Differences in the Upscaling Procedure for Compositional Reservoir Simulations of Immiscible and Miscible Gas Flooding

Compositional reservoir simulation is essential to represent the complex interactions associated with gas flooding processes. Generally, an improved description of such small-scale phenomena requires the use of very detailed reservoir models, which impact the computational cost. We provide a practical and general upscaling procedure to guide a robust selection of the upscaling approaches considering the nature and limitations of each reservoir model, exploring the differences between the upscaling of immiscible and miscible gas injection problems. We highlight the different challenges to achieve improved upscaled models for immiscible and miscible gas displacement conditions with a stepwise workflow. We first identify the need for a special permeability upscaling technique to improve the representation of the main reservoir heterogeneities and sub-grid features, smoothed during the upscaling process. Then, we verify if the use of pseudo-functions is necessary to correct the multiphase flow dynamic behavior. At this stage, different pseudoization approaches are recommended according to the miscibility conditions of the problem. This study evaluates highly heterogeneous reservoir models submitted to immiscible and miscible gas flooding. The fine models represent a small part of a reservoir with a highly refined set of grid-block cells, with 5 × 5 cm2 area. The upscaled coarse models present grid-block cells of 8 × 10 m2 area, which is compatible with a refined geological model in reservoir engineering studies. This process results in a challenging upscaling ratio of 32 000. We show a consistent procedure to achieve reliable results with the coarse-scale model under the different miscibility conditions. For immiscible displacement situations, accurate results can be obtained with the coarse models after a proper permeability upscaling procedure and the use of pseudo-relative permeability curves to improve the dynamic responses. Miscible displacements, however, requires a specific treatment of the fluid modeling process to overcome the limitations arising from the thermodynamic equilibrium assumption. For all the situations, the workflow can lead to a robust choice of techniques to satisfactorily improve the coarse-scale simulation results. Our approach works on two fronts. (1) We apply a dual-porosity/dual-permeability upscaling process, developed by Rios et al. (2020a), to enable the representation of sub-grid heterogeneities in the coarse-scale model, providing consistent improvements on the upscaling results. (2) We generate specific pseudo-functions according to the miscibility conditions of the gas flooding process. We developed a stepwise procedure to deal with the upscaling problems consistently and to enable a better understanding of the coarsening process.

A New Simulation Layer Optimization and Permeability Upscaling Method for Preserving Critical Reservoir Heterogeneity

Running a fine grid model with 107 - 109 of cells is possible using a supercomputer with 103 - 106 of CPUs but may not be always cost-effective. The most cost-effective way is to use a coarse grid model that is much smaller but with static/dynamic profiles very close to the fine grid model. This paper proposes a new layer optimization and upscaling method with the aim for creating a consistent coarse grid model. Unlike the industry's existing layer optimization and upscaling methods, the proposed method performs layer optimization and upscaling fully integrated with the Lorenz coefficient and curves (LCC). Coarse grid layers and their permeabilities are created by minimizing the difference between fine and coarse grid LCCs. The process consists of static and dynamic optimizations. The former is measured by LCC while the latter by pressure, GOR, and water-cut. A new LCC-based permeability upscaling method is developed to preserve the fine grid multiphase flow behaviors. A satisfactory coarse grid model is achieved when both static and dynamic criteria are met. The proposed method has been successfully applied to a giant carbonate oil field in the Caspian Sea that consists of a matrix dominated platform and a fracture/karst dominated rim. Due to the field's complex geology and high H2S content (15%), a dual porosity, dual permeability compositional model has been created to model compositional sour crude flow within and between the matrix and fracture/karst features. The reservoir drive mechanisms are fluid expansion, miscible gas injection and aquifer drive. The reservoir is undersaturated and has an abnormally high initial reservoir pressure. The fine-grid static model contains 104 million cells (370×225×625×2) and the optimized upscaled coarse-grid dynamic model has 8.3 million cells (370×225×50×2). The upscaled model can be run efficiently on the company's existing HPC infrastructure with a maximum of 64 CPUs. Excellent matches of the Lorenz coefficient maps for reservoir total/zones and Lorenz curves at all wells between the fine and coarse grid models have been achieved. Matches on the dynamic variables, e.g., pressure, gas breakthrough time, and GOR growth, in all producers are within the defined acceptable tolerances. The high quality of the static and dynamic matches between the coarse- and fine-grid models confirms that the reservoir properties of the coarse-grid model is very close to the fine-grid model and can be used a base model for history matching and uncertainty analysis.

The Application of Machine Learning Algorithm in Relative Permeability Upscaling for Oil-Water System

Traditional two-phase relative permeability upscaling requires the fine-scale two-phase flow simulation over the target regions/blocks. It can be very computationally expensive especially for cases with multiple (hundreds of) geological realizations (as commonly used in subsurface uncertainty quantification or optimization). In this paper, we develop a machine learning assisted relative permeability upscaling procedure, in which the full numerical upscaling is performed for only a portion of the coarse blocks, while the upscaled functions for the rest of the coarse blocks are calculated by the machine learning algorithm. The upscaling procedure was tested for generic (left to right) flow problems using 2D models for scenarios involving multiple realizations. Numerical results have shown that the coarse-scale simulation results using the newly developed machine learning assisted upscaling procedure are of similar accuracy to the coarse results using full numerical upscaling. Because the fine-scale numerical simulation is only performed for a small fraction of the model, significant speedup is achieved.