Академический Документы
Профессиональный Документы
Культура Документы
Depth GEOLOGICAL
SEISMIC SOURCE
Km STRUCTURE
Courtesy GX Technology
Forward & Inverse Seismic Problem
START END
ACQUISITION Structure
SYSTEMS Velocity
Rock, Fluid, Lithology
t
MODELLING
WAVELET
INVERSION
RECOVER
GEOLOGY
EARTH PROPOGATION
EFFECT
t
* DECONVOLVE
CONVOLVE
*
EARTH MODEL
• Vp, Vs, ρ
• Porosity SEISMIC RESPONSE
• Structure
• Saturation
Seismic Trace
MODEL ING
I NV E R S I O N
Well AI spectrum
Seismic Bandwidth
0 20 40 60 80
Frequency (Hz)
Impedance Information
Band-Limited vs Full Bandwidth
Importance of Trend Model
Benefits of Inversion vs Reflection
Data
1. P-Imp is a Layer Property
2. Removal of Artifacts
3. Enhanced Resolution
5. Better Drilling Decisisons
Full Stack or Partial Stack Inversion?
Amplitude Versus Angle (AVA)
Simultaneous Inversion of AVO
Seismic Data
Simultaneous Inversion Benefits
Resolution of Full Stack Vs
Simultaneous
Lithology Separation with
Simultaneous Inversion
Inversion history
Early inversions were limited to post-
stack data.
In recent years, inversion has been
combined with pre-stack AVO analysis,
known as Simultaneous Inversion.
Current inversion technology has
shifted attention to the quality of the
input seismic data, and the model
building.
Seismic Inversion Methods
Impedance Displays
Post-stack seismic inversion transforms an
input seismic volume into a volume of
acoustic impedance.
General Forward Model for
Inversion
Reflectivity Seismic
Notes
There are no multiples modeled.
Transmission loss and geometric spreading
are ignored.
Frequency-dependent absorption is ignored.
The wavelet may be time varying.
General Forward Model for
Inversion
The effect of
convolving the
wavelet with the
reflectivity is to
remove much of the
high frequency detail:
Convolution in the
time domain is
multiplication in the
frequency domain.
As we can see from
these plots, the effect
of the wavelet is to
remove both high and
low frequencies from
the trace spectrum.
Theoretically, inversion
attempts to recover
these lost regions.
Non-Uniqueness in Inversion
inversions
Model Based Inversion Flow
Load wells
Gather necessary inputs: Import seismic
Import or pick seismic horizons
Inversion Analysis: Use well logs and horizons to replace the low
frequencies not recorded by the seismic.
Over a limited
frequency range, the
phase spectrum may
often be approximated
by a straight line.
These wavelets both have the same amplitude
spectrum, but different (constant) phase spectra:
Effect different wavelet on seismic
Wavelets in the earth vary in space (including by
offset/angle) and in time for a several reasons:
◦ Near surface effects
◦ Frequency-dependent absorption i.e. decreasing
bandwidth with increasing depth
◦ Inter-bed multiples
◦ NMO stretch
◦ Offset or angle
Wavelets are used for 4 types of processes:
◦ Creating synthetics.
◦ Log correlation.
◦ Inversion.
◦ Inserted items in seismic displays.
Wavelet is very important
Procedure:
◦ Use statistical wavelet extraction to determine
a preliminary wavelet.
◦ Stretch/squeeze the logs to tie the seismic data.
◦ Extract a new wavelet using the well logs.
◦ Possibly repeat the previous two steps.
P-impedance log
QC inversion result
To evaluate the
accuracy:
1. The fit (or error)
between the real
well impedance in
blue and the
inversion
impedance in red.
2. The difference
between the real
seismic and the
synthetic seismic
that is created from
the inversion
impedance.
QC using well log data
Iteration Number
Sparse Spike inversion
Maximum Likelihood Sparse Spike Inversion builds up the reflectivity sequence one
spike at a time. Spikes are added until the trace is modeled accurately enough.
The amplitudes of the impedance blocks are determined using the Model Based
Inversion algorithm; with the locations of the ‘blocks’ determined by MLSS.
Spike Detection Threshold
As each spike is added, its amplitude is compared with the average
amplitude of all spikes detected so far. When the new amplitude is less
than a specified fraction of the average, the algorithm stops adding spikes
Inversion result comparison
Sparse Spike Inversion produces a result which is
similar to Model Based Inversion.
A difference is that the very thin layers may be missing or
discontinuous
Colored Inversion
Colored Inversion is a modification of Recursive
Inversion, which was originally described by Lancaster
and Whitcombe of BP at the 2000 SEG Convention.
In this process, there is a single operator, O, which
is applied to the seismic trace S to transform it directly
into the inversion result Z:
Well spectrum
Seismic spectrum
Operator Result
Inversion result
Issues in Colored Inversion:
No dependence on an initial model, but
needs a well to determine the impedance-
frequency trend.
Very fast to apply.
Assumes the data is zero-phase.
Produces a result similar to Recursive
Inversion, but with higher frequency
content and better scaling.