Geomechanic for Hydraulic Fracturing
in Unconventional Reservoirs
1
Raul German Rachid
Production Stimulation Engineer
Schlumberger
Argentina Bolivia Chile
04-Jul-11
Agenda
2
� Introduction to Unconventional Reservoirs
� Consequences of Heterogeneity and Lamination
� Stress Profile Modeling in Anisotropic Media
� Horizontal Wells
� Fracture Geometry Simulation
� Conclusions
04-Jul-11
Sch
lum
berg
er Pu
blic
Introduction to
Unconventional
3S
chlu
mb
erger P
ub
lic
Unconventional
Reservoirs
04-Jul-11
Sch
lum
berg
er Pu
blic
The Industry Challenge
� Orders of magnitude reductions in permrequire orders of magnitude increase inreservoir contact
� Efficiency key to economic successConventional ReservoirsConventional ReservoirsConventional ReservoirsConventional Reservoirs
Small volumes that are
easy to develop
Unconventional Unconventional Unconventional Unconventional
1md +1md +1md +1md +
0.01 md
4S
chlu
mb
erger P
ub
lic
Unconventional Unconventional Unconventional Unconventional
Large volumes
difficult to
develop
0.001 md
0.00001 md
04-Jul-11
Sch
lum
berg
er Pu
blic
Shales are unconventional reservoirs S
chlu
mb
erger P
ub
lic
504-Jul-11
Sch
lum
berg
er Pu
blic
Shale Gas Introduction
What are they?– Organic-rich shale– Source rocks
– Adsorbed and free gas
– Very low permeability
6S
chlu
mb
erger P
ub
licCommon traits of gas shale reservoirs– Abundant gas (20 to 400 BCF/mi2)
– Large developments (economies of scale)– Large and numerous hydraulic stimulations
– Long well life (60-year reserves common)
04-Jul-11
Sch
lum
berg
er Pu
blic
Trap, Reservoir & Source Rock
Trap/Seal
Trap
Conventional Unconventional
Sch
lum
berg
er Pu
blic
04-Jul-11
7
Reservoir
Source
Reservoir
Source
Rock is too tight to let go of Hydrocarbon
so source rock acts as the trap
and the Reservoir
Hydrocarbon leaves source and settles
in the reservoir because it cannot
pass the trap
Sch
lum
berg
er Pu
blic
Heterogeneous Rock at Fine Scale
Matrix
8S
chlu
mb
erger P
ub
lic
SPE 131772
Kerogen
04-Jul-11
Sch
lum
berg
er Pu
blic
Sal
t
Ara
b-D
Car
bona
te
Jona
h La
nce
Fm
. (T
ight
GA
S)
Cem
CR
ET
E
Org
anic
Sha
le
Bric
k
Ber
ea S
and
Shale in Perspective: Permeability
9S
chlu
mb
erger P
ub
lic
1000 10100 1.0 0.1 0.001 0.0001 0.00001 1e-060.01
Bric
k
Ber
ea S
and
md
Unconventional
04-Jul-11
Sch
lum
berg
er Pu
blic
Earth Scale
SEM Scale
Mechanical Properties for ShalesThe Consequence of Laminations
10S
chlu
mb
erger P
ub
lic
Core Scale
Thin Section Scale
SEM Scale
Log Scale04-Jul-11
Sch
lum
berg
er Pu
blic
Unconventional Shale Gas Reservoirs
�Hydraulic fracture containment is often either unknown or perceived as uncertain.
�Traditional stress modeling in shale gas reservoirs has
11S
chlu
mb
erger P
ub
lic
�Traditional stress modeling in shale gas reservoirs has lead to inefficient fracturing or unexpected height growth.
�However by considering anisotropic rock properties……
04-Jul-11
Sch
lum
berg
er Pu
blic
Heterogeneity
A heterogeneous material is one consisting of
dissimilar or diverse constituents
Homogenous Heterogeneous Heterogeneous
12S
chlu
mb
erger P
ub
lic
04-Jul-11
Sch
lum
berg
er Pu
blic
Anisotropy
Isotropic Anisotropic Anisotropic
Anisotropy is defined as the variation of a property
with the direction in which it is measured.
