Accreditation process for testing laboratories according to international standard ISO/IEC 17025 (1) requires estimation of measurement uncertainty, which is in certain areas, a quite demanding and complicated process. Measurement of occupational exposure with passive personal dosimeters is one of these areas with nonlinear model function and therefore numerical methods of analysis as Monte Carlo modelling are proposed. This paper describes simplified uncertainty analysis performed for Panasonic TLD system composed from UD-716 automatic TLD reader and UD-802 personal dosimeters used in two dosimetry services in Croatia. The model function and sources of uncertainty with their probability density distributions, all based on literature data (2), type testing and work experience, are presented and discussed. It was found that expanded uncertainty of less than ± 30% can be achieved using very simple reconstruction algorithm. System stability testing in the scope of quality assurance (QA) program provides data on system stability over long time periods. It was found that average system response over long time periods can be maintained in the range of ± 2% with reader critical components routine maintenance. Main contributions to uncertainty in given simple model, advantages and disadvantages in comparison with more complicated algorithms and possibilities for improvements are also discussed.

Javno zdravstvo i zdravstvena zaštita