In this study, discrete fracture network (DFN) modelling was performed for Triassic–Jurassic analogue reservoir units of the NW Lurestan region, Iran. The modelling was elaborated following a multi-scale statistical sampling of the fracture systems characterising the analysed succession. The multi-scale approach was performed at two different observation scales. At the macro-scale, a digital outcrop analysis was carried out by means of a digital line-drawing based on camera-acquired images, focussing on the distribution of major throughgoing fractures; at the meso-scale, the scan line method was applied to investigate the background fractures of the examined formations. The gathered data were statistically analysed in order to estimate the laws governing the statistical distribution of some key fracture set attributes, namely, spacing, aperture, and height. The collected dataset was used for the DFN modelling, allowing the evaluation of the relative connectivity of the fracture systems and, therefore, defining the architecture and the geometries within the fracture network. The performed fracture modelling, confirmed, once again, the crucial impact that large-scale throughgoing fractures have on the decompartmentalization of a reservoir and on the related fluid flow migration processes. The derived petrophysical properties distribution showed in the models, defined the Kurra Chine Fm. and, especially, the Sehkaniyan Fm. as good-quality reservoir units, whereas the Sarki Fm was considered a poor-quality reservoir unit.
An optimization approach for large scale simulations of discrete fracture network flows
