ABSTRACTS

Seismicfaultinterpretationbasedonimprovedholistically-nestededgedetection.LIUNaihao1,2,LIShizhen3,HUANGTeng2,GAOJinghuai2,DINGJicai1,andWANGZhiguo4.OilGeophysicalProspecting,2022,57(3):499-509.

The accuracy and efficiency of fault interpretation greatly affect the exploration and development of oil and gas reservoirs. The traditional manual fault interpretation method relies on the experience of interpreters and takes a long time; the conventional automatic fault interpretation method mainly interprets faults by discontinuity analysis of seismic data and often contains multiple parameters, and thus its accuracy in fault interpretation mostly depends on the selected parameters. With the development of deep learning in recent years, the convolutional neural networks (CNNs) with nonlinear properties can also describe the discontinuous characteristics of seismic data. Therefore, an edge detection technology in deep learning, i.e., the holistically-nested edge detection (HED) network, is introduced in this study, and the network is improved and optimized on the basis of the cha-racteristics of seismic data and seismic faults, which leads to the improved HED (IHED) network suitable for intelligent seismic fault interpretation. The main steps are as follows: ① The original two-dimensional (2D) HED network is extended to a three-dimensional (3D) version, and thus a 3D HED network is constructed; ② the architecture of the 3D HED network is adjusted considering the multi-scale property of the network; ③ the 3D HED network is trained with 3D synthetic seismic data and corresponding label data for a 3D IHED model, and then the 3D IHED model is applied to field data for seismic fault interpretation. Compared with the coherence cube algorithm and U-Net model, the 3D IHED model features higher accuracy in the prediction of faults and better continuity. The proposed model provides an efficient and reliable new idea for intelligent fault interpretation.

Keywords： holistically-nested edge detection, deep learning, intelligent fault interpretation, convolutional neural network, U-Net

1. CNOOC Research Institute Co., Ltd., Beijing 100028, China

2. School of Information and Communications Engineering, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China

3. College of Artificial Intelligence, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China

4. School of Mathematics and Statistics, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China

Numericaldispersionsuppressionbasedonjointdeeplearninginthespaceandwavenumberdomains.ZHANGYan1,CUILinqi1,SONGLiwei2,andDONGHongli3,4.OilGeophysicalProspecting,2022,57(3):510-524.

The finite-difference scheme is commonly applied in seismic prospecting for numerical simulation of wavefields. The simulation accuracy, however, is affected by the serious numerical dispersion caused by spatial coarse grids or low-order operators of difference. In this paper, a numerical dispersion suppression method based on the joint learning of deep convolutional neural networks (CNNs) is proposed, which uses CNNs to adaptively extract wavefield features for dispersion correction. Firstly, the sparse features of the wavefield data in the space and wavenumber domains are used to build a CNN based on residual learning for the extraction of the main features of the wavefield data. Secondly, the L1norm is used for the sparse optimization of the network model, which can reduce the complexity of the model and enhance the generalization ability of the network. Finally, a joint objective optimization function is constructed to enable the network to learn the non-linear approximation capability of dispersion suppression under the semantics of the joint space-wavenumber domain constraints. The proposed method is applied to wavefield data from different forward models, and the results reveal that the method can effectively protect seismic signals and suppress dispersion; the combination of the network with migration learning is applied to the data from the new forward model, and good results can be achieved. Compared with similar algorithms, the proposed method boasts higher computational accuracy of coarse grids, lower computational costs, and a higher signal-to-noise ratio (SNR) of the obtained wavefield snapshot.

Keywords: numerical dispersion suppression, con-volutional neural network, joint learning, sparse constraint, residual network

1. School of Computer and Information Technology, Northeast Petroleum University, Daqing, Heilongjiang 163318, China

2. School of Physics and Electronic Engineering, Northeast Petroleum University, Daqing, Heilongjiang 163318, China

3. Artificial Intelligence Energy Research Institute, Northeast Petroleum University, Daqing, Heilongjiang 163318, China

4. Key Laboratory of Networking and Intelligent Control of Heilongjiang Province, Daqing, Heilongjiang 163318, China

Asparserepresentationmethodforseismicdata:adaptivemultilayereddictionarylearning(AMDL).YONGHao1,2,HANDuo1,2,ZHANGJunjie1,2,andWANGJunqiu1,2.OilGeophysicalProspecting,2022,57(3):525-531.

The accuracy of seismic data reconstruction by compressed sensing (CS) largely depends on the performance of the dictionary used for sparse representation. The sparsity level of each training sample in theK-singular value decomposition (K-SVD) is fixed, which may lead to under-fitting or over-fitting of the original sample. Moreover, it only uses the features of the original samples as the training dictionary and cannot utilize the implicit features generated in the dictionary learning process, which affects the reconstruction accuracy. In this paper, we adopt the adaptive multilayered dictionary learning (AMDL) method for the sparse representation of seismic data to improve theK-SVD method. It not only makes full use of the features at different levels in the dictionary learning process but also adaptively determines the number of atoms chosen for each layer. The experimental results show that the method can provide a more accurate sparse representation for CS-based reconstruction of seismic data than theK-SVD method.

Keywords：CS, AMDL, sparse representation, seismic data, sparse coding

1. College of Instrumentation and Electrical Engineering, Jilin University, Changchun, Jilin 130026 China

2. Key Laboratory of Geophysical Exploration Equipment, Ministry of Education, Changchun, Jilin 130000, China

Anintelligentmethodofinstantlyoutputtingsingle-shotrecordswithgeologicalhorizonsinseismicdataacquisitionfields.HUFeng1，WOYukai2，LIANGShunjun1，CHENJiangli1，ZHANGXiaobin1，andJINGLongjiang1.OilGeophysicalProspecting,2022,57(3):532-539.

