ABSTRACTS

PROCESSINGTECHNIQUE

Random noise suppression on seismic data based on structured-clustering dictionary learning

ZhangYanet al (1119)

An adaptive filtering denoising method based on generalized S-transform and Gaussian smoothing

CaoPengtaoet al (1128)

Seismic data quality improvement with the anisotropic diffusion filter based on fault operator

LiFuqianget al (1137)

SEISMICSIMULATION

Elastic wave propagation features in tilted ellipsoidal anisotropic media

LiangKaiet al (1142)

Numerical simulation of 2D visco-acoustic wave equation with an optimized combined compact difference scheme

WangYonget al (1152)

A linear traveltime perturbation interpolation method for seismic ray tracing

LiTongyuet al (1165)

A linear traveltime interpolation method with hybrid meshes for VTI medium under rugged interface conditions

WangQiet al (1175)

MIGRATIONANDIMAGING

Least-squares reverse time migration with a hybrid stochastic conjugate gradient

LiChuanget al (1188)

An efficient forward modeling with the low-rank finite-difference algorithm for complex TTI media and its application in reverse time migration

HuangJinqianget al (1198)

Full 3D double-square-root prestack depth migration in visco-acoustic medium and its application in offshore seismic data

HeJianweiet al (1210)

First break traveltime tomographic imaging based on the eikonal equation in VTI medium

ZhangKaiet al (1218)

PETROPHYSICS

Modified continuous random patchy-saturation model in tight gas detection

WeiXianet al (1227)

Frequency-dependent elastic properties in fractured porous media and impact parameters

GouBinbinet al (1235)

Resistivity characteristics of core samples containing natural gas hydrates in Shenhu Area,the South China Sea

ChenYufenget al (1241)

MICROSEISMIC

Focal mechanisms and moment tensor in orthorhombic anisotropic media

TangJieet al (1247)

COMPREHENSIVERESEARCH

Quantitative characterization of syndepositional fault activity intensity with the sequence vector slip distance: An example in Panyu Low-uplift,Pearl River Mouth Basin……

LiuYonget al (1256)

Key techniques for subsalt reservoir prediction in the central and east parts of Ordos Basin

CaiKehanet al (1263)

Thin reservoir prediction based on seismic segmented-frequency band tune and Wheeler transformation

YinWenet al(1269)

Narrow-channel sand prediction with seismic principal components and Fisher discriminant analysis

JiangYanet al (1283)

Impedance inversion based on small-angle stacking seismic data

LiMenget al (1291)

Spectrum decomposition based on three-parameter wavelet

ZhuZhenyuet al (1299)

GEO-SOFTWARETECHNIQUE

Multi-level parallel design and optimization for reverse time migration based on OpenACC programming model

ZhaoHuet al (1307)

SEISMICGEOLOGY

Evaluation of fault lateral sealing and reservoir accumulation model in Daliuquan Area,Langgu Sag

HuXinleiet al (1314)

Regional stratigraphic framework establishment in South China Sea with pluralism

SongShuang(1326)

NON-SEISMIC

3D resistivity inversion with multiple priori-information constraint

GuoLaigonget al (1333)

Influence of IP effects on marine controlled-source electromagnetic responses

DingXuezhenet al (1341)

Discretization of normalized total gradient based on power mean

WangYanguoet al (1351)

ABSTRACTS

Randomnoisesuppressiononseismicdatabasedonstructured-clusteringdictionarylearning.ZhangYan1,2,RenWeijian2andTangGuowei3.OGP,2018,53(6):1119-1127.

Since seismic waveforms vary greatly in different spatial positions,the sparse representation based on the global dictionary learning is not enough to provide optimal sparse representations of local seismic data features.Therefore we propose a random noise suppression based on sparse representations of structured-clustering dictionary learning.First the regularity and redundancy in the coefficients distribution are represented by the self similarity of seismic data block structures and the global dictionary sparse representation,the K-means method is used to cluster seismic data blocks.Then according to the structural features,the overcomplete dictionary is obtained by singular value decomposition (SVD) for a class of data blocks.So the seismic data blocks are recorded according to every clustering center,and the original seismic data are more sparsely represented and described.After that,the regularization model is established to update centroid and estimated values of seismic data.Finally,the dual variable iterative threshold algorithm is used to solve the optimization problem of the double L1norm in the model,and the noise components are removed.Based on our experiments,the proposed algorithm can obtain higher peak signal to noise ratio,and better local seismic textures,which proves the effectiveness of the proposed algorithm in the random noise suppression.

Keywords: seismic data denoising,sparse representation,structured clustering,dictionary learning

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

2.School of Electrical Engineering & Information,Northeast Petroleum University,Daqing,Heilongjiang 163318,China

3.Modern Educational Technology Center,Northeast Petroleum University,Daqing,Heilongjiang 163318,China

AnadaptivefilteringdenoisingmethodbasedongeneralizedS-transformandGaussiansmoothing.CaoPengtao1,2,ZhangMin1,2andLiZhenchun1,2.OGP,2018,53(6):1128-1136,1187.

The conventional time-frequency filtering denoising method based on generalized S-transform needs human-computer interaction to determine the suppression range of high-frequency noise in the time-frequency domain.To overcome this shortcoming,an adaptive denoising method is developed with jointing the generalized S-transform adaptive time-frequency filtering and the Gaussian smoo-thing denoising algorithm.First,signals are processed with the generalized S-transform into time-frequency domain data.Then in the process of reconstructing the time domain signal by the inverse S-transform,the data adaptive time-varying filtering function is used to remove most of high-frequency random noise.Finally the Gaussian smoothing filter function is used to remove the residual high-frequency random noise of the signals.Tests on model and real data verify the correctness and effectiveness of the proposed method in removing high-frequency random noise on seismic data.Compared with the conventional method,the proposed method is less affected by processing parameters and the time-frequency resolution of the signal in the processed data is also higher.

