Introduction

Drinking water is essential for life and is an important source of essential elements. Exposure to trace and toxic elements often occurs through drinking water. Concentrations of major and trace elements (essential and toxic) in drinking water have been investigated in several previous studies (Avino et al. 2011; Birke et al. 2010; Da Silva et al. 2000; Dinelli et al. 2012; Lee et al. 1999; Stetzenbach et al. 1994). Studies have been carried out recently on drinking water quality in Riyadh, KSA (Al-Saleh and Al-Doush 1998; Hashem 1993; Khanfar 2010; Al Nouri et al. 2014); however, only a limited range of parameters were investigated.

Worldwide, bottled water consumption has greatly increased (Datamonitor 2014), which means that there should be regular independent tests on its quality. To ensure this, worldwide standards for bottled drinking water have been set by different organizations; for example, the European Union (EU) (through the Drinking Water Directive), Food and Agriculture Organization (FAO), United States Environment Protection Agency (USEPA) and the World Health Organization (WHO) (Birke et al. 2010).

Bottled drinking water and treated drinking water are the main types of drinking water available in Najran city. Groundwater wells in the Najran valley are the main source of treated drinking water. After extraction, this groundwater is transported by tankers to water treatment stations across the entire city for treatment. Untreated groundwater from wells is transported directly to homes and businesses for domestic use, where it is stored in reservoirs under houses and businesses, and then is later pumped upstairs to large tanks, from where it is distributed through pipes for domestic use. Each house in the city has a complete, independent, isolated water distribution system.

The main aim of this study was to investigate the levels of major elements, non-metal elements (anions), trace elements (essential and toxic) and other parameters (physicochemical properties) in drinking water, and then to compare them with national and international guidelines. A further aim was to relate drinking water quality to public health. The findings of this research may provide further opportunities for drinking water research and treatment, which could be of benefit for public health in the KSA.

Materials and methods

Water sample collection

All samples were collected during May–July 2012. The following water samples were collected: (1) groundwater samples (n = 11) from different wells in the Najran valley, (2) treated drinking groundwater (n = 13) from commercial stations distributed throughout the city, and (3) bottled drinking water (n = 24) of different brands, purchased from local markets and supermarkets.

Samples were collected in polyethylene bottles that were soaked overnight in 15 % nitric acid. The soaked polyethylene containers were washed with deionized water and dried at room temperature. Afterwards, the containers were rinsed several times with the water source to ensure sufficient flushing before collection. The method of collection was similar to those reported in previous studies (Reimann et al. 2003; Barati et al. 2010; and Birke et al. 2010). All samples were stored in a refrigerator before transferred to other locations for analysis.

Analytical methods

pH and EC were measured on samples at room temperature upon arrival in Najran University, using a Lovibond sensodirect pH-200 potentiometer and a Lovibond sensodirect-200 conductometer.

Cations were measured by inductively coupled plasma-mass spectrometry (ICP-MS) at Nottingham University, Nottingham, UK. TH was calculated from Ca to Mg concentrations: Hardness in mg/L = 2.497 (Ca in mg/L) + 4.118 (Mg in mg/L) (Bryant and McPhilliamy 2002; APHA 1999). The samples were shipped by courier to Nottingham.

Anions were measured by ion chromatography (IC) at the General Directorate of Water, Ministry of Water and Electricity, Abha, KSA. TDS was also measured at the same laboratory. Samples were transferred to Abha in cold condition.

Inductively coupled plasma-mass spectrometry

Multi-element analysis of water samples was carried out using ICP-MS (X-SeriesII; Thermo-Fisher Scientific Inc., Waltham, MA, USA) with collision cell technology with energy discrimination. A hexapole collision cell (7 % hydrogen to helium) was used to remove polyatomic interferences. Samples were introduced to the ICP-MS plasma by an autosampler (Cetac ASX-520 with 4 × 60 place sample racks), through a concentric glass venturi nebulizer (Thermo-Fisher Scientific; 1 mL/min). Details of the ICP-MS operating conditions are shown in Table 1.

