The Research And Implementation Of 3D Reservoir Modeling Technology Based On Self-Avoid Random Walks

Fluvial reservoir plays an important role in petroleum industry of our country. However, because of its fast changes in horizontal and the serious heterogeneity, it is still difficult to describe it in detail. Based on time constrained modeling, applying a random walk method for three-dimensional fluvial reservoir modeling could make the best of its random feature and forecast characteristics to confirm river channel morphology in 3D space and the distribution principle of channel sand, especially in the case of limited geological data. Consequently, it is significant for expanding the ideas of geologists and guiding the exploration and development of oil and gas field.This dissertation put emphases on the algorithm, mathematic principle of self-avoid random walks and its application mode and effect in fluvial reservoir modeling. Meanwhile, the influence of lithofacies of wells on random walk fluvial reservoir modeling, for example, the attracting interactions between sand facies wells, the excluding interactions between mud facies wells and the thickness of sand in small layers of single-well, was investigated further. On the basis of the research above, the 3D fluvial reservoir modeling combined with the geological conditions, such as the principal direction constraint, the geological constraint and the pseudo-well constraint.The structural ideas were used in designing 3D Reservoir Modeling System, and this system finally completed by using of C++language.Based on the actual data of fault block Ken 71 of Ken Xi Oilfield, a random walk modeling was applied for checking. The corresponding results suggested that:all of the three-dimensional river channels come from the results of fluvial reservoir modeling pass by the sand facies well locations and avoid the mud facies well locations by using the optimized algorithm of self-avoid random walks. Furthermore, it is clear that the changes of the thickness of river channels are in accordance with the sandstone thickness. Analysis on numerous simulation results can know that the well's sandstone thickness louder, the probability the river passes is heavier. Combined with the restriction of the principal direction constraint, the geological constraint and the pseudo-well constraint, that the basin of simulations with the river deposition of layer picture tally better.