The study was conducted in Mbeere South Constituency, Embu County in Kenya located between latitudes 0˚50.40'S and 0˚37.80'S, and longitudes 37˚35.40'E and 37˚48.0'E occupying an area of approximately 1312 km². The constituency is comprised of five wards namely, from Mwea, Makima, Mbeti South, Mavuria, and Kiambeere wards. According to the 2019 Kenya Population and Housing Census, Mbeere South had a population of 163,476 residents [15]. The constituency has a predominantly rural economy supported by agriculture, livestock keeping, trade, and natural resource exploitation including artisanal mineral extraction. The climate is largely semi-arid. Mbeere South, like much of Embu County, has a diverse geological composition that comprises the Precambrian Mozambique belt, which is part of the basement rocks dominated by gneisses, schists, granites, and pegmatites, together with localized sedimentary deposits [16][17]. These lithological characteristics make the area suitable for investigating the distribution of naturally occurring radionuclides and evaluating associated radiological hazards. The map of the sampling sites is shown in Figure 1Figure 1 [18].

Figure 1. Location of Mbeere South and sampling sites in Embu County, Kenya.

This study employed an experimental and observational research design to determine the activity concentrations of naturally occurring radionuclides in rock samples collected from Mbeere South, Kenya. Field investigations involved in situ radiation measurements using a portable radiation survey meter, while laboratory analyses were performed using high-resolution HPGe gamma-ray spectrometry.

In-situ ambient dose rate screening was conducted at potential sites using a calibrated portable BR-9B XR1 Geiger-Müller (GM) detector (background ∼0.10 µSv∙h⁻¹). Screening was used to guide selection and ensure inclusion of a range of lithologies and exposure conditions. Sampling was purposive, not random, based on accessibility, active quarrying status, and lithological variation observed after GM screening. Thirty rock samples were purposively selected with rock samples preferentially collected from sites showing relatively elevated or representative radiation levels, as well as from different geological settings, including quarries, seasonal riverbeds, hillsides, and mineralized zones in Mbeere South. The samples were collected from the five wards as follows: four samples from Mwea, one sample from Makima, thirteen samples from Mbeti South, three samples from Mavuria, and nine samples from Kiambeere. Results are therefore limited to the sampled sites and are not extrapolated to the entire Mbeere South constituency. The geographical coordinates of each quarry were recorded using a handheld GPS.

Approximately 0.5 - 1.0 kg of fresh representative rock was collected from each selected site. The samples were immediately placed in clean, well-labelled polyethylene bags. Each sample was assigned a unique identification code (001-030) corresponding to its sampling location and carefully packaged to prevent cross-contamination and loss of material during transportation to Chuka University. At Chuka University Science Laboratory, the samples were pulverized into a fine homogeneous powder using hardened steel hammers, with cleaning between samples to avoid cross-contamination.

The pulverized powders were transported to the Institute of Nuclear Science and Technology [INST], University of Nairobi, oven-dried at 80˚C for 8 h to constant mass, sieved through 1 mm mesh to obtain uniform particle size, homogenized, weighed, and packed into 500 g (500 ml) Marinelli beakers to ensure identical counting geometry. Beakers were hermetically sealed and stored for 30 days to allow ²²⁶Ra and its short-lived progeny (⁻1214Pb, ²¹⁴Bi) and ²³²Th series to attain secular equilibrium before measurement on HPGe [19][20].

The activity concentrations of ²²⁶Ra (as a proxy for ²³⁸U under secular equilibrium), ²³²Th and ⁴⁰K in the 30 purposively selected rock samples from Mbeere South, screened with a handheld GM counter, were determined by high-resolution gamma-ray spectrometry at the Institute of Nuclear Science and Technology (INST), University of Nairobi. The system is equipped with a p-type coaxial High Purity Germanium detector (Baltic Scientific Instruments, Latvia) model GCW-30180 housed within a cylindrical multilayer graded shield and supported by a low-background lead shield to minimize external background radiation. The detector was operated at 3500 V bias, with energy resolution (FWHM) of 1.91 keV at 1332.5 keV of ⁶⁰Co and relative efficiency 27.8% [21].

To attain secular equilibrium between ²²⁶Ra and its short-lived progeny, and between ²³²Th and its daughters, sealed 500 ml Marinelli beakers containing 500 g of pulverized rock were stored for 30 days prior to counting. Each sample was counted for 36,000 s.

