Assessment of Radioactivity Concentrations and Associated Radiological Health Risk in Natural Spring Mineral Bottled Drinking Water from South Africa

1. Introduction

The availability and accessibility of safe drinking water is a global critical public health issue, especially in regions where natural and anthropogenic factors may contribute to the quality of water. In the republic of South Africa, natural spring mineral bottled drinking water is a popular alternative to tap and other sources of drinking water, as obtainable globally, often perceived to be safer and of higher quality [1,2,3,4]. With the rising global population, sanitation and global warming effect, the demand for and consumption of bottled drinking water is estimated to increase in the coming years [5,6,7]. However, the activity concentrations of the naturally occurring radionuclides in the commercial bottled drinking water can pose some potential radiological health risks to consumers, if not regularly checked and regulated, especially with the increased demand for drinking water [8,9,10].

Naturally occurring radioactive materials (NORM) such as uranium-238, thorium-232, and potassium-40, along with their progenies, are ubiquitous in the earth and can be present in drinking water sources, as well [2,11]. These radionuclides get into water supplies through the geological formations and human activities, potentially leading to long-term exposure when ingested. Continuous consumption of water with elevated levels of radionuclides may increase the risk of adverse health effects, including cancer and other radiation-induced diseases [12,13].

Studies on the activity concentration of natural radionuclides in bottled drinking water have been reported [9,14,15,16], using various analytical procedures with varied activity concentrations. However, careful search of literature showed that such work has not been report on the South African bottled drinking water, hence, this baseline study aims to assess the activity concentrations of the radionuclides in natural spring mineral bottled commercial drinking water brands mostly consumed in South Africa. Using the High-Purity Germanium (HPGe) gamma spectrometry technique, we measured the levels of radioactive contaminants in the various bottled drinking water samples. By finding the levels of natural radioactive contaminants, we estimated the associated radiological health risks because of the consumption of bottled drinking water and compared with national and international safety standards as well as other published related data. The findings in this research will provide valuable insights into the safety of consuming natural spring mineral bottled drinking water in South Africa and inform regulatory bodies to ensure public health protection.

In this article, we will discuss the methodology used for sampling and analysis, present the results of the activity concentrations, and analyze the potential health risks based on the national and international guidelines and standards. Our sole objective is to contribute to the ongoing efforts to safeguard public health by highlighting the importance of monitoring and regulating the levels of natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K, in drinking water, for sustainable growth and development, aligning with goal 6 of the sustainable development goal.

2. Materials and Methods

The twenty-one different brands of water samples analysed in this study are made up of the most available and topmost consumed bottled natural spring mineral commercial drinking water, collected from the grocery stores after mini survey, in south Africa [6]. The samples were prepared according to the specified standards by the South African Water Quality Guidelines [2] and world health organization [10]. The water samples were poured into a 0.5 L cylindrical polyethylene vials based on the gamma detector geometry and then sealed using black polyethylene adhesive tape, to make it airtight and stored for a period of thirty days for the purpose of secular equilibrium to be attained between the parent and daughter natural series radionuclides of ²³⁸U (²²⁶Ra) and ²³²Th, in Figure 1.

A Canberra n-type well High-Purity Germanium (HPGe) gamma detector, model GCW2021 (USA), connected to an MCA of model DSA 1000, incorporated with software for analysis known as Genie 2000, at the Centre for Applied Radiation Science and Technology (CARST), North-West University (Mahikeng Campus) was used in this study for data acquisition and spectrum analysis, as shown in Figure 1. The spectrometer has energy resolution of 2.1 keV FWHM at 1.33 MeV of ⁶⁰Co gamma line, 20 % relative efficiency. A multinuclide Europium-152 standard source(s) which can emit a wide range of gamma rays between 46.54 - 1836 keV, applied to calibrate of the detector, in terms of the energy (keV) and efficiency (%) with the curve shown in Figure 2. The prepared water samples were each placed into a cylindrical enclosure made of lead with low-background, 11.50 cm thick on the detector, and counted for 12 hours. Prior to the determination of the activity, the value of the background was subtracted from the gamma radiation spectrum.

