Basic oxygen furnace (BOF) slag is considered as a potential alternative construction material and is used here on an experimental plot to accurately quantify the risk of pollutant release. Since pollutant release depends on flow, this initially requires characterizing BOF slag hydraulic properties. These were monitored and estimated at plot scale by carrying out water infiltration experiments and inverse numerical modeling. Monitoring the plot showed that the BOF slag studied crusted at the surface as a result of weathering processes. Numerical inversion proved that the crusted material differed from the unaltered slag in terms of water retention and hydraulic conductivity functions. Although all the data pointed to a decrease in saturated hydraulic conductivity with crusting, the trends depended on the infiltration devices used for the capillary length (tension disc vs. Beerkan). Scanning electron microscope (SEM) microanalysis of laboratory weathering cells and lysimeter measurements were monitored in parallel to study the microstructure more precisely and highlighted a reduction of porosity by clogging. On the basis of SEM observations, two conceptual models of pore reduction, based on two different pore clogging hypotheses, were applied to predict hydraulic properties. This step demonstrated that the effect on water retention and hydraulic conductivity strongly depended on the way precipitated phases form and coat grains and could explain the evolution of the transport properties observed. This study contributes to knowledge on the hydraulic properties of BOF slag and their evolution due to carbonation