Calibrating the geological horizon of seismic profiles by VSP and synthetic records，which has plenty of well-studied methods and well-established standards，is a critical step in conventional seismic data interpretation. However，the method of calibrating the geological horizon of single-shot records has rarely been explored. Generally，the quality of single-shot records is measured by the “phase surface method” in the field of seismic data acquisition, but this method fails to provide a clear picture of the geological horizon corresponding to the event in the reflection layer of single-shot records，which affects the scientific and rational qua-lity evaluation of single-shot records. Considering this，we propose building the four-dimensional seismic data volume of “the figure of buried timeT0u, geological horizon，topography，and source coordinates” and applying an intelligent method that instantly calibrates the geological horizon of each single-shot record. In other words, by the recor-ding geometry for seismic data acquisition, the computer automatically outputs the single-shot record with its geological horizon as each shot is explo-ded. The proposed method is conducive to analyzing the reflected energy and signal-to-noise ratio in the target layer, achieving a real-time response in che-cking the quality of each single-shot record，and thus ensuring that high-quality seismic data can be acquired.

Keywords：seismic acquisition，single-shot record，real-time automatic horizon calibration，quality analysis

1. Southwest Branch, BGP Inc.，CNPC，Chengdu，Sichuan 610213，China

2. School of Geoscience and Technology，Southwest Petroleum University, Chengdu，Sichuan 610500，China

DesigntechnologyofnonlinearsweepsignalofvibroseisbasedondampedRickerwavelet.XULeiliang1,ZHANGJian1,andZHAOGuoyong1.OilGeophysicalProspecting,2022,57(3):540-549.

At present, the correlation wavelet of the commonly used vibroseis sweep signal has many large sidelobes, which produces strong correlation noise, especially in the area with a low signal-to-noise (S/N) ratio in the west, and thereby affects the quality of seismic data seriously. Ricker wavelet is an ideal wavelet with small sidelobes and strong resolution. However, most of its energy is concentrated in the main frequency band, which leads to few low-frequency components and a narrow frequency band. As a result, Ricker wavelet fails to meet the requirements of broadband exploration. To solve the problem, this paper proposes a damped Ricker wavelet featuring small sidelobes, rich low frequency and broadband and defines its calculation formula. Then, the damped Ricker wavelet is used to design a broadband nonlinear sweep signal of vibroseis. With the characteristics of broadband and small sidelobes, the sweep signal is utilized to perform forward simulation. The application test shows that this method can improve the vibroseis exploration effect and the quality of seismic data.

Keywords：vibroseis, damped Ricker wavelet, sweep signal, broadband, high S/N ratio

1. SINOPEC Geophysical Corporation Shengli Branch, Dongying, Shandong 257100, China

Adynamicseasurfacemodelingmethod.WANGZhaoqi1,FANGuozhang1,DINGLiangbo1,ZHANGWei2,YEYueming1,andWANGZong-ren1.OilGeophysicalProspecting,2022,57(3):550-556,581.

Due to the influence of natural factors such as gravity and wind, the actual sea surface exhibits dynamic fluctuations in shape. The rough sea surface lowers the estimation accuracy of the ghost wave operator in data processing, thereby affecting the deghosting effect. To study the influence of ghost waves generated under the dynamic sea surface on seismic records, this paper first combines the static rough sea surface at different moments by using the Pierson-Moskowitz (PM) wave spectrum model to generate a dynamic sea surface at continuous time which approximates a real dynamic sea surface. Then, by using the Kirchhoff approximation, the paper puts forward the modeling methods of non-pulse source ghost waves and geophone ghost waves under the dynamic sea surface. The synthetic data test shows that the proposed modeling method can effectively simulate the seismic records more in line with the actual dynamic sea surface and test the actual effect of deghosting algorithms.

Keywords：dynamic sea surface, ghost waves, Kirchhoff approximation, PM wave spectrum, modeling method

1. PetroChina Hangzhou Research Institute of Geo-logy, Hangzhou, Zhejiang 310023, China

2. Shanghai Investigation, Design and Research Institute Co., Ltd., Shanghai 200335, China

CompressedsensingdatareconstructiontechnologyinjointFKandShearletdomain.YANHai-yang1,2,3,4,ZHOUHui1,2,3,LIUHaibo4,XUZhaohong4,SUNZandong4,andLIUZhao4.OilGeophysicalProspecting,2022,57(3):557-569.

The irregularity of seismic data caused by sparse source-receiver acquisition or field acquisition affects the imaging quality of seismic data. The reconstruction method based on compressed sensing theory can effectively reconstruct seismic data under the condition of limited sampling. Since the spatial random absence of seismic traces is shown as spatial aliasing in the wavenumber domain, we transform the reconstruction of seismic traces in the spatiotemporal domain into random noise suppression in the frequency-wavenumber (FK) domain. Specifically, the multi-scale and multi-directional Shearlet transform is performed on FK-domain data, and by iterative inversion to eliminate spatial aliasing in the FK domain, the spatial reconstruction of seismic traces is realized. The method in this paper performs the Shearlet transform after the FK transform, which can be viewed as a new sparse basis transform. Since the spectrum of the global random sampling factor is characterized by white noise, and the spectrum of the piecewise random sampling factor is characte-rized by blue spectra, the interference of the effective signal and aliasing of the piecewise sampling data is relatively reduced, which is more conducive to data reconstruction. The reconstruction experiment indicates that the reconstruction accuracy in the FK + Shearlet domain for piecewise random sampling is higher than that in the Shearlet domain for global or piecewise random sampling as well as that in the FK + Shearlet domain for global random sampling.

