Relative Fidelity Processing of Seismic Data E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 465

Veröffentlichungsjahr: 2017

Beschreibung

This book presents a comprehensive overview of relative fidelity preservation processing methods and their applications within the oil and gas sector. Four key principles for wide-frequency relative fidelity preservation processing are illustrated throughout the text. Seismic broadband acquisition is the basis for relative fidelity preservation processing and the influence of seismic acquisition on data processing is also analyzed. The methods and principles of Kirchhoff integral migration, one-way wave equation migration and reverse time migration are also introduced and illustrated clearly. Current research of relative amplitude preservation migration algorithms is introduced, and the corresponding numerical results are also shown.

RTM (reverse time migration) imaging methods based on GPU/CPU systems for complicated structures are represented. This includes GPU/CPU high performance calculations and its application to seismic exploration, two-way wave extrapolation operator and boundary conditions, imaging conditions and low frequency noise attenuation, and GPU/CPU system based RTM imaging algorithms. Migration velocity model building methods in depth domain for complicated structures are illustrated in this book. The research status and development of velocity model building are introduced. And the impacting factors are also discussed. Several different velocity model building methods are also represented and analyzed. The book also provides the reader with several case studies of field seismic data imaging in different kinds of basins to show the methods used in practice. 

Leseprobe

Table of Contents

Cover

Title Page

Preface

1 Study on Method for Relative Fidelity Preservation of Seismic Data

1.1 Introduction

1.2 Discussion on Impact on Processing of High‐resolution, High SNR for Seismic Acquisition and Observation Mode

1.3 Discussion on the Cause of Notching

1.4 Discussion of Impact on Processing of Relative Fidelity Preservation Seismic Data for Seismic Acquisition and Observation Mode

1.5 Comparison of Results of High‐resolution, High SNR Processing and Relative Fidelity Preservation Processing

1.6 Elastic Wave Forward Modeling

1.7 Conclusions

References

2 Method and Principle for Seismic Migration and Imaging

2.1 Kirchhoff Integral Prestack Depth Migration

2.2 Amplitude Preservation Fourier Finite Difference Prestack Depth Migration Method

2.3 Reverse Time Migration

References

3 Study of Reverse Time Migration Method for Areas With Complicated Structures Based on the GPU/CPU System

3.1 Introduction

3.2 The GPU/CPU High‐performance Calculation and Its Application in Seismic Exploration

3.3 Study on the Two‐way Wave Extrapolation Operator and Its Boundary Conditions

3.4 Study on the Imaging Condition and Low‐frequency Noise Suppression Method

3.5 Study and Application of RTM Prestack Imaging Algorithm based on the GPU/CPU System

3.6 Conclusions

References

4 Study and Application of Velocity Model Building Method for the Areas with Complicated Structures

4.1 Introduction

4.2 Status Quo and the Development of the Velocity Model Building Method

4.3 Impacting Factors for the Velocity Model Building

4.4 Study and Application of the Seismic Velocity Model Building Method

4.5 Quality Monitoring and Accuracy Discussion of the Seismic Velocity Model Building

4.6 Velocity Analysis Method for Reverse Time Migration in Angle Domain

4.7 Study of the Full Waveform Inversion Method

References

5 Case Study

5.1 Application of 3D Prestack Reverse Time Migration in Subsalt Imaging

5.2 Application of High‐density All‐round Seismic Data Processing in the Carbonatite Region

5.3 Application of Seismic Imaging Method for Complicated Structures in the Tuha and Jiuquan Basins

5.4 Application of the Seismic Prestack Imaging Method in the Buried Hill Structural Zone in the Nanpu of Jidong Oilfield

References

Index

End User License Agreement

List of Tables

Chapter 02

Table 2.1 Spatial location of quasi‐P wavefield components and elastic parameters.

List of Illustrations

Chapter 01

Figure 1.1 Comparison between (a) conventional seismic profile and (b) relative fidelity seismic profile through wells S302 and S303.

Figure 1.2 The t

0

contour of the sand bottom boundary in Cretaceous Qinshuihe Formation of Shidong area.

Figure 1.3 Seismic data high resolution and high SNR processing.

Figure 1.4 Single shot records at same shot point of different seismic sources. (a) Analysis on 0 ~ 20 Hz scanning. (b) Analysis on 10 ~ 120 Hz scanning.

Figure 1.5 Analysis results on single shot record and frequency spectrum of target layer processed by static correction of first arrival refraction.

Figure 1.6 Analysis result on single shot record and frequency spectrum of target layer processed by prestack noise elimination.

Figure 1.7 Analysis results on single shot record and frequency spectrum of target layer processed by prestack zero phase deconvolution.

Figure 1.8 Seismic profile and frequency spectrum of target layers processed by high resolution and high SNR processing.

Figure 1.9 Results comparison of random noise data processed two times with RNA method repeatedly.

Figure 1.10 The location of notch frequency band in seismic frequency spectrum.

Figure 1.11 Filter frequency spectrum comparison between pre‐pack and post‐pack deconvolutions in single deep well.

Figure 1.12 Filter frequency spectrum comparison between prepack and post‐pack deconvolutions in combined well.

Figure 1.13 The processing flow of relative fidelity preservation seismic data.

Figure 1.14 Single shot record of single deep well and seismic spectrum of target layer (SD4).

Figure 1.15 Single shot record of combined well and seismic spectrum of target layer (SD4).

Figure 1.16 Single shot record of controllable seismic source excitation and seismic spectrum of target layer (SD4).

Figure 1.17 Seismic profile and target layer spectrum processed by relative fidelity preservation.

Figure 1.18 Spectrum comparison of single deep well prestack and poststack deconvolution filters.

Figure 1.19 Spectrum comparison of both combined well prestack and poststack deconvolution filters.

Figure 1.20 Sand interpretation based on results of high‐resolution (a1, b1) and relative fidelity (a2, b2) seismic data processing. (a1,a2) single deep well, 1200 receiving channels, channel interval of 5 m; (b1,b2) combined well, 1200 receiving channels, channel interval of 5 m.

Figure 1.21 Forward model design.

Figure 1.22 Forward model 1 (100 Hz, channel interval of 10 m) prestack time migration profile (depth domain).

Figure 1.23 Forward model 2 (100 Hz, channel interval of 10 m) prestack time migration profile (depth domain).

Chapter 02

Figure 2.1 Single shot illumination analysis results by Marmousi model. (a) Based on conventional Fourier finite difference operator. (b) Based on fidelity‐preserved amplitude Fourier finite difference operator.

