Upload
others
View
14
Download
0
Embed Size (px)
Citation preview
Patrick Connolly Associates Ltd
Blueback Workflows for Seismic Reservoir
Characterization
written and presented by Patrick Connolly and supported by Cegal
This course will provide participants with the skills needed to design and implement
workflows for seismic reservoir characterization based on the Cegal Blueback suite of QI
applications. The course covers both the theory and practice of coloured inversion, AVO
including elastic and extended elastic impedance, DHIs, seismic net pay and inversion including
the new stochastic application ODiSI.
This is a classroom based course using examples, laptop based exercises and discussion
plus the opportunity to allow participants to immediately apply the theory to real data examples
within the Petrel environment using the Blueback Seismic Reservoir Characterization, ODiSI,
Rock Physics and Investigator plug-ins. The course can also include a workshop session to
discuss your own projects and data.
The core components of the course form an integrated workflow and hence we
recommend a minimum of three days to cover the key elements with five days required for a full
development of all the elements.
The course will be delivered at an intermediate level. Participants should have a basic
knowledge of the seismic method, including acquisition and processing with a minimum of three
years working with seismic data. However, the subject matter of this course, AVO and inversion,
is covered from basic principles. A complete copy of the slides is provided to each participant.
Summary
The primary objective of the course is to help make your seismic more like geology. This is
achieved through a number of conceptually simple steps;
1. Transforming from reflectivity to relative impedance and improving resolution (Coloured
Inversion)
2. Coordinate rotating intercept and gradient data to optimize correlation with desired
reservoir properties (Extended Elastic Impedance)
3. Removing tuning effects and calibrating amplitudes (Seismic Net Pay), or
4. Inverting optimized relative impedances to reservoir properties.
This last step will be illustrated using the new stochastic inversion application ODiSI that
performs the inversion by matching the seismic data to very large numbers of pseudo-wells.
The key message is that, rather than putting the seismic data into one complicated application,
it is better to proceed through a series of simpler, comprehendible steps. The user can then
develop an intuitive feel for the relationship between the input and output and thereby optimize
the results from each component of the workflow. This leads to an appreciation of the level of
confidence that can be placed in the final results.
Judgment by eye is critical. There is no way to calculate optimum parameters and there is no
way to measure the uncertainty of the final results; there are always too many unknowns. For
this reason it is always best to apply transformations to the seismic rather than, say, analysing
cross-plots of extracted seismic measurements where context is lost. Staying within the seismic
domain provides broader geological context allowing the user to best exercise judgement.
This course provides the theory behind each step of the workflow, reinforced with Excel
exercises, and discusses practical aspects of implementation with participants able to apply
ideas to real data within the Petrel environment using Cegal plug-ins.
Coloured Inversion
• The theory behind the characteristic form of impedance and reflectivity spectra
based on the statistics of bed thickness distributions.
• The advantages of relative impedance over reflectivity.
• Maximizing bandwidth and optimizing wavelets; practical aspects of Coloured
Inversion and Blueing.
• Stabilizing the low frequencies though frequency slice filtering and structurally
conformable filtering.
• Controls on resolution; stratigraphic filtering, Q and ghosts.
Exercises
• Excel: 1D geological modelling with different bed-thickness distributions
• Excel: Wavelet modelling
• Petrel: Dynamically design a CI operator and run the CI processes
• Petrel: Frequency Spectra Analysis
AVO
• AVO theory; Zoeppritz equations and linearizations; Aki &
Richards, Wiggins, Fatti, etc.
• Measuring AVO, measurement errors, what can be reliably
measured and what are the implications of errors.
• Intercept-gradient cross-plots, coordinate rotations and the
relationship with elastic moduli. Fluid and lithology projections.
• AVO in the impedance domain; Elastic Impedance and
Extended Elastic Impedance
• Cross-plot domains and rock physics templates
• Optimizing the correlation of seismic with reservoir properties
Exercises
• Excel: Half-space modelling with Wiggins, Dong, Fatti and Rüger equations
• Excel: AVO measurements and associated uncertainties.
