INTERCOMPARISON BETWEEN CONIMETRIC AND GR AVIMETRIC SAMPLING OF RESPIRABLE DUST IN URANIUM MINES IN SAXONY AND THURINGIA - RESULTS AND CONCLUSIONS IN VIEW OF RETROSPECTIVE EXPOSURE ASSESSMENT

In a research project funded by the Hauptverband der gewerblichen Beru fsgenossenschaften the dust concentrations in workplaces of former SDA G Wismut mines will be estimated retrospectively. The evaluation is to be performed on a gravimetric base. A large amount of konimetric meas urement data for respirable dust is available from the successor compa ny, the Wismut GmbH, beginning in the year 1960. Until the early eight ies konimetry was the only measurement procedure applied on a routine base for the grouping of work procedures and job categories respective ly. Even until 1990 it was the predominat procedure. Prerequisite for the recalculation of particle number concentrations into gravimetrical respirable dust concentration is the availability of a sufficient sta tistically sound recalculation function between the two methods. For t his purpose a large number of intercomparison experiments between grav imetry and konimetry were performed in underground workplaces of the W ismut GmbH, steadily increasing the number of measurement instruments and the frequency of konimetric samples taken during the measurement i ntervals in the course of the experiments. Intercomparisons were perfo rmed in different lithologies representative for the conditions of ura nium mining and under varying ventilation velocities in the range betw een 0.1 and 2.2 m/s. The preparation and treatment of the konimeter di scs before the counting and the counting itself were performed accordi ng to the original standards of the former GDR and the requirements of the SDAG Wismut respectively. Nonplausible deviations between two ide ntical konimetric measurements within the same experiment could be rec ognized and eliminated eventually by use of registrating tyndallometri c devices in parallel. This was only done in proven single cases and t hus only 64 of the total of 4300 konimetric data (i. e. 1.5 %) had to be disregarded. This small number of not included singular data had no consequences for the statistical evaluation. While using all pairs of intercomparison measurements a non-linear regression curve was found to be the most appropriate fit. The correlation factor was found to be r = 0.832. For the Thuringian uranium mining district a special recal culation function could be deviated with a regression factor of r = 0. 96. A further differentiation, for example according to different lith ologies, was not possible as the number of measurement pairs was not s ufficient to allow for the deduction of a satisfying specific recalcul ation function between particle number concentration and gravimetric c oncentration.