These are planning estimates rather than forecasts of reactor construction, and the study did not evaluate economics, licensing, fuel-fabrication complexity or detailed plant siting. The study examined the possibility that small modular reactors could provide firm, dispatchable electricity for expanding data-center loads in Brazil. Data centers have high and largely continuous electricity requirements, which can strain transmission networks, reserve margins and operating flexibility as renewable generation grows.

Small modular reactors, generally defined by the International Atomic Energy Agency as reactors with electrical outputs of up to 300 megawatts, are being considered as one possible source of low-carbon, controllable power. Their modular construction and potential proximity to large electricity users could reduce some transmission pressures, although the research does not establish an SMR supply chain, plant deployment or commercial operation for Brazilian data centers. To construct a reference demand scenario, the researchers screened Brazilian ministerial ordinances issued between 2020 and 2025 concerning data-center connections to the National Interconnected System.

They identified 21 projects whose grid access had been formally recognized, then found power-demand information for 18 of them using regulatory records, developer and operator information, industry databases and news reports. The resulting projects totaled 2,525 megawatts: 1,574 megawatts, or 62 percent, in São Paulo; 776 megawatts, or 31 percent, in Ceará; 95 megawatts, or 4 percent, in Rio Grande do Sul; and 80 megawatts, or 3 percent, in Bahia. The authors stressed that this was an illustrative input assembled from identified projects and publicly available information, not a projection of Brazil’s total future data-center demand.

The comparison used three pressurized-water-reactor designs, selected because Brazil already operates this reactor class and because the available designs supplied sufficiently consistent public data on core geometry, enrichment and burnup. ACP100 is a Chinese design with a nominal output of 125 megawatts per unit; BANDI is a South Korean 60-megawatt design; and RITM-200N is a Russian design whose two-reactor plant has a combined nominal output of 110 megawatts. The authors represented each core in the community-developed OpenMC Monte Carlo transport code and calculated initial heavy-metal mass from the modelled materials, densities, volumes and number of fuel assemblies.

Because the purpose was a representative inventory estimate rather than a depletion calculation, the models did not simulate neutrons changing the fuel composition as it burned. Fuel-cycle calculations then combined the core inventories with average discharge burnup taken from technical literature, the assumed thermal efficiency of the reactors and the reference electricity demand. The authors distinguished the heavy metal initially loaded into a fleet from cumulative fuel consumption.

Initial heavy-metal mass represents the uranium loaded at the start of operation, whereas burnup measures the thermal energy produced per unit mass of initial heavy metal. Natural-uranium requirements were derived through an enrichment mass balance using product enrichment, natural-uranium assay, depleted tails assay and an allowance for conversion losses. The analysis assumed continuous full-power operation, producing a 100 percent capacity factor and an upper-bound estimate of annual electricity and uranium use.

Reducing the capacity factor to 95 or 90 percent lowered annual uranium demand by 5 or 10 percent without changing the rankings among the designs. The designs differed sharply in their fuel characteristics. ACP100 uses three uranium-dioxide enrichment zones and had an effective enrichment of 2.99 percent in the study, with 15.95 tonnes of heavy metal in one core, a net electrical output of 125 megawatts and average burnup of 16.44 megawatt-days per kilogram of heavy metal.

BANDI uses conventional low-enriched uranium dioxide in 52 assemblies, with 4.95 percent enrichment, 11.89 tonnes of heavy metal per core, a 60-megawatt output and an average burnup of 29.4 megawatt-days per kilogram. RITM-200N uses uranium dioxide in a silumin matrix at 19.75 percent enrichment. Its much more compact core contained 2.48 tonnes of heavy metal, while the two-reactor configuration produced 110 megawatts and had an average burnup of 96.4 megawatt-days per kilogram.

These comparisons concern reactor inventories and fuel-utilisation assumptions, not the practical fuel-manufacturing readiness of the three concepts. When the reference demand was assigned entirely to nuclear generation, ACP100 required 21 units providing 2,625 megawatts of installed capacity. Its fleet-wide initial heavy-metal inventory was 345.24 tonnes, its estimated specific natural-uranium requirement was 55.956 grams per megawatt-hour, and continuous operation would require 1,286.71 tonnes of natural uranium per year.

