Access the full text.
Sign up today, get DeepDyve free for 14 days.
M. Widess (1982)
Quantifying resolving power of seismic systemsGeophysics, 47
R. Wu, M. Toksöz (1987)
Diffraction tomography and multisource holography applied to seismic imagingGeophysics, 52
G. Vermeer (1998)
Factors affecting spatial resolutionGeophysics, 64
Ralph Knapp (1990)
Vertical resolution of thick beds, thin beds, and thin-bed cyclothemsGeophysics, 55
G. Clarke (1968)
Time-Varying Deconvolution FiltersGeophysics, 33
J. Robertson, H. Nogami (1984)
Complex seismic trace analysis of thin bedsGeophysics, 49
N. Ricker (1945)
The Computation of Output Disturbances from Amplifiers for True Wavelet InputsGeophysics, 10
M. Brühl, G. Vermeer, Michael Kiehn (1996)
Fresnel zones for broadband dataGeophysics, 61
M. Safar (1985)
On the lateral resolution achieved by Kirchhoff migrationGeophysics, 50
J. Chen, G. Schuster (1999)
Resolution limits of migrated imagesGeophysics, 64
Zishun Li, Xuebin Guo (2007)
Predicting the distribution of thin bed reservoirs by broad frequency band seismicApplied Geophysics, 4
Ma Zai-tian (2005)
Theoretical Analysis of Reflection Seismic Imaging ResolutionJournal of Tongji University
D. Okaya (1995)
Spectral properties of the earth’s contribution to seismic resolutionGeophysics, 60
S. Treitel (1969)
PREDICTIVE DECONVOLUTION: THEORY AND PRACTICEGeophysics, 34
H. Deng (1992)
Seismic wave propagation in thinly-layered media with steep reflectorsSeg Technical Program Expanded Abstracts
Jing-Hua Gao, Wen-Chao Chen, Youmin Li, Fang-Bao Tian (2003)
Generalized S Transform and Seismic Response Analysis of Thin Interbedss Surrounding Regions by GpsChinese Journal of Geophysics, 46
J H Gao, W C Chen, Y M Li (2003)
Generalized S transform and seismic response analysis of thin interbedsChinese Journal of Geophysics, 46
D. Seggern (1991)
Depth Imaging Resolution of 3-D Seismic Recording PatternsSeg Technical Program Expanded Abstracts
Yanghua Wang (2002)
A stable and efficient approach of inverse Q filteringGeophysics, 67
A A Mateeva (2003)
Thin horizontal layering as a stratigraphic filter in absorption estimation and seismic deconvolution
J C Bancroft (2007)
SEG
R J Qian (2008)
Characteristics of the Seismic Waves and Related Technical Analysis
R. Knapp (1991)
Fresnel zones in the light of broadband dataGeophysics, 56
(1968)
Time-varying deconvolution filters. Geophysics
M. Tygel, J. Schleicher, P. Hubral (1994)
Pulse distortion in depth migrationGeophysics, 59
J. Bancroft (2007)
A Practical Understanding of Pre- And Poststack Migrations
C. Varela, A. Rosa, T. Ulrych (1993)
Modeling of attenuation and dispersionGeophysics, 58
M. Widess (1973)
HOW THIN IS A THIN BEDGeophysics, 38
P. Goupillaud (1961)
AN APPROACH TO INVERSE FILTERING OF NEAR-SURFACE LAYER EFFECTS FROM SEISMIC RECORDS*Geophysics, 26
P. Riel, A. Berkhout (1985)
Resolution in seismic trace inversion by parameter estimationGeophysics, 50
M. Schoenberger (1974)
RESOLUTION COMPARISON OF MINIMUM-PHASE AND ZERO-PHASE SIGNALSGeophysics, 39
R. Kallweit, L. Wood (1982)
The limits of resolution of zero-phase waveletsGeophysics, 47
E. Kjartansson (1979)
Constant Q-wave propagation and attenuationJournal of Geophysical Research, 84
N. Ricker (1953)
Wavelet Contraction, Wavelet Expansion, and the Control of Seismic ResolutionGeophysics, 18
K. Marfurt, R. Kirlin (2001)
Narrow-band spectral analysis and thin-bed tuningGeophysics, 66
Di Bang-rang, GU Pei-cheng (2005)
Quantitative analysis of resolution of seismic migration imagingJournal of the University of Petroleum,China
E. Robinson (1967)
Predictive decomposition of time series with application to seismic explorationGeophysics, 32
Junbin Huang, Li Gao, Y. Gao (2007)
Side lobes of wavelets impact identification of thin sand bodiesApplied Geophysics, 4
Fu-tian Liu, P. Xu, K. Chun, Jin‐Song Liu, Z. Yin, Xian-kang Zhang, Cheng-ke Zhang, Jin-ren Zhao (2003)
Crustal Velocity Structure of the Deep Continental Subduction Zone — A Wide Angle Reflection/Refraction Seismic Study on the Eastern Dabie OrogenChinese Journal of Geophysics, 46
