29 Sep 07 - Erice 1
Advances in the Prediction of Shear Events
Induced From Mechanical Stimulation
T. Kohl & T. MégelGEOWATT AG
Dohlenweg 28, CH-8050 Zürich
29 Sep 07 - Erice 2
Contents
EGS site Soultz-sous-Forêts
FE Kernel / HEX-S Reservoir Model
Interpretation GPK3 /Stimulation Forecast of GPK4
Stimulation Planning at Coso EGS Project
29 Sep 07 - Erice 3
INTRODUCING GEOWATT AG
Geothermal Energy utilization
Engineering
Hydrogeology
Numerics
• Resource Evaluation• Generic studies High- and Low-Enthalpy Systems• Courses at University and Engineering Schools
• Dimensioning Heating/Cooling (BHE-Felder)• Reservoir Engineering• Measurements
• Tunneling: Inflow scenario• Flow system
• Coupled 3D-FE calculations• Data bank• Specific simulation tools
29 Sep 07 - Erice 4
European EGS Project Soultz s.F.Location and tectonic setting
Basel
Strasbourg
Karlsruhe
Soultz-sous-Forêts
Major V
osgian
Fault
Wes
t. Rh
ine
Faul
t
Adapted from Genter, 1994
SoultzGPK2,3,4
GPK1
GPK2/GPK3/GPK4
GPK1
29 Sep 07 - Erice 5
EGS Concept(Enhanced Geothermal System)
Production Injection
fractured network
surface Installations
Production: ~3MWe (≈10% Pt)
Natural fractures, artificially enhanced through "hydraulic or chemical stimulation"
29 Sep 07 - Erice 6
EGS - Modeling Approaches
• Essentially fractured medium, Darcy flow; non-Darcy flow at high fluid velocities in fractures
• Hydraulic coupling: advection• Thermal coupling: buoyancy, density, viscosity• Mechanical processes play an important
role in reservoir development and assessmentFracture mechanics
Shear fracturing
Tensile fracturing
Matrix elasticityPoroelasticityThermoelasiticity
• Injected fluid and formation fluids are different;biphasic flow or multicomponent transport
• Geochemistry also play an important role in reservoir characteristics
( ) nc στ ⋅Φ+= tan
pf
f
PSP
SP
⋅++>
+>
ασ
σ
min
min
29 Sep 07 - Erice 7
Hydraulic Stimulations and Microseismicity
• GPK2 Stimulation (July 2000): 14'080 events• GPK3 Stimulation (July 2003): 21'600 events• GPK4 Stimulation (September 2004): 5'753 events• GPK4 Stimulation (February 2005): 2'966 events• GPK4 1st Step rate test (February 2005): 183 events• GPK4 Acidization test (March 2005): 304 events• GPK4 2nd Step rate test (March 2005): 256 events
29 Sep 07 - Erice 8
Soultz - Microseismicity
Seismic Events at GPK2 and GPK3 after 1 day injection
29 Sep 07 - Erice 9
Microseismic Density
Total Stimulation EventsDensity calculated for cube volumes: 50x50x50m3
29 Sep 07 - Erice 10
Analysis of Microseismic Density
Comparison of GPK4 Stimulations:September 2004 February 2005
29 Sep 07 - Erice 11
Indication for Hydromechanical Interaction
Injection Rate and located events GPK3
[h] after 27.05.2003, 12:50
Eve
nts/
hour
Inje
ctio
nra
te[l/
s]
0 10 20 30 40 50 60 70 80 90 100 110 120 1300
50
100
150
200
250
300
350
400
450
500
0
10
20
30
40
50
60
70
© GEOWATT AG CH-Zürich
29 Sep 07 - Erice 12
Stochastic vs. Deterministic Reservoir Models
Deterministic Reservoir Models+ Control of all physical processes- Complex meshing - Simplified geometries
Stochastic Reservoir Models+ Refined representation of fractures/fault zones- Neglecting often matrix interaction- Hybrid approach
29 Sep 07 - Erice 13
FE Code Kernel
ORIGINAL GOAL:Simulation of coupled thermal-hydraulic-mechanical interactions
fully 3-D very versatile code (i.e. transport of radionuclides, chemical subst.)continuously new (linear & non-linear) mechanisms (i.e. turbulence)special treatment of fractures1-D, 2-D & 3-D elements with different shape functionsplatform independent (standard Fortran90)