Keywords： compressed sensing, data reconstruction, Shearlet transform, piecewise random, sparse sampling

1. State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum (Beijing), Beijing 102249, China

2. CNPC Key Laboratory of Geophysical Exploration, China University of Petroleum (Beijing), Beijing 102249, China

3. College of Geophysics, China University of Petroleum (Beijing), Beijing 102249, China

4. BGP Offshore, CNPC, Tianjin 300457, China

Seismicrandomnoisesuppressionalgorithmwithautomaticdeterminationofthenumberofretainedsingularvalues.ZHUYuefei1,2,CAOJingjie1,2,andYINHanjun1,2.OilGeophysicalProspecting，2022,57(3):570-581.

The rank reduction-based method requires dividing the seismic data into different blocks during the noise suppression. However, each block corresponds to different numbers of singular values, and the number of valid singular values has to be estimated manually. It causes inefficient computation and cannot be industrialized. Therefore, we propose the adaptive damped multi-channel singular spectrum analysis (ADMSSA), which automatically determines the number of retained singular values by the Akaike information criterion during the seismic random noise suppression. Then the damped multi-channel singular spectrum analysis (DMSSA) is used for denoising. First, we introduce the denoising principle of the multi-channel singular spectrum analysis (MSSA) method, followed by the Akaike information criterion and empirical method to determine the number of valid singular values. The empirical method verifies the validity of the Akaike information criterion. Under the framework of DMSSA, ADMSSA just uses the range of dominant frequency to denoise automatically. The simulations and experiments with actual data show that ADMSSA can automatically determine the reliable number of singular values to achieve a high signal-to-noise ratio, breeding great potential for industrial applications.

Keywords： multi-channel singular spectrum analysis, Akaike information criteria, seismic data, denoising, number of singular values, block Hankel matrix

1. Key Laboratory of Intelligent Detection and Equipment for Underground Space of Beijing-Tianjin-Hebei Urban Agglomeration, Ministry of Natural Resources, Shijiazhuang, Hebei 050031, China

2. Hebei Key Laboratory of Strategic Critical Mineral Resources, Shijiazhuang, Hebei 050031, China

Researchonmultipleattenuationusing3Dhigh-precisionamplitude-preservingRadontransform.MAJitao1，LIUShiyou2,3，andLIAOZhen1.OilGeophysicalProspecting，2022,57(3):582-592.

3D seismic exploration has become a common method for seismic exploration. Conventional 2D Radon transform for multiple attenuation is only applicable for seismic data acquired in a 2D manner and does not consider the characteristic of 3D pro-pagation of a seismic wave field. Therefore，it is not suitable for 3D seismic data processing. It is urgent to explore a processing algorithm for 3D seismic data. After the systematic study of 3D Radon transform for multiple attenuation，iterative threshold shrinkage is adopted to improve the resolution of the 3D Radon transform domain. Consi-dering the characteristic of amplitude versus offset, orthogonal polynomial transformation is introduced to fit the amplitude variation of seismic data in different curvature directions. The results of synthe-tic data and real data show that the 3D high-precision amplitude-preserving Radon transform can achieve high-resolution data in the model domain and，effectively separate primaries and multiples. Moreover，the polynomial fitting can protect the amplitude of effective waves. The proposed me-thod enables multiple attenuation with high fidelity.

Keywords：3D Radon transform，iterative threshold shrinkage，high resolution，polynomial fitting，amplitude preserving

1. College of Geophysics，China University of Petroleum (Beijing)，Beijing 102249，China

2. Hainan Branch, China National Offshore Oil Corporation (CNOOC)，Haikou，Hainan 570100，China

3. School of Geoscience，China University of Petroleum (East China)，Qingdao，Shandong 266580，China

Qfactorestimationbyweightedspectralratiome-thod.YANGDengfeng1，LIUJun1，WUJing1，ZHANGWeiwei1，andBAIHaijun1.OilGeophysicalProspecting,2022,57(3):593-601.

The spectral ratio method is currently one of the most commonly usedQfactor estimation me-thods. When seismic data contains noise interfe-rence，the spectral ratio method has poor stability and the estimation result depends on the selected frequency band. This is because equal weights are used in theQfactor estimation by the spectral ratio method when the least-squares method is employed to fit a straight line. As a result, the fitting results are greatly affected by the abnormal values in the low and high frequency regions. To improve the stability of the spectral ratio method and reduce its dependence on the selection of frequency bands, this paper proposes the weighted spectral ratio method forQfactor estimation. During least-squares fitting， a Gaussian function is introduced as a weighting factor to reduce the weighting coefficients of signals with low signal-to-noise ratios. The peak frequency and variance of the Gaussian function are the centroid frequency and variance of the source wavelet，respectively. The model test indicates that the weighted spectral ratio method is more stable than the conventional spectral ratio method，and its dependence on frequency band selection is reduced. The application of actual VSP data further proves that the weighted spectral ratio method can estimate theQfactor stably and effectively.

Keywords：spectral ratio method，Qfactor，weighted spectral ratio，VSP

1. Shenzhen Branch, CNOOC China Limited，Shenzhen，Guangdong 518000，China

Spectral-elementmethodbasedonoptimalnumericalintegrationforseismicwaveformmodeling.MENGXueli1，2，LIUShaolin1，YANGDinghui3，WANGWenshuai2，XUXiwei1，andLIXiaofan4.OilGeophysicalProspecting,2022,57(3):602-612.

