Natural Fracture Characterization and Modeling of the Permo–carboniferous Carbonate Reservoir of the Usinsk Oil Field

In this work, the natural fracture characterization and modeling was performed in the highly fractured Permo-Carboniferous Carbonate Reservoir of the Usinsk Oil Field located in the Timano - Pechora region. The main conclusions of this study are:  Fractures are fold and fault related. They are organized in three main sets N150, N120 and N20. The fractures are mainly organized in clusters along the N150 set. The global fracture densities are moderate to low but can show high values in the clusters;  The main set N150, is mainly concentrated in the hinge zone of the anticline. The minor two other sets N120 and N20 are not clustered and represent oblique fractures sets to the fold axis;  The porosity seems to no have clear impact on the fractured interval as both vuggy and low porosity facies can appear fractured or not. Diffuse fractures related to lithology were not been characterized;  Curvature analysis showed that there is good fit between fracture orientation seen at wells and seismic faults and lineaments orientation. This allows to think that fault can have also an impact on the fracture present on the reservoir;  A fracture model was built in the reservoir grid available in the SW sector of the reservoir, containing the 1GS well;  The model generated in FracaFlow contains fold-related fractures along the three main sets, associated to the structural position. It contains also fault-related fractured associated to the damaged zone of the faults;  The dynamic data analysis performed in the SW sector confirmed the presence of fracture in the reservoir;  All the dynamic data analyses were combined to prepare a “Fracture synthesis map”. The map was used to pick small-scale deterministic fracture lineaments to be used in fracture modelling;  To cover uncertainty in the fracture modelling due to limited data, two fracture models were up-scaled and equivalent fracture parameters were generated. It was recommended to test the models and select the one which gives better result during the history match.