Optimization of Low Salinity Water/Surfactant Flooding Design for Oil-Wet Carbonate Reservoirs by Introducing a Negative Salinity Gradient

: Engineered water surfactant ﬂooding (EWSF) is a novel EOR technique to reduce residual oil saturation; however, it becomes quite challenging to obtain Winsor Type III microemulsion and the lowest IFT under actual reservoir conditions if only low salinity water is used. The main objective of this study was to design a negative salinity gradient to optimize the performance of the hybrid method. Three coreﬂoods were performed on carbonate outcrop samples. The injection sequence in the ﬁrst test was conventional waterﬂooding followed by optimum engineered water injection (2900 ppm) and ﬁnally an EWSF stage. The second and third tests were conducted using a varying negative salinity gradient. Engineered water for this study was designed by 10 times dilution of Caspian Sea water and spiking with key active ions. A higher salinity gradient was used for the ﬁrst negative salinity gradient test. A total of 4300 ppm brine with 1 wt% surfactant was injected as a pre-ﬂush after waterﬂooding followed by a further reduced salinity brine (~1400 ppm). The second negative salinity gradient test consisted of three post-waterﬂooding injection stages with salinities of 4600, 3700, and 290 ppm, respectively. Up to 8% and 16% more incremental oil recovery after waterﬂooding was obtained in the second and third tests, respectively, as compared to the ﬁrst test. The descending order of brine salinity helped to create an optimum salinity environment for the surfactant despite surfactant adsorption. This study provided an optimum design for a successful LSSF test by adjusting the brine salinity and creating a negative salinity gradient during surfactant ﬂooding. A higher reduction in residual oil saturation can be achieved by carefully designing an LSSF test, improving project economics.


