Effect of different injection fluids scenarios on swelling and migration of common clays in case of permeability variations: a micromodel study

Swelling and migration of present clays make damage to the oil reservoirs due to low salinity waterflooding (LSWF) can induce serious problems in the case of oil recovery improvement and researchers are trying to solve this problem. The purpose of this work is to investigate the mechanism of two phenomena of swelling and migration clays in the porous media of a reservoir rock by injecting a different composition of LSWF using a glass micromodel and providing the appropriate composition and pattern of injection with the removal of damage. Proper water flooding design, application of efficient swelling inhibitors, and migration control are among the most important methods to overcome the problem of formation damage due to swelling and migration of clays. A series of static (bulk or bottle test) and dynamic tests were carried out using a micromodel with a coating of kaolinite and montmorillonite clays in the vicinity and injection of different low salt water compositions. The type and amount of these clays were selected based on the results of XRD and SEM mineralogical tests on real reservoir rock, FW and diluted FW, SW and diluted SW, solution of 1% zirconium oxychloride in 20 times diluted seawater (SI), and composition of nanofluid MgO, SiO2, and Al2O3 in 20 times diluted. In the studies conducted by the micromodel, only the images taken were used in the analysis of the mechanisms, but here, the input and output pressures of the micromodel were recorded with high-precision pressure transmitters, and by using the differential pressure, the permeability was calculated and the formation damage index was introduced. The overlap of the interpretation of the captured images and the changes of the numerical parameter of the damage index in all stages of injection of smart water composition was considered to evaluate the simultaneous and separate mechanisms of swelling and migration of clays. The results of the experiments in this research show that clay swelling has destructive effects on permeability, and migration due to the transfer of clays from the porous medium can have promising effects on reducing the damage index in some conditions. And it is necessary to use the swelling control compound during the flooding process, but the migration inhibitor compound is not always suitable. Gradual reduction of salinity is also introduced as a pattern to prevent swelling damage or clay migration. In general, in this study, the best design and fluid engineering for smart water injection with the least damage in the micromodel scale was presented.


Introduction
Formation damage is a general term that refers to the disruption of the physical, chemical, or mechanical properties of the formation and especially the reduction in the permeability of hydrocarbon formations by various hydrocarbon recovery processes. Also, it is an unfavorable operational and economic problem that can occur during various stages of oil and gas recovery including drilling, production, hydraulic fracturing, and workover (Civan 2007). The changes in chemical, physical, and thermodynamic conditions associated with EOR techniques can lead to a variety of formation damage. In hydrocarbon research conducted for the formation damage mechanisms, various reasons have been reported: Particle and clay migration, clay swelling, fluid-fluid incompatibility, rock-fluid interaction, flow path blockage, formation collapse, thermal stress, dissolution reactions, and precipitation of organic and inorganic materials, net stress pore deformation effect of overburden layers, wettability alteration, changes in pore surface properties, pore structure, and mechanical properties (Al-Mohammad et al. 2012;Mikhailov et al. 2010;Bedrikovetsky et al. 2011;Mojarad and Settari 2005;and Scott et al. 2007). The definition states that formation damage may occur during field IOR operations such as water flooding. Low salinity water injection (LSWF) is one of the EOR methods performed as a tertiary recovery method (Gamage and Thyne 2011). In addition, many studies have shown that reducing the salinity of injected salt water below the critical salt concentration (CSC) causes the swelling or migration of clays and consequent reduction in permeability (Jafari Barnaji et al. 2016). Song and Kovscek studied the migration of clays in the presence and absence of oil and observed that at salinity below 4000 ppm, clay migration occurs under both conditions. They also stated that several mechanisms can be proposed to illustrate the increase in oil recovery due to low-salinity water injection, including fine migration, cation exchange, and emulsion formation (Song and Kovscek 2016). Despite the success of the main experiments and experimental studies of field-scale low-salinity water injection, the various mechanisms of LSWF impact on reservoirs and oil recovery remain poorly understood and controversial, and the effects of fine migration on oil recovery are not understood (Atthawutthisim 2012;Al-Shalabi and Sepehrnoori 2016). Petroleum researchers and engineers are trying to identify the causes of formation damage and control, reduce or eliminate it. In recent decades, due to the important role of clays in the water injection method to increase oil recovery, various studies have been done on the role of clay minerals in LSWF. The clays in the reservoir rock are classified