Nagaoka 4D Seismic Revisited - IEAGHG

1

Nagaoka 4D Seismic Revisited

Naoshi Aoki, ○Akihisa Takahashi , Kazuya Shiraishi, Takao Nibe (JGI, Inc.),

Ziqiu Xue (RITE)2011.6.7

IEAGHG 7th Monitoring Network Meeting

2

From RITE Website

CO2 injection demonstration test was conducted by RITE at Nagaoka FY 2000-2004

Total amount of 10,405 tons of CO2 was injected.

3

4D Seismic Extent

1.7 km

2.0 km

4D

Seismic

Extent

and

Well

Locations

4D Seismic

IW-1OB-2

OB-3OB-4

Tanase et al. (2008)Well Locations

Depth Structure Map and

4D Seismic Extent

4

Injection History

Xue

et al. (2006)Red Square: CO2 Breakthrough at the well

5

2800m/s

2200m/s

Modified from Tanase

et al.(2008)

Pink and Green shows simplified velocities before and after CO2 breakthrough

Velocity Difference before/after CO2 breakthrough OB-2

6

Revisit the 4D seismic

• Time lapse seismic surveys were conducted in  2003 and 2005.

• Two surveys were reprocessed using the up‐ to‐date processing software.

• Seismic response in the five‐meter‐thick  injected zone was re‐evaluated.

• Acoustic impedance inversion was conducted.

7Reprocessing Results 2005 Repeated Survey

Time Slice 1.006sec

IL82

East of the injection areais contaminated with

migration noises.

WNW ESE

500 m

8Reprocessing Results Time Slices (1.006sec)

Previous Processing Re-processing

S/N improvement around reservoir

2005 Monitoring survey500 m

9

Log‐Seismic Correlation

OB-2 OB-3 OB-4

Blue：Synthetic Seismogram Red：Seismic DataLog-seismic correlation is fairly good except for OB-2.

OB-2 location would be contaminated with migration noises.

10

Injected Zone and Seismic Resolution

OB-4 Acoustic Impedance

Ic：Zone2b4 msec (5 m)

λ＝65 m（43 Hz)

Zone2_top

Zone3_top

Extracted Wavelet

Injected zone of 5 meters corresponds to 1/13 of wavelength

No separability but still have visibilityunder the good S/N ratio.

Blue: Actual AI Red: Estimated AI

11

11

OB-4 Synthetic Seismogram： Before CO2 arrival

Density Velocity AI REFC

SyntheticSeismogram

Bfr CO2 Arrival

SeismicSection

2005

Seismic IL82

1000 ms

950 ms

12

12

OB-4 Synthetic Seismogram： After CO2 arrival

Density Velocity AI REFC Seismic IL82

1000 ms

950 ms

• Travel timedifference is

just an 1 msec.

•Amplitude differencearound the target

is visible

SyntheticSeismogram

Aft CO2 Arrival

SeismicSection

2005

13Comparison

of

migrated

sections 2003

and

2005

20052003

Similar phenomena (amplitude increase) can be seenin actual seismic section.

Note: Quality of the 2003 survey is not sufficient due to the shortage of shot points.

14AI Inversion Results 2005 Survey

Low impedance anomaly is observed a bit below the injected zone in AI inversion.

OB-3(Projected)

OB-4 IW-1 OB-2 High

Low

Injected ZoneLow AI Anomaly

Lc Top

1000 ms

980 ms

1020 ms

1040 ms

15

Mis-tie analysis in AI inversion No Mis-tie Initial model

4 ms Delay

Blue: Original AI Log Red: Estimated AI Log

Low impedance peak apparently

delayed by 8 msec

Injection zone

980 ms

1000 ms

1020 ms

Possible cause of the time delay of the anomaly would bemis-ties between seismic and initial model for inversion.

16Volume Attribute Average AI Value (2005 Survey)

Average AI Value (Lc top +6 to +26 msec)

Orange to Yellow: Low Impedance Area

OB-2

OB-3

OB-4

17

Summary of Results

• 4D response from such a thin layer (5 meters)  can be detected.

• Quality of the 2003 survey was not sufficient  for 4D processing due to the shortage of shot  points.

• Extent of CO2 can be recognized by amplitude  anomaly in the 2005 survey.

• Position of amplitude anomaly on AI inversion  would be estimated a bit downward due to 

mis‐ties.

18

Thank you for your kind attention

Evaluation of the Nagaoka test field is still ongoing.