Eagle Lake Viking Miscible Displacement Laboratory Studies

Abstract

Abstract A series of displacement tests was conducted on preserved reservoir core plugs to assist in estimating waterflood and miscible flood performance potential for the Eagle Lake Viking Reservoir in Saskatchewan. The reservoir rock is a heterogeneous sandstone, comprising thin interbeds of sand and shale with the sand lenses containing widely different amounts of smectite clays. To assess miscible recovery efficiencies in this complex sand, displacement tests were conducted with first-contact miscible propane and carbon dioxide solvents. Using standard waterflood relative permeabilities and a dual permeability model, all miscible displacements were matched by simulation with a Todd-Longstaff-type miscible mixing model. Introduction The Eagle Lake Viking Reservoir is located in southwest Saskatchewan as shown in Figure 1. It is a sandstone formation of Mississippian age. Discovered in 1957, one half of this 12.7 million m3 field was put on waterflood ten years later in 1967. Unfortunately, because of its low permeability, it has been possible to only inject 0.14 pore volumes of water in the subsequent 2O-year period. As a result, in the thirty years since discovery, only 15% of original oil- in-place has been produced. The cause of poor performance appears to be due to the abundant shales and days. Thin section analyses, however, show that bands of higher permeability are also found throughout the pay zone and often exist as thin to very thin sand stringers which are relatively free of days. In 1985, operations engineers decided that the half of the field which was still under primary depletion should be placed under some form of secondary recovery. Because the waterflood had been less than satisfactory, it was decided to examine miscible flooding as a potential alternative. Hence, a core flooding study using carbon dioxide as a flooding agent was designed with the following objectives;quantify carbon dioxide oil displacement efficiency;compare carbon dioxide with waterflood oil recovery; andcompare the potentially complex carbon dioxide miscible oil displacement process with a more conventional hydrocarbon solvent process using propane. To extend the initial objectives for reservoir design purposes, a fourth objective was to:obtain waterflood and miscible displacement parameters for use in numerical simulation of field applications, Reservoir Fluid and CO2 Miscibility Pressure The live oil used in this study was obtained by recombining separator gas and liquid to a bubble point pressure of 6509 kPa at 22 °C. Fluid densities were measured as a function of pressure at the reservoir temperatures of 22 °C and were matched with the Peng-Robinson(l) equation-of-state. The reference 2 characterization procedure was used and viscosity predictions were made with the Jossi, Stiel and Thodos correlation. Table 1 illustrates the accuracy or this prediction which involved proprietary correlations for heavy end parameters. This set of density and viscosity data was used in the simulation study. The oil composition is also shown in Table 1. To determine the first-contact miscibility pressure for CO2, oil and CO2 were blended in a visual cell at 22 °C. These results and the Peng-Robinson equation-of-state matches are presented in Table 2.