis. It has been observed that as per EWWQI, about 85.19% of railway stations have good drinking water, namely Teliamura, Sekerkote, Kumarghat, Agartala, Panisagar, Bishramganj, Jogendranagar, Nalkata, Churaibari, Jolaibari, Belonia, Nadiapur, Dharmanagar, Mungiakami, Bishalgarh, Manu, Santirbazar, Jirania, Ambassa, S. K. Para, Garjee, Udaipur, and Pencharthal railway stations (Table 8). On the other hand, Jawaharnagar, Thailik Twisa, Manu Bazar, and Sabroom railway stations have poor drinking water quality, according to EWWQI. No significant difference has been found by removing iron from EWWQI.

Table 8 Classification of drinking water quality of the railway station through EWWQI

To estimate the efficiency between two methods, the efficiency index (EI) has been carried out on variance differences of the corresponding indices with regards to EWWQI by using the following formula:

$$EI=\frac{{v}_{i}-{v}_{x}}{{v}_{x}}$$

where

v _i :

variance of WAWQI

v _x :

variance of EWWQI

It has been observed that EWWQI is 1.65 times more efficient than WAWQI.

Cluster analysis categorises 27 railway stations of Tripura into three clusters. In cluster 1, total 11 (40.74%) railway stations, i.e. Churaibari, Mungiakami, Dharmanagar, Agartala, Teliamura, Bishramganj, Panisagar, Kumarghat, Jogendranagar, Thalik Twisa, and Sabroom (Fig. 15). Maximum railway stations of this cluster are located in urban areas and relatively a large number of passengers used to travel from this railway station. Cluster 2 includes another 11 (40.74%) railway stations, namely Nadiapur, Nalkata, Manu, S.K. Para, Sanirbazar, Sekerkote, Bishalgarh, Belonia, Jolaibari, Udaipur, and Garjee. Most of the station railway stations are located in the forest area and comparatively fewer numbers of passengers are moving from the railway station. Cluster 3 has five railway stations Pencharthal, Jawaharnagar, Jirania, Ambassa, and Manu Bazar (Fig. 15).

Fig. 15

Cluster dendrogram (Ward’s method) showing the linkage between the different railway stations of Tripura (source: prepared by the authors, 2022)

Causality analysis

Drinking water quality varies due to natural, quasi-natural, and anthropogenic activities and causes (Akbar et al. 2021). The causality of diversified quality of drinking water varies due to geological structure (Pant et al. 2021; Shroder et al. 2021; Baba and Gundur 2017), physiographic characteristics (Qadir and El-Rawy 2021; Wu and Lu 2021; Taka et al. 2015), soil properties (Wu et al. 2022; Shahadha et al. 2019; George et al. 2017), micro-regional land use pattern (Rimba et al. 2021; Tahlru et al. 2020; Hua 2017), agricultural practices (Spitsov et al. 2020; Sasakova et al. 2018; Nikolaidis et al. 2008), industrial development (Zou et al. 2021; Edwards and Crabb 2020), and anthropogenic activities (Akhtar et al. 2021; Kulk et al. 2021; Khatri and Tyagi 2014). Railway stations of Tripura are located in different geological structures and geomorphic environment (Fig. 16). The state of Tripura exhibits a wide array of sedimentary rock characteristics of marine-mixed-fluvial-type origin ranging in age from the uppermost Oligocene (38 million years from present time) to the recent period (Dey et al. 2014). According to the Geological Survey of India, this sediment stratigraphy of state geology is divided into three groups: Dupitila formed during the Quaternary period, Tipam which bent between Pliocene to Pleistocene period, and Surma group have been laid down in the during Uppermost Oligocene (which lasted for 65 million years). The geological structure of the state is governed by local tectonic movement with a wide range of environmental conditions (Dey and Sarkar 2012). Tectonically, the region now comprises a series of sub-parallel arcuate, elongated, doubly plunging folds arranged in the north–south direction. Wide flat synclines separate these folds for anticlines.