BANDI required 43 units for 2,580 megawatts, with the largest initial fleet inventory at 511.27 tonnes, but the lowest specific and annual natural-uranium requirements at 48.405 grams per megawatt-hour and 1,093.99 tonnes per year. RITM-200N required 46 units for 2,530 megawatts. Its initial inventory was only 132.48 tonnes, yet its 19.75 percent enrichment and fuel composition produced the highest specific natural-uranium requirement, 59.274 grams per megawatt-hour, and an estimated fleet demand of 1,313.68 tonnes per year.

The results show why unit size, initial inventory, burnup and enrichment cannot be treated as interchangeable measures of fuel efficiency. Regional deployment was also dependent on the discrete size of each reactor. São Paulo’s 1,574-megawatt reference load would require 13 ACP100 units, 27 BANDI units or 29 RITM-200N units, with total initial inventories of 213.72, 321.03 and 83.52 tonnes, respectively.

Ceará’s 776-megawatt demand would require seven, 13 or 15 units. By contrast, the smaller loads in Rio Grande do Sul and Bahia could each be met with one ACP100 unit or two BANDI or RITM-200N units. Installed capacity exceeds the target load in these cases because reactor outputs cannot be divided into fractional units.

The counts describe generating units, not independent nuclear sites: manufacturers’ configurations may pair reactors in one plant, and multiple plants could potentially occupy one larger site. The estimated uranium requirements would be consequential for Brazil’s developing domestic supply. The currently operating Caetité facility has a nominal capacity of about 220 tonnes of uranium per year, while projected expansions associated with Caetité and Santa Quitéria could raise production to approximately 678 and 1,950 tonnes per year, respectively.

Against those figures, the study’s full-power annual requirements of roughly 1,094 to 1,314 tonnes would represent a substantial share of projected capacity, although Brazilian uranium production is not the same as the country’s full resource potential. The study also did not assess mine and conversion supply chains in detail, and the RITM-200N’s silumin-based fuel may introduce fabrication and processing issues not captured by kilograms of natural uranium. The findings therefore identify a trade-off relevant to fuel-cycle planning: compact, highly enriched cores can minimize the uranium loaded initially without minimizing the upstream natural uranium required per unit of electricity.

Overall, the modelling supports a more nuanced view of SMRs for data centers than a simple ranking by reactor size or uranium inventory. ACP100 minimizes the number of units in the reference case but has a higher annual natural-uranium requirement than BANDI. BANDI offers the lowest estimated annual uranium use but requires the most units and the greatest initial fleet inventory.

RITM-200N offers the smallest initial inventory but not the lowest upstream fuel requirement. Because the analysis assumes that SMRs supply the entire 2.525-gigawatt reference load, operates every reactor continuously and does not model economic costs, detailed plant arrangements, licensing pathways or fuel supply constraints, it cannot identify an optimal commercial choice. Its principal contribution is a physically grounded screening framework showing how the same data-center demand can produce very different deployment and material requirements depending on a reactor’s core design, enrichment and burnup.

Subject of Research: Comparative modelling of small modular reactor deployment and uranium requirements for Brazilian data centers

Article Title: Small modular reactors for data centers: deployment and nuclear fuel requirements

Article References: de Abreu, W. V., Martinez, A. S., Silva, A. S. E., Takashi, L. K., Ribeiro Lima, Á. C., de Pontes, J. D. B. B., Ramon, L. L. S., Gonçalves, A. D. C., Silva, A. C. D., Silva, F. C. D., & Lima, M. A. C. (2026). Small modular reactors for data centers: deployment and nuclear fuel requirements. Frontiers in Nuclear Engineering, 5, Article 1971830. https://doi.org/10.3389/fnuen.2026.1971830

Image Credits: AI Generated

DOI: 10.3389/fnuen.2026.1971830

Keywords: small modular reactors, data centers, nuclear fuel, uranium demand, reactor modelling, OpenMC, fuel burnup, Brazil, pressurized water reactors, energy planning, Small, modular

The problem LEACS addresses arises directly from the push toward tightly coupled multiphysics simulation. Modern reactor analysis increasingly iterates Monte Carlo transport solvers such as OpenMC, Serpent2, MCNP, and RMC together with thermal-hydraulics, fuel performance, and structural mechanics codes. When Monte Carlo transport is coupled to a finite element thermal solver, the temperature and density fields it receives are not flat. Fuel pellets can exhibit centerline-to-surface temperature gradients of several hundred Kelvin, and coolant density can change by tens of percent across an axial span. Those gradients feed back into neutron transport through temperature-dependent cross sections, and capturing that feedback is precisely the point of running a coupled calculation in the first place. Yet conventional Monte Carlo treats cross sections as piecewise-constant along a particle’s flight path, an approximation that quietly discards exactly the physics the coupling was meant to resolve.

