Fine Reservoir Geological Model Of Aggregate Coarse Degree Studies

The determination of rational upscaling degree of reservoir fine geologic model is very important in dealing with the reliability of numerical simulation and saving the time of computation simulations. Because the research work on the rational upscaling degree of reservoir fine geologic model is slow, and there is no systematic procedure of the determination of rational upscaling degree, we carried out the following research works:(1) We constructed a fine geologic model, and carried out the numerical simulation work based on the fine geologic model, and treat the results of simulation as a reference for comparison.(2) According to upscaling the fine geologic model method with different upscaling method, we finished the numerical simulation of different models that generated by different upscaling methods. Then, on the basis of analyzing the error of simulation results, the optimal coarsening methods are chosen.(3) Upscaling the fine geologic model with the optimal upscaling methods is performed in the paper, and then we generated the different models in different upscaling degree, and simulated the parameters for different coarsening model.(4) We have also analyzied the curve of Average error between the number of grid to determine the knee point and the corresponding grid size and to find a reasonable upscaling degree.In view of the above procedure, we obtained these conclusions as follows:(1) The determination of rational upscaling degree is effected by several factors, namely the plane upscaling degree, the vertical upscaling degree and the composition of fluid phase in simulation. The vertical upscaling degree and the composition of fluid phase are the critical effect factors on the reasonable upscaling degree.(2) The results show that the harmonic-arithmetic method for calculating permeability is optimal method, and the RMS algorithm is the worst method for upscaling algorithm with the maximum error.(3) The relations curve of average errors between the results of upscaling-model simulation and fine-model simulation and the grid numbers are stepped-like change, existing error knee point and reasonable upscaling extent.(4) The deepening of vertical coarsening degree will weaken the ladder-like feature of average error curve, and this phenomenon will be more obvious in three-phase flow to turn to the single knee point. Thus, vertical upscaling can not be more than three times of it or the error will exceed the permissible tolerance range for engineering.