Introduction
Over 50% of the proven conventional hydrocarbon reserves are in carbonate reservoirs.The average recovery factor in carbonate formation is below 35% due to the complex structures, formation heterogeneities, and oil-wet/mixed-wet conditions resulting from their depositional history and later diagenesis [1,2].However, there is a high potential to increase oil recovery in carbonate reservoirs by using enhanced oil recovery (EOR) methods.The majority of carbonate field development plans that improve oil recovery by water and gas flooding still do not provide high ultimate recovery.The performance of conventional waterflooding needs to be enhanced further considering reservoir features.Experimental studies and field experiences indicate that the efficiency of waterflooding depends on the salinity and composition of the injected water.Low salinity waterflooding (LSWF) is a relatively recent and promising EOR technique with a few pilot-scale applications [3,4].Low salinity water can be obtained through the dilution of high saline seawater or formation water and optimization of the active ions (such as Ca 2+ , Mg 2+ , and SO 4 2− ) [5,6].Low salinity waterflooding may alter the wettability of carbonates toward the water-wet condition through proposed mechanisms such as multi-component ionic exchange (MIE) [7,8], pH variation, and associated rock dissolution [8], and double layer expansion [9][10][11].Surfactant flooding (SF) is a chemical EOR method that provides IFT reduction and wettability alteration [12].Surfactant flooding, together with LSWF, is a new, promising hybrid method of EOR [13].In past years, the application of surfactants with the integration of LSWF has acquired widespread interest among researchers.Most of the studies focus on sandstone formations [5].LSWF and SF experiments were conducted in carbonates separately, and a few research studies attempted to utilize this hybrid combination in carbonates together.The success of this method depends on several factors that are discussed in this work.High incremental oil recovery and low retention of surfactant, which decreases the project cost, are the leading indicators of the effective hybrid injection process.High oil recoveries from the porous medium can be attained at a high capillary number due to the ultra-low IFT values [14,15].
Surfactants that provide better IFT reduction at low salinity conditions are more accessible and more cost-effective than those that are practical in high salinity water.In addition, surfactant retention intensifies as salinity increases [16].Alagic and Skauge [17] reported a hybrid EOR project combining the influence of low salinity brine and surfactant flooding in a combined low salinity brine and surfactant (LSS) flooding process.The concept of this hybrid EOR method is that during LSWF, oil layers destabilize in a low IFT condition, which is provided by SF, that also prevents re-trapping.Alameri et al. [18] reported an increase in incremental oil recovery by more than 10% after the injection of LSS into carbonate core samples.
Surfactants or surface-acting agents are amphiphilic organic compounds that are soluble in both water and organic solvent.Surfactants consist of a hydrocarbon (non-polar) chain, which is also called a hydrophobic tail, and a polar or ionic portion, which can be simplified as a hydrophilic head [12].The hydrophobic tail interacts with an organic solvent, and the hydrophilic head part interacts with water.This process leads to the formation of oil-in-water and water-in-oil microemulsions, which depends on the balance between the lipophilic and hydrophilic groups of a surfactant [19,20].The IFT between oil and surfactant solution depends on the salinity, temperature, surfactant type, surfactant concentration, and oil composition [21,22].The salinity of the water has an impact on the phase behavior of the surfactant solution [23][24][25].The solubility of anionic surfactant in water becomes lower when the salinity of the formation water increases.Surfactants start to move from the water phase to the oil phase as the electrolyte concentration becomes high.Hence, when the salinity of brine increases, the surfactants go to the oleic phase from the aqueous phase.The optimum salinity lies somewhere in between the low salinity and high salinity wherein Winsor Type III microemulsion is obtained that is in equilibrium with oil and water and yields ultra-low IFT between water and oil phases [26,27].For an experiment, anionic surfactant and mono-ethanolamine salts of alkyl-orthoxylene sulfonic acid (MEACNOXS) were used.The used oil was a mixture of 10% of aromatic oil (N) and 90% of paraffinic oil (I).At optimal salinity, the equal solubilization of phases occurred, and the IFT of the system reached an ultra-low value; as a result, the residual oil recovery had the highest value [21].
Surfactant loss has a negative effect on the feasibility of a surfactant flooding process and increases the required amount of chemicals for a successful EOR method [28,29].The reasons for surfactant loss are adsorption, phase trapping, and precipitation processes that lead to low displacement efficiency of hydrocarbon [30,31].The surfactant may adsorb on the rock surface because of the electrostatic interaction between the surfactant and the rock [29].The surface charge of carbonates makes the selection of the appropriate surfactant type difficult.The main mechanism for the surfactant adsorption in carbonates is van der Waals and electrostatic interaction between the mineral surface and hydrophobic tail of the surfactant [32,33].Besides adsorption, surfactant precipitation in high salinity and high temperature conditions is another challenge for the applica-tion of anionic surfactant in carbonate formations.One of the reasons for surfactant precipitation is the presence of multivalent ions, which causes active surfactant and cosurfactant phase dissociation [34,35].There is a possibility to avoid phase trapping and precipitation of surfactants by regulation of brine salinity.Hybrid EOR methods are targeted to recover more residual oil in an economic, ecological, and technically feasible way by stimulation of multiple mechanisms simultaneously.LSWF promotes a suitable environment for effective surfactant flooding due to the detachment of oil molecules, wettability alteration, and reduction in IFT [36][37][38].Moreover, the combination of LSWF with surfactant flooding leads to the minimization of surfactant consumption and improves its solubility and stability, which in turn reduces the project cost.Those are reasons for the prioritising of the hybrid over the single EOR technique [39].
Numerous experiments have confirmed the additional recovery of hydrocarbons from sandstone and carbonate reservoirs due to the application of the hybrid technique.For instance, Alameri et al. [18] declared more than 10% of tertiary oil recovery in carbonates and Shaddel [40] observed an additional 5-7% from sandstone core samples by using sodium dodecylbenzene sulfonate surfactant (SDBS) with LSW.The experiment conducted by Khanamiri et al. [41] on sandstone cores resulted in 2-6% of the additional oil production.Alagic et al. [42] determined the high effectiveness of the application of LSW and surfactant flooding for oil-wet sandstone core samples.Johannessen and Spildo [43] conducted laboratory coreflooding experiments to investigate the synergy of LSWF and LSSF in sandstone samples.
The optimal amount of injected pore volume (PV) and surfactant concentration was discussed by Todd et al. [44] and James J. Sheng [45], where the effectiveness of flooding of large PV with a low concentration of surfactant and small slug with a high concentration of surfactant were compared.As a result, the second case showed more preferable results and higher oil recovery.Studies by Gogarty [46] and Murtada and Marx [47] have proven the validity of the previous conclusion that recovery characteristics for high concentration surfactant flooding are more efficient than low concentration.Moreover, a simulation study was performed by Tavassoli et al. [48] using UTCHEM-Iphreeqc.Flooding of LSW and surfactant was modeled and confirmed the results of the experimental study of Tahir et al. [49].The purpose of the simulation was to prove the effectiveness of a properly selected surfactant and flood design with the combination of LSW and other injection parameters of pre-flush brine on oil recovery.Similarly, some other numerical modeling studies have also demonstrated the efficiency of the hybrid LSSF method [50][51][52].The selection of optimum salinity in LSSF design is a crucial step and contributes significantly towards the success of the project [53].
The destabilization of the adsorbed oil layers and the reduction in oil-water IFT lead to wettability alteration by LSSF which in turn increases the capillary number.As a result, permeability increases while threshold capillary pressure decreases, leading to a higher reduction in S or [54,55].Some studies have reported the benefits of utilizing a negative salinity gradient during surfactant/polymer flooding (SP) and alkaline/surfactant/polymer (ASP) flooding, mainly in sandstone reservoirs [23].Instead of injecting a pre-flush of optimum salinity, a higher salinity pre-flush, followed by a lower or optimum salinity chemical flooding results in obtaining Winsor Types II-III-I microemulsions inside the reservoir [56,57].However, if low salinity pre-flush is conducted first followed by low salinity surfactant flooding or chemical flooding, it can result in generating a Winsor Type I microemulsion that usually does not result in ultra-low IFT between oil and water phases.The idea of a negative salinity gradient has not yet been studied in detail for LSSF applications in carbonate reservoirs.Thus, the objective of this study is to assess and quantify the effect of negative salinity gradient design to optimize LSSF flooding performance in carbonate reservoirs.
Nelson and Pope [23] presented the concept of negative salinity gradient for the first time in 1978, as they observed maximum oil recovery in coreflood tests, by injecting pre-flush brine, chemical slug, and subsequent chase fluid, having salinities in descending order.According to work done by Nelson [58], the negative salinity gradient of the pre-flush water has a vital role in the establishment of an optimal salinity environment for surfactant flooding.Ideally, this concept functions when the pre-flush fluid of above-optimum salinity, chemical solution of optimum salinity, and chase fluid of below-optimum salinity are injected during a chemical EOR test.A higher capillary number and thus a higher reduction in IFT can be achieved by opting for a negative salinity gradient because of weaker rockfluid interactions [59].There are several reasons to use a negative salinity gradient design.First of all, it aids in maintaining the phase behavior conditions suitable for Winsor Type III for a longer duration inside the reservoir [60,61].Whenever surfactant concentration becomes lower due to retention and adsorption, the optimum salinity for surfactants starts to be unfavorable or too high for the lower amount of surfactant.That is why the salinity of pre-flush fluid is designed to be above optimum salinity.Secondly, various experimental and modeling studies have shown a delay in surfactant breakthrough and reduced surfactant loss as a result of surfactant partitioning into the oleic phase when the pre-flush fluid has a salinity higher than the optimum salinity [58,62,63].Finally, the postflush or chase fluid of below-optimum salinity can stimulate the partitioning of the retarded surfactant into the aqueous phase, generating Type III through to Type I microemulsion with ultra-low IFT and further reducing surfactant loss [64,65].
Chen et al. [66] performed coreflood tests on Berea sandstone to investigate the effect of different injection salinity profiles on the performance of ASP flooding in the Daqing oilfield.The results showed the highest oil recovery of 95.3% OOIP by injecting pre-flush brine of 3% salinity, ASP slug of 2% salinity (optimum salinity), polymer slug of 0.5% salinity, and post-flush water of 0% salinity.This injection scheme followed a negative salinity gradient design and yielded the microemulsion phase behavior of Type II-III-I during the test.The residual oil saturation in the case of negative salinity gradient profile was only 3.7% compared to residual oil saturation of 9% for the case of constant salinity profile where the salinities of all injection fluids were 2 wt% (optimum salinity), generating Type III-III-III microemulsion behavior.Thus, the results of this study confirmed the effectiveness of a negative salinity gradient design to reduce the residual oil saturation to a minimum value.
The experimental study conducted by Han et al. [67] also signified the benefits of negative salinity gradient during surfactant/polymer (SP) flooding in Berea sandstone cores.The incremental oil recovery by using a negative salinity gradient design during SP flooding was around 10% higher than that obtained from conventional SP flooding at a constant optimum salinity.The prevalence of a descending salinity profile from over-optimum to optimum to under-optimum salinity in the reservoir helps in the restoration of HLB of surfactant between the oleic and aqueous phases, preventing surfactant from becoming too hydrophilic under a low salinity environment and too lipophilic under high salinity conditions [45,68].As a result, a reduction in surfactant loss is observed due to a delay in surfactant breakthrough, and the Winsor Type III microemulsion is effectively generated over a longer duration during the chemical injection phase.Both these factors improve the performance of surfactant flooding while ensuring lower chemical consumption.
The novelty of this research study is the applicability of the results to formulate an optimum low salinity-assisted surfactant flooding design that will be augmented by following a negative salinity gradient scheme.Such a design has been shown to provide Winsor Type II-III-I phase behavior that is favorable for the efficient recovery of residual oil by surfactant flooding.Salinity is a principal factor in surfactant flooding, whereby maximum incremental oil can be recovered by retaining the salinity of the formation brine and the injected surfactant solution at optimum salinity.The extensive literature review shows that the negative salinity gradient design warrants enhanced performance compared to a constant optimal salinity design; however, the designed salinity gradient can be adversely influenced by rock and fluid characteristics.In this context, it becomes challenging to regulate optimal salinity gradient profiles, and thus experimental studies Given the efficacy of negative salinity gradient during chemical flooding in sandstone reservoirs, there is still a knowledge gap on the effect of negative salinity gradient on the performance of low salinity/surfactant flooding in carbonates.Analyzing the available literature reveals a deficiency of an in-depth study on the design and effectiveness of negative salinity gradient for hybrid low salinity water/surfactant flooding technique, particularly in carbonate reservoirs.This approach needs further investigation and detailed analysis to be considered to achieve a more efficient hybrid EOR design for oil-wet carbonate formations.Hence, the focus of the current research is to study the performance of LSWF in combination with SF in carbonates in the presence of a negative salinity gradient.Three coreflooding tests were designed to study the effect of negative salinity gradient on the oil recovery by hybrid LSSF.Thus, this study investigates the possibility of further enhancing residual oil recovery beyond that obtained by traditional LSSF by employing the negative salinity gradient concept to LSSF in carbonate cores.