• Excel: Intercept-gradient coordinate rotations
• Excel: Optimizing chi-angle stacks for different scenarios
• Petrel: Dynamic EEI analysis of well data
• Petrel: Dynamic EEI analysis of seismic data
• Petrel: AIGI Cross plotting
• Petrel: Designing rock physics templates
Seismic Net Pay
• Bandwidth and the causes and nature of impedance and
reflectivity tuning
• Detuning and calibration and the meaning of seismic net-to-
gross.
• Understanding the assumptions, uncertainties and practical
aspects of the SNP method
• Examples of application to conventional and unconventional
reservoirs
• Well ties, wavelet estimation and the limits of seismic calibration
Exercises
• Excel: Impedance and reflectivity tuning
• Excel: The effects and limitations of bandwidth on detuning
• Petrel: Seismic Net Pay Exercise
Seismic Inversion
• DHI’s, exploration risking and Bayes theorem
• The risks of using statistically driven attribute
relationships for reservoir characterisation
• Deterministic and stochastic inversion. The risk of bias
from ignoring uncertainties (especially from low
frequencies models)
• Bayes theorem with probability density functions.
• Sources of uncertainty, two-step inversion and partial
stochastic methods.
• Seismic uncertainty and the limits of formal approaches.
• Inversion methods; optimizing and sampling algorithms. Spatial correlation.
• Inversion landscape; categorising inversion applications. A review of published
methods. Criteria for selecting the appropriate method
• One Dimensional Stochastic Inversion; ODiSI. Theory and practice.
• Workflows for Seismic Reservoir Characterisation.
Exercises
• Excel: Spurious correlation from multi-attributes
• Excel: Uncertainty and How to Find the Answer to Anything
• Excel: The limits of optimisation algorithms; inversion by spreadsheet
• Petrel: full parameterisation and execution of ODiSI plug-in for Petrel.
Image Gallery: Coloured Inversion
Optimise wavelets to improve resolution
Petrel: Dynamically design and apply a coloured inversion operator
phase 39
F1 8 F0 0 0.00 weight r 0.8
F2 12 F1 10 0.00 weight i 0.6
F3 50 F2 30 0.00
F4 70 F3 70 0.00
F4 120 0.00
F5 500 0.00
Frequency Hermitian real Hilbert imag noise complex inverse FFT time (ms) real imaginary modulus dB
0.0 0.00 0.00 1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 8.06051587301237E-004 -512 0.00 0.00 0.00 0.00 -45
1.0 0.00 0.00 -1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 7.93396687531645E-004+9.89055656230725E-005i-511 0.00 0.00 0.00 0.00 -45
2.0 0.00 0.00 1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 7.57238080406619E-004+1.88911300385758E-004i-510 0.00 0.00 0.00 0.00 -45
2.9 0.00 0.00 -1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 7.02682663089593E-004+2.62570120613537E-004i-509 0.00 0.00 0.00 0.00 -45
3.9 0.00 0.00 1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 6.37260599603784E-004+3.1508234276717E-004i-508 0.00 0.00 0.00 0.00 -45
4.9 0.00 0.00 -1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 5.69640837147972E-004+3.45021811761659E-004i-507 0.00 0.00 0.00 0.00 -45
5.9 0.00 0.00 1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 5.08171754287981E-004+3.54469407386315E-004i-506 0.00 0.00 0.00 0.00 -45
6.8 0.00 0.00 -1.0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0 4.59508622356945E-004+3.4853856261613E-004i-505 0.00 0.00 0.00 0.00 -45
7.8 0.00 0.07 1.0 0.67 1.33 0.00 0.00 0.00 0.00 1.33 0.00 1.333333333333334.27570599068686E-004+3.34388113813699E-004i-504 0.00 0.00 0.00 0.00 -24
8.8 0.20 0.20 -1.0 -2.00 -4.00 0.00 0.00 0.00 0.00 -4.00 0.00 -4 4.13006328316395E-004+3.19909725942369E-004i-503 0.00 0.00 0.00 0.00 -14
9.8 0.40 0.40 1.0 4.00 8.00 0.00 0.00 0.00 0.00 8.00 0.00 8 4.13251191561936E-004+3.12333240866047E-004i-502 0.00 0.00 0.00 0.00 -8
10.7 0.60 0.60 -1.0 -6.00 -12.00 0.00 0.00 0.00 0.00 -12.00 0.00 -12 4.23149179862914E-004+3.17003417625306E-004i-501 0.00 0.00 0.00 0.00 -4
