ABSTRACTS

Electronicdelayer-controlledthree-stageseismicsourcessuperposition.LIUHaiping1,LIUHuimin2,andWANGHaitao3.OilGeophysicalProspecting,2021,56(2)：217-225.

Multi-stage delay stacked source is advanced for seismic data acquisition in thick loess area. However, using a mechanical striker or a screw-type detonating tube to delay the explosion of seismic sources, there are many problems, such as low accuracy of time delay, high cost, complex structure and low efficiency. Although such delay can be achieved by using an advanced electronic detonator, it has not been widely used because of high cost. In this paper, we introduce a method of controlling the delayed explosion of seismic sources using an electronic delay technology, and give the schematic of the control circuit on the surface (referred to as a high-voltage encoder) and the delay circuit under the ground (referred to as a delayer). It is a new type of explosion system controlled by a multi-stage seismic source circuit. Taking the exploration project in Shanxi Province, we tested the system. The results show that the system can stably control three explosive columns to detonate from top to bottom accurately. The distance between the explosive columns is 1.7m, and the detonating time differs by 3ms. Keeping hole depth, charge size, receiver patch and folds constant, from the point of shot static correction record, the effective reflection energy of vertical delay three-magnitude source is strong than that of plane three-combination source in, the frequency band of the former is wider, the surface wave of the former is weaker and the exploration cost of the former is lower than the latter; from the point of seismic time section, the effective reflection energy and frequency band of vertical delay three-magnitude seismic source are equivalent to plane five-combination seismic source，and the cost of the former is reduced by about 70% to 80% on the latter. The multi-stage delay stacking method is useful for exploring coal and oil resources in the thick loess area.

Keywords: multi-stage delay stack source, electronic time delay technology, effective reflected wave, energy, frequency band

1. Coal Geological Geophysical Exploration Surveying & Mapping Institute of Shanxi Province, Jinzhong, Shanxi 030600, China

2. Geological Technology Limited Company of Shanxi Jindiyuan, Jinzhong, Shanxi 030600, China

3. Houma Electric Depot of Daqin Railway Limited by Share Led, Houma, Shanxi 043000, China

Researchonthestaticcorrectionmethodfortheloessandgravelareainwesternfoothillbelt.CUIQinghui1，SHANGXinmin1，TENGHouhua1，JINChangkun1，ZHAOShengtian1，andSONGGuiqiao2.OilGeophysicalProspecting，2021，56(2)：226-233.

In the study area located in the foothill belt in the southern edge of the Junggar Basin, the near-surface structure is the mixture of loess and gravel which changes drastically. Therefore, conventional near-surface modeling and static correction is less effective. In order to build a model of the near-surface structure for shot depth design and static correction, first we studied the characteristics of the near-surface zone in the foothill belt based on the uphole data, and summarized the near-surface velocity and time-depth relation. Then we proved the loess layer and the gravel layer have a good spatial autocorrelation in thickness by geostatistics, modeled the thickness of loess and gravel by selecting a best Kriging interpolation, and correct the model by the similarity coefficient of the bottom of loess or gravel and the surface to improve the modeling accuracy. Finally，we proposed a layered time-depth curve static correction and applied it to real data. The efficiency and effect of this static correction method are better than conventional methods.

Keywords：foothill belt, loess and gravel area, near-surface modeling, time-depth relation curve, static correction

1. Geophysical Research Institute, Shengli Oilfield Branch Co., Sinopec, Dongying, Shandong 257022, China

2. Exploration & Development Affairs Department, Sinopec, Beijing 100728, China

SuppressionofrandommicroseismicnoisebasedoncompleteensembleempiricalmodedecompositionwithadaptivenoiseofTFPF.CHENYijun1,CHENGHao1,GONGEnpu1,andXUELin1.OilGeophysicalProspecting,2021,56(2):234-241.

Microseismic monitoring is widely applied in unconventional oil and gas fields, and supports the production and reserve increase of oil and gas fields. Because microseismic data are non-stationary, conventional denoising methods are not effective. This paper proposes a time-frequency peak filtering (TFPF) method of adaptive white noise based on the sample entropy (SE) complete set of empirical mode decomposition (CEEMDAN) to suppress noises while preserving effective signals. First raw microseismic data are decomposed into several IMFs of intrinsic modal components by CEEMDAN. Then after calculating the sample entropy, the IMFs are divided into two groups — one group will be filtered and the other will be left alone. The former group is TFPF filtered after selecting filter windows, and reconstructed with the latter to get final filtered signals. Application to theoretical model and field data has shown that the noise suppression method proposed in the paper is more effective than traditional EMD and constant-window TFPF denoising methods.