Methodology
The principal objective of this study was to design injection brine composition and sequence for LSSF at negative salinity gradient conditions.For this purpose, three coreflood tests were designed.The flow chart depicted in Figure 1 illustrates the research design followed to achieve study objectives.A discussion of the materials used in the study and the methods followed is provided hereafter.can be adversely influenced by rock and fluid characteristics.In this context, it becom challenging to regulate optimal salinity gradient profiles, and thus experimental studi are essential in designing salinity gradients suitable for hostile reservoir environmen encountered in carbonate formations.
Given the efficacy of negative salinity gradient during chemical flooding in san stone reservoirs, there is still a knowledge gap on the effect of negative salinity gradie on the performance of low salinity/surfactant flooding in carbonates.Analyzing the ava able literature reveals a deficiency of an in-depth study on the design and effectiveness negative salinity gradient for hybrid low salinity water/surfactant flooding techniqu particularly in carbonate reservoirs.This approach needs further investigation and d tailed analysis to be considered to achieve a more efficient hybrid EOR design for oil-w carbonate formations.Hence, the focus of the current research is to study the performan of LSWF in combination with SF in carbonates in the presence of a negative salinity gr dient.Three coreflooding tests were designed to study the effect of negative salinity gr dient on the oil recovery by hybrid LSSF.Thus, this study investigates the possibility further enhancing residual oil recovery beyond that obtained by traditional LSSF by em ploying the negative salinity gradient concept to LSSF in carbonate cores.

Methodology
The principal objective of this study was to design injection brine composition a sequence for LSSF at negative salinity gradient conditions.For this purpose, three c reflood tests were designed.The flow chart depicted in Figure 1 illustrates the resear design followed to achieve study objectives.A discussion of the materials used in t study and the methods followed is provided hereafter.

Rock Samples
Three Indiana limestone outcrop cores with 73 mm length and 38 mm diameter we used for coreflood tests.According to X-ray diffraction, the sample contained 99.97% c cite (CaCO3) and 0.03% quartz (SiO2).Porosities of the core samples were measured usi a helium porosimeter, and by saturation method, and were in close agreement.The phy ical properties of core samples are given in Table 1, whereby the porosities reported a those obtained from the saturation method.

Rock Samples
Three Indiana limestone outcrop cores with 73 mm length and 38 mm diameter were used for coreflood tests.According to X-ray diffraction, the sample contained 99.97% calcite (CaCO 3 ) and 0.03% quartz (SiO 2 ).Porosities of the core samples were measured using a helium porosimeter, and by saturation method, and were in close agreement.The physical properties of core samples are given in Table 1, whereby the porosities reported are those obtained from the saturation method.

Crude Oil, Brine, and Surfactant
Crude oil with an acid number of 4.3 mg KOH/g of oil and a viscosity of around 10 cp at a reservoir temperature of 80 • C is used in this study.The crude oil is medium-heavy in density (0.867-0.907 g/cm 3 ), has low sulfur content (0.1-0.14%), is slightly paraffinic (0.52-2.06%), and has a high amount of resins (15-21.5%)[69].This crude oil has displayed a high tendency to make the surface of carbonate rock oil-wet based on contact angle measurements, given a certain aging time between 1 week to 1 month.The results of the contact angle study can be found in another reference [70].
Formation water (FW) with a salinity of ~182,000 ppm was used to saturate the cores and mimic initial reservoir conditions.South Caspian Sea water (CSW) from West Kazakhstan with a salinity of 13,000 ppm was used as high salinity water (HSW) during the conventional waterflooding stage.Finally, the brine with an optimum salinity of 2876 ppm obtained by 10 times diluted Caspian Sea water with 3-and 6-times spiked calcium and sulfate ions, respectively (10 × SW-6SO 4 , Mg, 3Ca), was used as optimum engineered water (EW).According to a previous study, this LSW design affected the underlying crude oil/rock/brine system the most and resulted in the highest alteration in wettability to the water-wet state [71,72].The contact angles measured by Sekerbayeva et al. [71] showed almost 22% improvement in wettability towards the water-wet state using optimum EW.The ionic compositions of FW, HSW, and LSW are presented in Table 2. Soloterra-113H, an anionic surfactant provided by the Sasol company, is used in the study.The reason for choosing this surfactant is that, according to prior studies [71], Soloterra-113H in combination with the optimized engineered water (10 × SW-6SO 4 , Mg, 3Ca) provided the most effective microemulsion phase and low IFT between the oil and water phases.Moreover, the aqueous solution with Soloterra-113H showed the best stability at 20 • C and 80 • C. Furthermore, anionic surfactants are cost-effective compared to cationic surfactants.The optimum surfactant concentration, as obtained from aqueous stability and phase behavior results in a previous study by our team [71], was 1 wt%.The IFT between EW and surfactant was calculated using Nelson's correlation [58] given by Equation (1).For the first coreflood test, 1 wt% surfactant solution was prepared using EW as makeup brine.However, for the rest of the two experiments, a varying salinity gradient was designed to achieve the optimum in-situ brine salinity of ~2900 ppm.The details of the negative salinity gradient are discussed in a later section.
where σ mo,mw is the microemulsion/oil or microemulsion/water IFT in dynes/cm, V ow is the oil or water volume in microemulsion in ml, and V s is the surfactant volume in microemulsion phase in mL.An important consideration in the selection of a surfactant for carbonate reservoirs is the adsorption of the surfactant.Although anionic surfactant tends to show higher adsorption on negatively charged calcite surface, some studies have shown a similar adsorption behavior of anionic and cationic surfactants in carbonates.Kun Ma et al. [73] investigated the adsorption of the cationic surfactant (CPC) and the anionic surfactant (SDS) on carbonates.The study showed that CPC demonstrated low adsorption on a synthetic calcite surface but high adsorption on natural carbonate surfaces.Since anionic surfactants can provide better results in terms of IFT reduction compared to cationic surfactants, they are being used in carbonate reservoirs as well, despite having a higher adsorption tendency [53,74].In this study, the static adsorption of Soloterra-113H was determined by a UV spectrophotometer for the base case formulation with optimum engineered water.
Energies 2022, 15, x FOR PEER REVIEW 7 of 21 An important consideration in the selection of a surfactant for carbonate reservoirs is the adsorption of the surfactant.Although anionic surfactant tends to show higher adsorption on negatively charged calcite surface, some studies have shown a similar adsorption behavior of anionic and cationic surfactants in carbonates.Kun Ma et al. [73] investigated the adsorption of the cationic surfactant (CPC) and the anionic surfactant (SDS) on carbonates.The study showed that CPC demonstrated low adsorption on a synthetic calcite surface but high adsorption on natural carbonate surfaces.Since anionic surfactants can provide better results in terms of IFT reduction compared to cationic surfactants, they are being used in carbonate reservoirs as well, despite having a higher adsorption tendency [53,74].In this study, the static adsorption of Soloterra-113H was determined by a UV spectrophotometer for the base case formulation with optimum engineered water.