11.7 0.80 0.80 1.0 8.00 16.00 0.00 0.00 0.00 0.00 16.00 0.00 16 4.36009080157368E-004+3.36544289860422E-004i-500 0.00 0.00 0.00 0.00 -2
12.7 1.00 0.93 -1.0 -9.33 -18.67 0.00 0.00 0.00 0.00 -18.67 0.00 -18.66666666666674.44887568053942E-004+3.70550369978032E-004i-499 0.00 0.00 0.00 0.00 -1
13.7 1.00 1.00 1.0 10.00 20.00 0.00 0.00 0.00 0.00 20.00 0.00 20 4.43855083581524E-004+4.15842272661938E-004i-498 0.00 0.00 0.00 0.00 0
14.6 1.00 1.00 -1.0 -10.00 -20.00 0.00 0.00 0.00 0.00 -20.00 0.00 -20 4.29010409439018E-004+4.67217639959402E-004i-497 0.00 0.00 0.00 0.00 0
15.6 1.00 1.00 1.0 10.00 20.00 0.00 0.00 0.00 0.00 20.00 0.00 20 3.99064051336093E-004+5.18536898655938E-004i-496 0.00 0.00 0.00 0.00 0
16.6 1.00 1.00 -1.0 -10.00 -20.00 0.00 0.00 0.00 0.00 -20.00 0.00 -20 3.55397798241397E-004+5.63924571469095E-004i-495 0.00 0.00 0.00 0.00 0
17.6 1.00 1.00 1.0 10.00 20.00 0.00 0.00 0.00 0.00 20.00 0.00 20 3.01610954815311E-004+5.98851353317123E-004i-494 0.00 0.00 0.00 0.00 0
18.6 1.00 1.00 -1.0 -10.00 -20.00 0.00 0.00 0.00 0.00 -20.00 0.00 -20 2.42662582187709E-004+6.20892247706556E-004i-493 0.00 0.00 0.00 0.00 0
19.5 1.00 1.00 1.0 10.00 20.00 0.00 0.00 0.00 0.00 20.00 0.00 20 1.83794609483867E-004+6.30025201304385E-004i-492 0.00 0.00 0.00 0.00 0
noisewavelet
constant
phase
phase
noise
smoothed
noise
real +
noise
imag +
noise
amplitude
spectrum
smooth
spectrum
phase rotation
calculate
-180
-120
-60
0
60
120
180
0.0
0.5
1.0
1.5
2.0
0 20 40 60 80 100 120
amp
litu
de
frequency (Hz)
amplitude spectrum
wavelet
noiselevel
-200 -150 -100 -50 0 50 100 150 200
time (ms)
wavelet - zero phase & envelope
Petrel: Analyse the frequency spectra
Image Gallery: AVO
Half-space and gather modelling to understand the controls on AVO response and errors in
measurement
AIGI coordinate rotation to optimise impedance contrasts
Vp Vs rho K m A 0.057 0.056 1.8 0.001 0.040
t0 2.00 layer 1 2500 1200 2.25 9,742,500 3,240,000 B -0.381 -0.351 7.9 -0.030 plot bias x 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200 4400 4600 4800 5000
vi 3000 layer 2 3032 1857 2.08 9,557,791 7,172,774 C 0.096 measured % error difference 1000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
vrms 2800 1001 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
(Vs/Vp)^2 0.305 1002 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
-0.022 0.056 1003 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
DVp Dvs Drho -0.351 0.136 1004 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
532 657 -0.17 1005 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
x sin2 q q tan2 q R3 R2 1006 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
200 0.00 2.2 0.00 0.056 0.056 1332 1457 13 1007 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
400 0.01 4.4 0.01 0.055 0.055 1008 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
600 0.01 6.6 0.01 0.052 0.052 1009 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
800 0.02 8.7 0.02 0.048 0.048 1010 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1000 0.04 10.9 0.04 0.044 0.043 1011 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1200 0.05 13.0 0.05 0.038 0.038 1012 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1400 0.07 15.1 0.07 0.032 0.031 1013 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1600 0.09 17.1 0.09 0.025 0.024 1014 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1800 0.11 19.1 0.12 0.017 0.016 1015 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2000 0.13 21.1 0.15 0.009 0.007 1016 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2200 0.15 23.1 0.18 0.001 -0.002 1017 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2400 0.18 25.0 0.22 -0.007 -0.011 1018 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2600 0.20 26.8 0.26 -0.016 -0.021 1019 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2800 0.23 28.6 0.30 -0.024 -0.031 1020 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3000 0.26 30.4 0.34 -0.032 -0.041 1021 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3200 0.28 32.1 0.39 -0.040 -0.051 0 1022 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3400 0.31 33.8 0.45 -0.048 -0.061 0 1023 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3600 0.34 35.4 0.51 -0.055 -0.071 1024 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3800 0.36 37.0 0.57 -0.061 -0.081 1025 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4000 0.39 38.5 0.63 -0.067 -0.091 1026 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4200 0.41 40.0 0.70 -0.073 -0.101 1027 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4400 0.44 41.4 0.78 -0.077 -0.110 1028 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4600 0.46 42.8 0.86 -0.081 -0.119 1029 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
4800 0.49 44.2 0.95 -0.084 -0.128 1030 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
5000 0.51 45.5 1.04 -0.086 -0.137 1031 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1032 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1033 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1034 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1035 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1036 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1037 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1038 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1039 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1040 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1041 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1042 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1043 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1044 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1045 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1046 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1047 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
% angle error
time/depth reflectivity
% scaling error (at 5000m)-80
-60
-40
-20
0
20
40
60
80
-0.20
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0 1000 2000 3000 4000 5000 thet
a
rfc
offset (m)
Wiggins equation
2-term
3-term
theta
This worksheet calculates AVO curves using the
Wiggins equation.
Delta and r can be adjusted using the slider bars.
Copyright Patrick Connolly Associates Ltd. 2015 .
Incidence angles are estimated from offset using
the equation;
where = offset, = interval velocity, = RMS velocity and = zero offset TWT. Values may be changed by the
user.
1800
1850
1900
1950
2000
2050
2100
2150
2200
Gather
1800
1850
1900
1950
2000
2050
2100
2150
2200
-0.05 0.00 0.05 0.10
Intercept
1800
1850
1900
1950
2000
2050
2100
2150
2200
-0.5 0.0 0.5
Gradient
1800
1850
1900
1950
2000
2050
2100
2150
2200
-0.02 0.00 0.02
Stack
q 0 AI0 2100
ln(AI0) 7.65
c 0 -180
cos c 1.00 50.00 -50.00
sin c 0.00 0.00 0.00
AI 3300 EEI 3300
GI 2500lnAI 0.45
lnGI 0.17
phi 21.09
AB 0.48
P 0.45
x 0.45 0.45
y 0.00 0.17
AI 2700 EEI 2700
GI 2300lnAI 0.25
lnGI 0.09
phi2 19.90
A2B2 0.27
P2 0.25
x2 0.25 0.25
y2 0.00 0.09
AI 3050 EEI 3050
GI 2800lnAI 0.37
lnGI 0.29
phi3 37.63
A3B3 0.47
P3 0.37
x3 0.37 0.37
y3 0.00 0.29
shale EEI 3050 point point point
brine EEI 3300 -250 350 600
brine
sand
oil sand
shale
3050
3300
2700
3050
1000 1500 2000 2500 3000 3500 4000
EEI
brine
shale
oil
shale
-0.2
0.0
0.2
0.4
0.6
-0.2 0.0 0.2 0.4 0.6
lnG
I
lnAI
projection
brine sand
oil sand
shale
EEI(c) = AI0.(AI/AI0)cosc.(GI/AI0)
sinc
This worksheet demonstrates the effect of changing
chi projection on a simple oil-brine-shale system.
Adjust chi using slider bar.
AI and GI values of oil, brine and shale can also be edited.
A noise ellipse can be specified to allow calculation of signal-to-noise
Copyright Patrick Connolly Associates Ltd. 2015 .
Exercise:
1) What is the chi angle of the fluid projection?