The equipment validation using international atomic energy agency (IAEA) standardised reference materials such as ¹⁵²Eu, ¹³³Ba and IAEA-RGU-1, to ascertain the level of accuracy of the detector. It was by determining the activity of the sources at production, from their respective certificates and then at present after decay, using the online Rad Pro decay calculator, and comparing with the measured activity by the detector. The results showed agreement with calculated activity of ¹⁵²Eu to be 25.447 kBq while measured activity to be 26.027 kBq with percentage error of 2.281 %. The laboratory has recently participated in the IAEA interlaboratory proficiency exercise 2024.

In this study, the various gamma lines of the radionuclides together with their daughters used to determine the activity concentrations values by computing their weighted mean values with their respective uncertainties. The gamma line peak energy used for ²³⁸U (²²⁶Ra) progenies were ²¹⁴Pb (295.21 keV, P_γ = 19.2% and 351.95 keV, P_γ = 37.20%) and ²¹⁴Bi (609.31 keV, P_γ = 46.30% and 1120.29 keV, P_γ = 15.10%). For ²³²Th the gamma lines of its progenies, ²¹²Pb (238.63 keV, P_γ = 44.60% and 727.17 keV, P_γ = 11.80%), ²¹²Bi (727.17 keV, P_γ = 11.80%) and ²²⁸Ac (911.60 keV, P_γ = 27.70% and 969.11 keV, P_γ = 16.60%) were used, while for ⁴⁰K, its single gamma lined at 1460.81 keV, P_γ = 10.67% was used.

The activity concentration A (BqL^-1) was calculated using Equation 1, which was preceded by the subtraction of decay corrections [17].

$$\mathrm{A}(\mathrm{Bq}{L}^{-1})={\displaystyle \frac{{\mathrm{N}}_{\mathrm{a}}}{{\mathrm{P}}_{\mathsf{\gamma }}·{\mathsf{\epsilon }}_{\mathsf{\gamma }}·\mathrm{t}·\mathrm{V}}}$$

(1)

where ${\mathrm{N}}_{\mathrm{a}}$is the net peak area of a peak at a given energy line in the spectrum, P_γ is the emission probability of gamma rays at the given energy photopeak, ε_γ is the efficiency of detecting a radionuclide, t is the total time taken in seconds (s) for the sample counting and V is the sample volume, in litres (L).

The radium equivalent (Ra_eq) which describes the output of gamma rays, from a sample and the associated radiological hazard, provides a base guideline useful in the regulation of radiation safety standards. It is estimated by assuming that 370 BqL^-1 of ²²⁶Ra, 259 Bq L^-1 of ²³²Th and 4810 Bq L^-1 of ⁴⁰K produces gamma radiation of the same dose rate, given by Equation 2 [17,18,19,20]

$$R{a}_{eq}(\mathrm{Bq}{L}^{-1})={\displaystyle \frac{A_{Ra}}{370}}+{\displaystyle \frac{A_{Th}}{259}}+{\displaystyle \frac{A_K}{4810}}\times 370$$

(2)

where $A_{Ra}$, $A_{Th}$ and $A_K$ are the average activity concentrations of ²²⁶Ra, ²³²Th and ⁴⁰K, respectively.

Dose rate (D) is the radiation dose absorbed per hour due to terrestrial or NORMs sources of gamma radiation in air or material. Equation 5 is used to calculate the absorbed dose rate according to [20,21].

$$\mathrm{D}(\mathrm{nGy}{h}^{-1})=0{.462\mathrm{A}}_{\mathrm{Ra}}+0{.621\mathrm{A}}_{\mathrm{Th}}+0{.0417\mathrm{A}}_{\mathrm{K}}$$

(3)

where $A_{Ra}$, $A_{Th}$ and $A_K$are the mean activity concentrations of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K and their respective coefficients 0.462, 0.621 and 0.0417 as dose conversion factors in nSv/h/Bq/L.

