The Magic of Lamé

Bill Goodway

Location: University of Saskatchewan

Student Geophysical Society
Saskatoon.

Date: September 11th 2009

Acknowledgements
ƒ Society of Exploration Geophysicists
ƒ Shell Sponsorship

ƒ EnCana Corporation
 SEG Membership
ƒ Technical Journals in Print and Online
ƒ Networking Opportunities
ƒ Receive Membership Discounts on:
ƒ Continuing Education Courses
ƒ Publications
ƒ Workshops and Meetings Free 
Bookmark

SEG materials are available today!

Join Online x x x http://seg.org/join

Who was Lamé and what is the physical significance of his
parameters Lambda (λ) and Mu (μ) ?
ƒ Gabriel Lamé (1795-1870): French engineer,
mathematician and elastician.
ƒ Introduced λ and μ in 1828, named after himself,
in a series of lectures titled:
“Mémoire sur l’éqiuilibre intérieur des corps
solides homogènes”
ƒ Lamé formulated the modern version of Hooke’s
law relating stress to strain in general tensor form,
creating the basis for the science of materials,
including rocks.
ƒ Interestingly and most notably, only Lamé’s
moduli λ and μ appear in Hooke’s law and not
Young’s modulus, the bulk modulus, or any other
modulus or modulus ratio.

Any apparent likeness to myself is purely coincidental!

Disclaimer for use of numerous equations that follow:
“If geophysics requires mathematics for its treatment it
is the Earth that is responsible not the geophysicist.”
from Sir Harold Jeffreys, University of Cambridge

Assertions
ƒ Lamé moduli of rigidity μ and incompressibility λ allow the
fundamental parameterization of seismic waves used to
extract information about rocks in the Earth.

ƒ These parameters link many fields of Earth Science at

different scales, from Petroleum Exploration to Earthquake
Seismology.

ƒ Other common formulations result in contradictions which

are removed by restating equations using Lamé parameters.
Outline
„ P-wave, S-wave propagation, Hooke’s Law and Lamé moduli
„ Motivation from logs
„ AVO equations and methods to invert for Lamé moduli
„ Examples of AVO inversion for elastic parameters from 3D data
„ Rock Mechanical Properties and Closure Stress
„ Relating stress (fracture) intensity and orientation to Passive
Microseismology and Earthquake Seismology
„ Conclusions
Modes of Seismic Wave Propagation
Shear S Wave Uni-axial Compressional
P Wave
Undeformed cross-section
of porous sandstone

resistance
resistance
shear

shear
compressive resistance
Shear Velocity Compressional Velocity
(Dipole Log) (Sonic Log)
μ λ+2μ
Vs = Vp =
ρ ρ
Note λ+2μ, the unnamed P-wave modulus is a direct consequence
of bound rocks in the Earth
Basic relations: Hooke’s law, moduli and wave equations
Hooke’s law as formulated by Lamé:
σ ij = λδ ije v + 2μe ij
for shear stress σxz in x direction on the z normal face (i ≠ j) so δij = 0:
σ xz = 2μe xz
for normal axial stress in z direction (i = j) so δij = 1:
σ zz = (λ + 2μ )e zz + λe xx + λe yy

Wave equations for P-wave and S-wave propagation “sense” attributes of the
medium by equating Newton’s 2nd law (F = Ma) to Hooke’s law.

P-wave propagation of the volume strain eV term in Hooke’s law

d 2eV
ρ 2 = (λ + 2μ )∇ 2 e V
dt
S-wave propagation of the shear strain exz term in Hooke’s law :
d 2 e XZ
ρ 2
= μ∇ 2
e XZ
dt
Static Moduli and Moduli Ratio Definitions in Lamé terms
Lamé parameters:
Rigidity Mu (μ) and “Pure Incompressibility” Lambda (λ)
Common moduli resulting from medium’s measurement condition:
“Compressional P-wave Modulus” M = λ + 2μ
(Bound uni-axial compression)
Young’s Modulus E = μ(3λ+2μ)/(λ+μ) E = M – 2λν
(Unbound uni-axial compression)

Bulk Modulus K = λ + (2/3)μ K = M – (4/3)μ

„ Poisson’s ratio ν = λ /(2 λ + 2μ)

„ Vp/Vs ratio √(2 + λ/μ)
A given material has various moduli that are purely a function of
measurement conditions
Lamé parameters λ and μ are invariant and form the basic elements
within moduli, giving a simpler physical meaning
 Confusing implication of –ve Lambda from Bulk Modulus
Hydrostatic pressure P or tri-axial stress σ in Cartesian coordinates
− P = σ ij = λδ ije V + 2μe ij where volume strain : e V = e zz + e xx + e yy σzz
sum the 3 x, y, z stress directions
⎛ 2 ⎞ σxx
⇒ −3P = (3λ + 2μ)e V ⇒ − P = ⎜ λ + μ ⎟e V
⎝ 3 ⎠
σyy
2
where bulk modulus : Κ = λ + μ
3
Lower limits: as μ≥0 and Κ>0 gives a negative Lambda limit as λ> −2/3 μ
From theoretical rock physics based on log data Lambda is never negative (dry or wet)
Kappa, Mu and Lambda Moduli vs Porosity Kappa, Mu and Lambda Moduli vs Porosity
for 100% GAS saturation for 100% BRINE saturation
40000 40000
Log Kappa: Gas Log Kappa: Brine
35000 Layer 2: Gas 35000 Layer 2 Kappa: Brine
Log Mu Log Mu
30000 30000 Kappa Sw 1.0
Kappa gas
Mu Mu
Moduli [MPA]

Moduli [MPA]
25000 25000
Lambda Gas Lambda Sw 1.0
20000 20000

15000 15000

10000 10000

5000 5000

0 0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Porosity Porosity

Saturated moduli vs porosity: Gassmann fluid substitution, Lower Hashin-Shtrikman bound porosity
P, S Impedance vs. LambdaRho, MuRho Worldwide Log Data
Gas Sand discrimination Castagna mudrock Line and LambdaRho,MuRho threshold cutoff
Gas Sand Brine Sand or Shale -ve Poisson’s Positive Poisson’s Ratio
10 +ve Bulk Mod Positive Bulk Modulus
5
1.
S Impedance km/s.gm/cc

