Research And Application Of Carbonate Multi-scale Fractures Prediction

The hydrocarbon detection of carbonate fracture-cave reservoirs is more and more important in the oil and gas exploration.However,due to the complex composition of mineral and strong randomness of reservoir space distribution,carbonate reservoirs have large heterogeneities.And it is difficult to accurately determine favorable areas by conventional reservoir prediction methods.This paper takes the middle-lower Ordovician carbonate reservoirs in the TH oilfield of the Tarim Basin in western China as the research object.Considering the classification of fracture scales,the analysis of azimuth P-wave anisotropy and the shear wave splitting characteristics is done based on rock physics and forward modeling.And then,the understandings are applied to the detection of fractured hydrocarbon reservoirs in the TH field.First,according to the vertical length and the aspect ratio of the fractures,fractures are divided into four categories: micro-scale fractures,meso-scale fractures,macro-scale fractures and cave.Among them,meso-scale fractures are the key and difficult point in current research.According to actual well logging and core data,the multiple mineral components of background rock and the extensive development of fractures and pores are considered in the model.In this paper,Voigt-Ruess-Hill theory is combined with Chapman theory to construct a rock physics model suitable for carbonate rocks in the study area.Based on the rock physics model,forward modeling analysis of the relationship between P-wave azimuth properties and fracture parameters is carried out.It is believed that the P-wave attenuation is most sensitive to the meso-scale fractures in meter scale,and fractures are the main cause of azimuthal anisotropy.Therefore,the use of azimuth characteristics of P-wave attenuation can better identify the mesoscale fractures in meter scale.Then,calculatting accurate attenuation parameters by pre-stack azimuth gathers becomes an important aspect.However,in actual seismic data,it is often difficult to obtain the value of attuation.In this paper,smooth pseudo-Wigner-Ville distribution and least-squares nonlinear fitting method are introduced into the calculation of frequency attenuation gradient which represents high frequency absorption.At the same time,this method is extended from poststack seismic data to prestack azimuth gathers to obtain azimuth frequency attenuation gradient.The use of azimuth frequency attenuation gradient can better reflect the azimuthal anisotropy characteristics caused by the meso-scale fractures.In addition,S-wave(transformed shear wave)data provide more comprehensive reservoir information than P-wave data.Based on the above rock physics model of carbonate rock,the forward modeling and numerical analysis of seismic records show that the energy difference between fast and slow shear waves is small when the model saturated with water.While,the energy difference between the two waves is large when the model saturated with gas.That is to say,comparative analysis of amplitude energy can be used to distinguish gas and water bearing reservoirs.At the same time,it is proved that the frequency variation of shear wave splitting is not only related to the fracture density,but also related to the fracture scale.After fast and slow shear wave are sepatated perfectly,the gas-bearing reservoirs can be detected using the shear wave delay and amplitude energy ratios,which is in good agreement with the logging data.Based on the above research and analysis,a comprehensive multi-scale fracture prediction method combined with well data and seismic data is applied.Firstly,the prestack azimuth frequency attenuation gradient is used to predict the development intensity of mesoscale fractures.At the same time,variance,curvature and ant body of postastack seismic data are used to predict the distribution of macro-scale fractures,and the calculation results are constrained by trial production wells to determine the effective range.Then,the instantaneous peak frequency is used to detect cavities.Finally,the three prediction results are merged together to obtain multi-scale fracture comprehensive detection results.The actual results show that this method can well explain the spatial distribution of fractures in carbonate rocks and improve the accuracy of fracture prediction.