Targeting horizontal stresses and optimal hydraulic ... · PDF fileoptimal hydraulic...

Targeting horizontal stresses and optimal hydraulic fracturing locations through seismic data

David GraySenior Research Advisor CGGVeritas

What matters in Hydraulic Fracturing?

FracturesWill they open?Will they stay in zone?How wide?What type?What direction?Will they stay open?

What do we need?

Rock StrengthYoung’s modulus, Poisson’s ratio, Unconfinedcompressional strength

StressHorizontal and Vertical stresses, σHmax, σhmin, σz

Closure stress = σhmin

Hoop stress = σθθ and Fracture Initiation PressureStress state

What can we get from seismic?

Rock StrengthYoung’s modulus, Poisson’s ratio

StressVertical and horizontal stressesClosure stressHoop stressStress state

How can we get these from seismic?

Rock StrengthYoung’s modulus, Poisson’s ratio

AVO/LMR (Amplitude Versus Offset / Lamé’s Moduli)Multicomponent Seismic Analysis

StressAzimuthal AVO, Azimuthal VelocitiesMulticomponent Seismic Fracture Analysis

When can we get these from seismic?

Rock Strength – NOW!AVO (1982-present, 27 years)LMR (1997-present, 12 years)Joint and Simultaneous Inversion (~2004–present, ~5 years)Multicomponent Analysis(~1980-present, ~30 years)

Stress – NOW!Azimuthal AVO (1996-present, 13 years)Multicomponent Fracture Analysis (~1986-present, ~23 years)Horizontal and Vertical Stresses (Now)

We’ll show you later

Rock Strength

Generating Rock Strength Values from Simultaneous or Joint Inversion of Seismic DataExample from SE of Red Deer, AB

Rock Strength

http://www.kgs.ku.edu/Geophysics/OFR/2006/OFR06_01/index.html

Brinell Hardness Test Hammer & Plate Seismic

Density – Simultaneous Inversion

Colorado Shale2WS

Shear Modulus (μ=ρ∗Vs2)

Important according to Goodway (CSEG lunch, Feb. 09), it can serve as a proxy for differential stress

Colorado Shale2WS

MuGPa

Lame’s Modulus (λ=ρ∗Vp2−2μ)

Colorado Shale2WS

LambdaGPa

Poisson’s Ratio (ν)

( )μλλν+

=2

Colorado Shale2WS

PoissonRatio

Young’s Modulus (E)

( )νμ += 12E Brittleness

Colorado Shale2WS

YoungGPa

Rock Strength Conclusions

Rock Strength derived from seismic since LMR in 1997Translate to values Engineering want

Young’s modulusPoisson’s ratio

Seismic measures “dynamic” rock strengthCalibrate to static values

Usually done with a scalar (e.g. 0.65)Static closer to dynamic in older, harder rocks like these

Stress

Vertical, Maximum and Minimum Horizontal Stresses from SeismicExample from SE of Red Deer, AB

How do we get stress from seismic?

Hooke’s LawStrain = Compliance * Stress

Isotropic Stress

We know σy > σx, in generalAnisotropic stresses

⇒ Anisotropic Compliance

Hooke’s LawStrain = Seismic * StressFor a Horizontally Transverse Isotropic (HTI) medium

Linear Slip Theory

How do we get stress from seismic?

ZNZN

ZT

How? - Horizontal Stresses and Seismic AzAVO

Seismic AzAVO TermsE – Young’ s Modulusν – Poisson’s RatioZN – Normal Compliance

ZNZN

Hooke’s LawStrain = Seismic * StressMovement = Seismic * ForceΔMovement = Seismic * ΔForceFracture = Seismic * ΔPressure

How do we get stress from seismic?

Assumptions

Some assumptions have been made in order to produce these results. Many are due to the preliminary nature of this work and will be addressed in future work. The assumptions are:1. Shear-wave background velocities for the inversions are

estimated from the “Mudrock Line” of Castagna et al (1985).

2. Calibration of estimated stresses, assumed the WCSB Atlas stress measurements are valid for this area.

3. MZN ≈ μZT

4. Azimuthal anisotropy in the seismic is due to differential horizontal stresses.

5. Linear Slip theory and all its assumptions are valid for these rocks.

6. The rocks behave elastically.

Vertical Stress Gradient from g*ρ

Colorado Shale2WS

GradientMPa/m

Vertical Stress

σz ≈ Σi[g*ρ(i) *z(i)]

Colorado Shale2WS

StressMPa

Calibrated Minimum Horizontal Stress

σhmin = σx = Κx∗σz

Colorado Shale2WS

StressMPa

Calibrated Maximum Horizontal Stress

σHmax = σy = Κy∗σz

Colorado Shale2WS

StressMPa

σx (= Closure Stress) in 2WS

Note range varies from ~20 MPa - ~26 MPa over this 10 km2 area.

