CO2 plume migration and fate at Sleipner, Norway: Calibration of numerical models, uncertainty analysis, and reactive transport modelling of CO2 trapping to 10,000 years

摘要

The Sleipner Project in Norway is the world's first industrial-scale geological carbon dioxide storage project. Time-lapse seismic monitoring data have been collected, tracing CO2 plume development from 1996 to 2010. Therefore, the Sleipner Project provides a somewhat unique opportunity to simulate the dynamics of CO2 in a real geological system. The purpose of this study is to simulate CO2 plume migration dynamics and assess the impact of uncertain factors on short and long term migration and fate of CO2 for the uppermost layer (Layer 9) of the Utsira Sand. First, we applied a multi-phase compositional simulator to the Sleipner Benchmark model for Layer 9 and calibrated our model against the time-lapsed seismic monitoring data at the site from 1999 to 2008. By adjusting lateral permeability anisotropy, CH4 in the CO2 stream, and reservoir temperature, approximate match with the observed plume was achieved. Model-predicted gas saturation, thickness of the CO2 accumulation, and CO2 solubility in brine (none of them used as calibration metrics) were all comparable with interpretations of the seismic data in the literature. Second, hundreds of simulations of parameter sensitivity (pressure, temperature, feeders, spill rates, relative permeability curves, and CH4 content) were conducted for the plume migration, based on the calibrated model. The results showed that simulated plume extents are sensitive to permeability anisotropy, temperature, and CH4 content, but not sensitive to the other parameters. However, adjusting a single parameter within the reported range of values in the literature would not reproduce the north-south trending CO2 plume;; it took a combination of permeability, CH4, and temperature adjustments to match simulated CO2 plume with seismic monitoring data. Although there is a range of uncertain parameters, the predicted fate of CO2 fell within a narrow band, ~ 93±2% structural trapping and ~ 7±2% solubility trapping. The calibrated model is not unique. Many combinations of permeability anisotropy, temperature, and CH4 would produce similar matches. Other possibilities that would have improved the development of an N-S elongated CO2 plume, such as a slight tilting of the surface of Utsira top to the south, were not experimented in this study, but are worthy of exploration for future studies. Finally, we used coupled reactive mass transport model to investigate the effects of rate laws and regional groundwater flow on long-term CO2 fate in Layer 9. The mineral composition and brine chemistry for the Utsira sand were adopted from the literature, and we modelled 100 year injection and continued water-rock interaction to 10,000 years. The results indicated that: (1) The predicted fraction of CO2 mineral trapping when using the linear rate law for feldspar dissolution is twice as much as when using the non-linear rate law. (2) Mineral trapping is more significant when regional groundwater flow is taken into consideration. Under the influence of regional groundwater flow, the replenishment of fresh brine from upstream continuously dissolves CO2 at the tail of CO2 plume, generating a larger acidified area where mineral trapping takes place. In a Sleipner like aquifer, the upstream replenishment of groundwater results in ~ 22% mineral trapping at year 10,000, compared to the ~ 4% when the effects of regional groundwater are ignored.