Study Investigates Smart Waterflooding in Sandstone Reservoirs
In this paper, the authors consider the effect of water chemistry on water/rock interactions during seawater and smart waterflooding of reservoir sandstone cores containing heavy oil.
In this paper, the authors consider the effect of water chemistry on water/rock interactions during seawater and smart waterflooding of reservoir sandstone cores containing heavy oil. Oil recovery, surface-reactivity tests, and multicomponent reactive-transport simulation were conducted to understand smart waterflooding better. The authors determine that lowering magnesium-ion (Mg2+) concentration results in lower additional oil recovery, and that lowering calcium-ion (Ca2+) concentration leads to higher additional oil recovery.
“Smart water” (SW) can be defined as a water engineered by manipulating the ionic composition (adding or removing ions) regardless of the resulting salinity. An upward pH shift has been reported in the literature during low-salinity (LS) and/or smart waterflooding. That upward shift modifies the wettability of sandstone, thereby increasing oil recovery.
When high salinity is displaced by LS water, Ca2+ will be desorbed from the clay surface because ion exchange occurs between hydrogen (H+) and Ca2+ at the negative sites of the clay surface. The organic materials will be then desorbed from the clay because of the reaction between the protonated basic and the acidic materials with hydroxide. The authors’ investigations on free-clay sandstone provided a different insight into such reactions. The observations from chromatographic columns of quartz showed an increase in the acetate detachment from the quartz surface with a significant pH jump during LS waterflooding. The oil-recovery observations from the quartz column supported the proposed mechanism.
Previous systematic laboratory works indicated that the abundance of Mg2+ in the formation water (FW) is more effective than Ca2+ for a higher LS enhanced-oil-recovery (EOR) effect. Conversely, the abundance of Ca2+ in the injected LS water has a larger effect on LS EOR. When Mg2+ was present in the LS water, no oil-recovery improvement occurred.
Brines. Synthetic FW and SW were prepared by dissolving sodium chloride (NaCl), calcium chloride (CaCl2), and magnesium chloride (MgCl2) in deionized water.
Crude Oil. A reservoir crude oil was delivered from eastern Kansas from the Bartlesville sandstone reservoir. The crude oil had a viscosity of approximately 600 cp at ambient temperature. The density is 0.83 gm/cc at 20°C.
Reservoir Rock Description and Core Preparation. Five Bartlesville sandstone reservoir cores were drilled from the original 3-in. diameter to a 1-in. diameter preserved core from eastern Kansas. The preserved reservoir sandstone cores were delivered with 3-in. diameter. The most abundant mineral was quartz, followed by clays.
Results and Discussion
Coreflood Experiment 1. The reservoir core (RC2a) was successively flooded with FW and SW. FW comprised 1802, 420, and 39 mM/L of CaCl2, MgCl2, and sodium sulfate (Na2SO4), while SW comprised 135, 1260, and 316 mM/L of CaCl2, MgCl2, and Na2SO4. Seawater salinity is approximately half the salinity of the FW. The ultimate oil recovery during FW was 50% of original oil in place (OOIP). Flooding SW into the RC2a resulted in an additional oil recovery of 5.2% of OOIP. During FW flooding, the injection pressure increased quickly until reaching 135 psi. The injection pressure then dropped and stabilized at 54 psi. The injection pressure was then kept at that value after the injected fluid was switched to SW. The effluent pH for the FW was 6.55 and for the SW was 7.35. The pH (all experiments) of the effluent water is shown in Fig. 1.
Coreflood Experiment 2. The reservoir core (RC2b) was sequentially waterflooded with 5 pore volumes (PV) of FW followed by 3 PV of SMW1 (FW/2). The concentration of CaCl2, MgCl2, and Na2SO4 in SMW1 was reduced by half. The oil recovery during FW was 49.2% of OOIP. No additional oil recovery was observed during SMW1 injection. The injection pressure during FW flooding jumped to 141 psi; the pressure then declined dramatically and stabilized at 72 psi. After flooding the core with SMW1, the pressure stayed at that range.
Coreflood Experiment 3. SMW2 was depleted in divalent cations (Ca2+ and Mg2+) while keeping the salinity the same by dissolving NaCl. The oil recovery by injecting 5 PV of FW into RC2c was 51.4% OOIP, which was approximately the same as the previous two coreflood experiments. As expected, a significant increase in oil recovery of 10.35% OOIP [which was much higher than in SW flooding (10.35% vs. 5.2%)] was observed when switching the injected brine to 3 PV of SMW2. The injection pressure readings were lower than the previous two experiments. The pH jumped from 6.51 during FW flooding to 7.93 during flooding of SMW2.