2) What is the chi angle of the lithology projection?
1000
1500
2000
2500
3000
3500
4000
0 60 120 180 240 300 360
imp
edan
ce
chi angle
brine
oil
shale
brine point
oil point
shale point
-600
-400
-200
0
200
400
600
800
0 60 120 180 240 300 360
imp
edan
ce c
on
tras
t
chi angle
shale-brine
shale-oil
brine-oil
shale-brine point
shale-oil point
brine-oil point
Scenario modelling for selecting c angles
Petrel: Dynamic real data AIGI cross-plotting
Vp0 2500 90 chi 0 -180
Vs0 1200 0.00 cos 1.00
rho0 2.30 1.00 sin 0.00
K 0.23
Vp Vs rho AI GI lnAI lnGI EEI
Shale 2499 982 2.24 5598 8522 8.63 9.05 5598
Brine 2393 1097 2.08 4985 7115 8.51 8.87 4985
Gas 1811 1132 1.71 3097 6095 8.04 8.72 3097
Shale 2438 1006 2.25 5486 7920 8.61 8.98 5486
Brine 3110 1645 2.10 6531 4349 8.78 8.38 6531
Oil 2953 1774 2.04 6012 3697 8.70 8.22 6012
Gas 2848 1829 1.91 5448 3570 8.60 8.18 5448
Shale 3208 1593 2.40 7699 4209 8.95 8.34 7699
Brine 4054 2407 2.36 9567 2524 9.17 7.83 9567
Oil 3942 2437 2.31 9102 2448 9.12 7.80 9102
Gas 3923 2467 2.25 8815 2442 9.08 7.80 8815
Shale 2743 1273 2.38 6523 5485 8.78 8.61 6523
Brine 3111 1579 2.34 7280 4246 8.89 8.35 7280
Oil 2869 1555 2.19 6278 4282 8.74 8.36 6278
shallow high porosity sand
medium porosity sand
deep Jurassic sand
Angola Oligicene
Shale,
5598
Brine,
4985
Gas,
3097
Shale,
5598
2000 4000 6000 8000 10000 12000 14000EEI
shallow high porosity sand
Shale,
5486
Brine,
6531
Oil,
6012
Gas,
5448
Shale,
5486
2000 4000 6000 8000 10000 12000 14000EEI
medium porosity sand
Shale
Brine
Gas
8.0
8.5
9.0
9.5
7.5 8.0 8.5 9.0
lnG
I
lnAI
shallow high porosity sand
This worksheet contains 4 datasets illustrating different
systems (data 1-3 courtesy Rock Physics Associates Ltd.)
Each dataset is plotted with linked points on an AIGI
crossplot.
You should first identify the location of each facies on the
crossplot. Then, using the slider bar, investigate optimum chi angles to seperate the various facies.
Copyright Patrick Connolly Associates Ltd. 2015 .
Shale
Brine
OilGas
8.0
8.5
9.0
9.5
8.0 8.5 9.0 9.5
lnG
I
lnAI
medium porosity sand
Designing rock physics templates
Image Gallery: Seismic Net Pay
Modelling tuning effects and calculating the detuning correction
Petrel: Seismic Net Pay exercise
AI F1 F2 F3 F4 flag dt dt filteredamplitude
sum
average
amplitude
normalised
average
amplitude
seimic
net-to-gross
detuning
correction1 / dt filt
10000 4 12 40 80 0 1 14.5 8964 618 0.12 0.07 0.60 0.07
2 14.6 17845 1226 0.23 0.14 0.60 0.07
max 5353 3 14.6 26563 1816 0.34 0.21 0.60 0.07
4 14.8 35101 2376 0.44 0.27 0.61 0.07
5 14.9 43199 2901 0.54 0.34 0.62 0.07
6 15.1 51213 3383 0.63 0.40 0.63 0.07
7 15.4 58605 3813 0.71 0.46 0.64 0.07
