Assessment of test duration effect in indoor radon measurement by Monte Carlo simulations.

To better understand the effect of various test durations on indoor radon measurement results in Canada, Monte Carlo simulations were performed for test durations of 1 month (30 d), 2 months (61 d), 3 months (91 d) and 6 months (183 d). For each of the specified test durations, a total of 1500 Monte Carlo simulations were performed. Each simulation was compared with the result of a 1-y measurement. On average, the radon concentration estimated from a 30-d test differed by about ±22 % from the value of a 1-y measurement. The difference reduced to about ±17 % for a 61-d test, ±14 % for a 91-d test and ±9 % for a half-year test. Health Canada's recommendation of a 3-month radon test performed during the heating season resulted in an estimated radon concentration, on average, ∼20 % higher than the value determined from a 1-y measurement. This ensures a conservative estimate of the annual average radon concentration, as there is some risk at any radon level. Therefore, to avoid an underestimation of radon exposure and to ensure appropriate levels of precision and accuracy are met, the results from this study suggest that a radon measurement duration of 3 months or longer during the heating season (from October through to April) is needed.