into two groups: swelling such as montmorillonite and non-swelling such as kaolinite. Non-swelling clays such as kaolinite may migrate due to repulsive forces between clay and the pore walls, stopping at narrower pore throats and plugging the flow paths of the porous media. Swelling clays also in contact with brines can swell to numerous times bigger than their original volume, making the flow paths narrower. Consequently, it results in a reduction of permeability or formation damage. Therefore, the control of swelling and migration of clays is of interest to researchers. To control particle migration, Habibi et al. (2011) investigated the effect of various nanoparticles such as MgO, SiO 2 and Al 2 O 3 on the reduction of migration using artificial core samples and injection of distilled water with fine particles. The results showed that when the porous media was already treated with Nanofluid, the fine particles could adhere to the grains and prevent them from migrating. In addition, to investigate the mechanisms of this modification method, the effect of nanoparticle concentration and fluid injection rate in the medium on grain detachment was studied. A theoretical model for calculating the interaction energy of total surfaces was used to evaluate the experimental results, and was validated using SEM images for artificial core samples. Finally, MgO nanoparticles were found to be more effective 1 3 even at high fluid injection rates, when used at higher concentrations. They concluded that MgO nanoparticles were an effective factor in controlling particle migration (Habibi et al. 2011). Sameni et al. (2015) utilized glass micromodels to study the effect of nanoparticles and reported that nanoparticles are not useful in controlling swelling because they cannot penetrate the interlayer spacing of the clay structure but can control migration. Swelling clays such as montmorillonite may swell up to 20 times their original volume in the presence of injected water and may even migrate, both of which can reduce the porosity and permeability of the hydrocarbon formation (Sameni et al. 2015). In 2015, Aksu et al. studied the effect of the different amounts of swellable clay (montmorillonite) and non-swellable clay (kaolinite) on permeability alteration. They utilized X-ray imaging and tumor computed tomography during flooding tests in the sandstone core samples to detect the permeability changes over time. In these experiments, the effect of the amount and type of the clay mineral mixtures was investigated to provide information on morphological changes of clay grains before and after water saturation. To understand the extent of swelling-induced permeability reduction in natural samples with different porosities, the volume, swelling quantity, porosity model, and permeability relationships were changed (Aksu et al. 2015). In recent years, micromodels have been used for a better understanding of the pore scale behavior in various hydrocarbon recovery techniques and mechanisms. The ability to observe events that occur in a porous medium is a major advantage of using micromodels. The system of micro-models can be used to visualize the swelling and migration of clays in the porous medium due to incompatible water injection. In addition, the effect of clays on permeability reduction and hydrocarbon recovery can be detected. Microfluidic model porous media, more commonly known as micromodels, are devices with a transparently connected porous network that allows direct visualization of the dynamics of complex fluid flow inside the reservoir rock Anbari et al. (2018). Sharifpour et al. (2017) stated that the formation damage in the vicinity of the well is due to the reduction of permeability resulting from foreign fluid penetration into the reservoir rock. This damage is due to many factors, including the mineralogical composition of the rock and the effect of injected water on the swelling of the clay, which prevents further oil from being swept. They visually examined the swelling of the clay using homogeneous glass micromodels with different patterns and presented a model for calculating the permeability using the clay content and pore dimensions (Sharifipour et al. 2017(Sharifipour et al. , 2019. The crucial role of water injection in oil recovery enhancement has been proven in most studies, but the composition of reservoir rocks containing clay minerals is influenced by the composition of injected water. The swelling and migration of clay minerals result in the blocking of the hydrocarbon flow paths and permeability reduction. On the other hand, further studies show that low salinity water injection (LSW) increases oil recovery compared to high salinity water injection. Due to its simplicity, environmental compatibility, and low cost, it attracted attention. In other words, the injection of low salinity water (LSW) into reservoirs improves the microscopic displacement efficiency compared to high salinity water injection (HSW). To reduce the risk of damage, selecting the appropriate composition and further studies on the phenomenon are needed. Swelling and migration of clay will play a key role in the success of LSW projects. In this EOR method, the recovery factor can be improved by up to 30% of the oil in place when the appropriate water composition is selected (Snosy et al. 2021;Shahmohammadi et al. 2022;and Zaidel et al. 2022). Mehdizad et al. used clay stabilizers instead of inhibitors to test their ability to simultaneously control swelling