Delineation Of Seismic Attribute Anomalies Based On Fractal Theory

In the oil and gas exploration, seismic method is an important way to identify the deep underground structure. The seismic attribute extraction is an essential step of data processing, and changing property value corresponding to certain changing physical parameters, which reflecting the stratum structure of different positions and different depths. However, in the real seismic exploration, relationship between seismic attributes and the predicted objects is always very complicated. When we use seismic attributes to predict reservoir parameters underground, people often select seismic attributes only according to his/her personal experience. Thus, the selected seismic attributes often vary, with great arbitrariness. When geological conditions are ideal, or objects to be forecasted are relatively simple, or the raw seismic data collected has high SNR, this approach can works well. However, if geological structure is more complicated, predicted effect is greatly reduced.The difficulty is: in different work area or the different reservoirs of the same work area, the most sensitive, or the most effective, or the most representative seismic attributes of predicting the object is often not identical. Even in the same reservoir of the same work area, if the predicting object is different, the corresponding sensitive seismic attributes is different. So selecting appropriate properties can provide more accurate basis for geological interpretation, so as to provide technical assurance for oil and gas field development.In the traditional seismic data processing methods, linear models are used to describe the seismic waves. However, modern scientific research shows that seismic waves are chaotic, with significant nonlinear characteristics. To study the underground structure accurately, it is needed to be described by nonlinear models. Fractal geometry is a powerful tool for solving nonlinear problems. It objects to study the irregular object. Using the widespread self-similarity in the nature, fractal achieved great success in describing the fine structure of objects. In seismic exploration, due to underground sedimentary sequence contains a number of sedimentary cycles at different levels. Larger cycle contains smaller cycle, and in some small cycles contains smaller sedimentary cycle, which constituting a multi-level self-similar nested structure, so the fractal dimension properties of seismic data can describe characteristics of underground strata.Using fractal dimension to deal with seismic data is based on: the value of the fractal dimension correspond to the changing spectrum of signal amplitude along with frequency. High fractal dimension indicating porous rock underground that richly contains oil and gas. When seismic waves passing through them, spectrum of amplitude decays rapidly with the decaying frequency; In comparison, low fractal dimension indicates the compact formation underground, attenuation of amplitude slower than frequency spectrum.This paper presents methods for calculating capacity dimension and correlation dimension of seismic signals. Due to the influence of various noise, real seismic data is not strictly self-similarity, which just have scale-free nature in some of the scales. So, it is needed to try several times of calculating to choose proper parameters in order to reflect the real structure underground. Finally, we take a small developed oil and gas seismic profiles as example, calculating the capacity dimension and correlation dimension curve, and instantaneous capacity dimension profiles. According to the calculation results, the method this paper proposed can identify underground oil and gas effectly, which has some guidance to development of oil and gas fields.