Figure 2.2 Illumination comparison results of all shots by Marmousi model. (a) Based on conventional Fourier finite difference method. (b) Based on preserved‐amplitude Fourier finite difference method.

Figure 2.3 Migration comparison results of all shots records by Marmousi model. (a) Based on conventional Fourier finite difference method. (b) Based on fidelity‐preserved amplitude Fourier finite difference method.

Figure 2.4 Test result of foothills model data migration method. (a) Kirchhoff integral method. (b) Preserved‐amplitude Fourier finite difference method.

Figure 2.5 Test result of foothills actual seismic data migration method. (a) Kirchhoff integral method. (b) Preserved‐amplitude Fourier finite difference method.

Figure 2.6 Diagram of PML absorbing boundary.

Figure 2.7 Flowchart of excitation time imaging conditions.

Figure 2.8 Flowchart of cross‐correlation imaging condition.

Figure 2.9 NWGI velocity model.

Figure 2.10 Reverse time migration results by NWGI model.

Figure 2.11 Marmousi velocity model.

Figure 2.12 Reverse time migration results by Marmousi model.

Figure 2.13 Staggered grid diagram.

Figure 2.14 Pulse response test data.

Figure 2.15 Pulse response based on downward wave maximum energy method imaging condition.

Figure 2.16 Pulse response based on cross‐correlation imaging condition.

Figure 2.17 Pulse response based on normalized cross‐correlation imaging condition.

Figure 2.18 Horizontally layered medium model and synthetic record. (a) Horizontally layered medium model. (b) Figure a model’s synthetic record in every four groups.

Figure 2.19 Migration results under two imaging conditions. (a) Migration results by downward wave maximum energy method imaging condition. (b) Migration results by the normalized cross‐correlation imaging conditions.

Figure 2.20 η model.

Figure 2.21 The reverse time migration profile of Quasi‐P wave equation in VTI media. (a) Imaging condition migration profile by downward wave maximum energy method. (b) Imaging condition migration profile by normalized cross‐correlation method.

Figure 2.22 Krichhoff prestack time migration profile (from Alkhalifah, 1998).

Figure 2.23 Prestack reverse time depth migration profile.

Figure 2.24 Reverse time migration profile of isotropic medium acoustic wave equation.

Chapter 03

Figure 3.1 Floating‐point computation efficiency development comparison between GPU and CPU from 2003 to 2007.

Figure 3.2 Reverse time depth migration result of the Marmousi model on GT200 GPU (from Tongji University).

Figure 3.3 Reverse time migration imaging result of 3D salt dome model on GT200 GPU (from Tongji University).

Figure 3.4 Diagram of network subdivision.

Figure 3.5 NWGI velocity model.

Figure 3.6 Profile comparison before (a) and after (b) low‐frequency suppression of RTM imaging result (parameters [32, 0, 8], wavelet dominant frequency of 20 Hz).

Figure 3.7 Profile comparison before (a) and after (b) low‐frequency suppression for RTM imaging result of SEG/EAGE Sigsbee model (parameters [32, 0, 8]).

Figure 3.8 The Marmousi model RTM migration result.

Figure 3.9 The Marmousi model RTM migration result by applying Laplace filtering.

Figure 3.10 The Marmousi model RTM migration result by applying Muder noise elimination method, parameters [32, 0, 2].

Figure 3.11 Comparison of (a) the CPU structure features and (b) the GPU structure features.

Figure 3.12 Architecture model of GPU hardware.

Figure 3.13 Reverse time migration flowchart.

Figure 3.14 GPU flowchart of reverse time depth migration.

Figure 3.15a Inline velocity model.

Figure 3.15b The CPU migration result.

Figure 3.15c The CPU migration result.

Figure 3.16a The GPU migration result.

Figure 3.16b The CPU migration result.

Figure 3.16c The GPU migration result.

Figure 3.17a Depth slice velocity model.

Figure 3.17b The CPU migration result.

Figure 3.17c The GPU migration result.

Figure 3.18 NWGI model (a) forward wavefield and (b) forwarded zero migration domain profile.

Figure 3.19 Comparison between (a) NWGI velocity model and (b) imaging results of different migration methods.

Figure 3.20 Velocity‐depth model of multi‐well restrained inversion.

Figure 3.21 Imaging result comparison of (a) subsalt structure one‐way wave migration and (b) reverse time migration.

Chapter 04

Figure 4.1 The coverage plat.

Figure 4.2 Distribution of pre‐regularization offset.

Figure 4.3 Distribution of post‐regularization offset.

Figure 4.4 The migration effects comparison of pre/post‐regularizations. (a) Pre‐regularization migration. (b) Post‐regularization.

Figure 4.5 The schematic map of subsurface media.

Figure 4.6 Inversion coherence capacity of interval velocity.

Figure 4.7 The inverted interval velocity spectrum.

Figure 4.8 Schematic diagram of coherence inversion.

Figure 4.9 Obtaining interval velocity with coherence inversion method.

Figure 4.10 Interval velocity inversion of the first interval. (a) Model data. (b) Actual data (anisotropy assumption).

Figure 4.11a Successive interval velocity inversion of the second interval.

Figure 4.11b Interval velocity inversion of the second interval.

Figure 4.12a Interval velocity inversion of the second interval.

Figure 4.12b Successive interval velocity inversion of the second interval.

Figure 4.13 Flow chart of prestack depth migration velocity analysis.

Figure 4.14 Transformation from depth error to time error.

Figure 4.15 Vertical slowness angle at reflection point.

Figure 4.16 CRP gathers (a) before model optimization and (b) after model optimization.

Figure 4.17 Comparison of horizontal delay before and after model optimization.

Figure 4.18 Field example of CFP modeling and imaging, where (a) is the velocity field after CFP modeling, and (b) is the CFP prestack migration profile of the corresponding velocity field.

Figure 4.19 Flow chart of comprehensive modeling combining first arrival wave with reflection wave.

Figure 4.20 Inverted shallow layer velocity model Velocity unit:m/s.

Figure 4.21 (a) True velocity model and (b) depth migration result.

Figure 4.22 (a) Velocity model of inversion shallow layer and (b) depth migration result.

Figure 4.23 (a) Shallow layer velocity inversion model and (b) depth migration result.

Figure 4.24 (a) Conventional velocity model and (b) velocity modeling integrating first arrival wave inversion with reflection wave inversion.

