2. Ødegaard
FEASABLE TECHNOLOGY PUT INTO PLAY
AT THE RIGHT TIME
Well log
analysis
Rock physics
diagnostics
Seismic
modelling
Field studies Exploitation
Feasibility
studies
Lithology &
fluid prediction
4D timelapse
Reservoir
characterisation
Reservoir
simulation
SEISMIC
INVERSION
Acoustic
impedance
Poisson’s
Ratio
Density
ATTRIBUTE
CALCULATION
NEURAL
NETWORK
3 WEEKS 6 WEEKS 1 YEAR
5. Ødegaard
4D Loop Seismic-2-Simulation-2-Seismic
3D Seismic
Geological
Model
Reservoir
Model
Seismic
Acquisition
Time
4D Body
Identification
New Data
Difference
Cube
Baseline
Time-lapse
Depth
Geological
Model
Geological
Model
Geological
Model
Geological
Model
Reservoir
Model
Reservoir
Model
Reservoir
Model
Reservoir
Model
Predictions
Wells
Wells
6. Ødegaard
Understand acoustic vs physical properties
in inverted seismic data.
Apply rock physics in 4D modelling
Reservoir engineering aspects
Multiple wells with sonic and shear logs
3 vintages of 3D seismic data (near and far offset)
AVO inversion and lithology prediction
NELSON
4D AVO seismic
8. Ødegaard
Comparison of the conventional far-offset difference data
to the inverted far offset difference data. In the lower section bright
colours indicate a positive impedance change. The difference signal
is restricted to the lower Z3 reservoir interval.
Top Forties
Top Z3
Top Z2
11. Ødegaard
Crossplot of Poisson’s ratio versus acoustic impedance for well log data
from the Nelson reservoir interval. Oil and water filled sands as well as
shale fields can be clearly defined.
13. Ødegaard
Oil sand Brine sand
Oil sand Brine sand
1990
1997
4D Rock physics Diagnostics
Defining lithology and fluid fields from inversion data to carry
out a probabilistic prediction of fluid and lithology volumes.
14. Ødegaard
A perspective view to the northeast of the Oil sand volume
prediction from the 1990 baseline acoustic impedance and Poisson’s
ratio data.
15. Ødegaard
Isometric perspective to view to the north of a volume detection of
high oil sand probability (bright colours) masked by a detection of
positive impedance change (2000-1990) in grey. Bright colours
indicate potential unswept oil at the top of the reservoir at mid-2000.
16. Ødegaard
Oil Sand Probability Volume
Far Offset Impedance Difference
Positive impedance change (red) indicating sweep (2000)
SW NE
Oil Sands in Red (1990)
Cone around production well
Edge Drive
Basal Rise
4D Target
500m
Oil sand probability and far-offset impedance difference sections
through the N30 target, between two nearby production wells. It can
be seen from the sweep pattern on the far-offset impedance section
that oil is not being swept effectively from the Z4 section between the
two producers, in an area which the oil sand probability shows to be
good reservoir.
18. Ødegaard
Far offset impedance difference section intersecting the N30y production well.
Bright colours show a high positive impedance contrast (sweep).
The pilot hole for the well encountered an 25 meters oil column before penetrating the moved
OWC as prognosed by 4D.
Simulation had indicated that the area would be almost completely swept (orange horizon).
OWC from 4D
Seismic (FO Inv Diff) OWC from SimulatorTop Forties
19. Ødegaard
ABSTRACT:
Integrated analysis of 4D seismic data and petrophysical data is used
to produce probabilistic fluid and lithology volumes for monitoring
reservoir performance on the Nelson Field.
Petrophysical analysis of log data shows distinct fields for oil sand, water
sand, shale and heterolithic ‘lithologies’ in acoustic impedance - Poisson’s ratio
space.
Elastic inversion techniques applied to conventional 4D AVO datasets convert
the reflectivity data to acoustic impedance, shear impedance, Poisson’s ratio
and angle impedances. The elastic inversion datasets are used to quantify oil
water contact movements through volume sculpting techniques.
Well derived relationships are used to predict 3D volumes of oil sand probability
from three different seismic survey vintages, 1990, 1997 and 2000. Changes in
oil sand probability due to production are verified by comparison with repeat
production logs.
Integrated volume interpretation of 4D far offset inversion difference
(oil water contact movement) and oil sand probability show areas of
unswept oil, highlighting infill opportunities. Early results from infill
drilling have validated the method realising the potential economic
benefits of 4D seismic technologies.
