Predictive modelling and evaluation of the Phase C tracer tests using the Posiva Streamtube approach: Äspö Hard Rock Laboratory, TRUE Block Scale project, Posiva/VTT Processes modelling team

Predictive modelling and evaluation of the Phase C tracer tests performed as part of the TRUE Block Scale project were performed using the Posiva Streamtube approach. The predictive modelling was based on the concept of a single flow channel that connects the source and the sink sections. Properties of the flow channels were estimated using the results of the Phase B tracer tests. Matrix diffusion and retardation properties of the different tracers were imported from results of the corresponding evaluation of the TRUE-1 tracer tests, the latter carried out in an interpreted single fracture. The TRUE-1 tracer tests were interpreted to provide the matrix diffusion and sorption properties along single fracture flow paths. The evaluated matrix diffusion and sorption properties were applied directly to predict the TRUE Block Scale tracer tests performed in a network of fractures. The import is based on the assumption that the characteristics of the immobile pore spaces at the TRUE-1 and TRUE Block Scale sites are similar. A comparison with in situ test results show that the predicted breakthrough curves are reasonably accurate for the non-sorbing tracers. For the sorbing tracers the discrepancy between the measured and predicted breakthrough curves increases with increasing sorptivity of the tracer. Following the predictive modelling, the breakthrough curves were modelled anew relative to the measured breakthrough curves to identify the processes causing the observed tracer retention. The most consistent understanding of the Phase C tracer tests was provided by a model that is based on surface sorption on the fracture surfaces combined with matrix diffusion into fault gouge. The least sorbing tracers also show retardation caused by diffusion into stagnant zones of the flow field.