Keywords: generalized S-transform (GST),time-frequency filtering and denoising,adaptive,Gaussian smoothing,high-frequency random noise

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

2.National Laboratory for Marine Sciences and Technology,Qingdao,Shandong 266071,China

Seismicdataqualityimprovementwiththeanisotropicdiffusionfilterbasedonfaultoperator.LiFuqiang1,ZhouDonghong1,MingJun1andXiaTong-xing1.OGP,2018,53(6):1137-1141.

Faults and fractures have weak response on seismic sections and its edge structure information is easy to be weakened during the data acquisition and processing.Therefore,we analyze the nonlinear anisotropic diffusion filtering theory on the basis of previous studies.By introducing a fault operator,the anisotropic diffusion filtering is analyzed.The fault operator is a bright spot to judge the discontinuity of seismic axis,its size reflects the degree of discontinuity.When the bright spot is close to a fault and crack,the value of the fault operator is small with a weak diffusion,and vice versa.So using the fault operator can not only ensure the filtering effect,but also protect the effective geological information.Both numerical simulations and real data applications show that the anisotropic diffusion filter based on fault operator can effectively remove the noise and enhance the break-point energy.Seismic attributes can be easily extracted from the data processed by the proposed method.

Keywords: anisotropic diffusion filter,fault operator,edge structure information,break-point energy

1.Tianjin Branch,CNOOC,Tianjin 300459,China

Elasticwavepropagationfeaturesintiltedellipsoidalanisotropicmedia.LiangKai1,2,CaoDanping1,2,YinXingyao1,2andWuGuochen1,2.OGP,2018,53(6):1142-1151.

The propagation features of elastic wave include velocity and polarization.They are important parameters in the researches of seismic wave propagation,and play important roles in the theory and application.The tilted ellipsoidal anisotropic (TEA) media is a special case of transversely isotropic media.Based on stiffness matrix of TEA media in the observation coordinate system,we derive analytic expressions of velocity and polarization for elastic wave in TEA media with Christoffel equation.Results of theoretical analysis and numerical examples indicate: ①In TEA media,iso-surfaces of phase velocity for qP wave and SH wave are anisotropic,but not ellipsoidal,and their iso-surfaces of group velocity are both ellipsoidal with the tilted symmetry axis; phase velocity and group velocity of qSV wave are the same,and isotropic; ②The polarization of qP wave,qSV wave and SH wave are perpendicular to each other,and the polarization of qP wave and qSV wave is related to anisotropic parameters; the polarization of SH wave is independent of the anisotropic parameters and is perpendicular to the plane determined by the propagation direction and the symmetry axis.

Keywords: tilted ellipsoidal anisotropic (TEA) media,phase velocity,group velocity,polarization,elastic wave

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

2.Laboratory for Marine Mineral Resources,Qingdao National Laboratory for Marine Sciences and Technology,Qingdao,Shandong 266071,China

Numericalsimulationof2Dvisco-acousticwaveequationwithanoptimizedcombinedcompactdifferencescheme.WangYong1,2,XuYoude3,GaoGang1,2,GuiZhixian1,2,ChenYing4andWangYanan5.OGP,2018,53(6):1152-1164.

The numerical simulation of seismic wavefield has great significance in geophysical exploration and seismology.In this paper,a combined compact difference scheme is proposed for the numerical simulation of visco-acoustic wave equation.First the second-order discrete scheme of the displacement field is established according to the Taylor series expansion and visco-acoustic wave equation,and a combined compact difference scheme is used to calculate the spatial derivative of the displacement field.Then the accuracy,dispersion and stability of the difference scheme are analyzed.Finally the optimization of the combination compact difference scheme is discussed based on the idea of dispersion-relation-preserving.The following understandings are obtained from our theoretical research: A.3-point 6-order combined compact difference scheme has smaller truncation errors and lower simulation dispersion than the conventional 7-point 6-order center difference and 5-point 6-order compact difference; B.The dispersion relation and the stability of the visco-acoustic wave equation difference scheme are not only related to the space grid size and time step,but also to the media quality factor and the seismic dominant frequency; C.The numerical wavenumber of the optimized combination compact difference scheme is more close to the true wavenumber than that before optimization,which is more beneficial to suppress the numerical dispersion and improve the calculation efficiency.The numerical simulation and wavefield of visco-acoustic wave equation are analyzed on uniform and Marmousi models based on the perfectly matched layer (PML) boundary condition.Numerical simulation results show that the proposed scheme has higher accuracy and computational efficiency in complex media.

Keywords: combined compact difference,visco-acoustic equation,numerical simulation,numerical dispersion,stability condition,perfectly matched layer (PML)

1.Key Laboratory of Exploration Technologies for Oil and Gas Resources,Ministry of Education,Yangtze University,Wuhan,Hubei 430100,China

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

3.West Branch,Research Institute of Exploration and Development,Shengli Oilfield Branch Co.,SINOPEC,Dongying,Shandong 257001,China

4.Surveying Service Center,Equipment Service Department,BGP Inc.,CNPC,Zhuozhou,Hebei 072751,China

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

Alineartraveltimeperturbationinterpolationme-thodforseismicraytracing.LiTongyu1,3andZhangJianzhong1,2,3.OGP,2018,53(6):1165-1174.

