ABSTRACTS

Seismicfaultinterpretationbasedondeepconvolutionalneuralnetworks.CHANGDekuan1, 2,YONGXueshan2,WANGYihui2,YANGWuyang2,LIHai-shan2,andZHANGGuangzhi1.OilGeophysicalProspecting,2021,56(1):1-8.

Seismic fault interpretation has always been a key task in the process of oil and gas exploration and development. Conventional fault interpretation is mainly based on human-computer interaction, which is of low efficiency and causes the results with many uncertainties. In addition, conventional methods for fault interpretation usua-lly set multiple parameters, whose controls accuracy of the predicted faults. This paper proposes a method using seismic data based on convolutional deep neural networks. Taking the advantages of ResNet for effectively training deep convolutional neural network and U-Net architecture for characterizing multi-scale and multi-layer characteristic information, this method combines deep residual neural network and U-Net architecture to construct a network architecture (SeisFault-Net) for fault interpretation based on seismic data. The U-Net architecture consists of an encoding sub-network and a decoding sub-network. They enable the SeisFault-Net to train models in an end-to-end manner. The residual neural network can suppress the gradient dispersion of deep network, and effectively improve the training efficiency of the SeisFault-Net. After trained, the SeisFault-Net can perform fault interpretation based on seismic data without setting any parameters. This avoids the empirical error and uncertainties caused by parameters artificially set in conventional methods. Applications to raw data have proved that the SeisFault-Net me-thod can effectively and accurately detect fault loca-tions, and the faults have good vertical continuity and clear outlines. The detailed information of faults interpretated by the SeisFault-Net method is more abundant and accurate than the coherent algorithm. And the calculating efficiency of the SeisFault-Net method is very high in seismic fault interpretation.

Keywords：fault interpreation, deep learning, resi-dual neural network, U-Net, seismic data interpretation

1.China University of Petroleum (East China), Qingdao, Shandong 266555, China

2.Research Institute of Petroleum Exploration & Development - Northwest, PetroChina, Lanzhou,Gansu 730020, China

Randomnoisesuppressionofseismicdatabasedonjointdeeplearning.ZHANGYan1,LIXinyue1,WANGBin1,LIJie1,andDONGHongli2.OilGeophysicalProspecting,2021,56(1):9-25，56.

Random noise suppression is the key task of seismic data processing. To improve SNR (signal-to-noise ratio) and increase the efficiency and accuracy of following processing and interpretation, appropriate suppression methods should be used for noises induced by different mechanisms and with different characteristics. Applicable denoising methods based on deep learning usually focus on feature extraction in time or frequency domain, which result in over-smoothed or blurred textures in local zones. In addition, the kernel of a traditional convolution neural network is usually set to be a small and fixed block, which limits the size of the receptive field and reduces the diversity of the target characteristics extracted from seismic data. This paper proposes a method of random noise suppression based on joint deep learning. Firstly, features in both time domain and frequency domain are considered, and the joint error is used to define the loss function to improve effect of various extracted features. Secondly, by analyzing the influence of the kernel size and network depth on the size of the receptive field, the method of expanding convolution is used to extract more diverse features and reduce the loss of details of seismic data. Thirdly, according to the similarity between the input and output samples of the network, a residual learning strategy is introduced. Finally, the batch normalization (BN) algorithm is used to accelerate the convergence of the model and improve denoising efficiency. Compared with similar algorithms, the method proposed in this paper has a better effect on preserving the features of events and provides higher SNR.

Keywords: noise suppression, deep learning, joint loss function, expanded convolution, residual network, texture, signal-to-noise ratio (SNR)

1. Institute of Computer and Information Techno-logy, Northeast Petroleum University, Daqing, Heilongjiang 163318, China

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

PermeabilitypredictionusingPSO-XGBoostbasedonloggingdata.GUYufeng1,ZHANGDaoyong1,andBAOZhidong2.OilGeophysicalProspecting,2021,56(1):26-37.

Models for permeability prediction generally can be classified into two major types, physical and fitting models. Universally, physical models are wel-comed by geophysicists since the predicted values are calculated on the basis of logging theory, but they show bad generalization on application due to strict requirements on logging data. Fitting models repre-sented by stepwise regression are capable to make quick prediction, but they are difficult to accurately and analytically explain the relationship between permeability and logging curves because of their cal-culation mechanisms, thus also presenting bad gen-eralization. In order to create a new and more pow-erful fitting model, XGBoost, a widely used fitting model at present, is selected and modified by PSO to optimize hyper-parameter tuning. Then the hybrid model PSO-XGBoost is proposed. In this paper, taking the tight sandstone reservoirs of the Chang 4+5 members as a case, the prediction capability of the PSO-XGBoost mo-del are validated by three well-designed experiments. The experiment results show that:①Compared with physical models, fitting models utilize a fewer parameters to complete prediction, and present better applicability on permeability prediction when modeling data are insufficient, but they have limits on generalization since the prediction is sensitive to the quality of mode-ling data and thereby usually unstable; ②SVR, GBDT, and XGBoost can be improved by PSO, and the formed PSO-SVR, PSO-GBDT and PSO-XGBoost can figure out permeability rapidly. In comparison, PSO-SVR and PSO-GBDT show relatively unstable prediction due to their sensitivities on the quality of learning samples, while PSO-XGBoost displays better performances in predicting efficiency, reliability of predicted results, and prediction stability. Therefore, PSO-SVR is deemed to be unsuitable on permeability prediction, and PSO-XGBoost suitable; ③The prediction capabilities of stepwise regression, PSO-SVM, PSO-GBDT, and PSO-XGBoost can be enhanced when more learning samples are trained.

