Seismic inversion

Sonja Maultzsch

Seismic inversion course, UPPA, 2008

Course Outline
General aspects of seismic inversion
ƒ Concept and purpose
ƒ Petrophysical basis
ƒ Data requirements
ƒ Overview of methods and basic inversion process

Poststack inversion – practical workflow

ƒ Data QC
ƒ Wavelet extraction
ƒ A priori model building
ƒ Inversion parameterization and QC of results

Prestack inversion – practical workflow

ƒ Data QC
ƒ Wavelet extraction
ƒ A priori model building
ƒ Inversion parameterization and QC of results

Advanced inversion techniques

ƒ Geostatistical inversion
ƒ Joint PP/PS inversion
ƒ 4D inversion

2 - Seismic inversion course, UPPA, 2008

 General aspects of seismic inversion

Concept and purpose

Petrophysical basis
Data requirements
Overview of methods and basic
inversion process

3 - Seismic inversion course, UPPA, 2008

Scope of seismic inversion
SEISMIC ACQUISITION
SEISMIC ACQUISITION and
and PROCESSING
PROCESSING
Hypothesis:
Earth
Earth ** Wavelet
Wavelet = = Seismic
Seismic After processing, the seismic trace is the
BANG
result of a convolution between a wavelet
BANG
and reflectivities at interfaces that result
from contrasts in impedance.

SEISMIC INVERSION Objective:

Seismic / Wavelet = Impedance Compute an impedance model of the
subsurface from seismic amplitudes
Via
An inverse process using a wavelet
estimated from the seismic data and
calibrated at wells.

RESERVOIR CHARACTERIZATION
Benefits:
Impedance Earth Model The impedances are a natural link between
petrophysical characteristics of the
subsurface and the seismic data.

4 - Seismic inversion course, UPPA, 2008

Inversion hypothesis: convolutional model

Seismic Modelling

Acoustic Impedance Zero offset

impedance Log Reflectivity Seismic trace

« derivation »

Ri=(Ei+1-Ei)/(Ei+1+Ei)
convolution

* + =
wavelet noise
« integration »

Ei+1 = Ei(1+Ri)/(1+Ri)

Seismic Inversion

5 - Seismic inversion course, UPPA, 2008

Reflectivity versus Impedance
From an interface to a layer property
Seismic 'sees' subsurface in terms of interfaces between geological strata

Impedance "sees" the subsurface in terms of layers

Impedances: properties of the rock relation to other petrophysical parameters of interest,

such as porosity or fluid saturation
Increase in resolution by inversion for impedances

6 - Seismic inversion course, UPPA, 2008

 From impedances to reservoir characteristics:
Lithology and fluid discrimination
35

30

25
Porosity %

20

15

10

0
6000 7000 8000 9000 10000 11000 12000 13000 14000 15000 16000

Acoustic Impedance

Empirical relationships between

impedances and petrophysical
parameters established from well data
(cross-plotting)
Relationships must be valid when
upscaled into the seismic bandwidth

7 - Seismic inversion course, UPPA, 2008

Definition of Poisson’s ratio

2
1 ⎛ Vp ⎞ Important parameter that contains P-wave
⎜ ⎟ −1 and shear-wave information
2 ⎝ Vs ⎠
ν= 2 Sensitive to fluid saturation and lithology
⎛ ⎞
Vp
⎜ ⎟ −1
⎝ Vs ⎠

ε xx
ν =−
ε zz
Ratio of lateral to axial strain z
in a uniaxial stress state

8 - Seismic inversion course, UPPA, 2008

Petrophysics in shallow siliciclastics

Discrimination between the

various lithologies/fluid
content requires two
petroelastic parameters:
P impedance and Poisson’s
ratio
Water
2
1 ⎛ Vp ⎞
⎜ ⎟ −1
2 ⎝ Vs ⎠
Shales ν =
Sands
2
⎛ Vp ⎞
⎟ −1
Poisson’s ratio