.5
8 Vs

MuRho G.Pa.gm/cc

=1
/
Vp 60

Vs
Vp/
6 2
Vs=
/
Vp 40
4

20 = 2
2 p/ Vs
V

0 0
0 2 4 6 8 10 12 14 -10 0 10 20 30 40 50
P Impedance km/s.gm/cc LambdaRho G.Pa.gm/cc
+ve Poisson’s
(Castagna & Smith 1994) -ve Bulk Mod
 LambdaRho vs.MuRho cross-plots for in-situ wet and gas substituted logs
In-situ wet sand log porosity
high
MuRho G.Pa.gm/cc

wet sand
cloud

les
a
sh
und
gro
ack
b
0 low
LambdaRho G.Pa.gm/cc
 LambdaRho vs.MuRho cross-plots for in-situ wet and gas substituted logs
Gas substituted log porosity
high
MuRho G.Pa.gm/cc

gas sand cloud

les
a
sh
+ve κ, und
-ve or zero λ gro
ack log for fluid substitution
b
0 low
LambdaRho G.Pa.gm/cc
Negative LambdaRho Histograms of Vp/Vs ratios comparable log
values reveal log errors (wet sand zones)
1.5 1.75 2.0
 Motivation from well logs

Alberta Cretaceous P Impedance S Impedance LambdaRho MuRho Bulk

Clastics/Carbonate facies Rho m/s. gm/cc m/s. gm/cc G.Pa. gm/cc G.Pa. gm/cc modulus
(Glauconite) Vp m/s Vs m/s gm/cc (Rayls) (Rayls) (Rayls2) (Rayls2) G.Pa. Vp/Vs Poisson Lambda/Mu
Ostracod Limestone 5464 2186 2.70 14753 5902 147.97 34.84 63.41 2.50 0.40 4.25
Ostracod Shale 2851 2115 2.30 6557 4865 -4.33 23.66 4.98 1.35 -0.11 -0.18
Channel Shale Plug 4372 2342 2.60 11367 6089 55.06 37.08 30.68 1.87 0.30 1.48
Regional Shale/Silt 4098 2186 2.55 10450 5574 47.05 31.07 26.58 1.87 0.30 1.51
Porous Channel Sand 4098 2342 2.35 9630 5504 32.16 30.29 22.28 1.75 0.26 1.06
Tight Channel Sand 4684 2623 2.55 11944 6689 53.19 44.74 32.55 1.79 0.27 1.19
AVG. % CHANGE
POROUS/TIGHT SAND 21% 19% 49% 39% 37% 2% 5% 11%
Ostracod shale has negative Lambda and Poisson ratio but positive Bulk modulus
This is an error and indicates a log problem
Motivation from well logs
Averaged log values for Shales, and Gas Sands from Alberta
Vs Vp ρ Relations to transform velocities, density
to Lamé parameters λ and μ
Shale 1290 m/s 2898 m/s 2.425 gm/cc
λ = Vp2*ρ - 2 Vs2*ρ
Gas Sand 1666 m/s 2857 m/s 2.275 gm/cc μ = Vs2*ρ
2
Vp/Vs (Vp/Vs) ν λ+2μ μ λ λ/μ
Shale 2.25 5.1 0.38 20.37 4.035 12.3 3.1
Gas Sand 1.71 2.9 0.24 18.53 6.314 5.9 0.9
% change 27 55 45 9.2 44 70 110
North Sea P-wave and Shear wave Logs
Alba Field Seismic P-P and P-S comparison
Shear wave P-wave
Vs Vp
Towed streamer
Shale Shale P-wave data

Sand
Sand
OWC OWC

Shale Shale Converted wave

P-S OBC data

From MacLeod et al 1999 Chevron/WEGCO

Petroleum industry log tracks for Sands, Shales and Carbonates
showing improved LambdaRho, MuRho crossover discrimination
of gas zones and lithologies compared to P-, S-impedance
P-, S-impedance LambdaRho, MuRho
Impedance = Velocity*Density
metres metres
P-impedance = Vp*ρ
1240 1240 Silt
S-impedance = Vs*ρ

Relations to transform impedances

1260 1260
to Lamé parameters λρ, μρ
Shale Shale
λρ = (Vp*ρ)2 - 2 (Vs*ρ)2
Gas Sand B
1280 1280 Tight Streak μρ = (Vs*ρ)2
Gas Sand A
Carbonates Carbonates
1300 1300

1320 1320
S Impedance MuRho
P Impedance LambdaRho
10000

18000
14000
2000

6000

120

160
40

80
0

(GPa.gm/cc)
(m/s.gm/cc)
Fluid,
Fluid, Porosity
Porosity &
& Lithology
Lithology directions
directions in
in LambdaRho
LambdaRho (λρ),
(λρ), MuRho (μρ) space
con
sta Dolomite

on
nt
P

rat Poiss
-im
ped
150 an

io
ce

t
tan
5%

ns
co
Sa
125

nd
Quartz
μρ (GPa.gm/cm3 or Rayls2)

s
10%
Limestone

Li Car
t h bon
Gas Sand Zone

it y
100 5% 15% 5%

ol ates
Wet Sand Zone
ros

og
Marine Shale above Gas Zone

y
10%
Po

10%
20% Tight Sandy Shale between Gas & Wet Sand Zones
75
15%
Calcareous Shale/Silt
15%
Shale
20%
50 20%
Gas Oil Water Sh
ale
Fluid
25
Porosity % bubble
20%

0 15 30 45 60 75 90 105
λρ (GPa.gm/cm3 or Rayls2)
(Adapted from Hoffe, Perez and Goodway
CSEG convention 2008)
Sensitivity of Vp/Vs, Poisson’s ratio vs. Lambda/Mu ratio
4.50 0.45
4.25
Comparison to Vp/Vs 4.00
3.75
0.40

3.50 0.35
−1
⎛ Vp ⎞
3.25
d (Vp / Vs)
= 0.5⎜ ⎟
3.00 0.30

Poisson’s ratio
d (λ / μ )

Vp/Vs ratio
2.75

⎝ Vs ⎠ 2.50
2.25
0.25

2.00 Vp/Vs Poisson ratio 0.20

1.75
Comparison to Poisson’s ratio 1.50
1.25
0.15

1.00 0.10
dν 0.75
= 0.5(1 − 2ν) 2 0.50 0.05

d (λ / μ )
0.25
0.00 0.00

0.0

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

3.6

4.0

4.4

4.8

5.2

5.6

6.0

6.4

6.8

7.2
Lambda/Mu ratio

Relative sensitivity to water saturation: “Fizz Water” (Low Gas Saturation) Discrimination
(data points from Han et al 2001)
Ip/Is, Poisson’s, Lambda/Mu ratios 170%
160% LambdaRho
and LambdaRho, “Fluid Factor” 150%
Lambda/Mu ratio
Normalized avg. % change

140%
130% Poisson ratio
120% Ip/Is ratio
110%
Fluid Factor Smith&Gidlow
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Water Saturation (Sw)