Keywords：microseism, denoising, time-frequency peak filtering (TFPF), empirical mode decomposition (EMD), sample entropy (SE)

1. Key Laboratory of Ministry of Education on Safe Mining of Deep Metal Mines, Northeastern University, Shenyang, Liaoning 110819, China

AmethodofsignalextractinginlowSNRseismicdata.XUZilong1，SONGLin1，andXIAHongrui2.OilGeophysicalProspecting，2021，56(2)：242-248.

The combination of improved wavelet threshold denoising and stacking technologies is useful to extract signals from low SNR seismic data. After wavelet transform, the spatial-temporal threshold of the wavelet coefficient of each seismic trace is calculated in every wavelet scale. In order to diminish over-smoothing or over-preserving in classic wavelet threshold denoising, both median filtering and stacking are used as constraints in denoising. Classic wavelet threshold denoising is modified based on the above method to realize wavelet threshold denoising. Finally, denoised seismic data can be obtained after wavelet reconstruction. Using multiple wavelets to repeat the above operations, multiple denoised traces can be obtained. These traces can be stacked to extract signals from low SNR data. Both theoretical and real data show that this method is able to achieve accurate extraction of signals from low SNR seismic data, and the processing result is significantly better than conventional methods.

Keywords：low SNR data, signal extracting, median filtering, improved wavelet threshold denoising, suppressing noise, stacking technology, enhancing signal

1. Sinopec Geophysical Research Institute, Nanjing, Jiangsu 211103, China

2. Geophysical Research Center, Geophysical Department, Jianghan Petroleum Administration Bureau, Sinopec, Qianjiang, Hubei 433100, China

EstimationofresidualphaseofseismicdatabyusingthephaseshiftcharacteristicsofRickerwavelet.LIUChuanqi1,ZHOUJianke1,LIBin1,andWANGTeng1.OilGeophysicalProspecting,2021,56(2):249-256.

Zero-phase seismic data has the advantages of high resolution and easy interpretation. However, the deconvolution processing under the assumption of minimum phase is difficult to ensure that the seismic data is zero-phase, so it is necessary to estimate the residual phase through subsequent technical means for zero-phase correction. By studying the phase-shift characteristics of Ricker wavelet, it is found that there is a time difference at the peak time of Ricker wavelet with different dominant frequencies after rotating the same phase, which has a linear relationship with phase rotation. Under the assumption of constant phase, the seismic data with unknown phase are filtered twice, and the time difference of the same reflecting interface is calculated based on the filtered data. Then the phase estimation is realized by the linear relationship between the time difference and the phase. This method is free from the restriction of discriminant criterion. The feasibility and effectiveness of the method have been verified by synthetic records and real data.

Keywords：Ricker wavelet, constant phase, phase rotation, peak time difference, residual phase estimation

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

Seismicdatade-noisingmethodbasedonVMDintime-frequencydomain.HURuiqing1，HEJunjie1，2，LIHuafei1,ZHANGXiaoli1,PEIJiading1,andLIUYiwei1.OilGeophysicalProspecting,2021,56(2):257-264.

Strong noise interference is the primary factor that causes poor imaging of deep seismic data. A new idea applies a variable mode decomposition algorithm to noise suppression. Firstly, the analytical signals of seismic data are constructed by Hilbert-Huang transform (HHT), then the seismic data are converted into time-frequency domain where time-frequency slices are decomposed as instrinsic mode functions (IMFs) by the variable mode decomposition algorithm；then the energy distribution of effective signals and noises on the time-frequency slices is analyzed，and the time-frequency slices are reconstructed by the effective IMFs; and finally the slices are transformed back to the space-time domain to achieve the goal of noise suppression. The control of key parameters on the denoising effect of the algorithm has been analyzed on model data. The results of actual data have verified that the algorithm can effectively suppress strong random noises,and it is also effective for suppressing linear noises.

Keywords：variable mode decomposition, time-frequency analysis, noise attenuation

1. Korla Branch, Geophysical Research Institute, BGP Inc., CNPC, Korla, Xinjiang 841001, China

2. College of Earth Science and Engineering, Xi’an Shiyou University, Xi’an, Shaanxi 710065, China

Double-phase-shiftfilteringmethodforeliminatingharmonicdistortioninprocessingslip-sweepvibroseismicsignals.LIBolin1，WANGYanchun1，LIUXueqing2，ZHANGJingdong3，LIYunzhu4，andSUNTingbin3.OilGeophysicalProspecting，2021，56(2)：265-272.

