Seismic monitoring of a unique CO2 injection site at Sleipner

1
Seismic monitoring of a unique CO2 injection site
at Sleipner

• Ola Eiken, Statoil

2
Increasing atmospheric CO2 causes global warming.
The Kyoto agreement put limits on emissions.A
CO2 free vision
3
CO2 physical properties
Measured at HARC 2000
4
Sleipner A
5
CO2 injection well in Utsira Fm.
6
Sleipner CO2 injection

• CO2 is injected into a thick sandstone layer at
  800-1100 m depth below sealevel
• 97.5-98 of the injected gas is CO2
• The sandstones have porosities of 35-40 and 1-8
  D permeability

seismic survey
7
Seismic data acquisition and processing

• Reprocessing of base survey together with
  processing of monitor survey.
• Comprehensive processing sequence with aim of
  maximum repeatability

• Deterministic zero-phasing using supplied
  far-field signatures from 1994 and 1999.
• Tidal correction based on model
• Swath consistent static correction
• Global frequency-amplitude match between surveys
• New velocity analyses in CO2 injection area

8
difference
1994
october 1999
after injecting 2 mill. tons CO2 since 1996
no change above this level
top Utsira Fm.
100 ms
injection point
1000 m
9
200 m
crossline 1122
east
east
west
west
horizon A0
horizon A
100 ms
horizon A1
horizon B
horizon C
horizon D
B
D
C
.
A
A1
1994 vintage
1999 - 1994 difference
10
southwest
northeast
500 m
100 ms
B
C
D
A
.
A1
1994 vintage
11
amplitude Horizon B time
500 m
.
.
16000 8000 0 amplitude
929 949 969 ms
12
Model of reflection from top and bottom of a thin
CO2-filled layer (with velocity 1450 m/s).
Bottom of layer is at constant depth.
Maximum amplitude
Pulse 0 1 2 3 4 5 6 7 8
9 10 11 12 13 14
Layer thickness in meters
13
Estimated mass based on seismic time-delay
Time delay Gas column height
Uncertainty distribution assessed in each
factor. Monte Carlo simulation.
14
Possible causes of discrepancies in mass
calculations

• CO2 is partly accumulated in thin layers and
  partly distributed at low saturations between the
  layers.

• The reservoir is warmer and CO2 less dense than
  assumed ?
• The Gassman theory is not valid ?

15
Fundamental limitations on seismically based mass
estimates

• The non-linear relation between Vp and Sg.
• Causes gas saturations between 0.1 and 1.0 to be
  unresolvable
• Lack of information in the 2-10 Hz frequency
  band.
• Saturation gradients are not discernible

16
Conclusions

• Injected CO2 is well imaged by time-lapse seismic
  data.
• The geometry of the CO2 bubble is well defined.
• There is no evidence of leakage through the seal.
• A portion of the injected CO2 is trapped at high
  saturations beneath thin shales.
• The observed time-delays is not sufficiently
  explained by high-saturation CO2 layers.
• There is a basic limitation in time-lapse seismic
  data on resolving long-wavelength variations and
  on quantifying CO2 saturations.

17
Acknowledgement

• Rob Arts from TNO, Andy Chadwick from BGS and
  Erik Lindeberg from SINTEF, who all have
  contributed a lot in the multi-diciplinary SACS
  team.
• Statoil and the Sleipner field partners Exxon,
  Norsk Hydro and TotalFinaElf.
• Statoils SACS partners BP-Amoco, Norsk Hydro,
  Exxon, TotalFinaElf, Vattenfall, British
  Geological Survey, Bureau de Recherches
  Gèologiques et Minières, Geological Survey of
  Denmark and Greenland, Institut Francais du
  Pètrole, Netherlands Institute of Applied
  Geoscience, Sintef Petroleum Research.
• WesternGeco, who acquired and processed the
  seismic data.
• The European Union RD programme Thermie.