Energy and efficiency calibrations were carried out using certified mixed radionuclide standard BB-3591 covering 60 - 3500 keV containing ¹³⁷Cs, ²⁴¹Am, ⁶⁰Co, ⁸⁸Y, ⁵⁷Co, ¹¹³Sn and ¹³⁹Ce, traceable to international standards. Spectral acquisition was via SpectraLine-GP software. Background correction was performed by counting an empty Marinelli beaker filled with deionized water under identical geometry for 12 h and subtracting the background spectrum.

²³²Th was evaluated from weighted mean of ²¹²Pb (238.6 keV) and ²²⁸Ac (911.1 keV), ²²⁶Ra (proxy for ²³⁸U) from ²¹⁴Pb (351.9 keV) and ²¹⁴Bi (609.3 keV), and ⁴⁰K directly from 1460.8 keV [22].

Quality assurance was confirmed by concurrent analysis of Certified Reference Material IAEA-447 Natural and Anthropogenic Radionuclides in Moss-Soil with certified values: 25.1 ± 2.0 Bq∙kg⁻¹ for ²²⁶Ra, 22.2 ± 0.8 Bq∙kg⁻¹²³⁸U, 21.8 ± 0.8 Bq∙kg⁻¹ for ²³⁴U, with information 550 ± 18 Bq∙kg⁻¹ for values ⁴⁰K, 37.3 ± 2.0 Bq∙kg⁻¹ for ²³²Th, 37.3 ± 1.5 Bq∙kg⁻¹ for ²¹²Pb, 37.3 ± 2.0 Bq∙kg⁻¹ for ²²⁸Ac. Measured vs certified recoveries were 96.2% for ²³⁸U (as ²²⁶Ra), 98.5% for ²³²Th and 102.3% for ⁴⁰K within acceptance 90% - 110%.

The Minimum Detectable Activity (MDA) was 1.18 Bq∙kg⁻¹ for ²³⁸U (as ²²⁶Ra), 0.85 Bq∙kg⁻¹ for ²³²Th, 12.4 Bq∙kg⁻¹ for ⁴⁰K. Uncertainties combined counting statistics and calibration, reported as ±1σ. Assessment is external gamma screening only; radon exhalation was not assessed.

The radium equivalent activity (Ra_eq) was calculated to assess the combined radiological hazard arising from ²²⁶Ra, ²³²Th and ⁴⁰K. The activity concentration of ²²⁶Ra was assumed to be equal to that of ²³⁸U under secular equilibrium. It was determined using Equation (2) [20][21].

where C Ra , C Th , and C K are the activity concentrations of ²²⁶Ra, ²³²Th and ⁴⁰K in Bq∙kg⁻¹ respectively. The recommended safety limit of Ra_eq is 370 Bq∙kg⁻¹ [7].

The absorbed dose rate (D) in air was calculated from the activity concentrations of ²²⁶Ra, ²³²Th, and ⁴⁰K to assess external gamma radiation exposure from the rock samples, assuming secular equilibrium between ²³⁸U and ²²⁶Ra following 30 days of sample storage. It was determined using the Equation (3) [23].

where C Ra , C Th and C K are the activity concentrations of ²²⁶Ra, ²³²Th, and ⁴⁰K in Bq∙kg⁻¹ respectively. The worldwide average outdoor absorbed dose rate is approximately 59 nGy∙h⁻¹, and values significantly higher than this may indicate elevated natural background radiation levels [7].

The outdoor annual effective dose (AED_out) was estimated to evaluate the potential radiological health risk to members of the public from external gamma radiation emitted by natural radionuclides in the samples. The absorbed dose rate in air was converted to outdoor annual effective dose using the Equation (4) [24].

where AED_out is the outdoor annual effective dose, D is the absorbed dose rate calculated in Equation 3, According to UNSCEAR 2000, the DCF (dose conversion factor) and O f (outdoor occupancy factor) values are 0.7 Sv/Gy and 0.2, respectively, T is the time factor (8760 hours per year). An outdoor occupancy factor, O f = 0.2 was adopted in accordance with UNSCEAR recommendations, assuming that humans spend 20% of their time outdoors. The annual effective dose equivalent should not exceed 1 mSv∙y⁻¹ for safety purposes [7]. Values exceeding this level may indicate increased radiological exposure to the population and thus require further assessment. All radiological hazard indices were evaluated against internationally recommended reference limits to assess the suitability of the investigated rocks for use as building materials.

