Potential risks of spectrum whitening deconvolution — Compared with well-driven deconvolution

Petroleum Science, Jun 2009

Deconvolution is widely used to increase the resolution of seismic data. To compare the resolution ability of conventional spectrum whitening deconvolution to thin layers with that of well-driven deconvolution, a complex sedimentary geological model was designed, and then the simulated seismic data were processed respectively by each of the two methods. The amplitude spectrum of seismic data was almost white after spectrum whitening, but the wavelet resolution was low. The amplitude spectrum after well-driven deconvolution deviated from white spectrum, but the wavelet resolution was high. Further analysis showed that if an actual reflectivity series could not well satisfy the hypothesis of white spectrum, spectrum whitening deconvolution had a potential risk of wavelet distortion, which might lead to a pitfall in high resolution seismic data interpretation. On the other hand, the wavelet after well-driven deconvolution had higher resolution both in the time and frequency domains. It is favorable for high resolution seismic interpretation and reservoir prediction.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://link.springer.com/content/pdf/10.1007%2Fs12182-009-0023-y.pdf

Potential risks of spectrum whitening deconvolution — Compared with well-driven deconvolution

Pet.Sci. Potential risks of spectrum whitening deconvolution ---- compared with well-driven deconvolution Li Guofa 0 1 Zhou Hui 0 1 Zhao Chao 0 1 0 Key Laboratory of CNPC Geophysical Exploration, China University of Petroleum , Beijing 102249 , China 1 State Key Laboratory of Petroleum Resource and Prospecting, China University of Petroleum , Beijing 102249 , China Deconvolution is widely used to increase the resolution of seismic data. To compare the resolution ability of conventional spectrum whitening deconvolution to thin layers with that of welldriven deconvolution, a complex sedimentary geological model was designed, and then the simulated seismic data were processed respectively by each of the two methods. The amplitude spectrum of seismic data was almost white after spectrum whitening, but the wavelet resolution was low. The amplitude spectrum after well-driven deconvolution deviated from white spectrum, but the wavelet resolution was high. Further analysis showed that if an actual reflectivity series could not well satisfy the hypothesis of white spectrum, spectrum whitening deconvolution had a potential risk of wavelet distortion, which might lead to a pitfall in high resolution seismic data interpretation. On the other hand, the wavelet after welldriven deconvolution had higher resolution both in the time and frequency domains. It is favorable for high resolution seismic interpretation and reservoir prediction. Well-driven; high resolution; spectrum whitening; deconvolution; seismic wavelet - 1 Introduction Spectrum whitening deconvolution is a widely-used method for increasing the resolution of seismic data (Mou et al, 2007; Jia et al, 2002) . It is supposed that the spectrum of a reflectivity series is approximately white. Its objective is to whiten the spectrum of seismic data in an effective frequency band (Li et al, 2001; Chen and Zhou, 2000) . Using the basic idea of well-constrained impedance inversion, the well-driven high resolution seismic data processing method successfully introduces well-constraint before the stage of reservoir inversion to the seismic data processing stage (Kaderali et al, 2007; Spikes et al, 2008) . By introducing the reflectivity series from well log data, the deconvolution problem is transformed from statistical to deterministic (Baan, 2008; Wang, 2006) . Well-driven deconvolution, which ensures more reliable resolution after deconvolution, evaluates and determines the deconvolution operator by wavelet. Although the application of well-driven deconvolution has not been as popular as spectrum whitening deconvolution, its advantages over spectrum whitening have gradually attracted more attention. For comparing the two methods, referring to the actual sedimentary structure in the Dagang Oilfield, Bohai Bay Basin, China, a continental sedimentary thin interbedded model was designed, and the simulated seismic data were respectively processed by spectrum whitening deconvolution and well-driven deconvolution. The results illustrated the advantages of well-driven deconvolution, and meanwhile r e v e a l e d t h e p o t e n t i a l r i s k s o f s p e c t r u m w h i t e n i n g deconvolution. 2 Spectrum whitening deconvolution and its problems The basic idea of spectrum whitening deconvolution is to whiten the amplitude spectrum of seismic data in a specific frequency band. Assuming that the spectrum of a reflectivity series is approximately white, adjusting the amplitude spectrum of seismic record is equivalent to adjusting that of the wavelet (Li et al, 2008) . Spectrum modeling and spectrum modulation techniques are often used to improve the resolution of seismic records in order to control the amplification of noise after spectrum whitening (Rosa and Ulrych, 1991; Zhao et al, 1996) . The amplitude spectrum of the wavelet is simulated by smoothing that of the seismic record, and then a spectrum whitening operator is estimated in the frequency domain. In fact, these methods can be treated as spectrum whitening deconvolution in the frequency domain, and they also need the same basic hypotheses as the time domain methods (Baan and Pham, 2008; Velis, 2008) . The two amplitude spectra shown in Fig. 1 are quite different and it is difficult to simulate one by the other. In fact, Fig. 1(a) is the amplitude spectrum of a 50 Hz Ricker wavelet, and Fig. 1(b) is the amplitude spectrum of a seismic record which is generated by convoluting the Ricker wavelet with four 20 ms spaced reflection coefficients with equal 20 40 60 80 100 20 40 60 80 100 Frequency, Hz (a) Frequency, Hz (b) Another problem of spectrum whitening deconvolution is the neglect of phase effects on wavelet resolution. The wavelet resolution can be most reliably evaluated on the basis of its waveform. If it is evaluated in the frequency domain, amplitude and phase spectra of wavelet should be taken into account together. Now the spectrum whitening deconvolution is applied to a 30 Hz Ricker wavelet, and its phase spectrum is changed. Fig. 2 is the waveform and its amplitude spectrum after the deconvolution. It is seen that the amplitude spectrum is whitened, but the wavelet has more side lobes since the wavelet is no longer zero-phase. Spectrum whitening does not substantially increase the resolution of seismic wavelet. 