Diffusive and convective transport of radon through cracks in the building understructure

The objective of this paper is to present a theoretical evaluation of soil and house related factors that may affect the transport of radon from the soil into houses. A two-dimensional mathematical model was used to simulate the diffusive and convective transport of radon into the house through idealized openings in the understructure. With the help of model predictions we tried to find out whether diffusion or convection predominates and under which circumstances. Radon transport through cracks in the house understructure is influenced mainly by the soil permeability, radon concentration at the soil-crack interface, the total area of cracks and the pressure difference across cracks. Because of its large range of variability, the soil permeability appears to have the greatest effect on the radon transport through cracks. At permeabilities below 1x10-12 m2 diffusive transport predominates and is almost invariable with the soil permeability. For permeabilities between 1x10-12 m2 and 1x10-11 m2, diffusion equals convection, and above 1x10-11 m2 convective transport predominates and is directly dependent on the soil permeability. The radon flux from perimeter cracks will be probably in all cases lower than that from central cracks. The difference between these fluxes will be influenced by the foundation type (slab or continuous footers) as well as by the house type (with or without basement) and by the location of the cracks in the house understructure. Generally, minimal differences will be observed in houses with basements and founded on continuous footers. Our findings could have several practical applications. The most important concerns our former hypothesis that a drainage layer beneath the slab could contribute to the dilution of the soil gas radon concentration and thus to reduction in the amount of radon entering the house. The recent placing houses on highly permeable drainage layers, which are not ventilated, should not be recommended in the future because such layers create conditions favoring appreciable convective transport. The simulated predictions have shown that our 2D model performs consistently with the 3D model presented by Loureiro in 1990