Although simple and efficient for eliminating harmonic distortion of vibroseismic data, the single-phase-shift filtering method is less effective for dealing with complex signals such as white noise and signals recorded by slip-sweep vibroseis survey, so it failed to be put into actual application. In view of this situation, we propose a double-phase-shift filtering method. It manipulates and polishes the original phase shift from the traditio-nal phase shift method, and at the same time carries out another phase shift to process the signal that the single-phase-shift filtering method cannot process. Applied on model data, the double-phase-shift method can suppress harmonic distortion more effectively than the single-phase-shift me-thod. Then we applied the double-phase-shift filtering method to slip-sweep signals, and analyzed the applicable conditions. In view of the fact that raw seismic data are composed of various reflections with different intensity, a subsurface model with multiple reflections should be established, and the frequency ratio of signals should be consi-dered in order to discuss the applicability of the double-phase-shift method. Applied for raw data and compared with the single-phase-shift filtering method, the double-phase-shift method is better in suppressing harmonic distortion and worthy promoting.

Keywords：double-phase-shifts,harmonic distortion,harmonic elimination, slip-sweep technique, vibroseis

1. School of Geophysics and Information Techno-logy, China University of Geosciences(Beijing), Beijing 100083, China

2. Beijing Jingneng Oil & Gas Resources Development Co. Ltd, Beijing 100022, China

3. BGP Geological Research Center, CNPC，Zhuo-zhou, Hebei 072751, China

4. CCDC Geological Exploration & Development Research Institute, CNPC, Chengdu, Sichuan 610051, China

Multi-channelabsorptioncompensationmethodbasedonglobalconstrainedinversion.LIUJintao1，WANGXiao1,WANGXiaowei1,SUQin1,LIUMengli1,andYUANHuan1.OilGeophysicalPro-specting，2021，56(2)：273-282.

Inverse Q filtering based on inversion is a commonly used method of seismic wave absorption compensation. This method improves the stability of inversion by using L2norm regularization as a vertical constraint. However，it is easily disturbed by noises, and noises would reduce the accuracy and stability of inversion. This paper proposes a global-constrained absorption compensation me-thod based on inversion. The method first calculates an error operator to predict seismic signals, and then adds it as a lateral-constrained operator to the vertical-constrained absorption compensation system. This expends the single-trace approach to a multi-trace algorithm. The inversion-based global-constrained absorption compensation method has stronger noise suppresion and stability, and it can effectively reduce the amplification effect of compensation on high-frequency noises, improve the seismic resolution and preserve the spatial conti-nuity of seismic event. Tests on model data and actual data have confirmed the effectiveness of this method.

Keywords：absorption compensation, vertical-constrained, global-constrained, error operator for signal prediction

1. Northwest Branch, Research Institute of Petroleum Exploration & Development, PetroChina, Lan-zhou, Gansu 730030, China

Animprovedone-step3DCRSstackingmethodbasedonhybridoptimizationalgorithm.SUNXiao-dong1，2，HOUMengrui1,RENLijuan3,WANGWeiqi1,andLIZhenchun1.OilGeophysicalProspecting,2021,56(2):283-288.

Common reflection surface (CRS) stacking makes full use of seismic data within the range of a Fresnel radius, and maximizes the signal-to-noise ratio without reducing the resolution. It is a po-werful imaging means of the seismic data with low signal-to-noise ratio. At the same time, CRS stacking takes into account the inclination and local curvature of the underground reflector, so the imaging accuracy is higher. In conventional three-dimensional CRS stacking, the accuracy of the eight parameters obtained sequentially in multiple steps is low, and affects the final stacking effect. We propose a one-step three-dimensional CRS stacking strategy which combines the rapid search of genetic algorithm and the global convergence of simulated annealing algorithm. It uses a multi-group hierarchical hybrid parallel optimization algorithm that mixes the two algorithms, that is, a thermal slot method is used to generate initial population in the top layer, the middle layer executes a parallel genetic iterative algorithm to achieve population evolution, and the bottom layer uses a simulated annealing algorithm to achieve global optimization, and the genetic and simulated annealing hybrid algorithm is used for hierarchical parallel computing. The CRS stacking method and the design of the hybrid algorithm significantly improve the optimization cost and accuracy of parameters. Tests on real data have verified the practicability of the one-step 3D CRS stacking method.

Keywords：common reflection surface stack, para-meter coupling, genetic algorithm, simulated annea-ling, parameter optimization

1. Key Laboratory of Deep Oil & gas, China University of Petroleum(East China), Qingdao, Shandong 266580, China

2. Shandong Provincial Key Laboratory of Reservoir Geology, China University of Petroleum(East China), Qingdao, Shandong 266580, China

3. Nanhai Western Petroleum Research Institute, Zhanjiang Branch, CNOOC, Zhanjiang, Guangdong 524000, China

ComparisionofazimuthalanisotropyofP-waveandS-wavevelocitybasedonwideazimuthseismicdata.HEFeng1,LONGFan2,HANGang1,CHENXiao-zhi1,LIJun2,andWANGHuandi3.OilGeophysi-calProspecting,2021,56(2):289-294,301.

