of faults
Geomechanical analysis of sealing capacity and seismicity of faults Jan ter Heege
Glarus thrust fault, Swiss Alps
2/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Faults determine hydrocarbon accumulation S
N
TWT Buang-1 S
N
Corallina-1 BPLI
BPLI
TM1
high strain reactivated bounding faults
TM1
TM3 TM3
TE TE
From Gartrell et al., APPEA Journal 2005, modified for DFS by Clennell (CSIRO)
Juxtaposition seals, SGR > 60% KC 77 m oil column
no oil
KA
JO JO
non-leaking bounding fault 2000m 2000m
Example Timor Sea, Australia
KC KA
3/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Conventional techniques to determine the sealing capacity of faults and some of their limitations
4/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Shale gouge ratio can be used to determine sealing capacity of faults without considering stress state Analysis for entire fault section across slipped interval
Throw, T
Shale
SGR= (Vsh.z) / T x 100%
Sand
Vsh5, z5
Slipped interval (T) Photo: G Skerlec
Vsh4, z4 Vsh3, z3 Vsh2, z2 Vsh1, z1
Outcrop data (meter scale)
after Yielding et al. 1997
5/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
SGR extensively calibrated using reservoir well pressures, but poorly constrained (SGRcrit =10-20%)
Yielding et al. 1997
6/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of the sealing capacity of faults using integrated experimental-modelling studies
Brecht Wassing
Silvio Giger
Bogdan Orlic
Ben Clennell
7/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults ~4 m
Large rig to produce experimental clay smears in large samples (24x12x15 cm) at realistic stress conditions (n= 8 MPa) from Giger (CSIRO)
8/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Experiments simulated by discrete element models to determine SGR vs. shale properties and normal stress normal stress
normal stress
“fault zone”
displacement
synthetic shale: Vshale + mechanical properties synthetic & natural sandstone
matching mechanical properties
9/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults
10/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults 1 0.9
shale content (V shale )
SGR=20%
dip angle fault: 60º shale layer thickness: 10 m
transition zone
0.7 faults sealing
0.6
SGR=10%
shale1
n=20 MPa
0.8
n=10 MPa shale2
schematic! ?
0.5 0.4
faults leaking
0.3 0.2 0.1 0 0
20
from geological model
40
60
80
fault displacement (D)
100
120
11/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Conventional techniques to determine fault reactivation and induced leakage and seismicity
(sources: TNO, KNMI – Royal NL Meteorological Institute, Ministry of Econ. Affairs)
12/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of strength and sealing capacity of faults by integrating experimental data and models 0.4
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
m=0.6
m=1.0
Mohr diagram
t/Sv
0.3
0.2
high n low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s'v 0.5 s n/Sv
0.6
0.7
0.8
Shmin
SHmax
4
4
slip tendency: shear stress/normal stress (/n)
0.9
1
Sv
3
13/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Slip tendency can be mapped on fault planes to determine critically stressed (leaking) fault segments 0.4
m=0.6
m=1.0
t/Sv
0.3
0.2
high n
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s' 0.5v s n/Sv
0.6
0.7
Shmin
0.8
SHmax
0.9
1
Sv
14/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Full field finite element models can account for local geology in predicting fault reactivation Feasibility of CO2 storage in the depleted De Lier gas field (onshore NL)
Structural geological model
Input to FEM
- rock properties - constitutive material models - in situ stress
NE GWC
Pressure load from reservoir simulator Pressure load for DIANA Static pressure
SW
N
Extrapolated pressure
Pressure [bar]
2D FE Model
Simulated pressure Pressure in DIANA
Time [year]
3 km
Enlarged FE model
10 km
Location of monitoring points Fault
Stacked reservoirs
Fault
Caprock
143 145 148
Aquifers
Bogdan Orlic (TNO)
15/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Variation in local stress due to CO2 injection can result in reactivation of fault segments Possible stress paths for the Bounding Fault for injection (el.143)
Location of monitoring points Fault
143 145 148
Shear traction [MPa]
25
Reversible stress path (elastic response) Partially reversible stress path Irreversible stress path (rigid response) Stress before injection Stress after injection MC: c=0, ф=35°
20 15
Fault re-activation? 10 5 0 0
5 10 15 20 Normal effective traction [MPa]
25
Bogdan Orlic (TNO)
16/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of fault reactivation and induced seismicity using experimental + modelling studies
17/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better methods to determine strength (friction coefficient) of faults 0.4
m=0.6
m=1.0
?
t/Sv
0.3
0.2
high n
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s'v 0.5 s n/Sv
0.6
0.7
Shmin
0.8
SHmax
0.9
1
Sv
18/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability from geological model
log k f 4 SGR
Crawford et al. 2008
1 4
log( D )(1 SGR )
5
19/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability
Strong leaking fault segments
map in geological model to determine spill points
Weak sealing fault segments Crawford et al. 2008
20/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability
Gerco Hoedeman (TNO)
21/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of induced seismicity velocity weakening faulting regimes (experiments) in models fault reactivation ≠ seismicity !! more “clay”
no seismicity
seismicity
But, does this apply to reservoir-bounding faults?
