attribute expression of the mississippi lime
TRANSCRIPT
1Marfurt
Attribute Expression of the Mississippi Lime
Kurt J. Marfurt (The University of Oklahoma)
Attribute expression of the Mississippi Lime
• Lithology
• Fractures
o Strain
o Thickness
o Lithology
Diagenesis
Impedance
Curvature
Spectral components
Impedance
Textures
2Marfurt
Outline
• Correlation of fractures to curvature
o Clay models
o Outcrop analogues
o Hunton Limestone – seismic with image logs
A proposed workflow using post-stack data
o Kansas fractured carbonate
• Attribute images of Osage County surveys
• Recommendations for future seismic acquisition
Clay model experiments
(Staples, 2011)
kpos from
scanned
surface
photo
3Marfurt
(Staples, 2011)
photo
Clay model experiments
Clay model experiments
(Staples, 2011)
kpos from
scanned
surface
4Marfurt
Clay model experiments
(Staples, 2011)
Outline
• Correlation of fractures to curvature
o Clay models
o Outcrop analogues
o Hunton Limestone – seismic with image logs
A proposed workflow using post-stack data
o Kansas fractured carbonate
• Attribute images of Osage County surveys
• Recommendations for future seismic acquisition
5Marfurt
Deformation of mixed competency rocks (Glen Rose fm)
(Ferrill and Morris, 2008)
positivenegative
negative
Deformation of less competent rocks (e.g. Eagleford fm)
(Ferrill and Morris, 2008)
6Marfurt
Lidar correlation of fractures on outcrops to curvature (UK)
(Pearce et al., 2011)
Carboniferous Eelwell limestone
Lidar correlation of fractures on outcrops to curvature (UK)
(Pearce et al., 2011)
7Marfurt
Lidar correlation of fractures on outcrops to curvature (UK)
(Pearce et al., 2011)
“I endorse this correlation.”
(Saint Aiden)
Outline
• Correlation of fractures to curvature
o Clay models
o Outcrop analogues
o Hunton Limestone – seismic with image logs
A proposed workflow using post-stack data
o Kansas fractured carbonate
• Attribute images of Osage County surveys
• Recommendations for future seismic acquisition
9Marfurt
Most-positive principal curvature, k1,
co-rendered with image log fracture density
(Staples, 2011)
Strike of most-positive principal curvature, k1,
co-rendered with image log fracture density
(Staples, 2011)
10Marfurt
Azimuthal Intensity Correlations to Fracture Density
Well Number AzimuthAzimuthal Intensity vs. Fracture Density
Correlation (r)
Measured Length Along
Wellbore (ft)
1 --- No correlation ---
2 45 0.69 5800 - 6290
3 75 0.77 5400-6230
3 45 0.88 7275-7605
4 75 0.80 7120-7560
5 45 0.70 7890-8495
5 -75 0.57 6405-6955
6 -75 0.66 7065-7505
7 -75 0.69 5800-6185
Correlation of fractures with curvature
(Staples, 2011)
Correlation of Geometric Attributes to Production
The problem• Geometric attributes are an indirect measure of the paleo
rather than the present-day stress regime
• Since the time of fracture initiation
• the direction of Shmax has probably rotated
• fractures may have been diagenetically altered
(cemented, filled, or enlarged)
11Marfurt
Seismic Depth Map of Top Mississippian
• Discovered in 1962
• Cumulative production >1.7 million bbl oil
• High water cut (>94%)
1 kmC.I. = 3 m
A
A’
1
2
3
