Long-term gas permeability properties of concrete in wet repository conditions

Abstract The permeability properties of concrete have been of growing concern, especially over the last decade, as interest in the longevity, or service life estimation, and high technology applications of concrete have continued to increase. The permeability of porous media is by nature a highly complex phenomenon. The scatter, biases, and ranges of permeability test results are great and often inexplicable in simple terms. This is especially true of the gas permeability of porous solids in general, and of the complex, history-dependent material that is concrete in particular. The generation of gas (hydrogen) due to steel corrosion in moist or wet repository conditions has created a special interest in gas transfer and gas permeability problems. A survey of the literature shows that high-class, large-scale research on the gas permeability of concrete has been largely lacking. Vital factors in real applications, such as aging, history-dependence, moisture, and threshold pressures have received little attention. With the usual gas permeability testing equipment, achieving large pressure difference ranges has been virtually impossible. Roughly speaking, with a small pressure difference scale the pressure difference ( Δp ) range in the function k — Δp (where k = permeability) is practically non-existent and does not reveal the true nature of gas flow. The influence of the pressure gradient on the flow process has not been discussed. At the Concrete and Silicate Laboratory of the Technical Research Centre of Finland, we used a special gas-testing apparatus between 1986 and 1989 and obtained a relatively large number of test results within a pressure difference range of 0–100 bar (atm). This seems to have exposed the real features of the gas permeability behavior of concrete, indicating a dependence on pressure differences and aging, and the existence of ‘surprising’ threshold pressures. The highest pressure difference gradients applied (probably well above anything encountered in practice) were up to 100 bar over a 5 cm thick specimen, or 20 atm/cm (2 MPa/cm). The effect of measurement time, duration of the applied pressure difference, thickness of specimens and pressure gradient used at testing may also be relevant factors in practice. As wet conditions and the long-term behavior of concrete play vital roles in the storage of nuclear waste, this paper focuses particularly on these issues. In continuously wet conditions concrete structures (silos etc.) do not suffer significantly from drying cracks. This report therefore deals chiefly with solid, non-cracked concrete. Crack flow is another complicated subject, and is discussed briefly as a sideline. Two nomograms are introduced illustrating air flow through porous media and cracks.