The linear traveltime interpolation (LTI) method assumes that traveltime is linear along a cell boundary,and the traveltime at any point on a cell boundary can be computed by interpolating the known traveltimes at adjacent grid nodes.In reality,however,the traveltime is not linear along a cell boundary,and the linear assumption may result in big calculation errors especially in large grid cells.To solve this problem,we calculate the traveltime with a linear traveltime perturbation interpolation (LTPI) method.In this method,the traveltime at a grid node is decomposed into two parts: a traveltime perturbation and a reference traveltime propagating in an equivalent homogeneous medium.The traveltime perturbation is much less than the reference traveltime.We assume that the traveltime perturbation is linear along the cell boundary,while the reference traveltime is nonlinear.We discretize the complex media with undulant topography and interfaces into a series of irregular cells.We also derive the LTPI formula for 2D irregular cells.Based on the LTPI,we propose a ray tracing method.Numerical experiments indicate that the LTPI is more accurate and stable than the LTI in computing wavefront traveltimes and ray paths.The LTPI performs better than the LTI in complex media.Under a certain accuracy requirement,the LTPI is more efficient than the LTI.

Keywords: wavefront traveltime,ray tracing,traveltime perturbation,traveltime interpolation,irregular cell

1.Key Laboratory of Submarine Geosciences and Prospecting Techniques,Ministry of Education,Qingdao,Shandong 266100,China

2.Laboratory of Marine Mineral Resources Evaluation and Detection Techniques,National Laboratory for Marine Science and Technology,Qingdao,Shandong 266061,China

3.College of Marine Geosciences,Ocean University of China,Qingdao,Shandong 266100,China

AlineartraveltimeinterpolationmethodwithhybridmeshesforVTImediumunderruggedinterfaceconditions.WangQi1,2,LiQingchun1andWangZhiqi1.OGP,2018,53(6):1175-1187.

In order to calculate seismic wave traveltime in VTI medium under rugged interface conditions,we divide a model with a hybrid mesh composed by rectangular meshes and irregular quadrilateral meshes to better fit rugged interface conditions.In the hybrid mesh,the relationship between group velocity and group angle is transformed into the relationship between group velocity and interpolated point coordinates to find the local traveltime.Then with the multi-step partitioning technique,the multi-step linear traveltime interpolation (LTI) ray tracing algorithm is extended to VTI medium to realize the traveltime calculation of seismic waves (first arrival,reflection,multiple reflection,multiple transmission conversion,and multiple reflection conversion) in VTI medium.Compared with the finite difference,fast marching method,and irregular cell shortest path method,numerical simulations show that the proposed algorithm is more accurate and can better adapt complex-structure models.

Keywords: vertical transverse isotropic,ray tracing,rugged interface,hybrid mesh,seismic wave traveltime

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

2.China National Aviation Fuel North-west China Company,Xi’an,Shaanxi 710082,China

Least-squaresreversetimemigrationwithahybridstochasticconjugategradient.LiChuang1,2,HuangJianping1,2,LiZhenchun1,2,WangRongrong3andSunMiaomiao1,4.OGP,2018,53(6):1188-1197.

To improve the efficiency of the least-squares reverse time migration (LSRTM),a hybrid stochastic conjugate gradient method is derived with the stochastic optimization theory.The correlation-based shot sampling method is proposed to reduce the amount of migrated data set,which also reduces stochastic characteristics of the gradient.After sampling shot data,the stochastic gradient descent iteration and the conjugate gradient iteration are used alternatively to update the migration imaging.The conjugate gradient iteration improves the convergence efficiency and image quality of LSRTM.Numerical tests on the synthetic data of SEG/EAGE rugged topography model and 2D field data from the exploration area M show that the proposed method can improve the computational efficiency of standard LSRTM,and has higher image quality and faster convergence compared with the stochastic gradient descent method.

Keywords: least-squares migration,stochastic optimization,stochastic gradient,conjugate gradient,rugged topography

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

2.Laboratory for Marine Mineral Resources,Pilot National Laboratory for Marine Science and Technology,Qingdao,Shandong 266580,China

3.Institute of Electronics,Chinese Academy of Sciences,Beijing 100190,China

4.Shengli Branch,Petroleum Engineering Geophysical Limited Corporation,SINOPEC,Dongying,Shandong 257088,China

Anefficientforwardmodelingwiththelow-rankfinite-differencealgorithmforcomplexTTImediaanditsapplicationinreversetimemigration.HuangJinqiang1,LiZhenchun2andJiangWen3.OGP,2018,53(6):1198-1209.

The computation efficiency is the key factor to restrict the practicality of anisotropic reverse time migration.In addition,pseudo shear-wave artifacts,numerical dispersion,and instability are also inherent problems of TTI medium qP-wave forward modeling and reverse time migration.The low-rank wavefield extrapolation algorithm is basically free from pseudo shear-wave artifacts,numerical dispersion,and instability.However,this method is a time-consuming and inefficient because its computing speed is controlled by model parameters.To improve the computation efficiency,a new compact differential template is proposed based on the idea of mixed-grid finite-difference,and the adaptive difference coefficient matched with the model is obtained by means of low-rank decomposition.Then,a low-rank finite-difference forward modeling and reverse time migration for TTI media are implemented.Numerical tests show that the proposed scheme is an anisotropic reverse time migration practical approach which combines high imaging accuracy and high calculation efficiency.The proposed scheme possesses the flexibility and efficiency of finite-difference method and pure-qP-wave precise calculation of low-rank wavefield extrapolation algorithm.So it can remove the pseudo shear-wave artifacts and numerical instability while improving the computation efficiency.

