Uncertainty analysis of the weighted equivalent lung dose per unit exposure to radon progeny in the home

A parameter uncertainty analysis has been performed to derive the probability distribution of the weighted equivalent dose to lung for an adult (w(lung) H(lung)) per unit exposure to radon progeny in the home. The analysis was performed using the ICRP Publication 66 human respiratory tract model (HRTM) with tissue weighting factor for the lung, w(lung) = 0.12 and the radiation weighting factor for alpha particles, wR = 20. It is assumed that the HRTM is a realistic representation of the physical and biological processes, and that the parameter values are uncertain. The parameter probability distributions used in the analysis were based on a combination of experimental results and expert judgement from several prominent European scientists. The assignment of the probability distributions describing the uncertainty in the values of the assigned fractions (ABB, Abb, AAI) of the tissue weighting factor proved difficult in practice due to lack of quantitative data. Because of this several distributions were considered. The results of the analysis give a mean value of w(lung) H(lung) per unit exposure to radon progeny in the home of 15 mSv per working level month (WLM) for a population. For a given radon gas concentration, the mean value of w(lung) H(lung) per unit exposure is 13 mSv per 200 Bq.m(-3).y of 222Rn. Parameters characterising the distributions of w(lung) H(lung) per unit exposure are given. If the ICRP weighting factors are fixed at their default values (ABB, Abb, AAI = 0.333, 0.333, 0.333; w(lung) = 0.12; and wr = 20) then on the basis of this uncertainty analysis it is extremely unlikely (P approximately 0.0007) that a value of Hw/Pp for exposure in the home is as low as 4 mSv per WLM, the value determined with the epidemiological approach. Even when the uncertainties in the ABB, Abb, AAI, values are included then this probability is predicted to be between 0.01 to 0.08 depending upon the distribution assumed for describing the uncertainties in the ABB, Abb, AAI, values. Thus, it is concluded that the uncertainties in the HRTM parameters considered in this study cannot totally account for the discrepancy between the dosimetric and epidemiological approaches.