Lambda/Mu Ratio vs. LambdaRho-MuRho difference crossplots
10.0
Gas Sands

Worldwide Log Data 8.0

Shales
Brine Sands
(Castagna and Smith 1994)

ratio
La m bda /M u R a tio
6.0

Lambda/Mu
4.0

trend
linear
2.0

0.0
-10 -6 -2 2 6 10 14 18 22 26 30 34
LambdaRho-MuRho difference
LambdaRho-MuRho Difference

λ/μ ratio vs. λρ−μρ difference

with lines of constant Lambda/Mu ratio decreasing Ip

P- and Shear Impedance

(Dave MacKidd 1996) μρ
1/
∝
e
lop
s

LambdaRho-MuRho difference
Lambda/Mu ratio vs Lambda-Mu difference crossplot templates:
Lithology and Porosity discrimination Fluid discrimination
3.20 1.20

Shale 1.00

2.40 λ/μ
Carbonates
0.80

0.60
1.60

λ/μ Glauconitic Gas Sand

Glauconitic Wet Sand
0.40
-15 -10 -5 0 5 10

Sandstone line λ−μ (GPa.)

0.80 6 1 - Water Saturation Oil
Wet
0 1
Target
Porous Sand Sand Lines
4
Alberta Glauconitic Gas Sand Log
0.00
-20 0 20 40 60 λ/μ
λ−μ (GPa.) 2

(from T. Chen et al 1998) Hydrocarbons

(in yellow)
GOM log data
0
2 4 6 8 10
λ (GPa)
 Geology, Logs and Classes of AVO methods
Earth’s seismic Reflectivity Inversion
Real Earth reflection response methods methods
rock layers from log measured Seismic Trace Impedance model
(impedances) impedance contrasts of the earth

Seismic
Wavelet INVERSION
CONVOLUTION

Aki-Richards
Shuey Hampson
Smith-Gidlow-Fatti Goodway
Hilterman-Verm Connolly
Wang-Goodway
AVO Reflectivity Methods
Far Offset Near Offset
Geophones
Shots

Seismic Ray Paths

Common Reflection Point

Multiple Shots and Receivers
Offset Gather with Reflection Amplitude Variation with Offset

Increasing Offset
 Walkaway VSP Geometry AVO “Gathers” from VSP
Direct proof of AVO Near
P-P reflection
Far
200m Offset 1700m
Source Positions 0.5

Near Far 0.6

Offset Offset
0.7

0.8
3C Geophones

Pi Pj
0.9
Pi~ Pj
P-Pr Gas Sand
Rp-p=P-Pr /Pi AVO 1.0
Reflector response
1.1
Linear P-P AVO three term reflectivity equations and
approximations following Aki & Richards 1980

ΔVp ΔVs 1 ⎛ ⎞ Δρ Aki & Richards

1
(
Rpp(θ ) = 1 + tan θ
2
) −4
Vs 2
sin 2
θ + ⎜
⎜ 1 − 4
Vs 2
sin 2
θ ⎟⎟
⎠ ρ 1980
2 2
2 Vp Vp Vs 2 ⎝ Vp
Rpp(θ) = f1(θ)∗(Vp change) - f2(Vp/Vs,θ)∗(Vs change)+ f3(Vp/Vs,θ)∗(density change)

Gidlow,Smith,Fatti 1992 ignore Δρ/ρ 3rd term

2
⎛1 2
⎞
ΔIp ⎛ Vs ⎞ ΔIs ⎜ ⎛ Vs ⎞ 2 ⎟ Δρ
Rpp(θ ) = (1 + tan θ )
2
− 8⎜ ⎟ sin θ
2
− tan θ − 2⎜⎜
2
⎟⎟ sin θ
2Ip ⎜⎝ Vp ⎟⎠ 2Is ⎜ 2 ⎝ Vp ⎠ ⎟ ρ
⎝ ⎠
Rpp(θ) = f1(θ)∗(Ip change) - f2(Vp/Vs,θ)∗(Is change)+ [f(Vp/Vs,θ)∼0]∗(density change)

where the link is:

ΔIp 1 ⎛ ΔVp Δρ ⎞ ΔIs 1 ⎛ ΔVs Δρ ⎞
= Rpp(0) = ⎜⎜ + ⎟⎟ = Rss(0) = ⎜⎜ + ⎟⎟
2Ip 2 ⎝ Vp ρ ⎠ 2Is 2 ⎝ Vs ρ ⎠
Other P-P reflectivity AVO equations and approximations
General form: Rpp(θ) ≈ A + B sin2(θ) A= intercept, B=gradient

⎛ Δν ⎞ 2 ΔVp Shuey 1985

Rpp(θ) = Rpp(0) + ⎜⎜ ARpp(0) + (
⎟⎟ sin θ + tan 2 θ − sin 2 θ )
2Vp ignore ΔVp/Vp term
⎝ (1 − ν ) 2 ⎠
Δν Hilterman & Verm 1995
Rpp(θ) = Rpp(0) cos 2 θ + sin 2 θ
(1 − ν ) 2
ignore ΔVp/Vp term
NI = Rpp(0) normal incidence reflectivity, PR = Δν/(1−ν)2 Poisson reflectivity

Note ΔVp/Vp ~ 4Δρ/ρ Gardner 1974

⎛ ⎛ Vs ⎞
2 ⎞ Wiggins 1984
⎜
Rpp(θ ) = Rpp(0) + ⎜ Rpp(0) − 8⎜⎜ ⎟⎟ Rss(0) ⎟ sin 2 θ
⎟ ignore Δρ/ρ term
⎝ ⎝ Vp ⎠
⎠

Wang 1999 (Goodway 1998)

2 Δμ
⎛ ⎞
2
Δρ ΔVp Vs
Rpp(θ ) = + (1 + tan θ )
2
− 4⎜⎜ ⎟⎟ sin θ ƒ ignore Δρ/ρ term
2ρ 2Vp ⎝ Vp ⎠ 2μ ƒ wide angle gradient approximation
ƒ introduced Mu reflectivity into AVO
Comparison of 3 term Aki & Richards to 2 term approximations

0.2
0.19
0.4
0.18
0.17
0.35 Aki & Richards 3 term
0.16
0.15
Gidlow, Smith Fatti
P-P reflectioncoefficient

