The arithmetic of the fuel cycle explains why demand rises so steeply. A single 1-gigawatt light-water reactor consumes approximately 200 to 250 tonnes of natural uranium per year, depending on burn-up and fuel management strategies. Every new large reactor added to the grid therefore locks in decades of fuel purchases, and the current construction pipeline spans continents. Egypt is building four VVER-1200 units at El Dabaa. Bangladesh is completing the two-unit Rooppur plant. The United Arab Emirates’ four-unit Barakah complex, producing around 5.6 gigawatts, is now fully operational. Poland is advancing a Westinghouse-led AP1000 programme, Hungary is continuing Paks-II, the Czech Republic has begun early small modular reactor work with Rolls-Royce at Temelín, and Vietnam, Indonesia, Saudi Arabia, Argentina and Brazil are all expanding or reviving nuclear programmes. South Africa, operator of the continent’s only commercial reactors at Koeberg, is planning new capacity and life extensions.

Two technological trends will shape how this demand materialises. The first is the small modular reactor, or SMR, which promises factory fabrication, lower capital entry barriers and flexible siting. Most SMR concepts still burn low-enriched uranium, so their roll-out would add to, rather than substitute for, conventional uranium demand, while geographically dispersing fuel offtake contracts into many smaller and more frequent agreements. The review treats SMR uptake as a sensitivity rather than a certainty, noting that deployment depends on licensing timelines, the transition from first-of-a-kind to nth-of-a-kind manufacturing, supply-chain maturity and financing competitiveness. Cost reductions are possible but not guaranteed, and the pace of dispersal remains scenario-dependent.

The second trend is the rise of high-assay low-enriched uranium, or HALEU, enriched to between 5 and 20 percent uranium-235, which many advanced reactors and SMRs require to achieve higher burn-up and more compact cores. The review identifies HALEU production as a pivotal bottleneck: meeting projected demand will require significant investment in conversion and enrichment capacity, and the fuel form brings new logistical, regulatory and non-proliferation challenges. A related analysis cited in the study, published in Science in 2024, highlighted the weapons-relevant concerns surrounding HALEU, underscoring why its supply chain will attract intense safeguards scrutiny. Meanwhile, centrifuge technology has already displaced energy-hungry gaseous diffusion for enrichment, and laser-based concepts under pilot study could reshape future capital, regulatory and safeguard frameworks.

Fast reactors and closed fuel cycles offer a longer-term hedge. Fast-spectrum systems operating in a closed cycle can extract many-fold more energy from the same uranium inventory, converting depleted uranium stocks and recycled actinides into fissile material and reducing long-lived radiotoxic waste. If widely deployed later this century, they could substantially lower fresh-uranium mining needs. But the review is careful to treat fast reactors as a separate long-term sensitivity, since commercial deployment faces formidable technical, economic, proliferation-safeguard and institutional hurdles. Fusion, by contrast, is dismissed as a near-term offset: because it relies on the deuterium-tritium cycle rather than uranium, it sits outside the front-end fuel cycle entirely, and authoritative assessments, including the IAEA’s World Fusion Outlook, indicate it is unlikely to materially affect reactor-related uranium demand within the next two decades given the ITER-to-DEMO timeline and unresolved tritium-breeding challenges.

The market picture that emerges is one of tightening supply. Primary mine production has historically ranged between roughly 55,000 and 65,000 tonnes per year, well below the projected requirements of the 2030s and 2040s. Utilities have historically bridged this gap with secondary supply, including commercial inventories, civil stockpiles and re-enrichment of tails, but those buffers are finite. The joint NEA and IAEA Red Book reports identified recoverable resources exceeding 7.9 million tonnes, yet stresses that timely investment in exploration, permitting and processing is needed to prevent shortfalls. New mines typically require many years to move from exploration through permitting to construction, and under-investment in exploration since the 2010s has limited near-term flexibility. The commercial consequence is already visible: utilities are signing longer-term supply contracts rather than chasing spot-market bargains.

