Hollow-fiber membranes (HFMs) have been widely applied to many liquid treatment applications such as wastewater treatment, membrane contactors/bioreactors and membrane distillation. Despite the fact that HFMs are widely used for gas separation from gas mixtures, their use for mechanical filtration of aerosols is very scarce. In this work, we compared mathematical models developed for the prediction of air filtration efficiency by applying them on the structural parameters of polypropylene HFMs. These membranes are characteristic of pore diameters of about 90 nm and have high solidity, thus providing high potential for nanoparticle removal from air. A single fiber/collector and capillary pore approach was chosen to compare between models developed for fibrous filters and capillary-pore membranes (Nuclepore filters) based on three main mechanisms occurring in aerosol filtration (inertial impaction, interception and diffusion). The collection efficiency due to individual mechanisms differs significantly. The differences are caused by the parameters for which the individual models were developed, i.e., given values of governing dimensionless numbers (Reynolds, Stokes and Peclet number) and also given values of filter porosity and filter fiber diameter. Some models can be used to predict the efficiency of HFMs based on assumptions depending on the conditions and exact membrane parameters.

The main effort of this work was to evaluate the situation of the atmosphere in selected regions of Brno during the years 2009-2013 and to estimate health risks which might come up due to the increased concentrations of airborne particulate matter. PM₁₀ samples were collected in four areas varying in degree of automobile traffic using automatic and gravimetric sampling methods. PM₁₀ concentrations were assessed using Spearman's rank correlation coefficient. Health risks were estimated based on calculation of relative risks and population for four health endpoints. The selected health outcomes were premature mortality, cardiovascular disease, respiratory disease, and chronic bronchitis. The highest PM₁₀ concentrations were measured in two regions with high traffic loads T1, T2 and background region B2. The values were 34.33 ± 11.52 µg·m^-3 in 2010, 34.87 ± 12.03 µg·m^-3 in 2013 and 34.52 ± 8.81 µg·m^-3 in 2009, respectively. The highest correlation was between T1 and T2 having Spearman's correlation coefficient 0.888 followed by T1-B1 pair with coefficient 0.886. For all health outcomes, the highest health effect of PM (E) was determined for T2 site in 2010 which was 48 ± 14, 49 ± 21, 44 ± 19 and 24 ± 10 for premature mortality, cardiovascular disease, respiratory disease, and chronic bronchitis, respectively. The concentrations are highly correlated, especially in traffic regions. The annual concentrations did not exceed the legislation limit but 24-hours limit was exceeded more than two times in several cases. The highest number of cases with a given health outcome was estimated in traffic regions especially for cardiovascular disease and premature mortality.