Aging of Core Samples
To mimic the initial conditions of core samples and to measure the absolute permeability by brine, the core samples were flooded with the formation water using the Aging Cell Apparatus by Vinci-Technologies, which is shown in Figure 2. To attain reservoir conditions, the heating mantle temperature was set at 80 °C, the unit confining pressure was regulated between 1000-1200 psi (6.9 × 10 6 -8.2 × 10 6 pa), and the back pressure was settled at 500 psi (3.4 × 10 6 pa).Pressure data were recorded for each pore volume at 3, 5, 7, 9, and 12 cc/min (5.0 × 10 −8 , 8.3 × 10 −8 , 1.2 × 10 −7 , 1.5 × 10 −7 , and 2.0 × 10 −7 m 3 /s) flow rates.Recorded results were used to estimate the absolute permeability of each core sample using Darcy's equation.Filtered crude oil was injected into the core sample at flow rates of 0.5, 2, and 5 cc/min (8.3 × 10 −9 , 3.3 × 10 −8 , and 8.3 × 10 −8 m 3 /s) to measure the effective permeability of oil.Each flow rate continued until the effluent water cut was less than 0.1%.The produced water volume was used to calculate initial water saturation (Swi).Table 3 shows the absolute permeability, effective permeability, and Swi for the core samples.After the injection test, the core samples were inserted into specialized aging cells filled with crude oil, which were placed in the oven at 80 °C for several days to achieve the initial wettability conditions in the reservoir.Recorded results were used to estimate the absolute permeability of each core sample using Darcy's equation.Filtered crude oil was injected into the core sample at flow rates of 0.5, 2, and 5 cc/min (8.3 × 10 −9 , 3.3 × 10 −8 , and 8.3 × 10 −8 m 3 /s) to measure the effective permeability of oil.Each flow rate continued until the effluent water cut was less than 0.1%.The produced water volume was used to calculate initial water saturation (S wi ).Table 3 shows the absolute permeability, effective permeability, and S wi for the core samples.After the injection test, the core samples were inserted into specialized aging cells filled with crude oil, which were placed in the oven at 80 • C for several days to achieve the initial wettability conditions in the reservoir.

Coreflooding
All the coreflood tests were performed at real reservoir conditions, whereby outcrop carbonate cores saturated with high-salinity and high-hardness formation brine were used.To restore the initial wettability of the cores and mimic real reservoir conditions, a viscous crude oil was injected to reach S wi and the cores were then aged in specialized aging cells for one month at a reservoir temperature of 80 • C. The coreflood tests were performed at a high reservoir temperature of 80 • C.
The final stage of the experimental part is the coreflooding oil displacement test.A similar procedure was followed for all three coreflood tests.The accumulators were filled with the injection fluids, and the system temperature was set at 80 • C. Aged core samples were loaded into the core holder and about 1000-1200 psi (6.9 × 10 6 -8.2 × 10 6 pa) of confining pressure was applied to act as the overburden pressure.In addition, the backpressure was set at approximately 500 psi (3.4 × 10 6 pa).After the required installation, 1 h was given to the system to reach 80 • C and to stabilize the system pressure.All three core flooding experiments were conducted at a reservoir temperature of 80 • C.
Before starting the injection of HSW, the core was flooded with oil at injection rates of 0.5, 2, and then 5 cc/min (8.3 × 10 −9 , 3.3 × 10 −8 , and 8.3 × 10 −8 m 3 /s) to make sure that there is no movable water and S wi is reached.All designed coreflooding experiments were started from HSW injection at the rate of 0.5 cc/min (8.3 × 10 −9 m 3 /s), 2 cc/min (3.3 × 10 −8 m 3 /s), and 5 cc/min (8.3 × 10 −8 m 3 /s) until zero oil production, i.e., residual oil to waterflood (S orw ).The next injection fluids followed the same procedure.In the end, the engineered water was injected to check that there was no more oil production.The produced effluent was collected into graduated tubes and the oil production and pressure drops as a function of injected pore volume (PV) were recorded.

Results and Discussion
This section presents an analysis of the results obtained from this study.The first step was to determine the optimum surfactant concentration that would result in Winsor Type III microemulsion in combination with the optimized engineered water.Aqueous stability and phase behavior studies were carried out to obtain the optimum combination for LSSF and the findings are presented in the following section.

Aqueous Stability and Phase Behavior Results
Based on the aqueous stability results, 1 wt% Soloterra-113H provided the most stable surfactant solution after 1 week aging at a reservoir temperature of 80 • C, as can be seen in Figure 3a.The phase behavior results were also similar, in that Winsor Type III microemulsion was observed at a reservoir temperature of 80 • C for 1 wt% surfactant solution prepared in EW (Figure 3b).The IFT calculated by Nelson's correlation [58] for the base case of optimum EW and surfactant system was around 0.02 dynes/cm which is almost three orders of magnitude lower compared to the typical oil-water IFT value of 30 dynes/cm [75].The static adsorption of surfactant was found to be 1.02 mg/g of rock which is within the acceptable surfactant adsorption limit for EOR applications [76].The details for static adsorption measurement are beyond the scope of this study and have been reported elsewhere [77].Thus, the optimum surfactant concentration of 1 wt% was used in all coreflood tests.Similar results have been reported in the literature, where 1 wt% surfactant concentration provided a higher reduction in IFT and almost 32% more incremental oil recovery compared to recovery obtained with 0.4 wt% surfactant concentration in an LSSF study [42].

Negative Salinity Gradient Design
To study the role of salinity profile on residual oil recovery and to optimum salinity profile to maximize the enhanced oil recovery by LSSF, t tests were designed.The first coreflood (CF-1) in the series of experiment case, having constant salinity (optimum salinity) of 2876 ppm in both LSW SF.Two core flooding experiments, namely CF-2 and CF-3, were designe effect of negative salinity gradient on the low salinity/surfactant flooding Primarily, this concept was suggested by Nelson and Pope [23].The highe was obtained during the coreflooding experiment, where salinities in the p chemical flood, and post-flush water drive were in descending value.Surf tration decreases during propagation through the core sample, because of th adsorption, thus optimal salinity is lowered [78].Hence, the optimum sali by phase behavior may change and be lowered in the porous media, leadi ification of the salinity and generation of the Winsor Type III microemulsio media.Through the application of the negative salinity gradient injection s linity alters the optimum condition in the porous media to maintain the p microemulsion phase as long as possible [66,67].Moreover, surfactant early can be excluded during surfactant partition into the oil phase under an salinity state [56,68].
Following this line of investigation, the effect of the negative salinity gr was studied and compared to the normal surfactant flooding at the optimum negative salinity sequences were designed, as shown in Figure 4.The diffe Figure 4a,b is the slope of the graph, where a sharper negative salinity gradi for CF-2 compared to the gradient designed for CF-3.The first surfactant so after HSW flooding in CF-2 (Injection 1) has a salinity of 4314 ppm which the optimum salinity, while the second surfactant solution (Injection 2) h 1438 ppm, yielding optimum salinity in-situ when it comes in contact with