Geopolitics adds a further layer of fragility. Many newcomer nuclear programmes are vendor-financed with bundled fuel-supply contracts that channel demand into the vendor’s national fuel cycle. Russia currently controls around 46 percent of global enrichment capacity and roughly 20 percent of conversion capacity, alongside substantial upstream mining interests through joint ventures. The review suggests that scenarios should assume 30 to 50 percent of incremental uranium demand from vendor-financed builds will be met through the vendor’s home-country fuel cycle unless diversification or counter-contracts are negotiated, a structural dependency that recent analyses of European reliance on Russian nuclear fuel services have made uncomfortably concrete.

On the supply side, the review emphasises that uranium deposits are extraordinarily heterogeneous, and that geology dictates both extraction technology and environmental footprint. Canada’s Athabasca Basin hosts the world’s highest-grade ores, at McArthur River and Cigar Lake, where uranium concentrations can exceed 10 to 15 percent, orders of magnitude above typical global deposits, allowing large recoveries from small volumes of rock using specialised freeze-wall and water-jet underground methods. Kazakhstan, Namibia and Australia, by contrast, mine lower-grade but volumetrically vast deposits. Sandstone-hosted roll-front deposits are amenable to in-situ recovery, in which permeable ore horizons are leached underground, avoiding surface disturbance and tailings but imposing strict hydrogeological restoration requirements. Large low-grade open-pit operations such as Husab and Rössing in Namibia, and polymetallic mines like Olympic Dam in Australia, where uranium is a co-product of copper-gold mineralisation, illustrate the economic diversity of the resource base. Because mining energy intensity, water demand and tailings production scale inversely with ore grade, the choice of deposit and method carries direct sustainability consequences.

Exploration technology is also evolving rapidly. Modern campaigns combine classical geological mapping with airborne gamma-ray spectrometry, magnetics, gravity and electromagnetic surveys, followed by radon sampling, borehole logging and confirmatory drilling under IAEA protocols. Machine learning is increasingly used to assimilate radiometric, geochemical and structural datasets into probabilistic mineral-potential maps, improving target ranking, though the review cautions that robust training data, transparent feature selection and geological validation remain prerequisites for reliability. Non-destructive assay techniques such as neutron resonance transmission analysis, which can identify and quantify uranium and other actinides isotope-specifically, support fuel-cycle inventory verification in line with IAEA safeguards practice. A further unconventional option is recovering uranium as a by-product from phosphate rock used in fertiliser production; under favourable conditions this could supply on the order of 10 to 20 percent of global reactor-related demand, with modular recovery plants capable of delivering 4,000 to 8,000 tonnes per year within a few years of commissioning, though scale-up beyond that remains speculative and hinges on fertiliser economics rather than uranium prices alone.

The review’s bottom line is that the nuclear resurgence is fundamentally underpinned by a uranium renaissance. The resource base is judged adequate through mid-century, but only if exploration, mining, conversion, enrichment and HALEU production receive timely investment under strict regulatory oversight, with occupational doses managed under ALARA principles and transparent community engagement. Advances in in-situ recovery, ore-grade modelling and hydrometallurgical processing will be key to minimising environmental impact, while innovation across conversion, enrichment, fabrication and reprocessing will redefine the efficiency and circularity of the fuel cycle. Whether the world’s newly announced reactors get their fuel, in other words, depends less on geology than on decisions being made now about capital, technology and governance, across a supply chain that has spent a decade running on standby.

Subject of Research: Projected global uranium demand trajectories and front-end nuclear fuel-cycle technologies in a renuclearising world

Article Title: Uranium at the nexus of energy security and sustainable development in a renuclearised world: demand trajectories and technological pathways

Article References: Guembou Shouop, C. J. (2026). Uranium at the nexus of energy security and sustainable development in a renuclearised world: demand trajectories and technological pathways. BMC Environmental Science, 3(1), Article 3. https://doi.org/10.1186/s44329-025-00045-3

Image Credits: AI Generated

DOI: 10.1186/s44329-025-00045-3

Keywords: uranium, nuclear energy, nuclear renaissance, small modular reactors, HALEU, fuel cycle, uranium mining, in-situ recovery, energy security, enrichment, fast reactors, sustainable mining

