Relationships between Personal Measurements of 'Total' Dust, Respirable, Thoracic, and Inhalable Aerosol Fractions in the Cement Production Industry.

AIMS The aims of this study were to examine the relationships and establish conversion factors between 'total' dust, respirable, thoracic, and inhalable aerosol fractions measured by parallel personal sampling on workers from the production departments of cement plants. 'Total' dust in this study refers to aerosol sampled by the closed face 37-mm Millipore filter cassette. METHODS Side-by-side personal measurements of 'total' dust and respirable, thoracic, and inhalable aerosol fractions were performed on workers in 17 European and Turkish cement plants. Simple linear and mixed model regressions were used to model the associations between the samplers. RESULTS The total number of personal samples collected on 141 workers was 512. Of these 8.4% were excluded leaving 469 for statistical analysis. The different aerosol fractions contained from 90 to 130 measurements and-side-by side measurements of all four aerosol fractions were collected on 72 workers.The median ratios between observed results of the respirable, 'total' dust, and inhalable fractions relative to the thoracic aerosol fractions were 0.51, 2.4, and 5.9 respectively. The ratios between the samplers were not constant over the measured concentration range and were best described by regression models. Job type, position of samplers on left or right shoulder and plant had no substantial effect on the ratios. CONCLUSIONS The ratios between aerosol fractions changed with different air concentrations. Conversion models for estimation of the fractions were established. These models explained a high proportion of the variance (74-91%) indicating that they are useful for the estimation of concentrations based on measurements of a different aerosol fraction. The calculated uncertainties at most observed concentrations were below 30% which is acceptable for comparison with limit values (EN 482, 2012). The cement industry will therefore be able to predict the health related aerosol fractions from their former or future measurements of one of the fractions.