Evaluation of modelling of the TRUE-1 radially converging and dipole tests with conservative tracers. The Aespoe task force on modelling of groundwater flow and transport of solutes. Tasks 4C and 4D

The 'Aespoe task force on modelling of groundwater flow and transport of solutes' is a forum for the international organisations supporting the Aespoe HRL Project. The purpose of the Task Force is to interact in the area of conceptual and numerical modelling of groundwater flow and solute transport in fractured rock. Task 4 of the Aespoe Modelling Task Force consists of modelling exercises in support of the TRUE-1 tracer tests. In this report, the modelling work performed within Tasks 4C and 4D is evaluated, which comprised predictive modelling of the radially converging tracer tests and dipole tracer tests performed within the TRUE-1 tests using non-sorbing tracers. The tests were performed between packed off boreholes penetrating a water-conducting geological feature with a simple structure (Feature A). These tests are to a great extent preparatory steps for the subsequent tests with sorbing radioactive tracers. In Tasks 4E and 4F of the Aespoe Modelling Task Force predictive modelling of the sorbing tracer tests is performed. Eight modelling teams representing seven organisations have performed predictive modelling using different modelling approaches and models. The modelling groups were initially given data from the site characterisation and data on the experimental set-up of the tracer tests. Based on this information model predictions were performed of drawdown, tracer mass recovery and tracer breakthrough. The performed predictions shows that the concept of Feature A as a singular well-connected feature with limited connectivity to its surroundings is quite adequate for predictions of drawdown in boreholes and conservative tracer breakthrough. Reasonable estimates were obtained using relatively simple models. However, more elaborate models with calibration or conditioning of transmissivities and transport apertures are required for more accurate predictions. The general flow and transport processes are well understood, but the methodology to derive the necessary parameters for predictions needs development. The present understanding of the heterogeneity of the feature is limited and needs to be further evaluated. An issue that came in focus in this study is how to find a suitable relationship between the hydraulic fracture aperture derived from the hydrological tests and transport aperture derived from the tracer experiments. Furthermore, the use of extrapolation in assigning the boundary conditions used in the modelling gave rise to some uncertainty in the flow conditions in the outer parts of the feature. The performed tracer experiments and the modelling work have increased the understanding of the groundwater flow in Feature A and have contributed valuable knowledge and experience for the subsequent more elaborate experiments within TRUE-1, e.g. the tests with sorbing tracers. The continued experimental work within TRUE-1 and the continued modelling work within Task 4 shows that a sufficient understanding has been obtained of at least the upstream part of the structure