Technical Report NTB 91-33

This report describes the modelling of break-through curves from a series of two­tracer dynamic infiltration experiments, which are intended to complement larger scale experiments at the Nagra Grimsel Test Site. The tracers are ⁸²Br, which is expected to be non-sorbing, and ²⁴Na, which is weakly sorbing. The ²⁴Na concentration is well below the natural Na concentration in the infiltration fluid, so that sorption on the rock is governed by isotopic exchange, exhibiting a linear isotherm. The rock specimens are sub-samples (cores) of granodiorite from the Grimsel Test Site, each containing a distinct shear zone. Best-fits to the break-through curves using single-porosity and dual-porosity transport models are compared and several physical parameters are extracted. It is shown that the dual-porosity model is required in order to reproduce the tailing part of the break-through curves for the non-sorbing tracer. The single-porosity model is sufficient to reproduce the break-through curves for the sorbing tracer within the estimated experimental errors. Extracted K_d values are shown to agree well with a field rock-water interaction experiment and in situ migration experiments. Static, laboratory batch-sorption experiments give a larger K_d, but this difference could be explained by the larger surface area available for sorption in the artificially crushed samples used in the laboratory and by a slightly different water chemistry.