Differential Horizontal Stress Ratio DHSR=(σH- σh)/σH

Colorado Shale2WS

Estimates require only seismically derivable parameters as σz cancels.

DHSRRatio

Stress State

What do these stress measurements tell us?Example from SE of Red Deer, AB

Stress variations with Depth (after Zoback)

σH > σz > σh

σz > σH > σh

σH > σh > σz

Directio

n o

f M

ovemen

t

Norm

al to

Fractures

Thrust = Horizontal fractures that move horizontally w/o much overburden allows them to go up

Shear = Vertical fractures that move horizontally

Normal = vertical fractures that move vertically

Stress and Fractures

If σHmax ≈ σhminTensile cracks any direction

|| rock weaknessFracture network

If σHmax >> σhmin (>5%)Fractures || σHmax

Shear FracturesTensile Fractures

Connect to existing fracture network for production σHmax

σHmax

Pressure

σhmin = Closure Stress

σhmin

DHSR with Orientation on Young’s Modulus

Young’s Modulus

DHSR

Cross-plot DHSR vs. E

E > 17 GPa, DHSR < 6%

Hydraulic fracture swarms where rocks are brittle (E > 17 GPa) and stress ratio is small (DHSR < 6%)

Aligned FracturesE > 17 GPa, DHSR > 6%Ductile

E < 12.5 GPa

Good and bad areas for Hydraulic Fracs in the Colorado

Colorado Shale

Areas in green indicate zones where hydraulic fractures will be optimal. In yellow, where aligned fractures are more likely to occur. In red, where hydraulic fracturing won’t work because the rock is too ductile.

2WS

Differential Horizontal Stress Ratio (σH- σh)/σH

Colorado Shale2WS

DHSRRatio

Good and bad areas for Hydraulic Fracs in the Colorado

Colorado Shale

Areas in green indicate zones where hydraulic fractures will be optimal. In yellow, where aligned fractures are more likely to occur. In red, where hydraulic fracturing won’t work because the rock is too ductile.

2WS

Young’s Modulus (E)

Colorado Shale2WS

YoungGPa

Probable zones of better Hydraulic Fractures in the 2WS Median

Map of 2nd White Speckled Shale showing zones highlighted in seismic attribute crossplot. Green is where fracture swarms will form, red is where the rocks are more ductile and yellow is where aligned fractures will occur. Note that only about ¼ of this reservoir is optimal for hydraulic fracturing.

Differential Horizontal Stress Ratio Slice

Probable zones of better Hydraulic Fractures in the 2WS Median

Map of 2nd White Speckled Shale showing zones highlighted in seismic attribute crossplot. Green is where fracture swarms will form, red is where the rocks are more ductile and yellow is where aligned fractures will occur. Note that only about ¼ of this reservoir is optimal for hydraulic fracturing.

Young’s Modulus in 2WS

Fracture Initiation Pressure Vertical Borehole

PFIS ≈ 3σhmin - σHmax

Colorado Shale2WS

Fracture Initiation Pressure Horizontal Borehole Side

PFIS ≈ 3σHmax - σv

Colorado Shale2WS

PressureMPa

Fracture Initiation Pressure Horizontal Borehole Top

PFIT ≈ 3σv - σHmax

Colorado Shale2WS

PressureMPa

Conclusions

What can we get from seismic?Rock Strength

Young’s modulus, Poisson’s ratioDynamic needs to be calibrated to static

StressVertical and Horizontal stresses, σz, σHmax, σhmin

Closure stress = σhmin

Hoop Stress, Fracture Initiation PressureStress state

Better places for hydraulic fracturesFracture type – vertical or horizontalFracture motion – vertical or horizontal

σH > σv > σh

σv > σH > σh

σH > σh > σv

Thrust = Horizontal fractures that move horizontally w/o much overburden allows them to go upShear = Vertical fractures that move horizontally

Normal = vertical fractures that move vertically

Acknowledgements

CGGVeritas, EnCana, Apache, Talisman, Magnitude, FairborneJonathan Miller, Paul Anderson, Bill Goodway, John Varsek, Louis Chabot, Doug Anderson, Ron Larson, John Logel, Eric Andersen, Jared Atkinson, Chris Irvine, Darren Schmidt, Franck Delbecq, Ron Schmid, Mark Houston, Christophe Maison, Dragana Todorovic-Marinic, Christian Abaco, Lee Hunt, Scott Reynolds

Targeting horizontal stresses and optimal hydraulic fracturing locations through seismic data

David GraySenior Research Advisor CGGVeritas