The total annual effective ingestion dose (TAEID) because of the various radionuclides in mSvy^-1 in the drinking water estimated for the members of the public for the age categories of infant (≤1y), children (1-2y), teenager (12-17y) and adult (>17y). It is the radiation dose absorbed or received in a year by the ingestion of bottled water and computed using Equations 4 and 5, based on the data from Table 1 and Table 2 [10,12,22].

$$AEID \left(mSv{y}^{-1}\right)=A_c\times I\times F_{dc}$$

(4)

$$TAEID \left(mSv{y}^{-1}\right)={\displaystyle \sum }\left(A_c\times I\times F_{dc}\right)$$

(5)

where AEID is the annual effective ingestion dose in mSvy-1 because of radionuclide $A_c$ is the radionuclide activity concentration in the bottled drinking water sample in BqL^-1, I is the drinking water annual intake rate shown in the 5th row of Table 1 and $F_{dc}$ is the effective dose coefficient of the relevant radionuclide for the various age group as provided for by the [20,22].

Due to the sensitivity of the gonads (testes and ovaries) and organs such as the bone surface and active bone marrow to ionising radiation, the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) has identified them as organs of interest because if they are exposed to radiation, reproductive health could be affected. Hence, the index, annual gonadal equivalent dose (AGED) has been set to serve as a measure of ionising radiation that the gonads received for one year. The AGED (mSvy^-1) because of radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K can be determined by Equation 6.

$$\mathrm{AGED}\left(\mathrm{Sv}{y}^{-1}\right)=3{.09\mathrm{A}}_{\mathrm{Ra}}+4{.18\mathrm{A}}_{\mathrm{Th}}+0{.314\mathrm{A}}_{\mathrm{K}}$$

(6)

where ${\mathrm{A}}_{\mathrm{Ra}}$, ${\mathrm{A}}_{\mathrm{Th}}$ and ${\mathrm{A}}_{\mathrm{K}}$ are the activity concentrations of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K (BqL^-1) with the respective coefficients as conversion factors.

The internal and external hazard indices (H_in and H_ex) used to quantify the level of risk due to internal and external exposure to radon which is carcinogenic and the short-lived progenies. Both H_in and H_ex calculated using Equations 7 and 8, either of which must be less than one for a negligible level of radiation risk [17,21,23,24].

$${\mathrm{H}}_{\mathrm{in}}={\displaystyle \frac{{\mathrm{A}}_{\mathrm{Ra}}}{185}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{Th}}}{259}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{K}}}{4810}}\le 1$$

(7)

$${\mathrm{H}}_{\mathrm{ex}}={\displaystyle \frac{{\mathrm{A}}_{\mathrm{Ra}}}{370}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{Th}}}{259}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{K}}}{4810}}\le 1$$

(8)

where A_Ra, A_Th and A_K are the activity concentrations of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K, in (BqL^-1) with the respective coefficients as conversion factors.

The representative alpha gamma level indices (I_α and I_γ) are used to estimate levels of risks in relation to exposure to annual effective ingested doses by humans from alpha and gamma radiations of radionuclide source. I_α and I_γ are determined using Eqs. 9 and 10 such that, the value of annual effective ingested dose increases by 0.3 µSv if I_γ ≤ 2, but for 2 ≤ (I_γ) ≤ 6, it will increase by 1.0 µSv. Therefore, (I_γ) can be used for materials identification purpose [17,24].

$${\mathrm{I}}_{\mathsf{\gamma }}={\displaystyle \frac{{\mathrm{A}}_{\mathrm{Ra}}}{300}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{Th}}}{200}}+{\displaystyle \frac{{\mathrm{A}}_{\mathrm{K}}}{3000}}$$

(9)

$${\mathrm{I}}_{\mathsf{\alpha }}={\displaystyle \frac{{\mathrm{A}}_{\mathrm{Ra}}}{200}}$$

(10)

where A_Ra, A_Th and A_K are the specific activity concentrations of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K, in (BqL^-1).

Based on the International Commission on Radiological Protection (ICRP) and the South African Water Quality Guidelines [2] recommendations for adult members of the public, the stochastic effects of ionizing radiation due to the ingestion of bottled drinking water sources were estimated using Eqs 11 – 14. Additionally, other health risk parameters because of exposure from radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K yielding low dose of radiation considered to be chronic risk of effects on the somatic as well as hereditary organs, have also been estimated using lifetime fatal cancer risk and hereditary effects coefficient [25,26].

$$FCR=TAEID\times CR{F}_c$$

(11)

$$LFCR=TAEID\times A_g\times CR{F}_c$$

(12)

$$SHE=TAEID\times HE{F}_c$$

(13)

$$ELH=TAEID\times A_g\times HE{F}_c$$

(14)

where FCR is the fatal cancer risk, TAEID is the total annual effective ingestion dose, ${\mathrm{CRF}}_{\mathrm{c}}$ is the lifetime fatal cancer risk factor $\left(5\times {10}^{-2 }S{v}^{-1}\right)$, LFCR is the lifetime fatal cancer risk to adult, $A_g$is the adult lifetime age of 70 years [27] of continuous exposure to low level of radiation by the population, SHE is the severe hereditary effects, ${\mathrm{HEF}}_{\mathrm{c}}$ is the hereditary effect conversion factor and ELH is the estimated lifetime hereditary effects for the adult group.