Pet.Sci.(2013)10:195-204 195 DOI 10.1007/s12182-013-0267-4 A joint high-resolution processing method and its application for thin inter-beds 1, 2 1, 2 Liu Zhiwei and Wang Yanchun School of Geophysics and Information Technology, China University of Geosciences, Beijing 100083, China .H\/DERUDWRU\RI*HRGHWHFWLRQ0LQLVWU\RI(GXFDWLRQ&KLQD8 © China University of Petroleum (Beijing) and Springer-Verlag Berlin Heidelberg 2013 Abstract: Seismic processing characterizing thickness and borders of thin inter-beds has gradually evolved from post-stack migration to pre-stack migration, and the latter considers both vertical and lateral resolutions. As the key processing methods for improving vertical and lateral resolution, conventional GHFRQYROXWLRQDQGSUHVWDFNWLPHPLJUDWLRQ3670 DUHQRWVLPSO\GRPLQDWHGE\WKHHVWLPDWLRQDQG compression of the wavelet because of its instability. Therefore, considering the variations of wavelet IUHTXHQF\EHIRUHGXULQJDQGDIWHU3670FDQREWDLQJRRGFRPPRQUHIOHFWLRQSRLQW&53 JDWKHUV and imaging profiles of thin inter-beds. Based on the frequency characteristics of the wavelet before, GXULQJDQGDIWHU3670DMRLQWKLJKUHVROXWLRQSURFHVVLQJGPHWKRIRUWKLQEHGVLQWHULVSURSRVHGLQWKLV paper, including inverse Q¿OWHULQJIRUKLJKIUHTXHQF\FRPSHQVDWLRQEHIRUH3670RSWLPXPZHLJKWLQJ .LUFKKRII3670IRUSUHVHUYLQJKLJKIUHTXHQFLHVGXULQJ3670DQG ZDYHOHWKDUPRQL]HUGHFRQYROXWLRQ IRUFRQVLVWHQWSURFHVVLQJDQGIUHTXHQF\EDQGEURDGHQLQJDIWHU3670$QDSSOLFDWLRQWRUHDOGDWD characterized by mudstone beds in the Oriente Basin proved that the joint high-resolution processing method is effective for determining the thickness and borders of thin inter-beds and is favorable for subsequent reservoir prediction and seismic inversions. Key words: Thin inter-bed, seismic wavelet, inverse Q¿OWHULQJRSWLPXPZHLJKWIXQFWLRQKDUPRQL]HU deconvolution Seismic vertical resolution refers to the ability to 1 Introduction identify reflected interfaces from wavelets. Rayleigh Generally, thin inter-beds refer to a geologic body which is studied Fraunhofer optical diffraction and proposed that composed of two or more thin layers intermingled with beds two neighboring peaks can be visually distinguished if the of more normal thickness. The lithology difference results in time interval of the wavelet is not less than the peak-trough different wave velocities in neighboring layers, so the bottom interval ' W , and the resolution limit is a quarter of the and top interfaces of one layer have opposite reflection wavelength, commonly called the Rayleigh criterion (Fig. FRHI¿FLHQWVDQGWKHWUDYHOWLPHLVVKRUW%HFDXVHRIOLPLWHG 5LFNHU VWXGLHGWKHFRPSRVLWHZDYHIRUPE\ IUHTXHQF\EDQGRIWKHZDYHOHWWKHVHLVPLFUHÀHFWLRQRIWKLQ convolving a zero-phase wavelet with two pulses of equal inter-beds is a result of superposition and interference of amplitude and polarity, and showed that the composite seismic wavelets (Widess, 1973). Vertically, the superposition ZDYHIRUPEHFRPHVÀDWEHWZHHQWKHWZRSXOVHVZKHQWKHWLPH and interference weaken the reflectivity coefficients of interval of the two pulses is 2 , which is defined as the ' W interfaces, and decrease the resolution of thin inter-beds. A vertical resolution limit, i.e., 1/4.6 of wavelength, now called seismic interpreter cannot properly distinguish the boundaries WKH5LFNHUFULWHULRQ)LJ LGHVV: VWXGLHGD of a single layer (Knapp, 1990). Laterally, some wedges composite waveform by convolving a zero-phase wavelet exist in the positions with lithologic changes in the seismic with two pulses of equal amplitude and opposite polarity. SUR¿OHVHYHQWKHPXGVWRQHEHGV7KHVHVWURQJHYHQWVDSSHDU He concluded that the composite waveform converges to the discontinuous because the diffracted amplitudes of two sides derivative of the wavelet with a decreasing interval between of the diffracted point cannot be weakened after migration for the two pulses and indicated that the resolution limit is 1/8 different velocities (Robertson and Nogami, 1984). Therefore, of the wavelength if the amplitude of the composite wave thin inter-beds imaging is complex and requires high vertical is considered. Kallweit and Wood (1982) made a summary and lateral resolution. of vertical resolution and showed that the Ricker criterion can be applied to both equal and opposite polarity situations of one doublet. Van Riel and Berkhout (1985) pointed the *Corresponding author. email: zwliu007@sina.com disadvantages of the above criteria: 1) only the interference Received September 18, 2012 QLYHUVLW\RI*HRVFLHQFHV%HLMLQJ&KLQD 196 Pet.Sci.(2013)10:195-204 of the central peak with the first trough is considered, but investigated seismic imaging resolution using geophysical the influence of side lobes of the wavelet on resolution is models. Generally, the lateral resolution is evaluated by the ignored. 