Comparation of the azimuthal anisotropy of P-wave and S-wave velocity is very important for determining the sensitive parameters to azimuthal a-nisotropy. Based on prestack gathers in OVT domian in area B,Easten China, anisotropy inversion of P-wave and S-wave velocity was carried out to determine the anisotropic intensity and oritentation. The anisotropy of S-wave velocity is consis-tent with the fracture data (undeveloped fractures) revealed by known well data, such as logging data, and satisfies the relationship between faults and fractures (where faults are developed near the zone with high anisotropic intensity, and the anisotropic direction is roughly parallel to the fault strike). The anisotropy of P-wave velocity is obviously weaker than that of S-wave velocity, and does not satisfy the relationship between faults and fractures in the area, i.e. the anisotropy intensity is small around the fractures, or anisotropy oritentation is not approximately parallel to the fractures, or the two cross with acute angle. So,it is concluded that S-wave velocity is more sensitive to azimuthal anisotropy and more reliable for fractures prediction.

Keywords：P-wave velocity, S-wave velocity, azimuthal anisotropy, fracture, inversion of elastic parameters, ellipse-fitting

1. CNOOC Research Institute, Beijing 100027, China

2. Shanghai Borui Energy Technology Company, Shanghai 202150, China

3. Petroleum Industry Press, Beijing 100029, China

Implementationandapplicationofcompressedsensingalgorithmforseismicspectruminversion.XIAHongmin1,LIULanfeng1,ZHANGXianhui2,3,andCHENShuangquan2,3.OilGeophysicalProspecting,2021,56(2):295-301.

Seismic spectrum inversion is a method of decomposing the reflection coefficient into odd and even components, and then using part of spectrum information for inversion. In this study, a spectral inversion method based on compressed sensing algorithm is developed in conjunction with the quadratic spectral calculation method of seismic data. Seismic spectrum inversion is carried out by the compressed sensing algorithm that implements gradi-ent projection and sparse reconstruction, and better improves the stability of conventional spectrum inversion algorithm. At the same time, the second spectrum of seismic data is used to estimate the seismic wavelet spectrum in the inversion, which well realizes the application of spectrum inversion of seismic data. Application to synthetic and raw data has proved that the spectrum inversion method based on compressed sensing algorithm is very adaptable and practicable. In raw data processing, this method is less influenced, and its inversion result has high signal-to-noise ratio and good lateral continuity.

Keywords：spectrum inversion, compressed sensing, sparse reconstruction, seismic wavelet, high resolution

1.SINOPEC Petroleum Exploration and Production Research Institute, Beijing 100083,China

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

3.CNPC Key Laboratory of Geophysical Prospecting, China University of Petroleum, Beijing 102249, China

FullelasticwaveforminversioninLaplace-Fourierdomainbasedontimedomainweighting.LIUZhangju1,2,3,4,TONGSiyou1，2,FANGYunfeng5,andJIAJunlian6.OilGeophysicalProspecting，2021，56(2)：302-312,331.

Full elastic waveform inversion is a strong nonlinear problem and easy to fall into the local minimum due to strong dependence on the precision of the initial model or the low-frequency component of seismic record, which lead to failed inversion. In this study, we set up full elastic waveform inversion in Laplace-Fourier domain based on time domain weighting. This method reduces the dependence of full waveform inversion on low-frequency components by introducing a Laplace atte-nuation factor into the full waveform inversion in time-Fourier domain, and forms time-Laplace-frequency domain elastic inversion with the advantages of three domains by combining the advantages of time domain and frequency domain. And finally the negative effect of the Laplace attenuation factor is eliminated by the flexible form of weighting seismic records in time domain. Application to model data has proved that, in the absence of low-frequency components, this method can obtain more accurate inversion results with low dependence on the initial model.

Keywords：full waveform inversion, elastic wave, time domain weighting, Laplace-Fourier domain

1. Key Lab of Submarine Geosciences and Prospecting Techniques, MOE, Ocean University of China, Qingdao, Shandong 266100, China

2. Functional Laboratory for Marine Mineral Resources Assessment and Prospecting，Qingdao National Laboratory for Marine Science and Techno-logy，Qingdao，Shandong 266061，China

3.School of Earth Sciences, Zhejiang University, Hangzhou, Zhejiang 310027, Chian

4.Key laboratory of Submarine Geosciences, Se-cond Institute of Oceanography, Ministry of Natural Resoures, Hangzhou, Zhejiang 310012, China

5. BGP Recearch & Development Center，CNPC，Zhuozhou，Hebei 072751，China

6. Geophysical Research Institute，BGP，CNPC，Zhuozhou，Hebei 072751，China

Diffractionwaveseparationandimagingbasedonhigh-resolutionRadontransformonaniterativemodelshrinkingapproach.LUOTengteng1，XUJi-xiang1,andSUNXiping1.OilGeophysicalProspecting,2021,56(2):313-322.