Fig. 16

Relationship among a geology, b geomorphology, c physiography, and d oil with diversified drinking water quality in different railway stations of Tripura (source: prepared by the authors, 2022)

It has been observed that about 59.26% (16) of railway stations of Tripura have been established in Tipam formation and the remaining 40.74% of railway stations of the state developed over the Dupitila formation (Table 9). In Tripura, about 37.34% of land has been identified with Surma stratigraphic group sub-divided in for formation, i.e. Bokabil Formation, Upper Bhuban Formation, Middle Bhuban Formation, and Lower Bhuban Formation, but this formation is not exposed in Tripura where no railway station has been observed (Table 9). Local unconformity, lithic characteristics (oldest rock), and the stratigraphic position of the Surma group are the primary causality of the diversified water quality of the station. It has been observed that 62.50% of railway stations situated on Tipam Sandstone have good drinking water quality, whereas 45.45% of railway stations situated on the Dupitila structure have good drinking water quality, respectively. About 9.10% of railway stations of the Dupitila structure have inferior and 45.45% have poor quality of drinking water, respectively (Table 10).

Table 9 Relation between drinking water quality, geological structure, and geomorphic features

Table 10 Relation between drinking water quality and geological structure of different railway stations in Tripura

Maximum stations (37.04% of total railway stations) located on Tipam sandstone structure have a good quality of water because rock-water interactions highly influence the water quality of the railway stations of Tripura (Fig. 16a).

It has been observed that five predominant geomorphic features are found in Tripura: the alluvial plain, highly dissected hills and valleys, moderate dissected hills and valleys, low dissected hills and valleys, and pediments pediplain complex (Fig. 16b). It has been observed that 37.04% (10) railway stations are located in the alluvial plains, whereas 50% of railway stations have good drinking water quality (Table 11). About 40% and 10% of railway stations in this geomorphic landscape have poor and very poor, respectively. Only four railway stations have been found with poor infrastructure, namely Panisagar, Jirania, Bishalgarh, and Udaipur. Located off the Panisagar railway station on the alluvial plain of river Juri, Pencharthal, Bishalgarh, and Udaipur are situated alluvial plain of rivers Juri, Deo, Bijoy, and Gomati, respectively. Manu railway station has been observed with very poor drinking water in this category.

Table 11 Relation between drinking water quality and geomorphic features of different railway stations in Tripura

Maximum railway stations (51.85%) are found in highly dissected hills and valleys. In Tripura, about 61.95% of land fall under highly dissected hills and valleys, so naturally, railway stations are proportionate to the land area. A total of 33.33% of railway stations within this geomorphic class have good drinking water quality, namely Churaibari, Mungiakami, Jogendranagar, Agartala, Sekerkote, Bishramganj, Belonia, Jolaibari, and Thailik Twisa railway stations (Table 11).

The nature of soil plays a crucial role in drinking water quality (Smith et al. 2021). As per FAO’s taxonomical classification, Tripura has five soil types, namely ferric acrisols, dytric cambisols, dystric glaysols, and two variants of eutric glaysols (Fig. 16d). According to World Reference Base (WRB), ferric acrisols are red and yellow weathered soils with metal oxides, particularly iron and aluminium, found in the humid tropical region. Tripura 14 (51.85%) railway stations are located over ferric acrisols soil (Table 12).

Table 12 Relation between drinking water quality and soil characteristics of different railway stations of Tripura

About 42.86% of railway stations of Tripura located over ferric acrisols soil, i.e. Jirania, Bishalgarh, Udaipur, Santirbazar, Manu Bazar, and Sabroom, have poor quality of drinking water where the average concentration of iron is 0.57 mg/L which also most double from BIS standard. On the other hand, 37.50% of railway stations, i.e. Nalkata, S. K. Para, and Jawaharnagar located on Dytric Cambisols, have poor quality drinking water where the average value of iron is 0.39 mg/L which depicts that soil properties directly influence drinking water quality.

Relief structure and land-use patterns influence the water quality by affecting the pollution process (Wang et al. 2014). However, land use variables might affect the pollution process differently with the spatial scale and season variations.

It has been observed that physiography has a significant negative correlation (r =  − 0.438) with drinking water quality (EWWQI). This depicts that with the increase in height, there is a significant correlation between drinking water quality and composite land use pattern (WPI) (Table 13).

Table 13 Correlations between physiography (elevation), land use (WPI), and water quality (WAWQI and EWWQI)

But the relation between water quality indices with individual land -use variables like built-up area, agricultural land, forest cover, fallow land, and water bodies, as well as NDVI (vegetation cover) and NDBI (urbanisation), has been calibrated through the Pearson correlation matrix (Table 14). It has been found that drinking water quality (WAWQI) has a significant negative correlation with urbanisation (NDBI). The result depicts the urbanisation process that induces pollutant materials to the drinking water. Agricultural land also has a negative relation with water quality indices because, during agricultural practices, some manures are infused with soil, and later on, due to the leaching process, manure contaminates the water.