Table 1 ICP-MS Thermo-Fisher Scientific X-Series II operating conditions
Full size table

Internal standards used included scandium (Sc, 100 µg/L), rhodium (Rh, 20 µg/L), germanium (Ge, 20 µg/L) and iridium (Ir, 10 µg/L), prepared with 2 % trace analysis grade HNO3 (Fisher Scientific, UK). These were introduced to the sample stream via a T-piece. External multi-element calibration standards (Claritas-PPT grade CLMS-2 from SPEX Certi-Prep Ltd, Stanmore, Middlesex, UK) included Al, As, Ba, Cd, Co, Cr, Cs, Cu, Fe, Mn, Mo, Ni, Pb, Rb, Se, Sr, U, V and Zn, concentrations of which were in the range blank (0) up to 100 µg/L (0, 20, 40, 100 µg/L). A tailored external multi-element calibration solution (PlasmaCAL, SCP Science, France) was used to create Ca, Mg, Na and K standards in the range 0–30 mg/L (0, 10, 20, 30 mg/L). Ti was determined semi-quantitatively.

Plasmalab software (version 2.5.4, Thermo-Fisher Scientific) was used for sample processing by an average calibration from two blocks, run before and after samples. Internal cross-calibration was applied.

Ion chromatography

Anions (Cl−, F−, SO4 2− and NO3 −) were measured with a Metrohm 850 Professional IC. The IC was equipped with a Metrosep A Supp 7—250/4.0 column, self-regenerating suppressor that used sodium carbonate (3.6 mM) eluent with a flow rate of 0.8 ml/min and conductivity detection. The recording time was 31 min, pressure was 10.36 MPa and the operating temperature was 45 °C. MagIC Net software was used for the calculations. The calibration standards were in the range (0, 10, 20 and 40 mg/L for Cl− and NO3 −), 0, 1, 2, and 3 mg/L for F− and 0, 20, 40 and 60 mg/L for SO4 2−. Recoveries for random blind samples (n = 3) were 106, 90, 104 and 111 % of Cl−, F−, SO4 2− and NO3 −, respectively.

Quality control

Several blanks were included in the measurement process to assess quality control, which included checking with international certified water standard reference material and in-house standard measurements.

The ICP-MS performance was checked for sensitivity [using Lithium (Li), Indium (In) and Uranium (U)], stability (using CeO/Ce ratio and Ba2+ ions) and the background signal in standard mode before converting to ‘Collision Cell technology with Kinetic Energy Discrimination’ mode (CCT-KED). Ten operational blanks were run to determine the limits of quantification (LOQ = 10 × SD), and three replicates were used for limits of detection (LOD = 3 × SD). Two samples of the CRM NIST 1643E (trace elements in water) were run after the 10 operational blanks for quality assurance (QA) purposes. The results of limit of detection (LOD), limit of quantification (LOQ) and certified reference material (CRM) from National Institute for Standards and Technology (NIST); CRM NIST1643e are shown in Table 2.

Table 2 The LOD, LOQ of the 24 elements and the results of CRMNIST 164E using ICP-MS Thermo-Fisher Scientific Series II
Full size table

Results

The guideline values of WHO, USEPA, GSO and SASO (WHO 2008; USEPA 2009; GSO 2008; SASO 1984) for drinking water are summarized in Table 3. A summary of the parameters (Physicochemical properties and the concentration of anions, major and trace elements) for the three types of water samples (ground water treated drinking water and bottled drinking water) is shown in Table 4. Levels of all parameters are presented in mean, standard deviation (SD), minimum (Min) and maximum (Max).

Table 3 WHO and SAS guideline values for drinking water
Full size table
Table 4 Summary of the parameters measured in different water samples collected from Najran city
Full size table

Figure 1 shows comparison of the levels of major elements, anions and physicochemical properties in bottled drinking water, between labeled values (mean) and measured values (mean).

Fig. 1
figure 1

Comparison of the levels of major elements, anions and physicochemical properties, in bottled drinking water, between labeled values (mean) and measured values (mean)

Full size image

Discussion

Physiochemical properties

The average pH for ground water was slightly lower than the average values recorded for both treated drinking water and bottled drinking water (Table 4). The difference may be due to the coagulant material used for water purification. The pH of the treated drinking water samples’ range overlaps with the range for bottled drinking water samples, and is within the WHO guidelines (WHO 2008), including the ground water samples. Average pH values for drinking water in this study were generally higher than those for groundwater samples, indicating a positive outcome from water treatment processes and/or the presence of carbonate. Drinking water pH values should be controlled at all stages of water treatment and distribution to avoid any harmful effects to human health, and to minimize corrosion of water distribution systems (WHO 2003a).