Figure 4.25 (a) Velocity modeling migration integrating first arrival wave inversion with reflected wave inversion and (b) comparison of geological outcrop interpretation profile.

Figure 4.26 Diagram of external drift Kriging principle.

Figure 4.27 Multi‐model external drift Kriging interpolation.

Figure 4.28 (a) Well constrained interval velocity inversion plane and (b) well velocity error.

Figure 4.29 Inverted interval velocity model under multi‐well constraint.

Figure 4.30 Well data constraint and velocity interface.

Figure 4.31 Velocity modeling under geological structure constraint.

Figure 4.32 (a) One‐way wave of PSDM profile and (b) reverse time migration.

Figure 4.33 The relationship of incident angle, propagation angle and dip angle.

Figure 4.34 The Marmousi II velocity model.

Figure 4.35 Reverse time migration results of the model Marmousi II.

Figure 4.36 Computation results of dip angle of the model Marmousi II.

Figure 4.37 Reverse time migration results of the simple interval model.

Figure 4.38 Computation results of the incident angle of the simple interval model.

Figure 4.39 Computation results of the incident angle at CDP points of the model Marmousi II. (a) CDP = 150. (b) CDP = 350. (c) CDP = 650.

Figure 4.40 Single shot reverse time migration results of the model Marmousi II. (a) Single shot reverse time migration before Laplace filtering. (b) Single shot reverse time migration after Laplace filtering.

Figure 4.41 Angle domain CIP gathers at different locations of the model Marmousi II.(a) CDP = 150. (b) CDP = 650.

Figure 4.42 The flat bed velocity model.

Figure 4.43 Angle domain CIP gathers gained by flat bed velocity model at various velocities.(a) Accurate velocity. (b) Lower velocity. (c) Higher velocity.

Figure 4.44 The real velocity model and parameters for numerical experiment of angle domain velocity analysis.

Figure 4.45 Reverse migration results from the constant velocity initial velocity model.

Figure 4.46 The velocity model updated through the angle domain velocity analysis method.

Figure 4.47 Reverse time migration results obtained from the updated velocity model.

Figure 4.48 Comparison of the updated velocity model and real velocity model. (a) CDP = 200. (b) CDP = 450. (c) CDP = 600. (d) CDP = 750. (e) CDP = 1000. (f) Locations of five CDPs in the velocity model.

Figure 4.49 (a) Tomography inversion velocity. (b) Inversion results after five times iteration with 2–10 Hz.; (c) Inversion results after five times iteration with 2–3 Hz, 2–4 Hz, 2–6 Hz, 2–8 Hz.

Figure 4.50 Flow chart of gradient full waveform inversion.

Figure 4.51 The Marmousi real velocity model.

Figure 4.52 Inversion results of the model Marmousi. (a) The 100th iteration; (b) The 200th iteration.

Figure 4.53 Velocity value of the 120th channel with various iterations. (a) The 100th iteration. (b) The 200th iteration. The green line represents real value while the red line represents inversion results.

Chapter 05

Figure 5.1 Seismic time profile through salt dome.

Figure 5.2 Reflection wave characters of main geologic horizons on seismic profile.

Figure 5.3 Geologic structural map of Pre‐Caspian Basin Region.

Figure 5.4 Regional structure division map of the Middle Block.

Figure 5.5 Quality of original single shots. (a) Acquired in 2005. (b) Acquired in 2007.

Figure 5.6 Correlation of single shot at the same location acquired in different years.

Figure 5.7 Profile correlation on the same CDP location in different years.

Figure 5.8 Time domain‐velocity profile (anomalous subsalt velocity).

Figure 5.9 Forward geological model.

Figure 5.10 Relation between salt dome time thickness and upward pull height.

Figure 5.11 Building time‐domain model with interval control method.

Figure 5.12 Space carving of salt dome.

Figure 5.13 Velocity‐depth model built on multiple‐well constraint.

Figure 5.14 Correlation of (a) time migration and (b) depth migration crossing Well AL‐1 line (east‐west).

Figure 5.15 Comparison of low frequency noises from RTM of Marmousi Model (a) before suppression and (b) after suppression.

Figure 5.16 Comparison between (a) one‐way wave migration and (b) reverse time migration results of salt dome model.

Figure 5.17 Comparison between (a) velocity slice and (b) reverse time migration velocity slice of salt dome model.

Figure 5.18 Comparison of imaging results of subsalt structure by (a) one‐way wave migration and (b) reverse time migration.

Figure 5.19 Comparison of imaging slices of subsalt structure by (a) one‐way wave migration and (b) reverse time migration.

Figure 5.20 (a) Iso‐t

o

graph and (b) structural map of the Lower Permian top in mid block.

Figure 5.21 (a) Iso‐t

o

graph and (b) structural map of KT‐I top in mid block.

Figure 5.22 (a) Iso‐t

o

graph and (b) structural map of the Carboniferous KT‐II top in mid block.

Figure 5.23 (a) Iso‐t

o

graph and (b) structural map of the Carboniferous MKT top in mid block.

Figure 5.24 (a) Iso‐t

o

graph and (b) structural map of the Carboniferous Visean top in mid block.

Figure 5.25 Distribution map of traps on KT‐I and KT‐II tops in mid block.

Figure 5.26 Deployment diagram for high‐density 3D seismic survey.

Figure 5.27 Distribution map of high‐density 3D surface types.

Figure 5.28 Plane variation diagram of high‐density 3D surface elevation.

Figure 5.29 Thickness variation diagram of high‐density 3D low‐velocity belt.

Figure 5.30 Analysis of interference wave.

Figure 5.31 Analysis of spectrum.

Figure 5.32 Comparison chart of (a) single shot and (b) spectrum of acquired data by conventional method and high‐density full azimuth method.

Figure 5.33 (a) SNR attribute and (b) original stacked profile.

Figure 5.34 Comparison of velocity spectrum and gather of high‐density full azimuth data at different azimuths.

Figure 5.35 Attribute graph of folds, azimuth and offset in work zone.

Figure 5.36 Single shot before and after frequency division anomalous amplitude processing.

Figure 5.37 The profile before and after frequency division anomalous amplitude processing and noise profile.

Figure 5.38 Cross‐array observation system chart.

Figure 5.39 3D f‐k

x

‐k

y

before and after single shot prior to denoising on cross‐array gather.

Figure 5.40 The figure of refracted wave about different arrays of lateral offset.

Figure 5.41 Distribution of group interval before and after linear moveout correction.