3. Results

3.1. Radioactivity Concentration of Radionuclides

The radioactivity concentrations (BqL^-1) with their respective uncertainties (± Standard Deviation) of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K present in the mineral spring bottled water measured is presented in Table 2, columns 2 to 4. Additionally, the total annual effective ingestion dose (TAEID) (mSvy^-1) for the various age groups infants (≤ 1 year), children (1-12 years), teenagers (12-17 years) and adults (> 17 years) due to the intake of radionuclide in the natural spring mineral bottled water from South Africa, is in the same table but columns 5 to 8, computed using the information from Table 1. The mean activity concentration of ²²⁶Ra ranged from 1.060 ± 0.067 BqL^-1 in sample W-11 to 2.571 ± 0.143 BqL^-1 in sample W-2 with a mean value of 1.766± 0.399 BqL^-1 ; ²³²Th varied from 1.736 ± 0.112 BqL^-1 in sample W-14 to 7.807 ± 0.099 BqL^-1 in W-8, with a mean value of 3.688 ± 1.371 BqL^-1 and ⁴⁰K ranged from 149.000 ± 38.480 BqL^-1 in W-17 to 242.900 ± 59.700 BqL^-1 in W-5 with a mean of 220.229 ± 22.297 BqL^-1. The TAEID for the infants (≤ 1 y) varied from (1.303 - 4.639) ×10^-2 mSvy^-1, with a mean value of (2.422 ± 0.768) ×10^-2 mSvy^-1, children (1-12 y) range between (4.012 - 11.597) ×10^-3 mSvy^-1, with a mean of (6.669 ± 1.771) ×10^-3 mSvy^-1, teenagers (12-17 y) is between (0.945 - 3.305) ×10^-2 mSvy^-1, with a mean of (1.745 ± 0.549) ×10^-2 mSvy^-1, while for the adult (> 17 y) ranged from (2.074 - 5.168) ×10^-3 mSvy^-1, with a mean of (3.218 ± 0.749) ×10^-3 mSvy^-1. The distribution of the activity concentration levels of the natural radionuclides are also presented in Figure 3. The results of both the activity concentration of the natural radionuclides and the total annual effective ingestion dose are variable across the samples and comparable to similar studies in the literatures [21,28]

The estimated contribution of each of the analysed natural radionuclides to the total annual effective ingestion dose for the different age groups in the natural spring mineral bottled drinking water is shown in Fiture 4. The results show that ²³²Th contributed the highes percentage to the dose of 83.44 % for infants (≤ 1year), followed by 81.89 % for teenagers (12-17 years), 68.68 % for children (1-12 years), and 57. 78 % for the adults (> 17 years) group. ⁴⁰K contributed average percentage to the dose of 31.00 % for the adults (> 17 years), followed by 25.33 % for children (1-12 years), 10.30 % for infants (≤ 1year), and 7.01 % for teenagers (12-17 years) group. ²²⁶Ra contributed the lowest percentage to the dose of 11.23 % for the adults (> 17 years), followed by 11.10 % for teenagers (12-17 years), 6.26 % for infants (≤ 1year), and 6.00 % children (1-12 years) group.

Figure 4. Contribution of each of the analysed natural radionuclide to the total annual effective ingestion dose for the different age groups in the natural spring mineral bottled drinking water from South Africa.

Figure 4. Contribution of each of the analysed natural radionuclide to the total annual effective ingestion dose for the different age groups in the natural spring mineral bottled drinking water from South Africa.