High-accuracy seismic waveform modeling for complex media is a difficult issue in the geophysics community，and developing a high-accuracy and efficient numerical algorithm is crucial to the research on the forward modeling and inversion of seismic waveforms. At present, the spectral-element method (SEM) has been successfully applied to seismic wave simulation by models on different scales. However，the Gauss-Lobatto-Legendre (GLL) numerical integration algorithm used by the conventional SEM is not able to accurately calculate the polynomial integration involved in the mass and stiffness matrices, which thus decreases the accuracy of SEM. Here we propose an optimal numerical integration algorithm to solve the abovementioned problem. We first construct the least-square formation of the objective functions for numerical integration and exact integration. After that，we utilize the conjugate gradient method to solve the weight coefficient of optimal numerical integration，which increases the accuracy of the numerical integration and thereby improves the numerical accuracy of SEM. Theoretical analyses and numerical examples verify that the spectral-element method based on optimal numerical integration performs better in suppressing numerical dispersion and increasing calculation accuracy.

Keywords：spectral-element method， motion equation of seismic wave，forward modeling，numerical integration

1. National Institute of Natural Hazards，Ministry of Emergency Management of China，Beijing 100085，China

2. School of Mathematics and Statistics，Ningxia University，Yinchuan，Ningxia 750021，China

3. Department of Mathematical Sciences，Tsinghua University，Beijing 100084，China

4. Institute of Geophysics & Geomatics，China University of Geoscience，Wuhan，Hubei 430074，China

Reverse-timemigrationofoptimizedpureacousticequationinanisotropicmediabycombiningPoissonalgorithmandwavefielddecompositionimagingcondition.XUShigang1,BAOQianzong1,RENZhi-ming1，andLIUYang2.OilGeophysicalProspecting,2022,57(3):613-623.

The traditional anisotropic reverse-time migration mainly adopts pseudo-acoustic wave equations (PWEs), which can easily lead to pseudo-shear wave interference on imaging profiles and numerical instability. Developing anisotropic pure acoustic wave equations (PAWEs) can solve the aforementioned problems. Therefore, this paper first reviews two commonly used PWEs in TTI media, and then an optimized pure acoustic wave dispersion relation based on least squares is obtained. On this basis, the high-accuracy PAWE is solved by the Poisson algorithm and finite-difference. In the conventional cross-correlation imaging condition, the full wavefield information is involved, which is prone to cause strong low-frequency noise. Given this issue and anisotropy, the isotropic imaging condition based on complex wavefield decomposition is further extended to anisotropic media. The wavefield can be decomposed to the different directional components, and the components with opposite propagation directions are selected for the final imaging. The basic theory and model examples verify that the combination of the optimized pure acoustic waves and the imaging condition for wavefield decomposition can effectively suppress pseudo-shear wave interference and low-frequency noise in anisotropic reverse-time migration, and high-quality imaging profiles can be produced.

Keywords：anisotropy, wave equation, pure acoustic wave, reverse-time migration, wavefield decomposition

1. Department of Geophysics, School of Geological Engineering and Geomatics, Chang’an University, Xi’an, Shaanxi 710054, China

2. State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum (Beijing), Beijing 102249, China

Analysisofquantitativerelationsbetweendifferentexactadjointoperatorpairsinacousticleast-squaremigration.WANGJiansen1,RENYuxiao2,CHENLei2,YANDong3,YANGChuangen3,andXUXinji2.OilGeophysicalProspecting，2022,57(3):624-637.

Least-square migration (LSM) is frequently mentioned in high-resolution imaging, whose successful application depends on the adjoint characteristic of forward-migration operator pairs. Normally, a forward-migration operator pair can be designed according to the Born approximation theory or/and reverse time migration (RTM) process. Its adjoint characteristic can be affected by the discretization and numerical implementation methods of wave equations, and the dot-product test can be used for the numerical test of this characteristic. However, the relations and the diffe-rences between the imaging results of different adjoint operator pairs are not clear. Considering this, three pairs of exact adjoint operators are derived by starting from the matrix expression of the second-order acoustic wave equation, two of which are constructed only on the basis of the Born approximation theory and the RTM process separately, and the third one is based on the self-adjoint discretization of the acoustic wave equation. They are named as Born-AdjBorn, DeRTM-RTM, and self-adjoint Born-RTM operator pairs, respectively, and the corresponding LSM processes are called LSBM, LSRTM, and self-adjoint LSBRTM, respectively. The matrix analysis and mathematical derivation indicate that a series of quantitative relations exist between the imaging results using the three operator pairs, and all these quantitative relations are validated by numerical experiments.

Keywords： acoustic LSM, wave equation discretization, exact adjoint operator pair, quantitative comparison

1. School of Qilu Transportation, Shandong University, Jinan, Shandong 250061, China

2. Geotechnical and Structural Engineering Research Center, Shandong University, Jinan, Shandong 250061, China

3. Huaneng Tibet Hydropower Safety Engineering Technology Research Center, Linzhi, Tibet 860000, China

Simultaneousinversionofpetrophysicalparametersofreservoirbasedoncuckoosearchalgorithm.LIUShiyou1,DUANZhichuan2,ZHOUFan1,andWANGRui1.OilGeophysicalProspecting,2022,57(3):638-646.

The inversion of petrophysical parameters is a significant method for reservoir prediction and evaluation, which can directly describe the abundant information contained in a reservoir. Due to the nonlinearity of geophysical inversion, it is difficult to reduce the ill-posed problems caused by this characteristic in the inversion of petrophysical parameters by local optimization, and thus the inversion results have multiple solutions. Given the above problems, this paper proposes a simultaneous inversion method based on the cuckoo search (CS) algorithm for petrophysical parameters of reservoir. On the basis of the relationship between elastic impedance and petrophysical parameters of reservoir, the inversion objective function is constructed, and the CS algorithm is introduced to find the optimal solution to the objective function. As a new meta-heuristic algorithm, the CS algorithm includes the Levy flight mechanism that can effectively solve the problems of local extreme values faced by conventional methods, and thus it can achieve high-precision prediction of petrophysical parameters of reservoir. The theoretical model and actual data test reveal that this method can effectively perform the inversion of petrophysical parameters, which can provide data support for re-servoir description.