In ground water, the average values of EC, TH and TDS were very high compared with those obtained for treated drinking water and bottled drinking water samples (Table 4). The TH was calculated from Ca to Mg concentrations, which are the highest in ground water samples. The TDS and EC values have correlation, as a rough rule one ppm of TDS correlates to a conductivity of two µS/cm (Hydranautics 2014). The mean value of TDS in ground water samples is approximately fourfold compared to mean values recorded for treated drinking water samples and bottled drinking water samples (Table 4). The mean values of the EC, TH and TDS in treated drinking water samples were the lowest, which may be due to the softening process. This also may be attributed to ion removal during the treatment process. The values of EC, TH and TDS in both types of drinking water samples were much lower than those recommended by the SASO (1984) and WHO (2008) guidelines (Table 3). It is worth mentioning that some TH and TDS values for ground water samples (three samples only) exceeded the guideline values set by WHO (2008).

Anions

As shown in Table 4, average concentrations of anions (F−, Cl−, NO3 − and SO4 2−) in ground water were higher than those in both treated drinking water and bottled drinking water samples. The following range of concentrations for anions (Cl−, NO3− and SO4 2− (10–100 mg/L) and F− (0.1–1 mg/L) in drinking water is considered essential for human health (Foster et al. 2006). One sample among the ground water samples had the highest average concentrations for Cl− (364 mg/L). The highest value of Cl− was below SASO guideline value, and higher than WHO and USEPA guideline values. However, the anions’ levels in the rest of the samples were below the national and international guideline thresholds (Table 3).

The average Cl− concentration in bottled drinking water was almost twice that of treated drinking water, and showed greater variation than that of treated drinking water (Table 4). There are different sources (brands—different aquifers) of bottled drinking water while treated drinking water samples came from one source (Najran valley—same aquifer). Despite this variation, the Cl− concentrations in both types of drinking water do not exceed the international and national guideline values (Tables 3, 4). It is noteworthy that, to date, no health-based values have been proposed for Cl− in drinking water (WHO 2003b).

Only one sample among the ground water samples had high concentration for SO4 2− (1132 mg/L), exceeding the recommended values for all organizations. If SO4 2− concentrations in drinking water are high, there is a noticeable taste, and concentrations of 1000–1200 mg/L have a laxative effect on humans. However, there is no health-based guideline value for SO4 2− in drinking water (WHO 2004). SO4 2− concentrations in both types of drinking water were below the USEPA (2009), WHO and SASO (1984) guideline values.

Occurrence of F− and concentrations in samples varied to such a degree that samples fell into two distinct groups. F− was not detected in seven of the treated drinking water samples (n = 13), and four of the bottled drinking water samples (n = 24). However, the fact that the F− ion is present in all but two ground water samples, the source of the treated drinking water, suggests that this anion is depleted by the treatment processes. These findings agree with results of a recent study that reported average F− concentrations of 0.35–0.82 mg/L in ground water in Najran area (Alabdulaaly et al. 2013). More importantly, our data suggest that the population of Najran, even if only ingesting 3L of the available drinking water daily, is exposed to low levels of F−. Indeed, the drinking water intake requirements for adults in normal conditions are in the range 2.5–3 L/d; in hot climates, it is greater than 3L/d (Popkin et al. 2010), while in a fluoridated area in a temperate climate, the daily intake of fluoride in drinking water is 0.6 and 2 mg/adult (Fawell et al. 2006). It is very difficult to find a study on the adverse health effects associated with fluoride deficiency in drinking water. Dean et al. (1939) showed that, in areas where fluoride levels in drinking water were low; there was a high incidence of dental caries. Further, side effects from a fluoride-deficient diet include weak bones and teeth and increased cavities (MedlinePlus 2014). Conversely, several other studies, undertaken in 28 different countries worldwide, have indicated that exposure to high drinking water fluoride concentrations is associated with human dental fluorosis (Fawell et al. 2006). In particular, studies carried out by Akpata et al. (1997) and Albdulaaly et al. (2013) in the Hail region of the KSA reported incidences of dental fluorosis among children, and well water fluoride concentrations ranging between 0.3 and 4.0 mg/L. Another series of (more than 20) papers concluded that children exposed to high levels of fluoride in drinking water had lower IQs (Choi et al. 2012).