In order to make full use of the diffractions in seismic record to describe small underground structures (faults, fractures, holes, etc.), it is necessary to separate weak diffractions from seismic full-wave field and image diffractions separately. In dip-angle common imaging point gathers (CIGs), a separating and imaging method of diffractions based on high-resolution Radon transform on an iterative model shrinking approach is proposed. This method considers the morphological difference in diffracted and reflected events. It uses an iterative shrinkage threshold algorithm (ISTA) to obtain the sparsity of Radon model through model shrinkage steps in time domain and solve the problem that the separated effect of diffraction wave field caused by poor focusing of reflection and diffraction energy group is not ideal in conventional Radon transform. On the basis of programming algorithm, the method can effectively separate diffracted waves and has strong anti-noise ability on model data. On real data, the computing efficiency of the method is about twice than that of the conventional IRLS at more than 20 iterations, proving it applicable for real seismic data.

Keywords：Radon transform, iterative model shrin-kage, high resolution, dip-angle CIGs, diffraction separation and imaging

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

Geology-orientedintegratedseismicdataprocessingandinterpretationforlithologicreservoirs.ZHANGMing1,SUNXiping1,CUIXingfu1,ZHANGXin1,LILinggao1,andDUWenhui1.OilGeophysicalProspecting,2021,56(2):323-331.

After routine seismic data processing, it is difficult to obtain amplitude-preserved seismic data with high resolution for characterizing lithologic reservoirs due to their small thickness and large lateral changes. A better alternative is scheme geology-oriented integration of seismic data processing and interpretation in the way of “double-helical parallel operation” instead of “serial operation”. The flow includes ① At the stage of project planning, efforts focus on establishing key techniques and workflow for seismic data processing and interpretation in accordance with geologic setting and tasks; ② At the stage of seismic data processing and interpretation, using well-tie calibration and attribute analysis, the effects of data processing on geologic characterization are analyzed and fed back to data processing to improve techniques and parameters; ③ At the stage of reservoir characterization, seismic prediction combined with regional geologic study is carried out for a detailed understanding of lithologic reservoirs. The method emphasizes synchronous processing and interpretation instead of processing followed by interpretation. The integration of seismic data processing and interpretation yields good results in practice.

Keywords：lithologic reservoir, seismic prediction, integrated seismic data processing and interpretation, amplitude preservation, resolution enhancement

1.Research Institute of Petroleum Exploration & Development, CNPC, Beijing 100083, China

Studyonrecognizingthechannelwallofthedeep-watercentralvalleyintheSouthChinaSea.LIAOYi1，LIUWei1,MAGuangke1,DENGHaidong1,andLILei1.OilGeophysicalProspecting,2021,56(2):332-339,355.

With high cost and risk of developing the L gas field in the deep-water central valley in the Qiongdongnan Basin, it is necessary to precisely prove the geological reserves of the gas field. However, the calibrated results from wells A and B to the channel wall are completely opposite, and affect the defination of the structure boundary and the reserves. Based on a variety of data, we studied the sedimentary evolution of the canyon channel wall, determined its seismic response characteristics through forward simulation on three models, and finally confirmed its rationality on field data. The results show that there is a pressure isolating layer with high velocity, high density, high resistivity and high gamma between two different formation pressure systems inside and outside the central canyon channel wall. The top of the isolating layer is the canyon channel wall, and the bottom of the isolating layer is the formation pressure boundary. They are, respectively, the responses of seismic crest and trough. The determination of seismic response characteristics of the channel wall and the formation pressure boundary in the South China Sea is conducive to proving the geological reserves of natural gas and effectively predicting key points of formation pressure before drilling operation in the future.

Keywords：central canyon, channel wall, pressure prediction, forward modeling, pressure isolating layer

1.Hainan Branch of CNOOC Ltd, Haikou, Hainan 524057, China

Applicationof4Dmulti-componentseismicsurveyindynamicreservoirmonitoring.WANGBo1，NIEQihai1，CHENJin’e1,WANGChunyan1,GUOJingru1,andLIUYuan1.OilGeophysicalProspecting,2021,56(2)：340-345.