3.2. Radiun Equivalent and Absorbed Dose Rate

The radium equivalent (Ra_eq) in unit of BqL^-1, which is the index of the combined activity concentration levels of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K in the natural spring mineral bottled drinking water, was evaluated using Equation 2, and result is presented in Figure 5. The result of Ra_eq showed maximum activity value of 28.252 ± 0.613 BqL^-1 and minimum activity value of 15.929 ± 0.268 BqL^-1, which are the respective radium equivalent activity values of sample W-8 and sample W-17, with an average value of 23.976 ± 0.446 BqL^-1. The absorbed dose rate D_Ab in nGyh^-1 was also estimated using Equation 3, and the result is presented in Figure 5. The result show that the maximum value of absorbed radiation dose rate is 14.092 ± 0.436 nGh^-1 from sample W-5, while the minimum value is 8.152 ± 0.207 nGyh^-1 from sample W-17 with a mean of 12.232 ± 1.445 nGyh^-1.

3.3. Annual Gonadal Equivalent Dose and Hazard Indices

The annual gonadal equivalent dose (AGED), internal and external hazard indices (H_in and H_ex), representative alpha and gamma indices ((I_α and I_γ) estimated results are presented Figure 6a. The results show that, the annual effective ingestion dose (AEID) varies between 0.040 ± 0.038 mSvy^-1 to 0.069 ± 0.054 mSvy^-1 in samples W-17 and W-5, with an average value of 0.060 ± 0.007 mSvy^-1. The AGED has maximum value of radiation dose from drinking water as 0.103 ± 0.037 mSvy^-1 while minimum value is 0.060 ± 0.017 mSvy^-1 in samples W-5 and W-17 respectively, mean is 0.090 ± 0.027 mSvy^-1. The mean values of the radiological health hazard indices, H_ex, H_in, I_γ and I_α are 0.065 ± 0.008, 0.070 ± 0.009, 0.098 ± 0.012 and 0.009 ± 0.002 respectively, as shown in Figure 6a.

3.4. Cancer Risks and Hereditary Effects

Figure 6b presents the results of the estimated cancer risks and hereditary effects due to the consumption of natural spring bottled drinking water, containing natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K. The fatal cancer risk (FCR) varies between 1.141 x 10^-4 to 2.842 x 10^-4, with a mean value of 1.769 x 10^-4 ± 4.119 x 10^-5. The lifetime fatality cancer risk (LFCR), taking into cognizance the reference Man age of 70 years, varies between 7.983 x 10^-3 to 19.897 x 10^-3, with a mean value of 12.389 x 10^-3 ± 2.883 x 10^-3. The severe hereditary effects (SHE) to adults per year range between 4.147 x 10^-6 and 10.336 x 10^-6, with an average value of 6.436 x 10^-6 ± 1.498 x 10^-6, while the estimated lifetime hereditary effect (ELHE) varies from 2.903 x 10^-4 to 7.235 x 10^-4 with a mean of 4.505 x 10^-4 ± 1.049 x 10^-4. These results are lower but comparable to those reported by [26]

3.5. Statistical Analysis

The descriptive statistical analysis (minimum, maximum, mean, standard deviation (STD DV), range, variance, skewness and kurtosis) of the measured activity concentration levels of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K together with the estimated radiological health risk parameters in the bottled drinking water is shown in Table 3 while the frequency distribution with the Gaussian distribution curve is presented in Figure 7. The statistical analysis was performed using the following statistical packages MS Excel, SigmaPlot version 15.0 and IBM SPSS Statistics 29.0.0.0.

Table 4 presents the statistical analysis of the Pearson correlation coefficient between the activity concentration levels of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K and the estimated radiological health risk parameters in this study. The purpose is to ascertain the order of mutual relationships between the pairs of the variables determined. It also test the interrelationship between the different models used in the estimation of the radiological health risks as well as the cancer risk and hereditary effects. The correlation could be positive or negative whereas the level of significance is based on whether the P-value is < 0.050. The interrelationship between the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K showed a weak positive correlation (P Values in parenthesis) of 0.331 (0.142), 0.598 (0.004), and 0.137(0.554), implying that although the pairs increase together, there is no significant relationship between them with P Values > 0.050 except for ²³²Th. This explains the fact that both ²²⁶Ra and ²³²Th are from different natural radionuclide decay series whereas ⁴⁰K is a non-decay series radionuclide. However, a strong positive correlation was observed between the 3 natural radionuclides and radium equivalent index of 0.684, 0.767 and 0.733, with respective P Values < 0.050, evidence of significant correlation between the pairs, implying that the radionuclides all contributed to the index, proportionately. In a similar vein, a positively strong correlation was observed between the 3 natural radionuclides and the corresponding radiological health risk variables, in the range of 0.881 to 1.000, except for ²²⁶Ra and ⁴⁰K which are low having uniform values of 0.327 and 0.496, which reflects their individual contributions to the parameters as shown also in Figure 4.