Table 14 Correlations between water quality (WAWQI and EWWQI) and land use variables

Passenger mobility and station infrastructure are important factors for drinking water quality policy intervention in railway stations. A strong positive significant correlation between station infrastructure and passengers’ mobility has been observed, but those factors do not associate with safe drinking water (Table 15). However, both drinking water indices show a poor negative correlation with passenger mobility. This reveals that the railway stations of Tripura cannot provide safe drinking water to the passengers.

Table 15 Correlations between station infrastructure (CII), monthly passenger mobility, and water quality (WAWQI and EWWQI)

Major findings

From the above analysis, the following are the significant findings:

• The drinking water quality of the railway stations varies from one another. Chloride, E. coli, turbidity, fluoride, and iron are the principal components of the drinking water of railway stations of Tripura, which cumulatively contributes 72.51%.

• TDS have a strong positive correlation with turbidity. The physical properties of drinking water have a significant positive correlation with the biological properties of drinking water.

• EWWQI is a more efficient (173%) drinking water quality index measure than WAWQI.

• Drinking water quality can improve from 55.55 to 81.48% by removing iron content from the water.

• The geological structure and geomorphic attributes influence drinking water quality of the railway stations of Tripura.

• Soil properties, especially iron, directly influence the drinking water quality of different railway stations in Tripura.

• Urbanisation is an essential cause for poor drinking water quality.

Policy and practice interface

Since its establishment, different railway companies were provided drinking water to the railway colonies and the railway stations. Railway Act, 1890 (Section-8) admired the provision of drinking water at railway stations; the same thing was endorsed by the Railway Act, 1989. On August 3, 1980, for the first time, Railway Board issued a notification (Vide No. 1980/W2/25/O dated 03.08.1980) regarding the quality of drinking water at the railway station and railway colonies. As per notification vide no. 1986/W2/25/18 dated September 16, 1986, the Ministry focuses on building an adequate number of taps, tube wells, open wells, tanks, etc., to supply safe drinking (Railway Board 1986). Ministry of Railways instructs all zonal railways to supply safe drinking water to the passengers (Letter vide no. 1994/LMB/09/05 dated February 8, 1995). Once again, on September 26, 1997, the Ministry of Railways advised the zonal railways to provide safe drinking water at the railway station (Letter vide. No. 1997/LMB/09/18 dated September 26, 1997). In 2003, the Ministry of Railways adopted Rain Water Harvesting Systems in the railway establishments to provide safe drinking water by recycling harvested water (Letter vide. no. 2003/LMB/09/01 dated June 12, 2003). To supply of hygienically safe drinking water, the Ministry of Railways was instructed to install a programmable Logic Controller based Automatic Bacteriological Treatment Plant in all the railway stations of India (Letter vide. no. 2004/LMB/09/01 dated March 7, 2014). Ministry of Drinking Water and Sanitation, Government of India, has issued the “Uniform Drinking Water Quality Monitoring Protocol” in February 2013 (Ministry of Drinking Water and Sanitation 2013). The UDWQMP mention thirteen basic minimum parameters, i.e., pH, turbidity, TDS, total hardness, alkalinity, fluoride, chloride, sulphate, nitrate, arsenic, iron, total coliforms, and E. coli that need to be tested for drinking water quality monitoring and verified with BIS standard. The engineering department is the concerned authority to monitor the drinking water quality with the help of the medical department. In the meantime, India has a National Water Policy (NWP), which was adopted to administrate the planning and development of water resources and optimal utilisation (Central Water Commission 2012). In response to NWP, in 2017 Ministry of Railways, the Government of India formulated the “Indian Railway Water Policy” (IRWP) (Ministry of Railways 2017). As per IRWP, every zonal railway should have a Water Management Plan (WMP). The smallest unit of WMP is a railway station. Under WMP, Indian Railways is planning to upgrade the water supply system, the introduction of automatic valves, compulsory provision of solar water heaters, rain water harvesting (RWH) on station premises, revive old water bodies on railway premises, and water recycling plants on a BOOT basis, use of water-efficient fittings, and water audit (WA).