and migration. Although the application of inexpensive mineral inhibitors is still common in field operations, it is a good alternative to cationic mineral inhibitors and polymers, which are used before the EOR method or after acid stimulation. They also stated that the most important factor for salt selection is the charge density of its cation, for example, K + is a monovalent cation it can perform equal or even better than divalent cations such as Ca 2+ in case of swelling and migration control due to its smaller ionic radius (Mehdizad et al. 2022a, b). Disagreements in field-scale research and experiments suggest that a basic understanding of the mechanisms that dictate the effect of low salinity is needed. Therefore, for reservoir rocks containing clay, which is a candidate for the water injection scenario, it is necessary to identify the mechanism and ways to control or direct the swelling and migration of clay particles to the further success of this IOR method. As stated in most studies, injection of low saline water is very effective in the secondary recovery of hydrocarbons. However, the effect of water composition on the clays swelling and migration still needs further studies. Therefore, success in waterflooding significantly depends on the performance of the injected fluid in the vicinity of the clays. So, the composition of water in the IOR process must be engineered and designed to minimize the formation damage caused by the injection. In most studies, two samples of kaolinite and montmorillonite clays have been identified as non-swelling and swelling clays causing damage. In this research, the presence of these types of clays was proven by the XRD results of a real carbonate reservoir rock therefore they have been selected as the migration and swelling agents. In recent years, micromodels have been used to better understand the pore-scale behavior of various oil recovery mechanisms, and visual and direct study of phenomena occurring in porous media is the main advantage of using micromodels in water 1 3 flooding studies. Glass micromodels have been used to study different EOR methods, including LSWF Mehdizad et al. (2022a, b).
The study of different aspects of EOR methods can be further improved by the use of microfluidic devices, because of their transparency. Visual observation of the mechanisms that take place in the porous medium is the main advantage of these devices over the core flooding experiments which are like black boxes. Usually, results of micromodel experiments are obtained using an image processing technique that isn't completely reliable. Therefore, making micromodels with realistic patterns which are coated with actual constituents of the real reservoir rocks, can be monitored both visually and quantitatively. Recording the pressure variations during the flooding can assist in further performance enhancement of such experiments. The present study aimed to investigate the formation damage caused by swelling and migration due to the presence of clay particles within the reservoir rock. The extent of formation damage was assessed through permeability variations and visual study using transparent porous media. The degree of formation damage is dependent on the swelling and migration of the clay particles and is not affected by the reservoir rock type. In other words, the type and percentage of the clay minerals are the main factors that influence the formation damage. The type and percentage of the clays that were used in the micromodel were based on the real carbonate reservoir data. Based on a CT-Scan image and XRD analysis of real rock, a heterogenous pattern was designed and coated by clays their presence is proven by XRD analysis. Kaolinite and Smectite were used to coat the micromodel which are representative of migration and swelling, respectively. A high-precision pressure transmitter was used for quantitative analysis of permeability variations and formation damage. In addition, the effect of the priority of different fluid injections under various injection plans was evaluated.

Materials and methodology
In this study, average porosity and permeability, formation water composition, type, and amount of clay based on field data from a reservoir in the southern region of Iran were employed, and experimental design and material selection was done accordingly. The properties of the aforementioned reservoir are listed in Table 1. Two categories of experiments were designed: (1) A Static experiment (swelling index test) designed to identify the amount of clay swelling (Montmorillonite, kaolinite) in the presence of different solutions with different compositions. The swelling index test was conducted to select suitable injecting fluids for formation damage analysis during micromodel tests. (2) Dynamic experiments were designed in 7 series of micromodel tests.

Materials
Kaolinite and montmorillonite were selected as swelling and migration agents based on the results of XRD and SEM analysis of core samples from the mentioned rock reservoir. Also, the solution compositions used in this study are Formation water of the nominated reservoir, (with dilution factors of 2, 6, 10, 20, and 40 times), Seawater (with dilution factors of 2, 6, 10, and 20 times), solution of 1% zirconium oxychloride in 20 times diluted seawater (SI), and nanofluid with dispersed nanoparticles such as MgO, SiO 2 , and Al 2 O 3 in 20 times diluted seawater. Compositions of utilized fluids in static tests can be seen in Table 2.

