Radon concentration in soil gas: a comparison of the variability resulting from different methods, spatial...

From the end of 1996 through March 1999, the spatial and temporal variability of the soil 222Rn concentration was investigated at a 20 m x 20 m test field with porous soil in 0.5 m and 1.0 m depth at nine positions each and at 1 m x 1 m plots at four positions each. For this, soil gas was collected weekly into evacuated scintillation cells and was analysed subsequently for radon activity. In the 20 m x 20 m field the spatial variability was characterized by coefficients of variation (C.V.) of 26% at 0.5 m, and 13% at 1.0 m depth. Within the 1 m x 1 m plots the C.V. values were 4% and 2%, i.e. within the uncertainty of the method. Time series analysis (TSA) of the soil radon data shows seasonal variations with maximum concentrations in the winter months. Radon concentrations ranged from 6 to 50 kBq m(-3) in 0.5 m depth, and from 8 to 34 kBq m(-3) in 1.0 m depth. Mostly, the concentrations were higher in 0.5 depth than in 1.0 m depth. However, seasonal variation of the 0.5 m to the 1.0 m concentration ratio has been verified by TSA. To test the variability resulting from different methods, additional procedures and instruments were investigated at the 20 m x 20 m field and at a second test field with a different soil type. Soil gas sampling into evacuated scintillation cells was selected as the reference procedure. Soil radon concentrations obtained with the different sampling procedures and detection methods at the 20 m x 20 m field essentially agreed within the limits of uncertainty of the methods tested. At the second test field, i.e. in a largely impermeable soil, deviations up to a factor of two related to the reference procedure were observed.