Keywords：permeability prediction, tight sandstone reservoir, machine learning, stepwise regression, SVR, GBDT, XGBoost, PSO technique

1.Strategic Research Center of Oil and Gas Resources, Ministry of Natural Resources, Beijing 100034, China

2.China University of Petroleum (Beijing), Beijing 102249, China

Seismicfaciesanalysisbasedoncepstrumcharacte-risticparametersandspectralclustering.SANGKaiheng1,ZHANGFanchang1,andLIChuanhui2.OilGeophysicalProspecting,2021,56(1):38-48.

Stochastic simulation, neural network, clustering and deep learning are always used to seismic facies analysis. However, stochastic simulation results are easily affected by stochastic models, and it is difficult to get accurate seismic facies division in complex geological areas. Neural network and deep learning methods have strong fault tolerance and generalization ability, but they require massive training samples and high computing costs. Classical clustering algorithms such as K-means clustering and C-fuzzy clustering can obtain ideal clustering results on simple data, but cannot achieve global optimization for non-convex data. To overcome these problems, we propose a seismic facies analysis method based on cepstrum characteristic parameters and spectral clustering. In this me-thod, seismic cepstrum characteristic parameters are calculated as input variable of spectral clustering, and then the corresponding relationship is established between seismic facies and geological body after calibrating by well data. The spectral clustering method based on graph theory transforms data clustering into graph segmentation, which achieve accurate clustering through optimal graph segmentation. And we also construct a sparse similarity matrix through optimizing the similarity matrix calculation method, by which the storage and calculation problems caused by large matrix dimensions can be solved. Therefore, spectral clustering is more suitable for 3D seismic facies division. The advantages of cepstrum characteristic parameters are as follows: on the one hand, it can reduce data dimension and computational complexity; and on the other hand, it can eliminate the influence of waveform, and improve division accuracy. The applications to model and real data show that the seismic facies divided by the proposed method are in better agreement with the paleogeomorphology than the facies division based on instantaneous seismic amplitude and multiple seismic attributes, showing clearer boundaries and better interpretability. The results are reliable data for oil exploration and reservoir evaluation.

Keywords：spectral clustering, seismic cepstrum characteristic parameters, seismic facies, reflection coefficient

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

2.School of Geophysics and Information Technology, China University of Geosciences (Beijing), Beijing 100083, China

First-breakpositioningandaccuracyevaluationbasedonvectorsuperposition.YANGHaishen1,XULijun1,MAJie1,HOUKunpeng1,andXIAOYongxin1.OilGeophysicalProspecting,2021,56(1):49-56.

Influenced by current, tide, cable，etc., the final position of submarine cable or node geophone would drift from the designed position in offshore seismic survey, so that it is necessary to determine the exact position of the receiver point on the seabed through secondary positioning. A first-break positioning method, namely a vector superposition method, is proposed in this paper. The exact coordinates of the receiver point can be obtained by constructing the vector of each source-receiver pair and vector superposition. And integrating with the quantitative evaluation index of positioning accuracy, we get a positioning method involving first break picking, positioning calculation and posi-tioning accuracy evaluation. Applications to raw data have proved that the method can provide accurate secondary positioning results of submarine cable or node geophone, and quantitatively evaluate the positioning accuracy. It has a wide application in marine seismic survey.

Keywords：marine exploration, first-break positioning, vector superposition, positioning accuracy evaluation, linear normal moveout (LMO)

1. Acquisition Technique Center, BGP Inc., CNPC, Zhuozhou, Hebei 072751, China

BlendednoisesuppressionbasedonSVDconstrainediterativeinversion.DONGLieqian1,ZHANGMugang1,LUOFei1,ZHANGYimeng1,WANGZe1,andWEIGuowei1.OilGeophysicalProspecting,2021,56(1):57-61，117.

The blending acquisition technology with ultra-high productivity emerged as a promising way of significantly increasing the efficiency of seismic acquisition. However, blended noises from simultaneous sources smear effective energy. We propose a blended noise suppression approach based on singular value decomposition (SVD) constrai-ned iterative inversion. In the proposed approach, the key step is that the maximum singular value of the blended noise in a user-defined window is used as a constraint to iteratively update the singular value vector of the blended data in common offset gathers or the common midpoint gathers after normal moveout, where the SVD has a superior capability to represent the coherency of the seismic data. In this way, the singular value vector corresponding to the effective energy is recursively po-lished by the iteratively updated constraint in the inversion framework to obtain final de-blended data. Applications to simulated field data have proved that the method is effective for suppressing noises while protecting signals.

