Fracture network modeling using petrophysical data, an approach based on geostatistical concepts

Fracture network modeling for hydrocarbon reservoirs plays an important role in dynamic simulation within the porous media. This modeling is traditionally performed using stochastic methods, which often are based on image logs run in very few wells. Accordingly, using such stochastic methods cannot satisfactorily provide reliable models for the fracture networks of the reservoir, which in turn result in unrealistic predictions of either fluid flow or production behaviour. In this paper, an algorithm is developed based on the initially and the secondary variables involving upscaled fracture attributes, including a new introduced variable named Cumulative Aperture (CA), and petrophysical data, respectively. Our introduced variable (CA) corresponds to the fracture volume which is defined as the multiplying of the fracture density by the fracture aperture. The introduced CA is correlated with the secondary porosity obtained from the conventional well logs. In addition, based on our experimental results, the other correlated variables are the aperture and the resistivity obtained from the image logs and the conventional logs, respectively. Following that, all correlated variables are jointly simulated using the turning bands co-simulation method at the target grid cells. The outcomes, at the end, are post-processed which led to a fracture network model. The resulting fracture network agrees with geological concepts and consequently is deemed reliable.