0.3
0.14 Shuey
coefficient

0.13
0.25
Wang
0.12
0.11 30° 45°
0.20.1
P reflection

0.09
0.15
0.08
0.07
0.06
0.1
0.05
0.04
0.05
0.03
0.02
0
0.01
0 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65
0
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Angle
Angle
Decomposition of AVO reflectivity equations into Lambda,Mu,Rho
(Gray et al 1999, Goodway 1997, 2002) Note behavior at 45º
Δρ ƒΔρ/ρ term disappears
R (θ) = 0.25 1 + tan θ
2
(
Δ(λ + 2μ )
λ + 2μ
)
− 2 sin θ
⎛ 2⎞
2 ⎜ VS ⎟ Δμ
⎜ V2 ⎟ μ
+ 0.25 1 − tan 2 θ(ρ
)
with “polarity” flip
⎝ ⎠
P
ƒΔμ/μ term has an
⎜1 − 2 VS ⎟ Δλ + (cos 2θ) ⎜ VS ⎟ Δμ + cos 2θ Δρ
1 ⎛ 2 ⎞ 2 ⎛ 2 ⎞
inflection and is zero
R (θ) =
4 cos 2 θ ⎜⎝ VP2 ⎟⎠ λ 2 cos 2 θ ⎜⎝ VP2 ⎟⎠ μ 4 cos 2 θ ρ ƒ Δλ/λ alone remains
Aki & Richards AVO reflectivity curve with separate Lambda, Mu and Rho terms
R(lambda)= -0.1 R(mu)=0.1 R(rho)=-0.03
0.016
0.014 30° 45°
0.012 Rμ(θ)
0.010
0.008
0.006
0.004
0.002
0.000
-0.002
-0.004
-0.006
Rρ(θ)
Reflectivity

-0.008
-0.010 Rfull(θ)
-0.012
-0.014
-0.016
Rλ(θ)
-0.018
-0.020
-0.022
-0.024
-0.026
-0.028
-0.030
-0.032
-0.034
-0.036
-0.038
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Angle
Log based AVO model inversion for density
(Gidlow method with error term retained in constrained AVO inversion)

45º angle requirement and polarity flip

AVO gather AVO gather AVO gather
log model inversion error
P-impedance S-impedance Density 45º
polarity flip

Base
of zone

Log track Low frequency Inversion result Density term is rejected in

background model the error plot as seen in
the polarity flip at 45º
Biot-Gassmann fluid replacement equation in Lamé terms
and relation to “Pore Space Modulus” (Hedlin, Russell, Hilterman and Lines 2003)
2
Biot-Gassmann Equation: ⎛ K ⎞
⎜⎜1 − dry ⎟⎟
K is bulk modulus, ⎝ K solid ⎠
K sat = K dry +
“sat” is saturated rock, φ is porosity ⎛ K ⎞ ⎛ φ ⎞
⎜⎜1 − φ − dry ⎟⎟(K solid )−1 + ⎜⎜ ⎟⎟
⎝ K solid ⎠ ⎝ K fluid ⎠
Biot-Gassmann Equation in Lamé terms
From μ dry = μ sat and substituting Δλ = λ sat − λ dry and ΔK = K solid − K dry
λ fluid ΔK 2
⇒ Δλ ≈ Remember λ sat can be obtained directly from AVO at 45 o
φ K solid
2

Relating “Pore Space Modulus” ρf to Biot-Gassmann Equation in Lamé terms

Pore space modulus ρf = I 2P − CIS2 = (λ sat + 2μ )ρsat − Cμρsat
⎛ VP ( dry ) ⎞
where C = ⎜
2
λ + 2μ
⎟ = dry
( )
⎜ VS( dry ) ⎟ μ
⎝ ⎠
⇒ ρf = (λ sat − λ dry )ρsat so equating Δλ from above
λ fluid ΔK 2
⇒ ρf ≈ ρsat
φ K solid
2
 AVO Classes in Rp, Rs Reflectivity Space Rutherford&Williams
0.20 Castagna
(Goodway, Hoffe, Perez 2002, 2008) Young&LoPiccolo
0.10 I
+ΔIs/Is = Rs

ΔIp/Ip(θ)
II
0.00
+0.3
III
II I Zero Gradient Line -0.10 IV
(ΔIp/Ip=2ΔIs/Is)
V
1
III pe
-0.20
Ty θ
e
ls
Decrease in Vp/Vs Fa
Grinsburg and Castagna:

S-wave velocity (km/s)

-ΔIp/Ip +ΔIp/Ip = Rp Vp(sand)=1065 +1.25 Vs
-0.3 reference pt 0.3 3 Vp(shale)=1126 +1.3 Vs
s
III Increase in Vp/Vs (velocities in m/s) sand
s
ga
background mudrock line 2
IV nd
relations (Vp=A+BVs) a
t s hale
V Background e
1 w s
Rs = 1.43 Rp d ry s
e ar ation rock
Lin rel ud ne
-Δ Is/Is nd M li
sa
-0.3 1 2 3 4 5 6
P-wave velocity (km/s)
100% φ intercepts:
gas sand (Vp≈0)
wet sand (Vp≈1500m/s) Vp=A+BVs where Vs=0
Mac Delta gas discovery: pre-drill AVO model reflectivity prediction
Rp,Rs synthetic: Brine (in-situ) vs Gas (substituted)
Brine Gas Rs vs. Rp xplots: angle range 5-40o
(in-situ) (substituted)
Rp Rs Rp Rs GAS Rotation of gas sands
away from background
Top gas sand
anomaly

et
Well: 300F4869301340

le w
17-Nov-2002 01:44:38

T-C11

ha d
/s un
0
AEC.APPLY_3
10
Base

nd gro
DepthTWTIME_GAS.DTSM_1
TWTIME_GAS.RHOB_1

gas sand
Metres 1200 US/M 200 1950 K/M3 2950

TWTIME_GAS.DT_1 AEC.VSH_3
500 US/M 100 0 V/V 1
2900.0

sa ack
anomaly
2900

T-C11
TGLU

Taglu

B
3000

Substituted T-C11 gas sand zone

T-C9
3100
T-C11 identified from blue (top) and orange
T-C8A
3200
(base) x-plot polygons of AVO anomaly

BRINE
3300

nd
3400

tre
3500

ck
m et
ro
le w
ud
ha d
/s un
nd gro
sa ck
Ba
AVO Inversion Methods
Angle Angle
Gathers Gathers
Gidlow, Smith, Fatti method Connolly method
AVO extraction of Rp, Rs Rp(θ) inversion to Ip(θ)

P & S “Reflectivity” Elastic Impedance

AI, EI
Goodway method
Rp,Rs inversion to Ip,Is

P & S Impedance
λ, μ, ρ
λρ = Ip2 - 2Is2 Elastic Impedance
μρ = Is2 extended in Lamé
λρ, μρ parameter terms
Elastic Impedance in Lamé (LMR) terms following Connolly 1999
∫ Rp
(1+ tan 2 θp ) −8 γ sin 2 θp (1− 4 γ sin 2 θp )
θ = ln Ip θ ⇒ Ip θ = Vp Vs ρ = EI
At 30° and 45° the EI (θ°) values are:

EI(30) = Vp 4 / 3 Vs −2 γ ρ (1− γ ) Density is present in all EI(θ°)

−4γ (1− 2 γ ) Density and Vs always functions of a power of γ = Vs2/Vp2
EI(45) = Vp Vs 2
ρ

Formulated in Lamé and density terms:

(1− 2 γ ) 2 cos 2 θ γ (cos 2 θ )2 cos 2 θ cos 2 θ 2 cos 2 θ
Ipθ = λ μ ρ = Lamé Impedance (LMRI)
At 0°, 30°, 45° and 60° the LMRI (θ °) values are:
LMRI(0) = λ(1− 2 γ ) / 2 μ γ ρ 1 / 2 Density independent of γ in all LMRI(θ°)
LMRI(30) = λ2 (1− 2 γ ) / 3 μ γ / 3 ρ 1 / 3 and changes sign at 45°
Density and Mu absent in LMRI (45°)
LMRI ( 45 ) = λ (1− 2 γ )
At 45° only Lambda “Pore Space Fluid Modulus”
LMRI ( 60 ) = λ2 (1− 2 γ )μ γ ρ −1 (scaled to power of γ)
Assuming meaningful LMRI(45°) LMRI(60°) following elastic parameters are extractable:
1 / 3γ
3 ⎡ LMRI ( 0 ) 2 LMRI(60) ⎤
LMRI ( 45 ) 1 / (1− 2γ ) = λ ,
LMRI(30)
= ρ, ⎢ ⎥ =μ
LMRI ( 60 ) ⎢⎣ LMRI ( 45 ) 3 ⎥⎦
AVO Classes in Reflectivity and LMR Space Rutherford&Williams
(Goodway, Hoffe, Perez 2002, 2008) 0.20 Castagna
Young&LoPiccolo
+ΔIs/Is=Rs 0.10 I

ΔIp/Ip(θ)
+0.3 II
0.00
II I Zero Gradient Line III
(ΔIp/Ip=2ΔIs/Is) -0.10 IV
1
III Ty
pe
V
l se -0.20
Decrease in Vp/Vs Fa θ

-ΔIp/Ip +ΔIp/Ip=Rp Dolomite

io
on t
reference pt

iss tan
-0.3 0.3

r at
Po ons
III Increase in Vp/Vs

C
5%

IV Quartz
Background Lih
V 10%

Porosity
Rs = 1.43 Rp μ·ρ ( Pa·kg/m3)
olo I Limestone
gy
5% 15% 5%
-Δ Is/Is
-0.3 10% II 10%
20%
15 15%
% III reference pt Shale
20%
20% IV
LMR space adds insight to V
C
P-im onsta
ped nt
AVO classes anc
e Ip
Fluid shift

λ·ρ (Pa·kg/m3)
MacKenzie Delta gas discovery
Pre-drill Fluid Replacement Log Model of Rp,Rs and LambdaRho vs. MuRho
crossplots of original wet zone in legacy well.
Note improved separation and fluid/porosity discrimination in λρ vs. μρ that is
not possible with Rp vs. Rs
100 100

101
LambdaRho vs MuRho • Shale Gas sand anomaly Rp vs. Rs
90 • Brine Sand 90 All AVO classes 1,2,3
• Gas Sand Fluid, porosity
80 Tight 80 confusion?
streaks
μρ
70 70

le
ha
Color: Volume Shale
60 60

d /s
Large consistent fluid separation

n
Rs

sa
50 discriminates oil (green oval) from gas 50

ren wet-
d
oc n d
40 40

dr o u
kt
φ

mu ckgr
g

30 30
sin

Ba
ea
cr

20 20
In

10 10
Rp
0 0
λρ
0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

40.0

45.0

50.0

55.0

60.0

-1
 Walkaway VSP
CMP Gather with Offset Balancing AVO gather
Far Offset Near Offset 200m 1700m

R2
40º
R1
Zone of 35º
Interest
30º

25º
Seismic LambdaRho,MuRho vs. P-,S-impedance crossplots and “fluid factor” stacks
P vs. S Impedance LambdaRho vs. MuRho

Cemented
Gas Sand
S Impedance (m/s. gm/cc)

Tight Sand Cemented

MuRho (Gpa. gm/cc)

Gas
Tight Sand
Sand
Shaly
Shaly
Gas
Gas Silt Shaly
Sand Gas Silt
Sand
Sand

Shale Shale
Cluster
Threshold cutoff for
Threshold cutoff for
Lambda/Mu Stack
Ip/Is “Fluid Factor” Stack
P Impedance (m/s. gm/cc) LambdaRho (GPa. gm/cc)

P-, S-impedance “fluid factor” LambdaRho, MuRho “fluid factor”

sec
0.8
Ambiguous Gas Sand Zone Isolated Gas Sand channel zone
0.9

1.0

1.1
Alberta Gas Sand A B C well A wet, wells B,C gas
Channel

Migrated stack
amplitude ambiguity unconformity
(white boxes)

LambdaRho stack
clear gas sand
anomalies
(white boxes) at wells unconformity

MuRho stack
clear sand channel layer
above unconformity
(green zones) at gas wells unconformity

Low High
 Alberta Upper/Lower Basal Quartz Sand Channel Discrimination
Log Crossplot
62 Density Overlay
2.00 Upper BQ
MuRhoG.Pa. g/cc (Rigidity)

Porosity = 25.6%
Upper BQ Permeability = 566 MD
42
2.42
Core Analysis
Lower BQ
22
2.85 Porosity = 13.8%
coals Permeability = 14.6 MD
Lower BQ
12
20 50 80 110 140 LMR lithology from crossplot polygons
LambdaRho G.Pa. g/cc (Incompressibility)
Seismic Crossplot 600
45
MuRho

700
35 Upper BQ
msec
Time

25 Lower BQ Upper BQ
800

15
0 40 80 120 160
LambdaRho 900
Lower BQ
Alberta Deep Basin: Carbonate gas/condensate discovery
75 MMCF/day (AAPG Best Paper 2005)
LambdaRho MuRho anomaly (3D Line)
LambdaRho MuRho

High

Gas/condensate
anomaly

Low
Alberta Deep Basin: Carbonate gas/condensate discovery
75 MMCF/day (AAPG Best Paper 2005)
LambdaRho Anomaly Map Vp/Vs Anomaly Map
Strong anomaly delineates Limited Vp/Vs attribute response
gas/condensate discovery

High λρ High Vp/Vs

Low λρ Low Vp/Vs

LambdaRho/MuRho Crossplot Errors from Mu
(S-impedance or velocity) Correlated Error Alone
100

