Assessment of natural radioactivity and radiation risks in soil near the Manyoni uranium deposit, Tanzania

Abstract

This study reports on natural radioactivity levels and radiation risk estimates from soil samples collected in five villages near the Manyoni Uranium Deposit in central Tanzania. The mean activity concentrations of ⁴⁰K, ²³²Th, and ²³⁸U were 207 Bq kg⁻¹, 78 Bq kg⁻¹, and 302 Bq kg⁻¹, respectively. While ⁴⁰K levels were below the world average, the levels of ²³²Th and ²³⁸U were more than 1.7 and 9.2 times higher than world average values. The average absorbed dose rate for the area was 196 nSv h⁻¹. Annual effective dose estimate values for outdoor and indoor exposure were 0.24 mSv and 0.96 mSv, respectively, exceeding global averages by 3.4 and 2.3 times. Based on the average external hazard index (H_ex = 1.19) and excess lifetime cancer risk (ELCR = 0.46%), there is a potential risk of radiological effects from radiation exposure, with Mwanzi Village identified as a high background radiation area.

Introduction

Natural radionuclides are major sources of radiation exposure and are responsible for ~70% of the dose received by the population from all radiation sources [1]. Thorium and Uranium radionuclides and their progenies are present in plant tissues, animal tissues, soil, air, and water and contribute significantly to the ingestion dose to the population [2]. The main contributors to background radiation in the soil are gamma rays emitted from radionuclides such as the ⁴⁰K, ²³²Th, and ²³⁸U series [3]. High uranium and radium concentrations in soils have been reported to cause elevated radiation levels with absorbed dose rates that may exceed the average world background level of 60 nGy h⁻¹ [4]. Typically, higher ²²⁶Ra concentrations are found in uranium mine sites, which may lead to increased radon concentrations in nearby buildings and settlements as well as contribute to the population inhalation dose [5, 6]. Several studies have reported high radioactivity levels in the regions near uranium deposits [7–10]. A study by Gan et al. (2018)⁷, in the area around the Xiangshan uranium deposits in Jiangxi Province, China, reported relatively enhanced activity concentrations of ²³⁸U (119 ± 97 Bq kg⁻¹ and ²³²Th (93 ± 28 Bq kg⁻¹). These values are higher than the world average activity concentration of ²³⁸U in soil (33 Bq kg⁻¹) [1]. Similar observations were made in Mounana, Gabon [11] and Homa Mountain, Kenya [12].

The Manyoni uranium deposit, located in central Tanzania, is estimated to contain 12,000 metric tons of uranium [13]. It is characterized by an extensive closed drainage system that developed over uranium-rich, weathered granites. This system captures the dissolved uranium leached from the underlying rocks and transports it to precipitation traps, referred to as mbuga or playa lakes. The uranium targets in these areas are primarily calcrete-hosted mineralization near the surface, with sandstone-hosted deposits located within buried fluvial channels [13]. Classified as surficial uranium deposits, the uranium in these formations is sourced from uranium-enriched granites. These deposits are considered young mineralizations, typically found in surface sediments or soils, and are often associated with secondary precipitates such as gypsum, dolomite, and calcite [14]. The geological structure underlying the deposit consists of mbuga clay, granites, and Kilimatinde cement, with uranium mineralization linked to nearby uranium-enriched granites [13].

The area is surrounded by human settlement areas, and the land is used mainly for growing food crops and grazing. Many shelters in the area are poorly ventilated, often featuring only a small door and window, or no windows at all, and are constructed from mud walls using local soil. Traditional clay pots are also crafted from the same soils. In addition, expectant women commonly consume anthill soils. Consequently, there is potentially elevated radiation exposure in local communities. The aim of this study was to assess the activity concentrations of ⁴⁰K, ²³²Th, and ²³⁸U in soil sampled from five villages located near the deposit and to estimate the associated radiological risks. A spatial approach was used to present the radiometric results from the five villages to facilitate easier interpretation and visual comparison. The findings may help inform policies designed to protect members of the general public from radiation risks.

Materials and methods

Description of the study area

Manyoni is a district located in the Singida region in central Tanzania and is located between 5° 45′ 0" S and 34° 50' 0" E, with an area of 28 620 km² and a population of 296 763.

