Seismic Monitoring of a Small‐Scale Supercritical CO2/CH4 Injection:CO2CRC Otway project Case Study
R. Pevzner1,2, M. Urosevic1,2, K. Tertyshnikov1,2, B. Gurevich1,2 , S. V. Shulakova2,4, S. Glubokovskikh1,2, D. Popik1,2, J. Correa1,2, A. Egorov1,2, H. AlNasser1,2, A. Kepic1,2 , B.M. Freifeld3, M. Robertson3, T. Wood3, T.M. Daley3 and R. Singh11CO2CRC, 2Curtin University, 3Lawrence Berkeley National Lab,4CSIRO
11th IEAGHG Monitoring Network MeetingTraverse City, Michigan June 13th – 15th, 2017
ACKNOWLEDGEMENTS
We would like to acknowledge the funding provided by the Australian government to support this CO2CRC research project.
We also acknowledge funding from ANLEC R&D and the Victorian Government for the Stage 2C project.
Funding for LBNL was provided through the Carbon Storage Program, U.S. DOE, Assistant Secretary for Fossil Energy, Office of Clean Coal and Carbon Management through the NETL.
We thank the National Geosequestration Laboratory (NGL) for providing the seismic sources (INOVA Vibrators) for this project. Funding for NGL was provided by the Australian Federal Government.
Paaratte
Stage Iinjection
Stage IIinjection
STAGE I: An 80/20 % of CO2/CH4 stream produced from Buttress, transported and injected into CRC‐1 well (previous CH4 production well) ‐65 Kt.
STAGE II: CO2/CH4 stream injected into CRC‐2 well – 15 Kt.
Otway Basin Pilot Project (Victoria, Australia)
Stage 2C Project goals
• Detect injected Buttress gas in the subsurface: ascertain minimum seismic detection limit
• Observe the gas plume development using time‐lapse seismic
• Verify stabilisation of the plume in the saline formation using time lapse seismic
Otway site aerial photo
CRC‐1
Naylor‐1
CRC‐2
Seismic surveys @ Otway site2000 2006 2007 2008 2009 2010 2012 2013 2015 2016 2017
2D seismic Sodas Ln, 7 surveys, various sources and seasons
PROTECT, SH, several lines
3D&4D seismic
CurdieVale
Baseline, Otw 4D Otw 4D, M1
Otw 4D, M2
Shallow3D/3C survey
4D with buried receivers
Test array,Soda RdGeophones + iDAS
Stage 2C, baseline
Stage 2C, 3xMonitor surveys, M1‐M3)
Stage 2C, M4
Zero‐offsetand offset VSP
Naylor 1, Z, O, WA
CRC‐1(Z, O, 4DBaseline)
CRC‐2 (Z)
CRC‐1(Z, O, 4D Monitor)
CRC‐1 (Z, O, Hydrophones)CRC2 (iDAS)
CRC‐1 4xOffsets
CRC‐2(iDAS)
CRC‐1 4xOffsets
CRC‐2(iDAS)
CRC‐1 4xOffsets
CRC‐2(iDAS)
Walk‐away and 3D/4D VSP
CRC2 (iDAS)
CRC‐1 (4D/3CBaseline)CRC‐2 (iDAS)
CRC‐1 (4D/3CMonitor)CRC‐2 (iDAS)
CRC‐1 (4D/3CMonitor)CRC‐2 (iDAS)
Near surface Various refraction and micro‐VSP surveys
Site characterization
Source tests
Stage 1Stage 2 prepar
Stage 2C
Stage 2C monitoring strategy
Full 4D finite‐difference time domain (FDTD) synthetic dataset was generated prior commencement of the first monitor survey and used to pre‐define and validate processing flows (Glubokovskikh et al., IJGGC 49, 2016)
4D seismic with buried receiver array acquired concurrently with 4D VSPBaseline: March 2015Monitor surveys: 5 kt, 10kt, 15 kt of injection (January‐April 2016), 1&2 years post injection (January 2017&2018)
Offset VSPs
Passive seismic using buried receiver array
LBNL group lead: Trialing 4D seismic with buried DAS array, 4D VSP in CRC‐2 (optical fiber on the tubing) and continuous seismic sources
TimelineFebruary 2015 – Receiver array installedMarch 2015 – Baseline data acquiredSeptember 2015• LBNL installs permanent vibroseis sources on site, baseline acquired;• passive seismic acquisition testedNovember 2015• Passive seismic data acquisition commences, including iDAS (8000 s / day)January 2016 – Monitor 1 (5122 t CO2) acquired, new foundations for permanent vibes builtFebruary 2016 • Both permanent vibes became operational• Monitor 2 (10000 t) acquiredApril 2016 – Monitor 3 (15000 t) acquiredJanuary 2017 – Monitor 4 (1 year post‐injection) acquired
Acquisition geometry
1 km
General Survey ParametersTotal Number of Source
Lines
26+1 Lines
Total Number of Sources 3003
Points
Source Line Spacing from 50 m
to 100 m
Source Point Spacing 15 m
Total Number of
Receiver Lines
11 Lines
Total Number of
Receivers
909 Points
Receiver Line Spacing 100 m
Receiver Point Spacing 15 m
Max Offset 2480 m
Sample Interval 1 ms
Receivers
RECEIVER PARAMETERSReceiver Type Sercel SG‐5Recording Pattern
Orthogonal cross‐spread pattern
Receiver Line Spacing
100 m
Receiver Point Spacing
15 m
Receiver Depth
4 m
Cables Depth 0.8 m
geophone
Receivers
trench
crosslineunit
FDU
6 wireless stations above the buried geophones on line 5
SOURCE PARAMETERS
Source Type INOVA UniVibe
26000 lbs
Sweep frequency 6‐150 Hz
Tapers 0.5 s
Sweep Length 24 s
Listening Time 5 s
Source
Line 5, receiver 46, common receiver gatherBuried Surface
Noise floor reduction ~ 25 dB
Fast‐track processing flowchartProcedure Parameters