80 Porous gas sand

Silt shale
60
Mu error = Mu (true) x2
MuRho

40
Mu error = Mu (true) x1/2

20

0
-25 0 25 50 75 100 125 150 175
LambdaRho
Gas sand channel seismic LMR crossplot shows clear relative isolation of gas
sand points despite inversion error skew (note -ve LambdaRho values)
Yellow oval in grey cloud mask of known gas sand points,
LMR cross-plot gas sand
clearly isolates best gas sands from background.
Correlated Rp, Rs extraction errors mean gas sand mask on λρ 3D map
detection in LMR cross-plot space is unaffected
60 13-
13-16

2.0
09-
09-17 12-
12-16

λ/μ ratio overlay

50
1.6 05-
05-16
MuRho

01-
01-17 04-
04-16
1.2
40
14-
14-09
16-
16-08
0.8 09-
09-08

11-
11-08 09-
09-08
30
0.4
08-
08-08

01-
01-08
20 0

1 km
-20 0 20 40 60
LambdaRho
Geomechanics and Moduli
0.350

Barnett shale properties (Grigg 2004)

involved in fracture prone rocks 0.300

0.27
0.250

Poison’s ratio decreases with increased 0.235

.
POISSON'S RATIO
Young’s modulus (Grigg SPE 2004)
0.200

0.150

This implies a simple increase in a

single modulus of rigidity μ.
0.100

E
= 2μ Poissons_ratio = - 0.045700 *E + 0.452206
1+ ν
0.050

-
3.00 3.50
3.9 MMpsi=27GPa
4.00 4.50 4.75 5.00 5.50
YOUNG'S MODULUS ( MMpsia)

Lambda vs. Young’s modulus cross-plot 75

overlaid with: 70

Youngs modulus (GPa)

0
ƒ curves of constant Mu 2
65
60 Brittle
4
ƒ lines of constant Poisson’s ratio 6 55
ƒ Barnett Shale trend 8
10
50
12
45
14 40
Mu (G.Pa)

Barnett Trend 16 35
18
0.1 30
20
Poisson’s ratio

0.14 25
0.18 22
24 20
0.22
0.26 26 15 Ductile
28
0.3 10
30
0.34 5
32
0.38
0.42
34 0
36
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
50
0.46
38
Lambda (GPa)
Closure Stress Equation (Engineering to Geophysics)
Fracture generation and moduli: E,ν,μ and λ

σ xx =
ν
1− ν
[σ zz − B V PP ] + B H Pp + E 2 e xx + νe yy ( )
Where 1− ν
σxx = min. horizontal closure stress to open an existing fracture
σzz = overburden stress σzz - BvPp
Pp = pore pressure acting vertically (Bv) and horizontally (Bh) σxx=σmin
σxx, exx
But horizontal strains are zero exx= eyy = 0
Uniaxial strain in ez only, as rock is bound in the earth Confining stress eyy
So Poisson’s ratio ν = - exx/ezz= 0 !
σyy = σmax
Hooke' s law for uniaxial strain
σ zz = (λ + 2μ )e zz and σ xx = λe zz
So in λ and μ terms
λ ⎡ ⎛ e2 yy - e2 xx ⎞⎤
⇒ σxx = ⎢σzz − BV PP + 2μ⎜ ⎟⎥ + BHPp
λ + 2μ ⎢⎣ ⎜ eyy ⎟⎥
⎝ ⎠⎦
Effective stress Tectonic strain energy
= Overburden – Pore Pressure % anisotropy = e2yy - e2xx
 Fluid, Porosity, Lithology, Geomechanics and Closure Stress
reduction directions in LMR Space

o
⎡ ⎛ e 2 yy - e 2 xx ⎞⎤

a ti
λ re
σ xx = ⎢ σ zz − B V P P + 2 μ ⎜ ⎟ ⎥ + B H Pp

r
du

.17 n’s
λ + 2μ ⎢⎣ ⎜ e yy ⎟⎥ ce
⎝ ⎠⎦ d

= 0 isso
150 Tectonic strain energy am E f
Effective stress pl fec

o
= Overburden – Pore Pressure % anisotropy = e2yy - e2xx ifi tiv

tP
ca e
io Str

ta n
tDolomite
n es

ns
125 s

co
Quartz 5%
MuRho (G.Pa.gm/cc)

Sand Limestone

Li
10%
100 5% 5%

th
it y

ol
ros

og
c
10% 15% 10% ct oni
Te

y
Po

75 e d t io n
Gas shales W.Canada uc
20% Calcareous shales red lifica
Barnett
15% analogue 15%
am
p

20% Carbonates
50 20%

Gas Oil Water Shale co

53 Barnett shale ns
tan
Fluid53 30 tP
30 relative EUR’s 25 - im
pe
da
Ductile shales nc
Porosity % bubble e

20%
0 15 30 45 60 75 90 105
LambdaRho (G.Pa.gm/cc) adapted from Hoffe et al
CSEG convention 2008
N. Texas Barnett Shale Fracture Generation and Anisotropy Maps
Good production high EUR Poor production low EUR
Weak non-aligned anisotropy Strong aligned anisotropy
(weak 3D attribute-green/blue tick marks) (strong 3D attribute-red/orange tick marks)
Broad fracture network Narrow fracture network

Grey/green overlay of microseismic event monitoring of fracture stimulated perm

From Y. Simon M.Sc. U of H 2005

 Lithology and geomechanical attribute mapping using LMR
(from Eric Keyser)
Log guided seismic cross plot polygons and resulting seismic section
3D line with colour overlay from cross-plot
polygons (discovery well rate 16 mmcfpd)
d y
an sit
s s oro
ga er p

Carbonates
gh
Hi
MuRho

Sands

Sands
ale
sh

Gas shales
e
e

al
ittl

sh
Br

Gas shales
ile
ct
Du

LambdaRho
Gamma log colour overlay
 Seismic AVO Gathers, 3D Azimuths and Fractures
Common Reflection Point Multiple Shots (S) and Receivers (R)
Far Offset Near Offset Far Offset

S S S S R R R R

Seismic Reflection Ray Paths

3D seismic acquisition and fractures

S R
Increasing Offset Core fracture orientation
Seismic Record of offset
gather with amplitude azimuth sectors
variation with offset Si, Ri

