Comparison of different methods for the assessments of effective dose due to gamma radiation emitted by building materials

• EN 17637 calculates indoor gamma dose from construction products. • EN 17637 was compared to point-kernel and Monte Carlo methods. • EN 17637 aligns with point-kernel methods for areal densities below 500 kg m –2 • The EN 17637 dose plateau assumption (> 500 kg m –2 ) is not supported by the study. • MCNP yields higher doses, with a >50% divergence for Th-232. The transition to a circular economy promotes using industrial by-products, such as Naturally Occurring Radioactive Materials (NORM) residues, in building materials, necessitating a proper evaluation of radiological risks. To this end, the European Basic Safety Standards Directive (EU-BSS) established a regulatory framework for natural radioactivity, setting a reference level of 1 mSv/year for indoor gamma exposure. It introduced the activity concentration Index (I); if I > 1, a detailed dose assessment considering material characteristics (e.g., density, thickness) is required. Since the EU-BSS does not prescribe calculation methods, the EN 17637:2022 standard was developed to provide a harmonized approach. However, alternative methods exist, including point-kernel codes (MicroShield, RESRAD-BUILD) and Monte Carlo simulations. This study critically compares these methodologies to assess their reliability. Effective dose rates were calculated per unit activity concentration over a wide range of mass per unit area (up to about 1500 kg m –2 ) for all relevant radionuclides to facilitate direct comparison. The analysis reveals that EN 17637 standard aligns well with other point-kernel methods for mass per unit area below 500 kg m –2 , with differences generally remaining within 10–20%. However, beyond this threshold, the standard assumes a dose plateau, whereas all other models indicate a continued increase in dose. Monte Carlo (MCNP) simulations yield dose estimates up to 54% higher specifically for Th-232, primarily driven by the choice of dose conversion factors and the inclusion of cross-wall scattering, which is inherently omitted in point-kernel models. Notably, when assuming isotropic geometry or excluding specific conversion factor influences, these overall differences generally fall below 15–20%. The study demonstrates that while EN 17637 remains an essential tool due to its simplicity, its accuracy for high-density materials can be improved by applying some adjustment factors proposed in this work.