8 15.7 65690 4194 0.78 0.51 0.65 0.06
9 16.0 72434 4515 0.84 0.56 0.67 0.06
10 16.4 78561 4778 0.89 0.61 0.68 0.06
11 16.9 84261 4987 0.93 0.65 0.70 0.06
12 17.5 89843 5143 0.96 0.69 0.71 0.06
13 18.1 94970 5249 0.98 0.72 0.73 0.06
14 18.8 99732 5313 0.99 0.75 0.75 0.05
15 19.5 104207 5345 1.00 0.77 0.77 0.05
16 20.3 108451 5353 1.00 0.79 0.79 0.05
17 21.1 112497 5343 1.00 0.81 0.81 0.05
18 21.9 116493 5326 1.00 0.82 0.83 0.05
19 22.7 120353 5304 0.99 0.84 0.85 0.04
20 23.5 124022 5276 0.99 0.85 0.86 0.04
21 24.3 127488 5245 0.98 0.86 0.88 0.04
22 25.1 130738 5208 0.97 0.88 0.90 0.04
23 25.9 133762 5166 0.97 0.89 0.92 0.04
24 26.7 136559 5118 0.96 0.90 0.94 0.04
25 27.5 139134 5063 0.95 0.91 0.96 0.04
26 28.3 141502 5002 0.93 0.92 0.98 0.04
27 29.1 143678 4936 0.92 0.93 1.01 0.03
28 29.9 145737 4867 0.91 0.94 1.03 0.03
29 30.8 147736 4797 0.90 0.94 1.05 0.03 50
30 31.7 149572 4725 0.88 0.95 1.07 0.03
31 32.5 151249 4652 0.87 0.95 1.10 0.03
32 33.4 152769 4578 0.86 0.96 1.12 0.03
33 34.2 154127 4504 0.84 0.96 1.15 0.03
34 35.1 155316 4429 0.83 0.97 1.17 0.03
35 35.9 156331 4354 0.81 0.97 1.20 0.03
36 36.7 157165 4277 0.80 0.98 1.23 0.03
37 37.6 157821 4199 0.78 0.98 1.25 0.03
38 38.4 158305 4119 0.77 0.99 1.28 0.03
39 39.3 158627 4037 0.75 0.99 1.32 0.03
40 40.2 158804 3954 0.74 1.00 1.35 0.02
41 41.0 158852 3870 0.72 1.00 1.38 0.02
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 50 100 150 200
aver
age
amp
litu
de
true thickness (ms)
band-limited impedance tuning
Calculate
0
20
40
60
80
100
0 20 40 60 80 100
app
aren
t th
ickn
ess
(ms)
true thickness (ms)
apparent thickness
0.00
0.01
0.02
0.03
0.04
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 50 100 150 200
aver
age
amp
litu
de
apparent thickness (ms)
band-limited impedance tuning
tuning curve
1 / dt filt
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 50 100 150 200
seis
mic
net
-to
-gro
ss
apparent thickness (ms)
seismic net-to-gross
0
2
4
6
8
10
12
0 20 40 60 80 100
det
uin
ing
corr
ecti
on
apparent thickness (ms)
detuning correction
This worksheet calculates band-limited impedance
tuning effects that can be used to model the seismic net pay method.
Given a trapezoidal filter specification it will calculate the filtered dt (apparent thickness) and the summed
amplitude. From these, average amplitude, seismic net-to-gross and the detuning correction can be
derived.
Set the F1-F4 values and press Calculate.
Copyright Patrick Connolly Associates Ltd. 2015 .
Exercise:
1) What is the effect on the tuning curves of changing the high and low-end filter values?
2) If you reduce the low frequency content (i.e. increase F1 and F2) what happens to the values
of the detuning correction?
