Integrated approach to improve numerical and geostatistical performance on a naturally fractured carbonate reservoir

The numerical simulation of carbonate reservoirs, while keeping the heterogeneous behavior in a reasonable development time, is a challenge. This study introduces a new method to improve simulation of complex carbonate reservoirs, balancing speed and accuracy by using multiple stochastic techniques and hierarchical upscaling. The methodology has five main steps: (1) develop the geological model, (2) hierarchical upscaling and numerical validation, (3) define the probabilistic workflow, (4) Discretized Latin Hypercube combined with geostatistical realizations, and (5) computational cost. The methodology is applied to a Brazilian pre-salt field. The time consumption for the numerical simulation and geomodelling of the proposed workflow was compared against conventional procedures. Notably, the hierarchical upscaling approach allowed the use of a reference model for the dynamic matching procedure. Innovative elements include implicit modeling of small-scale fractures within the matrix domain using analytical averaging techniques, the incorporation of pseudo-functions, and the direct integration of the discrete fracture network into the simulation grid. These advancements result in a remarkable reduction of 20% in simulation time and a 60% decrease in the time required to generate new geostatistical images. The innovative approach presented herein promises to address critical challenges in carbonate reservoir simulation, advancing the field's understanding and practice.

Automated summary

This study introduces a new method that improves the simulation of complex carbonate reservoirs by balancing speed and accuracy using multiple stochastic techniques and hierarchical upscaling. The method is applied to a Brazilian pre-salt field and achieves significant results.