A study of biases, assumptions and practical considerations for the use of discrete fracture networks in geomechanical practice

Abstract

The use of Discrete Fracture Networks (DFNs) is becoming increasingly common in geomechanical practice in addition to their continuing role in hydrogeology. These models can serve as useful tools for estimating interconnectedness of fractures, leading to estimates of probable block sizes and shapes for a set of input parameters. However, the development of these models is reliant on assumptions made about collected field data and while constructing the model themselves. The implications of these biases and assumptions are not well documented. This work investigates the variables involved in building a Discrete Fracture Network model in order to provide insight into the decisions and assumptions made during the modeling process. Select assumptions required within the FracMan DFN software pertaining to model selection and construction are evaluated; biases and assumptions relating to field data and how it is collected that may impact the development of DFN input parameters are investigated and limits of the effects of these models on block sizes are determined. The parameters determined to be critical in determining the overall geometry of the fracture network are ranked according to their relative importance in DFN modelling and according to the relative accuracy of each parameter.