⎜
⎝ Vs ⎠
Brine
HC
Poisson’s ratio or S
Sandstones
impedance can be obtained
Class III II I by inverting pre-stack data
P-wave Impedance at several angles of
Sa ist

incidence (sub-stacks):
nd og
h

/ s ram

Pre-stack inversion
ha s
le

9 - Seismic inversion course, UPPA, 2008

Reflectivity as a function of Incidence Angle
(made simple / moderate angle situation ~40 deg.)
Acoustic “Elastic”
cos²(θ) x + sin²(θ) x
contrast contrast
Reflectivity

¾At Near offset / angle, the reflectivity is dominated

Near Mid Far UFar by the “acoustic” contrast
Angle of Incidence: θ
( a function of offset) ¾At any other offset / angle, the reflectivity is the
result of an “acoustic” and an “elastic” contrast,
whose net effect depends on angle and the relative
strengths and polarities
Class I
Reflectivity

¾The “elastic” component can only be recovered by

combining reflectivities at different angles
Class II

¾This is possible only if angles / offsets are large

Class III enough and when seismic data quality is sufficient

10 - Seismic inversion course, UPPA, 2008

 Petroelastic study
Reflectivity modelling for different fluids
PR vs. IP Crossplot
Well: ACA-001 ACA-002 ACA-002_G1
Initial situation 0
15469

0
15469

10.0

11.0
0
5.0

6.0

7.0

8.0

9.0
0.500 0.500

0.450 0.450

0.400 0.400

0.350 0.350
PR initial situation

0.300 0.300

0.250 0.250

0.200 0.200

0.150 0.150

0.100 0.100

0.050 0.050

0.000 0.000
10.0

11.0
5.0

6.0

7.0

8.0

9.0

IP initial situation

Well Legend:
ACA-001 ACA-002 ACA-002_G1

In situ Gas Oil Water

Petro-elastic study on actual logs AVO modelling from in-situ & fluid replaced logs

⇒ Is it worth carrying-out an inversion on a specific area?

⇒ What type of inversion should be performed?

11 - Seismic inversion course, UPPA, 2008

 Fluid substitution: Gassmann equations

K sat K dry K fl
= + μ sat = μ dry
K 0 − K sat K 0 − K dry φ (K 0 − K fl )

K sat = Saturated Rock bulk modulus

K 0 = Mineral bulk modulus
μ dry = Dry Rock shear modulus
K fl = Fluid bulk modulus
K dry = Dry Rock bulk modulus
Φ = Porosity

Assumptions:
ƒ Pore space completely connected, statistically isotropic
ƒ Frequency is sufficiently low that wave-induced pressure gradients can equilibrate

12 - Seismic inversion course, UPPA, 2008

 Benefits of seismic inversion

Layer property (geology-petrophysics)

ƒ Related more easily to geology / petrophysical properties
ƒ Mapping of reservoir / flow units

Integration of various data sources to increase resolution

ƒ Seismic data, log data and a-priori geological knowledge are put together to obtain a
broad bandwidth impedance result

Reservoir characterisation
ƒ Lithology indicator
ƒ Porosity indicator
ƒ Fluid indicator

13 - Seismic inversion course, UPPA, 2008

 When is an inversion useful?

Exploration stage: to help the interpretation, the understanding of

depositional systems or to do some quick reserve estimates
ƒ There are several « quick and dirty » inversion techniques available which
can provide 80% type of answer

Reservoir characterisation: mapping of rock properties, lithologies,

fluid saturations
ƒ Requires feasibility study
ƒ Understand, which parameters the impedances are sensitive to
ƒ Data QC
ƒ Determine, which type of inversion is required

14 - Seismic inversion course, UPPA, 2008

 Feasibility study

15 - Seismic inversion course, UPPA, 2008

Well data QC and Cross-plotting

Seismic scale

Inverted Is
8
7
6

L ING 5
A 4
SC
UP filtered log scale 3
2
1
EI filtered
log scale Inverted Ip Facies
group
EI

AI filtered

AI
16 - Seismic inversion course, UPPA, 2008
 Basic inversion process

Inverse problem

Optimal impedance model (m)