In the study area located in the Athabasca oil sand region of Canada, the target layers are extremely shallow, the surface conditions are complex, and the time of reservoir development is short. Therefore, the difference of seismic data before and after the reservoir development is small, and it is difficult to do 4D seismic consistency processing, extract reliable reservoir change and record secondary logging data. This limits the popularization and application of 4D seismic inversion. This paper proposes a 4D multi-component joint inversion method based on a low-frequency model. The model takes into account the difference of converted seismic data, adds converted seismic data into the inversion process. By using 4D3C seismic data and fully integrating two frontier seismic exploration technologies (“4D” and “multi-component”), the 4D multi-component joint inversion method realizes fine reservoir characterization and dynamic monitoring. The following understandings are obtained: ①Real-time combination of processing and interpretation and step by step quality control improve the consistency of non-reservoirs, and retains and highlights the real difference of reservoirs; ②rock physics analysis shows that the velocity ratio of compressional and shear wave is most sensitive to reservoir change and can be used as a sensitive parameter for dynamic reservoir monitoring; ③by making full use of time-shift information of compressional seismic data and converted seismic data, the low-frequency model reflecting reservoir variation can be established, which avoids the limitation of secondary logging data. The method is simple and easy to use, and the prediction of steam chamber is accurate and reliable.

Keywords：4D multi-component seismic, data matching, rock physics, joint inversion, velocity ratio of compressional and shear wave, reservoir monitoring

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

Seismictime-frequencyanalysisbasedonVMDandenvelopederivativeoperatorforfractured-vuggyreservoirprediction.WUDi1,SONGWeiqi1,LIUJun2,CHENLihao1,CHENJu’nan2,andYANGZipeng1.OilGeophysicalProspecting,2021,56(2):346-355.

Deep fracture-vuggy carbonate reservoirs are characterized by scattered spatial distribution and quite different internal structures. Time-frequency analysis methods such as Hilbert-Huang transform, ensemble empirical mode decomposition and complete ensemble empirical mode decomposition always encounter problems like modal aliasing and endpoint effect when applied to reservoir prediction. In this paper, we propose a high-accuracy time-frequency analysis method which combines Variational Mode Decomposition (VMD) with Envelope Derivative Operator (EDO). The VMD method can adaptively and non-recursively decompose original seismic signal into a series of band-limited quasi-orthogonal IMFs. The EDO operator with good non-negative characteristics and anti-noise ability can calculate the instantaneous amplitude and frequency of selected IMFs which contain effective information in every frequency band instead of Hilbert transform method. The final time-frequency distribution can track energy change to predict fractured-vuggy reservoirs. Application of the VMD-EDO method in real seismic data has verified its effectiveness in improving time-frequency resolution, and characterizing energy anomalies hidden in broadband seismic data. Considering the reservoir characteristics of “low-frequency energy strengthened and high-frequency energy attenuated”, the VMD-EDO method is valid for oil and gas detection.

Keywords：variational mode decomposition(VMD), envelope derivative operator (EDO), Teager-Kaiser operator, time-frequency analysis, fractured-vuggy reservoir

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

Predictionofthinreservoirswithlesswelldatabasedonsparse-layerreflectivityinversion.DUXin1,2,FANTing’en1,2,FANHongjun1,2,ZHANGXianwen1,2,ZHANGJingyu1,2,andCAIWentao1,2.OilGeophysicalProspecting,2021,56(2):356-363.

Sparse spike inversion aims at restoring the top and bottom of a thicker reservoir. It is less effective for interpreting thin reservoirs by focusing on seismic tuning effect. To predict the reservoir in the study area with only a well, we propose a method for predicting the thin reservoirs. The procedures include:①Enhance the effective frequency of PSTM data based on sparse-layer reflectivity inversion; ②Split the wide frequency band data into low-frequency and high-frequency components through frequency division processing. Track the primary sequence stratigraphic framework on the low-frequency data and make stratal slices. Use the 90° phase shift method to convert the high-frequency data to lithology estimation, and extract the seismic attributes on the stratigraphic framework.③ On the histogram of the attributes, use integral operation to convert the seismic attributes to a ‘best reservoir probability estimation.④Sum up the ‘best reservoir probability estimation’ in each stratigraphic unit and establish a ‘cumulative frequency of favorable thin reservoir area estimation’. This attribute characterizes the spatial overlap of thin reservoirs, and can be used to guide design and location of wells for exploring the thin reservoirs.

Keywords：prediction of thin reservoir, less well data, sparse-layer reflectivity inversion, seismic sedimentology, frequency division processing, cumulative frequency of favorable thin reservoir area

1.CNOOC Research Institute Ltd.,Beijing 100020,China

2.State Key Laboratory of Offshore Oil Exploitation,Beijing 100020,China

UsingStackingmodelfusiontoidentifyfluidinhigh-temperatureandhigh-pressurereservoir.QINMin1,HUXiangyang1,LIANGYunan1,YUANWei1,andYANGDong1.OilGeophysicalProspecting,2021,56(2):364-371.

