MULTIDIMENSIONAL FLOW MODELING USING DISCREATE FRACTURE NETWORK (DFN) IN FRACTURED ROCKS

Abstract

Fluid flows through complex network of fractures in a fractured rock mass. Equivalent porosity model approach is conventionally used for flow simulation through such types of porous media which considers fractured rock as a single porosity system where every fracture contributes to flow.However, fracture connectivity which affects flow pathway significantly is ignored in such types of simulation. The other modeling approach like dual porosity/permeability considers rock masses as mobile-immobile media considering fracture network as a highly permeable media and surrounding porous matrix as an immobile domain. But these methods require more detailed input data information for their development. Also the entire study domain is considered as a whole, including the soil matrix and fractures, for such types of models. Thus in this paper, DFN approach is followed to simulate the water flow through a fractured rock mass by considering each fractures individually and has flexibility of modeling using deterministic and stochastic approaches. As it is not feasible to map fracture networks in rock masses because accurate field measurement of single fracture is difficult, deterministic approach is rarely applicable in field. On the other hand, stochastic method uses data collected from rock outcrops, drill cores, borehole imaging, satellite imaging, geophysical surveys. Therefore, stochastic modeling of discrete fracture network is adopted here for modeling fracture locations, geometries and their orientation by respective probability distributions. Data required for discrete fracture network modeling are maximum and minimum fracture length, fracture orientation, and total number of fractures. FraNEP software which evaluates fracture length by applying power law equations using cumulative distribution functionand plots fractures orientation is used in this study. The simulator also classifies fractures into different sets according to their orientation which are used in further modeling stages. In addition, it also provides information like fracture density, intensity and their mean length. After creating fracture network model using DFN, all fractures are converted into pipe model using the Polygon method. Flow simulation is then performed by applying finite difference method for obtaining output in form of variation of pressure head across the connected fractures in multi-dimensional domain. The developed modeling approach is applied well to Jabal Akhtar dome in Oman mountains. Findings of this study are of direct use in predicting accurate flow and solute transport through fractured porous media at field scale level.