Background

Groundwater quality is important to human health, agriculture, aquaculture and industry (Vanloon and Duffy 2005). In the last decades, groundwater resource has become the potential source of domestic water supply in Ghana and the world at large (Hynds et al. 2014). Interestingly, many public health surveys and water quality analysis has shown that groundwater is not immune to contaminants such as waterborne pathogens, toxic elements (Asamoah and Amorin 2011).

Over the last few years, surface water and groundwater resources are among the most important environmental issues due to heavy metals contamination and human industrial activities (Vodela et al. 1997; Öztürk et al. 2009; Marcovecchio et al. 2007; Khodabakhshi et al. 2011; Ghasemi et al. 2011). Some of the heavy metals are essential for growth, development and health, whiles others are categorized as toxic species on living organisms (Underwood 1956). Heavy metal contamination has become a significant problem in several community and agricultural areas over the years due to the application of commercial agrochemicals on agricultural production (Vodela et al. 1997; Rattan et al. 2005). However, heavy metals originating from anthropogenic sources have been found in all components of the environment (Idris et al. 2007; Idris 2008; Ayni et al. 2011). In recent years more attention has been devoted to pollutants in the environment due to increase in anthropogenic contribution by heavy metals (Edmund et al. 2003; Marengo et al. 2006). Heavy metals can eventually dispersed and accumulated in the soil as well as surface and groundwater and may therefore impact adverse human health effect to living organisms (Rashed 2010; Chotpantarat et al. 2011; Chotpantarat and Sutthirat 2011; Taboada-Castro et al. 2012).

Groundwater quality, water type and sources of contaminants have been explored using Piper, Stiff plots and multivariate statistical techniques (Affum et al. 2015). Multivariate statistical techniques has extensively been used in literature to characterise and assess groundwater sources (Hussain et al. 2008; Shihab and AbdulBaqi 2010; Mahmood et al. 2011; Okogbue et al. 2012; Narany et al. 2014; Masoud 2014; Uddameri et al. 2014; Yadav et al. 2014; Asare-Donkor et al. 2015). However, the use of multivariate statistical technique and pollution evaluation index to assess the safety and quality of hand-dug well water in the Ejisu-Juaben Municipality is lacking.

Ejisu-Juaben Municipality is an agricultural region and have a high density of farmers that uses agrochemicals for producing various agricultural products, such as rice, cassava, chili and rubber trees (Anornu et al. 2009). The main chemically treated water is obtained from the Barekese Dam, which is produced by the Ghana Water Company Limited (GWCL) and distributed to individual households through pipes in the peri-urban communities of Ejisu. The demand for treated water by Ejisu-Juaben Municipality has increased over the years without any corresponding rise in quantity and quality. As a result of the above problems, individual households have resulted to the use of hand-dug wells as an alternative source of water supply, but the awareness of the health implications associated with the use of water from hand-dug wells have not been considered. This work is aiming to study the suitability and health risk associated with heavy metal contamination of groundwater and surface water for drinking purposes in the Ejisu-Juaben Municipality. The study will also employ multivariate statistical technique and pollution evaluation indices as complementary tool to identify the various possible sources of pollution that influence the water quality.

Methods

Study area

Geographically, Ejisu–Juaben Municipal Assembly lies within Latitudes 1°15′N and 1°45′N and Longitude 6°15′W and 7°00W. Ejisu-Juaben municipal area is one of the 27 administrative districts of Ashanti Region. Ejisu is the main capital of the municipality. The municipality has 4 main urban settlements namely: Ejisu, Juaben, Besease and Bonwire. The municipality can boast of about 88 communities. Ejisu stretches over an area of 637.2 km2 with population around 143,762. The municipality lies in the central part of the Ashanti Region and shares boundaries with six Districts in the Region. To its north east and west are Sekyere East and Afigya Kwabre, respectively, to the south; Bosomtwi and Asante Akim South, to its east; Asante Akim North and to the west; Kumasi metropolitan assembly. The mean monthly temperatures vary between 20 °C in August and 32 °C in March. The relative humidity ranges from 65 % in January to 85 % in August. The area usually experiences a wet semi-equatorial climate. Majority of the rainfall is from March to July and again from September to the later part of November, with a mean of 1200 mm which is ideal for minor season cropping as a result of climatic changes and seasonal drought. Groundwater is an important water resource for drinking, agriculture, and industrial uses in the study area. The area is characterized by rural setting and the major occupation of the people is agriculture. The main food crops grown are plantain, cassava, maize and cocoyam. One of the major cash crops grown includes cocoa, which is the driving force of the economy. Oil palm is also another cash crop that is widely grown in the district.