Seismic data (dobs) Forward problem:

g
Approximation using a
Inverse problem:
1D convolutional model
There is no simple inverse function g-1 dsynth= g(m)
that allows to go from dobs to the optimal
model m
Ædobs are insufficient (limited bandwidth)
and not exact (noise…)
ÆThe model m and the forward function
g are too simplistic
Synthetic data (dsynth)
ÆThe inverse problem has to be solved
iteratively and it has no unique solution

Iterative search of the optimal impedance model

Convergence towards the optimal model is driven by a cost (objective) function
The space of possible models can be reduced by constraining the inversion
17 - Seismic inversion course, UPPA, 2008
 Various inversion methods
Recursive: integration of seismic trace => pseudo-impedance

Coloured Inversion: convolution by operator from well to broaden spectrum =>

pseudo-impedance (BP)

These two are only for qualitative interpretation purposes: no low frequency part,
problem with thick layers.

Sparse Spike: algorithm to find an impedance model matching the seismic bandwith,
but with the smallest number of non zero reflectivities (limited number of
homogeneous layers)

Model-based: uses an initial low-frequency impedance model which is perturbed

Layer based: similar to model based, except that the model is layered and inverted
both for thicknesses of layers and impedances

Stochastic: multiple realizations of impedance models that can reproduce the seismic
amplitudes within a given error

The last four can be post-stack (inversion of a full stack) or pre-stack (inversion of
several sub-stacks)

18 - Seismic inversion course, UPPA, 2008

 Iterative scheme of model-based inversion

Interpretation
Structural model
Wavelet

Adjust L.F. AI Log

model
Synthetic seismic A priori model

Estimation of
cost function
Seismic Data

no Error
Data
Computation ?

Yes

Results Résiduals

Absolute AI Residuals

19 - Seismic inversion course, UPPA, 2008

Constraining the inversion with a-priori information
ÆBrings additional information which is not present in seismic data (well data, geological knowledge)
ÆConstrains the inversion towards a solution that takes into account the different types of information and the data quality

Information on a-priori model: Information on seismic data:

-distribution of reflectivities (ex: sparse spike inversion) -S/N ratio
-low frequency trend of impedances (well logs) Æstandard deviation of amplitudes
-Standard deviation of impedances from initial model
-Vertical and lateral variability of the impedances
Geological information:
-body geometry
-2D variograms for facies distribution

20 - Seismic inversion course, UPPA, 2008

 Required input data for an inversion

Well logs: Vp, Vs, Rhob, Phie, caliper, GR, ….

Check-shots, VSP or T=f (depth)

Seismic data: full stack, angle or offset stacks,

stacking or migration velocities

Smoothed seismic interpretation (horizons, faults),

geological tops

21 - Seismic inversion course, UPPA, 2008

 Frequency content of inversion schemes
RTI: DC from baseline
Recursive Trace (RTI)

Amplitude
TRACE BASED

20 40 60 80 Frequency (Hz)

CSSI: Trend model & Constraints Constrained Sparse Spike (CSSI)

(Invertace+)
Amplitude

20 40 60 80 Frequency (Hz)
MODEL BASED

InverMod & StatMod: Model & Constraints

Model based Inversion (Invermod)
Geostatistical Inversion (Statmod)
Amplitude

20 40 60 80 Frequency (Hz)

Courtesy P. Mesdag (Jason)

22 - Seismic inversion course, UPPA, 2008

 Summary – Quizz 1

Give 3 good reasons for carrying out an inversion

Improve resolution, get volume attribute, obtain a petrophysical
indicator

What is the use of a feasibility study?

Know if inversion will provide useful attributes for reservoir
characterisation; assess quality of seismic data, well data, and seismic
to well ties; determine most suitable inversion method
Which approximation is the forward function in a seismic inversion
based on?
1D convolutional model
Is the inversion for impedance unique?
No, therefore stochastic methods have been developed
What is the purpose of the a-priori model?
Provide low-frequency part, incorporate geological information

23 - Seismic inversion course, UPPA, 2008