The activity concentrations of the primordial radionuclides ²³⁸U, ²³²Th and ⁴⁰K in 30 rock samples from selected sites in Mbeere South are presented in Table 1 while the variations in the concentrations of the naturally occurring radionuclides

Table 1. Activity concentrations of ²³⁸U, ²³²Th and ⁴⁰K in rock samples.

are represented graphically as shown in Figure 2Figure 2. The values ranged from 3.23 ± 1.13 to 54.59 ± 3.87 Bq∙kg⁻¹ for ²³⁸U, 4.82 ± 1.58 to 122.14 ± 5.61 Bq∙kg⁻¹ for ²³²Th and 76.63 ± 13.43 to 1223.44 ± 50.12 Bq∙kg⁻¹ for ⁴⁰K. The mean activity concentrations were 19.02 ± 14.00 Bq∙kg⁻¹ for ²³⁸U, 38.66 ± 34.43 Bq∙kg⁻¹ for ²³²Th and 634.88 ± 302.94 Bq∙kg⁻¹ for ⁴⁰K.

Figure 2. Activity concentrations of U-238, Th-232 and K-40 in the investigated samples.

The mean of ²³⁸U was 45.7% below the world average of 35 Bq∙kg⁻¹ for rocks [UNSCEAR 2000]. Only 3 samples (005, 010, and 019) exceeded the world average, indicating generally low uranium content in Mbeere South basement. The relatively high value at 010 (54.59 Bq∙kg⁻¹) may be consistent with localized enrichment in felsic intrusives, as uranium tends to concentrate in late-stage magmatic differentiates. These are presented as hypotheses based on regional geology.

The mean of ²³²Th was 28.9% above the world average of 30 Bq∙kg⁻¹, indicating a slight thorium enrichment in the study area. The large standard deviation (SD) (34.43) shows high heterogeneity. Elevated values > 70 Bq∙kg⁻¹ at 010 (100.30 Bq∙kg⁻¹), 014 (122.14 Bq∙kg⁻¹), 015 (107.58 Bq∙kg⁻¹), 021 (74.80 Bq∙kg⁻¹), 026 (76.37 Bq∙kg⁻¹) and 029 (85.32 Bq∙kg⁻¹) are hypothetically consistent with the reported occurrence of accessory minerals such as monazite and zircon in gneissic rocks of the Mozambique Belt. Thorium is immobile and concentrates in resistate minerals, which could explain these hotspots. Samples with <10 Bq∙kg⁻¹ (004, 008, 009, 016, 020, 022, 025 and 030) may reflect quartzitic lithologies depleted in heavy minerals.

The mean of ⁴⁰K was 58.7% above the world average of 400 Bq∙kg⁻¹. Potassium is the dominant contributor to total activity in Mbeere South. 70% of samples exceeded 400 Bq∙kg⁻¹. The highest values at 018 (1223.44 Bq∙kg⁻¹), 026 (1138.68 Bq∙kg⁻¹), 029 (1058.88 Bq∙kg⁻¹), 027 (1057.67 Bq∙kg⁻¹), 015 (1023.07 Bq∙kg⁻¹) and 021 (986.31 Bq∙kg⁻¹) are consistent with K-feldspar-rich lithologies such as granitic gneisses and pegmatites, which are commonly mapped in Mbeere South. The very low value at 006 (76.63 Bq∙kg⁻¹) suggests a mafic enclave or quartz vein. Again, these lithological links are hypothetical and based on the geology of the Embu area [17].

The order of mean activity is ⁴⁰K > ²³²Th > ²³⁸U, which is typical for crustal rocks where potassium is a major rock-forming element. The low ²³⁸U/²³²Th ratio (mean 0.49) suggests thorium preferential enrichment, a common feature in weathered basement terrains. The mean activity concentrations obtained in this study were compared with mean activity concentrations from other regions of the world as shown in Table 2.

Table 2. Comparison of mean activity concentrations of ²³⁸U, ²³²Th and ⁴⁰K in rocks from Mbeere South with selected studies from other regions.

The mean activity concentration of ²³⁸U was 79.9% lower than Kerio Valley. Kerio Valley is underlain by alkaline syenites and granites with zircon and apatite that concentrate uranium. Mbeere South, though also in the Mozambique Belt, is dominated at sampled sites by quartzo-feldspathic migmatites where uranium U has been depleted by weathering and leaching, explaining the low values in samples 024 (3.81 Bq∙kg⁻¹) and 030 (3.23 Bq∙kg⁻¹). It is 66.7% lower than Northern Uganda 57.1 ± 5.4 Bq∙kg⁻¹, 94.7% lower than Oyo, Nigeria 355.95 ± 3.91 Bq∙kg⁻¹, 70.0% lower than Anka, Nigeria 63.29 ± 13.87 Bq∙kg⁻¹, and 81.7% lower than Aja, Saudi Arabia 103.80 ± 78.96 Bq∙kg⁻¹. Oyo and Aja are documented High Natural Background Radiation Areas developed on uraniferous charnockites and peraluminous granites respectively [27][29]. The fact that Mbeere South is 70% - 95% lower confirms the absence of such highly differentiated U-rich granites in the sampled sites.