Keywords: TTI media,forward modeling,reverse time migration,low-rank decomposition,pure quasi-P-wave

1.College of Resources and Environment Engineering,Guizhou University,Guiyang,Guizhou 550025,China

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

3.Geophysical Branch,Daqing Drilling Exploration Company,CNPC,Daqing,Heilongjiang 266071,China

Full3Ddouble-square-rootprestackdepthmigrationinvisco-acousticmediumanditsapplicationinoffshoreseismicdata.HeJianwei1,DengYong1,RenTing1,LiXiaozhang1,ZhangWenxiang1andLuTongxiang1.OGP,2018,53(6):1210-1217.

The multi-azimuths/wide-azimuth and high-density seismic acquisition in marine exploration could enhance the subsurface imaging in terms of illumination and energy focusing.However,the huge amount of shot gathers poses great challenge for seismic imaging.Besides,as the exploration target becomes deeper and deeper,the inelastic absorption affects the imaging quality of deep reservoirs.To solve these problems,we propose in this paper a full 3D double-square-root (DSR) prestack depth migration (PSDM) method in visco-acoustic medium.The full 3D DSR PSDM method has several advantages.Firstly,it does not need to select the migration aperture.Secondly,it is suitable for the multi-azimuths/wide-azimuth and high-density marine seismic data with a large number of shots.Thirdly,it can extended to the absorption medium.Fourthly,ADCIGs of all azimuths can be easily generated.With large memory and the high-dimensional fast Fourier transform,this method can largely improve the migration efficiency.Synthetic data and real marine data processing demonstrate the feasibility of the proposed method.

Keywords: prestack depth migration (PSDM),visco-acoustic medium,full 3D,attenuation compensation,multi-azimuths/wide-azimuth,high density

1.CNOOC Hainan Energy Co.LTD,Haikou,Hai-nan 570100,China

FirstbreaktraveltimetomographicimagingbasedontheeikonalequationinVTImedium.ZhangKai1,2,ChenJunyi1,2,ZhuBaihang1,LiZhenchun1,2andDingPengcheng1.OGP,2018,53(6):1218-1226.

The first break traveltime tomography is an effective method for the inversion of anisotropic parameter,its difficulties are the calculation of first break traveltime and joint inversion of anisotropic parameters in VTI medium.In this paper,we use three parameters including the vertical velocity,NMO velocity,and non-elliptic coefficient to characterize the anisotropy.The ray tracing based on Runge-Kutta algorithm is used to calculate Fréchet kernel.According to group slowness approximate expressions,Fréchet kernels of these three parameters are derived.In order to overcome the magnitude between the three parameters,a double-round strategy is employed.The least squares orthogonal decomposition (LSQR) algorithm is used to carry out the joint inversion of the two methods.Numerical simulations show that the proposed method has a better inversion accuracy than the Thomsen method,and the ray tracing based on the Runge-Kutta algorithm is not only applicable to isotropic medium but also to VTI medium.

Keywords: VTI medium,eikonal equation,first break traveltime tomography,Fréchet kernel,Thomsen parameter

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

2.Laboratory of Marine Mineral Resources Evaluation and Detection Techniques,National Laboratory for Marine Science and Technology,Qingdao,Shandong 266061,China

Modifiedcontinuousrandompatchy-saturationmodelintightgasdetection.WeiXian1,2,YangZhifang1,YanXinfei1,LuMinghui1,LiXiaoming1andRenYan1.OGP,2018,53(6):1227-1234.

The heterogeneous distribution of gas and water in tight-sand gas reservoirs may cause complex relationships between elastic and reservoir parameters,which results in strong uncertainty in the quantitative seismic interpretation.According to characteristics of complex fluid distribution in tight-sand gas reservoirs,we modify the continuous random patchy-saturation model.This modified model can well investigate the influence of petrophysical properties and pore fluids on wave responses.A multi-scale rock physical template is established based on the modified model,and is calibrated by seismic-,logging-,and ultrasonic-scale data.The multi-scale rock physics template is applied to the tight-gas reservoir prediction in Sichuan Province.The quantitative estimation of rock porosity and gas saturation matches very well with well gas production tests.

Keywords: rock-physics model,patchy-saturation model,multi-scale rock-physics template,porosity prediction,gas reservoir identification

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

2.Beijing Vocational College of Labour and Social Security,Beijing 100083,China

Frequency-dependentelasticpropertiesinfracturedporousmediaandimpactparameters.GouBinbin1,ChenXuehua1,QuLei2,LüBingnan2,JiangWei2andZhangJie2.OGP,2018,53(6):1235-1240.

When seismic wave propagates through pores with different sizes and shapes,it will lead wave-induced local flow,which causes frequency-dependent dispersion and attenuation.In this paper,we analyze the effects of fluid substitution in fractured porous media based on the Collet squirt flow model.First we calculate the velocity of P-wave and inverse quality factor of the media at different incident angles.The velocity dispersion and attenuation vary with the frequency.There is no change in velocity at the low- and high-frequency limits,and the inverse quality factors increase and then decrease with the increase of frequency.Then we analyze effects of the background porosity variation on the P-wave velocity and anisotropic parameters.The obvious decrease occurs in velocity with the increase of porosity at lower frequencies (<1kHz) while no obvious change occurs at higher frequencies (>100kHz),and the attenuation increases with the frequency and media anisotropy becomes stronger.In addition,fluid types also affect the P-wave velocity.The velocity dispersion and attenuation are not obvious when the media filled with gas.The velocity dispersion and attenuation become stronger within brine-bearing media,and they are the strongest within oil-bearing media at the whole frequency band.Finally,we analyze the effects of the fracture density on the velocity and anisotropic parameters.The increasing fracture density can increase the P-wave velocity attenuation and media anisotropy.

