Miscible gas injection in oil fields in general results in a better microscopic displacement efficiency compared to water injection; however, higher mobility ratio and lower macroscopic sweep of gas is a major disadvantage compared to water injection. Considering this situation, water alternating gas (WAG) was proposed and implemented as a method to improve macroscopic sweep efficiency of gas using cyclic water injection to reduce its high mobility ratio.
In this simulation study, a systematic parameter study was carried out to investigate the effect of different parameters in a miscible WAG process using a horizontal sector model and both black-oil and compositional simulators. The base model was first built using a black-oil simulator but a compositional model was later implemented in order to investigate the differences in results for the two models.
This study shows that WAG injection is advantageous compared to both water and gas injection in term of oil recovery factor. Changes of cycle time did not affect the recovery factor significantly while in this particular model recovery is sensitive to the changes of gas half cycle time. It is not possible to find a simple dependency between half cycle time and oil recovery factor but it is obvious that recovery factor is a function of both gas slug size and half cycle time and that an optimum value exists for each case. The effect on recovery by changes in water half cycle time is on the other hand not significant.
Injection of water through all layers improved the advance of stable front and improved final oil recovery for all cases. However, the gas injection interval needs to be optimized for each case based on the ratio of vertical and horizontal permeabilities of the reservoir, and thus the injection perforation strategy for gas must take this into consideration.
In the model used, which has no aquifer or gas cap, increased dynamic forces by simultaneous increase of field production and injection rates resulted in increased oil recovery.
There is good consistency between black-oil and compositional base models with only ±4% OIIP differences in term of field oil recovery.