Swelling index experiments
In general, the swelling index test is performed according to the ASTM D5890 (West Conshohocken, PA, USA, 2011) standard to measure the swelling tendency of clays, as shown in Fig. 1. In this work, a set of swelling index experiments were performed to determine the swelling tendency of existing clays in the reservoir rock in the vicinity of different water compositions to select water samples suitable for use in dynamic or micromodel experiments (Table 3). To perform swelling index tests, 2 gr samples of dry and finely ground smectite and kaolinite clays were added to different fluids in a 100 ml cylinder, as shown in Fig. 2. The cylinder was then placed in a fixed place and did not move, and in time steps of zero, 15, 30 min, and 1, 2, and 3 h from the start of the test, the height of the clays was read and recorded. The rate of swelling for the clay in the cylinder was obtained by dividing the volume of sediment by the total volume of fluid at different times. The test conditions were atmospheric in case of temperature and pressure. This experiment was performed in 10 experimental groups, the figures and results of which are given in Appendix A.

Micromodel experiments
Generally, micromodel experiments have been used to observe the swelling and migration of clay particles in numerous studies, and the quantitative analysis of results was based on an image processing technique. The micromodel provided and used in the experiments was constructed from two glass slides of the same size with 20 * 8 * 10 dimensions. One of the glasses was lasered to etch a heterogeneous pattern on it which was taken from a CT-Scan of a cross-section of a rock which was extracted by Corel-Draw software as a bitmap and two Flow Dispensers were added to the beginning and end of it ( Fig. 3). Micromodel's characteristics are presented in Table 4 and a schematic diagram of the experimental setup is shown in Fig. 4. The micromodel setup was composed of four parts comprising fluid injection section, micromodel, imaging system, and pressure transmitters. The initial permeability of the bare micromodel (without any clay coating) was measured during formation water injection in some micromodel to provide a glass of micromodel close to average porosity and rock reservoir permeability. In this study, in addition to visual investigations, the permeability of the porous media was calculated during the injection of different fluids into the micromodel by recording the inlet and outlet pressure. In this regard, seven groups of experiments were conducted with different fluid compositions selected from the results of the swelling index test and different injection sequences. Details of the selected fluids are listed in Table 3. Montmorillonite and Table 2 Specifications of solutions used in the swelling index test kaolinite were used in this study as swelling and migration agents. Two pressure transmitters were installed in the inlet and outlet of the micromodel to calculate the pressure drop across the micromodel. The effect of fluid injection with different salinities and compositions on the permeability was evaluated as an effective factor in structural formation damage due to swelling and migration of clays. All injection tests were carried out at ambient temperature under five flow rates of 1.2, 5, 10, 15, and 20 ml/hr. The 1.2 ml/hr rate was carried for 5 h to reach a stable condition in case of swelling and migration and the others were carried on until the same amount of fluid as the 5 h injection with 1.2 ml/hr was injected into the micromodel to obtain the permeability of the micromodel. These rates were selected as the highest rates which didn't induce hydrodynamic migration and were obtained from some preliminary tests with different flow rates. The initial permeability of the micromodel (without any clay coating) was measured during formation water injection. The formation water was injected at different rates and the pressure drops caused by each flow rate were recorded using pressure transmitters. Therefore, the initial permeability   was calculated according to the flow rates and their corresponding pressure drops. As shown in Table 5, thirty-two experiments were designed in 7 categories with 4 clay contents: (1) kaolinite with a concentration of 5% wt. (2) smectite with a concentration of 5% wt. (3) a mixture of 5%wt. kaolinite and smectite (4) a mixture of 10%wt. kaolinite and smectite. These clay contents were prepared as suspensions in 40,000 NaCl brine. Then clay suspensions were injected into the micromodel which was placed in an ultrasonic bath to ensure homogenous distribution of the clays in the micromodel. Then micromodel was placed in the oven at a temperature of 140 °C for 1 h as the final stage of the clay coating. Formation water (FW) was injected to simulate the initial conditions of the formation in the first experiment of each category, then Diluted Formation water (DFW) and Diluted seawater (DSW) was injected to impose damaging conditions in the micromodel by simultaneously causing swelling and migration of clays. Then, by changing the composition of water in two steps: (1) adding nanoparticles and (2) 1% zirconium oxychloride, in 20 times diluted seawater, the effect of these two aqueous solutions on the simultaneous control of swelling and migration was investigated. In the fourth category, seawater (SW) was injected after the formation water to investigate the gradual reduction of salinity effect on swelling and migration. In the 6 th category after the sequence used in the first 3 categories, swelling inhibitor solution was also injected to investigate its effect on swelling. Finally, the order of injected liquids in the last group (seventh) was changed for a mixture of 5% clay, and the effect of swelling and migration changes was compared with changing the injection priority compared to the conditions of previous experiments.