Keywords：blending acquisition with ultra-high productivity, singular value decomposition (SVD), blending noise, iterative inversion

1. BGP, CNPC, Zhuozhou, Hebei 072751, China

AnalysisandprocessingtechnologyofazimuthanisotropyinOVTdomainofwide-azimuthseismicdata.LIAng1,ZHANGLiyan1,YANGJianguo1,LINa2,LIShichao1,andYAOYulai1.OilGeophysicalProspecting,2021,56(1):62-68.

In order to improve the imaging quality of wide-azimuth seismic data, wide-azimuth anisotropy should be corrected. Picking coherent spectra is an anisotropic processing method in OVT (offset vector tile) domain. Based on azimuth gathers after migrating in OVT domain, it picks up time difference information on stacked sections, and then inverts anisotropy parameters. In other words, it obtains the azimuth and strength of azimuth anisotropy, and then corrects the azimuth anisotropy.Applied to theoretical models and raw wide-azimuth seismic data, the coherence spectra picking method eliminated the influence of azimuth anisotropy time difference, and improved the quality of seismic imaging, providing high-quality data for subsequent interpretation and reservoir prediction.

Keywords：wide-azimuth seismic, OVT (offset vector tile) domain, azimuth anisotropy, time difference correction, “snail” gathers

1. Shenyang Center of Geological Survey, China Geological Survey, Shenyang, Liaoning 110000, China

2. Research Institute of Exploration and Development of Daqing Oilfield Company Limited, Daqing, Heilongjiang 163712, China

Randomnoisessuppressionbasedonoverlappinggroupsparsity,non-convexLp-pseudo-normregula-rizationandhigher-ordertotalvariation.LIANGShanglin1，HUTianyue1，CUIDong2，andSUIJingkun1，2.OilGeophysicalProspecting，2021，56(1)：69-76.

Suffering serious staircase effects, the higher-order total variation model provides unsatisfactory denoising results. In this paper, we introduce a technique of overlapping group sparsity, and utilize the non-convex Lp-pseudo-norm for preserving weak signals. Our new model can fully exploit local similarity of signals instead of unrelated individual data. To solve the multi-constrained problem, we adopt the alternating direction method of multipliers to divide the whole problem into four sub-problems. The iteratively re-weighted alternating direction L1-norm and majorization minimization algorithm are added into the algorithm to improve the efficiency and accuracy. Applied to synthetic and field data, the improved method not only reduced staircase effects and attenuated random noises, but also effectively preserved weak signals.

Keywords：overlapping group sparsity, higher-order total variation, non-convex Lp-pseudo-norm, staircase effects, random noise suppression

1. School of Earth and Space Sciences, Peking University, Beijing 100871, China

2. Research Institute of Petroleum Exploration and Development, PetroChina, Beijing 100083, China

Amethodofcombiningmulti-channelsignalstosuppressthestrongreflectionthroughmatchingpursuit.YANGZipeng1，SONGWeiqi1，LIUJun2，CHENJunan2，LIUqun2，andWUDi1.OilGeophysicalProspecting，2021，56(1)：77-85.

It is usually difficult to identify useful signals in and around strong reflections. Relying on the sparse representation ability of the matching pursuit algorithm, we propose a method for suppressing strong reflections through matching pursuit (TC-MMP) under the constraint of multiple traces. To determine the condition for terminating iteration in the process of signal decomposition, we introduce a residual ratio threshold into the ma-tching pursuit algorithm. Compared with fixed itera-tions or fixed residual threshold, this algorithm can improve the efficiency of signal decomposition. It is a matching pursuit algorithm under the constraint of a termination condition. First, seismic signals with strong reflections are decomposed by TC-MMP, and matching wavelets with the strongest energy are searched and regarded as the best matched strong reflections. Then the weighted energy coefficient is obtained based on the strong reflections extracted from multiple traces. It can increase the energy proportion of effective signals in strong reflections. Finally, the optimal parameter for suppressing strong reflections is obtained through tests, making the useful signals in strong reflections retained while the suppressed signals near the strong reflections highlighted. Applications to a theoretical model with a high-resistance layer and raw seismic data with well tops show that the proposed method is more reliable and effective than conventional methods. It can provide technical support to accurate reservoir description.

Keywords：multi-channel extract, suppress the strong reflection, energy weighted, TC-MMP algorithm, useful signal

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

2. Research Institute of Exploration and Development, Northwest Oilfield Branch Co., SINOPEC, Urumqi, Xinjiang 830011, China

DeblendingmassiveOBNdataacquiredbyefficientandblendedshootingmethod.CHENYingpeng1,ZHANGHongjun1,LIUYong1,ZHAOMin1,SONGJiawen2,andQIQunli1.OilGeophysicalProspecting,2021,56(1):86-91.