Data Input SEG‐D data inputCorrelation with sweep signal Linear sweep 6‐150 Hz, length of sweep 24 s, output trace length 5 sBinning Bin size 7.5 m x 7.5 mTrace Editing Kill bad traces/seismogramsElevation Statics Final datum elevation – 30 m (MSL),Replacement Velocity – 1800 m/sRadon Filtering Number of P‐values – 700, Modelled noise subtraction, Applied in cone windowAutomatic Gain Control 500 ms, applied before radon filter and removed afterStatics 500 ms shift, applied before radon filter and removed afterAir Blast Attenuation Energy with velocity of 330 m/s was attenuatedSurface Waves Noise Attenuation 900 m/s, 6‐35 HzSpiking Deconvolution Zero‐phase spiking, Decon Operator length – 200 ms, Operator ‘white noise’ level – 0.1 %Automatic Gain Control 500 ms, applied before deconvolution and removed afterInteractive Velocity Analysis 2 iterations, VA Grid – 100 m x 100 m, 30% ‐ NMO mutingResidual Static Correction 2 iterations, Max Power AutostaticsAutomatic Gain Control and NMO AGC window ‐500 ms, NMO muting – 30%CDP stacking Stacking method – Mean, Power scalar for stack normalization 0.5Pad 3D Stack Volume INLINES 1‐219, XLINES 1‐ 266FXY‐deconvolution Wiener Levinson filter, 4‐180 HzFK Filter Applied in polygonMigration Phase shift Time Migration / Explicit FD Time Migration
2010 2015
Buried receiver array – preliminary results
Buried array higher resolution~25 dB ambient noise floor reductionVirtually all‐weather acquisitionLower impact on the land occupiers with no cables on the ground
Overall higher quality of the datahigher resolution – better source + sensitivity of the geophonesmore energy compared to 2009/2010 surveys
Time shifts computed between B and (top‐left to bottom‐right): M1, M2, M3, M4
Histograms of the time shifts
Histograms of NRMS values computed between the baseline image and each of monitor images: M1 (red), M2 (blue), M3 (green), and M4 (black). Computations: 200 ms window centred at 1000 ms
Distribution density of NRMS values computed in (from left to right): 60 ms windowcentred at 1136 ms, 200 ms window centred at 1000 ms and 400 ms window centred at 900 ms.
Survey area map
Intersection along the arbitrary line
Intersection along the arbitrary line
RMS amplitudes of the differences computed in 24 ms window centred at the plume level (1210 ms). The differences are computed between B and (top‐left to bottom‐right): M1, M2, M3, M4
VSP in CRC‐1Sercel SlimWave 3C VSP tool (10 levels, 15 m spacing)3D VSP with tool @ 760‐880 m MD4 offset VSPs
Comparison of baseline 4D VSP and surface seismic data
3D surface seismic 3D surface seismic 3D VSP
CRC‐1
Target intervalT (m
s)
Stage 2C 4D VSP results, xline 122
B M2 M3 M2‐B M3‐B
Offset VSP, SP1, M2‐B
Baseline Monitor 2 Difference
~1500 m MD
30 © Copyright Silixa Ltd 2016
• Standard optical fibre acts as the
sensor array
– Typical sampling at 10kHz on
10,000m fibre
– Standard gauge length of 10m
– Spatial sampling of 25cm
– DAS measures change in average
elongation per 10m gauge length
per 0.1ms acoustic time sample,
sampled every 0.25 m in distance
Distributed Acoustic Sensing
z, t Parker et al., Distributed Acoustic Sensing – a
new tool for seismic applications, first break (32), February 2014
FAT Helical Wound Cable• Anderson and Shapiro – HWC on soft mandrel 1980 US Patent 4375313 • Hornman et al. (2013 75th EAGE) introduced a helical wound FO cable • LBNL trialed multiple designs with varying physical properties• Line 5 installed one length of HWC for comparison to straight fiber
30° spiral wound on 58 Shore A rubber mandrel.
Normal Telecom Cable used in all trenches
Lessons learned – acoustic impedance of cable and surrounding soil is important
Surface Orbital Vibrator – VFD Controlled AC Induction Motor
Max Frequency 80 Hz, Force (@80Hz) 10 T‐fPhase stability is not maintained. Operate 2.5 hr/d
Force is adjustable
F=mω2r
Deconvolved SOV Data
• Helical Cable shows good sensitivity to reflected P. • Straight telecom less sensitivity
DAS after Post Stack
Time Migration
Strong reflection at
500 ms (related to a
carbonate layer)
Far offsets were
included in the stack
(due to directionality)
DAS 3D cube
DAS Monitor 2
Geophone Monitor 2
Stage 2C of the Otway project ‐ Conclusions
15,000 t were injected into the subsurface and an extensive seismic monitoring program was rolled out to detect itThe data is likely to be sufficient to claim detection & observation of the plume evolutionBuried receiver array‐ Better S/N, higher repeatability‐ Lower impact on landowners‐ Passive recording capability + ability to pair it with permanent
sourcesVSP data is inline with the surface seismic dataiDAS (in trenches) can be used to image subhorizontal reflectorsPermanent vibes ‐ operational
Next steps for DAS – Improving sensitivity
Stage 3: Comparison Carina DAS cable vs standard telcom in CRC‐3, SP0, 700 m offset, 5 sweeps
Geophones, Z component
iDAS v2 Carina
Government, Industry and Research Partners