Sampling

All solvents and reagents used were of high analytical grade supplied by BDH Chemical Ltd, UK. Double-distilled deionized water was used throughout the experiments. A total of nineteen (19) samples were obtained from the Ejisu-Juaben Municipality. The sampling stations are shown in Fig. 1. The sampling standard methods prescribed by APHA (2005) were followed carefully for the groundwater collection and analytical techniques. Samples were collected in 1.5 l high-density polyethylene (HDPE) containers. Prior to sampling, the bottles were rinsed with the water to be sampled and the samples were preserved by acidifying to pH <2 with HNO3 and kept at a temperature of 4 °C until analysis.

Fig. 1
figure 1

Map of the study area with sampling locations

Full size image

Heavy metal analysis

The preserved sample was taken out from the refrigerator and kept at room temperature until the attainment of thermal equilibrium. Prior to AAS analysis, water samples were digested with 5 mL of concentrated HNO3(aq). The mixture was heated on a hot plate and filtered using Whatman filter paper into a volumetric flask. Distilled water was added to mark to the top the volumetric flask. The heavy metals (Mn, Pb, Fe, Cd, Cu and Zn) was analysed using atomic Absorption Spectrophotometer (AAS) (VARIAN SpectrAA 220) with air acetylene flame.

Quality assurance and quality control

Replicate blanks and reference materials, NIST 1640(a) (USA: National Institute of Standards and Technology) were used for the method of validation and quality control. Replicate analysis of these reference materials showed good accuracy (relative standard deviation, RSD, ≤3 %) and recovery rates ranged from 94.5 to 105.8 %. The concentrations of heavy metal in the groundwater were reported in mg/L.

Pollution evaluation indices

Generally, pollution indices are estimated for a specific use of the water under consideration. The heavy metal pollution index (HPI), heavy metal evaluation index (HEI) and degree of contamination (Cd) were used to evaluate the drinking water quality. The HPI and HEI methods provide an overall quality of the water with regard to heavy metals using the ratios of monitored values of the desired number of parameters and the maximum admissible concentrations of the respective parameters. In the Cd method, the quality of water is evaluated by computation of the extent of contamination and computed as the sum of the contamination factors of each component exceeding the upper permissible limit. Therefore, the Cd summarizes the combined effects of a number of quality parameters regarded as unsafe to household water.

Heavy metal pollution index

The HPI method was developed by assigning a rating or weightage (W i ) for each chosen parameter and selecting the pollution parameter on which the index was to be based. The rating is an arbitrary value between 0 and 1 and its selection reflects the relative importance of individual quality considerations. W i is defined as inversely proportional to the recommended standard (S i ) for each parameter (Horton 1965; Reddy 1995; Mohan et al. 1996). In this study, the concentration limits (i.e., the highest permissible value for drinking water (S i ) and maximum desirable value (I i ) for each parameter) were taken from the WHO (2011) standard. The uppermost permissive value for drinking water (S i ) refers to the maximum allowable concentration in drinking water in absence of any alternate water source. The desirable maximum value (I i ) indicates the standard limits for the same parameters in drinking water. According to Mohan et al. (1996), HPI is determined according to Eq. (1):

$${\text{HPI }} = \frac{{\mathop \sum \nolimits_{i = 1}^{n} W_{i} Q_{i} }}{{\mathop \sum \nolimits_{i = 1}^{n} W_{i} }}$$
(1)

where Qi and Wi are the sub-index and unit weight of the ith parameter, respectively, and n is the number of parameters considered. The sub-index (Qi) is calculated according to Eq. (2):

$${\text{Q}}_{\text{i}} = \mathop \sum \limits_{i = 1}^{n} \frac{{\left\{ {M_{i} - I_{i} } \right\}}}{{S_{i} - I_{i} }} \times 100$$
(2)

where M i , I i , and S i are the monitored heavy metal, ideal, and standard values of the ith parameter, respectively.