Compared with land blended data with simple noises, OBN blended data have many types of noises due to the special method of OBN acquisition. Two primary causes for the multiple types of noises are heavy overlapped shots on the logical coordinate system and severe geometry deformation. After analyzing the types of noises in OBN data, we studied iterative dynamic mapping, pre-processing of logical shot coordinates and sparse inversion deblending technology. The sparse inversion deblending method is based on FKK. It globally maps the logical positions of all overlapped shots in a dynamic and iterative way to recongnize noises, and then it accurately and quickly deblends the massive seismic data severely blended. The application in Block A shows that the deblended OBN data are of high fidelity, indicating that the method can provide optimal results while improving deblending efficientcy.

Keywords：efficient and blended acquisition of OBN data, noises, iterative dynamic mapping, pre-processing of logical shot coordinates, sparse inversion deblending

1. Overseas Business Department,GRI,BGP Inc., CNPC, Zhuozhou, Hebei 072750,China

2. Research & Development Center,BGP Inc.,CNPC, Zhuozhou,Hebei 072750, China

Wave-equationtraveltimetomographyusingthege-neralizedRytovapproximation.FENGBo1，2,WURu-Shan3,LUOFei1,XURongwei1,andWANGHuazhong1.OilGeophysicalProspecting,2021,56(1):92-99.

The conventional finite-frequency tomography is usually derived from the Born or Rytov approximation which implies weak-scattering assumption. Therefore, the linearized forward problem in the finite-frequency theory is not satisfied for strong velocity perturbations. In the case of forward scattering and small-angle propagation, the generalized Rytov approximation (GRA) method recently developed can achieve improving phase accuracy of forward-scattered wavefield, making it more suitable for traveltime tomography. In this paper, we combine the conventional finite-frequency theory with GRA and propose a GRA-based traveltime sensitivity kernel, which works well regardless of the magnitude of velocity perturbations. Numerical examples show that the traveltime perturbation of forward-scattered waves can be correctly handled by the GRA-based traveltime sensitivity kernel. Then we propose an implicit matrix-vector product strategy which can calculate the Hessian matrix-vector product without explicitly forming the Hessian matrix, making it more attractive for 3-D problems. We solve the traveltime inverse problem with the Gauss-Newton method, where the Hessian matrix-vector product is obtained by the proposed implicit matrix-vector product method. Consequently, the Gauss-Newton method can be rea-lized in a matrix-free fashion, reducing the compu-ter memory and disk occupancy significantly. Numerical tests demonstrated that the proposed GRA-based traveltime tomography can estimate the near-surface velocity model with high resolution and at a fast convergence rate.

Keywords：finite-frequency traveltime sensitivity kernel, generalized Rytov approximation, first-arrival traveltime tomography, Gauss-Newton algorithm

1. Wave Phenomena Intelligent Inversion Imaging Group(WPI), School of Ocean and Earth Science, Tongji University, Shanghai 200092 China

2. Institute for Advanced Study, Tongji University, Shanghai 200092 China

3. Modeling and Imaging Laboratory, University of California, Santa Cruz, California 95060, USA

Low-amplitudestructuralimagingbasedonseismicnumericalsimulationandphysicalmodeling.TIANYancan1,2,SHIWenwu3,WANGGuoqing1,2,XUZhonghua1,2,andJIANGChunling1,2.OilGeophysicalProspecting,2021,56(1):100-108.

Exploration cases show that low-amplitude structures can form small but high-yield reservoirs with favorable conditions of source-reservoir-cap and oil-gas migration. However, traditional me-thods of velocity modeling cannot eliminate lateral variation of velocity usually caused by the varying proportion of sand in shallow formations and produce a velocity model with low precision, giving rise to poor imaging precision of low-amplitude structures in Matouying area in Jidong oilfield with big drilling deviation. In this paper, the velocity modeling technology for low-amplitude structural imaging based on seismic numerical simulation and physical modeling is proposed. A simplified model that highlights the major characteristic in this area is designed for seismic numerical simulation and physical modeling. Seismic simulation experiments show that correct imaging results can be obtained with the real velocity. At the same time, real velocity and stratigraphic form are used as the prior information for velocity modeling of low-amplitude structures to validate the technology.The results show that the tomographic inversion along the formation can create a precise initial velocity model, and the high-density structure-constrained grid tomographic imaging technology can build a highly precise model that indicates the trend in velocity.Therefore, the joint application of two modeling technologies leads to more precise imaging of low-amplitude structures.

Keywords：physical simulation, numerical simulation, low amplitude structure, velocity modeling, grid tomography

1. Key Laboratory of Reservoir Description, CNPC, Lanzhou, Gansu 730020, China

2. Research Institute of Petroleum Exploration & Development-Northwest, PetroChina, Lanzhou, Gansu 730020, China

3. Exploration and Development Research Institute, Jidong Oilfield Company, PetroChina, Tang-shan, Hebei 063000, China

Researchonmodelingandimagingofdeeplow-amplitudestructuresinthenorthwesternSichuanBasin.WANGYanxiang1,SUQin1,LEXingfu1,ZHANGJunduo1,LIUWei1,andYUANHuan1.OilGeophysicalProspecting,2021,56(1):109-117.