Keywords：cuckoo search algorithm, global optimization, elastic impedance, petrophysical parameter of reservoir, simultaneous inversion

1. Haikou Branch of CNOOC Limited, Haikou, Hainan 570000, China

2. School of Geosciences, China University of Petroleum (East China), Qingdao, Shandong 266580, China

Identificationofstrike-slipfaultsinGaoshiti-MoxiareaofSichuanBasin.QIUZehua1,ZHOULu1,2,CHENXiao3,GUANXu3,WUYong1,2,andQIANYujie1.OilGeophysicalProspecting,2022,57(3):647-655.

Natural gas is continuously discovered in the Permian Qixia Formation in the Gaoshiti-Moxi area of the Sichuan Basin, and thus this area has great potential for oil and gas exploration. Transtensional strike-slip faults control the distribution of oil and gas, but there is little research on the characteristics of strike-slip faults, and it is insufficient to identify strike-slip faults only according to seismic profiles. Therefore, starting with the mecha-nical mechanism of the formation of strike-slip faults, this paper analyzes the structural styles and structural characteristics and guides the identification of strike-slip faults with the dolphin effect and ribbon effect. Moreover, the advantages and disadvantages of various seismic attributes are analyzed and compared, and the seismic attributes most suitable for strike-slip fault identification are selected. The study believes that the strike-slip faults in the Gaoshiti-Moxi area are mainly controlled by the single shear mode of Riedel shear. On the section, the structural styles include the positive flower structure, linear structure, Y-shaped structure, and the superposition of multi-stage flower structures, which are characterized by small fault displacement, unobvious horizon dislocation, and dominant event bending disturbance. On the plane, structural styles include the braided structure, en-echelon structure, and horsetail structure, and the dolphin effect and ribbon effect appear along the fault strike and the fault dipping direction, respectively. The seismic attribute ana-lysis by the waveform similarity-based coherence algorithm has a poor identification effect of strike-slip faults, while the curvature and ant body attributes can clearly identify the planar distribution characteristics of strike-slip faults. This method can provide a reference for the identification of strike-slip faults in other similar areas.

Keywords： Sichuan Basin, Gaoshiti-Moxi area, Permian Qixia Formation, strike-slip fault, seismic attributes, structural style

1. Sichuan Provincial Key Laboratory of Natural Gas Geology, Chengdu, Sichuan 610500, China

2. School of Geosciences and Technology, Southwest Petroleum University, Chengdu, Sichuan 610500, China

3. Exploration and Development Research Institute, PetroChina Southwest Oil & Gas Field Company, Chengdu, Sichuan 610041, China

Applicationofpre-stackdepthmigrationandfinereservoirpredictiontechnologyinwelltrajectoryadjustmentduringdrilling.LIChunmei1,PENGCai1,ZHANGXuan2,GUOHongxi3,DUANJie1,andLIWenqi1.OilGeophysicalProspecting，2022,57(3):656-665.

Due to the complex structure, the imaging accuracy of pre-stack time migration is not high in the SYS area, which results in a large difference between the actual drilled structure data and the structure data from seismic prediction. The gas reservoir of Dengying Formation in the GM area is restricted by geological conditions such as deep burial, strong heterogeneity and thin reservoir. Thus, fine reservoir prediction is difficult, and the high-quality reservoir drilling ratio is low. To improve the migration homing in the SYS area, this paper conducted high-precision anisotropic pre-stack depth migration. This technique fully considers the anisotropy of underground media and carries out pre-stack depth migration with accurate anisotropy parameters. With the ability to restore the underground structure to the exact position, it avoids the information loss of steeply inclined strata, improves the quality of seismic imaging, and reduces the structural exploration error. The success rate of drilling is thereby increased significantly. To improve the reservoir drilling ratio in the GM area, the paper suppressed the strong amplitude reflection at the bottom of the Cambrian System with the newly completed well and ongoing drilling well data. Moreover, it performed the high-resolution post-stack geostatistical inversion and multi-attribute fracture and cave prediction to adjust the drilling target and well trajectory in real time so that the well trajectory can be made along a high-quality reservoir. The results show that after the integrated tracking of drilling, seismic and geology, the structural depth error of the target zone of the deployment well decreases year by year, and increases are observed in the reservoir drilling ratio, the average testing production and the proportion of high-yield wells. This further confirms the importance of real-time tracking of well location.

Keywords： well location tracking, anisotropy, pre-stack depth migration, strong amplitude suppression technology, high resolution, geostatistical inversion, multi-attribute fracture and cave prediction

1. Southwest Geophysical Research Institute, BGP Inc., CNPC, Chengdu, Sichuan 610000, China

2. Exploration and Development Research Institute, PetroChina Southwest Oil & Gasfield Company, Chengdu, Sichuan 610000, China

3. Exploration Department, PetroChina Southwest Oil & Gasfield Company, Chengdu, Sichuan 610000, China

Amethodforporositypredictionofdeeplyburiedreservoirsbasedoniterativeinversion.TIANJun1,LIUYonglei1,XUBo1,BAIJianpu1,andLIQinglin1.OilGeophysicalProspecting,2022,57(3):666-675.