The Physico-chemical parameters, trace and heavy metals concentration levels of the natural spring mineral bottled drinking water from South Africa and the associated carcinogenic and non-carcinogenic risks have been reported [6].

The statistical analysis of the relationship between the natural radionuclides activity concentration levels ²²⁶Ra, ²³²Th and ⁴⁰K and the potential hydrogen (pH) and the total dissolved solids (TDS) (ppm) has been performed. The purpose is to evaluate the drinking water geochemical properties’ correlation with the natural radionuclides, and the result are presented in Figures 8a-c and 9a-c, respectively.

Figure 8. The relationship between pH and natural radionuclide activity concentration levels (BqL^-1) for (a) ²²⁶Ra, (b) ²³²Th and (c) ⁴⁰K for the bottled drinking water.

Figure 8. The relationship between pH and natural radionuclide activity concentration levels (BqL^-1) for (a) ²²⁶Ra, (b) ²³²Th and (c) ⁴⁰K for the bottled drinking water.

Figure 9. The relationship between total dissolved solids (TDS) (ppm) and radionuclide activity concentration (BqL^-1) for (a) ²²⁶Ra, (b) ²³²Th and (c) ⁴⁰K for the bottled drinking water.

Figure 9. The relationship between total dissolved solids (TDS) (ppm) and radionuclide activity concentration (BqL^-1) for (a) ²²⁶Ra, (b) ²³²Th and (c) ⁴⁰K for the bottled drinking water.

A positively very low correlation was observed between the natural radionuclides’ activity concentration levels of ²²⁶Ra and ²³²Th and the pH of the drinking water, which was insignificant, with the regression curve fit indicating that both variables increase together, while no correlation between ⁴⁰K and pH, results of this trends have been reported by [15,28]. The general trend observed concerning the relationship between the activity concentration of the radionuclides and the TDS, where the correlation coefficient of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K is found to be 0.2110, 0.0019 and 0.0136, respectively, showed that ²²⁶Ra seem to present more relations than others due to its ability to readily dissolve in water rich in minerals.

Table 5 presents the comparison of the activity concentrations of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K (BqL^-1) and the associated radiological health risk variable, the total annual effective ingestion dose (TAEID) (mSvy^-1) in this study and other nations of the world. The activity concentration levels in bottled drinking water of African countries including Morocco, Egypt, Nigeria ranged between 0.00194 to 11.100 BqL^-1, for ²²⁶Ra, 0.00246 to 4.600 BqL^-1 for ²³²Th and 0.237 to 28.500 BqL^-1 for ⁴⁰K, whereas the TAEID values varied from 0.991E-03 mSvy^-1 (for Morocco) to 3.300 mSvy^-1 (for Nigeria) [12,14]. The value for the radioactivity concentration for South Africa from this study is found to be within the range of those reported from other African countries, except for ⁴⁰K whose values are much greater. The activity concentration levels in bottled drinking water of countries other than Africa ranged between 0.105E-03 to 4.900 BqL^-1 for ²²⁶Ra, 0.016E-3 to 1.600 BqL^-1 for ²³²Th and 0.157 to 10.500 BqL^-1 for ⁴⁰K, whereas the TAEID values varied from 0.0016 mSvy^-1 (for Croatia) to 0.500 mSvy^-1 (for Spain) [9,16]. It can be observed that the activity concentration of the radionuclides in this work compares well with those reported by [21,28].