In practice, WMP is not available in many railway stations in Tripura. About 26% of railway stations, i.e. Churaibari, Nadiapur, Panisagar, Nalkata, S.K. Para, Sekerkote, and Thailik Twisa, do not have an iron removal plant. All the railway stations of Tripura except Nadiapur, Nalkata, S.K. Para, and Thailik Twisa have a pumping station to lift the ground water stored in the tank and supplied to the taps of the station. However, storage tanks are not cleaned frequently and sometimes use calcium oxychloride (CaOCl₂) to clean the water. Cleaning the tank is the responsibility of the station manager (SM) or station superintendent (SS), but due to lack of funds, SM or SS is not able to clean the tank every month.

For cleaning purposes, every station usually receives ₹5000 ($64.25) per month, which is inadequate to clean the water tank also. The most critical condition found in Nadiapur and S. K. Para railway is where passengers, including railway staff, drink the water from an uncovered well. Automatic valves in the pumping station have not been introduced in any railway stations in Tripura. Only two railway stations of Tripura, namely Dharmanagar and Agartala, have the compulsory provision of solar water heaters (SWHs), which help water quality control (Fan et al. 2016). At present, 92.59% of railway stations do not have SWHs. As per RWH plan concern, Dharmanagar has a better position with two lakes at 18,075 sq. m. and 21m828 sq. m, compared to the other railway stations. In 2009 an artificial water body (2584 sq. m.) was dragged into the railway premises of Agartala railway station (23°47′31.57″N and 91°16′39.51″E) for RWH. Expect Agartala and Dharmanagar railway stations; no other railway stations have RWH. However, all the railway stations of Tripura have sufficient land to accomplish RWH. Land resource crisis has been found in only Jogendranagar, Jawaharnagar, Kumarghat, and Churaibari railway stations. IRWP mentions reviving old water bodies on railway premises, but in Tripura, no such old water bodies have been found to be revived since the establishment of the railway stations. No railway stations in Tripura have water recycling plants. From the monthly passenger flow, railway stations usually estimated water requirements and no proper water audit (WA) has been calculated by the stations nor verified by higher authorities at the divisional or zonal level. The main reason of improper WA is the uses of a single source (groundwater) drinking water supply. The only proper WA can be accounted for if the railway stations induce multiple drinking water sources. So it has been found that there is a considerable gap in the policy-practice interface.

The gap impacts the passengers’ satisfaction with those for whom all services are provided. Satisfaction of the passengers with drinking water quality has been studied through 5-point Likert scale. The satisfaction index has been calculated by modifying Hall et al. (1975) method using the following formula:

$${MSI}_{j}=\frac{\left({N}_{s}-{N}_{d}\right)}{\left(N-{N}_{e}\right)}=\frac{\left({N}_{s}-{N}_{d}\right)}{\left({N}_{s}-{N}_{d}\right)}$$

where

MSI _j :

modifiaed satisfaction index of jth parameter

N :

number of tatal passengers

N _s :

number of passengers having response with very good and good

N _d :

number of passengers having response with very bad and bad

N _e :

number of passengers having response with moderate

Remarks: Range of ${MSI}_{j}$ is [1, − 1]

Proof: Let us consider an infrastructural parameter “J”; among all $\left({N}_{s}+{N}_{d}\right)$ passengers if all passengers have either very good or good response only, then ${N}_{d}=0$,

$${MSI}_{j}=\frac{\left({N}_{s}-0\right)}{\left({N}_{s}+0\right)}=1$$

In another way, if all $\left({N}_{s}+{N}_{d}\right)$ passengers have a very bad and bad response to the railway infrastructural parameters