Keywords: Collet model,frequency-dependent,fractured porous media,anisotropic parameters

1.State Key Laboratory of Oil & Gas Reservoir Geology and Exploitation,Chengdu University of Technology,Chengdu,Sichuan 610059,China

2.Key Laboratory of Earth Exploration & Information Techniques,Ministry of Education,Chengdu University of Technology,Chengdu,Sichuan 610059,China

ResistivitycharacteristicsofcoresamplescontainingnaturalgashydratesinShenhuArea,theSouthChinaSea.ChenYufeng1,LiangDeqing2andWuNeng-you3.OGP,2018,53(6):1241-1246.

The relationship between gas hydrate saturation (Sh) and sediment resistivity is the key to quantitatively estimate hydrate resource within submarine sediments in Shenhu Area,the South China Sea.The resistivity of core samples from the site SH7B in the area is measured during the natural gas hydrate formation by the temperature oscillations.At the temperature of 3.6℃,the methane became gas hydrate.The sample resistivity is 1.220 Ω·m at the 100% water saturation,and it increases to 2.150 Ω·m at the 37.7% hydrate saturation.Gas hydrate may first be cemented on grain particle of the unconsolidated core and the resistivity increases rapidly whenShis less than 18.0%.And it contacts possibly with sediment frame or loads in the pore fluid in the range ofSh18.0%～25.0%.Finally it may be cemented on sediment particles whenShexceeds 25.0%.As a result,the pore throat may be cemented with hydrate and a part of pore fluid cannot convert into hydrate.

Keywords: natural gas hydrate,resistivity,saturation,the South China Sea

1.Department of Chemical Engineering and Resource Recycling,Wuzhou College,Wuzhou,Guangxi 543000,China

2.Key Laboratory of Natural Gas Hydrate,Chinese Academy of Sciences,Guangzhou,Guangdong 510640,China

3.Key Laboratory of Natural Gas Hydrate,Qing-dao Institute of Marine Geology,Geological Survey of China,Ministry of Natural Resources,Qingdao,Shandong 266071,China

Focalmechanismsandmomenttensorinorthorhombicanisotropicmedia.TangJie1,WenLei1,WangHao1andZhangWenzheng1.OGP,2018,53(6):1247-1255.

Hydraulic fracturing zones have anisotropic characteristics on the whole.It is necessary to analyze the influence of anisotropic parameters on the focal mechanism and the moment tensor.This paper studies shear-tensile focal mechanism and seismic moment tensor when the source is located in anisotropic media,analyzes the influence of the focal anisotropy on double couple (DC) components,compensated linear vector dipole (CLVD) components,and isotropic (ISO) component.The seismic response characteristics of dry and saturated orthorhombic anisotropy media is also analyzed.Effects of crack parameters and fluid parameters on seismic response are discussed.The following understandings are obtained based on our research: A.The moment tensor in the anisotropic media can be obtained by anisotropic parameters and source quantity; B.The non-DC components caused by shear rupture in anisotropic media depend on the type and intensity of the anisotropy and the direction of the fault plane.High ISO components appear in dry crack media and high CLVD components in water saturation media; C.The far field P wave radiation patterns are different between isotropic and anisotropic media,source characteristics and media anisotropy have significant effects on seismic wave traveltime,amplitude,and polarity.

Keywords: orthorhombic anisotropy,microseismic,focal mechanism,moment tensor,shear-tensile source

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

Quantitativecharacterizationofsyndepositionalfaultactivityintensitywiththesequencevectorslipdistance:AnexampleinPanyuLow-uplift,PearlRiverMouthBasin.LiuYong1,DuXiaofang2,HuPeng2,3,LiXisheng4,HuoShoudong5andXuHai6.OGP,2018,53(6):1256-1262.

Conventional 2D scalar methods cannot comprehensively characterize the activity intensity and space-time variability of fault systems.Therefore in order to study the activity intensity and temporal-spacial difference of syndepositional fault systems under weak torsional environment and its control on paleo-geomorphology,the concept and calculation method of the sequence vector slip distance (SVSD) are proposed in this paper.According to the high-frequency sequence interfaces,the spatial distribution and activity properties of fracture systems within the intramural tectonic activity are quantitatively calculated.In the lower part of Zhujiang Formation,Panyu Low-uplift,the Pearl River Mouth Basin,4 sets of syndepositional fault systems and six 4th sequence boundaries are identified based on 3D seismic data.Then the cumulative activity of each fracture system at each sequence interface (total vector slip distance) are calculated.At the same time,multiple cross-correlation algorithms are used to spatially locate the breakpoints of different fracture systems at each sequence interface.With subtraction difference in space,the SVSD increment of the boundaries is obtained,that is the activity increment of the fracture systems in the single-sequence development period.Finally quantitative characterization of syndepositional fault activity intensity in different periods is achieved.The results show that the activity intensity of the fault systems in study area has a strong-weak-strong trend,and the paleo-geomorphology controlled by four groups of the fault systems gradually changed from the early north high south low to the late northeast high southwest low.The proposed approach can effectively quantify the activity of “tension or weak torsion” fracture systems.