4. Discussion

Natural radionuclides such as ²²⁶Ra, ²³²Th and ⁴⁰K abounds everywhere in the natural environment at various concentration levels, including foodstuff and water, and are ingested by humans releasing radiation dose to vital organs providing risk of developing various cancers [13,32]. This study discusses the radioactivity concentration levels of the natural radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K (BqL^-1) and the associated radiological health risks due to the consumption of mineral spring bottled water. From the findings all the radionuclides being studied were detected at various concentration levels in the bottled drinking water. It can be observed that the activity concentration levels of ⁴⁰K is greater than that of the ²²⁶Ra and ²³²Th, showing a pattern of activity concentrations in the order of ⁴⁰K > ²³²Th > ²²⁶Ra, comparable to that reported by [12,14]. Potassium-40 an alpha emitter, has been described as being found everywhere, both in all water supplies and in all living organism because it is a very useful intracellular mineral [2,20]. The mean activity concentration levels exceed the recommended guideline limit for drinking water by the South African Water Quality Guideline but are way less than the worldwide average value [2,14,20]. The reason could be due to the diverse geology of South Africa, which is rich in NORMs such as uranium and thorium, anthropogenic factors as well as industrial processing of the bottled drinking water. However, the estimation of the TAEID presents on the overall, results that are way below the recommended dose limit of 0.1 mSvy^-1, for all the age groups assessed. The radiation dose from the drinking water is received by the age groups in the order of infant (≤1y) > teenagers (12-17 y) > children (1-12 y) > adult (>17). It shows that ²³²Th or its progeny ²²⁸Ra which is radiotoxic usually contributes more proportion to the annual effective ingestion dose in water and especially, highest in infants age group. It therefore requires remediation action especially that the infants are more sensitive and vulnerable to radiation exposure. This result is also way below the ambient and cosmic (above sea-level) levels of background radiation reported for the South Africa and global average, of 1.5 – 3.0 mSvy^-1 and 0.3 mSvy^-1, respectively. It implies that the mineral spring bottled drinking water from South Africa is safe for drinking according to global guideline limits for radionuclides contamination [20] and may not pose any significant health risk to the populace in the interim.

Radiological health risk models usually estimate the impact of natural radionuclide activity concentration on human organs due to ingestion of foodstuff and water. The parameters utilized in this study such as radium equivalent, absorbed dose rate, annual gonadal equivalent dose, hazard indices and representative alpha and gamma indices, presented values that are within acceptable limits, with the following implications. The Ra_eq findings will provide a base guideline which will be used to regulate safety standards regarding consumption of bottled water [19]. The absorbed dose gives an idea of the risk of exposure as the water is consumed, bearing also in mind the stochastic effect from accumulation of small doses over prolong periods of consumption. The gonads and bone surface which are very sensitive to radiation dose will be receiving very minimal AGED based on the findings in this study. Meanwhile, the internal and external exposure to radon which is short-lived and carcinogenic, as well as the representative alpha and gamma indices, from the bottled water is less than 1 hence offers negligible levels of radiation risk.

The estimated lifetime cancer risks and severe hereditary effects to the adults due to consumption of bottled water could be interpreted to mean that 12 out of 1000 may likely suffer some kind of cancer fatality while 6 out of 1 million may suffer some form of hereditary effect. However, the findings presents values to be within the recommended guidance limit set by the United State Environmental Agency (USEPA), of 1.0 x 10^-6 to 1.0 x 10^-4, which is interpreted to mean 1 person out of 1 million to 10,000 persons may suffer from some form of cancer fatality [25,26].

