We present two versions of a Darcy-scale model for simulation of the solution gas drive in heavy oils.

Presence and behaviour of a foamy-oil effect appears to be critical to the cold production process. This process is not a well-understood production mechanism because a wide range of different petrophysical parameters and experimental factors interact in a rather complex way. Over the past few years, a number of efforts have been made, in many institutions, in order to understand and model the solution gas-drive mechanism in primary heavy oil recovery. Conventional simulations succeed in matching actual field productions but are not reliable for prediction forecast purposes (large uncertainties on recovery factors). We present an evolving nucleation and flow model for solution gas drive in heavy oil.

The model is built at the Darcy scale and involves the effect of capillarity on bubble flow and phase change. An ad hoc simulation tool is built to analyze carefully the properties of the model. The tool allows to reproduce laboratory experiment across a sample rock. In the limit of infinitely slow depletion, an asymptotic theory is constructed that allows to predict analytically the pressure difference across a sample. The theory predicts a pressure gradient at first order without additional calculation, and the gradient appears, at first order, independent on the gas formation (degassing) kinetics. No saturation gradient is predicted at first order. The results are compared to numerical simulations and experiment, and good agreement is found in cases where the pressure gradient is known.

In a second version of our model we introduce the effect of capillarity on bubble growth. Capillarity delays bubble growth by stabilizing small bubbles under a certain critical radius. This new term leads to nucleation fronts propagating rapidly through the sample.

It appears more difficult to predict a saturation gradient. The causes of this difficulty are analyzed.

You can access this article if you purchase or spend a download.