Numerical investigation on the sensitivity of sampling representativeness in a nuclear facility stack

The sampling representativeness in a nuclear facility stack was numerically investigated within the ISO 2889:2023 framework, using the coefficients of variation (COV) of velocity and aerosol concentration, together with the angle of flow, as key evaluation metrics. A four-factor sensitivity study—stack height, diameter, inlet velocity, and aerosol flow rate—was conducted to quantify their effects on sampling criteria at the 37–40 m sections. Increasing the inlet velocity from 2 to 20 m/s reduced the velocity COV by approximately 30–40 % owing to enhanced turbulent mixing. Enlarging the stack diameter (0.8–1.6 m) decreased the outlet velocity by about 55 % and reduced velocity COV by 15–20 %, though the improvement became negligible beyond 1.4 m. Within the height range of 9–15 m, the velocity COV remained around 5 %, suggesting a minor influence of height. In contrast, increasing the aerosol flow rate from 0.001 to 0.1 kg/s raised the concentration COV from 5–6 % to 13.4 %, substantially degrading representativeness. This demonstrates that excessive aerosol injection can markedly degrade sampling representativeness through localized enrichment and agglomeration effects. All simulated configurations met ISO 2889:2023 representativeness requirements (velocity COV≤20 %, angle of flow ≤ 20°). These quantitative results identify aerosol flow rate and inlet velocity as the dominant parameters controlling sampling performance, while height and diameter have secondary roles with diminishing returns once geometric development is sufficient.