At the heart of the challenge is the collision distance sampling step. In a standard Monte Carlo particle history, the code calculates the macroscopic cross section of the current material, samples a distance to collision from an exponential distribution, and moves the particle to whichever comes first: the sampled collision, a cell boundary, or a cutoff distance. When cross sections vary continuously in space, the cumulative distribution function from which the collision distance is sampled depends on the optical depth accumulated along the particle’s actual trajectory through the temperature field. Evaluating that integral exactly is impractical, so every practical method must approximate it somehow, and each existing approximation imposes its own tradeoffs in accuracy, generality, or speed.

The dominant existing strategies fall into three broad camps. Delta tracking, originally devised by Woodcock in 1965, samples collisions against a majorant cross section and then accepts or rejects each pseudo-collision probabilistically. It handles complex geometry gracefully, which is why it remains popular for problems like TRISO particle fuels, but it struggles when a region contains very high cross sections, such as gadolinium absorbers, and constructing a true majorant is difficult when temperature-dependent Doppler broadening shifts resonance peaks in an energy-dependent way. Adaptive binning, recently implemented in the CARDINAL code via the MoabSkinner user object, regenerates the geometry into cells of roughly equal temperature and density, restoring the simple sampling formula at the cost of a much larger geometry. Its resolution is also tied to the multiphysics mesh, which cannot be subdivided below the element size, and tetrahedral meshes are inefficient for geometries like fuel pins that are meters long but only centimeters wide. Numerical estimation methods, dating back to work by Brown, Martin, and Greisheimer, fit polynomial functions to cross sections sampled along the flight path and integrate them with quadrature rules; they are fast, but representing functional expansions on unstructured, potentially deformed meshes while preserving orthogonality is nearly impossible in three dimensions.

LEACS, which stands for Lagrange Error-Adaptive Collision Sampling, takes a different route. Rather than converting the finite element solution into a lower-order representation, it accesses the native element shape functions directly during particle transport. When a neutron enters a material flagged for LEACS treatment, the method places sample points along the flight path, queries the finite element solution at each point to obtain the local temperature, uses windowed multipole formalism to Doppler-broaden the cross sections on the fly, fits piecewise quadratic functions to the resulting cross-section profile, and inverts the reconstructed cumulative distribution function to obtain the collision distance. The windowed multipole approach is itself notable: it broadens cross sections at arbitrary temperatures with far higher accuracy than linear interpolation between library temperatures, and recent advances evaluating the Faddeeva function as a rational fraction have made it extremely efficient.

The genuinely novel element is how LEACS chooses where to place those sample points. Because querying the finite element solution is expected to be the most expensive part of the transport calculation, every sample must earn its keep. The naive choice, equidistant spacing, wastes effort: for large optical depths, the most likely collision distances cluster near the particle’s initial position, yet equal spacing devotes as many points to the unimportant region near the boundary as to the critical region near the start. LEACS instead exploits the Lagrange error bound, a classical result from Taylor’s theorem that bounds the remainder of a polynomial approximation. By requiring the approximation error to be uniform across all intervals, the method derives an analytic spacing function that concentrates points where the curvature of the collision probability distribution demands them. Every sample location is therefore analytically optimal for reconstructing the true collision distance distribution, which is how LEACS aims to limit the slowdown that finite element sampling inevitably introduces.

To test the idea, the researchers built two pseudo-pin test cases spanning the extremes of neutron flight distance in a fuel pin: a 0.95-centimeter path emulating a purely radial crossing and a 364.85-centimeter path emulating a purely axial one. The material was uniform-density uranium-238, chosen because resonance absorption in U-238 is the dominant temperature feedback effect in a typical light water reactor, and one-dimensional thermal simulations in the MOOSE framework supplied realistic temperature profiles. Comparing the reconstructed cumulative distribution functions against a finely sampled control, the analysis showed that LEACS spacing consistently outperformed equidistant spacing, with the advantage growing dramatically for long flight paths. When neutron energies were sampled from a realistic pressurized water reactor pin-cell spectrum computed in OpenMC, the error in average collision distance dropped substantially, and the gains from Lagrange spacing over equidistant spacing increased as more sample points were used.