2) Rayleigh and Ricker criteria are both based on an )UHVQHO]RQH&RQVLGHULQJDÀDWUHÀHFWRUDQGDSRLQWVRXUFH isolated doublet, but the multi-layer case is more important in practice. 3) A good result is obtained in the equal polarity RIWKH¿UVWRUGHU)UHVQHO]RQH7KHHQHUJ\IURPWKHKLJKHU situation, but in the opposite polarity situation, which is more order Fresnel zones cancel each other, that is, the contribution important in practice, the result is not analyzed in detail. As RIWKH¿UVWRUGHU)UHVQHO]RQHGHWHUPLQHVWKHDFWXDOUHÀHFWLRQ a result, Berkhout introduced the wavelet length criterion response, as shown in Fig. 2. According to the Rayleigh for the vertical resolution. He proved that the dominant criterion, the dominant frequency of a seismic wavelet frequency is a key parameter in the Rayleigh and Ricker determines the lateral resolution. In migration imaging, criteria. Knapp (1991) proposed the absolute resolution and with the downward continuation of the source and receiver, concluded that the vertical resolution is related with the the seismic wavefield increasingly approximates the actual frequency of the wavelet, that is, the higher the frequency, the higher the resolution. The vertical resolution of seismic data migration is an essential approach to improve seismic lateral can be improved in two ways, deconvolution and parameter resolution. Pre-stack migration gradually replaced post-stack LQYHUVLRQ0RVWSDUDPHWHULQYHUVLRQPHWKRGVDVVXPHWKDW migration given the advantage of enhancing velocity analysis the formation is composed of a number of strong reflectors accuracy and preserving relative amplitudes. Accordingly, the ZLWKLQDJLYHQGHSWKUDQJHDQGWKHUHÀHFWLRQVRIWKLQLQWHU improvement of lateral resolution of thin inter-beds relies on beds are considered as band-limited noise and are even the dominant frequency of the seismic wavelet and pre-stack ignored, so the method is not suitable for improving thin migration method. EHGLQWHULPDJLQJ'HFRQYROXWLRQDVVXPHVWKDWDQ\UHÀHFWLRQ SR is a convolution of the zero-phase wavelet with reflectivity coefficients. Deconvolution can obtain a series of smooth reflectivity coefficients and the key is wavelet compression and reshaping. As a result, the frequency bandwidth and the frequency of seismic wavelet play an important role in improving the vertical resolution of thin inter-beds. Amplitude Basic wavelet ' W 1.5 0.5 ms 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 -0.5 Fig. 25HÀHFWLRQLQWHUIHUHQFHRI)UHVQHO]RQHV -1 ' W Therefore, the imaging processing of thin inter-beds should consider the frequency variation of wavelet before, during and after migration. Conventional surface consistent Fig. 1 Resolution limit of the zero-phase Ricker wavelet deconvolution (SCDC) and Kirchhoff pre-stack time PLJUDWLRQ3670 KDYHEHHQSURYHGWRLPSURYHUHVROXWLRQ The lateral resolution has been discussed by many for structural processing. However, for thin inter-beds, some authors on post-stack and pre-stack migration and diffraction new problems appear such as wavelet instability, amplitude- WRPRJUDSK\X:DQGRNVऺ]7 6DIDU DQDO\]HG SUHVHUYDWLRQRIWKHZHLJKWIXQFWLRQRII.LUFKKRI3670DQG the limiting factors to the lateral resolution by Kirchhoff relative amplitude-preserved frequency expansion. We present migration and concluded that seismic lateral resolution is a joint high-resolution processing method for thin inter-beds, controlled not only by the migration aperture and frequency emphasizing analyzing the frequency variations of wavelet but also by the weighting function, velocity accuracy and EHIRUHGXULQJDQGDIWHU3670DQGVHOHFWFRUUHVSRQGLQJ spatial sampling. Von Seggern (1994) proposed visible techniques to improve vertical and lateral resolution, which UHVROXWLRQIRU'JHRORJLFERGLHV.QDSS GH¿QHGWKH DUHZDYHOHWKLJKIUHTXHQF\FRPSHQVDWLRQEHIRUH3670 Fresnel zone for lateral resolution study based on the relation ZDYHOHWIUHTXHQF\SURWHFWLRQGXULQJ3670DQGZDYHOHW of the Fresnel zone and wavelet length. Brühl and Vermeer IUHTXHQF\FRQVLVWHQWH[WHQVLRQDIWHU3670 SRLQWHGRXWWKDWV.QDSS¶GH¿QLWLRQLQDQDUURZEDQG situation cannot converge to the classical Fresnel zone and 2 Model & method JDYHDUHGH¿QLWLRQEDVHGRQWKHUDGLXVRIWKHUHÀHFWRUZLWK PD[LPXPHUPHHUUHÀHFWHG J\9HQHUDQDO\]HGIDFWRUV 2.1 Inverse Q¿OWHULQJ affecting seismic lateral resolution, including the velocity PRGHOPLJUDWLRQDSHUWXUHIROGDQGVSDWLDOVDPSOLQJ0D When a seismic wavefield propagates in thin inter-beds, (2005) and Di and Gu (2005) qualitatively and quantitatively high frequencies are lost because of spherical dispersion and UHÀHFWRUDQGWKH)UHVQHO]RQHEHFRPHVVPDOOHU8QGRXEWHGO\ WKHUHÀHFWHGHQHUJ\RIUHFHLYHULVPRVWO\IURPWKHLQWHUIHUHQFH Pet.Sci.