Estimation of nanoparticle emissions in indoor industrial environments using a grey-box modeling approach

Abstract

Estimating nanoparticle emission rates from industrial activities is essential for developing quantitative risk assessment tools and prediction models for indoor air quality and occupational exposure. However, determining them is challenging, particularly for incidentally generated nanoparticles (INPs), due to their calculation from concentration measurements in complex environments with polluted backgrounds. This study addresses the challenges of defining INP emission rates by proposing a reduced-order grey-box modeling approach. The method was tested in three industrial scenarios with different thermal spraying activities, evaluating 78 models based on mass-balance aerosol concentration equations. Convergence tests, statistical analyses, and physical feasibility studies revealed that 33 % of the models met all criteria. The simplest models, incorporating forced ventilation and particle generation while excluding natural diffusion, aggregation, and deposition, demonstrated the best performance and robustness, with two models reaching a 100 % successful performance on six applied datasets. Emission rates for the monitored processes were of similar magnitude, with minor variations around 4 × 10¹⁵ particles/min attributed to the materials and component morphology. Estimated ventilation airflow rates also aligned with the expected slight underperformance of the extraction systems between 1 and 22 × 10⁷ cm³/min depending on the monitored booth and the ventilation configuration, showing air change per hour rates within the 39–105 h^-1 range. The findings highlight that grey-box modeling combined with model reduction through lumped sum parameters provides a systematic and reliable approach to estimating INP emissions. This method could inform new standard procedures. Future research should apply this approach to diverse industrial activities and exposure applications.