Negative Salinity Gradient Design
To study the role of salinity profile on residual oil recovery and to determine the optimum salinity profile to maximize the enhanced oil recovery by LSSF, three coreflood tests were designed.The first coreflood (CF-1) in the series of experiments was the base case, having constant salinity (optimum salinity) of 2876 ppm in both LSW pre-flush and SF.Two core flooding experiments, namely CF-2 and CF-3, were designed to study the effect of negative salinity gradient on the low salinity/surfactant flooding performance.Primarily, this concept was suggested by Nelson and Pope [23].The highest oil recovery was obtained during the coreflooding experiment, where salinities in the pre-flush water, chemical flood, and post-flush water drive were in descending value.Surfactant concentration decreases during propagation through the core sample, because of the dilution and adsorption, thus optimal salinity is lowered [78].Hence, the optimum salinity measured by phase behavior may change and be lowered in the porous media, leading to the modification of the salinity and generation of the Winsor Type III microemulsion in the porous media.Through the application of the negative salinity gradient injection scheme, the salinity alters the optimum condition in the porous media to maintain the presence of the microemulsion phase as long as possible [66,67].Moreover, surfactant early breakthrough can be excluded during surfactant partition into the oil phase under an over-optimum salinity state [56,68].
Following this line of investigation, the effect of the negative salinity gradient scheme was studied and compared to the normal surfactant flooding at the optimum salinity.Two negative salinity sequences were designed, as shown in Figure 4.The difference between Figure 4a,b is the slope of the graph, where a sharper negative salinity gradient is designed for CF-2 compared to the gradient designed for CF-3.The first surfactant solution injected after HSW flooding in CF-2 (Injection 1) has a salinity of 4314 ppm which is higher than the optimum salinity, while the second surfactant solution (Injection 2) has a salinity of 1438 ppm, yielding optimum salinity in-situ when it comes in contact with Injection 1.On the other hand, three different salinities are chosen for CF-3 such that Injection 1 and Injection 2 have over-optimum salinities of 4601.6 ppm and 3738.8 ppm, respectively.The third surfactant solution (Injection 3) injected in CF-3 has the below-optimum salinity of 287.6 ppm.The objective of designing two different negative salinity gradients is to analyze if residual oil saturation can be further lowered by opting for a gradual salinity decrease instead of a sharper decrease in salinity.Table 4 provides injection brine compositions for two negative salinity gradient flooding designs.

Coreflood Tests
Table 5 shows the injection sequences and corresponding microemulsion order for the three corefloods.The first core flooding experiment is designed as a base case to compare its results with the subsequent experiments with a negative salinity gradient during the LSSF stage.It consists of three stages as shown in Table 5.The first stage is conventional waterflooding (HSW) in continuous mode.Then EW followed by the EW-surfactant solution at optimum salinity is injected.This injection scheme will yield microemulsion phase behavior of Winsor Type III-III [67].The last step is post-flush water (EW) to ensure that there is no more oil production.The second and third experiments are designed to check the negative salinity gradient effect and to compare it with the previous test.The difference between the second and third tests is the degree of the gradient.

Coreflood Tests
Table 5 shows the injection sequences and corresponding microemulsion order for the three corefloods.The first core flooding experiment is designed as a base case to compare its results with the subsequent experiments with a negative salinity gradient during the LSSF stage.It consists of three stages as shown in Table 5.The first stage is conventional waterflooding (HSW) in continuous mode.Then EW followed by the EW-surfactant solution at optimum salinity is injected.This injection scheme will yield microemulsion phase behavior of Winsor Type III-III [67].The last step is post-flush water (EW) to ensure that there is no more oil production.The second and third experiments are designed to check the negative salinity gradient effect and to compare it with the previous test.The difference between the second and third tests is the degree of the gradient.

Coreflood Tests
Table 5 shows the injection sequences and corresponding microemulsion order for the three corefloods.The first core flooding experiment is designed as a base case to compare its results with the subsequent experiments with a negative salinity gradient during the LSSF stage.It consists of three stages as shown in Table 5.The first stage is conventional waterflooding (HSW) in continuous mode.Then EW followed by the EW-surfactant solution at optimum salinity is injected.This injection scheme will yield microemulsion phase behavior of Winsor Type III-III [67].The last step is post-flush water (EW) to ensure that there is no more oil production.The second and third experiments are designed to check the negative salinity gradient effect and to compare it with the previous test.The difference between the second and third tests is the degree of the gradient.An over-optimum salinity surfactant solution is continuously injected in CF-2 after HSW injection until no more oil is produced at the effluent.Under-optimum salinity surfactant solution is then injected in continuous mode to recover maximum residual oil.The microemulsion phase behavior obtained under this negative salinity gradient design will be of Type II-III-I [67].After continuous injection of HSW in CF-3, surfactant solutions of two over-optimum salinities in descending order are injected subsequently, followed by injection of under-optimum salinity surfactant formulation.The second negative salinity gradient profile will result in microemulsion phases of Type II-II-III-I.The results of the three coreflood tests are analyzed hereafter to select the best hybrid LSSF design.

Coreflood 1 (CF-1)
The first oil displacement test was performed to check the effect of surfactant flooding in combination with EW on oil recovery while maintaining a constant optimum salinity during EWF and EWSF.Hybrid EWSF has the potential to significantly increase oil recovery over that achieved by standalone EWF or SF.The core sample used in this test has Swi of 0.2 and a pore volume of 13.3 mL.Once Swi was established and the core was aged for about a month, the oil displacement test was initiated by first injecting HSW to reach residual oil saturation by waterflooding (Sorw).Figure 5 provides differential pressure and oil recovery data versus PV injected.
HSW flooding recovered 65.40% of OOIP and the Sorw was around 35%. Engineered water was then injected continuously for about 18 PVs, recovering an additional 6.08% of OOIP.This additional oil recovery was mainly due to wettability alteration and mineral dissolution mechanisms of EW as some cloudy appearance was observed in the effluent during EWF [6,77,79].The total recovery before chemical flooding was 71.49%.The surfactant solution prepared in EW was then injected in continuous mode and it recovered 11.04% of OOIP.This additional oil was recovered due to the reduction of oil-water IFT by surfactant [37,71] and the microemulsion phase behavior generated under a constant optimum salinity profile was of the Type III-III.Such microemulsion profile is not the desired profile as it results in a lower reduction of Sor.Moreover, the dilution and adsorption of surfactant inside the porous media can cause the microemulsion type to transition to Type I, resulting in higher partitioning of surfactant into the aqueous phase and causing excessive surfactant loss due to early breakthrough [80].
HSW-Caspian Sea water IWS-1-Injection 1 with a surfactant of 1 wt% concentration II IWS-2-Injection 2 with a surfactant of 1 wt% concentration II IWS-3-Injection 3 with a surfactant of 1 wt% concentration III-I An over-optimum salinity surfactant solution is continuously injected in CF-2 after HSW injection until no more oil is produced at the effluent.Under-optimum salinity surfactant solution is then injected in continuous mode to recover maximum residual oil.The microemulsion phase behavior obtained under this negative salinity gradient design will be of Type II-III-I [67].After continuous injection of HSW in CF-3, surfactant solutions of two over-optimum salinities in descending order are injected subsequently, followed by injection of under-optimum salinity surfactant formulation.The second negative salinity gradient profile will result in microemulsion phases of Type II-II-III-I.The results of the three coreflood tests are analyzed hereafter to select the best hybrid LSSF design.

Coreflood 1 (CF-1)
The first oil displacement test was performed to check the effect of surfactant flooding in combination with EW on oil recovery while maintaining a constant optimum salinity during EWF and EWSF.Hybrid EWSF has the potential to significantly increase oil recovery over that achieved by standalone EWF or SF.The core sample used in this test has S wi of 0.2 and a pore volume of 13.3 mL.Once S wi was established and the core was aged for about a month, the oil displacement test was initiated by first injecting HSW to reach residual oil saturation by waterflooding (S orw ). Figure 5 provides differential pressure and oil recovery data versus PV injected.Another observation in CF-1 is that the pressure drop during EWSF at 0.5 cc/min (8.3 × 10 −9 m 3 /s) was only slightly higher than that observed during EWF, indicating an inefficient microemulsion generation [67].The overall incremental recovery by hybrid EWSF with a constant salinity profile was around 17%.To sum up, this design recovered 82.53% HSW flooding recovered 65.40% of OOIP and the S orw was around 35%. Engineered water was then injected continuously for about 18 PVs, recovering an additional 6.08% of OOIP.This additional oil recovery was mainly due to wettability alteration and mineral dissolution mechanisms of EW as some cloudy appearance was observed in the effluent during EWF [6,77,79].The total recovery before chemical flooding was 71.49%.The surfactant solution prepared in EW was then injected in continuous mode and it recovered 11.04% of OOIP.This additional oil was recovered due to the reduction of oil-water IFT by surfactant [37,71] and the microemulsion phase behavior generated under a constant optimum salinity profile was of the Type III-III.Such microemulsion profile is not the desired profile as it results in a lower reduction of S or .Moreover, the dilution and adsorption of surfactant inside the porous media can cause the microemulsion type to transition to Type I, resulting in higher partitioning of surfactant into the aqueous phase and causing excessive surfactant loss due to early breakthrough [80].
Another observation in CF-1 is that the pressure drop during EWSF at 0.5 cc/min (8.3 × 10 −9 m 3 /s) was only slightly higher than that observed during EWF, indicating an inefficient microemulsion generation [67].The overall incremental recovery by hybrid EWSF with a constant salinity profile was around 17%.To sum up, this design recovered 82.53% of OOIP.An EW-postflush was performed at the end of the test to ensure no more oil production.