(2013)10:195-204 197 VWUDWLJUDSKLF¿OWHULQJVRWKHVHLVPLFUHVROXWLRQUHGXFHVDQG source in homogeneous media, whereas the wavelet’s VLJQDOLQVWDELOLW\HQKDQFHVLQODWHUWLPH'HQJDUHOD9 GRPLQDQWIUHTXHQF\DQGEDQGZLGWKVLJQL¿FDQWO\GHFUHDVHVLQ HWDO0DWHHYD )LJD VKRZVWKDWWKHVHLVPLF a thin layer medium with linearly varying velocity, as shown wavelet retains the frequency characteristics of a vibrating in Fig. 3(b). Amplitude Amplitude -7 2.5 ×10 2.5 (a) 2.7 2.7 2.9 2.9 3.1 3.1 (b) 3.3 3.3 0 10 20 30 40 3.5 3.5 (a) (b) (c) Fig. 3 High-frequency absorption of a seismic wavelet (a) Homogenous medium, (b) Layer medium with linear velocities, (c) Spectra of (a) and (b) Deconvolution, spectral whitening and frequency where, and f is a cut vf() vf( ){1 1/ ʌ Q ln f / f } cc compensation based on wavelet transform are widely used frequency, and the phase velocity is determined by the in seismic processing (Gao et al, 2003). Constructing a time- wavelet frequency. Frequency attenuation increases with an LHQHUGRPDLQRSWLPDO:¿OWHULQGHFRQYROXWLRQKDVWRSURYLGH increase in frequency and propagation time. For the solution treatments for both early and late time seismic records, so of the Q matrix, the most stable methods are discrete Q in the present deconvolution methods mainly have a window- vertical seismic profile (VSP) and scanning Q analysis in dependent design. However, the design leads to destruction seismic data. Inverse Q¿OWHULQJFDQEHXVHGWRFRPSHQVDWH of seismic wavelets in the overlapping segments. Spectral frequency-dependent seismic energy attenuation and phase whitening flattens the amplitude spectra as a kind of zero- distortion, and to improve the stability of the seismic signal phase deconvolution, and when the reflection coefficient is (Kjartansson, 1979). We simulated inverse Q filtering with QRWZKLWHQRLVHVSHFWUDOZKLWHQLQJZLOOGDPDJHWKHUHÀHFWLRQ the following results. Assuming that the constant Q equals to FRHI¿FLHQWV+LJKIUHTXHQF\FRPSHQVDWLRQEDVHGRQZDYHOHW 150, the wavelet is trace 1 in Fig. 4. The results of applying transform decomposes seismic data into different scales and inverse QIHUHQW¿OWHULQJWRWUDFHZLWKGLI Q values are traces then operates gain processing. Because the wavelet function 2, 3, and 4 in Fig. 4, showing that when the Q value is close and scale function at any time are orthogonal, the modeled to 150, the result is the best. processing is disadvantageous for the orthogonality. Wavelet -100 0 200 stability and white reflection coefficients play a key role in Q = 150 the compensation of high frequency. A seismic signal’s instability results from inelasticity and inhomogeneity of subsurface media, and inverse spherical divergence and inelastic attenuation correction before Q = 300 deconvolution are performed. The high frequency attenuation can be described as a function of the quality factor Q and be compensated by inverse QDQJ¿OWHULQJ*RXSLOODXG: 2002). During propagating, the seismic signal suffers phase Q = 150 distortion related to the frequency f and quality factor Q, and the seismic signal can be written as Q = 50 ª ʌ W/( Qf iHº f) ¬¼ Wt (,WW ,z) IFT{W(f , ,z)e } (1) Assuming that Q does not vary with frequency, then -100 ms 200 ʌ fz Qv f Wt WW z IFTW f z (2) Fig. 4 Simulation of inverse Q Q values Time, s Time, s Amplitude ¿OWHULQJZLWKGLIIHUHQW 198 Pet.Sci.(2013)10:195-204 performed, and the results were shown in Fig. 5(b), (c), and 2.2 Optimum weighting Kirchhoff PSTM (d). We can see that the Kirchhoff migration has a great .LUFKKRII3670LVEDVHGRQWKH.LUFKKRIILQWHJUDO advantage in improving the vertical and lateral resolution. solution of the wave equation, and its migration operator (YHQWKRXJKWKH.LUFKKRII3670LPSURYHVWKHLPDJLQJ response is better than that of frequency–wavenumber (FK) accuracy of inter-beds, the problems of migration noise and and finite difference (FD) migrations (Wu and Toksöz, amplitude preservation still exist which can affect the lateral 1987). We used the inter-beds model as shown in Fig. 5(a) to resolution of thin inter-beds (Chen and Schuster, 