Recent Advances in Dynamic Modeling of Naturally Fractured Reservoirs

Abstract

Abstract Modeling of naturally fractured reservoirs (NFR) is a challenging problem whichoccurs frequently in reservoir engineering. Rock matrix, natural fractures, vugs, and also micro-fractures, constitute these NFR. Furthermore, it is wellknown that vugs effect on permeability is related to their connectivity. Likewise, in certain cases NFR exhibit fractures at different scales with poorfracture connectivity and disorderly spatial distribution of them. Thus, it isimportant to consider other alternatives for characterizing theseheterogeneities in a more efficient way than traditional approaches. One of thereasons which drive to find solutions to these challenges is because thesegeological formations contain more than 60% of the world's remaining oil. Therefore, in order to exploit in a more efficient way the remaininghydrocarbons reserves, it is frequently necessary to face greater reservoircharacterization and modeling challenges than those ones existing when thesereservoirs started their productive life. This is even more relevant when apressure maintenance project or an IOR/EOR process is planning to beimplemented. This work summarizes recent advances that address different dynamic modelingapproaches, taking into account NFR's features like the presence of vugs andfractures at multi scales with a non-uniform spatial distribution. Thus, it isdemonstrated the advantages of these continuum approaches to model anomalousbehavior that cannot be reproduced by the conventional Warren and Root dualporosity model in order to obtain good-quality history matches for thesesystems. Also, the characterization challenges that the proposed models pose tothe user are reviewed. Finally, some insights aimed to improve the proposedmodeling techniques are provided. Introduction Naturally fractured carbonate reservoirs have complex pore systems because theyare susceptible to diagenetic alteration including dissolution, dolomitizationand fracturing processes, which may enhance the reservoir quality or occludereservoir quality through cementation (Choquette and Pray 1970). In these fields stress varies significantly and it is expected that fracturingmay be more pronounced faults. Thus, it is expected that the degree offracturing and vugginess may be highest at the crest of an anticlinalstructure. Also, lithology drives fracture initiation and propagation, as wellas distribution of hydraulic properties, in fact in many cases; lithologicdistribution has a somewhat larger effect on fracture intensity than structuralposition (Nelson 2001). Fracture density is higher in limestones than indolomites, but dolomites may contribute more to production than limestones(Massonnat et. al. 2002), especially when limestones are dolomitized, theporosity may become vuggy or intercrystalline. In a fracture system, scaling relationships has a major role about how the flownetwork within the reservoir is organized (Massonnat et. al. 2002). Thus, thereare two categories, the large-scale fractures or faults at seismic andsub-seismic scale, and the small-scale fractures observed at wellbore level. The fracture population of the first category is mainly controlled by thefacies as well as the reservoir structure, whereas the facies and the proximityto faults are the factors which control the second category (Bourbiaux et. al.2005). Likewise, the above both groups, seismic and sub-seismic faults as wellas fracture corridors and small-scale fractures, may show fractal behavior (deKok et. al. 2009).