To disclose the details of structural traps in the Longmenshan piedmont belt in the northwestern Sichuan Basin, we carried out imaging and modeling of deep low-amplitude structures, mesh-tomography-based velocity modeling, and physical simulations of structures. Longmenshan piedmont belt presents severe surface fluctuations and serious problems of static correction. Then adaptive first-arrival tomographic inversion with micro-logging constraints is adopted to solve the problem of static correction that affects the imaging of low-amplitude structures. Besides, excessive dynamic calibration is avoided by anisotropic dynamic correction at large offsets, improving the accuracy of velocity analysis in shallow and steeply-dipping strata and deep low-amplitude structures. More-over, the near-surface velocity model of constrai-ned tomography and the mid-deep velocity model obtained by layered tomography are combined to guide the construction of an initial velocity model based on full use of well-logging data. Mesh tomography is used to optimize the velocity model and perform pre-stack depth migration. Therefore, seismic interpretation is supported by accurate and reliable data.

Keywords：northwestern Sichuan Basin, deep low-amplitude structures, adaptive micro-logging constrained tomographic static correction, layered mesh tomography, structure physical simulation

1. Research Institute of Petroleum Exploration & Development - Northwest，Petrochina, Lanzhou，Gansu 730020,China

3D3CVSPimagingtechnologybasedonscatteringtime-distancerelationship.YANYuanyuan1,QINLi2,LUOKun2,LUJun1,andWANGYun1.OilGeophysicalProspecting，2021，56(1)：118-126.

Acquired by special recording geometry, 3D3C (three-dimensional and three-component) VSP data can’t be processed by conventional processing methods for surface seismic data. In addition, there are less applicable 3D3C-VSP imaging technologies with high precision, and there is less processing software for 3D3C-VSP data. To process 3D3C-VSP data, many iterations should be run for velocity analysis, and it is very difficult to get 3D imaging results. In this paper, we propose a method for velocity analysis and migration for 3D3C-VSP data. It is based on wave scattering theory and without iterating. First, we set up the time-distance curve equation of reflected wave in VSP data. Based on this equation, we move the amplitude at each time sample of the seismic trace to the corresponding reflective surface and extract the common-imaging-point gathers by interference stacking. Then, for the asymmetric ray-paths and inconsistent migration velocities of up-going and down-going waves, we convert the NMO corrected up-going waves in common-imaging-point gathers to virtual offset gathers with the characteristics of hyperbolic time difference. Eventually, through velocity analysis and NMO correction for the virtual offset gathers, we obtain high-precision PP and PS wave imaging profiles. This method was applied to the 3D3C-VSP seismic data of Well A in Sichuan Basin, and produced better imaging results than the conventional VSP-CDP stacking method.

Keywords：3D3C-VSP, common-imaging-point ga-ther, virtual offset, migration, PS-wave

1.School of Geophysics and Information Technology, China University of Geosciences (Beijing), Beijing 100083, China

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

Investigationtoshear-wavevelocitypredictionme-thodfororganic-richrock.LIUZhishui1,LIUJunzhou2,DONGNing2,BAOQianzong1,WANGZhenyu2,andSHILei2.OilGeophysicalProspecting,2021,56(1):127-136，154.

Aiming at the disadvantage that the influence of kerogen distribution and pore structure on velocity is not taken into account in the rock physics model of organic-rich rock,we present a method for S-wave velocity prediction of organic-rich rock by integrating the rock physics model(Kuster-Toksöz) with a nonlinear global optimization algorithm. In this method, the organic-rich rock is equivalent to a mixture of minerals, kerogen particles and fluid-containing pores, in which kerogen particles and pores are both equivalent to ellipsoid-shaped inclusions. The effect of kerogen distribution and pore shape on S-wave velocity is described according to the change of the aspect ratio of ellipsoids. The error between the predicted and mea-sured P-wave velocities is applied to establish the inverse objective function. Then calculate two parameters, the equivalent kerogen particles and the equivalent pore aspect ratios, by the optimization algorithm. The S-wave velocity is predicted based on the inverted parameters. Compared with three single-adaptive parameter methods commonly used in the industry, the new method of S-wave velocity prediction based on kerogen and pore aspect ratios simultaneously inverted from P-wave velocity (or P-and S-wave velocities) is more effective.

Keywords：effective pore aspect ratio, effective kerogen aspect ratio, S-wave velocity prediction, rock physics model, organic-rich shale

1.College of Geology Engineering and Geomatics, Chang’an University, Xi’an, Shaanxi 710054, China

2.Research Institute of Petroleum Exploration and Development, SINOPEC, Beijing 100083, China

AVOresponsecorrectionconstrainedbylow-frequencycomponentsintime-frequency-spacedomain.ZHANGShengqiang1，ZHANGZhijun1，GUOJun1，andTANHuihuang1.OilGeophysicalProspecting，2021，56(1)：137-145.

