Measurement uncertainty in odor emission measurements - conclusions from 10 years of proficiency testing

Abstract

Dynamic olfactometry according to EN 13725 is widely used as the standard method for quantifying odor emissions, yet its measurement uncertainty remains debated. This study evaluates ten years (2015–2024) of proficiency testing data, including more than 1900 measurements on a stack simulator and compares these with over 1000 real-world emission analyses. The key findings are a repeatability standard deviation (s_r) of 0.53 dB_od and a reproducibility standard deviation (s_R) of 2.55 dB_od for general odors. The expanded uncertainty (95% confidence) is estimated at 5.10 dB_od. For the reference odorant n-butanol, s_R was lower (1.87 dB_od), reflecting assessor pre-screening. Importantly, the sample composition (single substances versus complex mixtures) did not impact variability. Our results demonstrate that proficiency tests on a stack simulator accurately reflect real-world variability, challenging previous assumptions of lower uncertainties (≤2 dB_od). The findings underscore the need for revised uncertainty estimates in regulatory frameworks. This work provides the most comprehensive empirical basis to date for optimizing the implementation of EN 13725 and guiding future standardization efforts.