Coreflood-2 (CF-2)
The difference between CF-1 and CF-2 is the salinity profile of the injection fluids after HSW flooding.The first negative salinity gradient test consists of two-step surfactant flooding with salinity in the descending order, so the salinity slope, in this case, is sharper.The experiment started with the continuous injection of HSW for around 22 PVs at 0.5, 2, and 5 cc/min, recovering 41.3% of OOIP.Each flow rate continued until the oil cut in the effluent was less than 0.01%.Figure 6 presents the oil recovery and pressure drop profiles for CF-2.The first surfactant solution with over-optimum salinity of 4314 ppm (IWS-1) was then injected in continuous mode and it produced an additional 14.6% of OOIP.This incremental recovery is attributed to the combined mechanisms of wettability alteration by LSW and IFT reduction by surfactant.Although IWS-1 had a salinity higher than optimum, this salinity lies in the range of low salinity applications, i.e., 2000 to 5000 ppm [5,81].Once there was no appreciable oil production by further injection of IWS-1, the core was flooded with a surfactant solution with a below-optimum salinity of 1438 ppm (IWS-2).The descending order of salinity during IWS-2 increased the recovery factor by 10.5% of OOIP.
The reason for further reduction in S or by IWS-2 is threefold.First of all, lower salinity water altered the rock wettability towards a more water-wet state and detached the acidic oil groups adsorbed on the positive calcite surface.Secondly, the mixing of IWS-2 with ISW-1 resulted in the optimum salinity in-situ and thus Winsor Type III microemulsion was generated that further increased the moveable oil saturation.Finally, ultra-low IFT achieved by Type III microemulsion also helped to lower the S or .The microemulsion Type II-III-1 was generated during chemical flooding in CF-2 that effectively recovered residual oil while minimizing surfactant loss [82].This negative salinity design has resulted in a total incremental oil recovery of 25.1%, which is higher than the conventional EWSF case.In terms of the remaining oil in the core (ROIC), this sharper negative salinity profile successfully recovered 43% ROIC.
The pressure drop observed during IWS-2 was considerably higher than the pressure drop during IWS-1, indicating effective microemulsion phase behavior obtained during this negative salinity gradient test.The test was terminated with an EW-post flush to ensure all the moveable oil volume was already produced.The reason for further reduction in Sor by IWS-2 is threefold.First of all, lower salinity water altered the rock wettability towards a more water-wet state and detached the acidic oil groups adsorbed on the positive calcite surface.Secondly, the mixing of IWS-2 with ISW-1 resulted in the optimum salinity in-situ and thus Winsor Type III microemulsion was generated that further increased the moveable oil saturation.Finally, ultra-low IFT achieved by Type III microemulsion also helped to lower the Sor.The microemulsion Type II-III-1 was generated during chemical flooding in CF-2 that effectively recovered residual oil while minimizing surfactant loss [82].This negative salinity design has resulted in a total incremental oil recovery of 25.1%, which is higher than the conventional EWSF case.In terms of the remaining oil in the core (ROIC), this sharper negative salinity profile successfully recovered 43% ROIC.
The pressure drop observed during IWS-2 was considerably higher than the pressure drop during IWS-1, indicating effective microemulsion phase behavior obtained during this negative salinity gradient test.The test was terminated with an EW-post flush to ensure all the moveable oil volume was already produced.

Coreflood-3 (CF-3)
The difference between CF-2 and CF-3 is the degree of the gradient.The second negative salinity gradient design has a gradual slope and consists of three surfactant flooding steps, as Figure 7 presents.The first step was, as in the previous two experiments, to reach Sorw by continuous injection of HSW.The recovery factor was 38.79% of OOIP after injecting almost 23 PVs of HSW.The reason for a lower recovery factor by HSW in CF-2 and CF-3 is due to the more oil-wet nature of these core samples as the aging time in the oven was slightly higher for Core-2 and Core-3 compared to Core-1.Therefore, only the incremental recoveries after HSW flooding are considered for the comparison of salinity profiles in three corefloods.

