Reservoir Stress- Sensitivity BGD Smart JM Somerville M Jin
Reservoir Stress-Sensitivity
BGD Smart
JM Somerville
M Jin
Reservoir Stress-Sensitivity
• Reservoir properties and therefore behaviour influenced by changes in stress
• Caused by either changes in pore pressure or temperature, or combination
• Properties = permeability, dimensions, integrity
Stress-Sensitivity Scales
• Near wellbore– permeability – (stress skin cf skin caused by
invasion)– failure
• Increasingly distant from the wellbore– permeability
• Whole reservoir– permeability, directional floods
• Field– compaction, subsidence, seal alteration
Stress-Sensitivity Scales
• Near wellbore – Influenced by UBD– permeability – (stress skin, no skin
caused by invasion)– failure
• Increasingly distant from the wellbore– permeability
Reservoir Stress-Sensitivity: a multi-disciplinary challenge
More Realistic Reservoir
Model
Better Decisions
Reservoir Stress-Sensitivity: a multi-disciplinary challenge
More Realistic Reservoir
Model
Better Decisions
Stress Sensitivity
Better Decisions Re:-• Reserves• Well design• PI• Well locations• Production strategy• Reservoir management (inc 4D seismic)• Seal integrity• Compartmentalisation• Facilities• Efficacy of UBD technology and methodology
All impacting recovery factor and costs
HWYH-399
Breakout from CBIL(A)
Key:
Drilling-inducedtension from STAR
HWYH-394
Key:
Drilling-inducedtension cracks
All from STAR
Bed boundary
Fracture
Unclassified, possiblestylolite
The Conceptual Model
• The reservoir consists of blocks or layers of intact rock bounded by discontinuities
• The reservoir is stressed in an anisotropic manner
• The whole system exhibits hysteresis
Thinly-bedded interval in the Annot Sandstone.This intervalis underlain and overlain by more ‘massive’ sandstones.
The Reservoir
“Intact” Rock
Discontinuities
Boundary and Local Stresses within the Reservoir
Reservoir
Boundary or Regional Stresses
hh H
v
hHh
Local Stresses
v
v
v
Intact Rock Properties(stress-sensitive values where appropriate)
Ambient porosity and permeability Elastic constants E and v Biot’s coefficient Failure (Fracture) Criteria Vp and Vs velocities Vp anisotropy at ambient conditions Permeability at reservoir stress
conditions Palaeomagnetic trial
Tests with Specimen in Triaxial Cell
Stress-Sensitive Values of:-
• Elastic Moduli• Biot’s Coefficient• Permeability• Vp,Vs• Failure Criterion
1
1
22
1
2
P and S waves
Fluid flowing at pressure
Single State Triaxial Testing
1
1
2 = constant
Failure
1
1
2 2
1
1
2
x1
1
2’
2’’
2’’’
2’’’’
x
x
x
xx
x
x
Tan = Triaxial Factor
Failure Criterion - Triaxial Factor
Multi-Failure State Triaxial Testing
1
2
x1
1
2’
2’’
2’’’
2’’’’
x
x
x
x
x
x
x
Tan = Triaxial Factor
P Wave Velocity at 27.5MPa versus Porosity
y = -111.63x + 6753
R2 = 0.7776
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0
Porosity (%)
P W
ave
Vel
ocity
(m
/s)
Series1
UTMN 1307
HRDH 704
HWYH 325
HWYH 394
HWYH 399
Vp at 27MPa vs Porosity
Modulus of Elasticity at 27.6MPa versus Porosity
y = -1.7701x + 68.839
R2 = 0.5076
0.00
20.00
40.00
60.00
80.00
100.00
120.00
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0
Porosity (%)
Mo
du
lus
of
Ela
stic
ity
(G
Pa)
Series1UTMN 1307
HRDH 704HWYH 325
HWYH 394HWYH 399
Young’s Modulus at 27 MPa vs Porosity
Angle of Internal Friction versus Porosity
y = -1.3045x + 49.54R2 = 0.7722
0
10
20
30
40
50
60
70
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0
Porosity (%)
Ang
le o
f In
tern
al F
rict
ion
Series1
UTMN 1307
HRDH 704
HWYH 325
HWYH 394
HWYH 399
Angle of Internal Friction vs Porosity
Sampling Rationale - Intact Rock
Petrophysical Property
Rock Mechanics Property
Sample Core,
then Test
Wireline Log
Correlation
Populating Model - Intact Rock
Correlation
Synthetic Rock Mechanics Log