The statistical analysis of the data in this study indicates that, for the 21 samples, the values of the standard deviation are less than the values of corresponding mean, of the 3 radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K as well as the radiological health risk parameters. This implies that the distribution of the data has a very high degree of uniformity, and it is comparable with result reported by [24] but differ from [26]. The frequency distribution of the natural radionuclides shows that ²²⁶Ra and ²³²Th together with TAEID for the various age groups are skewed to the right or are slightly asymmetric (positively skewed) having distribution tail pointing to the right with value of skewness greater than 0, while ⁴⁰K and Ra_eq are skewed to the left or slightly asymmetric (negatively skewed) also having skewness value less than 0, as shown in Table 3, column 9. Skewness of data indicates its nature or extent of symmetry with perfect symmetry at zero, hence, frequency distribution of the radionuclides and the corresponding radiological health risk parameters in this study are only slightly asymmetric. Additionally, the kurtosis of the radionuclide concentrations presents negative value for ²²⁶Ra while ²³²Th, ⁴⁰K, Ra_eq and TAEID for various age group showed positive value (Table 3 column 10). An indication that the frequency distribution curve of the radionuclides is less peaked than the normal curve known as platykurtic, for ²²⁶Ra while more peaked than normal curve called leptokurtic, for ²³²Th, ⁴⁰K, Ra_eq and TAEID for various age group. The behaviour of the statistical frequency distribution in this study could be due to the unequal distributions of the natural radionuclides in the natural spring mineral drinking water, especially from the source aquifers bedrock. The Pearson’s correlation coefficient used in this study presents positively very low correlation between the natural radionuclides’ activity concentration levels of ²²⁶Ra and ²³²Th and the pH of the drinking water, which was insignificant with the regression curve fit indicating that both variables increase together, while no correlation between ⁴⁰K and pH, findings with this kind of trend have been reported by [15,28]. The effect of pH on the variable behaviour of ²³⁸U and ²³²Th decay series as well as acid leaching could be responsible for the findings in this study. With increase in pH, uranium and thorium as well as other minerals are usually leached in the surrounding rocks likewise the sediment [33]. The general trend observed concerning the relationship between the activity concentration of the radionuclides and the TDS, where the correlation coefficient of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K is found to be 0.2110, 0.0019 and 0.0136, respectively, showed that ²²⁶Ra seem to present more relations than others due to its ability to readily dissolve in water rich in minerals.

The findings in this study as compared to those reported from other countries suggests that, the radioactivity concentration levels of the radionuclides ²²⁶Ra, ²³²Th and ⁴⁰K are comparable to similar studies from the African countries [14,28] with values lower than [12]. However, this study’s values are higher than those reported in non-African countries [9,16], yet all values are within the permissible limit set by [2,20], implying that the bottled water is radiological safe for consumption.

5. Conclusions

The assessment of radioactivity concentration levels in natural spring mineral bottled commercial drinking water in South Africa using HPGe gamma spectrometry, has provided significant insights into radiological quality of these water sources. The study revealed varying levels of naturally occurring radionuclides across different brands, with mean values of 1.766± 0.399 BqL^-1 for ²²⁶Ra, 3.688 ± 1.371 BqL^-1 for ²³²Th and 220.229 ± 22.297 BqL^-1 for ⁴⁰K. The radiological health risks associated with these concentrations were found to be below the recommended safety guideline limits set by the United Nations Scientific Committee on the Effects of Atomic Radiation, indicating that the consumption of natural spring mineral bottled water in South Africa is radiologically safe.

However, our findings also highlight the potential health risks associated with the prolonged consumption of bottled drinking water containing elevated radionuclide levels. The radiation dose from the drinking water is received by different age groups in the following order: infant (≤ 1 year) > teenagers (12-17 years) > children (1-12 years) > adult (> 17 years). Among the radionuclides, ²³²Th contributes the highest radiation dose, accounting for 83.44% in the infant group, while ²²⁶Ra contributes the least, with 6.00% in the children group. ⁴⁰K falls in-between, with contributions ranging from 7.01% in teenagers to 31.00% in adults.

Based on our findings, approximately 12 out of 1000 individuals may suffer cancer fatality, and 6 out of 1 million individuals may experience hereditary effects due to the ingestion of the bottled drinking water over their lifetime. While the immediate risk appears exceptionally low and provides overall safe water quality, the cumulative effects over time could pose significant health hazards. This emphasizes the need for public awareness and education on the importance of choosing safe drinking water.

The findings of this study are comparable to those reported in the literature, with levels lower than those found in some African countries but higher than non-African countries. This highlights the importance of geological location variations in radionuclide concentrations.

The results of this study call for a collaborative effort between regulatory agencies, water bottling companies, health organisations to strengthen the enforcement of guidelines that ensure the safety of bottled drinking water. Regular testing and transparent reporting of radionuclide levels should be mandated to prevent possible radiological health risks.

In conclusion, this research contributes to better understanding of the radiological quality of bottled drinking water in South Africa and serves as a foundation for future studies and policy development aimed at ensuring safe drinking water for all, in alignment with the global sustainable development goals. By addressing the identified radiological risks, we can take proactive steps towards safeguarding public health and enhancing the overall quality of bottled drinking water available to consumers.