1999). The simulate the seismic data with an FD acoustic wave equation. ZHLJKWLQJIDFWRULQWKH.LUFKKRII3670RSHUDWRUGLUHFWO\ The width of the mudstone bed is 40 m, and the wavelet is a Ricker wavelet with a dominant frequency of 35 Hz. Then, I.LUFKKRI3670RIWKLQEHGVLQWHUQHHGVPXFKFRQVLGHUDWLRQ WKH).PLJUDWLRQ)'PLJUDWLRQDQG.LUFKKRII3670ZHUH about the weighting factor. V =2500 m/s rms 1 181 361 541 721 V =2500 m/s rms 1.8 1796 V =2518 m/s rms 1.9 1811 V =2536 m/s rms V =2540 m/s rms 2.0 V =2545 m/s rms 1825 V =2552 m/s rms 1837 V =2565 m/s rms (ms) (b) (a) 1 181 361 541 721 1 181 361 541 721 1.8 1.8 1.9 1.9 2.0 2.0 (c) (d) Comparison of different migration methods for a thin inter-beds model Fig. 5 (a) Time model of thin inter-beds, (b) FK migration, (c) FD migration, (d) Kirchhoff migration The Kirchhoff integral solution of the wave equation can ZKHUHǻ xDQGǻ y are horizontal and vertical trace intervals, be expressed as, and cosș is the dip factor or direction factor, and 1/vr is the spherical spreading factor used for the corrections of wPP TT w ½ ªº ª º (3) amplitudes and phases. Eq. (4) implies that a reflector PxyzW P A ®¾ v³ «» « » ʌ ww zr vr t ¬¼ ¬ ¼ ¯¿ can be imaged at the positions with maximum amplitudes where [P] is the integral of the seismic wavefield in the derived from the summations of seismic wavelets across the propagating path from source to receiver. However, the range A at W tr /v7KH¿UVWWHUP depends on the >ww P / z@ vertical variation of the seismic wavefield, and the second summations lead to a high-frequency reduction in Kirchhoff term is generally called the near-field source and mainly pre-stack migration (Tygel et al, 1994). Theoretically, the varies with 1/r . The two terms in computation are often amplitude of any imaging point in the bin should equal to the LJQRUHG7KHWKLUGWHUPLVJHQHUDOO\FDOOHGWKH¿HOGIDUVRXUFH stacked value of different offset tracks in the bin. The final and constitutes the basis of Kirchhoff migration. So Eq. (3) imaging point lies in the point of tangency with the maximum can be discretized as follows: amplitude and the amplitudes of other points are weakened. Practically, the high-frequency attenuation causes an increase Px(,y ,z,W 2z/v) out out out of seismic wavelet duration and a decrease of vertical and ''xy coswT (4) lateral resolution. The diffraction stack principle is shown in P (,xy ,z 0,W t r/v) ¦ in in in Fig. 6. 4ʌ vr w t LQÀXHQFHVWKHUHÀHFWHGZDYHDPSOLWXGHVDQGIUHTXHQF\VRWKH Pet.Sci.(2013)10:195-204 199 2.3 Wavelet harmonizer deconvolution ss co Deconvolution is one of the standard processing steps. The basic assumptions of deconvolution is minimum ș r phase, invariance of seismic wavelet and white reflection coefficients, and the optimum Wiener filter based on minimum square norm can be designed for improving seismic YHUWLFDOUHVROXWLRQ,QDGGLWLRQVRPHQRQOLQHDU¿OWHUVDUHDOVR GHVLJQHGEDVHGRQRWKHUQRUPVJHU6FKRHQEHU2ND\D Summation verified that a single-trace deconvolution algorithm cannot adapt to relative amplitude-preserved lithologic processing of inter-beds, and multi-trace processing should be used. Fig. 6 Diffraction stack in Kirchhoff migration For example, surface consistent deconvolution (SCDC) and multi-trace predictive deconvolution can basically achieve a The influence of Kirchhoff pre-stack migration on vertical and lateral wavelet consistency and improve seismic wavelet frequency mainly focuses on the weighting factors UHVROXWLRQEHIRUH&03&53VWDFN5RELQVRQ3HDFRFN which include the wavefront spreading factor, dip factor DQGUHLWHO7/LDQG*XR (YHQLIFRQVLGHULQJ (direction factor) and wavelet shaping factor (Bancroft, the optimal weighting factor, diffraction convergence in 2007). The wavefront spreading factor represents the WKH.LUFKKRII3670ZLOOOHDGWRDGHFUHDVHRIIUHTXHQF\ amplitude attenuation when the wavelet propagates, and %DQFURIW DVVKRZQLQ)LJ0RUHRYHUWKHWHPSRUDO conventional preprocessing usually performs spherical and spatial inconsistency of wavelet in preprocessing will spreading compensation. Therefore, this factor does not need OHDGWRPL[LQJRIIUHTXHQFLHVVRDVWRLQÀXHQFHWKHWKLQLQWHU to be applied in Kirchhoff migration. The dip factor (direction beds imaging to a certain degree. factor) has a maximum value at the apex of diffraction and Therefore, wavelet consistent processing is necessary has the largest contribution to integral