Coreflood-3 (CF-3)
The difference between CF-2 and CF-3 is the degree of the gradient.The second negative salinity gradient design has a gradual slope and consists of three surfactant flooding steps, as Figure 7 presents.The first step was, as in the previous two experiments, to reach S orw by continuous injection of HSW.The recovery factor was 38.79% of OOIP after injecting almost 23 PVs of HSW.The reason for a lower recovery factor by HSW in CF-2 and CF-3 is due to the more oil-wet nature of these core samples as the aging time in the oven was slightly higher for Core-2 and Core-3 compared to Core-1.Therefore, only the incremental recoveries after HSW flooding are considered for the comparison of salinity profiles in three corefloods.
After HSW, the chemical flooding was initiated by injecting the first surfactant solution with over-optimum salinity of 4601.6 ppm (IWS-1) in continuous mode.17.2% of OOIP incremental oil was produced during IWS-1, attributed mainly to wettability alteration by low salinity water and IFT reduction by surfactant.The microemulsion phase of Type II was generated during IWS-1, as the salinity was higher than the optimum salinity.This resulted in a higher surfactant partitioning into the oleic phase, thereby reducing surfactant loss, and delaying surfactant breakthrough [66,68].
The core was then flooded with a second surfactant solution with an over-optimum salinity of 3738.8 ppm (IWS-2).An additional 8.2% of OOIP oil was recovered by IWS-2 by the combined action of LSW and SF.Finally, the third surfactant solution with underoptimum salinity of 287.6 ppm was injected to switch the microemulsion phase behavior to Type III and then Type II.The recovery factor increased by 7.7% of OOIP during IWS-3.The gradual decrease of salinity in CF-3 recovered the highest percentage of the remaining oil in the core after waterflooding which was around 54% of ROIC.The pressure fluctuations observed during the IWS-3 injection indicated the generation of Winsor Type-III microemulsion as IWS-2 and IWS-3 came into contact.The further injection of IWS-3 shifted the microemulsion to Type I, causing the movement of the trapped surfactant from the oleic phase to the aqueous phase.An interesting observation in CF-3 is an almost stable pressure drop of around 17-20 psi (1.4 × 10 5 pa) during all three injection stages of LSSF.This is a clear indication that the incremental oil recovery during ISW-2 and ISW-3 was obtained due to a higher reduction in capillary forces, mainly IFT, as the phase behavior transitioned to favorable Type III microemulsion.Thus, the surfactant actively reduced IFT in this design and generated the desired microemulsion behavior of Type II-II-III-1 during chemical flooding.Similar results have been reported by Han et al. in a surfactant/polymer flood study on Berea sandstone [67].After HSW, the chemical flooding was initiated by injecting the first surfactant solution with over-optimum salinity of 4601.6 ppm (IWS-1) in continuous mode.17.2% of OOIP incremental oil was produced during IWS-1, attributed mainly to wettability alteration by low salinity water and IFT reduction by surfactant.The microemulsion phase of Type II was generated during IWS-1, as the salinity was higher than the optimum salinity.This resulted in a higher surfactant partitioning into the oleic phase, thereby reducing surfactant loss, and delaying surfactant breakthrough [66,68].
The core was then flooded with a second surfactant solution with an over-optimum salinity of 3738.8 ppm (IWS-2).An additional 8.2% of OOIP oil was recovered by IWS-2 by the combined action of LSW and SF.Finally, the third surfactant solution with underoptimum salinity of 287.6 ppm was injected to switch the microemulsion phase behavior to Type III and then Type II.The recovery factor increased by 7.7% of OOIP during IWS-3.The gradual decrease of salinity in CF-3 recovered the highest percentage of the remaining oil in the core after waterflooding which was around 54% of ROIC.The pressure fluctuations observed during the IWS-3 injection indicated the generation of Winsor Type-III microemulsion as IWS-2 and IWS-3 came into contact.The further injection of IWS-3 shifted the microemulsion to Type I, causing the movement of the trapped surfactant from the oleic phase to the aqueous phase.An interesting observation in CF-3 is an almost stable pressure drop of around 17-20 psi (1.4 × 10 5 pa) during all three injection stages of LSSF.This is a clear indication that the incremental oil recovery during ISW-2 and ISW-3 was obtained due to a higher reduction in capillary forces, mainly IFT, as the phase behavior transitioned to favorable Type III microemulsion.Thus, the surfactant actively reduced To further confirm the better performance of negative salinity gradient design, the rock samples used in the three corefloods were analyzed for visual appearance.Figure 8 shows a comparison of the used core samples, and it can be observed that the samples flooded using negative salinity gradient profiles (Core-2 and Core-3) appeared clearer as if most of the crude oil had been washed away from these cores compared to sample used for CF-1 (Core-1).
Figure 9 shows the comparison of incremental oil recoveries for the base case and the two negative salinity gradient cases.It can be seen that negative salinity gradient cases recovered higher incremental oil compared to the base case.The incremental oil recovered by the first negative salinity gradient test is 8% of OOIC higher than the base case while the second negative salinity gradient test recovered 16% of OOIC more incremental oil compared to the base case.Figure 9 shows that the application of a negative salinity gradient can enhance the performance of the LSSF method and increase oil recovery beyond that achievable by normal LSSF designed at constant optimum salinity.Hence, the approach is recommended to be used in the design of the hybrid LSWF/SF EOR method in carbonates.
if most of the crude oil had been washed away from these cores compared to sample used for CF-1 (Core-1).Figure 9 shows the comparison of incremental oil recoveries for the base case and the two negative salinity gradient cases.It can be seen that negative salinity gradient cases recovered higher incremental oil compared to the base case.The incremental oil recovered by the first negative salinity gradient test is 8% of OOIC higher than the base case while the second negative salinity gradient test recovered 16% of OOIC more incremental oil compared to the base case.Figure 9 shows that the application of a negative salinity gradient can enhance the performance of the LSSF method and increase oil recovery beyond that achievable by normal LSSF designed at constant optimum salinity.Hence, the approach is recommended to be used in the design of the hybrid LSWF/SF EOR method in carbonates.Figure 9 shows the comparison of incremental oil recoveries for the base case a two negative salinity gradient cases.It can be seen that negative salinity gradien recovered higher incremental oil compared to the base case.The incremental oil reco by the first negative salinity gradient test is 8% of OOIC higher than the base case the second negative salinity gradient test recovered 16% of OOIC more incremen compared to the base case.Figure 9 shows that the application of a negative salini dient can enhance the performance of the LSSF method and increase oil recovery b that achievable by normal LSSF designed at constant optimum salinity.Hence, t proach is recommended to be used in the design of the hybrid LSWF/SF EOR met carbonates.A summary of the coreflood results is presented in Table 6.Both negative salinity gradient designs have exhibited higher incremental oil recoveries compared to constant optimum salinity design in terms of OOIC.However, the gradual negative salinity gradient profile shows the best performance in terms of the remaining oil in the core (ROIC), compared to the EWSF case and the sharper negative salinity gradient case.The reason for higher recovery of residual oil in case of a gradual negative salinity gradient is the reduced surfactant loss during IWS-1 and IWS-2 flooding stages, development of Winsor Type III microemulsion during IWS-3 injection stage for a longer duration providing ultra-low IFT between oil and water, and redistribution of surfactant between aqueous an oleic phases as phase behavior transitions from Type-III to Type-I during IWS-3 [80,82,83].The results of this study can be validated and supported by similar studies in the literature.Han et al. performed coreflood tests to select the optimum surfactantpolymer formulation for the Daqing ASP project and found that a negative salinity profile resulted in the highest incremental oil recovery [67].Gupta and Trushenski have also reported similar findings where a negative salinity gradient yielded higher oil recovery in a surfactant-polymer flood design [62].The coreflood results from the work of Riswati et al. also validate the effectiveness of a negative salinity profile as a negative salinity gradient design with phase behavior of Winsor Type II-III-I provided 13% higher incremental oil recovery and 4% more reduction in S or compared to the cases where microemulsion phase behavior was I-III-I or I-II-I [83].
This study has successfully shown improvement in oil recovery by using an anionic surfactant in combination with low salinity water while following a negative salinity gradient design.Although the surfactant concentration of 1 wt% used in this study is slightly on the higher side, the literature shows similar studies where 1 wt% or higher surfactant concentration provided better results in terms of IFT reduction and residual oil recovery [42,84].Increased adsorption of anionic surfactants on positively charged calcite surfaces, the presence of formation water with high salinity and hardness, and high reservoir temperatures are some of the precursors for using a higher surfactant concentration during surfactant flooding in carbonate reservoirs.Operational problems such as surfactant loss and early breakthrough of surfactant can be overcome by using a gradual negative salinity gradient formulation in the hybrid LSSF method.Brine salinity influences the phase behavior of anionic surfactant systems because it provides a source of cationic electrolyte [85].The salinity of the water has a significant impact on the phase behavior of the surfactant solution.When the aqueous phase salinity becomes higher, the solubility of anionic surfactants in the aqueous phase decreases, thus surfactants are driven out of the brine and contribute to the middle or upper phase, causing the transition of microemulsion from Type I to Type II through Type III, as schematically shown in Figure 10.
surfactants in the aqueous phase decreases, thus surfactants are driven out of th and contribute to the middle or upper phase, causing the transition of microem from Type I to Type II through Type III, as schematically shown in Figure 10.The transition behavior of microemulsion from Type II to Type III to Type ferred inside the reservoir to overcome early surfactant loss and for efficient an term generation of favorable Type III microemulsion.This behavior is achieved by ing a negative salinity gradient design and yields higher recoveries of residua demonstrated in CF-2, and CF-3.The application scope of this study encompasses all chemical flooding projects involving surfactants as the performance of surfac be greatly enhanced by designing a negative salinity gradient, as supported by the of the coreflood tests.This study considered the harsh reservoir conditions typical carbonate reservoirs and thus the results of this work can be successfully applied mations with harsh conditions of salinity and temperature.