Convert Reservoir Characterisation Model into a Geomechanical Model
The Process• Populate the Conceptual Model with
properties and data
• So create a Geomechanical Model of the reservoir (plus surrounding rock)
• Impose process-induced changes on the Geomechanical Model using analytical or numerical solutions
• Numerical offers more realism than analytical – hence coupled modelling
Coupled Modelling
Fluid Flow Simulator
Stress-Analysis Simulator
Change in Pore Pressure, Temperature,
Saturations
Change in Effective Stresses
Rock Movements, Change in Stress and Strain
Change in Permeability
More realistic simulation results
Reservoir and o/b stresses, strains and
displacements
Differentiating Filter (Synthetic)
Fluid Flow Simulator
Stress-Analysis Simulator
Change in Pore Pressure, Temperature,
Saturations
Change in Effective Stresses
Rock Movements, Change in Stress and Strain
Change in Permeability
Enhanced 4D Seismic Interpretation/Reservoir
Management
Saturation-Related changes in Impedance
Stress-Related changes in Impedance
Changes in Velocity and Density
Differentiating Filter (Synthetic)
Fluid Flow Simulator
Stress-Analysis Simulator
Change in Pore Pressure, Temperature,
Saturations
Change in Effective Stresses
Rock Movements, Change in Stress and Strain
Change in Permeability
Enhanced 4D Seismic Interpretation/Reservoir
Management
Saturation-Related changes in Impedance
Stress-Related changes in Impedance
Changes in Velocity and Density
More realistic simulation results
Reservoir and o/b stresses, strains and
displacements
Example 1
UKNS, Perm Stress Sensitivity
(ECLIPSE coupled with VISAGE)
Production Prediction: permeability reduction
The diagram shows the absolute reduction (k1-k18). The maximum reduction in permeability is in the central part of the field
Perm sensitivity modelled with hysteresis
0.37000
0.37500
0.38000
0.38500
0.39000
0.39500
42000 44000 46000 48000 50000
mean stress (kPa)
k/k0
Series1
(ECLIPSE Output)
Stress Sensitive Permeability with hysteresis
0.37000
0.37500
0.38000
0.38500
0.39000
0.39500
42000 44000 46000 48000 50000
mean stress (kPa)
k/k0
Series1
Injection in Miller induced unloading
Injection in South Brae induced unloading in Miller Field
Depressurisation in Miller
Oil Production Rate is sharply reduced because the permeability reduction in the area causes a reduction in BHP and leads to a increase in gas production
Comparison of GOPR Predictions
(ECLIPSE Output)
Horizontal Ground Displacements - 1
Horizontal Ground Displacements - 2
Horizontal Ground Displacements - 3
Stress Ratio vs. time
Close to well
Far from well
Between wells
kq p
q p
3
1
3
3 2
/
/
Stress Status in p-q terms (anisotropy)
close to wells
far from wells
Stress Path Distribution
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61
S1
S4
S7
S10
S13
S16
S19
S22
S25
S28
S31
S34
S37
S40
S43
S46
S49
S52
S55
S58
k/k0
%
q
p
K stress path sensitive for Unconsolid Sand
90-100
80-90
70-80
60-70
50-60
40-50
30-40
20-30
10-20
0-10
Permeability Stress Path Sensitivity
Normalised Permeability Contoursk(%):
MOBIL "U"- Field: Unconsolidated Sand
Mean Effective Stress, p' (MPa)
#C4C2P2
#C4C2P4
#C4C2P6
#C4C4P1A
#C4C5P1
0
10
20
30
40
0 10 20 30 40 50
<= 30.0
<= 50.0
<= 70.0
<= 90.0
<= 35.0
<= 55.0
<= 75.0
<= 95.0
<= 40.0
<= 60.0
<= 80.0
<= 100.0
<= 45.0
<= 65.0
<= 85.0
> 100.0
Diff
eren
tial
Str
ess,
q
(M
Pa)
N/A(UCMS)
MATLAB
Excel
1
11
21
31
41
51
61
S1
S4
S7
S10
S13
S16
S19
S22
S25
S28
S31
S34
S37
S40
S43
S46
S49
S52
S55
S58
0
100
k
q
P'
p-q-k 3D
100-200
0-100
Compaction and subsidence
XY
Z
Model: MODL01L005: TIME/MONTHS *******Nodal DISPLACE YMax/Min on model set:Max = .504E-3Min = -.461E-1
-.4E-1-.35E-1-.3E-1-.25E-1-.2E-1-.15E-1-.1E-1-.5E-20.5E-2.1E-1.15E-1.2E-1
21-JAN-2000 10:53 compac05.cgmFEMGV 6.1-02 : HERIOT-WATT UNIVERSITY
XY
Z
