Gravity Segregation Between Gas and Water During WAG Injection In Non-Horizontal Reservoirs

WAG (Water-Alternating-Gas injection) is a non-thermal EOR technique proposed to improve the volumetric sweep efficiency and consequently the oil recovery during a gas injection project. Though miscible gas injection renders brilliant displacement efficiency due to its

miscibility with oil, it usually shows very poor volumetric sweep efficiency due to its high mobility. The mobility of gas is lowered by an alternate injection mode of gas-water, and gas fingering effect is therefore mitigated. Aside from the efet of gas fingering, gravity segregation

is another destructive factor to gas sweep efficiency in the field. The injected gas water mixture disappears eventually when it goes further away from the injection well, due to gravity-driven gas-water segregation provided with large density contrast. Gas tends to segregate to the top of reservoir, and flows towards production well at immobile water saturations. This effect is what we call "gas-channeling". The oil displacement below this channel will not benefit from the effect of miscible gas. Therefore, production wells should be preferably placed within the range of complete segregation, aka, before the mixed zone of gas-water completely disappears.

Due to this particular reason, Stone and Jenkins derived a physical model aiming at dipicting the segregation behavior between gas and water. Stone's model (1982) predicts the distance of complete segregation at steady state, and Jenkins (1984) derived a correlation for the thickness of override and underride zones. Rigorous proof for Stone's model has been provided by van Duijn(2010). These models, as accurate as they are, only works for WAG injection in horizontal reservoirs. For WAG injection in a reservoir with tilting angle, however, segregation process along the dipping direction is to some degree interrupted and altered. Dr. Namani (2012), on basis of Stone's assumptions, derived a physical model for gravity segregation in tilting reservoirs. The main goal of this research is to examine the validity and accuracy of Namani's model with the assistance of numerical simulation. At the end of the research, we discovered that Namani's model is conditionally accurate. The behavior of gravity segregation shown from simulation results, exhibits a rather uniform yet very different pattern from our prior expectations. Additionally, some very interesting physical effects were observed during transient stage of WAG injection, which may very well enhance the benefits of miscible gas injection and altimately, lead to incremental oil production.

Guanqun (Brandon) Yu