Reconditioning Seismic Data to Improve Carbonate Stratigraphic Interpretation and Reservoir Characterization, Fullerton Field, West Texas

1
Reconditioning Seismic Data to Improve Carbonate
Stratigraphic Interpretationand Reservoir
Characterization,Fullerton Field, West Texas
by Hongliu Zeng Bureau of Economic Geology John
A. and Katherine G. Jackson School of
Geosciences The University of Texas at Austin
2
Carbonate Earth ModeHow Much Can Seismic Data
Help?
Seismic Data
Earth Model
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
3
Presentation Outline

• Value of seismic data as seen in petrophysical
  analysis
• Why seismic data needs reconditioning before
  interpretation?
• Basic seismic data conditioning (phase adjustment
  and high-frequency enhancing)
• Advanced seismic data conditioning (progressive,
  model-based inversion)

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
4
Fullerton Seismic Data (2-D 3-D)
Labrador
Phase 1
Phase 2
Southful
2 km
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
5
Synthetic Seismogram (Cal/Mon 1)
Tubb
Abo
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
6
Log Velocity vs. Log Porosity(FCU2723)
Velocity (ft/s) 15000 25000
Porosity 0 10 20
Tubb
Depth (ft)
Zone1
Velocity (ft/s)
Zone2
Wichita
Abo
Porosity
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
7
Original Seismic Poorly Tied toLog
Lithology/Porosity
f
f
f
DT
Tubb
Up. St.
Zone 1
Zone 2
Wichita
Abo
1 km
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
8
Basic Data Conditioning

• Adjust seismic phase to 90º - Standard
  zero-phase wavelet is not ideal for lithologic
  interpretation of seismically thin beds -
  900-phase data resemble the impedance section,
  which conditions amplitude to lithology

• High-frequency enhancing - Original data are
  dominated by low frequencies - Boosting
  high-frequency components improves seismic
  resolution (even in a price of reducing S/N
  ratio)

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
9
Interpretive Advantage of90º-Phase
Wavelet(Fullerton Carbonates)
200
300
400
500
600
700
800
900
100
00
Tubb
Zone 1
Zone 2
Wichita
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
10
High-frequency Enhancing(Spectral Balancing)
0 60
120 Hz
Tubb
Zone 1
Zone 2
Wichita
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
11
Data Conditioning Improves Resolution and
Well-Seismic Tie
Tubb
Up. St.
Zone 1
Original data
Zone 2
Wichita
Abo
Tubb
Reconditioned Data after phase Shifting and
high- Frequency enhancing
Up. St.
Zone 1
Zone 2
Wichita
Abo
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
12
Amplitude as Indicator of PHIH (L2100/Wichita)
PHIH (well)
Amplitude
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
13
Amplitude as Indicator of PHIH(L2200/Zone 2)
PHIH (well)
Amplitude
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
14
Advanced Data Conditioning

• Seismic data are limited in frequency (lt100 Hz)
  and Resolution (10 ms or 100 ft)
• The only source of high-frequency information
  beyond seismic is well logs
• Use of well logs as soft constrain may
  significantly improve resolution (2 ms or 20 ft)
• Tool model-based inversion (Hampson-Russell,
  Jason, CGG, etc.)

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
15
Model-Based Inversion
Initial Model
Impedance Logs
Seismic
Inverted Impedance
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
16
Challenges in Building Initial Model

• Impedance values between wells are interpreted
• Correlation of well logs are characterized by
  uncertainties of various degrees
• Correlation surfaces may not be accurate beyond
  wells
• First initial model is always not optimal

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
17
Well Correlation vs. Real Seismic
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
18
Progressive Inversion

• Build multiple initial models and do multiple
  inversions
• Start from the most prominent and reliable
  geologic boundaries and seismic horizons, adding
  additional horizons from intermediate inversion
  results
• Use more and more horizons to build better
  initial model, resulting in better inversion with
  higher resolution.

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
19
Start from Best Horizons
Abo
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
20
Progressive Inversion(First Round)
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
21
Progressive Inversion(Second Round)
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
22
Progressive Inversion(Third Round)
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
23
Improvement in Resolution
Seismic
3rd-Round Inversion
100 ft
H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology
24
Summary

• Original seismic data are often not optimal for
  geologic interpretation and need reconditioning
• Basic data conditioning (1) corrects seismic
  phase and (2) enhance high-frequency components
  in the data, improving resolution to 100 ft
• Advanced data conditioning combines geologic
  interpretation and seismic data in a progressive,
  model-based inversion procedure, further
  improving resolution to 20 ft

H. Zeng UL/PTTC May 2003 Workshop, Bureau of
Economic Geology