2
1
3
4
2
1
3
4
5
2
13
4
5
6
29 Sep 07 - Erice 14
FE Code Kernel:Implemented Coupling SchemeHydraulic PRESSURE FIELD
transient Darcynon-linear (fracture aperture, non-Darcian)ρ(T, P) buoyancyμ(T, P)
Thermal TEMPERATURE FIELDtransient diffusion advectionλ(T, P), ρ(T, P)
Matrix Elasticity STRESS FIELDSteady statelinearPoro-ElasticityThermo-Elasticity
Fracture Mechanics FRACTURE APERTUREShearingCompliance
29 Sep 07 - Erice 15
FE Code Kernel:Elastic Matrix Mechanisms
Injection of cold fluid in a hot rock matrix
Thermo-elastic matrix stresses
Injection of pressurised fluid in ambient matrix
Poro-elastic matrix stresses
T3 Δ⋅⋅⋅= TT
ii KS βwith K Bulk modulus
βT coeff. linear expansion
PΔ⋅= BP
iiS αwith αB Biot coeff.
29 Sep 07 - Erice 16
Experience on deterministic H-T-M simulations
X
Y
r 30 years
3D Reservoir4 fracture zonesPoro-Elasticity & Thermo-Elasticity (Kohl, 1995)Normal aperture complianceNo shearing
-> not suited for hydraulic stimulation
Temperature
S22 stress change
29 Sep 07 - Erice 17
3D- Deterministic Fracture GeometryDynamic Long Term Behaviour
Synthetic exampleaperture: a0=200μm
t=10yrs
t=30yrs
-5100 -5050 -5000 -4950y-100
-80
-60
-40
-20
0z
-5-10-15-20-25-30-35-40-45
Situation after 30 yearsΔT [K]
-5100 -5050 -5000 -4950 y-100
-80
-60
-40
-20
0z
Situation after 10 years
-5100 -5050 -5000 -4950-100
-80
-60
-40
-20
0
Darcy-velocity
1 m s-1
-5100 -5050 -5000 -4950-100
-80
-60
-40
-20
02.5E-042.4E-042.3E-042.2E-042.1E-042.0E-04
aperture [m]
29 Sep 07 - Erice 18
New Model: HEX-S
FE algorithm for characterization of massive hydraulic stimulations
3D simulator of hydro-mechanical processesCoupling
→ fracture shear and compliance→aperture variation→Permeability
Mapping deterministic and stochastic fractures to FE grid
(Kohl & Mégel, 2007)
29 Sep 07 - Erice 19
Mechanical Shearing
Penny-shaped cracks with individual slip patches
Coulomb Shear Criteria
Compliance
Jacking Aperture: σn,eff <0
( )( )dilS
s
dilbasiceffn
UaKU Φ⋅=Δ=
Φ+Φ⋅−=Δ
tan;
tan,
τ
σττ
refn
effn
aa
,
,
0
91 σσ⋅+
=
29 Sep 07 - Erice 20
Hydraulic Flow in fractured medium
Full 3D transient hydraulic model
Discretization of borehole with 1-D elements
Mapping of fracture to rock matrix
explicit permeability adjustment
required hydraulic boundary conditions
(here laterally)
29 Sep 07 - Erice 21
'HEX-S' Flow Chart
Model setupPrediction of reservoir development
Stepwise structural interpretationof the microseismic cloud
Stepwise structural interpretationof the microseismic cloud
Mapping of the deterministic structures to a FE-grid
Mapping of the deterministic structures to a FE-grid
Mapping of stochastic structures to the FE-grid
Mapping of stochastic structures to the FE-grid
Pressure development in the reservoir
Pressure development in the reservoir
Adaptation of FE-grid permeability distribution
Adaptation of FE-grid permeability distribution
Shearing and opening of fracturesShearing and opening of fractures
Injection flow foreseenInjection flow foreseen
29 Sep 07 - Erice 22
Discretization 5km Reservoir
Numerical FE GridElement size 25m~400‘000 elementsDepth: 3000-6000 mVertical OH-sectionsLateral borders: Dirichlet BC
-6000
-5000
-4000
-3000
z
-4000
-2000
0
2000
4000
x
-6000
-4000
-2000
0
2000
4000
y
X Y
Z
XY
Z
GPK2
GPK4GPK3
29 Sep 07 - Erice 23
Fracture Data
Deterministic Borehole DataGPK2: Only incomplete UBI measurement to 6 major fracture zones according to BRGM interpretation.GPK3: 11 major fracture zones from UBI log (BRGM)GPK4: 11 major fracture zones from R. Maurer