Conclusions
Hybrid low salinity surfactant flooding involving the synergy of low salinity flooding and surfactant flooding has proved its effectiveness in carbonate cores.perimental study shows that the performance of the method can be improved by lection of an appropriate negative salinity gradient during LSSF design.The nega linity gradient concept was tested in limestone core samples and demonstrated pr results.Coreflood experiments involving a negative salinity gradient revealed hi cremental oil recovery compared to EWSF, demonstrating the benefit of the applic a negative salinity gradient in the hybrid LSSF scheme.A three-step slightly slop ative salinity gradient design showed a higher reduction in residual oil saturatio pared to the conventional EWSF injection.This study suggests that the injection and alteration in the salinity of the injected brines lead to better oil recovery du The transition behavior of microemulsion from Type II to Type III to Type I is preferred inside the reservoir to overcome early surfactant loss and for efficient and long-term generation of favorable Type III microemulsion.This behavior is achieved by following a negative salinity gradient design and yields higher recoveries of residual oil as demonstrated in CF-2, and CF-3.The application scope of this study encompasses almost all chemical flooding projects involving surfactants as the performance of surfactant can be greatly enhanced by designing a negative salinity gradient, as supported by the results of the coreflood tests.This study considered the harsh reservoir conditions typical of most carbonate reservoirs and thus the results of this work can be successfully applied to formations with harsh conditions of salinity and temperature.

Conclusions
Hybrid low salinity surfactant flooding involving the synergy of low salinity waterflooding and surfactant flooding has proved its effectiveness in carbonate cores.This experimental study shows that the performance of the method can be improved by the selection of an appropriate negative salinity gradient during LSSF design.The negative salinity gradient concept was tested in limestone core samples and demonstrated promising results.Coreflood experiments involving a negative salinity gradient revealed higher incremental oil recovery compared to EWSF, demonstrating the benefit of the application of a negative salinity gradient in the hybrid LSSF scheme.A three-step slightly sloped negative salinity gradient design showed a higher reduction in residual oil saturation compared to the conventional EWSF injection.This study suggests that the injection design and alteration in the salinity of the injected brines lead to better oil recovery due to the development of favorable phase behavior inside the porous media.
Among three injection salinity profiles evaluated in this study, including constant salinity, a sharper negative salinity gradient, and a gradual negative salinity gradient, the results revealed a gradual negative salinity gradient (microemulsion phase behavior of Type II-II-III-I) provided maximum incremental oil recovery of 33% of OOIP (54% of ROIC).The reason for higher oil recoveries by negative salinity gradient designs is that the formulations generated during these tests provided a favorable in-situ hydrophilic-lipophilic balance during surfactant flow, yielding uniform partitioning of surfactant in oil and water phases.The phase behavior profiles of Type II-III-1 and II-II-III-I inside the reservoir result in reduced surfactant loss, long-term generation of Winsor Type III microemulsion as surfactant redistributes between oil and water phases, and attainment of ultra-low IFT between oil and water.As a result, a higher reduction in residual oil saturation is obtained by low salinity surfactant flooding due to a carefully designed negative salinity gradient.
designing salinity gradients suitable for hostile reservoir environments encountered in carbonate formations.

Figure 1 .
Figure 1.Flow chart showing main steps involved in the research design for the selection of b LSSF scenario.

Figure 1 .
Figure 1.Flow chart showing main steps involved in the research design for the selection of best LSSF scenario.

Figure 2 .
Figure 2. Schematic of coreflood apparatus used for oil displacement tests.

Figure 2 .
Figure 2. Schematic of coreflood apparatus used for oil displacement tests.

Figure 3 .
Figure 3. Results of (a) aqueous stability test and (b) phase behavior study at 80 °C terra-113H prepared in EW.

Figure 3 .
Figure 3. Results of (a) aqueous stability test and (b) phase behavior study at 80 • C for 1 wt% Soloterra-113H prepared in EW.

Table 5 .
Design of oil displacement tests for hybrid LSSF with different salinity profiles.water EWSF-Engineered water with a surfactant of 1 wt% concentration III-III HSW-Caspian Sea water EW-Engineered water EWSF-Engineered water with a surfactant of 1 wt% concentration III-III 2. CF-2 HSW > IWS-1 > IWS-2 > LSW Energies 2022, 15, x FOR PEER REVIEW 11 of 21 2. CF-2 HSW > IWS-1 > IWS-2 > LSW HSW-Caspian Sea water IWS-1-Injection 1 with a surfactant of 1 wt% concentration II IWS-2-Injection 2 with a surfactant of 1 wt% concentration III-I 3. CF-3 HSW > IWS-1 > IWS-2 > IWS-3 > LSW HSW-Caspian Sea water IWS-1-Injection 1 with a surfactant of 1 wt% concentration II IWS-2-Injection 2 with a surfactant of 1 wt% concentration II IWS-3-Injection 3 with a surfactant of 1 wt% concentration III-IAn over-optimum salinity surfactant solution is continuously injected in CF-2 after HSW injection until no more oil is produced at the effluent.Under-optimum salinity surfactant solution is then injected in continuous mode to recover maximum residual oil.The microemulsion phase behavior obtained under this negative salinity gradient design will be of Type II-III-I[67].After continuous injection of HSW in CF-3, surfactant solutions of two over-optimum salinities in descending order are injected subsequently, followed by injection of under-optimum salinity surfactant formulation.The second negative salinity gradient profile will result in microemulsion phases of Type II-II-III-I.The results of the three coreflood tests are analyzed hereafter to select the best hybrid LSSF design.10.1.Coreflood 1 (CF-1)The first oil displacement test was performed to check the effect of surfactant flooding in combination with EW on oil recovery while maintaining a constant optimum salinity during EWF and EWSF.Hybrid EWSF has the potential to significantly increase oil recovery over that achieved by standalone EWF or SF.The core sample used in this test HSW-Caspian Sea water IWS-1-Injection 1 with a surfactant of 1 wt% concentration II IWS-2-Injection 2 with a surfactant of 1 wt% concentration III-

Energies 2022 , 21 Figure 5 .
Figure 5. Pressure drop and oil recovery profile for CF-1 with constant salinity profile.

Figure 5 .
Figure 5. Pressure drop and oil recovery profile for CF-1 with constant salinity profile.

Figure 6 .
Figure 6.Pressure drop and oil recovery profile for CF-2 with a sharper negative salinity gradient.

Figure 6 .
Figure 6.Pressure drop and oil recovery profile for CF-2 with a sharper negative salinity gradient.

Energies 2022 , 21 Figure 7 .
Figure 7. Pressure drop and oil recovery profile for CF-3 with a gradual negative salinity gradient.

Figure 7 .
Figure 7. Pressure drop and oil recovery profile for CF-3 with a gradual negative salinity gradient.

Figure 10 .
Figure 10.Schematic illustration of middle phase formation and microemulsion phase tran a function of salinity.

Figure 10 .
Figure 10.Schematic illustration of middle phase formation and microemulsion phase transition as a function of salinity.

Table 1 .
Physical properties of the core samples used in the study.

Table 1 .
Physical properties of the core samples used in the study.

Table 2 .
Ionic compositions of formation water, high salinity water, and engineered water.

Table 3 .
Rock-fluid interaction properties for samples used in oil displacement tests.

Table 3 .
Rock-fluid interaction properties for samples used in oil displacement tests.

Table 5 .
Design of oil displacement tests for hybrid LSSF with different salinity profiles.
EW-Engineered water EWSF-Engineered water with a surfactant of 1 wt% concentration III-III

Table 5 .
Design of oil displacement tests for hybrid LSSF with different salinity profiles.