Keywords: sequence vector slip distance (SVSD),syndepositional fault system,quantitative characterization,Pearl River Mouth Basin,Panyu Low-uplift

1.Urumqi Branch,GRI,BGP Inc.,CNPC,Urumqi,Xinjiang 830016,China

2.Zhongneng Power-Technology Development Company Limited,Beijing 100034,China

3.Laboratory of Marine Geology and Hydrology,Guodian New Energy Technology Research Institute,Beijing 102209,China

4.Shenzhen Branch,CNOOC,Shenzhen,Guangdong 518054,China

5.Key Laboratory of Petroleum Resources,Institute of Geology and Geophysics,Chinese Academy of Sciences,Beijing 100029,China

6.Petroleum Exploration and Production Research Institute,SINOPEC,Beijing 100871,China

KeytechniquesforsubsaltreservoirpredictioninthecentralandeastpartsofOrdosBasin.CaiKehan1,LiuFeng1,ZhangNa1andGaoGai1.OGP,2018,53(6):1263-1268.

Dolomite gas reservoirs were well developed under gypsum and salt rocks in the Ordovician Majiagou Formation,the central and east of Ordos Basin,However,it is very difficult to predict this kind of gas reservoirs because they are characterized by thin beds,strong heterogeneity,laterally rapid change,andlittle impedance difference with surrounding rocks.Therefore,based on high-precision internal imaging in the Ordovician carbonate rocks,a series of key techniques are used to predict dolomite gas reservoirs,such as seismic forward modeling,rock physical modeling,and prestack inversion of sensitive elastic parameters.Good results have been achieved in the reservoir prediction and the well deployment.The following understandings are obtained: ①Favorable subsalt reservoirs show weak complex wave; ②Brie equation (e=2) can be used to calculate the bulk modulus of mixed fluids,and can also be used for rock physical modeling and shear wave prediction because it characterizes the relationship between elastic parameters and the gas-bearing property of subsalt reservoirs; ③If the Poisson ratio of gas-bearing beds is less than 0.27,gas reservoirs can be delineated by crossplots of the Poisson ratio with p-wave impedance,and “small negative anomaly” of high frequency of fluid mobility indicates gas-bearing reservoirs.

Keywords: subsalt,dolomite reservoir,forward modeling,Brie equation,prestack inversion,fluid mobility

1.Exploration and Development Research Institute,Changqing Oilfield Company,PetroChina,Xi’an,Shaanxi 710018,China

Thinreservoirpredictionbasedonseismicsegmented-frequencybandtuneandWheelertransformation.YinWen1,ZhuJianbing2,LiYuan3,GuoJiashu4andLiChanghong2.OGP,2018,53(6):1269-1282.

Conventional reservoir prediction methods are mainly carried out in the time domain and depth domain,which leads strongly multiple solutions,especially for continental complex geological bodies.Besides,conventional seismic tuning body analysis is based on the dominant frequency of seismic data without considering the contribution of high frequency components.Therefore we propose an approach to thin reservoir prediction based on the sequence stratigraphy and seismic sedimentology.First seismic tuning bodies in different segmented-frequency bands are extracted.Then after horizon automatic tracking constrained by isochronal stratigraphic framework,horizon tracking results are transformed into the Wheeler domain with characteristics of sedimentary cycles in the time domain,and the stratigraphic correlation and chronostratigraphic framework are constructed to realize the synchronous interpretation of chronostratigraphy and sedimentary system tract,which reveals spatial relationships of sedimentary assemblages.Then,stratigraphic slices of high-frequency tuning body are extracted to predict strata-bound thin reservoirs,and the transverse prediction accuracy of thin sand bodies is effectively improved with the isochronous interface.With high vertical resolution of seismic data and high transverse resolution of isochronous slices in the Wheeler domain,the reservoir interpretation multiplicity is reduced.The validity of the proposed approach is verified by seismic data applications.

Keywords: seismic segmented-frequency band tune,horizon automatic tracking,Wheeler transformation,synchronous interpretation in the system tract,thin reservoir prediction

1.Karamay Campus,Faculty of Petroleum,China University of Petroleum (Beijing),Karamay,Xinjiang 834000,China

2.Geophysical Research Institute,Shengli Oilfield Branch Co.,SINOPEC,Dongying,Shandong 257022,China

3.Information Center,Human Resource and Social Security Bureau of Dongying,Dongying,Shandong 257091,China

4.College of Computer and Communication Engineering,China University of Petroleum (East China),Qingdao,Shandong 266580,China

Narrow-channelsandpredictionwithseismicprincipalcomponentsandFisherdiscriminantanalysis.JiangYan1,YangChunsheng1,LiWenyan1,ZhaoMingzhu1andZhangXiuli1.OGP,2018,53(6):1283-1290.

Fuyu oil bearing beds,as tight hydrocarbon reservoirs with low-porosity and ultra-low-permeability,are widely developed in the peripheral of Daqing,Songliao Basin.Channel sandbodies there are characterized by a narrow,thin,and poor distribution.It is very difficult to depict this kind of sandbodies only on the current sparse well patterns.Therefore,seismic principal components (PCA) and Fisher discriminant analysis are used for Fuyu reservoirs’ research in the Block Fang 38.The following observations are obtained: ①The PCA analysis is used for the dimension-reduction optimization of seismic attributes.The three-principal-component information at borehole accounts for 95.5% of the total variance of the original data.So the dimension of seismic attributes is reduced and the redundant information in multiple attributes is eliminated; ②The Fisher discriminant theory of multiple statistics is applied to the analysis of narrow channel sandbodies.Based on sandstone thickness at boreholes and the three-principal components at boreholes,Fisher sand and shale discriminant model is established,which not only integrates well-point sandstone thickness but also takes into account the spatial distance influence of seismic data.So the channel sandbody prediction is improved; ③With sandstone thickness of 1.5,2,and 3 m,the accuracy rates of Fisher sand and shale discriminant model are 87.5% ,88.9%,and 90%.And the ability to discriminate channel sandbody is even stronger.