Due to the presence of the Neogene unconsolidated clastics with high porosity and complex fault systems, absorption and attenuation to seismic signals is very strong, resulting in serious attenuation of high-frequency energy and poor amplitude-preserved property of pre-stack gathers in the Bohai Bay Basin. Specifically, the AVO responses of pre-stack gathers are consistent with the corresponding synthetics at low frequency, but not at high frequency, so that the pre-stack hydrocarbon detection techniques based on AVO response cannot be effective. An AVO response correction method in time-frequency-space domain is proposed. It is constrained by low-frequency components, and combines high-resolution time-frequency analysis technology. Following the AVO trend, the method takes better low-frequency energy as a reference, and builds a data-driven three-dimensional time-frequency-space-domain correction factor according to the relationship between the AVO response trend of pre-stack gathers at low frequency and pre-stack gathers at other frequency. The correction factor can make up for the influence of the high-frequency attenuation difference between far and near offsets on AVO analysis, and effectively improve the AVO amplitude-preserved property of pre-stack gathers. Applications to synthetic data and actual data show that the AVO response correction method we proposed can restore the real AVO law of pre-stack gathers and provide reliable basic data for the application of pre-stack hydrocarbon detection methods in Bohai Oilfield.

Keywords：hydrocarbon detection, pre-stack gather, AVO response correction, time-frequency-space domain, complex spectral decomposition, absorption and attenuation

1. Bohai Petroleum Research Institute,Tianjin Branch, CNOOC China Limited, Tianjin 300459, China

Reservoirsensitivefluidfactorinversionanditsapplicationinhydrocarbondetection.WANGDi1,ZHANGYiming1,NIUCong1,ZHANGYuhua1,andHANLi2.OilGeophysicalProspecting,2021,56(1):146-154.

Deep-water turbidite sandstone reservoirs in the Niger Delta Basin have a great potential of oil and gas exploration. Drilling data from this area indicate that both oil and water layers with high porosity show “bright spots” and classes II-III AVO anomaly. However, it is difficult to distinguish oil from water using conventional amplitude and AVO attribute. In this study, we firstly investigated amplitude responses and their controlling factors, and found that high porosity is the key factor controlling the multiple solutions of amplitude and AVO response; then we performed a series of AVO modeling with different porosity and fluid substitution, and output the intercept-gradient crossplot whose variation with the porosity is different from that with the fluid saturation, and based on this conclusion, we introduced an exten-ded AVO attribute by rotating the axis, which can improve the precision of hydrocarbon detection; finally, after applying pre-stack inversion to further enhance the realiability of hydrocarbon detection, a new fluid factor quantitative evaluation method was proposed for selecting the best sensitive fluid factor while suppressing the effect of porosity. According to our nalysis,λ/μis the most applicable parameter for detecting hydrocarbon in this area. And application to real data has proved that both the extended AVO attribute andλ/μcould effectively distinguish water layers from oil layers with “bright spots”. The predicted results are well consistent with the drilling data, indicating the feasibility of the method.

Keywords：Niger Delta Basin, turbidity sandstones, bright spots, axis rotation, extended AVO attri-bute, fluid factor, hydrocarbon detection

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

2.CNOOC International Ltd., Beijing 100028, China

Combinationandapplicationofdetectingtechnologyforlateraldiscontinuityofsandstonereservoir.FANTing’en1，ZHANGJingyu1，WANGHaifeng1，ZHANGXianwen1，andDUXin1.OilGeophysicalProspecting，2020，56(1)：155-163.

At present, the detection of lateral discontinuity of seismic signals mostly aims at fault detection, recognition and automatic interpretation, involving edge-preserving filtering technology, coherent cube technology, ant tracking technology, etc., which have been researched and used widely, but less researches on the detection of small-scale discontinuous seismic signals. As a third-generation coherent algorithm, local structural entropy has better noise resistance and higher resolution than the previous two generations of coherent algorithms, and has higher detecting accuracy for small discontinuous structures inside sandstone reservoirs. Taking local structural entropy as a core, and edge-preserving filtering and ant tracking attribute enhancement as support, a technology combination are formed for detecting the lateral discontinuity of sandstone reservoir. The process includes interpretative processing of seismic data, calculation of coherent attributes, characterization and optimization, and result verification and application. The application to the target sandstone unit in the H oilfield in the South China Sea shows that the results provided by the technology combination are consistent with the mudstone encountered by a horizontal well, proving the effectiveness of the technology combination.

Keywords：reservoir discontinuity, local structural entropy, edge-preserving filtering, ant tracking,attribute enhancement

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

Nonlinearmulti-parameterhybridinversionofpre-stackandpost-stackseismicdatabasedonZoeppritzequation.ZHANGLingyuan1,ZHANGHongbing1,SHANGZuoping2,YANLizhi1,andRENQuan1.OilGeophysicalProspecting,2021,56(1):164-171.