13S
chlu
mb
erger P
ub
lic
Evaluate using core, logs and seismic04-Jul-11
Sch
lum
berg
er Pu
blic
What is a Transversely Isotropic medium?
Isotropic mediaIsotropic mediaIsotropic mediaIsotropic media
Same property in the 3 principal directions of space
Transverse isotropicTransverse isotropicTransverse isotropicTransverse isotropic
Property is the same in 2 principal directions:
- TIV same property in horizontal plane
- TIH same property in vertical plane
X
Y
Z
X
Y
Z
Isotropic
14S
chlu
mb
erger P
ub
lic
- TIH same property in vertical plane
OrthotropicOrthotropicOrthotropicOrthotropic
Property varies in 3 directions
XY
Z
X
Y
Z
TIV
TIH
Orthotropic04-Jul-11
Sch
lum
berg
er Pu
blic
Stress Modeling of Shales
Jaeger and Cook – Fundamentals of Rock Mechanics (1979)
This is the case of a sedimentary rock with z-axis perpendicular to the bedding,
and the increase of the number of elastic constants from two for the isotropic
case to five is formidable. There is no great difficulty in handling many
mathematical problems involving such materials, cf. Hearmon (1961), Savin
(1961); the difficulty for practical purposes is in obtaining and using realistic
values of the elastic constants.
15S
chlu
mb
erger P
ub
lic
values of the elastic constants.
( )pVph Pv
vP ασασ −
−=−
1
( )pVh
V
V
hph P
v
v
E
EP ασασ −
−=−
1
Isotropic
TransverseIsotropic
04-Jul-11
Sch
lum
berg
er Pu
blic
Traditional Stress Modeling: Isotropy� σ vertical
Stress Profile
Shale
Sandstone
Isotropy assumes that:
Horizontal = Vertical
16S
chlu
mb
erger P
ub
lic
σ h
σ vσ p
( ) tectonicspvph +−−
=− σσυ
υσσ1
Shale
Sandstone All conventional sonic tools !!
04-Jul-11
Sch
lum
berg
er Pu
blic
Stress Profiles in Anisotropic Rock
Where:
� E = Young’s Modulus Vertical
� v = Poisson’s Ratio Vertical
σ vertical
E, vE’, v’
17S
chlu
mb
erger P
ub
lic
� v = Poisson’s Ratio Vertical
� E’ = Young’s Modulus Horizontal
� v’ = Poisson’s Ratio Horizontal
( ) tectonicsh
v
Ev
Ehpvph +−
−=− σσ
υυσσ
1
Laminated Shale which is the
reservoir & source rock
04-Jul-11
Vertically Anisotropic Formation – Impact on Frac Height
Isotropic Stress Anisotropic Stress Sonic Scanner
Leads to more accurate mechanical properties in laminated shales
18
properties in laminated shales
04-Jul-11
Sch
lum
berg
er Pu
blic
Staging the
Stimulation
Anisotropic Stress
Profile
19S
chlu
mb
erger P
ub
lic
Isotropic Stress
Profile
04-Jul-11
Sch
lum
berg
er Pu
blic
Comparison of Isotropic and Anisotropic Models
Low stress predicted in shales using conventional earth model in high clay volume rocks
MineralogyGR
Stress
Gradients
Higher stress predicted in shales
20S
chlu
mb
erger P
ub
lic
Higher stress predicted in shales using anisotropic earth model in high clay volume rocks
Measured stresses via in-situ stress testing
04-Jul-11
Sch
lum
berg
