Computational Challenges for Finding Big Oil by Seismic Inversion.

Slides:



Advertisements
Similar presentations
Multisource Full Waveform Inversion of Marine Streamer Data with Frequency Selection Multisource Full Waveform Inversion of Marine Streamer Data with Frequency.
Advertisements

Warping for Trim Statics
Multi-source Least-squares Migration with Topography Dongliang Zhang and Gerard Schuster King Abdullah University of Science and Technology.
Utah Tomography and Modeling/Migration (UTAM) Consortium S. Brown, Chaiwoot B., W. Cao, W. Dai, S. Hanafy, G. Zhan, G. Schuster, Q. Wu, X. Wang, Y. Xue,
Multisource Full Waveform Inversion of Marine Streamer Data with Frequency Selection Multisource Full Waveform Inversion of Marine Streamer Data with Frequency.
Multi-source Least Squares Migration and Waveform Inversion
Adaptive Grid Reverse-Time Migration Yue Wang. Outline Motivation and ObjectiveMotivation and Objective Reverse Time MethodologyReverse Time Methodology.
Aerial View Eucalyi River, Peru Seismic View: 2 km Deep Meandering River.
First Arrival Traveltime and Waveform Inversion of Refraction Data Jianming Sheng and Gerard T. Schuster University of Utah October, 2002.
Multiscale Waveform Tomography C. Boonyasiriwat, P. Valasek *, P. Routh *, B. Macy *, W. Cao, and G. T. Schuster * ConocoPhillips.
Solving Illumination Problems Solving Illumination Problems in Imaging:Efficient RTM & in Imaging:Efficient RTM & Migration Deconvolution Migration Deconvolution.
Joint Migration of Primary and Multiple Reflections in RVSP Data Jianhua Yu, Gerard T. Schuster University of Utah.
Arbitrary Parameter Extraction, Stationary Phase Migration, and Tomographic Velocity Analysis Jing Chen University of Utah.
Applications of Time-Domain Multiscale Waveform Tomography to Marine and Land Data C. Boonyasiriwat 1, J. Sheng 3, P. Valasek 2, P. Routh 2, B. Macy 2,
1 Fast 3D Target-Oriented Reverse Time Datuming Shuqian Dong University of Utah 2 Oct
Demonstration of Super-Resolution and Super-Stacking Properties of Time Reversal Mirrors in Locating Seismic Sources Weiping Cao, Gerard T. Schuster, Ge.
Multisource Least-squares Reverse Time Migration Wei Dai.
Multisource Least-Squares Migration Multisource Least-Squares Migration of Marine Streamer Data with Frequency-Division Encoding Yunsong Huang and Gerard.
3D Tomography using Efficient Wavefront Picking of Traveltimes Abdullah AlTheyab and G. T. Schuster King Abdullah University of Science and Technology.
V.2 Wavepath Migration Overview Overview Kirchhoff migration smears a reflection along a fat ellipsoid, so that most of the reflection energy is placed.
Making the Most from the Least (Squares Migration) G. Dutta, Y. Huang, W. Dai, X. Wang, and Gerard Schuster G. Dutta, Y. Huang, W. Dai, X. Wang, and Gerard.
Overview of Multisource Phase Encoded Seismic Inversion Wei Dai, Ge Zhan, and Gerard Schuster KAUST.
An Efficient Multiscale Method for Time-Domain Waveform Tomography C. Boonyasiriwat 1, P. Valasek 2, P. Routh 2, W. Cao 1, G.T. Schuster 1, and B. Macy.
Angle-domain Wave-equation Reflection Traveltime Inversion
Utah Tomography and Modeling/Migration (UTAM) Consortium S. Brown, C. Boonyasiriwat, W. Cao, W. Dai, S. Brown, C. Boonyasiriwat, W. Cao, W. Dai, S. Hanafy,
Center for Subsurface Imaging and Fluid Modeling Shuyu Sun and GT Schuster 8 PhD students, 5 Research Fellows (Prof Sherif Hanafy, Dr. Chaiwoot.