Sample collection and preparation

A total of 31 soil samples were collected from villages in and around the Manyoni Uranium Deposit area (see Fig. 1) to assess natural radioactivity levels. Sampling was conducted at a depth of 0–20 cm from the surface to capture surface-level radionuclide distribution. To ensure representativeness, the composite sampling technique was employed. For each farm, four subsamples were collected—two from the edges and two from the center. These subsamples were thoroughly mixed to form a single composite sample representing each sampling site [15, 16]. The collected soil samples were then stored in labeled, sealed polythene bags, and transported to the laboratory. The samples were then dried in an oven at a constant temperature of 105 °C for 24 h [17]. The samples were ground, thoroughly mixed, and sieved (~2 mm in size). Finally, the samples were packed in airtight stainless-steel canisters, which were tightly sealed to prevent the escape of ²²²Rn gas [17]. The samples were stored for 4 weeks to allow secular equilibrium between ²²⁶Ra and its daughters in the ²³⁸U series.

Instrumentation

Gamma-ray spectrometry was used to determine the activity concentrations in the soil samples. A high-purity germanium (HPGe) detector with a relative efficiency of 49% and a resolution of 1.65 keV at an energy of 1.33Mev of the ⁶⁰Co gamma-ray source was used for the measurements. The analyses were performed using a coaxial detector (model GEM40–83-SMP, serial No. 57P51572A; Ortec, USA) [18]. The detector was cooled in liquid nitrogen at −196 °C and coupled with a high-performance multichannel analyzer (ORTEC DSPEC LF1, Ortec, USA) with 8192 channels. Data acquisition was performed using Gamma Vision software (version 8, Ortec, USA) following detector efficiency and energy calibration procedures.

The activity concentration of radionuclides was calculated using Equation 1. Detector efficiency was determined using selected calibration sources: ⁶⁰Co (1173.2 keV and 1332.5 keV), ¹³⁷Cs (661.7 keV), and ²⁴¹Am (59.5 keV). The spectra for these sources were acquired by accumulating them for 86 400 s. The count rates were adjusted appropriately to minimize the dead-time. Energy calibration was performed by identifying prominent gamma peaks with the corresponding channel numbers.

Where,

|$A(E)$| is the activity concentration of the radionuclide (Bq kg⁻¹).

|${C}_{net}$| is the net count rate under the photopeak at energy E, obtained by subtracting the background counts from the total counts.

|$\epsilon (E)$| is the Absolute efficiency of the detector at energy E.

|${I}_{\gamma }$| is the gamma-ray emission probability (intensity) for the specific energy transition.

|$t$| is the counting time (s).

|$m$| is the mass of the sample (kg).

The activity concentration of ²²⁶Ra was determined from the average concentrations of 𝛾-ray emission lines of ²¹⁴Pb (295.2, 351.9 keV) and ²¹⁴Bi (609.3, 1120.29 keV). ²³²Th was determined from ²²⁸Ac gamma lines (338.4, 911.3, and 969.16 keV) and ²¹²Pb (238 keV). ⁴⁰K was determined directly from its single gamma line energy of 1460.7 keV. The activity of ²³⁸U was determined using the gamma lines of ²¹⁴Pb (295.2 keV and 351.9 keV) and ²¹⁴Bi (1378 keV and 1238 keV). To ensure the reliability of the measurement data, quality assurance was conducted using International Atomic Energy Agency (IAEA, Austria) standard reference materials with known mass activities: IAEA-RGU-1, IAEA-RGTh-1, and IAEA-RGK-1. These reference materials were also used to check the calibration and correct the density differences, as their densities were comparable to those of typical soils. In addition, the IAEA Soil-375 standard reference material was analyzed under the same experimental conditions as the soil samples to evaluate both the accuracy and precision of the results. The experimental activity concentrations obtained for ²³⁸U, ²³²Th, and ⁴⁰K showed good agreement with the certified reference values after decay correction, with discrepancies within a 10% margin of accuracy.