0
5
10
15
20
25
30
0 5 10 15 20 25 30
app
aren
t th
ickn
ess
(ms)
true thickness (ms)
apparent thickness
Petrel: SNP mapping exercise
Image Gallery: Inversion
Model spurious correlations from random ‘attributes’
Inversion algorithms
1 2 3 4 5 new r 2 =0.96
well ntg 0.41 0.59 0.16 0.24 0.62 ? well ntg 0.38 0.29 0.51 0.66 0.15
best 0.54 0.16 0.78 0.78 0.22 0.23 0.47 0.96 weights
calibrated 0.38 0.64 0.20 0.21 0.60 0.60 attribute 1 0.89 0.29 1.00 0.40 0.38 0.72
attribute 1 0.62 0.55 0.00 0.46 0.14 0.06 0.27 0.27 attribute 2 0.88 0.44 0.41 0.26 0.08 -0.92
attribute 2 0.28 0.83 0.93 0.73 0.09 0.13 -0.57 0.57 attribute 3 0.67 0.14 0.25 0.84 0.96 -0.57
attribute 3 0.41 0.38 0.36 0.11 0.66 0.41 0.69 0.69 attribute 4 0.92 0.74 0.16 0.95 0.29 1.00
attribute 4 0.50 0.41 0.21 0.55 0.43 0.95 0.28 0.28 attribute 5 0.48 0.69 0.31 0.31 0.27 -0.87
attribute 5 0.71 0.01 0.97 0.97 0.34 0.55 -0.92 0.92
attribute 6 0.00 0.16 0.53 0.47 0.13 0.68 -0.79 0.79
attribute 7 0.43 0.20 0.26 0.81 0.47 0.66 -0.28 0.28 0.64 0.21 0.71 0.29 0.27
attribute 8 0.35 0.78 0.42 0.97 0.66 0.68 0.14 0.14 -0.81 -0.40 -0.37 -0.24 -0.08
attribute 9 0.54 0.16 0.78 0.78 0.22 0.23 -0.98 0.98 -0.38 -0.08 -0.14 -0.48 -0.55
attribute 10 0.43 0.28 0.13 0.43 0.58 0.54 0.56 0.56 0.92 0.74 0.16 0.95 0.29
attribute 11 0.17 0.48 0.24 0.97 0.80 0.83 0.16 0.16 -0.42 -0.60 -0.27 -0.27 -0.23 target x y
attribute 12 0.23 0.84 0.69 0.03 0.92 0.19 0.57 0.57 averaged -0.01 -0.03 0.02 0.05 -0.06 1.00 intercept -0.09 0.00 0.00
attribute 13 0.87 0.88 0.22 0.79 0.47 0.21 0.37 0.37 calibrated 0.38 0.29 0.51 0.66 0.15 gradient 0.21 1.00 1.00
attribute 14 0.94 0.41 0.95 0.01 0.97 0.19 0.17 0.17
attribute 15 0.44 0.86 0.80 0.22 0.26 0.98 -0.06 0.06
attribute 16 0.59 0.90 0.29 0.65 0.21 0.30 0.16 0.16
attribute 17 0.95 0.49 0.61 0.05 0.59 0.45 0.26 0.26
attribute 18 0.49 0.99 0.57 0.95 0.97 0.74 0.52 0.52
attribute 19 0.19 0.19 0.10 0.31 0.81 0.03 0.59 0.59
attribute 20 0.76 0.86 0.85 0.18 0.28 0.44 -0.05 0.05
max 0.69 0.98
match 3.00 9.00
gradient -1.41 -1.41
intercept 1.07 1.07
intercept
#VALUE!
wei
ghte
d
attr
ibu
tes
calibration
Potential risks when using seismic attributes
as predictors of reservoir propertiesCYNTHIA T. KALKOMEY, Mobil E&P Technical Center, Dallas,
Texas
0.0
0.2
0.4
0.6
0.8
1.0
0.0 0.2 0.4 0.6 0.8 1.0at
trib
ute
pre
dic
tio
ns
well data
weighted attributes-0.1
0.1
0.3
0.5
0.7
0.9
1.1
-0.1 0.1 0.3 0.5 0.7 0.9 1.1
attr
ibu
te p
red
icti
on
well result
r2 = 0.96
Generate
This worksheet demonstrates the risks of using attributes as
reservoir predictors based only on correlation.
On the left are five well net-to-gross values. Beneath are 20
'attribute' values at each of the well locations. These have been generated from a uniform random distribution.
The program selects the attribute with the highest correlation and calibrates the values to minimise the
difference with the well data.
The calibrated attributes are cross-plotted against the well
data together with the R2 correlation coefficient.
Press 'Generate' to create more random attributes.
The right-hand-side of the worksheet demonstrates the risks
of using weighted combinations of attributes.
Copy random well net-to-gross values into cells S2:W2.
Copy five sets of random attribute values into S4:W8. We'll use a weighted combination of these attrributes as a
property predictor.
