Abstract
Water drive in a homogeneous porous medium with a uniform permeability distribution has been extensively studied in the past, both theoretically and experimentally. In this paper equations are derived for a two-dimensional water drive in a porous system with either a continuously or a porous system with either a continuously or a discontinuously varying permeability distribution perpendicular to the layer. The influence of perpendicular to the layer. The influence of capillary forces bas not been taken into account. A necessary condition for the validity of the equations is that the water should underrun the oil. It is shown that the permeability distribution has much influence on the oil recovery. A large difference in recovery was apparent from a comparison of three systems, in which the oil production decreased as follows:(1)permeability production decreased as follows:(1)permeability increasing in an upward direction perpendicular to the layer,(2)homogeneous, uniform permeability throughout,(3)permeability increasing in a downward direction perpendicular to the layer.
Introduction
The recovery from a homogeneous, two-dimensional system in response to a normal water drive can be calculated with the help of theories developed by Beckers. He describes segregated flow of oil and water in which the water underruns the oil due to gravity and viscous forces. In his approach it is impossible to include capillarity, which means that the system does not describe a transition zone. However, we have a great deal of scaled experimental evidence which shows that moderate initial transition zones disappear during the displacement process where gravity and/or viscous tonguing effects occur. The experiments then show a sharp interface, and a transition zone is only found in the top of the water tongue. Neglect of this small zone affects breakthrough-time calculations, but on the other hand, gives the advantage of being able to treat the problem with the end-point permeabilities only. Although not immediately recognizable, there is a form of crossflow involved in this concept. It can be found from the material balance if the layer is divided into two sections parallel to the bedding plane. The theory describes experiments with a maximum initial transition zone of about one-third of the layer thickness. There are no restrictions on the mobility ratio. The production curve predicts too early a breakthrough, but shortly thereafter very satisfactory agreement is found.
For many practical cases, these theories can be reduced to the simplified formulation given by Dietz. Experimental verification shows that his theory describes horizontal scaled-model experiments satisfactorily for mobility ratios larger than 6. Although gravity forces are completely neglected in the theory but not in the experiments, only the viscous forces are now responsible for tongue forming. Neglect of gravity delays breakthrough in comparison with predictions from the Beckers theory, but generally gives a somewhat better though not correct prediction of the moment of breakthrough.
These findings have encouraged us to apply the same principles to inhomogeneous systems. To delay the theoretical breakthrough, gravity forces parallel to the bedding plane have been included, parallel to the bedding plane have been included, which results in a in a better fit with the over-all production curve for tilted layers. production curve for tilted layers. The equations derived in this paper can be applied to systems with either a continuous or a discontinuouspermeabilitydistributionperpendicular to the layer They are therefore applicable to a great number of inhomogeneous systems. Areal extension can be effected with sweep efficiency factors or by applying stream-tube models.
THEORETICAL RESULTS AND DISCUSSION
In order to arrive at an analytical expression for the oil production from an inhomogeneous system, the following assumptions have been introduced.
1. Water flows underneath the oil.
SPEJ
P. 211
Global relationships between two-dimensional water wave potentials
