The Application Of Reservoir Prediction With Cluster Analysis Of Seismic Attribute

The seismic data contains rich information of strata lithology, fluid characteristics andgeological structure, and it is high practical value to effectively extract reservoir propertiesand lithology information, which is an important technical means for reservoir prediction.Seismic attributes is that seismic data after a particular mathematical transformation ofseismic wave dynamics, kinematics, geometry, morphology, and the statistical characteristicsof the measure. With the development of seismic attribute, the number of seismic attributesbecomes more and more. Faced with a different work area, a different reservoir and predictobject, the choice of seismic attributes are not identical. The seismic attribute analysis basedon cluster analysis is effective ways to target reservoir prediction, which can correct reflectnon-linear relationship between the seismic data and geological information, in order toestablish relationship between the seismic data and reservoir distribution in the study area.On the basis of the construction interpretation analysis of Ng3-3, extracted along thelayer40kinds of seismic attributes. Combination of geological data logging, removed toreflect the insensitive seismic attributes of effective reservoir characteristics, optimizing asensitive seismic attributes to reservoir characteristics; optimize seismic attributes base on thecorrelation between the seismic attributes, get ride a strong linear correlation attributes, andoptimization of relatively independent of the attributes. The optimization of the aboveattributes, favorable reservoir characteristics of the study area is sensitive to the five attribute:the average energy attributes, the attribute of the mean instantaneous frequency, energyhalf-time attributes, the attributes of effective bandwidth, the correlation length attribute.Cluster analysis based on the five seismic attributes, Fuzzy C Means cluster analysis onthe five attributes of this broad category analysis of the data, combined with well data toidentify a class of data to reflect the effective reservoir, and the projection to the Ng3-3timemap, and got the range of reservoir prediction. Combination of geology and logging data, to analyze the horizontal distribution of the law of the reservoir range, in the study area, thedevelopment of two large northeast-trending faults, and layer in a slope belt, the layerdevelopment some faulted nose structure, so the region mainly fault oil and gas reservoirs.Based on the above analysis, we can get the favorable exploration area of Ng3-3reservoirprediction.