Launch the Solver from the Data tab. Set the objective to
maximise the target (X16). Set the variable cells to X4:X8 and set Constraints such that X4:X8 must lie between 1 and -
1. Select Solve.
The optimised set of weighted attributes is calibrated and
cross-plotted against the well data. Show this plot to your manager and tell him/her than you can now predict the
results at any well location (or perhaps not).
-0.179784 5,000 0 10,800 119 5,033 -5,767
-0.059997 5,000 0 11,203 120 5,226 -5,977
0.080695 5,000 0 11,222 121 5,216 -6,006
0.235671 5,000 0 10,784 122 4,961 -5,823
0.396784 5,000 0 9,839 123 4,431 -5,408
0.554917 5,000 0 8,365 124 3,608 -4,757
0.700652 5,000 0 6,371 125 2,491 -3,880
0.824989 5,000 0 3,900 126 1,096 -2,804
0.920042 5,000 0 1,026 127 -543 -1,570
0.979677 5,000 0 -2,145 128 -2,374 -229
1.000000 5,000 0 -5,485 129 -4,328 1,157
0.979677 1,990 -3,010 -8,849 130 -6,323 2,525
0.554917 1,990 -3,010 -19,510 134 -12,840 6,670
-0.059997 2,440 -2,560 -20,324 138 -12,971 7,353
-0.377750 3,330 -1,670 -10,307 142 -4,769 5,538
-0.302018 4,230 -770 4,407 146 7,761 3,354
-0.107506 4,780 -220 15,735 150 17,038 1,303
-0.041806 4,730 -270 18,868 154 17,846 -1,023
-0.093380 4,150 -850 13,964 158 11,521 -2,443
-0.127626 3,280 -1,720 4,697 162 3,552 -1,146
-0.097791 1,990 -3,010 -4,409 166 -3,079 1,330
-0.055568 1,990 -3,010 -10,249 170 -10,189 59
-0.038294 1,990 -3,010 -11,927 174 -18,635 -6,707
-0.031091 1,990 -3,010 -10,222 178 -22,992 -12,770
-0.017527 1,990 -3,010 -6,913 182 -16,687 -9,775
-0.006031 2,000 -3,000 -4,246 186 -2,493 1,754
-0.003910 3,040 -1,960 -3,507 190 7,590 11,096
-0.002347 2,360 -2,640 -3,257 194 6,245 9,502
0.005357 2,280 -2,720 -257 198 633 890
0.012580 3,230 -1,770 6,342 202 3,044 -3,298
0.012584 4,310 -690 12,392 206 13,807 1,415
0.008788 4,020 -980 12,166 210 20,030 7,865
0.007011 2,010 -2,990 4,440 214 12,287 7,847
0.007095 1,990 -3,010 -6,246 218 -3,471 2,775
0.006261 1,990 -3,010 -14,215 222 -13,882 334
0.004423 1,990 -3,010 -16,651 226 -13,417 3,234
0.002961 1,990 -3,010 -13,123 230 -7,049 6,074
0.002005 3,070 -1,930 -4,487 234 -928 3,559
0.001103 4,170 -830 5,594 238 3,305 -2,289
0.000408 4,590 -410 10,506 242 5,237 -5,269
0.000126 4,280 -720 6,085 246 3,051 -3,033
0.000049 3,340 -1,660 -3,653 250 -2,910 744
0.000006 2,000 -3,010 -9,182 254 -7,654 1,529
5,000 0 -5,218 258 -6,348 -1,130
5,000 0 -3,078 259 -5,080 -2,002
5,000 0 -735 260 -3,577 -2,842
reset value
0.1 damping factor
don't use slider bars
2,0005,249.11RMSE
Mean 2,923
0 5,000 10,000
50
100
150
200
250
300
350
impedance
50
100
150
200
250
300
350
-40,000-20,000 0 20,000 40,000
target
50
100
150
200
250
300
350
-40,000-20,000 0 20,000 40,000
residual
50
100
150
200
250
300
350
-40,000-20,000 0 20,000 40,000
model
Optimise
Reset
Petrel: ODiSI parameterisation
Petrel: ODiSI testing
Petrel: ODiSI application