To obtain the permeability after each experiment, the graph of ∆P vs. q is drawn for four rates of 5, 10, 15, and 20 ml/hr, and its trendline was extracted. Using the slope of this line and the Darcy formula (Eqs. 1 and 2), permeability can be obtained for each of the tests. Figure 5(a-f), presents an example of the pressure plots for different flow rates and ∆P vs. q of one of the tests. Pressures were plotted against test duration time.

Results
Using static and dynamic tests which were described in the previous section, we aimed to answer three questions: "Can the formation damage due to swelling and migration of clays in LSWF be controlled?" and "what is the optimal sequence and method to decrease formation damage due to swelling and migration clays?" and "can a migration of clays be useful in decreasing formation damage?" In the following, the results of the swelling index tests and the micromodel tests will be discussed to answer the above questions.

Swelling index tests
Swelling index experiments were performed to evaluate the swelling of clays present in the studied reservoir rock (kaolinite and smectite) in the presence of Formation water and Seawater with different dilution factors. Finally, suitable fluids were selected for the design of micromodel experiments. The results of 10 swelling index tests are described below and the images are shown in Appendix A.
• In the First experiments series, seawater with dilution factors of 2, 6, 10, and 20 times was placed in the vicinity of smectite clay according to the pre-expressed design, which indicates an increase in the swelling of this clay with increasing dilution, and from dilution 10 onwards, the rate of swelling is approximately constant, indicating a critical swelling concentration (CSC) of 4190 ppm for this combination of seawater. Figure 30 shows the percentage increase in swelling. (1) In the second experiments series, times different salinity sea water was placed in the vicinity of kaolinite according to the pre-expressed design. The swelling of kaolinite has not changed with decreasing salinity of seawater (Fig. 31). In all experiments, different water compositions showed the same results as expected, and kaolinite was observed to remain unchanged. Due to the structural features of kaolinite and the lowest CEC value, it showed almost negligible swelling potential.
Seawater was diluted 20 times as shown in Table 3. The divalent ions of magnesium, calcium, and sulfate were removed and the effect of swelling in the presence of smectite clay mineral was evaluated, respectively. It was observed that reducing salinity itself increases swelling and the goal is to control it, this combination was removed from the micromodel experiments.
As can be seen in Fig. 32, the effect of these ions on the swelling of bentonite clay is SO 2− 4 > Mg 2+ > Ca 2+ . This composition sowed no swelling effect in the presence of kaolinite clay minerals (Fig. 33). Formation water with dilution factors of 2, 6, 10, 20, and 40 was placed in the vicinity of smectite clay according to the pre-stated design, which indicates an increase in swelling of this clay with increasing dilution and decreasing salinity. The Smectite clay swelling index has not changed until the formation water has been diluted 4 times and since then the swelling rate has increased. This indicates the critical swelling concentration (CSC) of about 24283 ppm for the water composition of the studied formation water. Figure 34 shows the increasing percentages of smectite clay swelling. In addition, in the vicinity of kaolinite clay, the formation water composition has no swelling effect (Fig. 35). Series of static tests were conducted to investigate the effect of nanoparticle (MgO, SiO 2 , and Al 2 O 3 ) modification on swelling. The SEM images of the used nanoparticles are shown in Fig. 6. As mentioned in previous studies, nanoparticles are not effective in controlling swelling but are suitable for controlling migration. In this series of experiments, the presence of these nanoparticles in 20 times diluted seawater could not be effective in controlling the swelling of smectite clay compared to the swelling of 20 times diluted pure seawater, so in micromodel tests only for migration control will be analyzed (Figs. 36, 37 and 38). • Based on the last series of static experiments, the effect of zirconium oxychloride on swelling control has been expressed. Therefore, in this experiment, the presence of 1% zirconium oxychloride as a clay stabilizer in 20 times diluted seawater significantly reduced the swelling rate of Montmorillonite clay compared to its swelling in the presence of pure 20 times diluted seawater (Fig. 39).