summation. However, DIWHU.LUFKKRII3670$IWHUSUHSURFHVVLQJGHFRQYROXWLRQ most cases proved that the aperture in Kirchhoff migration DQG3670WKHZDYHOHWVRIGLIIHUHQWRIIVHWVDQGGHSWKV LVPRUHLPSRUWDQWWRLPDJLQJWKDQWKHGLSIDFWRU5HÀHFWLRQ in CRP gathers are consistent, so single-trace predictive imaging and diffraction convergence rely on diffraction deconvolution is used for further improving resolution. VWDFNLQJ,IZHDVVXPHDEDQGOLPLWHGZDYHOHWWKHUHÀHFWHG Conventional single-trace deconvolution based on the wave is identical to the left side of the diffracted wave in assumption of a minimum-phase wavelet generally obtains phase, whereas it has a 180° phase difference from the right the deconvolution operator from different time windows side of the diffracted wave. Then, whether or not to use a and leads to amplitude distortion in the boundaries of time ZDYHOHWVKDSLQJIDFWRULQ.LUFKKRII3670GHSHQGVRQWKH ZLQGRZV+XDQJHWDO&ODUNH DVVKRZQLQ contrast between the migrated trace and corresponding well- Fig. 8. In harmonizer deconvolution, the operator length ORJFXUYHV4LDQ ,QVXPPDU\.LUFKKRII3670IRU and predictive distance are determined first, and then the inter-beds imaging should eliminate the effect of spherical deconvolution operator is calculated for every sampling spreading factor, enhancing migration aperture, and point of the seismic trace, as shown in Fig. 8. The calculated determining the wavelet shaping factor from well-log curves. deconvolution operators may vary in time and space, which Amp. X Amp. t t t (a) (b) (c) Fig. 7 The effect of Kirchhoff migration on a seismic wavelet (a) Wavelet before migration, (b) Kirchhoff diffraction stack path, (c) Wavelet after migration Time (depth) 0DUIXUWDQG.LUOLQ $SSOLFDWLRQVLQSUDFWLFHKDYH 200 Pet.Sci.(2013)10:195-204 1 51 101 151 201 251 301 Window1 Windows Window2 Window3 101 Fig. 8 Time windows of conventional (left) and harmonizer (right) deconvolution (a) can be regarded as a correction of wavelet inconsistency. 1 51 101 151 201 251 301 3 Application The Oriente Basin in Ecuador, South America, is a foreland basin, with low-relief traps and sets of thin inter- EHGVLQWKH;EORFN7KHREMHFWLYHOD\HU<KDVDPXGVWRQH bed, whose origin is structural and sedimentary. The OLWKRORJ\IHUHQFHGLIKDVEHHQYHUL¿HGE\GULOOLQJ7KHVXUIDFH conditions of seismic survey are plain, swamp and jungle, and a stable and horizontal refraction surface (refraction velocity is 2,000 m/s) exists. Apart from strong ground- roll waves, seismic noise mainly consists of low and high frequency abnormal noise distributed irregularly. The valid frequency ranges from 8 Hz to 45 Hz in the Y layer, and the URRWPHDQVTXDUH506 YHORFLW\DSSUR[LPDWHVWR m/s. The processing task is to investigate contacting relations (b) of thin inter-beds and identifying the lateral boundaries of the mudstone bed. Fig. 9 0D[LPXPQHJDWLYHDPSOLWXGHVH[WUDFWHGIURPWKH<OD\HU In preprocessing, a zone filter is used for eliminating (Grid unit: 25m×50m). (a) Before applying inverse Q filter, (b) After applying inverse Q¿OWHU ground-roll waves and suppressing low-frequency abnormal amplitudes. Based on spherical spreading compensation, surface-consistent amplitude processing including surface- consistent amplitude compensation (SCAC), zone amplitude the above analysis of the Kirchhoff migration, we correct the SURFHVVLQJ=$3 DQGVXUIDFHFRQVLVWHQWGHFRQYROXWLRQ ZHLJKWLQJIDFWRUVRIIWKH3670.LUFKKRILQFOXGLQJLQYHUWLQJ (SCDC) are applied for globally preserving horizontal the spherical spreading compensation factor, decreasing the changes of amplitudes. Along the Y layer, negative maximum dip factor and adjusting the shaping factor with reference amplitudes are extracted from the stacked volume after WRWKHZHOOORJ$IWHU3670WKHDPSOLWXGHDWWULEXWHVLPLODU I.LUFKKRI3670DQGWKHPXGVWRQHEHGLQWKHZKLWHFLUFOH to Fig. 9 is extracted along layer Y as shown in Fig. 11 and appears unclear in Fig. 9(a). From Fig. 9(a), the horizontal the waveform of the Y layer is compared with the well- boundary of the mudstone bed is not clear, abnormal ORJLQ)LJ,WLVYHULILHGWKDWWKH.LUFKKRII3670IRU amplitudes exist and the whole attribute is not balanced. We thin inter-beds does not have to correct the shaping factor, use inverse Q filtering before SCDC, a similar amplitude while the spherical spreading and dip factor play a key role attribute extracted from the new migration volume shows that in preserving amplitude attribute and improving