Per Bergmo

The Long-Term Fate of CO2 Injected into an Aquifer

Publisher Summary More than 4 million tones of CO2 from the Sleipner gas processing plant has been injected into the Utsira formation in the North Sea and eventually 25 million tones will be injected before gas production has ceased. The CO2 is injected near the bottom of the approximately 200 m thick sand. The migration of injected CO2 is being monitored by 4D seismic, which shows that the CO2 is retained in large thin clouds under what is expected to be horizontal semi-permeable shales within the formation. Assuming that an underground aquifer is capped by a capillary seal, preventing the injected CO2 from migrating into the atmosphere, the CO2 will eventually accumulate under this seal. The topography of the seal will determine the CO2 migration on the 1000 years time-scale. CO2 diffusing into the underlying aquifer column will set up convective currents in the aquifer, enhancing the solution rate of CO2. Most of the CO2 will have dissolved into the aquifer after 5000–50,000 years, mostly dependent on the vertical permeability of the aquifer and the contact area between CO2 and brine. On a time scale of several hundred thousand years, molecular diffusion through the capillary seal will be the dominating transport mechanism and will eventually deplete the formation of CO2 by transporting it into the atmosphere if reactions between CO2 and rock minerals are neglected.

Study of CO2 EOR in a Sector Model from Mature Oil Field, Cambay Basin, India

Among various Enhanced Oil Recovery (EOR) methods, gas injection has been proven to be one of the effective ways of enhancing oil recovery from mature fields. The field under study has approached the economic limit of production under conventional recovery methods (primary and secondary recovery). Since start of production in sixties, the field has produced 48.5 % of the initial oil in place and the water cut has increased to 89 % in April 2011. Responding to the industry needs, initially a comprehensive study was performed to evaluate the potential of immiscible CO2 injection for the recovery of residual oil after water flooding in this mature field. This paper presents the preliminary results of immiscible CO2 injection on the basis of laboratory studies and detailed compositional simulations carried out on a sector model of the field. Based on the results obtained from laboratory studies it was found that CO2 injection yields significant incremental recovery. Simulation results show significant increase in field oil production, essentially from 200 to 1100 m3/day and considerable decrease in water cut were observed. In addition, detailed PVT simulations were carried out to obtain an equation of state (EOS) that would better describe the phase changes in the reservoir. These results would form the basis for carrying out CO2 EOR simulations on a field scale.

CO2 Storage Capacity Estimates for a Norwegian and a Swedish Aquifer Using Different Approaches – From Theoretical Volumes, Basin Modelling to Reservoir Models

Open dipping aquifers might offer a unique possibility to store huge quantities of carbon dioxide. Many different modelling approaches have been used to quantify possible storage capacities often giving very diverse results. In this study, we applied three different methods to calculate and model theoretical volumes, structural trapping volumes using a basin modelling tool and capacities obtained from dynamic reservoir simulations. We tested end-member scenarios for different critical parameters. The results for two stratigraphic confined open/semi-closed dipping saline aquifers, the Garn Formation (Norwegian Sea, Norway) and the Faludden sandstone (Baltic Sea, Sweden) show broad variations. For the Garn Formation CO2 storage capacities vary from 2.0 to 8.4 Gt. Taking into accounts all results, we estimated a representative storage capacity ranging between 2.0 and 3.5 Gt. In the case of the Faludden sandstone the different modelled scenarios give a spread from 10 to 836 Mt and a representative capacity of 250–435 Mt was defined. We will show and discuss how the different estimates are calculated, how they are related to each other and finally exclude unreliable results. Furthermore we compare our results with published data from the same areas. This will demonstrate the complexity and difficulty of a direct comparison of geological CO2 storage estimates and pinpoint to the need for a general strategy to compare modelling results for geological CO2 storage estimates.

A Method for Ranking CO2 Flow Models Using Seismic Modeling and Time-Lapse Data

A method for discriminating between different reservoir flow models using forward modeling and time-lapse seismic is presented. A rock-physical model is used in order to generate synthetic time-lapse acoustic responses based on flow model predictions. From the acoustic properties a pull-down caused by modifications in acoustic velocity is calculated and compared to real measurements. Full synthetic seismograms are also generated. The method has been applied to the Sleipner CO2 sequestration project where time-lapse seismic is used to monitor the injected gas. Different vertical migration processes of the CO2 may explain the observed time-lapse response. In this chapter, this new methodology is used in order to discriminate between these processes.