At present, Shuey or Gray approximate formula and basis pursuit theory are widely applied in inversion of Poisson ratio. The same is true for the velocity ratio which is calculated from P- and S- waves velocity obtained from pre-stack inversion. But it is not avoidable to reduce the accuracy of inversion parameters. We try to apply Zoeppritz equation in inversion for velocity ratio and Poisson ratio. First we developed a hybrid pre-stack and post-stack seismic inversion workflow by the convolution model based on the reflection coefficient of vertical incidence and Zoeppritz equation. Then we constructed a new objective function which includes edge-preserving regularization to reduce the adverse effect of inversion unsuitability. We applied a fast simulated annealing algorithm to achieve global non-linear optimization. To reduce the influence from the magnitude difference of multiple parameters and the poor stability of multi-parameter simultaneous random search, we used an improved Fatti equation in inversion, which introduces two fitting formulas, a density and P-wave velocity formula, and a S-velocity and P-wave velocity formula, to improve the accuracy and stability of inversion. The inverted results of field data indicate that the velocity ratio obtained by direct Zoeppritz equation inversion is better than that obtained by approximate formulas. The inversion results from the new workflow have good consistency with the logging data. In a word, post-stack inversion obtained more accurate P-wave velocity and density, while the accuracy of velocity ration obtained by pre-stack inversion is higher than that obtained by post-stack inversion, and the inversion of 18°～24° angular trace gathers is better than that of 3°～9° and 33°～39° gathers.

Keywords：pre-stack seismic inversion, hybrid inversion, Zoeppritz equation, edge-preserving regularization, velocity ratio, gas zone identification

1. College of Earth Science and Engineering, Hohai University, Nanjing, Jiangsu 211100, China

2. College of Mechanics and Materials, Hohai University, Nanjing, Jiangsu 211100, China

InfluenceofstructuralcharacteristicsonoilandgasdistributioninblockAoftheMugladBasin.ALIYEVAGunay1，HUANGZhilong1，JINZhenkui1，NIEQihai2，ZHANGXiuqiang2，andZHAOGang2.OilGeophysicalProspecting,2021,56(1):172-180.

Existing research results cannot fully support the exploration and development of block A in the Muglad Basin due to the special tectonic positions, complex faults and diversified structural patterns in this area. In this regard, structural zoning is re-divided to study the active periods of faults in depth. In addition, the activity differences among structural units are compared and the tectonic evolution history is restored to clarify the relationship between tectonic evolution and hydrocarbon accumulation. Results demonstrate the following points: ①The tectonic evolution of block A has experienced pre-rift, syn-rift and post-rift periods. Then the syn-rift period was further divided into three tectonic stages, namely the Early Cretaceous to the Late Cretaceous, the Late Cretaceous to the Paleocene and the Eocene to the Pliocene, controlling source rocks, structural traps and hydrocarbon distribution, respectively. ②Block A, from east to west, can be divided into two tectonic zones: an inheritably stable zone and a continuously active zone. The former is typical of the east and west slopes of the Kaikang trough, with a “strong-strong-weak” fault pattern during evolution, while the latter is typical of the Kaikang trough, with a “strong-strong-the strongest” pattern. ③Two patterns of hydrocarbon migration and accumulation in block A are presented, namely the deep play and the late-adjustment play. They are common in the east and west slopes of the Kaikang trough with a low level of fault activities and the Kaikang trough with multi-layer reservoirs and continuous fault activities, respectively. The above research can support oil and gas exploration in basins with similar geological settings.

Keywords：block A of the Muglad Basin, fault movement, structural characteristics, hydrocarbon migration, hydrocarbon accumulation system

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

2.GRI, BGP Inc., CNPC, Zhuozhou, Hebei 072750, China

WFEM3Dfiniteelementnumericalsimulationbasedonvectorpotentialandscalarpotential.ZHOUYinming1,2,3,WANGJinhai4,HUXiaoying3,HEZhanxiang5,andXIONGBin6.OilGeophysicalProspecting,2021,56(1):181-189.

The wide-field electromagnetic method (WFEM) puts forward a formula suitable for calculating apparent resistivity in the whole area. It fundamentally breaks through the shackles of the “far-field area” theory, effectively expands the observing range and depth of the artificial source electromagnetic method, and improves field observing accuracy and efficiency. Considering that the underground medium is a three-dimensional structure, a WFEM 3D finite element numerical simulation method was developed from the perspectives of vector potential and scalar potential based on Coulomb criterion. This method overcomes the problems of “pseudo solution” and discontinuity of model boundary in the calculation of electromagnetic field distribution directly based on Maxwell equations, and avoids more complex edge finite e-lement forward simulation. In addition, the background field is solved by a quasi analytical method, and the secondary field is solved by a finite element method, which overcomes the singularity of local shooting field. In a model example, a prism model was designed, and the analytical solution to the homogeneous layered medium was used to verify the correctness and accuracy of the method. Then the detection abilities of WFEM and CSAMT to typical 3D objects were compared and analyzed by using the forward algorithm. The results show that under the same conditions, WFEM can reflect the information of underground objects more accurately and has a higher resolution.

Keywords：WFEM, vector potential, scalar potential, 3D finite element, numerical simulation

1. School of Geosciences and Info-physciences, Central South University, Changsha, Hunan 410083, China

2. Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring(Central South University), Ministry of Education, Changsha, Hunan 410083, China

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

4. Bureau of Geological Exploration & Deve-lopment of Qinghai Province, Xining, Qinghai 810029, China

5. SUSTech Academy for Advanced Interdisciplinary Studies, Shenzheng, Guangdong 518055,China

6. College of Earth Sciences, Guilin University of Technology, Guilin, Guangxi 541006, China

Amultiplelevel-setmethodfor3Dboundaryinversionofmagneticdata.XIAOXiao1,2,3,4,DUANYa-ting3,HUShuanggui3,TANGJingtian1,2,3,4,XIEYong5,andLIUChangsheng5.OilGeophysicalProspecting,2021,56(1)：190-200，208.