Micromodel tests
To show the appropriate injection method and fluid composition, 7 experimental micromodel group tests were performed to investigate the simultaneous swelling and migration of clays in the presence of different injection fluids and the effect of changing the injection priority or sequence. We defined a formation damage index and investigated the simultaneous effect of swelling and migration on permeability during the flooding of different solutions. However, previous studies did not evaluate the effect of different solutions and various injection scenarios on the formation damage. It was tried to peruse different injection plans to present the best case with lower formation damage. Pressures and ∆P versus flow rate plots were drawn. The term ΔK is used to better describe the changes in the permeability relative to the initial permeability of the clay-free micromodel which is 334 md, and as an indication of how much the permeability of the micromodel can be altered from its original volume in case of clay presence, migration, and swelling and it is almost similar to damage index. It can be calculated as below (Eq. 3): K n : Permeability after injection of different fluids, K mic : 334 md∶ Permeability of the clay − free micromodel.
(3) ΔK% = K mic − K n K mic * 100 Experiment category-1 In the first batch, the micromodel was coated with 5% kaolinite clay and exposed to injected fluids according to Table 6 which are formation water (FW), 40 times diluted formation water (DFW), 20 times diluted seawater (DSW) and nanofluid (NF). As shown in Table 6, the results of kaolinite tests during fluid injection of FW, DFW, DSW, and finally NF, respectively, show an upward trend in permeability and a decrease in the damage index, which indicates migration of clay out of the micromodel (Fig. 7). Regarding the drastic and increasing permeability changes from DFW to DSW, it can be said that due to the slight salinity difference between these two injected fluids, clay migration needs time and doesn't happen spontaneously. As it can be seen from Table 6, after injection of DFW, the permeability of micromodel increased from 59 (initial state of the micromodel due to injection of FW) to 105 md, which is less than two times which indicates migration of clays out of the micromodel due to reduction of salinity of the injected water from 97,132 to 2428 ppm. After the injection of DSW and reduction of salinity from 2428 to 2095, we didn't anticipate much migration due to a slight reduction of salinity, however, permeability increased from 105 to 272 md which is more than two times. As if the migration started during the injection of DFW but its effects in permeability are more sensible due to the injection of DSW, which can be interpreted as a delayed effect of migration.
The resistance of the clays to the migration was observed more in the pressure plots related to DFW, but their release,  mobility, and migration out of the micro-model are seen during DSW injection. Finally, with the start of NF injection, the change in the decreasing slope of the formation damage index in Fig. 8 is the indication of controlling the migration  Experiment category-2 In this batch, the micromodel was coated with 5% montmorillonite clay and exposed to injected fluids including, FW, DFW, DSW, and NF. The results of permeability and formation damage index changes related to micromodel experiments containing smectite clay, which is representative of swelling and slight migration are given in Table 7. In the first stage of FW injection, in Figure 11, the change in clay volume and clay movement can be cited. Of course, as mentioned before, the purpose of this stage was to visualize the initial conditions of the reservoir with the presence of formation water. As can be seen from the process of permeability changes and damage index in Table 7 and Fig.12, by reducing the salinity from FW to DFW, we rapidly saw drastic changes at the beginning of injection due to swelling of smectite clays, and then, these changes are not so different and it can be said that Montmorillonite clay has reacted quickly to swelling (Fig. 13), so we see permeability reduction and a damage index increment. During the DSW injection, by keeping the salinity approximately the same, a slight increase in permeability was observed, which can be considered as the migration of clays out of the micromodel (Fig. 14).  Then, with the injection of nanofluid (NF), which does not perform well in controlling swelling, as expected according to the results of tests 7 to 9 of the swelling index, we see a sharp decrease in permeability even more than the case of FW to DFW because the nanoparticles are not small enough to penetrate inner layers of the clay. Certainly, this increased intensity cannot be attributed to more swelling and it is due to controlling migration out of the micromodel which has a positive effect on permeability. Therefore, the use of nanoparticles in the injection fluid has not been able to control the damage of Montmorillonite clay. In other words, using a clay migration controller in the porous medium with clay that has both abilities of swelling and migration is not useful. In Fig. 15, we see the swelling of the clays, although slight.
• Experiment category-3 In this group of experiments, a mixture of 5% montmorillonite and kaolinite clay with equal amounts of 2.5% each was prepared based on the results of XRD experiments on core samples, the actual range of clay in the study reservoir and the micromodel was coated with this mixture and as in the previous two experiments, was exposed to different fluids.
