Considering more than 50% of oil is left on average after primary and secondary methods in conventional reservoirs, techniques to recover the remaining oil from these proven and mature reserves is critically important to meet the future energy demand. Miscible solvent (usually gas) injection is performed in the form of water-alternating gas rather than continuous injection of expensive injectant. In this case, the process should be optimally designed for slug sizes and proper sequences of the solvent considering several controlling parameters including past history of waterflooding (water content) and wettability.
We performed a set of experiments on vertically situated sand pack models saturated with 14cp oil. Heptane was selected as the solvent phase due to high miscibility conditions. Tests were started with waterflooding or solvent injection. Different slug sizes of solvent and water were tested on fully oil- saturated water, oil-wet samples, and samples with connate water (10 and 30%). The amounts of oil and solvent retrieved were monitored using refractometer. Using the collected data, the recovery rates and ultimate recoveries were comparatively analyzed. In addition to the technical feasibility, an economic analysis was performed considering the amount of solvent injected, oil and solvent recovered, and time for recovery.
Starting the process with heptane was technically and economically feasible (pay-out time is shorter) in the short run for both the oil- and water-wet cases. This was true especially if the rock was oil-wet, which yielded faster recovery and higher ultimate recovery. Excessive water injection (up to the plateau level) preceding the solvent injection in the oil-wet case resulted in lower recovery factor whereas this design was very effective in the water-wet case. Therefore, the time to switch to solvent injection was critical in the oil-wet case and a short initial cycle of solvent injection followed by short waterflooding cycle is suggested. In the oil-wet case, initial waterflooding resulted in an inefficient process; while it yielded high ultimate recoveries, the process time was longer than other injection options. In the water- wet case, a greater amount of solvent was needed in the first cycle.