80 e 60 d u lit p Am 40 20 0 0.8 0.4 e d u litp 0 m A - 0.4 - 0.8 80 de 60 u lit p Am40 20 0 90 de 60 u lit p m A 30 amplitude. It is seen that if the reflectivity series is not a white spectrum, it is difficult to simulate the amplitude spectrum of wavelet from that of seismic record. Analysis of actual well log data indicates that the spectrum of a reflectivity series is close to the blue spectrum instead of the expected white one (Walden and Hosken, 1985) . In object-oriented seismic data processing, the time window for spectrum whitening is generally short around the target reflection. As a result, the spectrum of a reflectivity series in the window is far from white spectrum due to inadequacy of the statistics. 3 Basic principle of well-driven deconvolution The reason why spectrum whitening deconvolution needs the hypothesis of a white spectrum is that, when the reflectivity series is unknown, the wavelet can not be accurately estimated from the seismic record. However, if the well log data near seismic traces contain velocity and density information, reflectivity coefficients can be calculated from the well log data. Therefore, the wavelet can be estimated directly from the seismic record without the basic hypothesis above mentioned. It is convenient to improve the resolution with the known wavelet (Li et al, 2005) . Spectrum whitening deconvolution is first used to increase the resolution of Fig. 4. The result is shown in Fig. 5. Compared with Fig. 4, the resolution is improved to some extent. The amplitude spectrum after the deconvolution is shown in Fig. 6. It is seen that the energy of each frequency component below 85 Hz is almost uniform. Spectrum whitening result is satisfactory if it is evaluated by amplitude spectrum. CDP 806 926 1046 1166 CDP 806 926 1046 1166 The resolution can only be approximately evaluated by using the amplitude spectrum of seismic data. Strictly speaking, it should be evaluated by the waveform and spectrum of the wavelet. Therefore, the seismic wavelet after spectrum whitening deconvolution is extracted by using the well logs shown in Fig. 5. The extracted wavelet and its spectrum are shown in Figs. 7(a) and 7(b). It is seen from Fig. 7(a) that wavelet shape is not satisfactory since the main lobe is wide and the side lobes are strong. From Fig. 7(b) it is noticed that the amplitude under 32 Hz is strong, the amplitude above 32 Hz is weak, and the amplitude around 40 Hz is only half of the peak. As a result, the resolution after spectrum whitening deconvolution is not satisfactory if evaluated by wavelet. Baan M V. Time-varying wavelet estimation and deconvolution by kurtosis maximization . Geophysics . 2008 . 73 ( 2 ): 11 - 18 Baan M V and Pham D T. Robust wavelet estimation and blind deconvolution of noisy surface seismics . Geophysics . 2008 . 73 ( 5 ): 37 - 46 Chen C R and Zhou X X. Improving resolution of seismic data using wavelet spectrum whitening . Oil Geophysical Prospecting . 2000 . 35 ( 6 ): 703 - 709 (in Chinese) Jia L H , Wu C J , Luo Y X , et al. High resolution seismic data processing techniques . Geophysical Prospecting for Petroleum . 2002 . 41 ( 4 ): 484 - 488 (in Chinese) Kaderali A , Jones M and Howlett J. White Rose seismic with well data constraints: A case history . The Leading Edge . 2007 . 26 ( 6 ): 742 - 754 Li G F , Mou Y G and Wang P. A interactive technique for seismic wavelet extraction . Journal of China University of Petroleum (Edition of Natural Science) . 2005 . 29 ( 5 ): 33 - 36 (in Chinese) Li G F , Mou Y G , Liu Y Q , et al. Spectrum-modulated technique . Oil Geophysical Prospecting . 2001 . 36 ( 5 ): 597 - 601 (in Chinese) Li G F , Xiong J L , Zhou H , et al. Seismic reflection characteristics of fluvial sand and shale interbedded layers . Applied Geophysics . 2008 . 5 ( 3 ): 219 - 229 Mou Y G , Chen X H , Li G F , et al. Seismic Data Processing . Beijing: Petroleum Industry Press. 2007 . 70 - 75 (in Chinese) Rosa A L R and Ulrych T J . Processing via spectral modeling . Geophysics . 1991 . 56 ( 8 ): 1244 - 1251 S p i k es K , Dvorkin J and Schneider M. From seismic traces to reservoir properties: Physics-driven inversion . The Leading Edge . 2008 . 27 ( 4 ): 456 - 461 Velis D R. Stochastic sparse -spike deconvolution . Geophysics . 2008 . 73 ( 1 ): 1 - 9 Walden A T and Hosken J W J. An investigation of the spectral properties of primary reflection coefficients . Geophysical Prospecting . 1985 . 33 ( 3 ): 400 - 435 Wang Y H. Inverse Q -filter for seismic resolution enhancement . Geophysics . 2006 . 71 ( 3 ): 51 - 60 Zhao B , Yu S P , Nie X B , et al. Spectral-modeled deconvolution and its application . Oil Geophysical Prospecting . 1996 . 31 ( 1 ): 101 - 113 (in Chinese)


This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2Fs12182-009-0023-y.pdf

Guofa Li, Hui Zhou, Chao Zhao. Potential risks of spectrum whitening deconvolution — Compared with well-driven deconvolution, Petroleum Science, 2009, 146-152, DOI: 10.1007/s12182-009-0023-y