Figure 5.42 Implementation process schematic of linear moveout correction with 2D f‐k.

Figure 5.43 Comparison of CMP gathers (a, b) before and after 3D prestack random noise attenuation.

Figure 5.44 Comparison of profiles before and after serial deconvolution.

Figure 5.45 Comparison of common offset folds before and after data regularization.

Figure 5.46 Comparison of common offset folds before and after data regularization.

Figure 5.47 Comparison of profiles (a) before and (b) after data regularization.

Figure 5.48 Comparison of profiles (a) before and (b) after data regularization.

Figure 5.49 Distribution map of fractures and cavities.

Figure 5.50 Processing flow for PSTM.

Figure 5.51 Criterion for velocity rationality in velocity analysis. (a) Analysis of root mean square velocity on CRP gather. (b) Residual delay analysis to perform fine tuning of velocity.

Figure 5.52 Root mean square velocity model for final PSTM.

Figure 5.53 Building the time interval model.

Figure 5.54 Lateral delay spectrum of target zone.

Figure 5.55 S/P velocity analysis and quality control.

Figure 5.56 Velocity profiles of P‐velocity modeling and S‐velocity modeling.

Figure 5.57 Comparison of P‐modeling and P‐joint modeling imaging profiles. (a) Along‐interval lateral velocity adjustment result. (b) Fine adjustment result of longitudinal velocity.

Figure 5.58 Six azimuth folds and azimuth attribute chart.

Figure 5.59 Comparison of fractures on different azimuth profiles.

Figure 5.60 Remarkable amplitude difference of fractures on different azimuth gather.

Figure 5.61 Velocity difference of “beads” at different azimuths.

Figure 5.62 Plane graph of the obtained anisotropic parameter δ.

Figure 5.63 Anisotropic parameters δ body and ε body.

Figure 5.64 Comparison of (a) isotropic depth migration and (b) anisotropic depth migration.

Figure 5.65 Reverse time flow chart.

Figure 5.66 Comparison of (a) integral PSDM and (b) reverse time PSDM.

Figure 5.67 Comparison of PSTM of (a) conventional data and (b) PSTM of high‐density full azimuth data.

Figure 5.68 Comparison of (a) PSTM and (b) PSDM of high‐density full azimuth data.

Figure 5.69 Bead profile superimposition map for PSTM and PSDM of high‐density full azimuth data.

Figure 5.70 Comparison of (a) Kirchhoff depth migration and (b) reverse time migration.

Figure 5.71 Comparison of (a) Kirchhoff depth migration and (b) reverse time migration.

Figure 5.72 Comparison of (a) Kirchhoff depth migration and (b) reverse time migration.

Figure 5.73 3D seismic survey deployment map for the north foothill of Turpan‐Hami Basin.

Figure 5.74 Division of surface types in the research area.

Figure 5.75 3D seismic survey deployment map for the Qingxi Depression of the Jiuquan Basin.

Figure 5.76 Surface types division of the research area.

Figure 5.77 Geomorphologic map of the research zone.

Figure 5.78 Near‐surface inversion profile of the research area and original single shot record.

Figure 5.79 Original single shots in different locations of Baka region (a) single shot in western work zone (b) single shot in middle work zone (c) single shot in eastern work zone.

Figure 5.80 Profile comparison of static correction effect at different locations. (a) Tomographic static correction profile. (b) Refraction static correction profile.

Figure 5.81 Comparison of near‐surface model for tomographic inversion. (a) Constraint inversion model. (b) Inversion model by conventional method.

Figure 5.82 Comparison of tomographic inversion stacking profile effect. (a) Model‐constrained stacking profile. (b) Non‐model‐constrained stacking profile.

Figure 5.83 Stacking results correlation of various static correction methods for northern Turpan‐Hami foothills. (a) Stacking results of database static correction. (b) Stacking result of integrated chromatographic static correction by fitted reconstruction.

Figure 5.84 Stacking results comparison of various static correction methods for the Qingxi Region of Jiuquan. (a) Profile prior to static correction. (b) Static correction profile by field model. (c) Tomographic static correction profile. (d) Constrained integrated static correction profile.

Figure 5.85 Application effects analysis of residual static correction in the Baka Region. (a) Profile prior to residual static correction. (b) Profile after residual static correction to reflected waves. (c) Profile after global optimization static correction.

Figure 5.86 Application effects analysis of constraint integrated static correction in the Qingxi Region of Jiuquan Basin. (a) Original profile prior to static correction. (b) Profile after constraint integrated basic static correction. (c) Profile after integrated residual static correction.

Figure 5.87 Single shot comparison before and after anomalous amplitude and surface wave suppression. (a) Original single shot. (b) Single shot after noise suppression. (c) Suppressed noise.

Figure 5.88 Single shot linear interference wave step suppression in the northern Turpan‐Hami foothills. (a) Original single shot. (b) Single shot after mid‐low velocity interference suppression. (c) Single shot after high velocity suppression.

Figure 5.89 Step suppression of 3D single shot linear interference in the Qingxi Region of Jiuquan. (a) Original single shot. (b) Single shot after mid‐low velocity interference suppression. (c) Single shot after high velocity suppression.

Figure 5.90 Step suppression of 3D single shot linear interference in the Qingxi Region of Jiuquan. (a) Original single shot. (b) Original single shot spectrum. (c) Original single shot f‐k spectrum. (d) Single shot after suppression. (e) Single shot spectrum after suppression. (f) Single shot f‐k spectrum after suppression.

Figure 5.91 Suppressing of prestack domain noise. (a) Original shot gather. (b) Result after shot gather noise suppression. (c) Gather of common geophone points. (d) Results after geophone domain noise suppression.

Figure 5.92 Noise suppression comparison of prestack multi‐domain and cross array domain. (a) Original shot gather. (b) Result after noise suppression.

Figure 5.93 Comparison of prestack step‐by‐step and multi‐domain noise suppression. (a) Original single shot. (b) Result of shot‐domain noise suppression. (c) Result of geophone‐domain noise suppression. (d) Result of cross‐array surface wave suppression. (e) Result of cross‐array domain anomalous energy suppression. (f) Result of linear noise cross‐array domain suppression.

Figure 5.94 Comparison of PSTM profiles for high precision root mean square velocity modeling. (a) PSTM profile for conventional velocity modeling method. (b) PSTM profile for high precision velocity modeling method.

