Hydrogeological characterisation and stochastic modelling of a hydraulically conductive fracture system affected by grouting: A case study of horizontal circular drifts

• A water-conductive fracture system is conceptualised to evaluate grouting results.
• Flow properties are assessed with a stochastic model for transmissivity distributions.
• A subdivision between larger conductors and the flow-constricted host rock is made.
• Suitable input data and models to evaluate different flow properties are proposed.

Knowledge of the flow characteristics of the hydraulically conductive fracture system is important when planning grouting measures for underground openings, although investigations and data analyses may, for different reasons, provide uncertain information. The primary focus of this paper is to establish a conceptual description of the flow properties of a hydraulically conductive fracture system and to use it to evaluate the suitability of hydrogeological investigation methods and grouting designs proposed for a horizontal repository design (KBS-3H). Data from the KBS-3H experimental site, located at the Äspö Hard Rock Laboratory in Sweden, were gathered and analysed to characterise the hydraulically conductive fracture system and to produce a stochastic model for predicting fracture transmissivity distributions. The rock mass at the site was affected by pre-grouting activities. It was, however, still possible to recognise two categories of inflow features: the main hydraulic conductors, which could be identified reasonably using short-duration borehole tests, and the less conductive part of the fracture system (host rock) where the flow configuration could not be captured by the conducted borehole tests. Characterising the flow properties of both categories is deemed important when developing efficient grouting designs. Consequently, careful consideration must be given during the design phase to the choice of data acquisition method and its implication for characterisation and modelling. The collection of suitable, strategic data is identified as a prerequisite for the reliable prediction of the flow properties of a fracture system.