Heavy metal evaluation index

Heavy metal evaluation index HEI gives an overall quality of the water with respect to heavy metals (Edet and Offiong 2002) and is expressed using Eq. (3):

$${\text{HEI }} = \mathop \sum \limits_{i = 1}^{n} \frac{{H_{c} }}{{H_{\text{mac}} }}$$
(3)

where Hc and Hmac are the monitored value and maximum admissible concentration (MAC) of the ith parameter, respectively.

Degree of contamination (Cd)

The contamination index (Cd) summarizes the combined effects of several quality parameters considered harmful to domestic water (Backman et al. 1997) and is calculated as follows:

$${\text{C}}_{\text{d}} = \sum\limits_{{{\text{i}} = 1}}^{\text n} {{\text{C}}_{{{\text{f}}_{\text{i}} }} }$$
(4)
$${\text{where}}\, {\text{C}}_{{{\text{f}}_{\text{i}} }} = \frac{{{\text{C}}_{{{\text{A}}_{\text{i}} }} }}{{{\text{C}}_{{{\text{N}}_{\text{i}} }} }} - 1$$
(5)

where Cfi; CAi and CNi represent contamination factor, analytical value and upper permissible concentration of the ith component, respectively. N denotes the ‘normative value’ and CNi is taken as MAC.

Risk assessment

Risk assessment methods and parameters previously reported by Wongsasuluk et al. (2014) were used in the present study. Risk assessment is defined as the process of estimating the probability of occurrence of an event and the probable magnitude of adverse health effects on human exposures to environmental hazards over a specified time period (NRC 1983; Kolluru et al. 1996; Paustenbach 2002; Wongsasuluk et al. 2014). Risk assessment consists of hazard identification, exposure assessment, dose response and risk characterization (Lee et al. 2005). According to Lim et al. (2008), two toxicity risk indices reported are the slope factor (SF) for carcinogen risk characterization and the reference dose (RfD) for non-carcinogen characterization (Table 1).

Table 1 The toxicity responses (dose response) to heavy metals as the oral reference dose (RfD) (USEPA IRIS 2011; Wongsasuluk et al. 2014)
Full size table

Siriwong (2006) reported the estimations of the magnitude, frequency and duration of human exposure in the environment as average daily dose, for each water sample as:

$${\text{ADD}}_{\text{i}} = \frac{{({\text{C}}_{i}\,{\times}\,{\text{IR }}\,{\times}\,{\text{EF }}\,{\times}\,{\text{ED}})}}{{({\text{BW }}\,{\times}\,{\text{AT}})}}$$
(6)

All the parameters in Eq. (6) have been defined in Table 2.

Table 2 Input parameters to characterize the ADD value (Wongsasuluk et al. 2014)
Full size table

The health risk was assessed in relation to its non-carcinogenic as well as carcinogenic effects based on the calculation of ADD estimates and defined toxicity according to the following relationships (USEPA IRIS 2011; Wongsasuluk et al. 2014). The non-carcinogenic was computed as:

$${\text{Hazard quotient }}\left( {\text{HQ}} \right) \, = \frac{\text{ADD}}{\text{RfD}}$$
(7)

The risk assessments of a mixture of chemicals, the individual HQs are summed to form hazard index (HI):

$${\text{Hazard index }}\left( {{\text{HI}}_{\text{i}} } \right) \, = \mathop \sum \nolimits {\text{HQ}}_{\text{i}}$$
(8)

According Lim et al. (2008) an HI/HQ >1 means an unacceptable risk of non-carcinogenic effects on health, while HI/HQ <1 means an acceptable level of risk. Table 2 shows the principal exposure factors that have been taken into account to carry out the risk assessment calculations.