Naturally occurring radioactive material, known in the radiation-protection world as NORM, refers to materials containing uranium-238, thorium-232, and potassium-40, together with their decay products such as radium-226 and radon gas. These radionuclides are everywhere in the Earth’s crust, and in undisturbed settings they pose minimal risk. The concern arises when human activity, particularly mining and mineral processing, concentrates and mobilizes them. Excavation, crushing of ore, and the improper disposal of waste rock and tailings can bring elevated NORM levels into contact with soil, water, and air. Chronic exposure through ingestion, inhalation, or skin contact has been linked to increased risks of lung, bone, and kidney cancers, in part because specific radionuclides accumulate in specific organs: thorium in the lungs and bones, uranium in the lungs and kidneys, and radium in the skeleton, where it chemically mimics calcium.

The Geita district is geologically primed for such enrichment. The mine sits within Archean banded iron formations, felsic volcanic rocks, and intermediate lithologies such as andesite and diorite. Banded iron formations and associated volcanic and intrusive rocks frequently host uraninite and thorite, the primary minerals of uranium and thorium, while amphibole- and biotite-bearing rocks add further natural radioactivity. Earlier studies had already recorded slightly elevated NORM levels in tailings and waste rock at the site, and similar findings have emerged from artisanal mining areas such as Rwamagasa and Buhemba. Yet despite Tanzania’s robust national radiation-safety framework, no sector-specific guidelines exist for managing NORM from gold mining, leaving a regulatory gap that the study’s authors say can produce inconsistent safety practices and uninformed land use.

To close that gap, the research team, led by Jerome M. Mwimanzi of the Nelson Mandela African Institution of Science and Technology and colleagues from institutions in Tanzania, Austria, Croatia, and South Africa, combined two methods rarely deployed together in the region. The first was high-resolution gamma-ray spectrometry of agricultural soils collected from ten farms across six villages surrounding the mine: Machinjioni, Mgusu, and Nyakabale in the Mgusu ward, and Mpomvu, Nyamalembo, and Samina in the Mtakuja ward. Sampling points were chosen using a stratified random design informed by isopleths from an AERMOD atmospheric dispersion model, capturing gradients of exposure tied to the tailings storage facility, waste rock dumps, and modeled ground-level particulate concentrations.

The analytical work was carried out at the Tanzania Atomic Energy Commission laboratory in Arusha. Soil samples were taken from the top 30 centimeters with a stainless-steel auger during the September 2023 dry season, oven-dried at 80 degrees Celsius, ground, and sieved through a 200-micrometer mesh before being sealed in canisters for four weeks. That waiting period is not arbitrary: it allows secular equilibrium to be established between radium-226, radium-224, and the short-lived daughters of radon-222 and radon-220, ensuring that gamma emissions from progeny nuclides faithfully reflect parent activities. Each sample was then counted for 50,000 seconds in a high-purity germanium coaxial detector housed in lead shielding, with radionuclides identified by their characteristic gamma energies: lead-214 and bismuth-214 lines for radium-226, actinium-228 and lead-212 for the thorium series, and the 1,460-keV line of potassium-40. Energy and efficiency calibration were performed with a multi-nuclide standard from the Czech Metrology Institute, and background corrections allowed quantification down to low minimum detectable activities.

The radiometric results were, on the whole, within internationally accepted limits. Average activity concentrations were 53 plus or minus 7 becquerels per kilogram for radium-226, 36 plus or minus 5 for thorium-232, and 205 plus or minus 38 for potassium-40. Radium-226 exceeded the global average of 35 becquerels per kilogram reported by the United Nations Scientific Committee on the Effects of Atomic Radiation, but thorium and potassium sat below their respective global benchmarks of 45 and 420. From these activities the team computed a battery of hazard indices. Radium equivalent activity averaged 121 plus or minus 12 becquerels per kilogram, well below the 370-threshold associated with a one-millisievert annual public dose. External and internal hazard indices averaged 0.30 and 0.5, both far under the limit of 1. The absorbed gamma dose rate averaged 55 nanograys per hour, just under the global population-weighted average of 59, and the outdoor annual effective dose equivalent averaged 67 microsieverts per year, slightly below the global figure of 70. The excess lifetime cancer risk estimates, 0.24 times ten to the minus three outdoors and 0.94 times ten to the minus three indoors, were also below global averages, though indoor risk was notably higher than outdoor, a pattern the authors link to occupancy time and possibly radon accumulation in dwellings.

