Gas liberation in porous media – experiments and modeling
Due to the Corona situation the seminar will be held virtually.
If you wish to follow the seminar you will have to sign up by sending an
e-mail to Christian Ove Carlsson cc@kt.dtu.dk
hereafter you will receive an invitation to join the virtual seminar.
Abstract
The liberated free-gas phase, generated due to reservoir pressure decline, affects the oil production negatively.
Disconnected immobile gas bubbles can block individual pore throats, lowering the reservoir's effective permeability to oil.
Unfavorable pore geometry and capillary forces/wettability balance are the main factors governing this phenomenon.
This work presents a novel experimental method to measure oil relative permeability outside the active saturation region.
The experiments involve gas liberation in low-permeability rock samples, induced by pressure decrease. We study the effect of immobile gas on the oil relative permeability.
Also we measure the critical gas saturation, after which the gas becomes mobile. The experiment was conducted on six North Sea chalk samples and two binary fluid mixtures.
The experimental results show that even a small amount of immobile gas for North Sea chalk rock may result in a significant reduction in oil relative permeability (up to 75%).
The relative permeability function outside the active saturation region is a power-law type function of the gas saturation.
Along with the experimental work, a model for relative permeabilities under two-phase flows in porous media is derived. The dependencies for the relative permeabilities on the saturation are derived based on a pore-level model of the porous medium, represented as a capillary network.
The distribution of the bubbles or droplets in the network is computed by applying a fundamental method from statistical physics. A model formula for single capillaries' conductivities, depending on the numbers of bubbles or droplets in them, is converted to the whole lattice's relative permeabilities by applying the effective medium formalism.
A comparison with the available experimental data indicates the excellent performance of the model. The experimental data may be fitted by variation of a single parameter, and the data from similar rocks have similar parameter values, which indicates the model's physical soundness.
Application of the model to the experimental dependences obtained in the first part of this work makes it possible to spread the oil relative permeabilities outside the immobile gas region, and moreover, to predict also the gas relative permeabilities for the whole range of saturations.
Acknowledgment
This study has been performed in the framework of a project supported by the Danish Hydrocarbon Research and Technology Center (DHRTC, Technical University of Denmark). Duc Thuong Vu and Tran Tsuong Dang are kindly acknowledged for invaluable help in the laboratory.