GEOCIENCIAS-VOL 13-2 2009.vp EARTH SCIENCES RESEARCH JOURNAL Earth Sci. Res. J. Vol. 13, No. 2 (December 2009): 134-139 CRS SEISMIC PROCESSING OF A GEOLOGICAL COMPLEX AREA Eduardo Jiménez1, Carlos A. Vargas2, 3 and Luis A. Montes3 1 Geoseimic Ltda. e-mail: Eduardo.jimenez@cable.net.co 2 Institute for Geophysics, University of Texas 3 Departamento de Geociencias, Universidad Nacional de Colombia, Bogotá. e-mail: lamontesv@unal.edu.co ABSTRACT We applied the NMO and CRS (Common Reflector Surface) approaches to a complex geological area in order to compare their performances for obtaining enhanced images. Unlike NMO, CRS does not depend on a previous time velocity model and uses a hyperbolic equation to estimate 2D travel times through three parameters (Normal ray emergence angle, NIP and N wavefront curvatures). To obtain the image a solution provided by coherence analysis algorithm was used. A low quality Colombian seismic line acquired in Middle Magdalena basin was used, where a foothill geological area is char- acterized by a thrusting fault. The CRS provided an enhanced image which allowed a new geological interpretation that is best constrained with other regional observations. Key words: Common Reflection Surface (CRS) stack, Zero-offset (ZO), Normal Moveout Correction (NMO), Common Mid Point (CMP) RESUMEN El propósito de esta investigación es comparar las técnicas de Superficie Común de Reflexión (CRS) y Normal Move Out (NMO) en su desempeño en zonas geológicamente complejas, procurando imágenes sísmicas de mejor calidad. A diferencia del NMO, el CRS no depende de un modelo previo de velocidad y usa una ecuación hiperbólica de tiempos de viaje 2D dependiente de tres parámetros (ángulo de emergencia del rayo normal, curvatura de la onda de punto de incidencia normal y curvatura de la onda normal). Para obtener la imagen se usó una solución provista por un algoritmo de análisis de coherencia. Se usó una línea sísmica de baja calidad adquirida en la cuenca del Valle Medio del Magdalena - Colombia, en una zona de pie de monte caracterizada por una falla de cabalgamiento. El CRS suministró una imagen mejorada que permitió una nueva interpretación geológica que se ajusta mejora con otras observaciones regionales. Palabras clave: superficie común de reflexión CRS, Zero-offset (ZO), corrección por sobre tiempo normal (NMO), punto medio (CMP) 134 Manuscript received: 10/06/2009 Accepted for publication: 26/10/2009 ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:24 p p p Composite 133 lpi at 45 degrees Introduction The CRS method has been considered an attractive stacking method to provide improved zero offset sections (Trappe et al., 2001; Hertweck et al., 2007). In Colombian a known example with the CRS applied to complex areas does not exist, although there is a reported pa- per with CRS applied to a mild topography and quiet tectoni- cally zone with relevant results (Cárdenas and Montes, 2006). In complex tectonically areas seismic tests are ham- pered by several factors that often lead to low quality seismic images with ravel reflectors, due to poor signal/noise ratio or weak seismic signal focusing. In order to compare NMO and CRS performances in complex zones, a low quality seismic line of the Middle Magdalena basin was used. The area includes a thrust fault in the foothill zone. Instead of stack velocities CRS used pa- rameters picked automatically by a coherence analysis algo- rithm (Birgin et al., 1999). As result, an enhanced CRS image allowed a new geo- logical interpretation. CRS method The CRS stack is a theoretically well established method (Jager et al., 2001; Mann et al., 1999; Tygel et al., 1997). It considers layers separated by curved reflectors whose re- flection comes from a reflecting segment. A second order approximation of transmitted and reflected travelime rays in seismic system was developed (Bortfeld, 1989). A multi- covering seismic data set is acquired over a set of homoge- neous and isotropic layers, with arbitrary velocities sepa- rated by smooth interfaces. The seismic system is defined by a zero offset ray, called Central ray, which incidences normally on the reflector. In a second order approximation around central point (x0) the travel time of the SRG ray is approximated by the hyperbolic function: t x h t x x t K x xN 2 0 0 0 2 0 2 0 2 2 ( , ) sin ( ) cos [ ( � � � � � � � � � � � � � � 0 2 2) ]� K hNIP (1) According Figure 1, x = (G + S)/2 is the common mid- point, h = (G – S)/2) is half offset, �0 is the near surface ve- locity and � and t0 are angle of emergence and travel time of the Central ray. KNIP is the wavefront curvature of a hypo- thetical wave with the source located at point NIP KN is the wavefront curvatures of an exploding reflector segment around point NIP. A more deep and complete theoretical development about CRS can be revised in others references (Bortfeld, 1989; Tygel et al., 1997) Procedure The NMO stacked section was obtained by a current seismic processing sequence in ProMAX®. Due to CRS does not re- quire a previous known velocity model the velocity analysis