Three-Phase Relative Permeability of Water-Wet Berea

Abstract

SPE Member Abstract Two-phase and three-phase relative permeabilities of water-wet Berea sandstone have been studied experimentally. Eight cases of saturation history have been investigated. About 1800 data have been collected using the steady-state method: about half of them for two-phase flow and the other half for three-phase flow. The experimental results are compared with predictions by Stone's methods. Stone's predictions of three-phase oil relative permeability are unsatisfactory. The experimental permeability are unsatisfactory. The experimental data do not support the assumption that two-phase relative permeability data can be used to predict three-phase relative permeability with a similar saturation history. The data suggest that displacement of trapped oil by gas is an important factor affecting three-phase oil permeability. It is also suggested that there exists a critical oil phase saturation to initiate three-phase flow from wateroil two-phase flow, which can be different from that in the two-phase flow. The prediction methods fail to account for these factors. Introduction Reservoir studies require three-phase relative permeabilities when three-phase flow takes place as permeabilities when three-phase flow takes place as in CO WAG flooding, steam flooding, or other enhanced recovery processes, but laboratory measurement is almost prohibitively difficult. As an alternative, Stone's prediction methods (commonly called Stone's first method and second method) are widely used in the oil industry to estimate three-phase relative permeabilities from the more readily available two-phase data. But the estimated values by Stone's methods are sometimes unrealistic and at best uncertain. This problem has been long recognized by many investigators, but has remained unresolved. It is the purpose of the present study to help resolve this problem. About a dozen groups of investigators ave made experimental measurements on three-phase relative permeabilities. Rut most of the data are not suited permeabilities. Rut most of the data are not suited to investigation of the prediction methods. One reason is that the data reported are incomplete on two-phase relative permeabilities. Another reason is that the saturation history is inconsistent or limited to a couple of saturation history cases. This paper presents two-phase and three-phase relative permeabilities of water-wet Berea sandstone collected for eight history cases, two cases of which have been presented elsewhere. The experimental results are compared with predictions by Stone's methods. Based on that, limitations of Stone's methods are discussed. EXPERIMENTAL APPARATUS AND PROCEDURE The steady-state method was used. It was automated to enhance accuracy and reproducibility of the data, as well as the speed of data collection. The apparatus and procedures, which have been detailed in a previous publication, are outlined below. previous publication, are outlined below. The core was placed between two end-section cores made of the same core material as the center core (Penn State method). Two sheets of unscented Kleenex were inserted between the core and the end sections to insure capillary contact. Relative permeabilities were measured on the center core to permeabilities were measured on the center core to minimize the capillary end effects. The water, oil and gas injection devices, and pressure and saturation measuring devices were pressure and saturation measuring devices were interfaced with and controlled by a personal computer (PC). P. 109