Two findings stand out from the environmental data. First, the annual gonadal dose equivalent averaged 380 microsieverts per year, above the global threshold of 300 but well under the 1,000-microsievert hazard limit, driven primarily by radium-226 and thorium-232. Second, internal organ doses were roughly five times higher than external ones across all organs, highlighting inhalation and ingestion as the dominant exposure pathways. The village of Machinjioni recorded the highest cancer-risk values, likely a combination of local geology and surrounding artisanal mining, while the highest organ doses clustered in samples from Nyakabale. A correlation analysis showed that radium-226 and thorium-232, not potassium-40, dominate the absorbed dose, with the absorbed dose rate and annual effective dose showing a perfect monotonic relationship, as expected from their linked equations.

The second half of the study turned from gamma spectra to human voices. Using Yamane’s formula on the 12,856 households of the two wards, the team calculated a required sample of roughly 390 respondents and interviewed exactly that number, 65 from each village, through structured questionnaires administered by trained assistants with ethical approval from the Tanzania Commission for Science and Technology. The demographics skewed female, 59.7 percent, because sampling occurred during the day when many men were at work, and educational attainment was modest: 62.6 percent had only primary schooling, while 7.4 percent had university education.

The awareness results were striking. Overall, 64.6 percent of respondents had never heard of NORM at all. Gender made no significant difference, but age, education, and occupation did. Younger residents were more aware, plausibly owing to greater access to digital media and environmental education. Education showed the strongest association with awareness, a chi-square of 89.37 with a p-value below 0.01, and occupation was nearly as powerful: awareness ranged from 16.7 percent among peasant farmers to 88.5 percent among government employees, presumably reflecting access to structured information channels. For a community whose livelihoods revolve around farming, livestock, and fishing on mine-adjacent land, the near-blind spot among agricultural workers is the study’s most consequential finding.

Yet among the 138 respondents who did know about NORM, concern ran high. On a five-point scale, residents rated the impact on wildlife highest at a mean of 4.48, followed by water quality at 4.41, air quality at 4.34, and the natural environment at 4.25, with human health at 4.21, yielding an overall “very high” perception score. Correlations among these perceptions showed that residents hold an integrated view of environmental degradation, linking water, air, wildlife, and agriculture, but perceived health impacts correlated with none of the environmental variables, suggesting people sense the ecosystem is damaged without understanding the pathways by which radiation might reach their own bodies. Qualitative accounts sharpened the picture: residents of Nyakabale, Mtakuja, and Mgusu reported rises in miscarriages, neonatal complications, and persistent respiratory symptoms attributed to dust from ore processing, concerns echoed in the Geita Town Master Plan’s documentation of escalating respiratory disease. Declining crop yields at Nyakabale were also attributed to contamination.

The authors are careful about limits. Ten soil samples cannot characterize an entire mining landscape; the survey captures one dry-season moment and self-reported perceptions, not causality; and the dose models rely on standard assumptions rather than personal dosimetry, radon monitoring, or food-chain measurements. But the policy message is clear. The soils around the Geita Gold Mine are, by standard hazard indices, safe under current conditions, yet the combination of localized hot spots, an unregulated NORM sector, and a community that is two-thirds unaware of the hazard leaves little room for complacency. The authors call for combining ongoing radiological surveillance with targeted education campaigns delivered through community meetings, schools, and local radio, aimed especially at older residents, people with limited schooling, and farming households. Radiation, being invisible and silent, only becomes dangerous when its existence is unknown; this study suggests that in Geita, closing the knowledge gap may matter as much as any measurement.