Absorbed dose rate in air

The absorbed dose rate in air (DR) from radioactivity concentrations of ²³⁸U, ²³²Th, and ⁴⁰K in air 1 m above the ground surface was determined using Equation 2 [1].

Where A_U, A_Th, and A_K represent the activity concentrations of ²³⁸U, ²³²Th, and ⁴⁰K, respectively. The conversion factors for these radionuclides are 0.462, 0.604, and 0.042 nGy h⁻¹ per Bq kg⁻¹, for ²³⁸U, ²³²Th, and ⁴⁰K, respectively.

Annual effective dose

The annual effective dose estimate (AEDE) for external terrestrial radiation in the outdoor and indoor environments was calculated using Equations (3) and (4), respectively [19]. Additionally, the total AEDE (D_tot) was calculated using Equation (5). This evaluation takes into account that the fact that people spend 80% and 20% of their time indoors and outdoors, respectively. The world average values for D_out, D_in, and D_tot are 0.07 mSv, 0.41 mSv and 0.48 mSv, respectively [1].

where D_out, D_in, and D_tot are the outdoor, indoor, and total annual effective dose equivalent, respectively.

Excess lifetime cancer risk

Excess lifetime cancer risk (ELCR) indicates the probability of developing cancer for a given population at a given level of radiation exposure. It was calculated using Equation (6) [20]:

Where AEDE is the annual effective estimate, LS is the average life expectancy (70 years), and PC is the probability coefficient for radiation-induced cancer risk, set at 0.055 Sv⁻¹ for the general population [21].

External hazard index for gamma-radiation

The external hazard index (H_ex) for γ-ray radiation from naturally occurring radioactive materials must be less than one to ensure the radiation hazard remains negligible. This corresponds to an annual dose limit of 1.0 mSv for the general population and is derived from the upper limit of radium equivalent activity (370 Bq kg⁻¹) [20]. H_ex is calculated as shown in Equation (7):

Spatial distribution of radionuclides using K-means clustering

To explore the spatial distribution of the activity concentrations of ⁴⁰K, ²³²Th, and ²³⁸U across the five villages, we employed the K-means clustering method to group the concentrations into three distinct clusters. This method was applied to each radionuclide to identify natural groupings based on their concentration levels in the soil samples. The optimal number of clusters (three) was determined through visual inspection of the data distribution. Each cluster represents a range of concentrations: low, moderate, and high. The clusters were then plotted on scatter plots, with the samples from different villages represented by colored points. The colors blue, green, and red correspond to low, moderate, and high concentration clusters, respectively, for each radionuclide.

Results and discussion

Radioactivity levels in soil

The results show that there is significant spatial variability in the activity concentrations of ⁴⁰K, ²³²Th and ²³⁸U across the five villages (Fig. 2 (a)–(c)). More than 90% of the samples exhibited activity concentrations below the world average level of 420 Bq kg⁻¹, while the activity concentrations of ²³⁸U and ²³²Th in all soil samples were 2–52 times and 1–4 times higher than the world average values, respectively. These findings strongly suggest the presence of uranium-bearing minerals in the area along with moderately elevated levels of thorium-bearing minerals. The activity concentrations of ²³⁸U (Fig. 2c) demonstrated the widest range of variability among the three radionuclides. Mwanzi exhibits the highest activity concentrations of ²³⁸U, with multiple samples in the range: 774–1736 Bq kg⁻¹ suggesting that this village may have significant uranium enrichment, possibly due to the underlying geological formations rich in uranium-bearing minerals. This uranium enrichment may also result from uranium mobilization and deposition processes, potentially influenced by local hydrological and soil conditions. In contrast, the levels were below 193 Bq kg⁻¹ in Majengo, Mitoo, and Mkwese.