Figure 5.95 Final interval velocity‐depth model. (a) Final interval velocity profile. (b) Final interval velocity body.

Figure 5.96 Final interval velocity‐depth model. (a) PSTM profile. (b) PSDM profile.

Figure 5.97 PSDM effects comparison between floating datum and fixed datum of some survey line. (a) PSDM of fixed surface. (b) PSDM of float datum.

Figure 5.98 Comparison of new and previous migration imaging results for highly steep structures. (a) Previous migration result. (b) New migration result.

Figure 5.99 PSDM results comparison of different methods. (a) Kirchhoff integral migration. (b) One‐way wave migration.

Figure 5.100 PSDM result comparison of different methods for the Qingxi area. (a) Kirchhoff integral PSDM. (b) Reverse time PSDM.

Figure 5.101 PSDM result comparison of different methods for the Qingxi area. (a) Kirchhoff integral PSDM. (b) Reverse time PSDM.

Figure 5.102 PSDM result comparison of different methods for the northern Turpan‐Hami mountain foothills. (a) Kirchhoff integral PSDM. (b) Reverse time PSDM.

Figure 5.103 Comparison between new and previous PSTM results. (a) Previous migration results. (b) New migration results.

Figure 5.104 Comparison between new and previous PSDM results (inline 644). (a) Previous migration results. (b) New migration results.

Figure 5.105 Comparison between new and previous PSDM results. (a) Previous migration results. (b) New migration results.

Figure 5.106 Representative profile of various structural patterns in the Baka region. (a) Monoclinal structure. (b) Fault structure. (c) Fault anticline structure.

Figure 5.107 Sand body forecast spread map of J

1

b interval.

Figure 5.108 Multi‐attribute fusion plane spread map of J

1

b interval.

Figure 5.109 Comparison between new and previous PSTM results (inline 740). (a) Previous migration results. (b) New migration results.

Figure 5.110 Comparison between new and previous PSTM results (inline 420). (a) Previous migration results. (b) New migration results.

Figure 5.111 Comparison between new and previous migration results (inline 410). (a) Previous PSTM profile. (b) New integral PSDM profile (depth domain).

Figure 5.112 Comparison between new and previous PSDM results (inline 420). (a) Previous integral PSDM profile. (b) New reverse time PSDM profile.

Figure 5.113 Structural map of coal bed top boundary in Badaowan Formation.

Figure 5.114 Favorable reservoir distribution map forecasted by multi‐information fusion.

Figure 5.115 Comparison between new and previous PSTM results. (a) Previous PSTM profile. (b) New PSTM profile.

Figure 5.116 Comparison between new and previous PSTM results before and after the research (I). (a) Previous PSTM profile. (b) New PSTM profile.

Figure 5.117 Comparison between new and previous PSTM results before and after the research (II). (a) Previous PSTM profile. (b) New PSTM profile.

Figure 5.118 Comparison between new and previous PSTM results before and after the research (III). (a) Previous PSTM profile. (b) New PSTM profile.

Figure 5.119 Comparison between PSTM and PSDM results after the research (I). (a) PSTM profile. (b) PSDM profile.

Figure 5.120 Comparison between PSTM and PSDM well‐tie profiles after the research (II). (a) PSTM profile. (b) PSDM profile.

Figure 5.121 Comparison of amplitude on prestack migration well‐time profile.

Figure 5.122 Proposed well site map on sub‐depression in the Qingnan region.

Figure 5.123 Seismic profile I of sub‐depression of the Qingnan Area.

Figure 5.124 Seismic profile II of sub‐depression of the Qingnan Area.

Figure 5.125 Schematic of work zone location.

Figure 5.126 Stacked profile.

Figure 5.127 Record of original single shot (development of high frequency noise).

Figure 5.128 Spectrum of original single shot record (development of high frequency noise).

Figure 5.129 f‐k spectrum of original single shot record (high frequency setting developed in tidal flat).

Figure 5.130 Separation of high frequency noise and signal.

Figure 5.131 Comparison between (a) shot gather sequencing as per trace number and (b) sequencing as per offset.

Figure 5.132 Comparison of (a) high frequency noise before suppression, (b) after suppression and (c) after denoising.

Figure 5.133 Effect graph of high frequency coherent noise suppression.

Figure 5.134 Effect graph of air‐bubbled secondary source suppression.

Figure 5.135 Folds comparison of (a) common‐offset profile bin and (b) after homogenization.

Figure 5.136 PSDM profile comparison (a) before and (b) after bin homogenization.

Figure 5.137 Folds comparison (a) before and (b) after bin homogenization.

Figure 5.138 PSDM profile comparison (a) before and (b) after bin homogenization.

Figure 5.139 Initial velocity model in depth domain obtained from longitudinal velocity modeling.

Figure 5.140 Determining velocity interface as per log data.

Figure 5.141 Building time horizon model.

Figure 5.142 Comparison of Ed

2

zone velocity before and after optimization. (a) Velocity prior to correction. (b) Depth domain residual moveout. (c) The velocity after reflected wave tomography.

Figure 5.143 P‐S velocity analysis and quality control.

Figure 5.144 Final velocity model. (a) Horizon depth. (b) Layer velocity (c) Final velocity‐depth model.

Figure 5.145 Flow chart for grid tomographic imaging.

Figure 5.146 Various attribute chart for seismic data.

Figure 5.147 Comparison of velocity profiles before and after grid tomographic imaging.

Figure 5.148 Comparison of depth migration profiles (a) before and (b) after grid tomographic imaging.

Figure 5.149 Comparison between (a) curved ray PSTM and (b) asymmetric travel time PSTM.

Figure 5.150 Profile comparison between (a) Kirchhoff PSDM and (b) reverse time migration (I).

Figure 5.151 Profiles comparison between (a) Kirchhoff PSDM and (b) reverse time migration (II).

Figure 5.152 Profile comparison between (a) previous migration result and (b) new migration results (I).

Figure 5.153 Profile comparison between (a) previous migration result and (b) new migration results (II).

Figure 5.154 Time slices comparison between (a) previous PSDM result and (b) new reverse time migration result.

Guide

Cover

Table of Contents

Begin Reading

Pages

ii

iii

iv

vii

viii

ix

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

Wiley Series in Petroleum Industry Press series list

Relative Fidelity Processing of Seismic Data

Xiwen Wang

March 2017

Sand Production Management for Unconsolidated Sandstone Reservoirs

Shouwei Zhou and Fujie Sun

January 2016

Relative Fidelity Processing of Seismic Data

Methods and Applications

Edited by Xiwen Wang et al

PetrochinaLanzhou, GansuChina

This edition first published 2017© 2017 John Wiley & Sons Ltd

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Xiwen Wang to be identified as the editor of this work has been asserted in accordance with law.