The existing linearized rock-physics inversion estimates porosity, shale content and saturation by using the P-wave velocity, S-wave velocity and density obtained by pre-stack elastic inversion. Compared with the P- and S-wave velocities, density contributes less to the reflection coefficient. Thus, the estimation of density requires a larger range of angles of seismic gathers. For deeply buried reservoirs, the estimation of density is unreliable due to the small reflection angles of pre-stack seismic gathers. In this case, the three-parameter inversion equation of rock-physics is underdetermined, and one cannot obtain a unique solution. This restricts the prediction of deeply buried reservoirs by existing linearized rock-physics inversion. Therefore, this paper proposes a method of porosity inversion for these reservoirs based on an iterative algorithm. First, the linear relationships of P- and S-wave velocities with porosity and shale content are deduced utilizing a linearized rock-physics model. Then, the objective function of iterative inversion is constructed in light of the Bayesian theory and solved by dichotomy. At last, the method is tested on synthetic seismic data and real data. Results show that it is independent of the density term and can well predict porosity with the iterative algorithm for deeply buried reservoirs with two phases of oil and water.

Keywords： rock-physics inversion, porosity inversion, deeply buried reservoir, iterative algorithm, dichotomy

1. Korla Branch of GRI, BGP, CNPC, Korla, Xinjiang 841000, China

AccumulationconditionsofHercynianperiodintheMaigaitislope,southwestdepressionofTarimBasinandexplorationdirection.LIUJun1,TIANLei1,ZHANGHuquan1,andZHANGNianchun1.OilGeophysicalProspecting，2022,57(3):676-685.

Oil in the Maigaiti slope, southwest depression of Tarim Basin is mainly accumulated in the Hercynian period, and Paleozoic strata are at the core of exploration. Exploring the Paleozoic strata of the Maigaiti slope has been continued for many years, but no significant progress has been made so far, the primary reason for which is the lack of understanding about accumulation conditions. This paper analyzes the differences in accumulation conditions (source rock evolution, oil and gas migration conditions, paleo-uplift evolution, and reservoir and preservation conditions) between the Maigaiti slope and surrounding areas, with the following results obtained. ① Different from the north and middle of Tarim Basin where hydrocarbon was generated on a large scale from the Lower Cambrian source rock in the Hercynian period, the Maigaiti slope did not witness the peak of hydrocarbon generation from source rock at in the same period. As a result, the hydrocarbon accumulation scale was limited. ② Accumulation conditions in several areas of the Maigaiti slope were also different. The east and west areas with paleo-uplift distribution were favorable areas for oil-gas migration, conducive to the accumulation of paleo-oil reservoirs in the Hercynian period. However, the paleo-oil reservoirs were damaged to different degrees in the late Hercynian period. The middle of the slope was located in the lower area of the structure in the period, and migration conditions above the gypsum-salt rocks were poor, so hydrocarbons were mainly accumulated under gypsum-salt rocks. ③ Different accumulation conditions led to the differences in oil-gas gathering characteristics in different formations and zones of the Maigaiti slope during the Hercynian period. Therefore, the exploration direction has been determined that the paleo-oil reservoirs of the Hercynian period are sought in the east and west of the slope, and the natural gas reservoirs of the Himalayan period are explored in the middle of the slope.

Keywords： accumulation conditions, Hercynian period, exploration direction, Maigaiti slope, southwest depression of Tarim Basin

1. Northwest Branch, PetroChina Research Institute of Petroleum Exploration & Development, Lanzhou, Gansu 730020, China

EvolutioncharacteristicsandcontrollingfactorsofshelfbreakzoneinsouthsubsagofBaiyunSaginnorthernSouthChinaSea.SHULiangfeng1,ZHANGLili1,LEIShenglan1,GAOZhongliang1,HANXiao1,andYUSa1.OilGeophysicalProspecting,2022,57(3):686-696.

The large-scale shelf-margin delta and deep-water fan systems developed in the Zhuhai Formation of the south subsag of Baiyun Sag are revealed by 3D seismic and drilling data. So far, researchers have not yet studied in detail the evolution characteristics of the Zhuhai Formation and clarified the evolution law of the shelf break zone during the transition from shallow-water shelf depositional environment to deep-water continental slope in the Zhuhai Formation and its relationship with deep-water deposition. Thus, this paper makes a detailed analysis of the sedimentary system structure of the shelf break zone in each sedimentary period with the new 3D seismic data over 2000km2in the study area and the fine interpretation results of more than 20 seismic profiles by referring to the latest research progress on the shelf break zone. Further, the paper identifies the evolution characteristics of each stage and summarizes the formation patterns and main controlling factors of the shelf break zone, with the following results obtained. ① The shelf break zone of the south subsag of Baiyun Sag is distributed in the NE-SW direction, and the progradation distance of the shelf break zone varied in different periods. From the fourth member to the first member of the Zhuhai Formation, the width of the shelf break zone decreased, which gradually advanced toward the SE. The average slope of the shelf break zone is 0.96°～4.76°, the seaward advance distance 21 ～28 km, and the progradation rate about 0.9～10.6 km/Ma. There are obvious differences in the characteristics of the shelf break zone between the north and the northwest. ② The evolution of the shelf break zone can be divided into three stages: the initial formation period (sedimentary period of the fourth member of the Zhuhai Formation), the development period (sedimentary period of the third member of the Zhuhai Formation) and the stable period (sedimentary period of the second and the first members of the Zhuhai Formation). During the development period, the shelf break zone advanced rapidly seaward, and its shape changed from bending to straight. In the stable period, the shape of the shelf break zone changed from jagged to smooth (sedimentary period of the second member of the Zhuhai Formation) and finally stabilized (sedimentary period of the first member of the Zhuhai Formation). ③ The evolution of the shelf break zone is influenced by the differential supply of sediments, sea level changes, paleo-uplifts, etc. Adequate provenance supply is the key to the formation of the shelf break zone. The differential supply of double provenance results in the difference in the shape of the shelf break zone in the north and northwest. The relative sea level rose first and then fell which modified the layout of the shelf break zone. The overall tectonic activity was stable during the depression period, and the paleo-uplift limited the range of the shelf break zone. The research is of important significance for the sedimentary evolution and oil and gas exploration in the deep-water area of the northern South China Sea.