NO3 − in drinking water come from environmental (geogenic) and anthropogenic sources, and include runoff from fertilizer, municipal sewage treatment systems, leaching from septic tanks, erosion of natural deposits and animal wastes (Ricker et al. 2001; EPA 2009). High NO3 − concentrations in drinking water reflect contamination of water sources (Azrina et al. 2011). The measured concentrations of NO3 − in all three types of water samples did not exceed the guideline values set by SASO, GSO and WHO. However, only three samples from both treated drinking water and bottled drinking water samples were exceeded the threshold of the USEPA (2009). Moreover, seven samples from ground water samples were also exceeded the guideline value of USEPA (Table 3).

Major elements

Actually the major elements with the range of concentrations Na and Ca (10–100 mg/L) Mg and K (1–10 mg/L) in drinking water are considered as essential to human health (Foster et al. 2006). In this study, the measured average concentrations of these major elements in ground water were higher than those for both treated drinking water and bottled drinking water (Table 4). These differences suggest that these elements are significantly depleted during the water treatment processes. The treatment seems to determine the element concentration ratios, as bottled drinking water has lower K and Ca concentrations, and higher Na and Mg concentrations, than treated drinking water. The measured concentrations for all major elements in all samples did not exceed the national and/or international guideline limits. Except only one sample for Ca concentration, among ground water samples, was exceeded the limit values for all organizations. Likewise, this sample was the same sample that recorded higher values for Cl−, SO4 2−, NO3 −, TDS, TH and EC as previously mentioned.

Trace elements

Only 1 % of naturally occurring dissolved constituents in groundwater is made up of trace elements. The trace elements are divided into two categories with range of concentrations—essential to human health: V, Se, Co, Ni and Cr (0.1–1 µg/L), Ba, Cu and Mn (1–10 µg/L), Fe and Zn (10–100 µg/L) and Sr (100–1000 µg/L); toxic elements: As, Cd, Pb and Al (0.1–1 µg/L), U (1–10 µg/L). (Foster et al. 2006).

The above-mentioned sample (ground water sample) that recorded higher values for Cl−, SO4 2−, NO3 −, TDS, TH and EC was also recorded highest value for Sr. In this case, Sr value did not exceed the recommended value (AWWA 2014a, b, Table 3). The concentrations of trace elements for the three types of water samples are presented in Table 4. Mean concentrations of 20 trace elements in the three types of the tested water samples were very low and did not exceed the guideline values of all organizations. Ground water samples were recorded higher values for V, Cr, Ni, Cu, Zn, As, Sr, Mo, Ba, Se and U than both treated drinking water samples and bottled drinking water samples. The presence of V, As, Sr, Mo, Ba and U in ground water was not surprising, because these elements are naturally occurring elements (WHO 2008; AWWA 2014a). Nevertheless, the occurrence of Cr, Ni, Cu and Zn could be due to containers and distributing system. Previous study reported that Cd, Fe, Cu and Zn in drinking water samples may be the result of these elements leaching from the storage tanks and/or distribution pipes (Al-Saleh and Al-Doush 1998). Fortunately, groundwater was not used for drinking purpose. This can be avoided by improving the water distribution system and by regular maintenance. At lower levels, these elements are considered safe, as shown in Table 3. Some elements such as Ti, Cd, Cs and Pb were less than LOQ for all tested water samples. This is promising because these elements considered to be toxic and could pose risk to human health (Foster et al. 2006; Holbrook et al. 2013). Fe and Co values were very low, just twice as LOQ for each element, in all tested samples. Both elements are considered essential to human health in drinking water with the following range of concentrations [Co (0.1–1 µg/L) and Fe (10–100 µg/L)] (Foster et al. 2006). Therefore, adequate quantity of these elements can not be obtained from such drinking water tested in this study. The range of mean values (1.3–2.5 µg/L) for Al was found to be in the following order: treated drinking water > bottled drinking water > groundwater. This measured range of Al in all tested samples is 80 times lower than the guideline set by WHO (2008) in drinking water. Thus, this is a promising result for such a toxic element in drinking water. The highest average value (0.5 µg/L) for Mn was recorded in bottled drinking water samples. Thus far, this measured value is lower than the essential level of Mn in drinking water. Average values of Rb were reported very low in all tested water samples. The values were far below the lower limit for Rb level (0.1–100 µg/L) in natural groundwater FWQP (2003).