Present algorithms for inverting the boundaries of magnetic targets use only two level sets. There are usually multiple magnetic geological bodies with different susceptibility in actual exploration. This paper proposes a new multiple level-set inversion algorithm for 3D inverting the position and geometry of magnetic targets with known susceptibility. First, based on the principle of multiple level sets, an objective function based on a multiple level sets function is established. Then, during the inversion process, an arbitrary tetrahedron element magnetic analytical solution algorithm is used for high-precision forward calculation, and the physical property mapping between forward and inverse grids is used to realize the independent operation of the forward grids and the inverse grids. The weighted essentially non-oscillatory scheme (WENO) is introduced to update and reinitialize the level set function, thereby improving the reliability and efficiency of the inversion. Finally, the effectiveness of the algorithm is verified by theoretical models with various numbers of level sets. The results show that the inversion based on the multiple level-set method has strong flexibility, and can automatically merge and separate regions in the model, thereby changing the model's topology with no need to manually reparameterize it. The accuracy of the inversion results is improved, and the boundary of the anomalous body obtained from the inversion accords well with the true boundary.

Keywords：multiple level-set, magnetic boundary inversion, susceptibility

1. Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring (Central South University), Ministry of Education, Changsha, Hunan 410083, China

2. Hunan Key Laboratory of Nonferrous Resources and Geological Hazards Exploration, Changsha, Hunan 410083, China

3. School of Geosciences and Info-physics, Central South University, Changsha, Hunan 410083, China

4. Technical Innovation Center of Coverage Area Deep Resources Exploration, Ministry of Natural Resources, Changsha, Hunan 410083, China

5. Changsha Aeronautical Vocational and Technical College, Changsha, Hunan 410124, China

Aquasi-3DTEMinversionbasedonlateralconstraints.YANGYunjian1,2,WANGXuben1,LIUXuejun2,HEZhanxiang3,MIXiaoli2,andTANGBiyan2.OilGeophysicalProspecting,2021,56(1)：201-208.

One-dimensional inversion is one of the main methods for processing of TEM data, but usually there is poor continuity on one-dimensional inversion profile. The principle of lateral constraint is introduced to construct quasi-3D transient electromagnetic method (TEM) inversion. Considering that the actual 3D measuring grid is often not strictly regular, the quasi-3D TEM data inversion is constructed with the distance-weighted adjacent stations lateral constraint, and the Gauss-Newton method is used to solve the inversion equation, which can not only effectively suppressed the irrational model mutation generated by single station 1D inversion, also has no the requirement on the regularity of the 3D measuring grid. The tests of the synthetic data and field data verified the effectiveness of this method.

Keywords：TEM, lateral constraint, quasi-3D inversion, irregular measuring grid

1. College of Geophysics, Chengdu University of Technology, Chengdu, Sichuan 610059, China

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

3. SUSTech Academy for Advanced Interdisciplinary Studies, Shenzhen, Guangdong 518055, China

Forwardmodelingof3DDCresistivitybasedonhigh-orderadaptivefiniteelementanditsapplicationinQinshuiBasin.ZHAONing1,2,HUANGMingwei1,SHENYahang1,TAODeqiang3,andQINCe1.OilGeophysicalProspecting,2021,56(1):209-216.

Considering the fact that the accuracy of an adaptive finite element solution is mainly affected by the cell size (h) and the order of a shape function (p), we should adaptively refine the meshes to be dense enough or apply a higher-order shape function in the meshes to acquire a high-accuracy finite element solution. However, this will greatly increase the time burden and require sufficient me-mory space of the computer. In light of these problems, in this paper, we combine an h-adaptive refinement algorithm with a higher-order shape function for the forward modeling of a 3D DC resistivity model. In the program, we use the tensor pro-duct of 1D polynomials to generate a shape function of any order in the 3D space and apply Kelly posteriori error estimation to guide the adaptive refinement of meshes. Numerical examples show that in the case ofp=3, our program has high accuracy, and the convergence of the finite element solutions is faster than that in the case ofp=1 or 2. This means that the most accurate finite element solution with the fewest degrees of freedom can be obtained when the order of the shape function is 3. Finally, the simulation of 3D DC resistivity is performed on the 3D modeling of a coalbed fracturing monitoring region in the southern Qinshui Basin, verifying the effectiveness of our program.

Keywords：DC resistivity method, finite element based forward modeling, high-order shape function, adaptive

1.School of Physics & Electronic Information Engineering, Henan Polytechnic University, Jiaozuo, Henan 454150, China

2.State Key Laboratory Cultivation Base for Gas Geology and Gas Control, Jiaozuo, Henan 454150, China

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