Least-squares Migration and Least-squares Migration and Full Waveform Inversion with Multisource Frequency Selection Yunsong Huang Yunsong Huang Sept.
Least Squares Migration of Stacked Supergathers Wei Dai and Gerard Schuster KAUST vs.
Impact of MD on AVO Inversion
Theory of Multisource Crosstalk Reduction by Phase-Encoded Statics G. Schuster, X. Wang, Y. Huang, C. Boonyasiriwat King Abdullah University Science &
Multiples Waveform Inversion
Migration Velocity Analysis 01. Outline  Motivation Estimate a more accurate velocity model for migration Tomographic migration velocity analysis 02.
Multisource Least-squares Migration of Marine Data Xin Wang & Gerard Schuster Nov 7, 2012.
Signal Analysis and Imaging Group Department of Physics University of Alberta Regularized Migration/Inversion Henning Kuehl (Shell Canada) Mauricio Sacchi.
Fast Least Squares Migration with a Deblurring Filter Naoshi Aoki Feb. 5,
Annual Meeting Utah Tomography and Modeling/Migration (UTAM) Feb. 5-6, 2009 N. Aoki, C. Boonyasiriwat, S. Brown, W. Cao, W. Dai, N. Aoki, C. Boonyasiriwat,
A Blind Test of Traveltime and Waveform Inversion Colin A. Zelt 1, R. Gerhard Pratt 2, Andrew Brenders 2, Sara Hanson-Hedgecock 1 and John A. Hole 3 1.
Multiscale Waveform Tomography C. Boonyasiriwat, P. Valasek, P. Routh, B. Macy, W. Cao, and G. T. Schuster * ConocoPhillips * **
Super-virtual Interferometric Diffractions as Guide Stars Wei Dai 1, Tong Fei 2, Yi Luo 2 and Gerard T. Schuster 1 1 KAUST 2 Saudi Aramco Feb 9, 2012.
G. Schuster, S. Hanafy, and Y. Huang, Extracting 200 Hz Information from 50 Hz Data KAUST Rayleigh Resolution ProfileSuperresolution Profile Sinc function.
Wave-Equation Waveform Inversion for Crosswell Data M. Zhou and Yue Wang Geology and Geophysics Department University of Utah.
Migration Velocity Analysis of Multi-source Data Xin Wang January 7,
Benefits & Limitations of Least Squares Migration W.Dai,D.Zhang,X.Wang,GTSKAUST RTM Least Squares RTM GOM RTM GOM LSRTM.
Fast Least Squares Migration with a Deblurring Filter 30 October 2008 Naoshi Aoki 1.
Fast 3D Least-squares Migration with a Deblurring Filter Wei Dai.
The Boom and Bust Cycles of Full Waveform Inversion: Is
LSM Theory: Overdetermined vs Underdetermined
Zero-Offset Data d = L o ò r ) ( g = d dr r ) ( g = d
Overview of Geophysical Research Research
Fast Multisource Least Squares Migration of 3D Marine Data with
Making the Most from the Least (Squares Migration)
Iterative Non-Linear Optimization Methods
Steepest Descent Optimization
Fast Multisource Least Squares Migration of 3D Marine Data with
Skeletonized Wave-equation Inversion for Q
Skeletonized Wave-Equation Surface Wave Dispersion (WD) Inversion
Efficient Multiscale Waveform Tomography and Flooding Method
Overview of Multisource Phase Encoded Seismic Inversion
Migration Intuitive Least Squares Migration Green’s Theorem.
Non-local Means (NLM) Filter for Trim Statics
Overview of Multisource and Multiscale Seismic Inversion
Least-squares Reverse Time Migration with Frequency-selection Encoding for Marine Data Wei Dai, WesternGeco Yunsong Huang and Gerard T. Schuster, King.
Overview of Multisource and Multiscale Seismic Inversion
PS, SSP, PSPI, FFD KM SSP PSPI FFD.
King Abdullah University of Science and Technology
Chaiwoot Boonyasiriwat
Non-local Means (NLM) Filter for Trim Statics
Inverse Crimes d=Lm m=L-1 d Red Sea Synthetics
Least Squares Migration
Wave Equation Dispersion Inversion of Guided P-Waves (WDG)
Presentation transcript:

Computational Challenges for Finding Big Oil by Seismic Inversion

JackBuckskin KaskidaTiber 35,055 Feet Motivation for Better Seismic Imaging Strategy ¼ billion $$$ well

Motivation for Better Seismic Imaging Strategy Oil Well Blowouts

Overpressure Zone Motivation for Better Seismic Imaging Strategy Oil Well Blowouts = Low Seismic Velocity Zone

Motivation for Better Seismic Imaging Strategy Mud Volcanoes 6.3 km 2 13 people killed30,000 people displaced May 29, 2006

Computational Challenge Seismic InversionComputational Challenge Seismic Inversion Outline Full waveform InversionFull waveform Inversion Multisource InversionMultisource Inversion

Given: d = Lm Seismic Inverse Problem Find: m(x,y,z) Find: m(x,y,z) Soln: min || Lm-d || Soln: min || Lm-d ||2 m = [L L] L d T T L d L dT migration waveforminversion

Given: d = Lm Computational Challenges Find: Find: m = [L L] L d T T 20x20x10 km 3 dx=1 m # time steps ~ 10 4 # shots > 10 4 m > 10 unknown velocity values Total = d > words

Computational Challenge Seismic InversionComputational Challenge Seismic Inversion Outline Full waveform InversionFull waveform Inversion Multisource InversionMultisource Inversion

Multisource Migration: m mig =L T d Forward Model: m =[L T L] -1 L T d Multisrc-Least FWI: Multisource Encoded FWI m’ = m - L T [Lm - d] f ~ [L T L] -1 f Steepest Descent Preconditioned d +Nd =[N L +NL ]m Nd +Nd =[N L +NL ]m multisource preconditioner

Multiscale Waveform Tomography 1. Collect data d(x,t) 2. Generate synthetic data d(x,t) by FD method syn. 3. Adjust v(x,z) until ||d(x,t)-d(x,t) || minimized by CG. syn To prevent getting stuck in local minima: a). Invert early arrivals initially a). Invert early arrivals initially mute 7 b). Use multiscale: low freq. high freq. b). Use multiscale: low freq. high freq.

0 km 20 km 0 km 6 km 3 km/s 6 km/s Boonyasiriwat et al., 2009, TLE

3 km/s 6 km/s Initial model 5 Hz 10 Hz 20 Hz Waveform Tomograms 3 km/s 6 km/s 3 km/s 6 km/s 3 km/s 6 km/s 0 km 6 km 0 km 6 km 0 km 6 km 0 km 20 km 6 km

Low-pass Filtering 18 (b) 0-15 Hz CSG (c) 0-25 Hz CSG

Dynamic Early-Arrival Muting Window Hz CSG 0-25 Hz CSG Window = 1 s

Hz CSG 0-25 Hz CSG Window = 2 s Dynamic Early-Arrival Muting Window

Depth (km) X (km) Traveltime Tomogram Velocity (m/s) Waveform Tomogram Depth (km) Results

Depth (km) X (km) Waveform Tomogram Velocity (m/s) Depth (km) Vertical Derivative of Waveform Tomogram

Kirchhoff Migration Images 22

Kirchhoff Migration Images 22

Computational Challenge Seismic InversionComputational Challenge Seismic Inversion Outline Full waveform InversionFull waveform Inversion Multisource InversionMultisource Inversion

1980 Multisource Seismic Imaging vs copper VLIW Superscalar RISC Aluminum Year CPU Speed vs Year

FWI Problem & Possible Soln. Problem: FWI computationally costlyProblem: FWI computationally costly Solution: Multisource Encoded FWISolution: Multisource Encoded FWI Preconditioning speeds up by factor 2-3 Iterative encoding reduces crosstalk

Multisource Migration: m mig =L T d Forward Model: Multisource Phase Encoded Imaging d +d =[ L +L ]m 1221 L {d { =[ L +L ](d + d ) TT = L d +L d TT L d +L d L d +L d212 1 Crosstalk noise Standard migration TT m = m + (k+1)(k)

