Measurements of the gravitational acceleration due to mass distributions within the earth may be used to detect variations in subsurface rock or fluid density. Although of much lower spatial resolution than the seismic method, gravimetry offers some important complementary adjuncts to time-lapse seismic monitoring. Firstly, it can provide independent verification of the change in subsurface mass during injection via Gauss's Theorem. This potentially important capability may enable estimates to be made of the amount of CO2 going into dissolution, a significant source of uncertainty in efforts to quantify free CO2 in the reservoir (dissolved CO2 is effectively invisible on seismic data). Secondly, deployed periodically, gravimetry could be used as an 'early warning system' to detect the accumulation of CO2 in shallow overburden traps where it is likely to be in the low density gaseous phase with a correspondingly strong gravity signature.
The possibility of monitoring injected CO2 with repeated gravity measurements is strongly dependent on CO2 density and subsurface distribution. A feasibility study of time-lapse gravimetry at Sleipner (Williamson et al., 2001) modelled plume scenarios with CO2 densities ranging from over 700 kgm-3 (corresponding to the lower reservoir temperature scenario) to less than 350 kgm-3 (corresponding to possible higher reservoir temperatures). The modelling indicated that future changes in the CO2 plume could theoretically be detectable by seabed gravimetry. For example it was shown that the addition of 2 million tonnes to the plume would produce a change in peak gravity signal of between -8 and -33 mGal, corresponding to CO2 densities of 700 kgm-3 and 350 kgm-3, respectively (Figure 7). Longer-term predictions suggest that the peak gravity signature of the plume will gradually decrease as it thins by lateral migration at the reservoir top. On the other hand, if CO2 leaked to shallower levels where it would have a still lower density, gravity changes could well exceed -100 mGal.
200 300 400 500 600 700 300
200 300 400 500 600 700 300
Figure 7. Peak gravity anomaly as a function of CO2 density, predicted for the Sleipner CO2 plume in 1999 (2.3 MT in situ). Gravity changes computed for the sea surface (solid symbols) and the seabed (open symbols).
A seabed gravity survey was acquired at Sleipner in 2002 (Eiken et al. 2003), with approximately 5 million tonnes of CO2 in the plume. The survey was based around pre-positioned concrete benchmarks on the seafloor that served as reference locations for the (repeated) gravity measurements. Relative gravity and water pressure measurements were taken at each benchmark by a customised gravimetry and pressure measurement module mounted on a Remotely Operated Vehicle (Figure 8).
Thirty concrete benchmark survey stations were deployed in two perpendicular lines, spanning an area some 7 km east-west and 3 km north-south and overlapping the subsurface footprint of the CO2 plume. Each survey station was visited at least three times to better constrain instrument drift and other errors. Single station repeatability was estimated to be 4 mGal. For time-lapse measurements an additional uncertainty of 1 - 2 mGal is associated with the reference null level. The final detection threshold for Sleipner therefore is estimated at about 5 mGal.
A repeat gravity survey is planned for the summer of 2005, with a projected 8 million tonnes of CO2 in the plume. The additional 3 million tonnes of CO2 are expected to produce a gravity change of between about -10 and -43 mGal depending on density. Such a change should theoretically be detectable. In the event that acceptably accurate measurements are obtained, it will be possible to derive the average density of CO2 in the plume. This will help to constrain plume temperatures, which will in turn reduce uncertainty in the seismic analysis.
Was this article helpful?