The practical payoff showed up most clearly in the spatial tallies. Using an AP1000-based pin cell with 1 percent mixed-oxide fuel, a composition chosen because it produces the largest Doppler effect in benchmark studies, and a C5G7 benchmark assembly for larger-scale cost testing, the team coupled OpenMC transport to MOOSE thermal solutions with volumetric heating derived from the peak AP1000 linear power rating. The results revealed a clear radial and axial variation in the neutron capture reaction rate that the constant-temperature treatment misses entirely. Strikingly, the root mean square difference between a three-sample LEACS run and a twenty-one-sample reference stayed below 0.35 percent, with average z-scores below 1.0, suggesting that as few as three sample points per flight segment are sufficient to capture the spatial impact of temperature gradients on the reaction rate distribution.

Computational cost remains the method’s principal hurdle. Default collision distance sampling takes roughly 0.068 microseconds per call; LEACS timing is orders of magnitude higher, with point location and inverse mapping within the finite element sampling routine dominating the bill. Slowdown factors depend strongly on the model: 4.7 to 9.7 for a two-dimensional pin cell, up to 16.4 for a three-dimensional pin cell with eleven samples, but only 1.3 to 2.1 for full assembly models, where the baseline calculation is already expensive and neutrons spend much of their history in regions that do not invoke LEACS. The authors note that a slowdown exceeding sixteen would likely be impractical for production work, while a factor of 1.3 may be acceptable in many applications. Future optimization will target the element search and inverse mapping algorithms, and planned extensions include density coupling, a capability that would matter for molten salt and boiling water reactors. The researchers foresee the method’s best home inside the CARDINAL coupling framework, where a direct head-to-head comparison with adaptive binning on both accuracy and cost would settle how this new chapter in Monte Carlo transport should be written.

Subject of Research: Error-adaptive collision sampling for Monte Carlo neutron transport coupled to finite element thermal solutions

Article Title: LEACS: lagrange error-adaptive collision sampling for multiphysics coupled Monte carlo transport

Article References: Kendrick, W. R., & Forget, B. (2026). LEACS: lagrange error-adaptive collision sampling for multiphysics coupled Monte carlo transport. Frontiers in Nuclear Engineering, 5, Article 1900800. https://doi.org/10.3389/fnuen.2026.1900800

Image Credits: AI Generated

DOI: 10.3389/fnuen.2026.1900800

Keywords: Monte Carlo transport, neutron transport, multiphysics coupling, LEACS, OpenMC, MOOSE, finite element methods, Doppler broadening, windowed multipole, reactor physics, collision sampling, nuclear engineering

The study, led by Sachin Shet and Ankan Charan of the Manipal Institute of Technology along with colleagues including S.V. Suryanarayana, S. Ganesan, Sabu Thomas, Srinivas Shenoy Heckadka, Bharath K Bhat and Sudha Kamath, focused on a 10 by 10 by 10 centimeter block of Ordinary Portland Cement grade 43 concrete with a measured density of 2.41 grams per cubic centimeter. The block was placed in the horizontal irradiation channel of a heavily shielded concrete bunker housing a sealed 16 curie americium-beryllium neutron source, an installation that has served the Manipal facility since its calibration in January 2007 and whose decay-corrected activity stood at 15.51 curies, or 573.87 gigabecquerels, as of May 2026.

The americium-beryllium source generates neutrons through the beryllium-9 (alpha, n) carbon-12 reaction, in which alpha particles from the decay of americium-241 strike a beryllium target. The result is a broad, continuous neutron spectrum extending up to roughly 11 megaelectronvolts, with a characteristic emission peak near 4 to 5 megaelectronvolts and a coincident 4.438 megaelectronvolt gamma ray from the first excited state of the carbon-12 nucleus. Critically, because the source sits inside a concrete bunker, neutrons scatter repeatedly from the surrounding walls before reaching the sample, introducing a substantial thermal component that is entirely absent from the bare-source spectrum described by the ISO 8529-1 standard used in the simulations.

To quantify how many neutrons survive the journey through the concrete, the researchers attached thin foils of high-purity gold-197, more than 99.99 percent pure, to the front and rear faces of the block. Gold is a classical neutron flux monitor because its capture cross-section for the gold-197 (n, gamma) gold-198 reaction is large and well characterized, with a thermal value of 98.6 barns and a resonance integral of roughly 1550 barns. The team irradiated the assembly for 19 hours, then counted the characteristic 411 kiloelectronvolt gamma ray emitted by the decaying gold-198 using a high-purity germanium spectrometer calibrated with a europium-152 source, applying careful corrections for true coincidence summing at the extremely close 5 millimeter source-to-detector distance.