er Pu
blic
Impact of Anisotropic Stress Profile
4300
4400
4500
4600
4700
4800
TV
D -
ft
Frac Half Length Propped Frac Half Length4300
4400
4500
4600
4700
4800
TV
D -
ft
Frac Half Length Propped Frac Half Length
Isotropic Stress Profile Fracture Geometry Anisotropic Stress Profile Fracture Geometry
Shale Reservoir Shale
Reservoir
21S
chlu
mb
erger P
ub
lic
4900
5000
51003500 3600 3700 3800 3900 4000 4100 4200
Stress - psi
-0.2 -0.1 0.0 0.1 0.2
Width at Wellbore - in @118.0 min
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Fracture Half Length - ft @118.0 min
4900
5000
51003000 3500 4000 4500
Stress - psi
-0.10 -0.05 0.00 0.05 0.10
Width at Wellbore - in @119.0 min
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Fracture Half Length - ft @119.0 min
Hydraulic fracture contained within the organic shale
Hydraulic fracture grows above the organic shale
Best barriers for organic shales are conventional, high clay volume inorganic shales
Fractures that grow out of zone will result in poor production regardless of the Reservoir Quality
04-Jul-11
Sch
lum
berg
er Pu
blic
Proppant Concentration
0.0 PPA 2.0 PPA 4.0 PPA 6.0 PPA 8.0 PPA
Fracture WidthFrac#1 MD = 13495.85ft
Lithology and StressSat. and Young's Modulus
Well
Depth
(TV
D) (ft)
13100
13200
13300
13400
13500
13600
Impact of Stress model on hydraulic fracture Isotropic Vs Anisotropic assumption
22S
chlu
mb
erger P
ub
lic
> 10.0 PPA
Length (ft)
0 100 200 300 400 500 600 700
Width (in)
-2.00 0 2.00
Stress (psi)
8000.0 10000.0 12000.0 14000.013700
Young's Mod. (psi)
4.0x1066.0x1068.0x1061.0x1071.2x107
Proppant Concentration
0.0 PPA 2.0 PPA 4.0 PPA 6.0 PPA 8.0 PPA
> 10.0 PPA
Length (ft)
0 500 1000 1500 2000
Fracture WidthFrac#1 MD = 13487.00ft
Width (in)
-2.00 0 2.00
Lithology and Stress
Stress (psi)
10000.0 12000.0 14000.0 16000.0
Sat. and Young's Modulus
We
ll D
ep
th (
TV
D)
(ft)
13100
13200
13300
13400
13500
13600
13700
13800
Young's Mod. (psi)
4.0x106 6.0x106 8.0x106 1.0x107
04-Jul-11
Sch
lum
berg
er Pu
blic
Anisotropy and Fracture Containment
( )pVph Pv
vP ασασ −
−=−
1
( )Vh PvE
P ασασ −=−
Isotropic Blue (v)
Anisotropic Red (Eh, EV, νh, νV)
Sch
lum
berg
er Pu
blic
( )pVh
V
V
hph P
v
v
E
EP ασασ −
−=−
1
Leads to more accurate mechanical properties inlaminated formations
Sch
lum
berg
er Pu
blic
`
24S
chlu
mb
erger P
ub
lic
04-Jul-11
Sch
lum
berg
er Pu
blic
Frac Breakdown
Pressure
4000
5000
6000
7000
Pre
ssu
re [p
si]
40
50
60
70
Slu
rry Rate [b
pm
]P
rop C
on
x 10 [PP
Ax10]
Koone 4-34H Stage 2
0
1000
2000
3000
4000
5000
6000
7000
8000
0 15 30 45 60 75 90 105 120
Time [min]
Pre
ssu
re [
psi
]
0
10
20
30
40
50
60
70
80
Slu
rry Rate [b
pm
]P
rop
Co
n x 10 [P
PA
x10]
Lithology σHmin
25S
chlu
mb
erger P
ub
lic
0
1000
2000
3000
0 15 30 45 60 75 90 105 120
Time [min]
Pre
ssu
re [p