Registered OfficesJohn Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USAJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office9600 Garsington Road, Oxford, OX4 2DQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of WarrantyWhile the publisher and authors have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Wang, Xiwen, 1956–Title: Relative fidelity processing of seismic data : methods and applications / Professor Xiwen Wang, Petrochina, Lanzhou, Gansu [Province], CH [China].Description: 1 edition. | Hoboken, NJ : John Wiley & Sons, Inc., 2017. | Series: Wiley series in petroleum industry press | Includes bibliographical references and index.Identifiers: LCCN 2016049496 (print) | LCCN 2016049872 (ebook) | ISBN 9781119052906 (cloth) | ISBN 9781119052937 (pdf) | ISBN 9781119052920 (epub)Subjects: LCSH: Seismic prospecting. | Petroleum–Prospecting. | Petroleum–Geology. | Petrology.Classification: LCC TN271.P4 W344 2017 (print) | LCC TN271.P4 (ebook) | DDC 622/.1592–dc23LC record available at https://lccn.loc.gov/2016049496

Cover Design: WileyCover Image: © Jackie2k/Gettyimages

Preface

In recent years, lithology reservoir has gradually become the major field of reservoir gain in China. The lithology analysis poses high requirements for amplitude preservation of seismic migration results, but the previous prestack depth migration methods mainly focus on structural imaging. Therefore, much study has been carried out in terms of the amplitude preservation prestack depth migration method using geophysics. Meanwhile, the reservoir of lithology formation is under the control of multiple factors, such as regional structures and depositional facies belts. It presents great exploration difficulties due to being complicated and concealed. On seismic profile, the display reliability of response to lithology formation is subject to the fidelity of processing.

Currently, it is difficult to realize absolute seismic fidelity preservation processing, but relative fidelity preservation processing is possible, which poses higher requirements for seismic data processing, requiring amplitude preservation. It needs to avoid adopting the module that will damage the amplitude relationship of neighboring seismic channels in the processing procedure. In view of the seismic processing problem of lithology reservoir exploration, the relative fidelity preservation processing method that is represented in this book has four critical principles for wide‐frequency relative fidelity preservation processing: (1) Protection of effective band. The widening of effective band should be based on SNR of data, high‐frequency widening and be conducted for the low SNR seismic data; the bandwidth and dominant frequency of deconvolution operator should be controlled. The main effect of deconvolution will be the elevation of energy of high‐frequency component within the effective band, to avoid the occurrence of serious notching in the high‐frequency end of effective band. (2) Protection of low frequency, especially that of 3 ~ 8 Hz. The scenarios of low frequency information suppressed due to high resolution, high SNR processing or significant loss of low frequency end effective information with adoption of strong j‐k denoising, should be avoided. (3) Amplitude preservation. Modification module such as RNA should not be applied, to avoid damaging the lateral relationship of seismic channel amplitude. (4) Phase preservation. The module that will damage phase position should be avoided during processing. Zero phase deconvolution and surface consistent deconvolution will not damage relationship of phases. Compared with high‐resolution high SNR processed profile, relative fidelity preservation processed profile could reflect seismic response of subsurface sand reservoir quite genuinely. While there is discussion on the processing method of relative fidelity preservation processing or high‐resolution high SNR, the impact on seismic processing by acquisition mode of seismic data is also elaborated in this book. Seismic broadband acquisition is the basis for relative fidelity preservation processing. Based on profile reflection traits and spectrum characteristics, the impact on the data processed with high‐resolution, high SNR method or data processed with relative fidelity preservation method by means of seismic acquisition source and observation approach is also analyzed and discussed.

In the previous seismic exploration, in order to satisfy the requirements of structural interpretation, SNR was given much more attention for seismic data imaging and the consideration of fidelity and amplitude preservation of data was insufficient. Nowadays, with the promotion of oil and gas exploration, the requirement for fidelity of seismic data processing is much higher. In recent years, the author has done much in this respect and the research idea and methodology for fidelity preservation imaging have gradually been developed.

The study of the relative fidelity preservation processing method, elaborates the method and principle of the Kirchhoff integral migration method based on ray theory, the one‐way wave equation migration based on wavefield extrapolation and two‐way wave equation reverse time migration. Meanwhile, the study status of respective amplitude preservation migration algorithm is introduced. The methods of critical technology, including reverse time migration imaging based on a GPU/CPU system for complicated structures and depth domain velocity modeling for complicated structural zone, are represented. Also, the pertinent fidelity preservation imaging processing measures and application performance are applied in practice for different fields, including subsalt imaging of 3D prestack reverse time migration, high‐density, all‐round seismic data processing for carbonate region, seismic imaging for complicated structures in the Tuha and Ji‐uquan Basins, and seismic prestack imaging of buried hill structural zone in the Nanpu of Jidong Oilfield. In view of the technical difficulties of seismic imaging and structural interpretation existing in these regions, in‐depth study on seismic prestack imaging and integrated interpretation has been carried out, with the formation of a pertinent technical series for seismic prestack imaging.

In terms of imaging method, it is difficult for the current regular integral seismic method to delineate subsalt imaging, while the two‐way wave equation method (WEM) could generate premium image for complicated deep salt structure. The reverse time migration imaging technology adopted in this book solves the problem of subsalt imaging, which significantly improves the imaging quality for the fairly large dip structure such as vertical fault, salt dome flank and salt domes, eliminating the distortion caused to the underlayer by abnormal velocity of salt dome, making the subsalt formation capable of being accurately imaged in depth field. Study on high‐density all‐round imaging technology in terms of carbonate seismic data is described in this book, with the formation of prestack imaging technical flow for high‐density all‐round seismic data, of which the critical technologies are: high fidelity all‐round noise suppression technology, enhancing prestack vertical resolution technology, all‐round data regularization technology, all‐round high accuracy migration velocity modeling technology, sub‐azimuth prestack depth migration technology. The result of prestack depth migration dominated by reverse time migration accurately delineates the subsurface structure pattern and location of fault for complicated block, and is especially capable of clearly delineating abdominal block of anticline, small displacement fault, and, consequently, provides a reliable basis of data for structural interpretation. By utilizing results of prestack depth migration, comprehensive study such as fine structural interpretation and reservoir prediction is conducted, structural details in the study area are reconfirmed, and the fault characteristics as well as the regularity of formation distribution are further understood. In this book, also, the asymmetrical travel time prestack time migration and prestack reverse time migration are combined, which solves the impact on imaging for lateral velocity variation. Where the structures are complicated and the dip is large, the performance of imaging is much better than the previous results from studies and the final processing results represent various geological characteristics, which are worth sharing. Currently, in theory, 3D reverse time depth migration is the most accurate imaging method, which is a surefire result of the development of prestack depth migration imaging technology and is the direction of development for accurate imaging when combined with FWI velocity estimation method.