Coreflood Experiment 4. FW was also diluted twice but depleted in MgCl2 and Na2SO4 to investigate the effect of Mg2+ and sulfate (SO42–). The resultant salinity of SMW3 was kept constant (52,275 ppm) by dissolving extra NaCl. RC2d was successively flooded with 5 PV of FW followed by 3 PV of SMW3. The same previous flooding procedure was conducted in this experiment. During the FW flood, 50.5% of OOIP was recovered. An additional 3.1% of oil recovery was observed during the SMW3 flood.
Coreflood Experiment 5. In this experiment, FW was kept constant at 52,275 ppm but the concentration of Ca2+ was diluted 100 times in SMW4. The oil recovery during FW flooding was 50.1% OOIP. The SMW4 flood resulted in an additional oil recovery of up to 8.5%, representing the second-highest improvement among all the previous experiments except when using SMW2. The pressure readings followed the same coreflood as in Experiment 3.
Discussion of Coreflood Results. Salinity of SMW2 was a bit higher than in SW, but because SMW2 was softened, the oil recovery using SMW2 was higher than that seen in SW. The same occurred with SMW1. SMW1 and SMW2 have the same salinity, but because SMW2 was depleted in divalent cations, the recovery was zero during SMW1 while it was significant during SMW2. There is no EOR effect during SMW1 because of the very high concentration of Ca2+ (901 mM/L CaCl2).
The high concentration of Mg2+ in SW (1260 mM/L MgCl2) does affect EOR. The pH measurements also supported the results of oil recovery. SMW2 and SMW4 exhibited the highest pH among the other injected water effluents. The injection pressure during SMW2 was much lower than that of the others because there was no Ca2+ in the injected brine. A previous study by the authors stated that high Ca2+ concentration led to edge-to-face agglomeration of the kaolinite plates to form higher-volume assemblages and reducing permeability. Injecting pressure during SMW4 flooding (containing 9 mM/L Ca2+) supported their theory about Ca2+ and pressure relationship. The authors aimed to keep Ca2+ concentration constant to show the effective role of Ca2+ in the injected SW, or even LS, water. Abundant SO42– in the sandstone prevents increase in pH, which is required to observe the EOR effect, but when SO42– is removed, a small amount of incremental oil recovery is observed. Mg2+ was observed to have a significant effect in the absence of Ca2+ in both FW and injected water.
Zeta Potential Measurements. Zeta potential for rock/brine and oil/brine interfaces was measured for the same sandstone, oil, and brines used in this study. The test results indicated that depleting or diluting Ca2+ in the SW results in higher negative charges, which increases the repulsive forces between the interfaces of rock/brines and oil/brines; thus, the rock turns further water-wet, meaning that reduction of Ca2+ concentration results in wettability alteration. Depleting Mg2+ in SW also results in a negative charge, but not as high as in the case of Ca2+.
Reactive Transport Modeling of Water/Rock Interaction
Recently, reactive transport modeling (RTM) has been used to anticipate weathering effect. A software package for simulating reactive transport was used that is based on governing coupled partial differential equations with a finite-volume discretization. The same experimental data of this study were used and a new code was developed. In this study, a reactive transport code was used to simulate the reaction and transport processes during flooding Bartlesville sandstone reservoir with SW, SMW1, SMW2, SMW3, and SMW4. The species concentration can then be calculated. This process is detailed in the complete paper.
In this study, tertiary seawater and many kinds of SW were investigated. The oil recovery observed was categorized on the basis of the ionic composition of the waters. The experimental studies confirmed that additional oil recovery can be obtained by diluting or depleting Ca2+ instead of the other ions. The authors believe that the more Ca2+ and Mg2+ (positively charged ions) are injected, the greater the increase of the oil wetness of the sandstone, and oil recovery will decrease. Decreasing Ca2+ will decrease the high surface complexes, and greater oil recovery will be obtained. Decreasing Mg2+ is also a beneficial factor that could provide more oil recovery, but not to the same extent as Ca2+ because Mg2+ cannot sorb as closely from kaolinite edges as does Ca2+. Reactive transport modeling supported the coreflood experiments. As the concentration of Ca2+ decreases, oil recovery increases. The same behavior was observed for Mg2+, but at a lower scale. SW is more effective and useful than LS water and could be a cost-effective EOR technique.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 193238, “Investigation of Smart Waterflooding in Sandstone Reservoirs: Experimental and Simulation Study, Part 2,” by Hasan N. Al-Saedi, SPE, Missouri University of Science and Technology and Missan Oil Company; Ralph E. Flori, SPE, Missouri University of Science and Technology; and Alsaba Mortadha, Australian College of Kuwait, prepared for the 2018 Abu Dhabi International Petroleum Exhibition and Conference, 12–15 November, Abu Dhabi. The paper has not been peer reviewed.