Deterministic fractured zones4 major planes (1. Interpretation)
Stochastic fractured zones>2000 distributed planes
29 Sep 07 - Erice 24
Soultz Fracture Model
Deterministic slip patches Stochastic realization
GPK2 GPK3 GPK4
29 Sep 07 - Erice 25
Initial apertures:Deterministic / stochastic slip patches
Aperture: "stochastic" slip patches: 20 – 300 μm"deterministic" slip patches generally largerDepending on orientation and depth (stress field)Calibrated with initial borehole transmissivity
29 Sep 07 - Erice 26
First AssessmentGPK3 Stimulation
Stimulation GPK3Focus on major hydraulic injection steps
Fit of downholepressure historyHighly non-linear processesPermeability variation due to shearing
time [s]
dP
dh[P
a]
Flow
rate
[l/s]
0 100000 200000 300000 400000 5000000
2E+06
4E+06
6E+06
8E+06
1E+07
1.2E+07
1.4E+07
1.6E+07
1.8E+07
2E+07
0
10
20
30
40
50
60
70
80
90
100
Measured (depth corr.)ModelFlow
IIdata : 1IImodl : 1
IIdata : 0.78IImodl : 0.88
IIdata : 0.70IImodl : 0.84
IIdata : 0.60IImodl : 0.80
IIdata : 0.80IImodl : 0.94
IIdata : 1.04IImodl : 1.20
IIdata : 1.01IImodl : 1.25
Identical time constants
29 Sep 07 - Erice 27
Dynamic reservoir processes
Microseismic events GPK3 Simulated shearing GPK3
-1000
0
1000
-2000-1000
01000
-6000
-5000
-4000
-3000
-2000
-1000
0
Easting [m]
t =1 day
Identical spatial constants
29 Sep 07 - Erice 28
Soultz 3D Hydraulic Field
Strongly anisotropic pressure distributionRadial field only around injection pointPressure envelope oriented along fracture orientationPressure wave propagation in areas with highest degree of fracturation
GPK3Open borehole section
5 MPa2 MPa
Pressure distribution around GPK3
29 Sep 07 - Erice 29
Y
X
Z
1.70E+071.60E+071.50E+071.40E+071.30E+071.20E+071.10E+071.00E+079.00E+068.00E+067.00E+066.00E+065.00E+064.00E+063.00E+062.00E+061.00E+06
Reservoir Pressure Distribution
Flow aligned along seismic structures
Seismicity connected to zones of high pressure
Simulated shearing events & pressure distribution GPK3
29 Sep 07 - Erice 30
Modeling Tool HEX-S: Prognosis GPK4 stimulation 04SEP13
Measurement
-5500
-5000
-4500
Depth
0
500
East
ing
-2000-1500-1000
Northing
XY
Z
-6000
-5000
-4000
-3000
x3
-100001000
x1
-2000-100001000x2
OH-sectionmodel
Path of development
Forecast10h
1stDay
29 Sep 07 - Erice 31
Forecast GPK4 stimulation:HEX-S Hydraulic Model
Predicted evolution of shearing events at GPK4 after
Plane View View from West10 hrs stimulation
1 day stimulation
Events are not location-corrected Path of development
OH-sectionmodel
29 Sep 07 - Erice 32
Stimulation of GPK4 (13-16 Sep. 04)Microseismic Locations
Measured evolution of microseismic locations at GPK4
View from West: t= 10h View from West: t= 3 days
XY
Z
Path of development
29 Sep 07 - Erice 33
Modeling Tool HEX-S: Prognosis GPK4 stimulation 04SEP13
Forecast
ΔPdh ≈ 21 MPa for 30l/s additional ΔPdh ≈ 1.5 MPawhen increasing to 45 l/smaximum Pdh at t=15500 sGeneral characterisic:Short transient P-behaviour
Measurements
ΔPdh ≈ 15.5 MPa for 30l/s additional ΔPdh ≈ 1.05 MPawhen increasing to 45 l/smaximum Pdh at t=16’600 sshort transients
time [s]
ΔPdh
[MP
a]
Qin
Om
ega
100000 200000 3000000
5
10
15
20
0
10
20
30
40
50
60
70
80
90
100
Ref.pressure45.0 MPa
time [s]
ΔPdh
[Pa]
0 100000 2000000
5E+06
1E+07
1.5E+07
2E+07
2.5E+07
IImodl : 6.9-7.4 IImodl : 9.7-10.7
29 Sep 07 - Erice 34
Simulation GPK4 Increase of aperture (≅permeab.)