Chronic daily intake (CDI) in the present study was calculated using Eq. (4) modified from Muhammed et al. (2011); Wu et al. (2009) and De Miguel et al. (2007).

$${\text{CDI}} = {\text{C}}i \,{\times}\, \frac{\text{DI}}{\text{BW}}$$
(9)

where, Ci, DI, and BW represent the concentration of heavy metal in the water samples (mg/L), average daily intake rate (2.2 L/day), and body weight (70 kg), respectively.

Statistical analysis

IBM Statistical Package for the Social Sciences (SPSS) ‘20 was used for the data analysis. Principal component analysis was used to identify the possible sources of heavy metals. Factor analysis was performed by varimax rotation (Howitt and Cramer 2005), which minimized the number of variables with a high loading on each component, thus facilitating the interpretation of PCA results. Cluster analysis was applied to identify groups of samples with similar heavy metal contents (Panda et al. 2006). CA was formulated according to the Ward-algorithmic method, and the rescaled linkage distance was employed for measuring the distance between clusters of similar metal contents. R-mode CA was used to determine the association of different water quality parameters and pollutant sources. Pearson’s correlation matrix was also used to identify the elements’ relationship.

Results and discussion

Descriptive statistics related to the heavy metal concentrations in the Ejisu-Juaben municipality are presented in Table 3. Coefficient of variation (CV) was the most important factor in describing the variability of groundwater properties. Data was ranked according to amount of variation as low variability (CV ≤ 15 %), moderate variability (CV 15–35 %), or high variability (CV, > 35 %) (Wilding 1985). The CV values in the groundwater ranged from 0 to 2 %. All the heavy metals in the study area were found to exhibit low variability. These low values of CV indicate a homogenous distribution of these metals in the corresponding sampling points and may increase the effect of non-point sources. The concentrations of these heavy metals are illustrated in the box plot (Fig. 2). In Fig. 2, iron has relatively large inter-quartile range than the other metals. The mean concentration of heavy metals in the groundwater samples follows the order: Fe > Zn > Mn > Pb > Cu > Cd.

Table 3 Descriptive statistics of heavy metals concentrations in groundwater obtained from Ejisu-Juaben Municipality (n = 19)
Full size table
Fig. 2
figure 2

Boxplot of log of trace metals concentration of hand-dug wells in Ejisu-Juaben Municipality. The white circles represent outliers

Full size image

The concentration of lead in groundwater ranged from 0.000 to 0.040 mg/L with a mean concentration of 0.010 mg/L (Table 3). About 26 % of the groundwater samples contained lead concentration above the levels permitted by WHO (2011). The main sources of lead contamination are industrial discharges from smelters, battery manufacturing units, run off from contaminated land areas, atmospheric fall out and sewage effluents. The levels of Pb from this study were lower than study reported by Kortatsi (2007), Wassa West district, Addo et al. (2013); Nassef et al. (2006), Sadat Industrial City. On the other hand the results in this study generally agree with those reported by Wongsasuluk et al. (2014), Ubon Ratchathani province, Thailand.

The concentration of iron in the groundwater ranged from 0.002 to 0.568 mg/L, with a mean of 0.166 mg/L. The maximum allowable limit of iron concentration in groundwater is 0.3 mg/L as per WHO (2011) classification. About 89 % of the samples were within the WHO permitted limit. Under reducing conditions, the solubility of Fe-bearing minerals increase in water, leading to the enrichment of dissolved iron in groundwater (Applin and Zhao 1989; White et al. 1991). Chronic consumption of water with iron overload may results in fatigue, weight loss, joint pains and ultimately heart disease, liver problems and diabetes (US-CDC 2011). Study done by Ansa-Asare et al. (2009) in Adidome district indicated that the levels of Fe were higher compared to this study. However, the present study generally agrees with those reported by Tay and Kortatsi (2007).