The measured spectrum-integrated capture reaction rates were 9.25 times ten to the minus nineteenth per atom per second at the front face and 5.42 times ten to the minus nineteenth at the rear face, each with combined relative uncertainties below 4 percent. The OpenMC Monte Carlo code, driven by the ENDF/B-VII.1 nuclear data library and 80 million neutron histories, predicted 9.884 and 5.710 times ten to the minus nineteenth respectively. The differences, 6.85 percent at the front and 5.39 percent at the rear, correspond to approximately 1.9 and 1.1 combined standard deviations, an agreement the authors attribute in part to localized density and compositional variations in the real block relative to the nominal model.

Even more striking was the agreement in the transmission ratio. The experimental ratio of rear-to-front reaction rates came out at 0.586, while the simulation gave 0.578, a difference of just 1.37 percent. Because systematic uncertainties largely cancel in a ratio, this figure represents the cleanest validation of the computational model. The researchers derived a transmission factor of 0.594, meaning the 10 centimeter block reduces the detectable neutron flux by roughly 41 percent beyond the geometric attenuation already expected from distance and environmental scattering. Simulations also confirmed that fast neutrons are attenuated more efficiently than thermal ones, consistent with hydrogen in the concrete moderating and capturing fast neutrons, and that room-return scattering keeps the flux at 20 centimeters at 51.33 percent of the 10 centimeter value, far above the 25 percent predicted by a simple inverse-square law.

The elemental composition of the concrete, drawn from the PNNL-15870 Rev. 2 compendium and consistent with Bureau of Indian Standards specifications, lists silicon, oxygen, calcium, aluminum, potassium, iron, sodium and trace manganese among its key constituents. Energy-dispersive X-ray spectroscopy confirmed the presence of the principal medium-Z elements on the actual block surface, with an elevated calcium-to-silicon ratio reflecting analysis of cement-paste-rich zones rather than any discrepancy with the bulk reference composition used in transport calculations.

Beyond shielding, the study probed what the neutron field does to the concrete itself. Gamma-ray spectrometry of the irradiated block, begun two hours after the 19-hour exposure, revealed prominent photopeaks from sodium-24 and manganese-56. Sodium-24 arises mainly from thermal capture on the concrete’s sodium content, with a secondary fast-neutron contribution from the aluminum-27 (n, alpha) reaction, while manganese-56 is produced by thermal capture on trace manganese impurities, present at only 100 to 500 parts per million but amplified by manganese-55’s large 13.3 barn capture cross-section, and by the iron-56 (n, p) pathway. A peak at 1460.8 kiloelectronvolts was traced to naturally radioactive potassium-40 rather than activation. These findings underscore that even trace constituents can dominate the induced radioactivity of concrete shielding, a consideration of real consequence for decommissioning and waste management.

The authors are candid about the scope of their work. Measurements were taken at only two positions, sufficient to determine a bulk transmission factor but not to resolve the spatial attenuation profile inside the block, and no secondary gamma-ray dose calculations were performed despite the significant induced activity from sodium-24 and manganese-56. Both limitations, they note, define the agenda for future campaigns at the Manipal facility, where multiple foil positions and dose assessments could extend the benchmark.

For a field racing toward compact reactors, fusion pilot plants and novel shielding composites doped with boron, heavy minerals and fibers, the message of this study is quietly reassuring: ordinary concrete behaves largely as the simulations say it should, and the open-source OpenMC framework paired with ENDF/B-VII.1 data is a trustworthy guide to its performance. As new materials are proposed to replace or augment it, this experimentally validated baseline provides the yardstick against which they will be judged.

Subject of Research: Experimental and Monte Carlo benchmarking of neutron shielding and activation in ordinary concrete using an AmBe neutron source

Article Title: Neutron shielding assessment of ordinary concrete in an AmBe mixed-energy neutron field

Article References: Shet, S., Charan, A., Suryanarayana, S., Ganesan, S., Thomas, S., Heckadka, S. S., Bhat, B. K., & Kamath, S. (2026). Neutron shielding assessment of ordinary concrete in an AmBe mixed-energy neutron field. Results in Physics, 89, Article 108753. https://doi.org/10.1016/j.rinp.2026.108753

Image Credits: AI Generated

DOI: 10.1016/j.rinp.2026.108753

Keywords: neutron shielding, ordinary concrete, americium-beryllium source, OpenMC, Monte Carlo simulation, gold foil activation, neutron transmission, HPGe spectrometry, induced radioactivity, sodium-24, manganese-56, nuclear reactor shielding