Naturally Fractured Basement Reservoir Characterization in a Mature Field

Characterizing the naturally fractured reservoir in a mature field is always a challenging task due to minimal subsurface data availability and the technology was not as advanced as nowadays. Therefore, this paper is proposed to provide an alternative solution to identify the presence of the fractures, classify them into the fractured quality related flowability, and distribute them vertically within the well interval and propose a lateral distribution method for reservoir modeling. This research was conducted based on a case study of basement fractured carbonate reservoir in Hungary. I used more than twenty development wells which mainly drilled during 1980-2000's. The fractures presence is simply identified by using gamma-ray and density logs. The relative movement of density log to the defined fractured baselines was directed to classify the fracture quality within three groups of macro-fracture, micro-fracture, and host-rock. These groups were validated by core data and the acoustic image log from the newest drilled wells. Furthermore, I implemented the self-organizing map (SOM) for distributing the fracture group to other wells which having limited subsurface data. Since the fracture classes were distributed along the well depth interval, then the well test (DST) results and production flow test data validated the flowability of them. As a result, the main flow contribution intervals of the fracture can be well-recognized. The macro-fracture consistently indicates the fracture class showing the main contribution of the liquid flowrate more than 10 m3/d along the perforated intervals. The rock properties of this class have porosity range around 1-2% with permeability dominantly more than 100 mD. In contrast, the host-rock class is defined as a protolith/non-fractured rock. The porosity and permeability are extremely low (tight rock). This class does not give any flow contribution due to the high content of the marl or clay, the absence of the fracture, or the fractures had been re-cemented by calcite or quartz minerals. Meanwhile, the micro-fracture denotes the group of rock with porosity range around 2-10% and permeability average between 1-10 mD. In general, the flowrate coming from this fracture class was lower than 10 m3/d of liquid during the flow-test. As a novelty, this proposed approach with the machine learning of SOM-clustering effectively assists us to recognize the fracture presence and its quality along the well-depth interval from the absence of the advanced technologies of image logs and production logging (PLT) measurement. Also, the defined fracture class here can take a role as a fracture facies or rock typing in terms of 3D reservoir modeling and distributed laterally based on fault-likelihood attribute and fault zone defined by distance-to-fault.