Cadmium concentration in the groundwater ranged from 0.000 to 0.002 mg/L with a mean concentration of 0.002 mg/L. The mean concentrations of cadmium were within the WHO guideline of 0.003 mg/L. Exposure to higher concentration of Cd may cause kidney damage as well as producing acute health effects (Momodu and Anyakora 2010). The mean concentration of Cd in this study was comparable to reports by Kortatsi (2007), Wassa West district; Asare-Donkor et al. (2015), Obuasi and Kuma (2004), Tarkwa.

In the study area, copper concentration ranged from 0.003 to 0.019 mg/L with a mean concentration of 0.010 mg/L. Zinc concentration in groundwater of the study area ranged from 0.000 to 0.047 mg/L with a mean concentration of 0.014 mg/L. Manganese concentration in the groundwater ranges from 0.005 to 0.020 mg/L with a mean value of 0.012 mg/L. The concentrations of manganese, copper and zinc in groundwater are within the maximum allowable limit as per WHO standard. Cadmium, manganese, copper and zinc are likely to be derived from the natural water–rock reaction processes since none of these metals exhibited concentration values outside the WHO (2011) limits. Study done by Nkansah et al. (2010) and Apau et al. (2014) indicated that the levels of Mn and Zn were comparable to this study. On the other hand, the levels of Zn and Mn from this study were higher than the study by Li et al. (2014), Henan-Liaocheng Irrigation Area; Zhang et al. (2013), Yellow River and Buschmann et al. (2008), Vietnam.

Pollution evaluation indices of water

The results of pollution evaluation indices (HPI) are presented in Table 4. The HPI values ranged from 319.20 to 688.05, with a mean of 374.47. The HPI results showed that all the samples were above the critical limit of 100 proposed for drinking water by Prasad and Bose (2001). The degree of contamination (Cd) was used as reference of estimating the extent of metal pollution (Rubio et al. 2000). The Cd values in the groundwater ranged from 0.26 to 26.80, with a mean of 7.79 (Table 4). According to Edet and Offiong (2002) and Backman et al. (1997), Cd may be classified into three categories as follows: low (Cd < 1), medium (Cd = 1–3) and high (Cd > 3) (Edet and Offiong 2002; Backman et al. 1997). As per the above classification, 21 % of the samples were classified as low zone, 16 % as medium zone and 63 % as high zone. The Cd and HPI indices show that most of the samples were highly polluted. The HEI values ranged from 2.25 to 29.88 with a mean value of 10.08. The proposed HEI criteria are as follows: low (HEI < 10), medium (HEI = 10–20) and high (HEI > 20). The HEI results show that 58 % of samples are within the low zone, 32 % fall within the medium zone and 10 % fall within the high zone. Cd, HPI and HEI values show similar trends at various sampling points (Fig. 3).

Table 4 Water pollution indices
Full size table
Fig. 3
figure 3

Spatial distribution of pollution evaluation indices

Full size image

Correlation study

Correlation analysis establish the relationships between heavy metal characteristics of water samples, which can reveal the sources and pathways of heavy metals that generated the observed water compositions (Azaza et al. 2011; Parizi and Samani 2013). A high correlation coefficient (near 1 or 1) means a good positive relationship between two variables and its value around zero means no relationship between them at a significant level of p < 0.05 (Varol and Davraz 2014). More precisely, it can be said that parameters showing r > 0.7 are considered strongly correlated whereas r between 0.5 and 0.7 shows moderate correlation (Manish et al. 2006). The correlation matrix of the heavy metals is given in Table 5. Mn shows a moderate negative correlation with Zn (r = −0.500, p < 0.05). Cu shows a moderate positive correlation with Zn (r = 0.580, p < 0.01). Pb shows a moderate positive correlation with Cd (r = 0.583, p < 0.01) and Fe (r = 0.459, p < 0.05). Fe does not show any significant correlation with other heavy metals, indicating that their distributions were not controlled by the same factor. The strong correlations between some of the heavy metals indicate the same input sources and similar geochemical behaviour (Lu et al. 2010; Saeedi et al. 2012). Therefore the associations of metals clearly indicate that the groundwater has assimilated various contaminants from the processes of chemical industries and landfill leachate or municipal sewage systems.

