In an international collaboration, a long-term radon concentration survey was carried out in schools of Southern Serbia with radon detectors prepared, etched and read-out in Italy. In such surveys it is necessary to evaluate measurement precision in field conditions, and to check whether quality assurance protocols were effective in keeping uncertainties under control, despite the complex organisation of measurements. In the first stage of the survey, which involves only some of the total number of municipalities, paired detectors were exposed in each monitored room in order to experimentally assess measurement precision. Paired passive devices (containing CR-39 detectors) were exposed for two consecutive 6-month periods. Two different measurement systems were used to read out CR-39s of the first and second period, respectively. The median of the coefficient of variation (CV) of the measured exposures was 8 % for 232 paired devices of the first 6-month period and 4 % for 242 paired devices of the second 6-month period, respectively. This difference was mainly due to a different track count repeatability of the two read-out systems, which was 4 and 1 %, respectively, as the median value of CV of repeated countings. The in-field measured precision results are very similar to the precision assessed in calibration conditions and are much lower than the room-to-room variation of radon concentration in the monitored schools. Moreover, a quality assurance protocol was followed to reduce extra-exposures during detector transport from Rome to schools measured and back.
Read full abstract