During the study for this book, the R&D was closely integrated with production, and pertinent technical series formed during study were applied in a timely way for oilfield exploration. With the application of new study results, the overall structural pattern and local structural characteristics of a series of blocks were accurately confirmed and favorable targets of exploration and drilling were optimally selected with the achievement of excellent application performance.

The preface is by Wang Xiwen et al. Chapter 1 by Wang Xiwen; Chapter 2 by Lv Bin, Wang Yuchao and Han Linghe; Chapter 3 by Wang Yuchao, Liu Wenqing, Wang Xiaowei, Xu Xingrong, Lv Bin, Zeng Huahui, Lu Lieqin and Zhao Yulian; Chapter 4 by Liu Wenqing, Wang Xiwen and Lyu Bin; section 5.1 of Chapter 5 is by Lv Wenqing, Zhang Jing and Shao Xichun; section 5.2 of Chapter 5 is by Wang Xiaowei, Tian Yancan and Zhang Tao; section 5.3 of Chapter 5 is by Su Qin, Xiao Mingtu, Li Hailiang, Lv Bin and Shao Xichun; section 5.4 of Chapter 5 is by Tian Yancan, Zeng Huahui and Zhang Tao. The whole book is modified and revised by Wang Xiwen.

The study projects involved in this book are supported by PetroChina E & P Company, Tarim Oilfield Subsidiary, Yumen Oilfield Subsidiary, Tuha Oilfield Subsidiary, Jidong Oilfield Subsidiary, Mid‐Asia and Russia Research Branch of Research Institute of Petroleum Exploration and Development (RIPED) of CNPC. The preparation of this book is supported and assisted by Director Yang Jie and other experts of Northwest Branch of Research Institute of Petroleum Exploration and Development (RIPED). We would like to sincerely acknowledge their support in publishing this book.

The authors would welcome comments and questions about the contents of the book.

1Study on Method for Relative Fidelity Preservation of Seismic Data

1.1 Introduction

The reservoir of lithology formation, under the control of multiple factors such as regional structures and depositional facies belt, shows some distribution regularity and also has significant exploration potential. However, the exploration challenge is relatively great due to their complex concealment. In recent years, China has made great breakthroughs and discoveries in this field, proving favorable and huge remaining resource potential; this has gradually become the major field of reservoir gain in China [1–4].

The lithology analysis of these lithology reservoirs poses high requirements for the seismic data processing [5–7]. The first challenge is to increase the resolution of seismic data, so as to identify the geologic features like thin sand layers and sand body pinch‐out points, on the processed seismic data. Second, to preserve the amplitude, and not damage the module of amplitude relation between adjacent seismic traces on processing flow. However, previous study mainly focused on identifying thin layers, increasingly widening the frequency band of seismic data and improving the dominant frequency. Therefore, a lot of high‐resolution seismic processes have been developed for the purpose of increasing seismic dominant frequency.

Increasing the resolution of seismic data is mostly realized by deconvolution method [8–19], which is an important means in seismic data processing. In order to increase the resolution of seismic data, an operator for high frequency deconvolution is devised to identify thin layers. However, which principles should be used to determine the frequency bandwidth and dominant frequency of operator for high frequency deconvolution? Is it possible to increase the dominant frequency of the operator for high frequency deconvolution without restriction? These are the questions that remain to be discussed.

Figure 1.1 presents the comparison of lithology reservoir seismic profiles through SN31 wellblock in the Luxi area of Junggar Basin. In Figure 1.1a, a high‐resolution and high‐S/N seismic processing technology was adopted. The lithology reservoir of J2t0 of SN31 wellblock was interpreted from the seismic profile shown in the figure (it also proved to be sand reservoir through drilling), but from the profile, the continuity of J2t0 reservoir was favorable (in the location of Well S302, the J2t0 extended for more than 1 km to the updip direction of the reservoir until the lithology reservoir pinched out). For this purpose, a number of wells were planned for this complete lithology reservoir. Wells S302 and S302 shown on the figure were among them, but their drilling effectiveness varied greatly. The reservoir in Well S303 is very good and the well is a high oil producer, while reservoir in Well S302 is poor (lithology varied, and J2t0 reservoir is mainly mudstone).

Figure 1.1 Comparison between (a) conventional seismic profile and (b) relative fidelity seismic profile through wells S302 and S303.

From the analysis of the data on Figure 1.1, it was believed that the over emphasizing on the high resolution and high SNR during processing deteriorated the fidelity of processed results, directly resulting in the distortion of lithology reservoir seismic response that was reflected on seismic profile.

On seismic profile, the display reliability of seismic response on lithology reservoirs is subject to the fidelity of seismic processing. Currently, it is very difficult to realize absolute seismic fidelity preservation processing, but relative fidelity preservation processing is possible. Figure 1.1b shows the relative fidelity preservation profile that was reprocessed as per the flowchart in Figure 1.13, in 2006. As clearly shown in the figure, the reflection of J2t0 reservoir in Well S302 is very weak, indicating that the reservoir is no longer present. The pinch‐out point of J2t0 reservoir identified on seismic profile is nearly 300 m to downdip direction of lithology reservoir in Well S302, which matches with the data of Well S302 that has been completed.

This has raised a question: the processing of seismic data is the key in exploration of lithology reservoirs. Only when the seismic response of reservoir properties is truly reflected on the profile after seismic processing to the greatest extent can the seismic data be effectively used to identify the lithology reservoirs.

In view of the above problems, we made analysis by selecting a 2D high‐resolution pilot line of 8 km from the Shidong area in the Junggar Basin across Wells Shidong‐2 and Shidong‐4, on the results of high resolution and high SNR processing and relative fidelity preservation processing. Based on this, we came up with a processing method for relative fidelity of seismic data.

