An approach to assessing the probability of unsatisfactory radon in air-conditioned offices of Hong Kong

Abstract

In order to maintain an acceptable Indoor Air Quality (IAQ), policies, strategies and guidelines have been developed worldwide and exposure concentrations of the indoor radon have been specified. Mapping indoor radon levels for a region could be done with intensive measurements on a large number of samples. To obtain the most accurate estimate of the levels with the uncertainties specified, a statistical model has been developed in this study to predict the fractions of samples in a region having an average radon level above the action levels of 150 Bq m −3 and 200 Bq m −3. The model was based on a transformation of the variation from a small sample set of data to a population geometric distribution via an estimator, known as the ‘sample correction factor’. Using a dataset from a cross-sectional measurement of indoor radon levels in 216 Hong Kong offices, where the mean was 37.2 Bq m −3 and the 68% range was from 17.3 Bq m −3 to 80.3 Bq m −3, the ‘sample correction factor’ was evaluated and tested by the Monte-Carlo simulations. The model estimates of the fractions above the indoor radon action levels 150 Bq m −3 and 200 Bq m −3 (1.2–7.7% and 0.4–4.1% for a sample size of 20, 2.8–5.1% and 0.8–2.4% for a sample size of 60) were demonstrated to be consistent with those determined from the dataset (3.5% and 1.4%). With the ‘sample correction factor’ thus quantified, it will be possible to provide the required data for the policymakers making appropriate decisions on resources and manpower management.