Title: Seismic monitoring of a unique CO2 injection site at Sleipner
1Seismic monitoring of a unique CO2 injection site
at Sleipner
2Increasing atmospheric CO2 causes global warming.
The Kyoto agreement put limits on emissions.A
CO2 free vision
3CO2 physical properties
Measured at HARC 2000
4Sleipner A
5CO2 injection well in Utsira Fm.
6Sleipner 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
7Seismic 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
8difference
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
9200 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
10southwest
northeast
500 m
100 ms
B
C
D
A
.
A1
1994 vintage
11amplitude Horizon B time
500 m
.
.
16000 8000 0 amplitude
929 949 969 ms
12Model 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
13Estimated mass based on seismic time-delay
Time delay Gas column height
Uncertainty distribution assessed in each
factor. Monte Carlo simulation.
14Possible 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 ?
15Fundamental 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
16Conclusions
- 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.
17Acknowledgement
- 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.