Niemeijer & Spiers 2007 (UU)
22/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Glarus thrust fault, Swiss Alps
2/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Faults determine hydrocarbon accumulation S
N
TWT Buang-1 S
N
Corallina-1 BPLI
BPLI
TM1
high strain reactivated bounding faults
TM1
TM3 TM3
TE TE
From Gartrell et al., APPEA Journal 2005, modified for DFS by Clennell (CSIRO)
Juxtaposition seals, SGR > 60% KC 77 m oil column
no oil
KA
JO JO
non-leaking bounding fault 2000m 2000m
Example Timor Sea, Australia
KC KA
3/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Conventional techniques to determine the sealing capacity of faults and some of their limitations
4/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Shale gouge ratio can be used to determine sealing capacity of faults without considering stress state Analysis for entire fault section across slipped interval
Throw, T
Shale
SGR= (Vsh.z) / T x 100%
Sand
Vsh5, z5
Slipped interval (T) Photo: G Skerlec
Vsh4, z4 Vsh3, z3 Vsh2, z2 Vsh1, z1
Outcrop data (meter scale)
after Yielding et al. 1997
5/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
SGR extensively calibrated using reservoir well pressures, but poorly constrained (SGRcrit =10-20%)
Yielding et al. 1997
6/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of the sealing capacity of faults using integrated experimental-modelling studies
Brecht Wassing
Silvio Giger
Bogdan Orlic
Ben Clennell
7/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults ~4 m
Large rig to produce experimental clay smears in large samples (24x12x15 cm) at realistic stress conditions (n= 8 MPa) from Giger (CSIRO)
8/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Experiments simulated by discrete element models to determine SGR vs. shale properties and normal stress normal stress
normal stress
“fault zone”
displacement
synthetic shale: Vshale + mechanical properties synthetic & natural sandstone
matching mechanical properties
9/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults
10/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrated experimental-modelling studies provide new data on the sealing capacity of faults 1 0.9
shale content (V shale )
SGR=20%
dip angle fault: 60º shale layer thickness: 10 m
transition zone
0.7 faults sealing
0.6
SGR=10%
shale1
n=20 MPa
0.8
n=10 MPa shale2
schematic! ?
0.5 0.4
faults leaking
0.3 0.2 0.1 0 0
20
from geological model
40
60
80
fault displacement (D)
100
120
11/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Conventional techniques to determine fault reactivation and induced leakage and seismicity
(sources: TNO, KNMI – Royal NL Meteorological Institute, Ministry of Econ. Affairs)
12/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of strength and sealing capacity of faults by integrating experimental data and models 0.4
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
m=0.6
m=1.0
Mohr diagram
t/Sv
0.3
0.2
high n low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s'v 0.5 s n/Sv
0.6
0.7
0.8
Shmin
SHmax
4
4
slip tendency: shear stress/normal stress (/n)
0.9
1
Sv
3
13/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Slip tendency can be mapped on fault planes to determine critically stressed (leaking) fault segments 0.4
m=0.6
m=1.0
t/Sv
0.3
0.2
high n
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s' 0.5v s n/Sv
0.6
0.7
Shmin
0.8
SHmax
0.9
1
Sv
14/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Full field finite element models can account for local geology in predicting fault reactivation Feasibility of CO2 storage in the depleted De Lier gas field (onshore NL)
Structural geological model
Input to FEM
- rock properties - constitutive material models - in situ stress
NE GWC
Pressure load from reservoir simulator Pressure load for DIANA Static pressure
SW
N
Extrapolated pressure
Pressure [bar]
2D FE Model
Simulated pressure Pressure in DIANA
Time [year]
3 km
Enlarged FE model
10 km
Location of monitoring points Fault
Stacked reservoirs
Fault
Caprock
143 145 148
Aquifers
Bogdan Orlic (TNO)
15/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Variation in local stress due to CO2 injection can result in reactivation of fault segments Possible stress paths for the Bounding Fault for injection (el.143)
Location of monitoring points Fault
143 145 148
Shear traction [MPa]
25
Reversible stress path (elastic response) Partially reversible stress path Irreversible stress path (rigid response) Stress before injection Stress after injection MC: c=0, ф=35°
20 15
Fault re-activation? 10 5 0 0
5 10 15 20 Normal effective traction [MPa]
25
Bogdan Orlic (TNO)
16/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of fault reactivation and induced seismicity using experimental + modelling studies
17/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better methods to determine strength (friction coefficient) of faults 0.4
m=0.6
m=1.0
?
t/Sv
0.3
0.2
high n
Mohr-Coulomb failure total stress effective stress more stable after injection less stable after injection stable fault orientations stable fault orientations stable fault orientations critically-stressed orientations
low n
0.1
-aP 0 0
0.1
0.2
s'hmin
0.3
s'Hmax
0.4
s'v 0.5 s n/Sv
0.6
0.7
Shmin
0.8
SHmax
0.9
1
Sv
18/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability from geological model
log k f 4 SGR
Crawford et al. 2008
1 4
log( D )(1 SGR )
5
19/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability
Strong leaking fault segments
map in geological model to determine spill points
Weak sealing fault segments Crawford et al. 2008
20/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Integrating experimental-outcrop-modelling studies to couple fault composition, strength, and permeability
Gerco Hoedeman (TNO)
21/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
Better predictions of induced seismicity velocity weakening faulting regimes (experiments) in models fault reactivation ≠ seismicity !! more “clay”
no seismicity
seismicity
But, does this apply to reservoir-bounding faults?
Niemeijer & Spiers 2007 (UU)
22/24 13 February 2012 Jan ter Heege TNO Petroleum Geosciences