The mean radioactivity concentrations of the soil in the villages are summarized in Table 1. ⁴⁰K exhibited the highest range of concentrations across the villages, with values ranging from 170 Bq kg⁻¹ in Mitoo to 393 Bq·kg⁻¹ in Majengo, suggesting substantial variability in its distribution. The mean activity for ⁴⁰K across all villages is 207 Bq kg⁻¹, but the large range and potential outliers in Majengo and Mwanzi indicate a more heterogeneous spread compared to the other radionuclides. ²³⁸U, like ⁴⁰K, shows considerable variability. The activity concentrations range from 99 Bq kg⁻¹ in Muhalala to 245 Bq kg⁻¹ in Mkwese, with a mean of 302 Bq kg⁻¹. The wide range of ²³⁸U concentrations suggests localized sources of uranium enrichment, possibly due to variations in soil mineralogy or environmental factors, such as water flow, which may concentrate uranium in certain areas. In contrast, ²³²Th showed the highest activity concentration across the villages. The values for ²³²Th range from 70 Bq kg⁻¹ in Mkwese to 84 Bq kg⁻¹ in Muhalala, with a mean of 78 Bq kg⁻¹. The relatively small range and low variance suggest that thorium is more evenly distributed in the soils, indicating a more uniform geological origin that may be influenced by local mineral content and natural background radiation levels.

Radiation risk assessment and radiological parameters

Radiation risk was assessed by quantifying the DR, AEDE, ECLR and H_ex (Table 2). The DR values varied significantly, ranging from 102 nSv h⁻¹ in Muhalala to 279 nSv h⁻¹ in Mwanzi, with an average dose rate of 196 nSv h⁻¹ for Manyoni. Notably, all villages exhibit dose rates significantly higher than the world average of 60 nSv h⁻¹, with Mwanzi showing an exceptionally high value, over 4.5 times the world average. In addition, the wide range of DR in Mwanzi (91–846 nSv h⁻¹) reflects the heterogeneous distribution of radionuclides within the area (see Fig. 2). The spatial distribution of DR across the surveyed villages is shown in Fig. 3. From the map, it may be observed that Mkwese and Mwanzi exhibited higher dose-rates compared to the other villages. Mwanzi, in particular, stands out with several high DR points (>240 nSv h⁻¹; more than 4 times the world average), which suggests that the area can be classified as a High Background Radiation Area (HBRA).

The AEDE values for outdoor exposure (D_out) range from 0.12 mSv in Muhalala to 0.34 mSv in Mwanzi, with an average of 0.24 mSv for Manyoni. For indoor exposure (D_in), the values are higher, ranging from 0.5 mSv in Muhalala to 1.37 mSv in Mwanzi, averaging 0.96 mSv for Manyoni. These values are still below the recommended public exposure limit of 1 mSv y⁻¹ set by ICRP, except in Mwanzi, where the indoor dose approaches or exceeds this threshold. The total AEDE (D_tot) ranges from 0.62 mSv in Muhalala to 1.71 mSv in Mwanzi, averaging 1.2 mSv for Manyoni, indicating that cumulative exposure over time could pose a health risk, especially in areas like Mwanzi.

ELCR for outdoor exposure ranges from 0.24% in Muhalala to 0.66% in Mwanzi, while for indoor exposure, it ranges from 1.7% to 4.59%, respectively. The elevated ELCR values in Mwanzi suggest a higher probability of cancer development over a lifetime due to long-term radiation exposure. The average ELCR for Manyoni is 0.46% for outdoor exposure and 2.66% for indoor exposure, indicating a notable risk of radiological health effects, particularly for residents who spend a significant amount of time indoors. The external hazard index (H_ex), which estimates the potential for radiation exposure from external sources, ranges from 0.62 in Muhalala to 1.65 in Mwanzi. A value exceeding 1 indicates a potential radiation hazard, which is the case for Mwanzi. The average H_ex for Manyoni is 1.16, highlighting a general radiological concern, particularly in areas with high DR like Mwanzi.