Keywords: seismic principal component,Fisher discriminant analysis,joint use of logging and seismic data,narrow channel,Fuyu reservoirs

1.Research Institute of Exploration and Development,Daqing Oilfield Company,PetroChina,Daqing,Heilongjiang 163712,China

Impedanceinversionbasedonsmall-anglestackingseismicdata.LiMeng1,LiuZhen2,LiuMinzhu3,MaYuehua4andShaoWenchao5.OGP,2018,53(6):1291-1298.

Impedance inversion is a key tool in reservoir prediction and characterization as it converts seismic reflection information in more visual and explainable strata information such as impedance,velocity,and density.Theoretical derivation and seismic data tests show that seismic reflection information at large incident angles damage greatly the resolution of full-angle stacking seismic data,and conventional poststack impedance inversion cannot well characterize thin reservoirs.In order to well predict thin reservoirs,we propose an impedance inversion based on small-incident angle stacking seismic data.The application of the proposed approach in reservoir prediction in a marine block,West Africa shows that this impedance inversion results match very well with logging data and depict clearly subtle reservoirs.

Keywords: incident angle,seismic resolution,impedance inversion.

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

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

3.Essca Company,Beijing 100101,China

4.Dagang Branch,GRI,BGP Inc.,CNPC,Tianjin 300280,China

5.Geophysical Research Institute,SINOPEC,Nanjing,Jiangsu 211103,China

Spectrumdecompositionbasedonthree-parameterwavelet.ZhuZhenyu1,GaoJialun2,JiangXiudi1,SunWenbo1,XueDongchuan1andWangQingzhen1.OGP,2018,53(6):1299-1306.

We first describe the basic principle of spectrum decomposition,then introduce three-parameter wavelet,study the influence of each parameter on wavelet.After that we conduct time-frequency analysis with three parameters based on forward modeling.Finally we perform reservoir characterization with three-parameter wavelet.It is found that subtle geological information can be highlighted.The following understanding are obtained based on model and real data tests: ①The three-parameter wavelet has high flexibility,the modulation frequencyσof the wavelet affects the vibration degree of the wavelet,the energy attenuation parameterτcontrols attenuation speed of attenuation function.When the value ofτis relatively big,theσhas less influence on the wavelet.The influence of the energy lag parameterβon the wavelet shape is more complex.Whenβis an integer multiple of trigonometric function period,only wavelet time shift occurs,it can be used to match the zero phase wavelet; when theβis not an integer multiple of trigonometric function period,wavelet time shift and deformation are generated,it can be used to match the non-zero phase wavelet.Deeply buried channels may be misinterpreted whenβis not equal to zero; ②Compared with Morlet wavelet,three-parameter wavelet can more precisely depict subtle sedimentary structures in thin interbeds; ③In real data processing,the optimal parameter combinations can be obtained according to the correlation of various forms of basic wavelets with real wavelet extracted from targets.

Keywords: spectrum decomposition,three-parameter wavelet,wavelet transform,time-frequency analysis,reservoir characterization

1.CNOOC Research Institute,Beijing 100028,China

2.China University of Petroleum (Beijing),Beijing 102200,China

Multi-levelparalleldesignandoptimizationforreversetimemigrationbasedonOpenACCprogrammingmodel.ZhaoHu1,2,WuSihai1,2,YinCheng1,2,TangZekai3andJiaPeng3.OGP,2018,53(6):1307-1313,1325.

The GPU parallel program models,OpenCL and CUDA,have been widely applied in seismic modeling and migration.However both of them may destroy serial codes and need to redesign parallel algorithms.This paper introduces a new model OpenACC for reverse time migration.We realize multi-level parallelism between nodes,and inside the node by adding compiler instructions,which forms a more convenient and high-performance technique.After the communication optimization between nodes,this approach is used for the reverse time migration.Numerical experiments show that multi-level parallels can solve computation and storage problems for large-scale models.It can also dramatically improve computing efficiency after communication optimization.

Keywords: GPU programming,OpenACC,reverse time migration,multi-level parallels

1.School of Earth Sciences and Technology,Southwest Petroleum University,Chengdu,Sichuan 610500,China

2.Sichuan Key Laboratory of Natural Gas Geology,Southwest Petroleum University,Chengdu,Sichuan 610500,China

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

EvaluationoffaultlateralsealingandreservoiraccumulationmodelinDaliuquanArea,LangguSag.HuXinlei1,LüYanfang1,CaoLanzhu2,YangDexiang2andSunYonghe1.OGP,2018,53(6):1314-1325.

To understand the oil and gas distribution law in Daliuquan area,Langgu Sag,we first classify the accumulation conditions such as sources,transporting faults,and sedimentary sands.Then we determine fault lateral sealing threshold based on the coupling relation between the minimum shale gouge ratio (SGR) value of certain fault surface and oil testing result in typical wells.Finally taking this threshold as criteria we evaluate fault lateral sealing or opening properties and analyze the accumulation model.The following understandings are obtained based on our research: A.The thickness and buried depth of targets are stable in Daliuquan area,and the faults activity stages are similar so that the effect of compaction diagenesis is not obvious; B.Fault sealing properties are mainly affected by fault rock SGR,and the fault sealing SGR threshold is 29% in the strata Es3x; C.The evaluation results of fault sealing or opening properties are well coincident with the oil and gas distribution,the obtained thickness ratio between mudstone and strata,pressure coefficient,and oil/gas-aqueous interface match well with the estimation,which verifies the feasibility of this proposed approach.For the accumulation model in Daliuquan area,oil and gas were generated from source rocks in strata Es3xand Es4s.The oil and gas migratied vertically along transporting faults,diverted laterally into sands with strata ratio greater than 18% when they encountered overburden sealing.Oil and gas accumulations were formed when sealing faults,anticline traps or unconformity surfaces were encountered.

