The Application Of Seismic Attributes With Varying Of Frequency To Reservoir Prediction

As the oil and gas exploration is more and more difficult, the complexity ofexploration is increased. Using seismic attributes to extract the hidden informationfrom seismic data can improve the accuracy of reservoir prediction, and has nowbecome one of the effective means of reservoir prediction. With decades ofdevelopment, the application of seismic attribute technology has formed certain processand approaches. But due to the complex relationship between seismic attributes andreservoir parameters, it is difficult to complete reservoir prediction and explanation taskby using the seismic attributes alone. It is necessary to seismic attributes analysis withthe comprehensive interpretation of logging and geological aspects of information.Therefore, seismic attributes which used to predict reservoir lithology and physicalproperties is still need to continue to develop and research.The seismic signal is non-stationary signal. In order to study the change offrequency information of seismic signal usually we adopt the method oftime-frequency analysis. By time-frequency analysis, we can connect theone-dimensional time domain with two-dimensional time and frequency domain, andobtain the information which is not display on seismic section. Using time-frequencyanalysis method, we can further study the connection between time-frequencycharacteristics and stratigraphic structure and oil and gas bearing. Therefore,time-frequency analysis technology is also one of the effective means of predicting oiland gas in the reservoir. Frequency is a kind of inherent attribute. The differentgeologic bodies have different reflection wave frequencies. It is very helpful to analysisand research the target of geological body by extracting the reflection wave spectruminformation. Spectral decomposition technique is to separate different frequencycomponents of the seismic signal. It can improve the credibility of reservoir predictionby analyzing the characteristic of amplitude of each frequency. In this article, we mainly discuss on the application of seismic attributes withvarying of frequency to reservoir prediction, conducted on multi-attribute with varyingof frequency analysis and prediction research. The data volume of different frequenciesis obtained by using spectral decomposition technology. Using different attributesextraction methods, we analysis attribute and do the prediction in view of the objectivelayer in the study area, and sum up a set of seismic attributes with varying offrequency analysis process. The results of research are applied to the PermianLucaogou Formation in Jimusaer Sag Junggar Basin sweet reservoir prediction,through the application of actual data, we achieve better results, and provide technicalsupport for reservoir prediction in the study area.