Isotropic AVO

600m
AVAZ:
Anisotropic AVO
AVAZ (Amplitude Variation with Azimuth): Ruger et al equation 1997
3D surface seismic geometry: Principal

an l
pl lle
multi-azimuth offsets Anisotropy Planes

c ara
e
Isotropic

ro e-p
fracture-perpendicular
layer

ot ur
symmetry plane φ=0º

pi
º i s ct
90 fra
P P(slow)
P(fast)
Sv (S2) slow

φ=
Isotropic layer Sv (S1)
P fast parallel perpendicular

Anisotropic layer Anisotropic layer

with vertical fractures
opposite sign
Rp (θ,φ) = Rp(0) + 0.5([ Δα/α − (2β/α)2 Δμ/μ] + [Δδ(v)+ 2(2β/α)2 Δγ]cos2φ) sin2θ
P-wave Isotropic (in-fracture) AVO
zero offset Anisotropic (cross-fracture) AVO terms
terms
+ 0.5(Δα/α + [Δδ(v) sin2φ cos2φ+ Δε(v)cos4φ]) tan2θ sin2θ

Anisotropy parameters:
θ = P-wave propagation angle (vertical plane)
φ=0: slow (weak) in symmetry axis (across fractures)
α, β = P- and S-wave velocity
φ=90: isotropic (parallel to fractures)
Δα/α = fractional change in P-wave velocity
Δμ/μ = fractional change in rigidity Mu Δγ = Sh-wave anisotropy ‘velocity splitting’
Δε(v) = P-wave anisotropy
Δδ(v) = P& Sv-wave anisotropy
Simplified VTI Anisotropy: Thomsen’s δ in Lamé Terms
Isotropic Modulus Tensor Matrix Anisotropic VTI Tensor Matrix
(Voigt Notation) (adapted from Sheriff’s Dictionary of Geophysics)
⎛σ xx ⎞ ⎡ C33 C33 − 2C44 C33 − 2C44 ⎤ ⎛ e xx ⎞ ⎡ C11 C11 − 2C 66 C13 ⎤
⎜ ⎟ ⎢ ⎥ ⎜e ⎟ ⎢ C − 2C ⎥
⎜σ yy ⎟ ⎢C33 − 2C44 C33 C33 − 2C44 ⎥ ⎜ yy ⎟ ⎢ 11 66 C11 C13 ⎥
⎜ σ ⎟ ⎢C − 2C C33 − 2C44 C33 ⎥ ⎜ e zz ⎟ ⎢ C 31 C 31 C 33 ⎥
⎜ zz ⎟ = ⎢ 33 44
⎥ ⋅⎜ ⎟ =⎢ ⎥⋅
⎜σ yz ⎟ ⎢ C44 ⎥ ⎜ e yz ⎟ ⎢ C 44 ⎥
⎜ ⎟ ⎢ ⎥ ⎜ e zx ⎟ ⎢ ⎥
⎜σ zx ⎟ ⎢ C44
⎥ ⎜⎜ ⎟⎟ ⎢
C 44
⎥
⎜σ ⎟ ⎢ C 66 ⎦⎥
⎝ xy ⎠ ⎣ C44 ⎦⎥ ⎝ e xy ⎠ ⎣⎢
⎡λ + 2 μ λ λ ⎤ ⎡λ ≡ +2 μ ≡ λ≡ λ13 ⎤
⎢ λ ⎥ ⎢ λ≡ λ ≡ +2 μ ≡ λ13 ⎥
⎢ λ + 2μ λ ⎥ ⎢ ⎥
⎢ λ λ λ + 2μ ⎥ ⎢ λ31 λ31 λ33 + 2 μ ⊥ ⎥
⎢ ⎥⋅ ⎢ ⎥⋅
⎢ μ ⎥ ⎢ μ⊥ ⎥
⎢ μ ⎥ ⎢ μ⊥ ⎥
⎢ ⎥ ⎢ ⎥
⎢⎣ μ ⎥⎦ ⎢⎣ μ ≡⎥⎦

Cross-section of gas So δ is generally

(C13 + C 44 ) 2 − (C 33 − C 44 ) 2 filled layers or fractures
δ= positive as
2C 33 (C 33 − C 44 )
λ13 > λ33
λ33
(λ13 2 − λ 33 2 ) + 2μ ⊥ (λ13 − λ 33 )
⇒δ= stiffer parallel to
2λ 33 + 6λ 33 μ ⊥ + 4μ ⊥
2 2
λ13 layers than across
In nearly all measurements δ
is either zero (within
measurement error)
or positive and correlated to
positive ε
ve
s iti tion
po rela
r
co

measurement
error
A better 3 term AVAZ equation: each term has equal significance
for robust inversion of Δα/α, Δμ/μ, Δγ, Δδ
Rp (θ,φ) = Δα/α + 0.5(Δα/α + [Δδ(v)+ Δη∗cos2φ]cos2φ) tan2θ
P-velocity Isotropic
zero offset (in-fracture) Anisotropic (cross-fracture) AVO terms
AVO terms
− 0.5((2β/α)2 Δμ/μ − [2(2β/α)2 Δγ − Δη∗cos2φ] cos2φ) sin2θ
Where η∗ = (ε−δ)/(1+2ε) obtained from NMO

Azimuthal Anisotropy from 3D seismic AVO over known fractures

AVO azimuth gradients
Offset vs. Azimuth ie nts
rad ular
angle cube o
g
ng ndic
r
st erpe
p

1800m offset slice

105o 285o
t

azimuth angle Fracture strike @ N10

AVO Classes for Fracture Reflectivity
Type ΔIpp/Ipp ΔIss/Iss Vpp/Vss
Zero Gradient Line
(assumes ΔIp/Ip=2ΔIs/Is) I × × Ø
+Δμ/μ Decrease in Vp/Vs
II Ú × Ø
+0.3
I III Ø Ú Ø
II
parallel to Increase in Vp/Vs IV Ø Ø Ø
fractures False
False Type
Type 11
III V Ø Ø Ø
Decrease in Vp/Vs
perpendicular
to fractures
0.20
-ΔVpp/Vpp parallel to +ΔVpp/Vpp
-0.3 0.3
fractures 0.10 I
III Increase in Vp/Vs

ΔIpp/Ipp(θ)
II
perpendicular
IV to fractures
0.00

Decrease in Vp/Vs III

V -0.10 IV
V
-0.3 -0.20
θ
Increase in Vp/Vs
-Δμ/μ (Hoffe, Perez, Goodway, 2002, 2008)
Zero Gradient Line
(assumes ΔIp/Ip=2ΔIs/Is)
Colorado B AVAZ fracture intensity and orientation map
with well A deviation path

Well A
deviation path

LOW HIGH
INTENSITY
Post-drill well A log tracks confirm 3D seismic attribute predictions
AVAZ 3D section with Well A gamma log FMI images Fracture identification: intensity and orientation
and horizons (Clrd A, Clrd B and SSPK) from FMI from crossed-dipole shear

0 M AXXENE_OV ERALL 150 0

( NONE )