Keywords： break zone, evolution characteristics, controlling factor, Zhuhai Formation, south subsag of Baiyun Sag, northern South China Sea

1. Research Institute, Shenzhen Branch of China National Offshore Oil Corporation Limited, Shenzhen, Guangdong 518054, China

ReconstructionofwelllogsbasedonXGBoost.ZHANGJiachen1,2,DENGJin’gen1,2,TANQiang1,2,andSHILin1.OilGeophysicalProspecting,2022,57(3):697-705.

Well logs play an important role in the formation evaluation. However, well log data might be missing or incomplete due to operational and geological issues in the logging process. As the reconstruction of well logs based on traditional empirical models and multiple regression method is less accurate, machine learning is proposed. Considering the limitation of the traditional neural network, XGBoost is utilized to build the reconstruction model for well logs. The directional wells in Bohai Bay Basin are exemplified to verify the model with the experiments of well log imputation and generation. The proposed model is compared with traditional machine learning models such as gradient lifting decision tree (GBDT), random forest (RF), and fully connected neural network (FNN) through K-fold cross-validation. The prediction effect is analyzed combined with the geological background. Results show that reconstruction of well logs based on XGBoost achieves high prediction accuracy, stability, and strong generalization ability.

Keywords： well logs, reconstruction method, machine learning, neural network, XGBoost

1. College of Petroleum Engineering, China University of Petroleum (Beijing), Beijing 102249, China

2. State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum (Beijing), Beijing 102249, China

Numericalsimulationandresistivityanisotropiccorrectionmethodofduallaterologresponseinhighlydeviatedwell.LAIQiang1,TANGJun2,WUYuyu1,XUWei2,XIEBing1,andJINYan1.OilGeo-physicalProspecting,2022,57(3):706-712.

This paper introduces the 3D finite element model (FEM) for simulating the dual laterolog response in the highly deviated well of carbonate re-servoirs. The effects of deviation angle and electrical anisotropy are considered in the numerical si-mulation. The correction charts of resistivity aniso-tropy in the deep and shallow investigation are proposed for the dual laterolog. These charts consider the investigation characteristics of dual laterolog tools and have more definite physical significance than the traditional anisotropic correction formula. To validate the effectiveness of the correction chart and method, core samples parallel and perpendicular to the carbonate reservoir of Dengying Formation in the Sichuan Basin are used in the high-temperature and high-pressure rock electricity experiment. The temperature and pressure correction charts are made to correct the horizontal and vertical resistivity, following which the anisotropy coefficient is obtained under certain temperature and pressure conditions. Then anisotropic correction is conducted on the dual laterolog data from the highly deviated well. Compared with those uncorrected, the corrected dual laterolog data from the highly deviated well are in better agreement with counterparts from the neighboring vertical well, which verifies the effectiveness of the correction method.

Keywords： dual laterolog, anisotropy, deviated well, numerical simulation, rock physics

1. Research Institute of Exploration and Development, Southwest Oil and Gas Company, CNPC, Chengdu, Sichuan 610041, China

2. College of Geophysics and Petroleum Resources, Yangtze University, Wuhan, Hubei 430100, China

AnewmethodforcalculatingclayboundwaterinNMRlogging.LUYunlong1，CUIYunjiang1，ZHUMeng1，andZHAOShuzheng1.OilGeophysicalProspecting,2022,57(3):713-718.

Clay bound water is one of the important indicators of reservoir logging evaluation. The theoretical formula reflecting the relationship between the clay bound water porosity and the cation exchange capacity (CEC) of clay is derived in light of the diffuse double layer theory, and then the clay bound water porosity is determined through the core cation exchange experiment. Depending on nuclear magnetic resonance (NMR) and cation exchange experiments，this paper analyzes the error of determining clay bound water porosity by a traditional method of utilizing theT2cutoff of 3ms clay bound water and clarifies the variation rules between the clay bound water porosity calculated byT2cutoff and different NMR cutoff. A model is established for the calculation of clay bound water porosity with variableT2cutoff of clay bound water. Further, a new method for its calculation in NMR logging is developed, with the specific calculation steps given in the paper. The application results show that the method is more reasonable and more suitable for core analysis than the traditional 3ms NMR logging method，which is an effective means of clay bound water evaluation.

Keywords：nuclear magnetic resonance(NMR)，clay bound water porosity，T2cutoff of clay bound water，cation exchange capacity(CEC)，porosity

1. Tianjin Branch of CNOOC Ltd.，Tianjin 300459，China

Finiteelementsimulationforfracturingmonitoringofoilandgasreservoirsbyusingcontactinducedpolarizationmethod.LIJinghe1,2,HEZhanxiang2,3,4,andMUTong1.OilGeophysicalProspecting,2022,57(3):719-727.

