The radionuclides at the center of the investigation, radium-226 and radium-228, are decay products of the uranium-238 and thorium-232 series that occur naturally in the Earth’s crust. They matter for drinking water because of their chemistry and their biology. Radium is relatively soluble compared with its parent elements, so it leaches readily from minerals into groundwater. Its two isotopes have half-lives long enough to persist in aquifers, 1,600 years for radium-226 and 5.75 years for radium-228, and once ingested the body treats radium much like calcium, depositing it in bone. There, the alpha and beta radiation emitted by the isotopes and their short-lived progeny can irradiate bone tissue, a pathway historically linked to elevated risks of bone cancer and other malignancies. The WHO sets guidance levels of 1.0 becquerel per liter for radium-226 and 0.1 becquerel per liter for radium-228, values designed so that lifetime consumption at those levels would deliver an annual committed effective dose of roughly 0.1 millisieverts.

The analytical backbone of the study was high-purity germanium gamma spectrometry performed at the Radiation Protection Institute in Accra. One-liter samples were sealed in Marinelli beakers and stored for three to four weeks, allowing radium-226 to reach secular equilibrium with its short-lived progeny lead-214 and bismuth-214, and radium-228 to equilibrate with actinium-228. The detector, an N-type HPGe with 25 percent relative efficiency and a resolution of 1.8 kiloelectronvolts at 1332 kiloelectronvolts, was calibrated with a mixed radionuclide standard in identical geometry, and each sample was counted for a demanding 36,000 seconds, ten full hours, to push detection limits as low as possible. Radon-226 was quantified through the 609.3 and 351.9 kiloelectronvolt lines of its progeny, while radium-228 was measured through actinium-228 gamma lines at 911.2 and 968.9 kiloelectronvolts. Even so, the minimum detectable activities were 0.12 becquerels per liter for radium-226 and 0.14 becquerels per liter for radium-228.

That second detection limit turns out to be one of the study’s most consequential technical findings. Because the detection limit for radium-228, 0.14 becquerels per liter, sits above the WHO guidance level of 0.1 becquerels per liter, a non-detect cannot demonstrate compliance. Every sachet and tap water sample in the study, across all three regions and both seasons, returned results below the detection limits for both isotopes. For radium-226, whose detection limit is an order of magnitude below its guidance level, that genuinely establishes compliance. For radium-228, the honest reading is more cautious: the true activity could lie anywhere between zero and 0.14 becquerels per liter, a range that spans the guideline. Treating both detection limits as upper bounds yields a screening dose of 0.095 millisieverts per year, which falls just below, but not comfortably below, the WHO individual dose criterion of 0.1 millisieverts per year. The margin is roughly 5 percent, smaller than the analytical uncertainty of the measurement itself.

The untreated sources told a dramatically different story. Borehole water, the primary drinking supply for peri-urban and rural communities without piped infrastructure, contained 5.03 to 5.16 becquerels per liter of radium-226 and 5.06 to 5.34 becquerels per liter of radium-228, exceeding the WHO guidance levels by factors of roughly five and about 51 to 53 respectively. River water was higher still. Samples from the Densu, Kakum, Pra and Anankwar rivers, along with the Weija and Kpong reservoirs, registered 27.02 to 28.39 becquerels per liter of radium-226 and 29.01 to 30.09 becquerels per liter of radium-228, overshooting the guidance levels by factors of roughly 27 to 28 and 290 to 301. These river values are unusually high for surface water anywhere in the world, and the authors themselves flag that they require independent confirmation before being used for regulatory decisions.

Converting activity into dose, the researchers applied standard ingestion dose coefficients, 2.8 times ten to the minus four millisieverts per becquerel for radium-226 and 6.9 times ten to the minus four for radium-228, multiplied by an assumed consumption of two liters per day, or 730 liters per year. Under the deliberately conservative assumption that an adult drinks exclusively from a single source for an entire year, borehole water delivered a screening dose of 3.58 to 3.73 millisieverts per year, roughly 36 times the WHO criterion, while river water delivered 20.14 to 20.96 millisieverts per year, about 200 times that benchmark and twenty times the 1 millisievert per year reference level of the International Basic Safety Standards. Notably, radium-228 contributed about 70 percent of the total dose in both cases despite similar activity concentrations, because its dose coefficient is roughly 2.5 times larger. The authors stress these are screening values, not measured exposures; anyone who switches between sources would receive less, and no age-resolved estimates were made for children, infants or pregnant women.