si]
0
10
20
30
Slu
rry Rate [b
pm
]P
rop C
on
x 10 [PP
Ax10]
04-Jul-11
Sch
lum
berg
er Pu
blic
Unconventional
Reservoir
Fracturing
26S
chlu
mb
erger P
ub
lic
Fracturing
Evaluation04-Jul-11
Sch
lum
berg
er Pu
blic
Hydraulic Fracturing Direction
27S
chlu
mb
erger P
ub
lic
04-Jul-11
Hydraulic Fracturing Direction
SH
S
ShSH
Sh
28
� Low stress anisotropy
� Lower seismic anisotropy
� Wide fracture fairway
� High stress anisotropy
� Higher seismic
anisotropy
� Narrow fracture fairway
SH
Sh
SH
Sh
04-Jul-11
Fracture Geometry Information from Horizontal Image LogsVariable Induced Fractures Infers Variable Stress
Transverse Fractures Only:
σσσσHHHH >> σσσσhhhh
No Fractures:
High High High High σσσσ’’’’
Long & Trans Fractures:
Low Low Low Low σσσσ‘ & ‘ & ‘ & ‘ & σσσσHHHH ~ σσσσh h h h
Long, Narrow Fracture FairwayLong, Narrow Fracture FairwayLong, Narrow Fracture FairwayLong, Narrow Fracture FairwayWide Fracture FairwayWide Fracture FairwayWide Fracture FairwayWide Fracture FairwayNo FracturesNo FracturesNo FracturesNo Fractures
29
04-Jul-11
Sch
lum
berg
er Pu
blic
SPE-90051 (HW)
30S
chlu
mb
erger P
ub
lic
SRV or ESV04-Jul-11
Sch
lum
berg
er Pu
blic
SPE-90051
31S
chlu
mb
erger P
ub
lic
04-Jul-11
Sch
lum
berg
er Pu
blic
SPE-90051
32S
chlu
mb
erger P
ub
lic
Effective use of the fracturing fluid
Volume optimization using the fracture acoustic volume for make real time decisions.
04-Jul-11
Sch
lum
berg
er Pu
blic
geophonesgeophonesgeophonesgeophones
Effective Stimulated VolumeEffective Stimulated VolumeEffective Stimulated VolumeEffective Stimulated Volume
(ESV) density based algorithm(ESV) density based algorithm(ESV) density based algorithm(ESV) density based algorithm
StimMAP* LIVE – Quantifying
Contact Volume
33S
chlu
mb
erger P
ub
lic
A Tool to make informed decisionsA Tool to make informed decisionsA Tool to make informed decisionsA Tool to make informed decisions
Event HistogramEvent HistogramEvent HistogramEvent Histogram
04-Jul-11
Sch
lum
berg
er Pu
blic
Hydraulic fracture mapping for evaluation
4150D
epth
(ft
)Microseismic Measurements
with prop-placement model
34S
chlu
mb
erger P
ub
lic
4850
-500 -250 0 250 500Along Fracture Length (ft)
Dep
th (
ft)
4850
SPE 38575 ( DOE-GRI MWX data )
04-Jul-11
Sch
lum
berg
er Pu
blic
Slickwater Fluids
– More ft2/$– Wider Fracture “Fairways”
Geled fluids
Fracturing Fluid Selection
35S
chlu
mb
erger P
ub
lic
Geled fluids
– Frac Initiation
04-Jul-11
Sch
lum
berg
er Pu
blic
Horizontal
Wells
36S
chlu
mb
erger P
ub
licWells
04-Jul-11
Sch
lum
berg
er Pu
blic
Fracture Treatments – Increase Surface Area & Flow
37S
chlu
mb
erger P
ub
lic
04-Jul-11
Sch
lum
berg
er Pu
blic
SurfaceSurface
Wellbore azimuth 90°
Which direction to drill?
Where to land?
Fracture height growth?
Fracture network width?
Fracture conductivity?
