ASPECTS OF RADON TRANSPORT IN CONCRETE

I. Cozmuta and E.R. van der Graaf

In the multi-phase time-dependent equation describing radon transport through a porous medium [1] parameters that are related to the microstructural properties of that medium are involved, the most frequently analyzed being the porosity and the moisture content. Their values can be individually determined using various experimental methods or by performing computer simulations based on some fundamental model. If the porous medium is concrete, the above mentioned parameters are also not constant in time because of the ongoing process of cement hydration. Hydration dynamically couples the development of the pore structure with the moisture transport, at any age the amount of free water determining the rate of hydration.

Two convincing examples can be given. First, radon release rates were previously shown to change with the age of concrete [2]. Considerable variations (a factor of 1.5) were registered in the first year after pouring, after this period the radon release rate was decreased to 0.3-0.6 of the maximum at the end of the observed period (8 years). Secondly, radon release rates were found also to be very sensitive to the moisture content of the concrete [2,3] consequently also being related to the ambient air humidity. This last fact is especially important for concrete durability as it can contribute to its various deterioration mechanisms: transport of contaminants, freezing and thawing cycles etc.

A detailed study of the dependency of radon release rates on the moisture content of a concrete cube performed at the NGD-KVI, Groningen, the Netherlands resulted the profile pictured in Fig. 1. The radon release rates for fresh and six month old concrete (storing conditions at constant relative humidity by wrapping in foil) were respectively measured [4], the obtained values (squares) indicating an increase with a factor of 1.7 (0.34 % loss of moisture from the initial mass) due to aging. Afterwards, the cube was immersed in water (for approximately one year) until completely wetted (it reached a constant mass) when its radon release rate was measured. The same cube was further dried in steps at 2500C (until totally dry), for each step the release rate being remeasured (circles). The radon release rate measurements corresponding to the extreme points were repeated (diamonds). A second set of measurements of the radon release rates of the same concrete cube as function of moisture was then recorded (triangle symbols). For both sets of data points the solid curves in Fig. 1 represent a fitted third order polynomial.

If one follows the path of radon production and transport, when the space between the emanating grains is entirely air-filled, the probability for the emanated radon to end up in one of the neighboring grains is very high such as only a small amount is available for transport through the pores space. When the pores are saturated with water, almost all emanated radon will be transported through this phase, however a factor of 104 slower (the difference between the radon diffusion coefficient in air and in water). Thus, an optimum thickness of the water layer covering the grains exists corresponding to a maximum radon release rate. The profiles indicated in Fig. 1 are a proof of these two competitive effects related to production and transport. The second profile being shifted is an indication of the alteration of the concrete structure during drying at high temperature (usually, water absorption techniques employed to determine the volume of the pore space in concrete require drying at 1050C). Future experimental investigations aim at repeating this type of measurements when drying of concrete occurs at a lower temperature (1050C).

A computer program that would be able to reproduce the experimental data aims at coupling a numerical modeling of concrete microstructure (linking the physical mechanisms of hydration, micro-structure development and moisture transport in real time) with the extensively validated NGD computational program [1] that models the radon concentrations (release rates) in real time. The assembling of the two programs will be accomplished at the MSC at Caltech.

[1] W.H. van der Spoel, Radon transport in sand, a laboratory study, Ph.D. thesis, Technical University Eindhoven, The Netherlands, 1998.

[2] L.M.M. Roelofs and L.C. Scholten, The effect of aging, humidity and fly-ash additive on the radon exhalation from concrete, Health Physics, 67 (3), pp. 266-271, 1994.

[3] E.Stranden, A.K. Kolstad and B.Lind, Radon exhalation: moisture and temperature dependence; Health Physics, 47 (5), pp. 480-484, 1984.

[4] E.R. van der Graaf, R.J. de Meijer and W.H. van der Spoel, KVI Annual Report, p.68, 1995