Abstract The year 2006 marks the first time the gas injection EOR production has surpassed that of steam injection processes in the US, signifying the growth of gas injection as a mature technology. In order to control the rising tendency of injected gas in horizontal floods, the water-alternating-gas (WAG) process has generally been the mode of operation in many fields. In spite of its wide application, the WAG process has not lived up to its expectations with reported recoveries in the range of 5–10% OOIP. In order to improve recoveries, we have been attempting to develop the gas-assisted gravity drainage process at LSU. This paper summarizes the effort of conducting scaled physical model experiments in a visual glassbead-packed model aimed at discerning the influence of some scaled dimensionless parameters, such as the capillary number, Bond number and gravity number, on the GAGD process performance.
A 2-D physical model, of 16" X 24" X 1" dimensions, packed with uniform glass beads, was used to conduct visual experiments. These experiments were so designed as to mimic the dimensionless parameters observed in some field projects. The secondary mode GAGD floods yielded recoveries up to 80% OOIP. Additionally, the recoveries displayed a semi-logarithmic relationship with gravity number (ratio of gravity to viscous forces). Interestingly, this relationship was observed to hold good for the high-pressure GAGD corefloods and even the field production data from gravity-stable gas injection projects conducted in pinnacle reefs. A multi-variable regression analysis of the laboratory as well as field data indicated that the Bond number, being the ratio of gravity to capillary forces, had a greater influence on GAGD performance than other parameters. In addition to the observed high recoveries, our attempts to relate the model run times to field project durations, through dimensionless time considerations, have indicated reasonably good rates of production when GAGD process is implemented in field projects.
1. Introduction 1.1 Background EOR surveys by the Oil and Gas Journal for the last two decades clearly show the increased popularity and production share of gas injection processes in the U.S. Among the gas injection EOR processes, CO2 as well as hydrocarbon processes, demonstrate higher potential as an effective tool to recover the 'left-behind' oil. The recent record crude oil/natural gas prices, as well as increased greenhouse gas emission concerns, tip the scales in favor of CO2 as the most favorable enhanced oil recovery tool.
The important functions that any EOR process needs to perform to be successful are:increase the microscopic displacement efficiency by increasing the capillary number, and
attain better volumetric sweep efficiency by improving the mobility ratio (M) (Green and Willhite, 1998).
The microscopic efficiency, defined as extent of mobilizing the trapped reservoir residual oil, is a function of the capillary number (Nc), where Nc is the ratio of viscous to capillary forces. On the other hand, the volumetric sweep, defined as the percent of reservoir rock contacted by the injected fluid, is governed by the mobility ratio and reservoir heterogeneity.