ASSESSING THE RISKS OF GEOLOGICAL STORAGE OF CO_2 IN MATURE OIL FIELDS

摘要

Geological storage of CO_2 in mature oil fields is an interesting solution to mitigate greenhouse effects in a planet with an increasing demand for energy. The precise assessment of the risks involved in the CO_2 storage is, however, necessary before such method be widely used. One of the main risks of the CO_2 storage is the dissolution, and eventual transport of CO_2-charged waters in the aquifer, away from the reservoir. To access the efficiency of the storage, a numerical model of the geological system involved is developed. The Forties Field (offshore North Sea) was selected within the NGCAS project because of the availability of suitable input data, absence of large faults, and good sealing capacity of the cap-rock. In order to produce an accurate model, capable to consider both the fluid flow and the dissolution rates of CO_2 in water, we defined a multi-scale approach: 1. Simulation of the fluid flow in large scale (area of tens of km), around the target area; and with the results of a 2D regional model as input data, 2. Simulation of fluid flow in greater detail, at reservoir scale, and of the interaction between CO_2 and water. Results of the numerical modeling enable to evaluate present-day and future risks of the CO_2 storage in Forties. Such an approach might also be used for the evaluation of risks of CO_2 storage in other mature oil fields as well. To assess the storage safety, several scenarios are investigated such as diffusion within the cap-rock or leakage through abandoned wells. The CO_2 transfer rate out of the storage is investigated. The key parameters for CO_2 long-term storage are the cap-rock capillary barrier (displacement pressure), which holds gas or supercritical CO_2 within the storage, and CO_2 effective diffusion coefficient within the cap-rock layers, which enables dissolved CO_2 to escape by diffusion within the aqueous phase. In the base case scenario, CO_2 could diffuse within first 50 meter of the cap-rock over the 1000-year period of the simulation. Whereas in a (extreme) worst case scenario approach, CO_2 would migrate (diffusion and convection) within 350 meter of the cap-rock over the 1000-year period of the simulation.