Table 5 Pearson correlation coefficient matrix for selected metals in groundwater samples (n = 19)
Full size table

Pollution source identification

Principal component analysis was used to further explore the extent of metal pollution and source identification following standard procedures (Dragovic et al. 2008; Franco-Uría et al. 2009). Varimax rotation (Gotelli and Ellison 2004) was used to maximize the sum of the variance of the factor coefficients which better explained the possible groups or sources that influenced the water system. Corresponding components, variable loadings, and the variances are presented in Table 6. Components loadings are classified by Liu et al. (2003) as “strong”, “moderate”, and “weak” corresponding to absolute loading values of >0.75, 0.75–0.50, and 0.50–0.30, respectively. R-mode analysis extracted two components with eigenvalues >1, which explained 64.24 % of the total variance. Positive scores in PCA indicate that water samples are affected by the presence of the parameters that are significantly loaded on a specific component, whereas negative scores suggest that water quality is essentially unaffected by those parameters. The 36.8 % of total variance is contributed by PC1 with higher positive loadings for Cu and Zn and moderate negative loadings for Mn. The occurrence of higher Zn concentration may be attributed to the greatest frequency of nearby sources like hazardous waste sites and the emission of industrial effluents. Components in PC1 are derived from mixed sources due to chemical induction of landfill leachate or municipal sewage (Bhuiyan et al. 2010). PC2 with high positive contribution of Pb and Cu, accounts for 27.4 % of the total variance. Despite the natural occurrence of Pb in the environment, anthropogenic sources such as discharge of various industrial effluents and public sewage also play a major role in the higher Pb loadings in the study area.

Table 6 Rotated component matrix of two-component model
Full size table

R-mode hierarchical cluster analysis

The R-mode HCA was used to determine the relationship among the various heavy metals using Ward’s method (Squared Euclidean distance as measure of similarity). Cluster analysis (CA) grouped the heavy metals into clusters on the basis of similarities within a group and dissimilarities between different groups. Parameters belonging to the same cluster are likely to have originated from a common source. The R-mode CA performed on the samples produced two clusters based on spatial similarities and dissimilarities (Fig. 4). The presence of Fe in cluster 1 reveals lithogenic contribution source of contamination and were identified in the higher contamination level. The second cluster includes Cd, Zn, Pb, Cu and Mn, which were derived from anthropogenic input.

Fig. 4
figure 4

Dendrogram of selected metals in water samples using ward’s method

Full size image

Human health risk assessment

Heavy metal pollution has being one of the important issues in environmental sciences. Ingestion of significant amounts of metal-containing drinking water will harm human health, resulting in several types of cancers (Wu et al. 2009; Yu et al. 2010). Using the measured data for the 6 heavy metals, the HQ were calculated using USEPA risk assessment models (US EPA 2011). The human health risk assessment of Cd, Fe, Cu, Pb, Zn and Mn showed HQ values to be <1 suggesting an acceptable level of non-carcinogenic adverse health risk (Table 7). The average HQ values of Pb in the water samples (0.4924) were highest, followed by Cd (0.3267), Mn (0.0101), Fe (0.0089), Cu (0.0029) and Zn (0.0007). The HI values of all the 6 metals (Cd, Fe, Cu, Pb, Zn, Mn) ranging from 0.0001 to 0.2022 were also <1, indicating acceptable risk for non-carcinogenic adverse health effect. The HQ values were found to be comparable with study by Muhammed et al. (2011), Kohistan region of Pakistan; Çelebi et al. (2014), Melenwatershed, Turkey and Sprang et al. (2009), European rivers. On the other hand, the levels of Cr were lower than the study by Keleperzis (2014), Thiva area of Greece.