The only comparable value is Jaintiapur, Bangladesh 25 ± 2 Bq∙kg⁻¹. Mbeere South is 23.9% lower than Jaintiapur, while Jaintiapur is only 31.4% higher than Mbeere. Jaintiapur is a sedimentary basin with sandstones and shales [30]. The similarity suggests Mbeere South accessible outcrops behave radiologically more like a low-U sedimentary terrain than a typical Kenyan basement granite. It is 73.9% lower than San Paulo, Brazil 72.87 ± 71.65 Bq∙kg⁻¹, which is developed on alkaline intrusives [31].

The mean activity concentration of ²³²Th is 40.5% lower than Kerio Valley 65.0 ± 23 Bq∙kg⁻¹, 68.4% lower than Northern Uganda 122.3 ± 11.2 Bq∙kg⁻¹, 94.6% lower than Oyo 709.17 ± 13.94 Bq∙kg⁻¹, 82.9% lower than Anka 226.67 ± 28.05 Bq∙kg⁻¹, and 91.9% lower than Aja 477.2 ± 354.7 Bq∙kg⁻¹. The extremely high Th in Oyo, Anka and Aja is attributed to monazite and thorite-bearing granites [27]-[29]. The 82% - 94% lower values in Mbeere South suggest limited monazite content in the sampled gneisses.

Importantly, Mbeere South is 4.5% higher than Jaintiapur 37 ± 4 Bq∙kg⁻¹ and essentially comparable to it. Jaintiapur is 4.3% lower than Mbeere. This again groups Mbeere South with low-Th sedimentary terrains. It is 65.9% lower than San Paulo 113.42 ± 102.88 Bq∙kg⁻¹. The large SD in Mbeere (±34.43) is due to two high samples 014 (122.14 Bq∙kg⁻¹) and 015 (107.58 Bq∙kg⁻¹) which hypothetically reflect local concentration of resistate Th-bearing minerals like zircon, versus low samples 030 (4.82 Bq∙kg⁻¹) and 023 5.34 Bq∙kg⁻¹ in quartz veins.

The mean activity concentration of ⁴⁰K is 36.8% lower than Kerio 1003.9 ± 217 Bq∙kg⁻¹, 30.6% lower than Northern Uganda 914.2 ± 30.2 Bq∙kg⁻¹, 42.4% lower than Oyo 1102.02 ± 22.05 Bq∙kg⁻¹, 23.8% lower than Anka 832.59 ± 241.53 Bq∙kg⁻¹, 18.7% lower than Aja 780.6 ± 570.5 Bq∙kg⁻¹, 28.2% lower than Jaintiapur 884 ± 41 Bq∙kg⁻¹ and 43.7% lower than San Paulo 1127.54 ± 329.46 Bq∙kg⁻¹.

Although elevated versus world average, Mbeere K is 18% - 43% lower than the compared basement areas. The elevation is hypothetically attributed to K-feldspar and microcline in the Mozambique Belt gneisses, observed as very high values in 018 (1223.44 Bq∙kg⁻¹), 026 (1138.68 Bq∙kg⁻¹) and 015 (1023.07 Bq∙kg⁻¹). The low value 006 (76.63 Bq∙kg⁻¹) likely represents a mafic band. High K is common in basement terrains and contributes less to dose than U and Th due to its lower dose conversion factor.

The radiological hazard indices calculated from activity concentrations are presented in Table 3 while the variations in the radium equivalent activity, absorbed dose rate and annual effective dose values among the investigated sample are represented graphically in Figure 3Figure 3. Radium equivalent activity (Ra_eq) ranged from 36.85 - 264.09 Bq∙kg⁻¹ with a mean of 123.19 ± 73.10 Bq∙kg⁻¹. The absorbed dose rate (D) ranged from 18.48 ± 1.47 to 122.18 ± 4.19 nGy∙h⁻¹ with a mean of 58.61 ± 33.21 nGy∙h⁻¹, while outdoor annual effective dose (AED_out) ranged from 0.0227 - 0.1498 mSv∙y⁻¹ with a mean of 0.0719 ± 0.0407 mSv∙y⁻¹.

Ra_eq values for all the 30 samples are below the screening limit of 370 Bq∙kg⁻¹. The mean is 66.7% below the limit. The maximum value from sample 015 (264.09 ± 9.91) Bq∙kg⁻¹ is still 28.6% below limit, while minimum sample 004 (36.85 ± 3.09) Bq∙kg⁻¹ is 90.0% below limit. High values 015 (264.09 ± 9.91), 010 (250.73 ± 9.53), 029 (227.37 ± 8.60), 026 (233.00 ± 8.96] and 018 (219.10 ± 8.49) hypothetically reflect local enrichment of Th-bearing resistate minerals in granitic gneiss,

Table 3. Radium equivalent activity, absorbed dose rate and outdoor annual effective dose.