Relative fidelity preservation processing[5] of seismic data should pay attention to the following items: (1) Protection of effective frequency band. The widening of effective band should rely on SNR of seismic data. In order to widen the low S/N seismic data to high frequency band, it is necessary to control the bandwidth and dominant frequency of the deconvolution operator. The main function of deconvolution is to increase the energy of high frequency components within the effective band so as to prevent notch frequency at effective high frequency band. (2) Protection of low frequency, especially that of 3–8 Hz. The scenarios that suppress low frequency data of 10 Hz and below in high resolution and high SNR processing, or give significant loss of effective low frequency information with adoption of strong f‐k noise elimination should be avoided. (3) Amplitude preservation. Modification modules like RNA should not be applied to avoid damaging the lateral relationship of seismic channel amplitude. (4) Phase preservation. The module should not be used as it may damage the phase position in processing, while the zero phase deconvolution and surface consistent deconvolution will not damage the phase relation. Compared with the profile processed by high resolution and high SNR, the profile processed by relative fidelity preservation could reflect the seismic effect on subsurface sand reservoirs quite genuinely.

While discussing the processing method of relative fidelity preservation processing or high‐resolution and high SNR processing, the effect of seismic acquisition manner on seismic data processing is also elaborated in this book. Seismic broadband acquisition is the basis for relative fidelity preservation processing. Based on the seismic pilot survey line, we analyzed and discussed the effects of seismic acquisition source and observation means on the data processed by high resolution and high SNR or by relative fidelity in terms of profile reflection features and spectrum features.

1.2 Discussion on Impact on Processing of High‐resolution, High SNR for Seismic Acquisition and Observation Mode

Figure 1.2 is the t0 contour of the sand bottom boundary in the Cretaceous Qinshuihe Formation of the Shidong area in the Junggar Basin. It is located in the distribution area of the oil layer of the Qingshuihe formation in Well Shidong No. 2 region and Well Shidong No. 4 region, which is highlighted by yellow in the figure. But based on the previous seismic data, it is really difficult to explain the reservoir distribution characteristics of these two wells. Therefore, in 2002, seismic pilot survey lines (blue line) across Wells Shidong No. 2 and Shidong No. 2 were deployed to study and find out a method for improving the quality of field seismic data acquisition, thus meeting the requirements of lithology reservoir exploration.

Figure 1.2 The t0 contour of the sand bottom boundary in Cretaceous Qinshuihe Formation of Shidong area.

The seismic pilot survey lines across Well Shidong No. 2 and Shidong No. 4 is 8 km long, with acquisition factors including: (1) Single well excitation: excitation well depth is 85 ~ 139 m; the dosage of explosive per well is 4 kg and the number of wells is 1, so the total dosage for single well excitation is 4 kg. The offset distance is 50 m. The number of receiving channels is 1200 and the distance between channels is 5 m (or 240 receiving channels with a 25 m interval). (2) Combined well excitation: excitation well depth is 6 m; the dosage of explosive per well is 2 kg and the number of wells is 10, so the total explosive dosage is 20 kg. The offset distance is 50 m. The number of receiving channels is 1200 and the distance between channels is 5 m (or 240 receiving channels with a 25 m interval). (3) Controllable seismic source excitation: 4 seismic sources multiplied by 6 times and the scan frequency is 8 ~ 90 Hz. The offset distance is 25 m. The number of receiving channels is 300 and the distance between channels is 25 m. High resolution and high SNR process (Figure 1.2) were used to increase the resolution of seismic data and find out the reservoir distribution characteristics in the Qingshuihe formation between Well Shidong No. 2 region and Well Shidong No. 4 region. In data processing, zero phase deconvolution was used both at prestack and poststack in order to increase the resolution; additionally, the random noise attenuation modification was employed to improve the signal‐to‐noise ratio.

We analyzed the processing progress based on the five control points shown in Figure 1.3: original single shot record frequency scan (control point ①, Figure 1.4); upon the completion of the first arrival refraction static correction process, we used the single shot record and spectrum analysis results of target layer (control point ②, Figure 1.5) to analyze the quality of data. After completing prestack noise elimination, we used the single shot record and the spectrum analysis results of target layer (control point ③, Figure 1.6) to analyze the effect of noise elimination. After finishing zero phase deconvolution process, we used the single shot record and the analysis results on spectrum of target layer (control point ④, Figure 1.7) to analyze the effect of zero phase deconvolution; after completing the migration process, we used the stacked migration profile and the spectrum analysis results of target layer (control point ⑤, Figure 1.8) to analyze the effect of increasing the resolution.

Figure 1.3 Seismic data high resolution and high SNR processing.

Figure 1.4 Single shot records at same shot point of different seismic sources. (a) Analysis on 0 ~ 20 Hz scanning. (b) Analysis on 10 ~ 120 Hz scanning.

Figure 1.5 Analysis results on single shot record and frequency spectrum of target layer processed by static correction of first arrival refraction.

Figure 1.6 Analysis result on single shot record and frequency spectrum of target layer processed by prestack noise elimination.

Figure 1.7 Analysis results on single shot record and frequency spectrum of target layer processed by prestack zero phase deconvolution.

Figure 1.8 Seismic profile and frequency spectrum of target layers processed by high resolution and high SNR processing.

In Figure 1.4, the original single shot record frequency scan (control point ① in Figure 1.3) is used to analyze the quality of original seismic data, and to analyze the single shot record of 0 ~ 20 Hz scanning at the same shot point from different seismic source.

A little signal (or noise) can be observed in the single shot record of 5 ~ 8 Hz scanning under single deep well excitation and in the single shot record of 3 ~ 6 Hz scanning under combined well excitation, while no signal (or noise) can be observed in the single shot record below 8 Hz scanning under controllable seismic source excitation, indicating the loss of low‐frequency component (the scanning frequency of controllable seismic source excitation is 8 ~ 90 Hz).

Figure 1.4b shows the analysis on single shot records of 10 ~ 120 Hz scanning at the same shot point of different seismic sources. Signals can be observed in the single shot record of 5 ~ 100 Hz scanning of single deep well excitation. Signals can be observed in the single shot record of 4 ~ 80 Hz scanning under combined well excitation. Signals can still be observed in the single shot record of 8 ~ 100 Hz scanning under controllable seismic source excitation.

From the perspective of scanning signal, single deep well and controllable seismic source excitation have wide frequency bandwidth, while combined well excitation has narrow bandwidth; low frequency component below 8 Hz is missing in controllable seismic source excitation.

Figure 1.5