4
Dickman Field (Kansas, USA)
(Nissen et al., 2005)
0.5 mile
Volumetric Curvature – Gilmore City HorizonFrequency-Azimuth Rose Diagram
Length-Azimuth Rose Diagram
fault
(Nissen et al., 2005)
12Marfurt
Shale-filled fractures intersected by horizontal well (Kansas, USA)
Karst-controlled
10-100 ft interval
Provide a barrier to fluid flow
(Nissen et al., 2005) Miissippian cave filled with Pennsylvanian shale, SW Missouri
Water production versus distance to nearest NW and NE lineaments
0.5 mile
No relationship
Increased water productionnear lineaments
(Nissen et al., 2005)
13Marfurt
0
2
4
6
8
10
12
14
0 100 200 300 400 500 600 700 800
distance to NE lineament (ft)
5 y
ea
r o
il p
rod
uc
tio
n (
x10
4 B
bl)
0.5 mile
Oriented lineaments -- Kansas Mississippian
0
10
20
30
40
50
60
70
80
90
100
0 100 200 300 400 500 600 700 800
distance to NW lineament (ft)
5 y
ear
wate
r p
rod
ucti
on
(x10
4 B
bl)
0
2
4
6
8
10
12
14
0 100 200 300 400 500 600 700 800
distance to NW lineament (ft)
5 y
ea
r o
il p
rod
uc
tio
n (
x10
4 B
bl)
Lineament trend vs. production
0
10
20
30
40
50
60
70
80
90
100
0 100 200 300 400 500 600 700 800
distance to NE lineament (ft)
5 y
ea
r w
ate
r p
rod
uc
tio
n (
x10
4 B
bl)
(Nissen et al., 2005)
Distance to NW lineament (ft)Distance to NW lineament (ft)
Distance to NE lineament (ft) Distance to NE lineament (ft)
5-yr
wat
er p
rod
uct
ion
5-yr
wat
er p
rod
uct
ion
5-yr
oil
pro
du
ctio
n5-
yr o
il p
rod
uct
ion
Fault plane
(permeability, κ >>0)
Flow ~κ/r
Methodology
Hypothesize that an anomaly at a
given azimuth corresponds to
an open fracture
Assume flow goes as κ/r
Correlate flow to production to test
the hypothesis
(Guo et al., 2011)
14Marfurt
1.0
0.5
0.0 rmax0
Filt
er
Am
plit
ude
distance, r
Flow as a function of distance, r, from a
natural point “injector”
(Guo et al., 2011)
Azimuthally-
limited subset φ1
Azimuthally-
limited subset φj
… …
Fluid flow φ1 Fluid flow φNFluid flow φj
… …
Azimuthal
filtering
Azimuthally-
limited subset φN
User thresholds for scaling
Seismic data
Curvature computation
k1, Ψ1 k2, Ψ2
Dip scan
Inline dip Crossline dip
Convolve with 1/r Convolve with 1/r Convolve with 1/r
Workflow to calculate
hypothesized fluid flow
(Guo et al., 2011)
15Marfurt
Computing azimuthally-limited curvature volumes
-900 00 +900
1.0
0.0
0.5
Filt
er
Am
plit
ude
Strike
(Guo et al., 2011)
Time
(ms)
888
860
A
A’
B’
B
N
0.5 mi
B’B
A A’
Gilmore City
Top Mississippian
Time-structure map of Gilmore City horizon and two seismic lines(Guo et al., 2011)
16Marfurt
0 +0.002-0.002
k2
N
0.5 mi
Most negative structural curvature, k2 , through the Gilmore City(Guo et al., 2011)
we
igh
t
1
0
0 +0.002-0.002
k2
klower kupper
N
0.5 mi
“Skeletonized” image
obtained by interactively
adjusting the color bar
(Guo et al., 2011)
17Marfurt
0
N
0.5 mi
k2 modulated by ψ2
Azimuthψ N
E
S
W
(Guo et al., 2011)
R= -0.1467 R= - 0.0499 R= 0.0752 R= - 0.3720
R= - 0.2062 R= - 0.1115 R= 0.1180 R= 0.1119
R= 0.1193 R= 0.0633 R= 0.0294 R= -0.0350
1 mi
Intensitylow highfluid flow with oil production(Guo et al., 2011)
18Marfurt
R= -0.1619 R= - 0.2503 R= - 0.2190 R= - 0.1516