The difference in logging responses of reservoirs with different fluid properties in Dongfang X gas field is not obvious, and it is especially difficult to determine the bottom limit of the resistivity of different fluids. In addition, it is susceptible to physical factors when using porosity logging curves to identify fluids, making different fluid properties of sample data overlap. A Stacking model fusion method is proposed. It includes various machine learning algorithms (i.e. decision tree, support vector machine, random forest and extreme gradient boosting) and has better effects on fluid identification at high temperature and high pressure. Through 10 fold and cross validations, the algorithms are iteratively optimized to achieve a final optimal model. Compared with a single machine learning algorithm, the Stacking model fusion algorithm can take into account the differences in data observation and training principles of different algorithms, and give full play to the advantages of each model. Tests on real data indicate that the Stacking model can improve prediction accuracy from 87.08% to 92%, compared with the best-performing single model XGBoost. It has a stronger learning ability, and more suitable for fluid identification in high-temperature and high-pressure reservoir. It provides a new idea for building models of logging interpretation.

Keywords：high-temperature and high-pressure reservoir, identification of fluid properties, machine learning, Stacking model merging

1.Hainan Branch of CNOOC Ltd, Haikou, Hainan 524057,China

Studyonsedimentaryfaciesbasedonunsupervisedneuralnetworkseismicattributeclustering.WANGTianyun1,HANXiaofeng2,XUHaihong2,SUNXiao-ping1,LITao1,andHOUYan1.OilGeophysicalProspecting,2021,56(2)：372-379.

The classification of seismic facies based on unsupervised neural network self-organizing analysis (SOMA) is a comprehensive attribute clustering method. The key to the application of this method is to optimize seismic attributes, determine the number of clustering types, and analyze the relationship between seismic facies and sedimentary facies. Under the guidance of seismic sedimentology theory, we use the SOMA (self-organizing analysis) technology for cluster analysis of attributes, carry out seismic- sedimentary facies analysis by combining basic geological data, and select four seismic attributes such as RMS amplitude, information entropy, chaotic Li and fractal correlation dimension for cluster analysis. Taking the Cretaceous Suhongtu Formation in the Aitgele sag as a case, and using the method, we found such sedimentary facies as fan delta, braided river delta, shallow shore lake and deep lake. Traditional seismic -sedimentary facies analysis can judge the type of seismic facies by artificially observing seismic reflection. In contrast, our technology can reduce the unreliability of sedimentary facies analysis in areas with less well data. It provides a new basis for sedimentary facies analysis for oil and gas exploration. Also it is a practical, objective and accurate technical means.

Keywords：SOM, clustering of attributes, seismic - sedimentary facies analysis, Aitegle sag, Suhongtu formation

1. BGP Inc., CNPC, Zhuozhou, Heibei 072751, China

2. Xi’an Geological Survey Center of China Geological Survey, Xi’an, Shaanxi 710054, China

CharacteristicsofthestructuraltransitionzoneanditscontrollonreservoirsinWulanhuadepression.CAOSijia1,LIPeng2,LIUDongmin2,HUShaohua2,GUOBo2,andCHENLin2.OilGeophysicalProspecting,2021,56(2):380-388.

Seismic, drilling and logging data show that two stages of transfer zones have developed in the Wulanhua depression in Erlian basin. During the initial period of faulting depression (K1ba), Tumuer and Hongjing transfer zones were developed. The Tumuer transition zone is opposite overlap, and divided into two sags by a large low uplift. The Hongjing transition zone, in the form of “a transversal fault transition zone”, was developed in the orthogonal part of the Hongge fault and Hongjing fault in the southern sag. During the second period of depression spreading (K1bt1) , Saiwusu,Tunan and Hongeer transfer zones were developed on the fault transform part along the sag-controlling boundary of the southern sag. The Saiwusu fault is syntropic unoverlap, while the Tunan fault and the Honggeer fault are syntropic overlap. During the third period of depression atrophying (K1bt2), the transfer zones were transformed by compression and inversion, and four reverse zones (Tumuer, Saiwusu, Hongjing and Honggeer) were developed. The transfer zones controlled the sedimentary structure, accordingly controlled the source and sedimentary system, and consequently controlled the reservoir distribution. The transfer zones reversed and uplifted during the transformation period are zones where hydrocarbon migrates and structural traps. The transfer zones are favorable for exploring stratigraphic-lithologic and structural- lithologic reservoirs in the Wulanhua depression.

