The Predicting Technique Of Carbonate Fractured Reservoir

Fracture can not only supply reservoir space but also can be regarded as migration pathway for oil and gas. Therefore, fracture prediction is of great significance in identifying carbonate reservoirs. In this thesis, taking the region of Sikou as the research target,fracture reservoir description technique research is discussed. It provides a good seismic prediction base for monographic study about the distribution of fracture. On base of collecting and studying the seismic, geological, logging and drilling data of the region, the sedimentary facies analysis has been carried out, and the spatial distribution of the reservoir has been predicted, which has established a good basis for the seismic prediction of the Es4 Member reservoir's fracture zone.On the basis of full understanding of the geology of Sikou area, we can learn that the thickness of limestone is very important in the formation of hydrocarbon reservoir. According to the analysis of'Shaojia'region, we know that the depth of the limestone is generally less than the tuning thickness. In view of this point, basing on the discussion of the frequency spectrum of laminar layer and alternate layer, this article has investigated the conditions for calculating layer depth by spectral decomposition. The spectral decomposition technology has difficulty in determining trap frequency cycle, so we propose the technology of cepstrum analysis for the prediction of limestone depth, and get a good result. Besides, we also have a discussion of predicting fracture reservoir in sedimentary facies of'Sikou'region by means of frequency domain amplitude attenuation. In case that the fractures in limestone contain fluid such as water, oil, or gas, the high frequency component will be absorbed, which leads the sharp attenuation of the energy of seismic wave in high frequency. With this method, the paper predicts the fractured limestone reservoir in Sikou area by analyzing the amplitude attenuation gradient of seismic record in high frequency. The theoretical model shows the feasibility and validity of the method in recognizing the fractured reservoir. We propose a new high-accuracy time-frequency analysis—Generalized S transform basing on exploration the advantages and disadvantages of Short-time Fourier transform, wavelet transform, S transform. Taking the advantage of Generalized S transform that it can overcome the influence of time window to the results of conventional spectral decomposition, so we get the instantaneous frequency of real sense by spectral decompose, which can be better to analyze the attenuation characteristic of frequency amplitude in fractured reservoir.Because the traditional coherence technique has defectiveness in recognizing the fractures of small scales, we propose multi-fractal technique to recognize the small scale fractures. On the basis of the multi-fractal theory, we research the point of the design of measure in multi-fractal spectrum ,in view of the feature of limestone fracture's seismic reflection. According to the calculating of routine deviation , correlation, sum measure, we find that the application is not ideal. So, we have designed synthesized measure arithmetic, and validate it by theoretical model and practical data. The result shows that this algorithm can identify the underground crack effectually, so it is an effective method to predict the limestone fracture reservoir.