The Reservoir Prediction Of Fuyu Formation In The Western Slope Of The Northern Xingshugang Oil Field

With the deepening exploration and development of the oil field, reservoir predictionaccuracy are increasingly demanded, the difficulty of the reservoir prediction is alsoincreasing. The well testing reports of Fuyu oil layer which is the main purpose formationshowed that it had good exploration prospects. But the thin oil layer also has a poor porosityand permeability and the low abundance, we need to further deepen the understanding ofregional geological features. Using3D seismic data, drilling data and geological data of thenorthern Xingshugang oil field, to compare the results of each inversion methods,high-resolution geostatistical inversion was proceeded on the basis of the sparse spike waveimpedance inversion to predict sand body distribution. By comparing the actual applicationwe found that P-impedance cannot distinct sand and shale, but electrical resistivity canseparate it well, it is the best parameter of this inversion; The sparse spike wave impedanceinversion had good correlation with seismic data and its results stability, but the lowerresolution of its results can not distinguish the thin layers, which can not meet therequirements of the prediction accuracy; The high-resolution geostatistical inversion introducegeological statistics information, considered reservoir parameter spatial structural andrandomness, took full advantage of the lateral continuity of seismic data and the verticalresolution of well data, improved the reservoir prediction accuracy, described the spatialdistribution of reservoirs in the study area, reflected the distribution of sandstone clearly,provided the basis for the further development.After corrected the logging data, calibrated purpose horizons by making syntheticseismograms, interpreted five seismic reflection horizons: the top surface of FⅠ1, FⅠ3,FⅠ6, FⅠ7and the undersurface of FⅠ7. According to the drilling data is abundant, wellnetwork density is large, uniform distribution of the actual well location, we usedInverTracePlusand StatMod methods from Jason software. Compared with testing data andlogging data, the results of this inversion can able to distinguish between sandstone andmudstone, the correlation of the reservoir position and well data is higher, the absolute errorand the relative error of thickness of the inversion layers and the practical drilling measuredsandstone thickness was less than1m and10%.