Transverse Aplication:
Place Multiple Fracutes
Hydraulic Fractures In Horizontal Wellbores
38S
chlu
mb
erger P
ub
lic
Vertical StressWellbore azimuth 0
Longitudinal fractures
Reservoir
Minimum horizontal stress
Vertical Stress
Maximum horizontal stress
Wellbore azimuth 90°Transverse fractures
04-Jul-11
Sch
lum
berg
er Pu
blic
Bedding– Lamination → Complexity
Quartz Rich Shales� Isotropic Behavior
Drilling/Stimulation Efficiency
Lateral Placement
39S
chlu
mb
erger P
ub
lic
� Expanding Clays
� Oil Based Muds
� Borehole Breakout
Closure Stress
04-Jul-11
Sch
lum
berg
er Pu
blic
Induced Stress due to a Horizontal Hole in σσσσh Direction
Drilling Process can induce Tensile Stress
Potential Initiation of Tensile Fractures
Longitudinal Fracture
Sch
lum
berg
er Pu
blic
Borehole
In-Situ Stress Field
Induced Stress Field
Sch
lum
berg
er Pu
blic
Factors Affecting Fracture GeometryFactors Affecting Fracture GeometryFactors Affecting Fracture GeometryFactors Affecting Fracture Geometry
Weijer 1994 – Fracture
initiate longitudinal when
OH drilled ┴ to σmax, then is
reoriented to transverse .
The effect of induced stress concentration
created by removing a cylinder of supporting
rock is known as “Hoop Stress”.
Sch
lum
berg
er Pu
blic
Sch
lum
berg
er Pu
blic
Microseismic Data and Fracture Orientation
42S
chlu
mb
erger P
ub
lic
04-Jul-11
Sch
lum
berg
er Pu
blic
SPE 11056243
Sch
lum
berg
er Pu
blic
04-Jul-11
Sch
lum
berg
er Pu
blic
Stage 1Stage 1Stage 1Stage 1 Stage 2Stage 2Stage 2Stage 2 Stage 3Stage 3Stage 3Stage 3 Stage 4Stage 4Stage 4Stage 4
σmax avg = .77.77.77.77
σmin ISIPISIPISIPISIP = .70.70.70.70
σM- σm = .07.07.07.07
σmax avg = .74.74.74.74
σmin ISIPISIPISIPISIP = .64.64.64.64
σM- σm = .10.10.10.10
σmax avg = .69.69.69.69
σmin ISIPISIPISIPISIP = .65.65.65.65
σM- σm = .04.04.04.04
σmax avg = .65.65.65.65
σmin ISIPISIPISIPISIP = .62.62.62.62
σM- σm =.03.03.03.03
SPE 110562S
chlu
mb
erger P
ub
lic
High High High High Stress Stress Stress Stress
AnisotropyAnisotropyAnisotropyAnisotropy
Low Low Low Low Stress Stress Stress Stress
AnisotropyAnisotropyAnisotropyAnisotropy
Sch
lum
berg
er Pu
blic
Stage 1Stage 1Stage 1Stage 1
Stage 2Stage 2Stage 2Stage 2
Stage 3Stage 3Stage 3Stage 3
Stage 4Stage 4Stage 4Stage 4
σM- σm = .07.07.07.07
σM- σm = .10.10.10.10
σM- σm = .04.04.04.04
σM- σm =.03.03.03.03
∆ P = P = P = P = ----191191191191
∆ P = 249P = 249P = 249P = 249
∆ P = 564P = 564P = 564P = 564
∆ P = 1109P = 1109P = 1109P = 1109
SPE 110562S
chlu
mb
erger P
ub
lic
σM- σm = .10.10.10.10 σM- σm =.03.03.03.03
Stage 2 Stage 3
Sch
lum
berg
er Pu
blic
Hydraulic
Fracturing
46S
chlu
mb
erger P
ub
lic
Fracturing
Simulation
04-Jul-11
Sch
lum
berg
er Pu
blic
Fracture geometry modelingDifferences between pseudo 3-D and planar 3-D
47S
chlu
mb
erger P
ub
lic
Propagation and fluid flow are 1-D
Assumes fracture length >> height (plane-strain assumption)
Propagation and fluid flow are 2-D
No assumption/restriction on the aspect ratio (length vs height)
Only constraint – fracture stays within one plane (no bending or turning)
Planar 3Planar 3Planar 3Planar 3Planar 3Planar 3Planar 3Planar 3--------D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3D models are more accurate in layered reservoirs than pseudo 3--------D, which will D, which will D, which will D, which will D, which will D, which will D, which will D, which will
maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and maximize benefit from petrophysical and geomechanicalgeomechanicalgeomechanicalgeomechanicalgeomechanicalgeomechanicalgeomechanicalgeomechanical datadatadatadatadatadatadatadata04-Jul-11
What is it?– Multi-stage stimulation design and evaluation software for conventional and
unconventional markets
– Integrated in the overall oil field services’ multi-disciplinary solutions;
…petrophysics, G&G, geomechanics, reservoir engineering
How does it work?– Implemented as a Plug-in for Petrel
Unconventional Reservoir Simulator (Mangrove*)48
– Implemented as a Plug-in for Petrel
What is the value?� Differentiate through technical solution rooted in reservoir
characterization (measurements and interpretation),
enabling reservoir centric stimulation design for specific
environments
� Reorient for the global shift to unconventional reservoirs
04-Jul-11
Multi-staging Advisors�Tight Gas Sandstone & Pilot Shale (Vertical)– 100 separate stacked sands over 3000 ft gross
– Differential depletion
– Starting point AutoFRAC (Denver)
� Shale (Laterals)
– Laterals through heterogeneous rock
– Ultra low permeability
–
49
Source: www.eandp.info
– Naturally Fractured
Completion challenges
– Consistent model
– Tedious process (2 days – 2 weeks)
Improved Efficiency, Consistency & Knowledge DisseminationImproved Efficiency, Consistency & Knowledge DisseminationImproved Efficiency, Consistency & Knowledge DisseminationImproved Efficiency, Consistency & Knowledge Dissemination
04-Jul-11
50
04-Jul-11
HF-NF Interaction (Crossing Criterion)
NF Arrest /
slippage
NF dilation
HF propagating
along NF
HF
51
Crossing
Fissure opening after
crossing
NF staying
closed
04-Jul-11
Example of UFM Results with micro-seismic data
52
04-Jul-11
Integrating Reservoir & Completion Quality
30000
40000
50000
3 M
on
th
BO
E
� 33% increase in 3 month average cumulative BOE on
new wells. Save $300k in frac costs
� New wells used Reservoir Quality and Completion
Quality to optimize completions.
Effective Porosity
Spectroscopy Volumes
Examine Reservoir and Completion Quality.
Recommend stages with optimal
properties. Variable
53
0
10000
20000
30000
Pre Optimization (6 Wells) Post Optimization (3 Wells)
3 M
on
th
BO
E
Perforations
Shale Stress Index
Poisson’s Ratio
Stress
with optimal properties. Variable number and lengths.
Recommend specific perforation location
− Combined logs and core measurements for the reservoir and completion quality assessment.
− Reservoir Quality technology routine: Triple Combo-Spectroscopy (PEX-ECS/ EcoScope), Di-Electric Scanner, NMR
− Completion Quality technology routine: Borehole Images (FMI, RAB, LWD Density), Sonic Scanner/Mangrove*
04-Jul-11
Conclusion & Summary
54
� Unconventional Reservoirs require special consideration related to the heterogeneity.
� Conventional Isotropic stress models can lead to erroneous evaluations.
� Geomechanical Models are becoming more important in the � Geomechanical Models are becoming more important in the process of the Reservoir completion.
� It’s not just about technology, it’s about integrating appropriate technology.
04-Jul-11
Sc
hlu
mb
erg
er P
rivate
Sch
lum
berg
er Pu
blic
Questions or
Comments?
Sc
hlu
mb
erg
er P
rivate
Sch
lum
berg
er Pu
blic
Comments?