The exploration target which can access currents through natural or artificial outcrops is an important research object in geophysical exploration, and it is common in minerals, oil and gas, hydrological engineering, and environmental geophysics. However, it is a challenging task to efficiently and accurately simulate the spatial response distribution of targets with access to currents. Therefore, on the basis of the fracturing monitoring problem of oil and gas reservoirs, this paper proposes a three-dimensional finite element (FE) numerical simulation technology for the contact induced polarization method. The contact induced polarization method adopts the observation system with contact power supply and ground measurement, and the direct power supply can effectively improve the intensity and precision of observed signals. The current distribution of a unit field source of discrete grids is calculated by tetrahedron subdivision of staggered grids, and the three-dimensional numerical simulation of a contact field source is realized by the FE method. The formula for calculating the apparent resistivity of the contact induced polarization method is defined. In addition, we design the field sources, physical soil trough test simulation, and fracturing monitoring models of coalbed methane (CBM) reservoirs in Qinshui Basin of different methods, and the numerical results are compared and analyzed with the observation data of traditional induced polarization me-thods. The results verify the feasibility and effectiveness of the application of the contact induced pola-rization method in oil and gas reservoir fracturing monitoring.

Keywords： contact induced polarization method, fracturing monitoring of oil and gas reservoirs, unit field source, finite element simulation, CBM of Qinshui Basin

1. College of Earth Sciences, Guilin University of Technology, Guilin, Guangxi 541004, China

2. Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Shenzhen, Guangdong 511458, China

3. Shenzhen Key Laboratory of Deep Offshore Oil and Gas Exploration Technology, Southern University of Science and Technology, Shenzhen, Guangdong 518055, China

4. Department of Earth and Space Sciences, Southern University of Science and Technology, Shenzhen, Guangdong 518055, China

Ground-airbornefrequencytippersoundingmethodandcharactersiticsanalysisofitsone-dimensionalforwardmodeling.FENGMin1,2,CHENHui1,2,DENGJuzhi1,2,YINMin1,2,YUHui1,2,andZHOUCong1,2.OilGeophysicalProspecting,2022,57(3):728-737.

To fully utilize the three-component magnetic field information of the ground-airborne frequency-domain electromagnetic method, this paper proposes a ground-airborne frequency tipper sounding method by referring to the magnetotelluric tipper measurement technology. Starting from the full-space magnetic field formula of the horizontal electric dipole in the one-dimensional (1D) layered medium, we derive the expressions of the two componentsTxandTyof the ground-airborne frequency-domain tipper, and the 1D forward modeling of a typical layered model is conducted to analyze the response characteristics of the ground-airborne frequency tipper. Through the analysis on the forward modeling results of the uniform half-space model and 1D layered model, the following conclusions are drawn: ① The tipper response is symmetric along the field source direction and vertical direction as a whole, and the distortion effect of the field source and the transition zone exists si-multaneously; ② With the increase in transmitter-receiver distance, the tipper response form basically remains unchanged, and the amplitude decreases rapidly; ③ The amplitude of the imaginary part of the tipper remains unchanged while the amplitude of the real part drops rapidly as the flight height changes; ④ The vertical variation of the resistivity of underground media can be effectively studied by measuring the tipper response of different frequencies; ⑤ The ground-airborne frequency tipper sounding method should be applied at a long distance to avoid the distortion effect area as far as possible, and componentsTxandTyshould be rationally selected for interpretation. The theoretical feasibility and effectiveness of this method are proved by simulation analysis, and this method could be a possible development way for the ground-airborne electromagnetic method.

Keywords： ground-airborne frequency-domain electromagnetic method, tipper sounding, 1D forward modeling, layered model, electromagnetic response characteristics

1. School of Geophysics and Measurement-control Technology, East China University of Technology, Nanchang, Jiangxi 330013, China

2. State Key Laboratory of Nuclear Resources and Environment, East China University of Technology, Nanchang, Jiangxi 330013, China

Researchoncharacteristicsof3Dtransientelectromagneticresponsethroughcasingincylindricalcoordinatesystem.JIANGHairong1,ZHAONing1,2,ZHAOHongtu1,QINCe1,XIAOZhanshan3,andMIXiaoli4.OilGeophysicalProspecting,2022,57(3):738-745.

In recent years, transient electromagnetic through-casing technology has become one of the hotspots in electromagnetic prospecting. Since the conductivity and magnetic permeability of inhomogeneous media in wellbores and formation change drastically on small scales (casing, mud, etc.), efficient forward modeling for three-dimensional (3D) transient electromagnetic through casing is greatly important to simulate and eliminate the influence of metal casing on the electromagnetic field. In the calculations of forward modeling, the solution region is discretized in the cylindrical coordinate system to effectively reduce the number of grids and thus reduce the solution scale of the equation system, and the 3D cased-hole transient electromagnetic forward modeling is realized on the basis of the finite volume method. By simulating and analyzing the influence of parameters such as casing conductivity, relative magnetic permeability, thickness, and inner diameter on the electromagnetic field, a foundation is laid for correcting the influence of metal casing. The comparative analysis of the response results of trapezoidal waves, half-sine waves, and triangle waves de-monstrates that the transmitting waveform using trapezoidal waves performs the best. Moreover, the analysis on the influence of the duration of each stage of trapezoidal waves on electromagnetic response results reveals that the shorter waveform stabilization time has a significant impact on the early response results; the impact of the turn-on time on the response results can be ignored, and shorter turn-off time leads to stronger electromagnetic response. The research conclusions can provide technical support for research on transient electromagnetic through-casing data processing technology.

Keywords： transient electromagnetic response through casing, cylindrical coordinate system, finite volume method, transmitting waveform, electromagnetic response characteristics

1. School of Physics and Electronic Information Engineering, Henan Polytechnic University, Jiaozuo, Henan 454000, China

2. Key Laboratory of Gas Geology and Gas Control in Henan Province-National Key Laboratory Cultivation Base Co-constructed by Provincial Departments, Jiaozuo, Henan 454000, China

3. China Petroleum Logging Co.,Ltd, Xi’an, Shaanxi 710077, China

4. GME & Geochemical Surveys, BGP Inc., CNPC, Zhuozhou, Hebei 072751, China