20. Ødegaard
TERN
4D full stack seismic
Reservoir Engineering aspects
Physical properties linked to acoustic properties
21. Ødegaard
Gamma
Sonic
Quartz matrix
0 % 100
Bulk volume water
100 % 0
Porosity
100 % 0
Clay volume
0 % 100
Neutron
Density
Acoustic Impedance
Depth(m)
WELL LOG FLUID SUBSTITUTION: 1980-1995
25. Ødegaard
From change in Sw From change in pressure
From combined effect of
change in Sw and pressure
ROCK PROPERTIES FROM SIMULATOR: 1980 - 1995
Changes in acoustic impedance
44. Ødegaard
Elastic properties of Brent group sand (6 wells)
AI-PR-SWAI-PR-PHI
PHIT evaluation
AI-PR (for sandflag data only)
Active Zone : 4:34/10-B-8 Z:2 Top Tarbert
3.
4.
5.
6.
7.
8.
9.
AI10^6(kg/m2s)
0.1 0.2 0.3 0.4 0.5
PR
0.2
0.4
PHIT
8730 points plotted out of 12340
Well Zone Depths
34/10-B-8 (2) Top Tarbert 2616.M - 2768.M
34/10-B-8 (3) Top Ness 2768.M - 2919.M
34/10-B-8 (4) Top NER 2919.M - 3094.M
34/10-C-33 (2) Top Ness 2095.M - 2116.M
34/10-C-33 (3) Top NER 2116.M - 2231.5M
34/10-B-15 T2 (1) Top Tarbert 2476.M - 2579.M
PHIT evaluation
AI-PR (for sandflag data only)
Active Zones : W:4 Z:2, 3, 4 W:6 Z:3, 2 W:1 Z:2, 3 W:2 Z:2, 3, 1 W:3
3.
4.
5.
6.
7.
8.
9.
AI10^6(kg/m2s)
0.1 0.2 0.3 0.4 0.5
PR
0.
1.
SW
8230 points plotted out of 11663
Well Zone Depths
34/10-B-8 (2) Top Tarbert 2616.M - 2768.M
34/10-B-8 (3) Top Ness 2768.M - 2919.M
34/10-B-8 (4) Top NER 2919.M - 3094.M
34/10-C-33 (2) Top Ness 2095.M - 2116.M
34/10-C-33 (3) Top NER 2116.M - 2231.5M
34/10-B-15 T2 (2) Top Ness 2579.M - 2670.M
•PP and Sw of four generations are extracted from the reservoir model.
•The RP model transform changes in SW and PP into changes in elastic rock properties.
49. Ødegaard
1. Generate suite of attributes
from the seismic
2. Extract attributes at well
locations
3. Investigate methods of
relating well log rock
properties to volume
derived data.
4. Apply derived relationships
to input volumes
5. Interpret the resultant
volumes
WORKFLOW
50. Ødegaard
Absolute acoustic impedance (AI):
• Real rock physics property
• Contains low frequency
information not present in
seismic
• Good ties with well log derived
acoustic impedance
• Hydrocarbon identification not
possible on AI alone.
ACOUSTIC IMPEDANCE
52. Ødegaard
• All the well data loaded into a GEOVIEW database.
• All the original and generated attribute volumes loaded.
• A target log specified – water saturation
• Attributes extracted along the well path
WELL LOG/ATTRIBUTE RELATIONSHIP
Sw synth Vp f fm Ia If phase dIa/dt filtseis IntIa
53. Ødegaard
• Neural networks used to generate non-linear transform between
attributes and target logs
• Transform applied to the inputs at the well locations produces very
good results
WELL LOG PREDICTIONS
WATER SATURATION
Well Prediction
56. Ødegaard
We would like to thank the following contributors to the
presentation:
STATOIL
NORSK HYDRO
CONCOPHILLIPS
AMERADA HESS
SHELL
57. Ødegaard
• Experience with this type of datasets
25 major 4D projects ‘under the belt’
Variety of PE Objectives Realised
Number of Different Geological Settings
Large group, Varied Disciplines
• Proprietary Technology
Simultaneous Inversion
4D ISIS
• Fast Project Start-up and Turnaround
• Cost Effective, High Quality
WHY USE INVERSION DATA 4D CAPABILITIES
58. Ødegaard
RA Geophysical Well Log Analysis
IP Rock Properties Modeling
OSIRIS Precise Seismic Modeling
EMERGE Neural Net
4D*ISIS Global Seismic Inversion
MAAT Seismic Attributes