1. Two reconstruction methods were tested. When compared to experimental data (absolute permeability and diffusion/dispersion), both revealed themselves as effective. A new NMR method (PGSE) allowing to characterize diffusion/dispersion in porous media was developed. Those measurements were interpreted by numerical simulations carried out in reconstructed structures.
2. Experiments of film thickness measurements put into evidence the effect of salinity on spreading behaviour. Flow experiments in capillary tubes allowed to determine capillary pressure threshold as a function of spreading.
3. Experiments in network micromodels were carried out under various conditions of spreading and wettability in 2D and 3D dimensions. They allowed to picture out the rôle of those two properties on fluid flow mechanisms and on core flood parameters.
4. A three dimensional three-phase network simulator accounting for viscous pressure drop in both bulk phases and films and for capillary effects was developed. It incorporates all the experimental findings at the pore scale. Relative permeabilities in drainage mode were calculated.
5. Core flood experiments lead to demonstrate that the Stone model was inappropriate to model three-phase flow in porous media. In particular, it was shown that three-phase relative permeabilities were functions of the three saturations. The effect of spreading and wettability was studied on relative permeabilities.
6. A new three-phase flow model was established based on a fractal representation of the pore structure. It was successfully compared to laboratory data. The basis of another model based on the pore scale distribution of fluid phases was developed. The specificity of this model is to account for hysteresis effects.
Recovery by gas injection seems to be very promising because of its efficiency, its moderate cost and its applicability in either thick or fractured reservoirs, such as those located in the North Sea. This leads to specific three-phase flow mechanisms that have to be carefully studied in order to better control and optimize recovery processes.
Beside the oil industry, the environmental industry is also interested by three-phase flow in porous media to control accidental pollution of the aquifers. Instead of developing empirical models with numerous adjustable parameters, it is intended to find a new modellig vased on physical phenomena and mathematically well-posed.
Our approach will be to carefully study, experimentally and numerically, the flow events at the pore scale and then derive a phenomenology on this scale. Then, up-scaling will be achieved, both theoretically and numerically, by developing a simulator integrating the pore scale phenomenology. It will provide, by comparison with experiments performed in real rocks and in synthetic 3D porous media models made in glass or resin, a new phenomenology firmly stated, physically and mathematically.
To reach these objectives, six partners (4 industries and 2 research centres) from four countries (NL, FR, GR and UK) are associated during 4 years, with a total budget of 4.3 MECU and a manpower of 38 years, funded at 58% by the Commission.
Funding SchemeCSC - Cost-sharing contracts
DT2 8DH Dorchester
TW16 7LN Sunbury On Thames