Table 7 Non-carcinogenic risk (hazard quotient, HQ) and overall toxic risk (hazard index, HI) of water from hand-dug wells in Ejisu-Juaben Municipality
Full size table

The health risk associated with drinking water depends on the volume of water consumed and the weight of the individual. In this regard, health risk assessment associated with the exposure duration (ADD) was determined using the concentration of Mn, Fe, Cu, Pb, Zn and Cd in the drinking water and the results are presented in Table 8. The ADD values ranged from 6.12 × 10−5 to 2.00 × 10−4, 2.45 × 10−5 to 8.04 × 10−2, 3.67 × 10−5 to 2.00 × 10−4, 1.22 × 10−5 to 5.00 × 10−4, 2.45 × 10−5 to 6.00 × 10−4 and 1.22 × 10−5 to 2.45 × 10−5 mg/kg/day for Mn, Fe, Cu, Pb, Zn and Cd, respectively (Table 8). Thus, the health risk assessment indicates an alarming situation within this study area. It is recommended that the use of the water from these contaminated hand-dug wells for domestic purposes must be discontinued, or appropriate remediation technology must be applied to save the health of the human population in the study area.

Table 8 Exposure duration, ADD (mg/kg-day) for selected hand-dug well in Ejisu-Juaben Municipality
Full size table

The average Chronic daily intake (CDI) levels for carcinogenic risk of Mn, Fe, Cu, Pb, Zn and Cd were found to be 3.68 × 10−4, 1.51 × 10−2, 2.89 × 10−4, 4.38 × 10−4, 5.02 × 10−4 and 4.19 × 10−5, respectively (Table 9). The CDI indices were in the order: Fe > Zn > Pb > Mn > Zn > Cu > Cd. In general, a CDI value of 1.0 × 10−6 is the limit for acceptable health risk (USEPA 2011). The total CDI of all the groundwater samples exceeds this acceptable value, indicating an increased cancer risk for individuals because of their lifetime exposure to these carcinogens (Iqbal and Shah 2013). Higher CDI values may be due to run-off from agricultural fertilization and fungicides, which intended to increase the concentration of these heavy metals and in turn can affect the water quality and ecosystem biodiversity (Li and Zhang 2010). Nguyen et al. (2009) determined that about 5 in 1000 people could suffer from cancer in Vietnam. In Nanjing, China, a study on six surface waters found carcinogenic value of 2.05–3.28 × 10−4 higher than the acceptable limits (Wu et al. 2009).

Table 9 Chronic daily intake in groundwater samples from selected hand-dug well in Ejisu-Juaben Municipality
Full size table

Conclusions

Pollution evaluation indices, principal component analysis, cluster analysis and correlation matrix have been used to assess the intensity and sources of pollution in groundwater samples from the Ejisu-Juaben Municipality, Ghana. The study is summarized as follows:

  1. 1.

    The mean values of the heavy metal contents in the groundwater follow the decreasing order: Fe > Zn > Mn > Pb > Cu > Cd. The mean values of Pb and Fe in the groundwater were generally high compared to the threshold limits allowable for drinking water.

  2. 2.

    The human health risk assessment showed hazard quotient (HQ) and Hazard index (HI) values to be <1, suggesting an acceptable level of non-carcinogenic adverse risk, however, the cancer risk exceed the acceptable limit of 1.0 × 10−6, indicating that cancer risk may occur.

  3. 3.

    Principal component analysis with the support of cluster analysis identified both natural source and anthropogenic activities as the main contributing factors of metal profusion in the groundwater.

  4. 4.

    According to correlation analysis, highly strong correlations were observed among some of the heavy metal pairs, suggesting common sources, mutual dependence and identical behaviour during transport.

  5. 5.

    The Cd, HPI, and HEI concentrations show that 63, 100, and 10 % as highly polluted due to the municipal sewage being the main sources in the study area. The Cd and HEI concentration indices values increased trend along the study area may be due to the agriculture runoff, discharge of industrial wastewater, and municipal sewage through the soil.