the vertical the abnormal amplitudes are suppressed and the sand blocks resolution. Additionally, Fig. 12 also shows that the optimal and mudstone bed become clear in Fig. 9(b). Fig. 10 shows weighting Kirchhoff migration protects the high-frequency the results of corresponding stack section and time-frequency and low-frequency reduction introduced by migration. spectra. We can see that inverse Q filtering can efficiently Although SCDC can be used to obtain lateral consistency compensate high frequency information and improve the RIZDYHOHWVDQGHOLPLQDWHSDUWVRIQHDUVXEVXUIDFHLQÀXHQFHV vertical and lateral resolution of thin inter-beds. on the wavelets, Fig. 14(a) shows that the vertical resolution In Fig. 9(b), there are some abnormal sand blocks and the of the inter-beds is not high enough to separate thin layers lateral resolution near the mudstone bed is low. According to and the amplitude spectra is still flat. In order to separate Pet.Sci.(2013)10:195-204 201 0 25 50 75 100 125 0 25 50 75 100 125 21 41 61 81 101 121 161 21 41 61 81 101 121 161 1.2 1.2 0.5 0.5 1.4 1.4 1.0 1.0 1.6 1.6 1.5 1.5 1.8 1.8 2.0 2.0 2.0 2.0 (a) (b) Fig. 10 Time-frequency spectra and sections of Inline No.50. (a) Before applying inverse Q$IWHUDSSO\LQJLQYHUVHE ¿OWHU Q¿OWHU 1 51 101 151 201 251 301 1 51 101 151 201 251 301 (a) (b) Fig. 11OD\HU*ULGXQLWPîP <0D[LPXPQHJDWLYHDPSOLWXGHVH[WUDFWHGIURPWKH (a) Conventional Kirchhoff migration, (b) Optimal weighting Kirchhoff migration (a) (b) (c) Fig. 12I3670UDFHVDIWHURSWLPDO.LUFKKRI7I3670F UDFHVZLWKRXWRSWLPDO.LUFKKRI7VORJE HOO1R¶:D 202 Pet.Sci.(2013)10:195-204 thin inter-beds, we must improve the vertical resolution of GLVWLQJXLVKHGDQGIUHTXHQF\VSHFWUDDUHDOVRÀDWWHQHG)LJ &53JDWKHUV%DVHGRQWKH¿[HGZDYHOHWRSHUDWRUOHQJWKDQG shows the comparison with well-log before and after applying predictive distance, the wavelet harmonizer deconvolution wavelet harmonizer deconvolution. GHVLJQVD¿OWHUYDU\LQJZLWKWLPHDQGVSDFHRQ&53JDWKHUV )LQDOO\WKHUHVROXWLRQRIWKH;EORFNLVLQFUHDVHGE\WKH for broadening bandwidth. The attribute of negative maximum joint method, and the thin inter-beds and boundaries of the amplitude is shown in Fig. 13(b). A large mudstone bed looks mudstone bed can be identified clearly. Additionally, some very clear and some small and narrow beds also appear in the small-scale geologic formations are well imaged, like thinner left bottom of the attribute map. Fig. 14 shows the results of inter-beds and low-relief traps. The final profiles and CRP stack section and spectra before and after applying wavelet JDWKHUVFDQPHHWWKHUHTXLUHPHQWVRI¿QHUHVHUYRLUSUHGLFWLRQ harmonizer deconvolution. It is noted that thin inter-beds are and seismic inversion. 1 51 101 151 201 251 301 1 51 101 151 201 251 301 201 201 151 151 101 101 51 51 1 1 (b) (a) Fig. 13OD\HU*ULGXQLWPîP <0D[LPXPQHJDWLYHDPSOLWXGHVH[WUDFWHGIURPWKH (a) Before wavelet harmonizer deconvolution, (b) After wavelet harmonizer deconvolution 221 241 261 281 301 321 341 221 241 261 281 301 321 341 1.2 1.2 1.4 1.4 1.6 1.6 1.8 1.8 2.0 2.0 138591 104979 0 50 100 0 50 100 (a) (b) Fig. 14 Stacked sections and amplitude spectra before and after wavelet harmonizer deconvolution. (a) Before wavelet harmonizer deconvolution, (b) After wavelet harmonizer deconvolution Pet.Sci.(2013)10:195-204 203 (a) (b) (c) Fig. 15 (a) Well No.2’s log, (b) Traces without wavelet harmonizer deconvolution, (c) Traces after wavelet harmonizer deconvolution data. Geophysics. 1996. 61(2): 600-604 4 Conclusions Q-&KH DQG6FKXVWHU*75HVROXWLRQOLPLWVRIPLJUDWHGLPDJHV Geophysics. 1999. 64(8): 1046-1053 Thin inter-bed imaging is a completely high-resolution Cla rke G K C. Time-varying deconvolution filters. Geophysics. 1968. lithologic processing method and only joint high-resolution 33(6): 936-944 processing can accurately identify the thickness and borders Den g H L. Seismic wave propagation in thinly-layered media with steep of thin inter-beds. In addition, imaging of thin inter-beds UHIOHFWRUV0DVWHU7KHVLV&HQWHUIRUDYH:3KHQRPHQD&RORUDGR is directly related to other steps of seismic processing 6FKRRORI0LQHV and will fail if a wrong geometry definition, inaccurate Di B R and Gu P C. Quantitative analysis of resolution of seismic VWDWLFFRUUHFWLRQLQDFFXUDWH506YHORFLW\DQGXQVXLWDEOH PLJUDWLRQLPDJLQJ-RXUQDORIWKH8QLYHUVLW\RI3HWUROHXP frequency broadening methods are used. In this paper, the 29(5): 23-32 (in Chinese) joint high-resolution processing method based on amplitude -+&KHQ:&0</LHWDO*HQHUDOL]HG6*DRWUDQVIRUPDQGVHLVPLF preservation can effectively increase the resolution of thin UHVSRQVHDQDO\VLVRIWKLQLQWHUEHGV&KLQHVH-RXUQDORI*HRSK\VLFV inter-beds. The key of joint high-resolution method for 2003. 