An Empirical probabilistic approach for constraining the uncertainty of long-term solute transport predictions in fractured rock using in situ tracer experiments

Uchida, Masahiro; Dershowitz, W.*; Lee, G.*; Shuttle, D.*

This paper addresses a residual uncertainty after the in-situ tracer experiment and try to demonstrate the usefulness of tracer experiments for safety assessment. There have been long debate on the usefulness of in-situ tracer experiment, because the difference in temporal scale, where in-situ tracer experiments are generally conducted for the period of a few days to several months with high velocity to recover the tracers, whereas safety assessment requires more than ten thousand years with much slower velocity. This paper addresses the issue by comparing breakthrough curves for safety assessment conditions (long-term, slow velocity) with and without calibration to in-situ tracer experiment called STT-1b test at the Aspo Hard Rock Laboratory in Sweden. Authors newly developed a conceptual model of internal structure of fracture, which explicitly addresses fault gouge, coating minerals, altered zone and intact rock, since especially fault gouge has high porosity and likely to affect the tracer migration behavior. The result was that calibrated model showed much tighter distribution of breakthrough curve statistics indicating the constraining power of in-situ tracer experiment.