Keywords: Langgu Sag,Daliuquan area,fault lateral sealing or opening,shale gouge ratio (SGR) threshold of fault rock,accumulation model

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

2.Exploration and Development Research Institute,Huabei Oilfield Company,PetroChina,Renqiu,Hebei 062552,China

RegionalstratigraphicframeworkestablishmentinSouthChinaSeawithpluralism.SongShuang1.OGP,2018,53(6):1326-1332.

Limited by the quality and distribution of exploration data,the south and north strata in the South China Sea have not been unified.Based on regional seismic sections covering the South China Sea,this paper uses the pluralism to synthesize a variety of data including magnetic data,and ocean drilling program (ODP) data to establish the stratigraphic framework in the South China Sea.The seismic geological strata calibration is carried out mainly based on the high exploration degree area (Zhujiangkou,Qiongdongnan,Zengmu,and Lile Basins),some other strata (in Beikang and Nanweixi Basins) are also calibrated.Six sets of seismic reflectors and their interfaces have been identified in the Cenozoic.According to our study,the Cenozoic strata in the South China Sea are controlled by the expansion.There are three sets of strata: pre-expansion stage,expansion stage and post-expansion stage.The research results provide significant references for the unification of the south and north strata,the partition of new schemes,and Moho surface research in South China Sea.

Keywords: South China Sea,regional seismic section,pluralism,stratigraphic framework,stratigraphic distribution

1.CNOOC Research Institute,Beijing 100028,China

3Dresistivityinversionwithmultiplepriori-informationconstraint.GuoLaigong1,2,DaiGuanglong2,YangBencai2,XueJunhua3andChenBenliang3.OGP,2018,53(6):1333-1340,1350.

The use of prior-information constraint can improve the inversion accuracy of 3D DC electrical resistivity.However,in most cases,single prior-information is used to modify the objective function and the results are inaccurate.In order to solving the problems,the inversion equation of iterative re-weighted least-squares is modified; parameters such as the structural metric,weight mean,and standard deviation are added; and the weight function of the model parameters is replaced by the error function and complementary error function.Then different prior-information can constrain the inversion with modifying parameters of weight function in different zones.The parallel conjugate gradient least-square method is used to effectively improve the inversion calculation.In synthetic data inversions,smoothness constraint,block constraint,and resistivity extreme-constraint are used.The results show that the inversion constrained with multiple priori-information is more accurate than that with single prior-information.In practice,parameters of weight functions can be selected based on geological structures to improve the accuracy of 3D resistivity inversion and the calculation efficiency.

Keywords: 3D inversion,resistivity,iteratively re-weighted least-squares,structural metric,parallel conjugate gradient least-square

1.School of Electronics and Information Engineering,Anhui University of Sciences and Technology,Huainan,Anhui 232001,China

2.School of Mining and Safety Engineering,Anhui University of Sciences and Technology,Huainan,Anhui 232001,China

3.State Key Laboratory of Deep Coal Mining & Environment Protection,Huainan,Anhui 232001,China

InfluenceofIPeffectsonmarinecontrolled-sourceelectromagneticresponses.DingXuezhen1,LiYuguo1,2andLiuYing1,2.OGP,2018,53(6):1341-1350.

In the numerical simulation and interpretation of marine controlled-source electromagnetic (MCSEM) datasets,IP effects are usually ignored.In this paper,we study the IP influences on the 1D and 2D MCSEM modeling responses based on the Cole-Cole model.A new parameter-IP strength is defined for evaluating the strength of IP effects.We introduce the influence of IP effects on 1D MCSEM inversion.Numerical results suggest that IP effects depend on polarization parameters,observing frequency and reservoir thickness.IP effects may lead higher reservoir resistivity,however this deviation is not very significant.

Keywords: marine controlled-source electromagne-tic (MCSEM),IP effects,Cole-Cole model,IP strength

1.College of Marine Geosciences,Ocean University of China,Qingdao,Shandong 266100,China

2.Key Laboratory of Submarine Geosciences and Prospecting Techniques of Ministry of Education,Qingdao,Shandong 266100,China

Discretizationofnormalizedtotalgradientbasedonpowermean.WangYanguo1,WuZiying1,DengJuzhi1,YangYaxin1,HuangSong1andLuoXiao1.OGP,2018,53(6):1351-1364.

The normalized total gradient algorithm is a commonly used method for data processing and interpretation of gravity and magnetic data.How-ever,this method cannot recognize the information of multi-sources and identify the geometry of geologic bodies.In order to improve the practicability of the normalized total gradient method,the iterative filtering is used for the stability calculation of downward continuation.And then the power mean function is used for normalized processing.Finally survey lines are discretized on the basis of curve characteristics of the ground total gradient anomaly.Thereout,we propose a discretization of normalized total gradient method based on power mean.Model tests show that the new method can effectively obtain the locations of field sources and the geometry of sources which can be estimated by the best power time.In application,the discretization of the normalized total gradient method based on power mean has higher inversion accuracy and stronger geological explanatory ability than conventional method.

Keywords: normalized total gradient,power mean,iterative method,discretization,geometry

1.Fundamental Science on Radioactive Geology and Exploration Technology Laboratory,East China University of Technology,Nanchang,Jiangxi 330013,China