Model: MODL01L018: TIME/MONTHS *******Nodal DISPLACE YMax/Min on model set:Max = .194E-1Min = -.339E-1
-.4E-1-.35E-1-.3E-1-.25E-1-.2E-1-.15E-1-.1E-1-.5E-20.5E-2.1E-1.15E-1.2E-1
21-JAN-2000 11:10 compac18.cgmFEMGV 6.1-02 : HERIOT-WATT UNIVERSITY
Compaction IN 1995 in which the result of injection is shown
Compaction in 1987 1
2
Example 2
UKNS, Seismic Stress-Sensitivity
(ECLIPSE, VISAGE, H-WU software)
Features of a 2D flow model grid embedded
for coupled geomechanical simulation
Overburden
Sideburden
Faults
Well
Gas , Water in the flow model grid
Caprock
Displaced shape of the geomechanical model
Surface subsidence
Differential compactionacross faults in reservoir
Typical location of shearstrain on faults
(VISAGE Output)
Mean effective stress distribution at the end of the simulation
Localized effectsat faults
Perturbed stress fieldabove and below reservoir
Unperturbed stress field(constant gradient) Apparent deepening of reservoir
due to decreasing pore pressure
(VISAGE Output)
Time-lapsed compressional acoustic impedance
Initial gas-water contact Changes in reservoirdue to fluid movement
Changes in reservoirdue to pore pressure decline
Changes in overburden/caprockdue to stress redistribution
Top of caprock
(VISAGE Output)
Initial Modelling: Before Production Begins
Time Lapse Model: Saturation Changes Only
Time Lapse Model: Saturation + Stress
Reflector at top of caprock
Reservoir base
Reservoir top
Time-lapsed seismic trace model
Pull-up in reflector eventdue to stress change effects
Perturbations at reflector event due to fluid change effects
Where are we now?
• Extreme examples of reservoir stress-sensitivity accepted: Ekofisk, HP/HT, Gulf of Mexico, Angola?
• The processes required exist in usable form
• Non-uniform levels of commitment• What about the more subtle
reservoirs?
Technical Challenges
• Discontinuity distributions• Discontinuity properties• Rel perm stress-sensitivity• In situ stress state• Coping with anisotropy• Seamless software
Organisational Challenges
• Realising the full value of the data we already have
• Cost vs value of the process
• Coping with multi-disciplinarity
Is this too much to ask for?
Shared analysis
Shared belief
Fully owned
decisions
Better performance
Decision Making• Straight from the geomechanical model,
aided possibly by some calcs, e.g.– fracture density = well locations for max PI– subsidence = yes or no
• With the aid of coupled modeling, e.g– improvement of appraisal– impact of perm sensitivity = recovery, GOR etc– Ground movements and subsidence = threat
to wells and facilities– 4D seismic enhancement = better management
Thank You
What do we want to achieve today?
• Overview of the main tasks of the project
• Select candidate reservoirs for study
• Set up communications
• Agree next meeting date 17th August?
Hysteresis
K
Increasing Stress
Hysteresis
K
Increasing Stress
X
Hysteresis
K
Increasing Stress
UBD site history very important
Effective Stress around the wellbore
Time
Drilling Completion Production
Failure Level
Multi-Disciplinary
Tasks assembling
data for Model
Deliverables
feedback toimprove
characterisation
feedback to improve
characterisation
Basin process simulations
*Genetic Units expertise Analogue studies
*Geomechanics of fracture genesis
*Published and proprietary studies
Stress-Sensitive Reservoir Modelling
and Coupled Simulations (Ground
movements, Fluid Flow and 4D seismic)
Characterise Reservoir Rocks Characterise Reservoir Faults & Fractures
Characterise Structural Setting of the Reservoir
*Log analysis
Reservoir Geomechanical Model
Better Decisions Reservoir Management
*Geomechanical Core Analysis*Structure and anisotropy analysis from Seismic
Creation of the Geomechanical Model
Stress-Sensitive Coupled Modelling
Building the Geomechanical Model