R= - 0.0465 R= - 0.1531 R= - 0.2217 R= - 0.0769
R= - 0.2040 R= 0.3406 R= 0.3350 R= 0.1891
1 mi
fluid flow with water productionIntensity
low high(Guo et al., 2011)
Cro
ss-c
orr
ela
tio
n c
oeff
icie
nt
Strike Strike
Oil production Water production
Unp
refe
rre
da
zim
uth
Pre
ferr
ed
azim
uth
(Guo et al., 2011)
19Marfurt
Outline
• Correlation of fractures to curvature
o Clay models
o Outcrop analogues
o Hunton Limestone – seismic with image logs
A proposed workflow using post-stack data
o Kansas fractured carbonate
• Attribute images of Osage County surveys
• Recommendations for future seismic acquisition
Attribute expression of the Mississippi Lime
• Lithology
• Fractures
o Strain
o Thickness
o Lithology
Diagenesis
Impedance
Curvature
Spectral components
Impedance
Textures
20Marfurt
2 mi
Original data Filtered data Rejected noise
Rejecting periodic components of seismic noise
A
600 ms
500 ms A’
A
A’
Original data Filtered data Rejected noise
Rejecting periodic components of seismic noise
0.0
0.4
0.8
1.2
Tim
e (
s)
A A’
2 mi
21Marfurt
Before data conditioning
Arbuckle horizon pick
0.600
0.580
0.620
Time (s)
After post stack data conditioning
Arbuckle horizon pick
0.600
0.580
0.620
Time (s)
22Marfurt
Top Mississipian SurfaceTime (s)
0.540
0.520
0.560
0.580
Top Mississipian Surface with k1 curvatureTime (s)
0.540
0.520
0.560
0.580
k1
Pos
Neg
0
Time slice at 0.6 s
24Marfurt
k1 curvature co-rendered with coherence
Time slice at 0.6 s
k1
Pos
Neg
0
k1 curvature vs. strike co-rendered with coherence
Time slice at 0.6 sStrikeψ
N
E
S
W
25Marfurt
k1 curvature vs. strike
Time slice at 0.6 sStrikeψ
N
E
S
W
k1 curvature vs. strike
Time slice at 0.6 sStrikeψ
N
E
S
W
26Marfurt
Attribute expression of the Mississippi Lime
• Lithology
• Fractures
o Strain
o Thickness
o Lithology
Diagenesis
Impedance
Curvature
Spectral components
Impedance
Textures
Texture analysis and self-organizing maps
applied to satellite imagery
(Matos et al., 2011)
Satellite imageSelf-organizing map
27Marfurt
Vertical slices through texture
attributes, Osage Co., OK
(Matos et al., 2011)256 Cluster color bar
Tim
e (
ms)
0
100
poor
?
good
Texture analysis and self-organizing maps
applied to Mississippi Lime, Osage Co, OK
(Matos et al., 2011)
28Marfurt
Outline
• Correlation of fractures to curvature
o Clay models
o Outcrop analogues
o Hunton Limestone – seismic with image logs
A proposed workflow using post-stack data
o Kansas fractured carbonate
• Attribute images of Osage County surveys
• Recommendations for future seismic acquisition
Recommendations for future seismic acquisition
Acquire high density, but less sweeps (more traces/mi2)
• reduce coherent noise (e.g. ground roll, shallow diffractors, …)
• reduce acquisition footprint that overprints fractures
Acquire long offsets
• facilitate λρ-μρ inversion to map lithology (tripolite, dolomite, tight chert,…)
•
Acquire wide azimuth to facilitate AVAz and VVAz analysis
• map natural fractures
• map maximum/minimum horizontal stresses
Calibrate seismic predictions with image logs and microseismic data
These are fairly routine parameters in other resource plays (i.e. shale gas)