2. 1.5 -1.5 6.0-6.0 2100 Depth (m) Time (s) Raw Seismograms 0 2100 Four-Layer Sand Channel Model 0 0 0.8 Midpoint (m)

3. Interferometric Seismic Imaging Jianhua Yu & Gerard T. Schuster University of Utah

4. Outline Passive data for IVSP while Drilling Imaging of Free Surface CDP Multiples Interferometric Principle: Interferometric Principle: Time Diff. between Arrivals Structure Diff. Interferometric Imaging Applications: Interferometric Imaging Applications:

5. Claerbout, Katz, 70’s xx Earthquake 1900’s xx Utah+LLNL 1997 Migrate Claerbout, Rickett 1999 xyxyxyxy xyxyxyxyMigrate Arbitrary Unknown Src V(z) V(x,y,z) W(t) Location W(t) Location NOYESYESYES NO YESYESYES YES/NO YESNO? NOYESYES YESNO YES YES YES SELECTIVE HISTORY PASSIVE IMAGING Validity

6. Interference Pattern Optical Lens LASER t

7. Interference Pattern Optical Lens ss t LASER Lens Deformation Time Differences

8. Outline Passive data for IVSP while Drilling Interferometric Principle: Interferometric Principle: Time Diff. between Arrivals Structure Diff. Interferometric Imaging Applications: Interferometric Imaging Applications:

9. Seismic Ghost Reflection Direct Ghost ? Find R(x,z) but not know source location

10. Direct Ghost 12Directx Directx Master Seismic Ghost Reflection Seismic Interferogram: Correlate Traces t }M m(x) = (g, t + t ) gx g MxMxMxMxgx M Kirchhoff Migrate psuedo-shot gathers Kirchhoff Migrate psuedo-shot gathers Ghost Direct has kinematics of primary reflection x M

11. Source Geophone Bit Position? Problems in RVSPWD Pilot signal? Wavelet ?

12. Problems with Drill-bit and RVSP Data No source wavelet No source initiation time Not easy to get pilot signal in deviated well Not easy to get pilot signal in deviated well and horizontal well

13. Static shift errors may exist Difficulty for separating primary and ghost waves from deviated or horizontal well Problems with Drill-bit and RVSP Data

14. Solution Cross- or Auto-correlogram Migration Why do we use autocorrelation of seismic data rather than the seismogram ???

15. Strengths of Autocorrelogram Migration No need to know initial time No limits to deviated well Reduce Static errors influence No need to know source wavelet

16. Weakness of Correlogram Migration Virtual Multiples & Loss of Resolution

17. Well Drill bit Receiver Primary Direct Wave Ghost What is Joint Migration

18. Final migration image Ghost migration Primary migration Seismic Data

19. Horizontal Well Model 0 Depth (m) 3 40 X (m) V1 V2 V4 V3 V5 V6

20. Shot Gather 1 200 0 4 Time (s) CSG 10 1 200 0 4 Time (s) Traces Autocorrelogram

21. 1.6 2.1 0 2.5 Depth (km) X (km) 1.6 2.1 X (km) Standard migration Joint migration Standard Migration with Joint Migration (3 CSGs) Source

22. 1.6 2.1 0 2.0 Time (s) X (km) 1.6 2.1 X (km) With only primaryJoint auto. migration Time Migration Results (39 CSGs) Source

23. 0 04.5 -5 East (kft) North (kft) Well Rig 10 Depth (kft) 0 9188 ft Offset=1135-4740 ft Recording Length: 20 s Sample Interval: 2 ms Main Acquisition Parameters

24. Main Processing Steps Trace editing and static shift Trace editing and static shift Frequency panel analysis and noise elimination Frequency panel analysis and noise elimination Velocity analysis Velocity analysis Amplitude balance and energy normalization Autocorrelograms, vertical stacking utocorelograms Joint migrating autocorelograms

25. Acquisition Survey Map Well Rig 3C Receivers Drill bit 0 0 150030004500 -5000 East (ft) North (ft) C Line AC4

26. 3.0 2.0 1.0 Time (s) SP1255 1235 1215 Drill hole Joint Migration ( insert) and CDP Section

27. 3.0 2.0 1.0 Time (s) SP1255 1235 1215 Primary Migration ( insert) and CDP Section Drill hole

28. 3.0 2.0 1.0 Time (s) SP1255 1235 1215 Drilling hole Joint Migration ( insert) and CDP Section JOINT

29. 3.0 2.0 1.0 Time (s) SP1255 1235 1215 Primary Migration ( insert) and CDP Section Drilling hole Primary

30. SUMMARY Need separating primary and multiple? NO Work for deviated or horizontal well ? YES Reduce the influence of static errors ? YES Need pilot signal ? NO Need source wavelet and initial time ? NO Suppress coherent noise ? YES Amplitude fidelity ? NO Virtual Multiple ? YES

31. 6.0-6.0 Thanks to 1997-2002 UTAM sponsors and DOE. and DOE. Thanks to Lew Katz and Fred Followill.

32. Primary Ghost 12 Free-Surface Multiple

33. Primary Ghost 12 x PrimaryxPrimary Seismic Interferogram: Correlate Traces Caution: Ghostx Primary = R 3 t } m(x) = (g, t + t ) gx g MxMxMxMxgx M Kirchhoff Migrate psuedo-shot gathers Kirchhoff Migrate psuedo-shot gathers Ghost Primary has kinematics of primary ref. x

34. Nine-Layered Model (J. Sheng) 00.6 1.2 1.8 3.0 Depth (km) 2.4 Model Crosscorrelogram image Distance (km) 0 3.0 Distance (km) 0 3.0 artifacts Kirchhoff Image

35. Nine-Layered Model Kirchhoff Image Product Image =Kirch*correl.00.6 1.2 1.8 3.0 Depth (km) 2.4 Distance (km) 0 3.0 Distance (km) 0 3.0 artifacts

36. SEG/EAGE Salt Model 0 0.6 1.2 1.8 2.4 Depth (km) 3.0 3.6 0 5.0 10.0 15.0 Distance (km)

37. Crosscorrelogram Image Crosscorrelogram Image0 0.6 1.2 1.8 2.4 Depth (km) 3.0 3.6 0 5.0 10.0 15.0 Distance (km)

38. Kirchhoff Image 0 0.6 1.2 1.8 2.4 Depth (km) 3.0 3.6 0 5.0 10.0 15.0 Distance (km)

39. Product Image 0 0.6 1.2 1.8 2.4 Depth (km) 3.0 3.6 0 5.0 10.0 15.0 Distance (km)