The decrease in permeability compared to the initial state of the non-clay coated micromodel is simply due to the occupation of the empty micromodel spaces by the clays. Then, with the start of DFW injection due to a significant reduction in salinity from FW to DFW, we see significant changes due to the swelling of smectite clays (Fig. 16). Some flow paths were blocked, which increased the damage index and greatly reduced the permeability compared to the initial conditions Table (8 and Fig . 17). These changes in the image taken from the micro-model (Fig. 16) are also evident from the beginning to the end of the micromodel; no changes were observed in the graphs and images taken at higher speeds, which indicate the stabilization of the clays in terms of swelling and displacement during injection of DFW.
In the continuation of this group of experiments and with the start of DSW injection, due to the slight decrease in salinity from DFW to DSW, no significant changes are observed in the pressure graphs and micromodel photos. Montmorillonite swelling clays stabilized during DFW injection and with the release of a significant amount of kaolinite clays as an indicator and representative of migrating clays and the exit of free clays from the micromodel, permeability of the system increased significantly, although their effects are unfortunately not well seen in the photos, these changes are evident in the numerical calculation of permeability and damage index Table (8 and Fig. 17).
In the final stage of injection of this group of experiments, the start of nanofluid (NF) injection, which has already been proven, does not affect controlling the swelling of the clay, in the changes seen at the very beginning of the injection with the flow rate of 1.2 ml/hr, it can be expected that there is swelling and damage to the formation, which can also be seen in the photos taken from the micromodel (Fig. 18). Also, in the numerical calculations of the damage index, the plot in Fig. 18 is evident. In general, at this stage, it can be said that the use of a migration controller (NF) is not useful because it prevents the clay from moving out of the micromodel.
Experiment category-4 In this experiment, a mixture of 10% clay smectite (montmorillonite) and kaolinite with equal amounts of each 5% were selected according to the results of the XRD test on the formation of core samples to evaluate the gradual decrease in salinity after FW injection. The formation water (FW) was injected to create the initial conditions of the reservoir, then seawater (SW), 40 times diluted formation water (DFW), 20 times diluted seawater (DSW), and finally, a combination of three types of nanoparticles MgO, SiO 2 and Al 2 O 3 in 20 times diluted seawater (NF) was injected. Migration of some kaolinite clays that are attached weaker to the pore walls has been observed. This can be seen in Fig. 19. At the beginning of SW injection into the micromodel, decreasing pressure changes were observed, which can be related to the migration of free clays, and has led to a reduction of 3% of the damage index and increased permeability Table (9); this issue in Fig. 20 is visible. It can be claimed that due to the lack of severe salinity shock to the clays in the micromodel, these changes and damage to the formation are not significant and it can be concluded that a gradual decrease in salinity can reduce damage to the formation. In previous experiments, after FW injection, we had a severe salinity reduction of 40 times. With a decrease in salinity to the salinity of seawater(SW) we did not observe significant damage. In Fig. 21, the migration and swelling of the clays are seen mildly, which according to the results can be said to have neutralizing effects on each other.
With the start of DSW injection, as can be seen from the table of permeability changes and the damage index Table (9), due to the slight decrease in salinity, it is not expected to see drastic changes in the formation damage.
Although swelling and migration are happening, according to the increasing trend of pressure, it can be concluded that swelling is more severe than the migration and exit of clays from the micromodel. (Fig. 22).
In the following, during nanofluid (NF) injection, the inability of the nanofluid to control swelling has caused a slight decrease in permeability Table (9). The main difference between this group of experiments and the previous series can be summarized in two cases: Fig. 16 Smectite swelling by injection of diluted Formation (DFW) water in 5% mixed coated micromodel with the rate of 1.2 ml/hr • The values of pressures recorded in this series of tests are significantly higher than the previous series of tests, which is normal due to the doubling of the amount of clays present in the micromodel. • In this series of tests, the reduction of salinity from FW to the final injection fluid, NF, has been done in a stepwise manner and with a lower slope of salinity reduction, which has led to more slight permeability changes than in the previous series of tests.