Keywords：Erlian basin,Wulanhua depression, transfer zone, tectonic evolution, hydrocarbon accumulation

1. School of Earth Sciences, Northeast Petroleum University, Daqing, Heilongjiang,163318, China

2. BGP Geological Research Center, CNPC, Zhuozhou, Hebei 072750, China

FissureeruptingmodelofPermianvolcanicrockinSichuanBasin.LIUPeng1,CHENKang2,HEQing-lin2,WANGPeng1,LIUZebin3,andLIUZhigang1.OilGeophysicalProspecting,2021,56(2):389-397.

Well YT1 obtained industrial gas flow from the Permian volcanic breccia lava. This set off a wave of exploring volcanic rock in Sichuan Basin in 2018. Affected by the quality of seismic data, there are still some deficiencies in understanding volcanic mechanism and eruption model, resulting in the lack of eruptive formation in wells such as ZJ2 drilled later. Starting from the characterization of volcanic channels, a joint study of gravity, magnetic and seismic data was carried out, and a volcanic mechanism model was established and the plane distribution of volcanic mechanism was effectively predicted after characterizing the characteristics of fissure-type eruption from the volcanic rock along the shear zone between the basement faults in Sichuan Basin. Used to guide well TF2, the drilling result is remarkable. The method provides a new idea for exploring special lithologic bodies of volcanic rock in Sichuan Basin.

Keywords：Sichuan Basin, volcanic rock, shear zone, eruption mode

1. Southwest Geophysical Research Institute, BGP, CNPC, Chengdu, Sichuan 610036, China

2. Research Institute of Exploration and Development, Southwest Oil and Gas Field Branch Company, PetroChina, Chengdu, Sichuan 610036, China

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

Benthicnodaltime-frequencydual-domainelectromagneticacquisitionsystemandtest.RENWenjing1,HEZhanxiang2,3,SUNWeibin4,ZHANGDongyang1,LUZhaoyang5,andLUYao4.OilGeophysicalProspecting,2021,56(2):398-406.

The 13th Five-Year Plan for scientific and technological innovation clearly requires the establishment of a technical system to safeguard national energy security and strategic interests, and the strengthening of strategic high-tech deployment in such fields as “Deep Sea, Deep Earth, Deep Space and Deep Blue”. Also it is of great significance to research and develop controllable source electromagnetic system for accelerating the research and development of offshore exploration technology. This paper introduces the working principle of the time-frequency dual-domain electromagnetic acquisition system with seabed nodes, describes the electromagnetic sensor, data acquisition station, acoustic unit and framework, analyzes the overall acquisition performance of the system, and finally demonstrates the performances and characteristics of the key links, such as data delivery, signal acquisition, release and recovery of seabed nodes. The interpretation result of the test section is consistent with the transverse change of the electrical property of the known gas-bearing reservoir, proving the system applicable. The system is suitable for industrial application and lays a foundation for promoting the industrial application of marine electromagnetic exploration technology in China.

Keywords：marine electromagnetic, node acquisition, time-frequency dual-domain, acquisition system

1.Xi’an Geophysical Prospecting Equipment Company of BGP, Xi’an, Shaanxi 710082, China

2.Shenzhen Key Laboratory of Deep Offshore Oil and Gas Exploration Technology, Shenzhen, Guangdong 518055, China

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

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

5.Wuhan University of Technology, School of Navigation, Wuhan, Hubei 430063, China

Multi-parameterconstrainedthree-dimensionalshapeinversionofmagneticbodies.LIJinpeng1,FANHongbo2,LIULi1,ZHANGYingtang2,LIZhining2,andWUBingyang3.OilGeophysicalProspecting,2021,56(2):407-418.

Traditionally, the magnetic and geometric parameters for shape inversion of magnetic bodies are defined manually. We propose a three-dimensional shape inversion method of magnetic bodies based on multi-parameter constraints. First, an objective function is established through L1norm and regularization constraint function, and the optimal solution is selected in the module to be updated to update the inversion model; then the method for estimating the multiple parameters of the magnetic target is proposed, which can directly estimate the position, magnetization direction and magnetization intensity of the magnetic target without prior information; and finally, the magnetic component (Bz) alongzand magnetic gradient tensor are used for joint inversion, and the inversion result is compared with normalized source strength and total magnitude. The simulation and experimental results show that, even without prior information, the method we proposed can effectively estimate the magnetic parameters of the magnetic target and accurately obtain the inversion results of the magne-tic target. At the same time, under the condition of remanent magnetization, compared with the traditional normalized source strength and total magnitude, this method can effectively improve the accuracy of three-dimensional inversion of magnetic bodies.

Keywords：three-dimensional shape inversion, multi-parameter, magnetic target, prior information

1. Unit 93114 of PLA, Beijing 100195, China

2. Army Engineering University(Shijiazhuang), Shijiazhuang, Hebei 050003, China

3. Beijing Institute of Remote Sensoring Information, Beijing 100192, China