and stacking steps were replaced by searching and optimiz- ing local and global CRS parameters through coherence analysis solution until obtain an optimized stacked section. This equivalent step was done using MPT® (a Numerica´s software trade mark). After stacking, the NMO and CRS im- ages were filtered and enhanced applying the same ProMAX sequences. A flow diagram of both sequences is shown in Figure 2. Search of CRS parameters This non conventional procedure replaces the NMO analysis and stacking currently used in seismic processing se- quences. The whole procedure is explained in the following sequence of steps with a visual description in the flow chart of Figure 3, as was defined by Muller (Muller, 1999) The first step is to estimate VNMO from CMP gathers at point x=x0, this reduces equation 1 to: t h t h CMP NMO 2 0 2 2 2 4 ( )� � � (2) 135 CRS SEISMIC PROCESSING OF A GEOLOGICAL COMPLEX AREA Figure 1. The Central ray is perpendicular to the reflector Ó and emerges at x0 with an � angle. The NIP and N waves reach the sur- face at point x0. ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:25 p p p Composite 133 lpi at 45 degrees This step is an NMO velocity analysis and stacking which provides an initial simulated ZO section (zero-offset). In the second step � is searched. For that, the reflectors in anterior ZO section are considered locally flat, i.e. case of small apertures (KN =0), that simplifies equation 1: t x h t x x( , ) sin ( )� � � 0 2 0 0 0 � � (3) In the third step � is known and with the equation in zero offset configuration, KN value is estimated from ZO section. Finally the KNIP is calculated using the � parameter and the expression K t NIP NMO � 2 0 0 2 � � �cos (4) Geological setting Geologically the area has been deformed by thrusting asso- ciated to compressive tectonic events. In the area a monocline dips to the East and the basement outcrops at West. The cretaceous rocks are below a gently dip Eastward Tertiary sequence, as seen in Figure 4. Different structural styles are observed: a first thrust with East vergence, a sec- ond one with West vergence and a lineation and NE-SW strike slip fault system. A low to high dip thrust fault emerges in surface with a high inclination. This fault causes the repetition of the creta- ceous sequence which was observed in the well sited at West. A second fault inferred from seismic section, allows the ele- vation of basement over the footwall block below the fault. A tertiary detachment fault interpreted in the seismic section was observed in the cartographic recognition of the area. 136 G.C. ONYEDIM, K.D. ALAGOA, I.O. ADEDOKUN, A.A. ADEROGBA AND C. OVURU Figure 2. The CRS sequence differs from NMO sequence in velocity analysis and stacking steps where instead, searching and optimiza- tion of CRS parameters are used. Figure 3. Flow charts with step sequence to obtain final CRS stack section. ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:26 p p p Composite 133 lpi at 45 degrees Discussion and results An E-W acquired seismic line was selected to be processed by the two methods in order to compare their performances in producing the seismic images. At East side of the thrust fault a well allowed to identify the stratigraphic units in the seismic section. This seismic line was previously processed using a con- ventional flow of processing to obtain a PSTM section, shown in Figure 5. In that figure is observed two different behaviors at left and right of the thrust fault. A set of clear and continuous re- flector with gentle dips can be seen from 2200 ms to 400 ms at right of the fault, whereas al right is almost impossible to identify a reflector. The seismic line was properly processed applying a con- ventional sequence flow with a careful selection of parame- ters, as result a new stacked section was obtained, providing also a more reliable NMO velocity model to be used as start- ing model in CRS method. The stacked section is displayed in the figure 6, where is evident the enhancement of the image compared with that other in figure 5. After the CRS processing a better image is showed the stacked section included in Figure 7. In figure 6 the reflec- tors appear weaker, strong and less coherent can be seen with a better continuity, with low noise and besides that the image owns more seismic events in the shallow part. Compared with NMO stack section the CRS is a more en- hanced and allow to constraints the monoclonal structure descripted in the geological setting section. To compare the quality of the images a S/N content evaluation was made. First the noise was isolated from the signal in the image through band pass filters and then re- spective spectral analysis´s were done. The results are de- picted in Figure 9, at top for NMO image and at bottom CRS image. Comparing the figures 9a and 9b the random noise level is reduced from 27.8% in NMO image to 17.9 % in CRS image. It represents a significant improvement in sig- nal to noise ratio and in consequence a rising in the seismic resolution. A post-stack CRS migrated section was interpreted us- ing the software Geographix (Landmark Graphics Co®). The change in the quality of the image permitted a new geo- logical interpretation matching the seismic section with the 137 CRS SEISMIC PROCESSING OF A GEOLOGICAL COMPLEX AREA Figure 4. Geological section of the zone with several structural styles observed and a well at West, from an interpreted stacked section (internal report). Figure 5. In the previous PSTM section CDPs 2100-2460, clear re- flectors are observed from CDP 2380 to the right but blurred at left. Figure 6. In the new NMO Stack section CDPs 2100-2460 the re- flectors are observed along the whole section, including at left of CDP 2380. ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:28 p p p Composite 133 lpi at 45 degrees lithological units identified in the stratigraphic column and the borehole logs (Figure 8). A schematic geological section was extracted from the interpretation and depicted in the Figure 10. The difference among the geological interpretation got- ten in Figure 3 and the interpreted in Figure 10 is evident. This new interpretation is due to the enhanced quality of the new obtained CRS image. Conclusions A low quality Colombian seismic line acquired in Middle Magdalena basin was processed using the NMO and the CRS techniques looking for an enhanced of the image on a thrust- ing fault. The CRS provided an improved image which al- lowed a new geological interpretation that is best constrained with other regional observations. Our observa- tions suggest that the CRS technique improves the imaging of complex areas and constitutes an alternative in seismic 138 G.C. ONYEDIM, K.D. ALAGOA, I.O. ADEDOKUN, A.A. ADEROGBA AND C. OVURU Figure 7. The CRS stack section CDPs 2100-2460 displays an en- hanced image with better continued and more clear reflectors along the section. Figure 8. NMO at top and CRS at bottom images with respective spectral analysis show reduction in the level of random noise. ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:28 p p p Composite 133 lpi at 45 degrees processing. CRS-2D method generated seismic sections of good quality without knowledge of subsurface velocity model, essential in expensive process like pre-stack migra- tions. Performance of CRS methods in thrusting zone, allowed a new interpretation and geologic model. Futures works should emphasize the CRS technology in 3-D data and use the information given by CRS, attributes in tomography in- version, AVO, migration and coherent cube analysis. Acknowledgements The authors express grateful to Landmark Graphics Co. Latin America to provide ProMAX and Geographix software and Numerica Ltda to permit the use of MPT (Multiparmetric Pro- cessing Tool) software implemented with CRS. Bibliography Birgin, E., Biloti, R., Tygel, M. and Santos, L. Restricted op- timization: a clue to a fast and accurate implementation of the Common Reflection Surface Stack Method. Journal of Applied Geophysics. Vol, 42, Issues 3-4. P 143-155. Bortfeld, R., 1989, Geometrical ray theory: Rays and travel times in seismic systems (second-order approximations of the travel times: Geophysics, Vol 54 No. 3 P 342-349. Cardenas, A., and Montes, L. 2006, Stacking on common re- flection surface with multiparameter traveltime. Vol. 10, No. 2, Earth Sci. Res. J. P 91-103. Hubral, P., 1983, Computing true amplitude reflections in a laterally inhomogeneous earth: Geophysics, 48 No. 8, P 1051 – 1062. Hertweck, T., Schleicher, J., Mann, J. 2007. Data Stacking Beyond CMP: The Leading Edge. V 26, No.7. P818-827. Jager, R., Mann, J., Hocht, G., Hubral P. 2001. Common re- flection Surface Stack: Image and attributes. Geophys- ics Vol 66 No. 1, P 97 – 109. Mann, J., Jager, R., Muller, T., Hocht, G., and Hubral, P., 1999, Common-reflection- surface-stack a real data ex- ample: Journal of Applied Geophysics, 42, 301 – 318. Neidel, N., and Taner, M., 1971, Semblance and other co- herency measures for multichannel data: Geophysics, 36, 482–497. Muller, T., 1999, The common Reflection Surface Stack Surface stack Method – Seismic imaging without ex- plicit knowledge of the velocity model: Ph. D. thesis, University of Karlsruhe. Tygel, M., Muller, T., Hubral, P., and Schleicher, J., 1997, Eigenwave base multiparameter traveltime expansion: 67th Annual Internat. Mtg. Soc. Expl, Geophys 1770 – 1773. Vieth, K., 2001, Kinematic wavefield attributes in seismic imaging. Ph.D. thesis, University at Karlsruhe. Trappe, H., Gierse, G. and Pruessmann, J., 2001. Case studies show potential of Common Surface stack structural resolu- tion in time domain beyond the conventional NMO/DMO stack. First Break, 625- 633. Special topic. Nov. 139 CRS SEISMIC PROCESSING OF A GEOLOGICAL COMPLEX AREA Figure 9. A new seismic interpretation resulted from the CRS im- age. WELL 1 Fault A Fault B WELL 1 Fault A Fault B Figure 10. A new schematic geological section provided by the in- terpretation of the CRS image. ENERO 30-GEOCIENCIAS-VOL 13-2 2009.prn D:\GEOCIENCIAS 13-2 DIC 2009\GEOCIENCIAS-VOL 13-2 2009.vp sÆbado, 30 de enero de 2010 18:16:29 p p p Composite 133 lpi at 45 degrees