2.7 h 5 h 20 h
36 h 53 h
Iso-Surface = 10 μm
29 Sep 07 - Erice 35
New fracture volume
Flow injection vs. new fracture volume: Injection during simulation period: 6'300m3
Fluid losses over lateral drainage systemsCalculated Volume Change reflects data
30 l/s injection during first 50'000 s: -> 1000 m3
45 l/s during last 12'000 s -> Additional 500 m3
80% of the new fracture volume is generated from compliance ΔV(t) Δa(r) at 10 / 53 hrs
time [s]
ΔPdh
[Pa]
V[%
]
Vto
t[m
3 ]
0 50000 100000 150000 2000000
5E+06
1E+07
1.5E+07
2E+07
2.5E+07
3E+07
0
10
20
30
40
50
60
70
80
90
100
0
500
1000
1500
2000
Model C3
Vc[%]
Vs[%]
Vtot
Injection vs. Volume Change never explicitly
defined
29 Sep 07 - Erice 36
Importance Shear Failure
Percentage of aperture improvement
next to the borehole: up to 70% of the total aperture enhancement.
Predicted displacements in reservoir: max. 1cmnear borehole: up to 1.5cm
29 Sep 07 - Erice 37
Conclusions
HEX-S new hydro-mechanical reservoir modelDeterministic/stochastic fracture distributionDynamic reservoir behavior by joint H-M solution
Requirements:Geological / fracture reconnaissance Stress field
Successful forecast of GPK4 stimulation
Physical Reservoir ModelBased on integrated data analysis
29 Sep 07 - Erice 38
Future Application:
Microseismic impactGeneric studies
• using local reservoir properties• Design calculation to minimize shear displacement
Design of optimum reservoir creation strategiesoptimization of lateral seismic extensionDual vs. single injections
Resolve fundamental questionsMax. flowrate vs. max. volume injection
On Site support during stimulationModel update jointly with seismic recordingMinimizing test duration
29 Sep 07 - Erice 39
Coso Project Area
38 pad38A-938B-938C-9
34-9RD2
29 Sep 07 - Erice 40
Coso EGS site
Borehole trajectoriesopen hole section (thick lines)
-600
-400
-200
0
200
[]
-200 0 200 400
38C-9
38b-9
38a-9
38-9
34-9RD2
-3000
-2500
-2000
-1500
-1000
z[m
]
0
500
x [m]-500
0
y [m]
38C-9
38b-9
38a-9
38-9
34-9RD2
38A-9
38C-938B-9
34-9RD2
29 Sep 07 - Erice 41
34-9RD2
38A/B/C-9
Coso Model:Deterministic FZ in HEX-S model
29 Sep 07 - Erice 42
10 km
Coso Model:Stochastic & Deterministic FZ
29 Sep 07 - Erice 43
Coso:3D Hydraulic Model
Strongly anisotropic pressure distributionRadial field only around injection pointPressure envelope oriented along fracture orientationPressure wave propagation in areas with highest degree of fracturation
Pressure distribution around 34-9 RD2
29 Sep 07 - Erice 44
Coso:First model calculationsPossible development of microseismicity