Then N_d = 0,

$${MSI}_{j}=\frac{\left(0-{N}_{d}\right)}{\left(0+{N}_{d}\right)}=-1$$

So, $-1\le {MSI}_{j}\le 1$

To identify the passengers’ demand for the betterment of railway infrastructural facilities in the state, about 735 samples have been collected from the different railway stations. For sample accuracy in estimating proportions by considering the standard normal deviation set at 95% confidence level (1.96), the percentage picking a choice or response (50% = 0.5) and the confidence interval (0.05 =  ± 5). Normality of sample distribution has been tested through Kolmogorov–Smirnov test (K-S test). S test (D = 0.15963, corresponding p-value 0.45071) found that the sample does not differ significantly, revealing that samples are normally distributed. It has been observed that in Tripura, about 67.35% of passengers are used to taking drinking water from different railway stations. The remaining 32.65% of passengers do not drink the water of railway stations mainly due to its bad quality. According to 2.42% of passengers, the quality of drinking water in the railway stations is very good. The perceived quality of drinking water is good, according to 46.26% of passengers (Fig. 17). It has been observed that about 8.08% and 3.23% of passengers said the quality of drinking water is bad and very bad, respectively (Fig. 17). Passengers of Ambassa (5.88%), Agartala (5.66%), and Manu Bazar (5.56%) stated that the quality of drinking water at the respective railway stations is very good. It has been observed that the perception of 100% of passengers of Nadiapur, Nalkata, S.K. Para, Jawaharnagar, and Sekerkote railway stations’ drinking water quality is very bad. The infrastructure of the drinking water facility of those railway stations is notably poor. All Santirbazar railway station passengers stated that the quality of drinking water is bad. More than 50% of passengers of Jogendranagar (54.05%), Agartala (56.60%), Bishramganj (66.67%), Belonia (58.06%), Jolaibari (100%), and Sabroom (66.67%) observed that drinking water quality of respective railway stations is good (Fig. 17).

Fig. 17

Perceived quality of drinking water in the railway stations of Tripura (source: prepared by the authors; data extracted from field survey during 2022)

In general, the overall satisfaction index on drinking water quality is 0.623 which depicts that a significant proportion of railway passengers of Tripura are satisfied with the provided drinking water of the railway stations (Table 16). Railway stations are classified based on drinking water quality-based satisfaction index.

Table 16 Classification of railway stations based on satisfaction index

It has been observed that the satisfaction level of railway passengers regarding drinking water quality is very in Mungiakami (1.000), Agartala (1.000), Jolaibari (1.000), Thailik Twisa (1.000), Sabroom (1.000), Dharmanagar (0.926), and Udaipur (0.876) (Table 17). According to the satisfaction index, about 25.96% of railway stations of Tripura, i.e. Jogendranagar, Belonia, Ambassa, Manu, Teliamura, Bishramganj, and Pencharthal, drinking water quality is good (Table 16). Poor drinking water was found in Jirania (0.429), Bishalgarh (0.429), Manu Bazar (0.385), Panisagar (0.333), Kumarghat (0.333), and Garjee (0.333). Passenger satisfaction regarding drinking water is very poor in Churaibari, Nadiapur, Nalkata, S. K. Para, Jawaharnagar, Sekerkote, and Santirbazar.

Table 17 Pairwise Spearman correlations among different models

Different experimental models (WAWQI and EWWQI) and the social science model (satisfaction index) of drinking water quality are positively correlated, indicating that passenger satisfaction is positively associated with drinking water quality. Indian Railways always emphasised passengers’ satisfaction (bottom-up approach) in planning and policy formulation. In this context, quality of drinking water at railway station is taken into account for better public policy.

Conclusion and recommendation

Spatial prioritisation based on drinking water quality of different railway stations of Tripura has been spotted using WAWQI and EWWQI models along with passengers’ statisfaction index, where passengers’ statisfaction is significantly associated with EWWQI. Causality of varied quality of drinking water is associated with physiography and land use pattern of railway station surrounding areas significantly. To resolve the issue of drinking water quality, a sustainable planning model is needed to improve the drinkability of water at railway stations; both technical and institutional issues need to be identified. It has already been proven that urbanisation is one of the major causes of the poor quality of water. Apart from that, agricultural practices and soil and rock characteristics also influence drinking water quality. In Tripura, there is a considerable gap in policy and practices regarding the quality of drinking water at the railway stations because the railway stations do not appropriately follow IRWP. All railway stations should prepare a WMP considering micro-regional factors like soil, rock, and land-use pattern because every station has a unique geophysical space with specific advantages and limitations. This study proposed an alternative technical strategy to resolve drinking water quality issues in the railway stations of Tripura.

The strategies and interventions are categorised into three segments. In the long term, a chlorination plant should be installed in all the railway stations of Tripura, which is needed to kill E. coli and total coliform, which are present in the drinking water of railway stations in Tripura. Reverse osmosis (RO) can remove salt, manganese, iron, fluoride, lead, and calcium (Welsh 2005). Every railway station has an ultraviolet (UV) water purifier that treats micro-biologically unsafe water with germicidal UV light (Times of India 2020). It could be encouraged to install all the railway stations of Tripura as a medium-level intervention (Table 18). For short-term interventions, cleaning and desilting of the storage tank are to be done frequently. Presently the railway stations do not have sufficient funds, workforce, and technologies to clean the storage tank (Table 18).

Table 18 Technical strategics to improve drinking water quality