46(4): 526-532 (in Chinese) Gou pillaud P L. An approach to inverse filtering of near-surface layer imaging thin inter-beds lies in the following aspects: effects from seismic records. Geophysics. 1961. 26(4): 754-760 1) Applying inverse Q filtering before SCDC can QJ+XD-% *DR/-DQG*DR<6LGHOREHVRIZDYHOHWVLPSDFW compensate for high-frequency loss and the amplitudes LGHQWL¿FDWLRQRIWKLQVDQGERGLHV$SSOLHG*HRSK\VLFV can reflect the lateral and vertical variations of subsurface 111-117 lithology. Kal lweit R S and Wood L C. The limits of resolution of zero-phase 2SWLPDOZHLJKWLQJ.LUFKKRII3670EDVHGRQ wavelets. Geophysics. 1982. 47(7): 1035-1046 amplitude compensation, dip, and well-log data can retain UWDQVVRQ(&RQVWDQW4ZDYHSURSDJDWLRQDQGDWWHQXDWLRQ-RXUQDORI.MD the structural and lithologic characteristics of thin inter-beds Geophysical Research. 1979. 82(1): 4737-4748 and reduce frequency mixing. Kna pp R W. Fresnel zones in the light of broadband data. Geophysics. 3) Using wavelet harmonizer deconvolution on CRP 1991. 56(3): 354-359 gathers can globally unify the seismic wavelets and further Kna pp R W. Vertical resolution of thick beds, thin beds, and thin-bed cyclothems. Geophysics. 1990. 55(9): 1183-1190 enhance the vertical and lateral resolution of thin inter-beds. =6DQG*XR;%3UHGLFWLQJWKHGLVWULEXWLRQRIWKLQEHGUHVHUYRLUVE\/L broad frequency band seismic. Applied Geophysics. 2007. 4(2): 118- Acknowledgements H:DUHYHU\JUDWHIXOWRVHQLRUHQJLQHHU/LX-LDQKRQJ IXUW.-DQG.LUOLQ50DU/1DUURZEDQGVSHFWUDODQDO\VLVDQGWKLQEHG senior engineer Duan Wensheng and senior engineer Liu tuning. Geophysics. 2001. 66(4): 1274-1283 HHYD$0DW$7KLQKRUL]RQWDOOD\HULQJDVDVWUDWLJUDSKLFILOWH ULQ -XQMLHIRUWKHLUVFLHQWLILFDGYLFHDQGDVVLVWDQFHGXULQJWKH absorption estimation and seismic deconvolution. Ph.D. Thesis. study. &RORUDGR6FKRRORI0LQHV =77KHRUHWLFDO0DDQDO\VLVRIUHIOHFWLRQVHLVPLFLPDJLQJUHVROXWLRQ References -RXUQDORIRQJML78QLYHUVLW\1DWXUDO6FLHQFH %DQ FURIW-&$3UDFWLFDO8QGHUVWDQGLQJRI3UHDQG3RVWVWDFN 1153 (in Chinese) ROXPH 6(*0LJUDWLRQV9 Oka ya D A. Spectral properties of the earth’s contribution to seismic KO%U HUPHHU09*-2DQG.LHKQ0)UHVQHO]RQHVIRUEURDGEDQ G resolution. Geophysics. 1995. 50(1): 241-251 204 Pet.Sci.(2013)10:195-204 Pea cock K L and Treitel S. Predictive deconvolution: Theory and 5LHO3DQGDQ%HUNKRXW9$-5HVROXWLRQLQVHLVPLFWUDFHLQYHUVLRQE\ practice. Geophysics. 1969. 34(1): 155-169 parameter estimation. Geophysics. 1985. 50(9): 1440-1455 Q5-4LD &KDUDFWHULVWLFVRIWKH6HLVPLFDYHV:DQG5HODWHGHFKQLFDO7 HOD&/5RVD$/DU5DQG98OU\FK7-0RGHOLQJRIDWWHQXDWLRQDQG Analysis. Beijing: Petroleum Industry Press. 2008 (in Chinese) dispersion. Geophysics. 1993. 58(8): 1167-1173 NHU5LF 17KHFRPSXWDWLRQRIRXWSXWGLVWXUEDQFHVIURPDPSOL¿HUVIRU PHHU*-2HU)DFWRUVIHFWLQJ9DIVSDWLDOUHVROXWLRQ7KH/HDGLQJ(GJH true wavelet inputs. Geophysics. 1945. 10(4): 207-220 1998. 17(8): 1025-1030 Ric ker N. Wavelet contraction, wavelet expansion and the control of Von Seggern D. Depth-imaging resolution of 3-D seismic recording seismic resolution. Geophysics. 1953. 18(4): 769-792 patterns. Geophysics. 1994. 59(4): 564-576 HUWVRQ-'DQG1RJDPL++&RPSOH[5REVHLVPLFWUDFHDQDO\VLVRIWKLQ Wan g Y H. A stable and efficient approach of inverse Q filtering. beds. Geophysics. 1984. 49(4): 344-352 Geophysics. 2002. 67(2): 657-663 Rob inson E A. Predictive decomposition of time series with application HVV0%LG:+RZWKLQLVDWKLQEHG"*HRSK\VLFV to seismic exploration. Geophysics. 1967. 32(3): 418-484 1180 DU0+6DI2QWKHODWHUDOUHVROXWLRQDFKLHYHGE\.LUFKKRIIPLJUDWLRQ HVVLG:0%4XDQWLI\LQJUHVROYLQJSRZHURIVHLVPLFV\VWHPV Geophysics. 1985. 50(8): 1091-1099 Geophysics. 1982. 47(8): 1160-1173 RHQEHUJHU6FK 05HVROXWLRQFRPSDULVRQRIPLQLPXPSKDVHDQG]HUR 56XDQG:RNV|]701'LIIUDFWLRQWRPRJUDSK\DQGPXOWLVRXUFH phase signals. Geophysics. 1974. 39(6): 826-833 holography applied to seismic imaging. Geophysics. 1987. 52(1): 11- HO06FKOHLFKHU-3XOVHDQGGLVWRUWLRQ+XEUDOLQ3GHSWK\JPLJUDWLRQ7 25 Geophysics. 1994. 59(10): 1561-1569 (Edited by Hao Jie)
Petroleum Science – Springer Journals
Published: May 18, 2013
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.