Figure 3. Variation of radium equivalent activity, absorbed dose and annual effective dose among the samples.

while low values 004 (36.85 Bq∙kg⁻¹), 016 (39.95 ± 3.09 Bq∙kg⁻¹), 030 (43.57 ± 3.62 Bq∙kg⁻¹), 024 (44.47 ± 3.75 Bq∙kg⁻¹) and 023 (48.06 ± 3.77 Bq∙kg⁻¹) hypothetically reflect quartzitic bands. The high SD of ±73.10 indicates high heterogeneity within the Mozambique Belt lithology.

The mean D is 0.66% below world average of 59 nGy∙h⁻¹. 12 of the 30 samples exceed world average. The highest D was in sample 015 (122.18 ± 4.40 nGy∙h⁻¹) is 107.1% above the world average, 010 (114.35 ± 4.21 nGy∙h⁻¹) is 93.8% above, 029 (106.70 ± 3.86 nGy∙h⁻¹) is 80.8% above. The lowest D is 004 (18.48 ± 1.47 nGy∙h⁻¹) is 68.7% below world average. This pattern follows Ra_eq, driven by ²³²Th contribution (D factor 0.604).

The mean AED_out is 2.7% higher than world average 0.07 mSv∙y⁻¹. This is an outdoor external gamma value only. The highest AED_out value 015 (0.1498 ± 0.0054 mSv∙y⁻¹) is 114% above world average, while the lowest 004 (0.0227 ± 0.0018 mSv∙y⁻¹) is 67.6% below.

From an external gamma screening perspective, the 30 purposively sampled rocks show low radiological hazard. Mean Ra_eq is well below 370 Bq∙kg⁻¹, mean D is at world average and mean AED_out is close to the world average. The rocks sampled may be considered suitable for use as building materials for external gamma. Suitability with respect to indoor radon exhalation cannot be concluded because radon was not assessed.

The mean radiological hazard indices from the present study are compared with values reported in Kenya and other parts of the world as presented in Table 4.

Table 4. Comparison of mean radiological parameters in the present study with values reported in other regions of the world.

Mbeere South samples are 53.5% lower in Ra_eq, 53.0% lower in D, and 53.0% lower in AED_out than those from Kerio Valley. Both are in basement complex, but Kerio values are more than double Mbeere South, hypothetically attributed to higher ⁴⁰K and ²³²Th in Kerio phonolites. This shows significant heterogeneity within Kenya and that Mbeere South’s sampled outcrops are not a High Natural Background Radiation Area (HNBRA). Mbeere South is 59.2% lower in Ra_eq, 57.6% lower in D, 57.7% lower in AED_out than Northern Uganda. Northern Uganda granite has been reported with ²³²Th up to 122.3 Bq∙kg⁻¹ vs Mbeere’s mean of 38.66 Bq∙kg⁻¹. The lower Th in Mbeere South hypothetically suggests less monazite enrichment. Mbeere South is 91.5% lower in Ra_eq, 90.8% lower in D, 90.8% lower in AED_out. Oyo exceeds Ra_eq limit by 293%, while Mbeere is 66.7% below. Mbeere South is 72.7% lower in Ra_eq, 70.8% lower in D, 70.8% lower in AED_out. Anka exceeds Raeq limit by 22%. This confirms Mbeere South’s sampled sites are low-background compared to Nigerian pegmatite/granite terrains.

Mbeere South is 85.4% lower in Ra_eq, 84.1% lower in D, 84.1% lower in AED_out than Aja. Aja is 129% above Ra_eq limit. Mbeere South is 15.6% lower in Ra_eq, 17.1% lower in D, 17.4% lower in AED_out. Both are low-background sedimentary-influenced signatures despite Mbeere being basement, hypothetically due to quartzitic dilution. Mbeere South samples are 61.7% lower in Ra_eq, 60.7% lower in D, 60.7% lower in AED_out than those from São Paulo, Brazil.

The 30 purposively selected rock samples from Mbeere South are 53% - 91% lower than High Natural Background Radiation Areas in Kenya, Uganda, Nigeria and Saudi Arabia. They are below the 370 Bq∙kg⁻¹ external gamma screening limit. Suitability of the rocks as building materials is concluded for external gamma screening only, as indoor radon exhalation was not assessed.