Season emerged as a meaningful variable. Wet-season activities exceeded dry-season values in every untreated source and region, with average increases of 1.8 to 4.9 percent. Statistical testing by one-way analysis of variance confirmed the seasonal effect was significant for borehole radium-228, river radium-226 and river radium-228, though not for borehole radium-226. The pattern is consistent with enhanced mineral dissolution, groundwater recharge and catchment runoff during the rains, which mobilize radionuclides from aquifer and catchment materials. Equally telling was what the regional comparison did not find: combined radium activities in river water were nearly identical across the three regions, 57.22, 57.26 and 57.24 becquerels per liter, with no statistically significant differences. The dataset therefore provides no basis for attributing regional variation to differences in underlying geology.

What explains the elevated activities in the untreated sources? The coastal belt is underlain by Precambrian crystalline rocks, Birimian metavolcanics, Tarkwaian sediments and coastal sedimentary basins, all of which host uranium- and thorium-bearing accessory minerals such as monazite, zircon and apatite. Weathering, dissolution and alpha-recoil release of radium from these minerals is the most plausible geogenic source, and the catchments of the Kakum, Pra and Anankwar rivers also host extensive artisanal gold mining that disturbs mineralized material and inflates suspended sediment loads. But the authors are careful: the study measured no hydrogeochemical parameters, performed no mineralogical characterization and conducted no source apportionment, so a geogenic origin remains a hypothesis rather than a demonstrated conclusion. A further technical caveat concerns the river samples themselves. They were collected as unfiltered whole-water samples and acidified to below pH 2, a procedure that leaches radium from suspended particles into solution. The reported values therefore represent total radium, dissolved plus particulate-bound, and likely overstate what a consumer actually absorbs.

The authors are equally explicit about what the exceedances do and do not mean. WHO guidance levels are conservative operational screening criteria, not thresholds of harm, and the 1 millisievert reference level is a trigger for planning protective action rather than a dose limit. The findings, they argue, call for confirmatory sampling by independent laboratories, source-specific investigation, and consideration of treatment, not a declaration of health detriment. Proven radium removal technologies exist, including ion exchange, reverse osmosis, activated alumina and lime softening, and the burden of exposure falls disproportionately on communities with the fewest alternatives to boreholes and rivers. The study also carries a regulatory message: Ghana’s national drinking water standards do not mandate routine radiological testing, and any future monitoring program must specify methods sensitive enough to detect activities below the guidance level, since a method whose detection limit exceeds the criterion cannot demonstrate compliance no matter how many samples are analyzed.

As a piece of baseline science, the work fills a genuine gap. Previous Ghanaian studies focused on mining-impacted areas or single municipalities and reported borehole radium-226 at or below 1 becquerel per liter, far lower than the values found here, which sit at or above the upper end of ranges typical of crystalline-basement aquifers worldwide. The recommended follow-up is concrete: paired raw and treated samples to quantify treatment removal, parallel filtered and unfiltered river sampling to separate dissolved from particulate radium, hydrogeochemical and mineralogical work to test the geogenic hypothesis, and an age-resolved dose assessment grounded in an actual consumption survey. Until then, the image that lingers is one of contrast, of city dwellers sipping radiologically quiet sachet water while, a few kilometers away, families draw water from boreholes and rivers that geology has quietly loaded with radium.

Subject of Research: Naturally occurring radium-226 and radium-228 contamination in drinking water sources along Ghana's coastal regions

Article Title: Assessment of 226Ra and 228Ra activity concentrations in drinking water sources along Ghana’s coastal communities

Article References: Agyeman, B. K., Adukpo, O. K., Adotey, D. K., & Serfor-Armah, Y. (2026). Assessment of 226Ra and 228Ra activity concentrations in drinking water sources along Ghana’s coastal communities. Discover Geoscience, 4(1), Article 397. https://doi.org/10.1007/s44288-026-00765-2

Image Credits: AI Generated

DOI: 10.1007/s44288-026-00765-2

Keywords: radium-226, radium-228, drinking water, Ghana, gamma spectrometry, WHO guidance levels, groundwater, natural radioactivity, radiological assessment, borehole water, river water, public health