Research On High Resolution Seismic Characterization Of Thin Oil And Gas Reservoirs

Thin reservoirs are an important field of oil and gas exploration and development in China,requiring seismic data with high resolution and amplitude preserving efficiency.Especially when the oilfield enters the development stage,the problem of high-resolution seismic characterization of thin-layer structures becomes particularly prominent and even more urgent.High-resolution seismic characterization of thin-layer structures is a complex system engineering,and the processing quality of seismic data plays a very important and fundamental role.Only when the signal-to-noise ratio,resolution and amplitudepreserving property of seismic data meet certain requirements,the subsequent attribute analysis and seismic inversion can obtain more reliable prediction results.Otherwise,the theoretical advantages of reservoir characterization by seismic inversion are difficult to be effectively exerted and fully accomplished.Based on the actual demands of high-resolution seismic characterization of reservoir structures in YS area of Dagang oilfield,a relatively systematic study is carried out on the main problems in seismic data amplitude recovery,high resolution processing,attribute analysis and seismic inversion.A directional deconvolution method is proposed to eliminate the influence of the air gun arrays and the combination of geophones.A highresolution processing method for regularization of reflection structures,and a composite seismic attribute analysis method for thin inter-bedded structures are proposed.On this basis,an exploratory study on rock physics seismic inversion under the Bayesian framework is carried out.Some research results have been practically applied in YS area of Dagang oilfield,and relatively reliable prediction results have been obtained.Although high-density single-point acquisition has achieved large-scale applications in the industry,in order to enhance the excitation energy and improve the signal-to-noise ratio of seismic records,combined excitation and combined receiving are still an important approach for seismic data acquisition in the field.This acquisition method not only improves the signal-to-noise ratio of the original seismic records,but also induces the directional effect of the seismic data.The directional effect not only reduces the resolution of seismic data,but also changes the function of reflection amplitude with incident angle and azimuth angle,which reduces the accuracy of reservoir prediction using pre-stack inversion to some extent.Therefore,this paper proposes an unsteady inversion-based wavelet directional deconvolution method for air gun arrays.Firstly,the method calculates the far-field wavelets related to the azimuth and exit angles according to the spatial patterns of the air gun arrays.Then,based on the velocity model,the exit angles of the seismic reflection at the source location is calculated,and a directional filter operator is constructed accordingly.Finally,the unsteady inversion method is used to shape the seismic wavelets in different directions into far-field wavelets directly below the arrays,so as to realize the wavelet directional deconvolution for the air gun arrays.The method in this paper directly performs directional deconvolution in the space-time domain,which can better avoid the influence of spatial aliasing on directional deconvolution.The contradiction between reservoir thickness and seismic resolution is very prominent in reservoir development.The existing high-resolution processing methods for seismic data are subject to the following two factors,one is the effective frequency band of the seismic data,and the other is the contamination of high-frequency noise for seismic signal.Aiming at the above problems,this paper proposes a multi-channel sparse pulse deconvolution method for regularization of spatial reflection structures.The method extracts the spatial reflection structure characterization operator from the seismic data itself,introduces the characterization operator into the regularization condition of multichannel sparse pulse inversion,and performs reflection coefficient inversion under the joint driven by the sparse structures and reflection structures.This method greatly improves the vertical resolution of seismic data,and provides high-quality seismic data for subsequent sequence interpretation and seismic inversion.Attribute analysis is an important technique for lateral distribution and thickness prediction for sand bodies.It makes up for the limitation of seismic data resolution in the time direction.However,most of the existing seismic attribute analysis methods are based on a single sand wedge model,which only considers the effect of top-bottom interference for sand bodies on resolution and reflection amplitude,and does not consider the interference effect between different sand bodies,which is not completely suitable for the attribute analysis of thin inter-bedded seismic reflection.Therefore,this paper proposes a composite seismic attribute analysis method suitable for thin inter-bedded geological models.Compared with the traditional amplitude attributes and frequency attributes,this composite attribute not only improves the estimation accuracy of the thickness for a single sand body,but also can better predict and analyze the cumulative thickness of sand bodies in sand-shale thin inter-bedded reservoir.After the geometries and spatial distributions of thin sand bodies have been clarified,it is more challenging to perform petrophysical seismic inversion for their physical parameters and hydrocarbon-bearing properties.Based on the rock-physics modeling and seismic response analysis for the mapping relationship between petrophysical parameters such as porosity,saturation and shale content of oil and gas reservoirs and elastic parameters such as P-wave velocity,S-wave velocity and density.An exploratory study is carried out on the rock-physics inversion method under the Bayesian framework,and the inversion results are preliminarily analyzed using the well logging data in the study area.Finally,taking YS area of Dagang oilfield as an example,focusing on the actual demands for thin-layer structure characterization in oilfield development.Conducting relatively systematic application research from major technical approaches such as highresolution seismic data processing,detailed structural interpretation,seismic attribute analysis and seismic inversion reservoir prediction.The validity and adaptability of the proposed method in this paper are investigated and analyzed,and the reference opinions and prediction graphs with relatively strong guiding significance for oilfield development are provided.