Multi-Source Waveform Inversion Strategy (Ge Zhan) Generate multisource field data with known time shift Generate synthetic multisource data with known time shift from estimated velocity model Multisource deblurring filter Using multiscale, multisource CG to update the velocity model with regularization Initial velocity model 144 shot gathers

3D SEG Overthrust Model (1089 CSGs) 15 km 3.5 km 15 km

3.5 km Dynamic QMC Tomogram (99 CSGs/supergather) (99 CSGs/supergather) Static QMC Tomogram (99 CSGs/supergather) 15 km Dynamic Polarity Tomogram (1089 CSGs/supergather) Numerical Results

Multisource FWI Summary (We need faster migration algorithms & better velocity models) IO 1 vs 1/20 Cost 1 vs 1/20 or better Resolution dx 1 vs 1 Sig/MultsSig ? Stnd. FWI Multsrc. FWI Stnd. FWI Multsrc. FWI

Multisource FWI Summary (We need faster migration algorithms & better velocity models) Future: Multisource MVA, Interpolation, Field Data, Migration Filtering, LSM

Research Goals G.T. Schuster (Columbia Univ., 1984) Seismic Interferometry: VSP, SSP, OBS Multisource+Preconditioned RTM+MVA+Inversion+Modeling: TTI 3D RTM, GPU: Stoffa+CSIM, UUtah K. Johnson SCI, PSU, KAUST Shaheen Cornea

Multisource S/N Ratio # geophones/CSG # CSGs L [d + d +.. ] d + d T d, d 2 1 L [d + d + … ] 1 2 T, …. +….

Multisrc. Migration vs Standard Migration # iterations Iterative Multisrc. Migration vs Standard Migration vs MS S-1 M ~ ~ # geophones/CSG # CSGs MS MI

Crosstalk Term Time Statics Time+Amplitude Statics QM Statics L d +L d L d +L d212 1 TT

Summary Time Statics Time+Amplitude Statics QM Statics 1. Multisource crosstalk term analyzed analytically 2. Crosstalk decreases with increasing w, randomness, dimension, iteration #, and decreasing depth dimension, iteration #, and decreasing depth 3. Crosstalk decrease can now be tuned 4. Some detailed analysis and testing needed to refine predictions. predictions. L d +L d L d +L d212 1 TT

Fast Multisource Least Squares Kirchhoff Mig.Fast Multisource Least Squares Kirchhoff Mig. Multisource Waveform Inversion (Ge Zhan)Multisource Waveform Inversion (Ge Zhan) Multisource Technology

0 Z k(m) 3 0X (km)16 The Marmousi2 Model The area in the white box is used for S/N calculation.

0X (km)16 0 Z k(m) 3 0 Z (km) 3 0X (km)16 Conventional Source: KM vs LSM (50 iterations) LSM (100x) KM (1x)

0X (km)16 0 Z k(m) 3 0 Z (km) 3 0X (km) source Supergather: KM vs LSM (300 its.) LSM (33x) KM (1/200x)

S/N 0 1 I 300 S/N = 7 The S/N of MLSM image grows as the square root of the number of iterations. MI

Fast Multisource Least Squares Migration ( Dai)Fast Multisource Least Squares Migration ( Dai) Multisource Waveform Inversion (Boonyasiriwat)Multisource Waveform Inversion (Boonyasiriwat) Multisource Technology

Comparing CIGs 23

Comparing CIGs 24 CIG from Traveltime Tomogram CIG from Waveform Tomogram

Comparing CIGs 25

Comparing CIGs 26 CIG from Traveltime Tomogram CIG from Waveform Tomogram

Comparing CIGs 27

Comparing CIGs 28 CIG from Traveltime Tomogram CIG from Waveform Tomogram

17 Data Pre-Processing 3D-to-2D conversion Attenuation compensation Random noise removal

17 Source Wavelet Estimation Pick the water-bottom Stack along the water-bottom to obtain an estimate of source wavelet Generate a stacked section In some cases, source wavelet inversion can be used.

17 Gradient Computation and Inversion Multiscale inversion: low to high frequency Dynamic early-arrival muting window Normalize both observed and calculated data within the same shot Quadratic line search method (Nocedal and Wright, 2006) A cubic line search can also be used.