The Research And Application Of Sequential Indicator Simulation In Reservoir Sedimentary Facies Modeling

At present, most of the old oil field in our country has entered the developmentof high water content stage. The most important work at this stage is the explorationand development of remaining oil. With the oil well pattern coming more and moreintensive, using the information provided by the dense well pattern, we can set up thereservoir sand body model through the reservoir sedimentary facies modeling. Andthe reservoir sand body model plays an important role in the exploration anddevelopment of remaining oil, which is distributed in the sand body.Combined with the oil field project, I have done some research on the relativelypopular modeling software in the international, such as the schlumberger company’sPetrel, Norway Roxar software company’s RMS geologic modeling software, and themodeling software researched and developed independently in China-GPTModelreservoir geological modeling software, Direct digital reservoir characterization tools,and I have understood the general steps of sedimentary facies modeling of thesemodeling software and the algorithms which have been used commonly. Eventually Ichoose Petrel software as the tools of studying modeling. Before the study, I havecollected the relevant data of a field block, including the location coordinates, bushingelevation, well trajectory log, seismic interpretation fault data, seismic interpretationcontrolling surface, etc. And I have completed the data format conversion throughprogramming, and corrected some of the wrong data. Then I applied the GPTMapsoftware to layer the well logging curve, and got the hierarchical data and thebreakpoints data. Then according to the seismic interpretation of fault data, I have setup the Fault model in the Petrel using the Fault modeling module; In accordance withthe requirement of the grid accuracy, I have set up the3d structure model using thePillar gridding module; According to the hierarchical data, the breakpoints data andseismic interpretation of the control surface data, I have set up the Stratigraphic modelusing the Make horizons module. Finally, I have imported the maps of sedimentaryfacies made in GPTMap software to the Petrel software using the deterministicmodeling methods, and set up the reservoir sedimentary facies model. I have alsoused the stochastic modeling method-the sequential indicator simulation method toestablish reservoir sedimentary facies model. I have focused on the sequentialindicator simulation algorithm, and discussed the algorithm and the basic knowledgeof geostatistics including the calculation of variation function and the mathematicaltheory of the Kriging estimates in detail, and have calculated the variation function byusing MATLAB tools, programmed to realize the ordinary kriging estimation methodand sequential indicator simulation method.The main result of this study is that two sedimentary facies models are established. Compared with the model using deterministic modeling approach, thesequential indicator simulation model and the sedimentary facies model determinedby the geological experts are consistent. In the case of large quantity data, thegeological experts can manually modify the figure of sedimentary facies according tothe result of the sequential indicator simulation, and save a lot of time to improve thework efficiency.