0 M INXENE_OV ERALL 150

( NONE )

0 XENEDIF 50
( NONE )
Microseismic downhole monitoring
P- and S-wave events recorded
Geometry of treatment and monitor wells across an 8-level 3C geophone array

weak early strong late

fast P-wave slow S-wave

Microseismic event
frequency ~ 500hz

P- and S-waves processed to locate

events based on arrival time, azimuth
polarization

Potential to characterize fracture plane

attributes from magnitude and focal
mechanism

Events displayed in cross-section and map view

to give quantifiable estimates of Stimulated Reservoir Volume
Surface Microseismic Buried Array Event Mapping
(from Pete Smith and Kevin Smith EnCana
& Chris Neale MSI 2009)

„ 100 stations
„ 3000’ grid spacing
„ Single 3C phone cemented at 250’
Example of P-wave event 1st arrival polarity reversal

~16 sq. mi. grid

Event Map: all 8 stages top 10% of events
ƒ Moment tensor solution suggests normal dip-slip faulting
ƒ Events colored by stage and sized relative to all stages
Circle size proportional to amplitude
(red positive, blue negative)
Squares are model amplitude fit to data.
In June a massive Earthquake (magnitude 7.8)
Frac event trends twisted South Island NZ. and moved its southern
tip 30 cm, toward Australia, a process that usually
takes ‘hundreds of years‘.
Its focus occurred in “soft rocks” between two
tectonic plates, muffling its power, as the rocks
lurched (ductile) rather than snapped (brittle),
causing a low-frequency rolling rather than more
damaging high-frequency waves
Moment tensor solution fracture planes
Stage1
Stage 2
Stage 2b
Stage 3

Stage 4A,B
(bridge plug flow)

Normal dip slip fault Stage 6

2000’ Stage 7

Stage 8
Barnett shale is more brittle than Piceance silts:
High rigidity (Mu), hence lower Closure stress, results in large fracture
planes creating high magnitude, low frequency events
Magnitude vs. distance for Barnett and Pieance

Barnett

Piceance
Earthquake moments and stress drop
Moment magnitude (Mw) related to seismic moment (M0) penny shaped cracks

M 0 = μD(πr 2 ) where D = average slip, r = fracture radius

log M 0
Mw = − 10.73
1.5
∂u x D
Strain: e xx = ≈ where L is fracture length
D r
∂x L

Stress drop Δσ across circular fracture plane

D M0
Δσ = μe xx ≈ μ =
2r 2πr 3

Stress drop proportional to moment 7 M0

and inversely proportional to area
Δσ =
16 3 r
Fracture radius inversely proportional to high corner frequency
of displacement spectrum r ∝ VS where f 0 = corner frequency
2 πf 0
Earthquake amplitude displacement
spectrum (1995 Chiapas, Mexico) Fracture surface area vs. moment and stress drop

f0 1.E+06

1.E+05

Surface area (sq.km )

1.E+04

1.E+03

1.E+02

1.E+01
1E+25

1E+26

1E+27

1E+28

1E+29

1E+30

1E+31
Moment (dyn.cm)

1 bar 10 bars 100 bars 1000 bars

Mohr-Coulomb theory related to seismic LMR
5000

4500 Lambda-Mu log cross-plot (scaled by strain)

overlaid onCohesion
Y Intercept = Linear Mohr Coulomb Failure Envelopes i on)
4000
ic t
Slope = Coefficient of Internal Friction
( fr
=μ
Mu x shear strain, psi.
(shear stress τ=μ∗eyy)

3500
l ope y = 0.5655x + 299.36
s
i lure
3000 fa

2500
Unstable
brittle gas shales with carbonates
critically stressed fractures
Area
2000
Stable 1st Confining Stress
2nd Confining Stress
1500
Barnett shale Area 3rd Confining Stress
4th Confining Stress
1000 analogue EUR 53 30
ductile shales
500
in-situ stress circle
0 σmax
σmin
0 1000 2000 3000 4000 5000 6000 7000 8000
frac stimulation Lambda+2Mu x normal strain, psi.
increases pore pressure
(horizontal normal stress σ=[λ+2μ]∗exx)
 3D AVAZ with coherence & Microseismic event overlays
Seismic azimuthal anisotropy (AVAZ) intensity
Microseimic events and orientation with microseismic overlay HIGH
with AVAZ overlay Coherence
lineations

% Reflection Anisotropy
stages
1 Orientation of
2
AVAZmax
3
inconsistent
4
with SH max?
1600 m

6
areas of
7
relatively high
8
reflection AVAZ
9 anisotropy near
horizontal
10
borehole
horizontal treatment well path monitor well LOW

(courtesy of Keith Young)

Conclusions
„ Fundamental Lamé parameters of rigidity μ and incompressibility λ
control seismic wave propagation used to extract information about
rocks in the Earth.

„ These parameters link many fields in Geoscience at different scales,

from Petroleum Exploration to Earthquake Seismology.

„ Common contradictory formulations are simplified by restating

equations in Lamé parameters e.g. incompressibility Lambda is less
confusing than Bulk Modulus in revealing log errors.

„ Improved petrophysical discrimination and insight linking rock reservoir

properties, pore fluid parameters and their seismic responses.

„ Insight into angle dependent reflectivity and impedance equations

reveals the significance of 45º.

„ More descriptive sensitivity in AVO crossplotting using inverted λρ, μρ

seismic attributes to isolate effects of lithology, porosity and fluids.
Conclusions
„ Thomsen’s delta involved in describing VTI and HTI anisotropy is a
function of the difference between diagonal and off-diagonal lambda
despite Thomsen’s claim that “…lambda has no common name, since it
is not useful for much in geophysics…” (DISC 2002)

„ More intuitive and appropriate for in-situ Geomechanical Closure Stress

description of rocks and fractures (brittle vs. ductile deformation) than
Young’s modulus and Poisson ratio.

„ Successful application of seismic attributes in predicting “sweet spots”

for unconventional gas shale drilling that intersected predicted fractures
and Stimulated Perm as confirmed by Microseismic monitoring.

„ Potential to extract Microseismic event attributes using Moment tensor

solutions from Earthquake Seismology for stress drop on fracture
planes created through well completions.
Acknowledgements
Dave Mackidd, Dave Cooper, John Varsek, Christian Abaco
Eric Keyser, John Parkin, Taiwen Chen, Weimin Zhang,
Keith Young, Pete Smith, Kevin Smith, Dan Potocki (EnCana)

Marco Perez (Apache Canada)

Brian Hoffe (Shell Canada)

Dave Gray, Jon Downton (CGGVeritas)

Peter Duncan, Chris Neale (Microseismic Inc.)

Shawn Maxwell (Schlumberger)