The Sr in naturally occurring element and the guideline value (4200 µg/L) for Sr is very high as shown in Table 3, compared to all trace elements, except for Zn, Cu and Fe. The increasing order of Sr-measured values is ground water > treated water > bottled water, which explain that the geographic However, the recorded values in all three water types did not exceed the guideline value.

The Sr mean value (74.82 µg/L) of this study in bottled drinking water was not high compared to other studies, median (179 µg/L) in bottled drinking water and median of European bottled water (406 µg/L) and mean of Sicilian bottled natural waters (298 µg/L) area is the source of this element. (Birke et al. 2010; Varrica et al. 2013). The reported average value of Sr in ground water and treated water samples could be attributed to the geogenic source of the water.

Groundwater could have good benefit after selective removal of harmful substances. This is because the ground water samples showed high levels of essential elements such as Zn and Se.

Zn is an essential element to humans and other living organisms, in terms of enzymatic function. Approximately, 200 zinc-containing enzymes have been identified, including dehydrogenases, phosphatases, peptidases and polymerases (WHO 1996; O’Dell 1984). Recommended Dietary Allowances (RDAs) for Zn are 8 and 11 mg for adult females and males, respectively (NIH 2014). The average daily intake for an adult from ground water source in Najran is calculated as 282.3 μg, if the adult ingested up to 3 L water per day.

Selenium (Se) is an essential element for humans. Therefore, the recommended daily intakes of Se were jointly set by World Health Organization and Food and Agriculture Organization as 26 μg per day for adult females and 35 μg of Se per day for adult males (WHO 2011). Based on the results of this study, the mean of Se in ground water samples is 2.83 μg/L (Table 4). Our data suggest that if an adult in Najran ingested 3 L of ground water per day, the daily intake will be 8.5 μg. This daily intake will be third of the recommended daily value for adult women and forth of the recommended daily value for adult males.

Results show that the drinking water quality in Najran generally meets the international standards. Toxic elements such as Cd, Al, As, and Pb were either present in minute quantities or not detectable at all, and were below the WHO, USEPA and GSO guideline values. As such, human consumption of treated drinking water or bottled drinking water in Najran should not lead to adverse health effects typically associated with these elements. The levels of trace elements in the bottled and treated drinking water samples do not suggest any implications for human health, based on international/national guideline values (Tables 3, 4).

It is noteworthy that none of the measured elements in drinking water samples in this study exceeded the international guideline values. Therefore, one can conclude that the drinking water quality in Najran city complies with international guidelines.

Quality measurement for bottled drinking water

Comparison of label contents, in all bottled drinking water samples, was established with measured values in this study (Fig. 1). A good agreement showed between average values of parameters on the labels of bottled water and measured parameters in this work. This provides reliable information on labels of bottled drinking water. However, TDS average value was slightly higher for measured results than the reported values on labels. This may be due to reporting the TDS values on bottles’ labels as range in some bottles rather than one single value.

Conclusions

Based on all measured parameters, drinking water quality in Najran city complies with national and international standards. However, some essential elements for human health such as fluoride iron and cobalt either hardly detected or near to limit of quantification. Fluoride was not detected in 53.8 % samples of treated drinking water and 16.7 % samples of bottled drinking water. Ion and cobalt cannot be quantified in 61.5 and 46.2 % of treated drinking water samples, 33.3 and 25 % of bottled drinking water samples, respectively.

Drinking water in Najran city should be given careful attention, especially for F−, Fe and Co in treated drinking water. This study is important as it will increase public awareness and knowledge about drinking water, and also serves as a contribution to, and promotion of, public health. More and regular studies are recommended, in this field, to enrich the demonstrated information in this study about the drinking water.