Comparison of natural radioactivity levels in Manyoni with similar studies in Tanzania and other countries

The mean activity concentrations of ²³⁸U, ²³²Th, and ⁴⁰K in the soil samples in this study were compared to those reported in similar studies from selected areas in Tanzania and other countries (Table 3). The present study reports 207 Bq·kg⁻¹ for ⁴⁰K, which is lower than other areas in Tanzania, such as Mkuju (564 Bq·kg⁻¹) and Bahi (496–808 Bq·kg⁻¹). The West of Manyoni has a significantly lower range for ⁴⁰K (47–205 Bq·kg⁻¹) compared to the present study. Overall, the ⁴⁰K levels in the present study were much lower than those in other areas of Tanzania. The activity concentration of ²³²Th in the present study was 78 Bq·kg⁻¹, which is comparable to that in Bahi (35–105 Bq·kg⁻¹). West of Manyoni, the range (12–49 Bq·kg⁻¹) was lower, while the level reported in Mkuju was also lower (36 Bq·kg⁻¹). Therefore, the levels in the present study are relatively higher than those in most other areas in Tanzania, except for Bahi. The present study found 302 Bq·kg⁻¹ for ²³⁸U, which is much higher than the concentrations reported in Mkuju (52 Bq·kg⁻¹) and Bahi (11–34 Bq·kg⁻¹). On the other hand, the levels reported in the West of Manyoni indicate high uranium concentrations (23–835 Bq·kg⁻¹), which are comparable to the values observed in the present study, confirming the presence of uranium-rich soils in Manyoni.

Compared to other countries and the world average, the activity concentrations of ⁴⁰K concentration in Manyoni (207 Bq kg⁻¹) is lower than the world average of 420 Bq kg⁻¹. Comparatively, areas such as Kapchorwa, Uganda (1339 Bq kg⁻¹) and Homa Mountain, Kenya (916 Bq kg⁻¹) report significantly higher levels. On the other hand, some regions, such as the phosphate-rock storage facility in South Africa (147 Bq kg⁻¹) and coastal areas of Ado-Odo, Nigeria (134 Bq kg⁻¹) have lower levels than the present study in Manyoni. The activity concentration of ²³²Th in Manyoni (78 Bq kg⁻¹) is higher than the world average (45 Bq kg⁻¹) but the levels are comparable to those in regions such as Mounana, Gabon (63 Bq kg⁻¹), and Metekel, Ethiopia (70 Bq kg⁻¹). However, there are much higher levels in areas such as the proposed uranium mining areas in India (231 Bq kg⁻¹) and Rize, Turkey (36–914 Bq kg⁻¹). The activity concentration of ²³⁸U in Manyoni (302 Bq kg⁻¹) is significantly higher than the world average (33 Bq kg⁻¹), which places Manyoni in the category of uranium-rich areas. These levels are comparable to those in areas such as the Kitongo Uranium Deposit, Cameroon (99 Bq kg⁻¹), Mounana, Gabon (2811 Bq kg⁻¹) and Xiangshan uranium deposits, Jiangxi in Eastern China (119 Bq kg⁻¹), where uranium deposits are present.

Conclusion

The activity concentrations of ⁴⁰K, ²³²Th, and ²³⁸U in the soil samples were quantified, and the radiation risks to villages near the Manyoni Uranium Deposit area in Central Tanzania were assessed. The measured activity concentration levels and radiation risk parameters were compared with world averages, recommended safety limits, and findings from similar studies. The results revealed significant spatial variability in radioactivity levels, with particularly high activity concentrations of ²³⁸U and ²³²Th compared to world averages. The potential for radiological health effects is notable, especially in Mwanzi, which has been classified as a HBRA. It can be concluded that there is a possibility of the radionuclide contamination being transferred to the other districts through food stuffs, and building materials. This study provides a crucial baseline for policymakers to develop strategies for the protection and management of radiation. A detailed assessment of radiological risks in the area encompassing the uranium deposit is recommended, especially before mining activities commence. Authorities should consider relocating the population to safer areas to protect them from radiation exposure. In addition, educating residents about radiation risks is critical, particularly for expectant women, who should be made aware of the dangers of consuming anthill soil that could increase radiation exposure to unborn children. Further studies are recommended to assess radioactivity levels and radiation risks in pastures and water, particularly in Mwanzi village, to ensure comprehensive understanding and mitigation of the risks posed.

Acknowledgements

The authors sincerely appreciate Mr. Simion Bartilol (Institute of Nuclear Science & Technology, University of Nairobi) for his valuable technical assistance and the support provided in this work.

Conflict of interest

The authors declare no conflicts of interest.

Funding

This work was funded by the Ministry of Education, Science and Technology, Tanzania.

References