Well test simulation through Discrete Fracture Network modelling in a fractured carbonate reservoir

Abstract

A Discrete Fracture Network (DFN) model was used to simulate the results of a production test carried out in a well drilled in a tight, fractured carbonate reservoir. Several static DFN models, depicting different geological scenarios, were built based on data from well logs, core analyses, PLT surveys and structural geology studies. Each of these models underwent a validation procedure, consisting of the simulation of the production test. The comparison between the simulated results and the actual data identified the scenario whose results most closely matched the actual well behaviour. In order to compensate for the lack of geological data, an iterative loop was performed between the static model and the dynamic simulation. Constraints-added flow simulations provided new information for use in modifying the DFN model, resulting in a step-by-step updating of the static model itself. Finally, a geologically sound model accurately matching the results of the production test was obtained. The final DFN model was used to calculate the equivalent petrophysical parameters that were transferred to the corresponding region of the full field dual-porosity fluid flow model.

A geometrically based method for predicting stress-induced fracture aperture and flow in discrete fracture networks

“…Kemeny and Post, 2003;Reif et al, 2011;Tavani et al, 2013). These studies provide size, spatial and orientation distributions for fracture systems that can be used for Discrete Fracture Network (DFN) analogue modeling in subsurface reservoirs (Cacas et al, 2001;Casciano et al, 2004;Wilson et al, 2011). Fracture aperture, on the other hand, cannot be accurately derived from outcropping rocks, as exhumation and weathering has changed the in-situ apertures in most outcrops.…”

Study of fracture-induced anisotropy from discrete fracture network simulation of well test responses

“…It has been shown (Bourdet et al , 1989 that the derivative of pressure (dp/dlogt) is more sensitive than the change of pressure itself to the properties of the reservoir. Indeed several authors (Ayoub et al 1983;O'Sullivan 1987, Wei et al 1998Wei 2000;Casciano et al 2004) have shown that the shape of the pressure derivative is sensitive to the properties of the fracture network through which diffusion is occurring. Different representations of the pressure disturbance geometry can be distinguished from the slope of the pressure derivative (Barker 1988), as can the time at which different boundaries are reached by the pressure disturbance.…”

“…The way in which pressure changes with time is known to be a function of the properties of the formation (Home 1995) and these data can also be used to infer properties of fracture networks. Although no unique solution exists for such inverse problems, it may be possible to discriminate between different equally plausible models (Wei et al 1998;Casciano et al 2004). Fracture networks are not random systems; they form in response to stresses in the brittle crust and depend on interactions with pre-existing structures.…”

Characterizing and Modelling Fractures and Karst in Carbonate Units - The Porocarste Project

“…First, dissolution along preferential zones creates anisotropy in karst systems, which are seen as heterogeneous and complex (Hopkins, 1999). Many current karst models, however, lack the capability to predict their complexities (Casciano et al, 2004). Second, quantification of porosity and permeability is missing from many karst studies, which tend to be descriptive or qualitatively oriented.…”