According to the explanations given and the permeability results of the Mix 10% test, it can be claimed that by reducing the slope of the salinity reduction of injected fluids, both migration and swelling can be controlled to an acceptable level. Also, another claim can be made based on the results of this test, which is that this trend has had a particularly greater impact on controlling migration than controlling swelling because, as the permeability data show, the swelling has been able to reduce the permeability, but the migration has not increased the permeability, and the migration that has been observed depends on the release of free clays with a weaker connection to pore walls as if the shock of reducing salinity has a greater effect on clay release and migration than on swelling. Changes in the formation damage index are shown in Fig. 23.
• Experiment category-5 In this experiment, the reproducibility of 5% kaolinite and montmorillonite results during injection of different fluids was investigated and in the last stage of injection, a solution of 20 times diluted seawater and one percent of zirconium oxychloride was used as a swelling inhibitor (SI).
By repeating the tests related to Mix 5% and calculating the permeability and the formation damage index Table  (10) and drawing the plots of changes in the formation damage index (Figure 24), it can be seen that the results are the same as the results of the experiment number-3. The change of the injected fluid from FW to DFW saw an earlier onset of swelling than migration, which showed itself in a decrease in permeability. With a slight decrease in salinity from DFW to DSW, swelling is more limited, but the released clays begin to migrate out of the micromodel with a delay, which is evident in the increase in  permeability. By injecting nanofluid and not controlling the swelling and controlling migration, we see a decrease in permeability and so an increase in the formation damage index. At the end of the injection, 20 times diluted seawater and 1% of zirconium oxychloride were injected as a swelling inhibitor (SI). The control effect of this compound on swelling and the formation damage index was investigated. It has been shown with the migration of abandoned clays in the previous stage, there is an increase in permeability and the formation damage index is less than in other experiments.
• Experiment category-6 In this experiment, the repeatability of a mixture of 10% kaolinite and smectite without the presence of seawater (SW) in the series of injected fluids of this group was performed to compare the effect of gradual reduction of salinity, and also investigate the effects of injecting swelling inhibitor (SI). In this group of tests, with a rapid decrease in salinity from FW to DFW, we see a sharp decrease in permeability due to swelling, and in the next step, the delayed migration in comparison with the swelling at the time of DSW injection, despite slightly salinity changes, we see the migration of clays out of the micromodel and which can be seen as an increase in permeability Table ( 11). With the start of NF injection, we still see an increase in the permeability, which can be due to the slight migration of clays and the stop of the swelling because of severe swelling in the injection of DFW. Swelling of montmorillonite clays, which has occurred significantly due to the shock of rapid salinity reduction, has also been somewhat effective and the flow paths have been opened, and finally, the result of these reactions has improved the formation damage index by 8%. Following the injection of the Swelling Controller or Inhibitor (SI), we see a slight decrease in permeability, which can be suggested for several reasons: (1) A large amount of swollen montmorillonite clay in this series compared to the previous test series (2) Migration of kaolinite clays and throat blocking because, during the previous fluid injection, the throat became narrower than before due to swelling and presence of more clays in the micromodel and provided more favorable conditions for throat blocking. It should be noted in this case the presence of more clay in this series of tests is not ineffective compared to the previous series.
• Figure 25 summarizes the comparison of Experiments 4 and 6, which, as can be seen with a 6% difference at the beginning of the experiment, show the initial conditions of montmorillonite and kaolinite mixed clays in the vicinity of the FW in the micromodel. Assuming that both experiments are with equal conditions, we follow the next fluid injection steps during the series of experiments. Finally, both methods can lead to the initial condition of the reservoir. Inherently, by injecting seawater with gradually decreased salinity and in the final stages of NF and SI injection as controllers for swelling and migration of clays, a very successful scenario in using the IOR by smart water injection is experienced. • Injection Priority Change or Sequence Change Experiments To investigate the effect of fluid injection sequence on formation damage caused by the Simultaneous effect of swelling and migration, one subsequent experiment (experiment 7) was performed by changing the fluid injection sequence for a mixture of 5% kaolinite and smectite clays. Experiment category-7 In this experiment, the priority of fluids injection was shifted in the micromodel with a mixture of 5% kaolinite and montmorillonite clays respectively to FW, NF, SI, DFW, and DSW. By injecting Formation water (FW), as in previous experiments, the initial conditions of the clays were created to simulate the reservoir in the micromodel. Then, it is time to inject the fluid containing nanoparticles, from which we expect migration control. Nevertheless, the noteworthy point in these graphs is the general trend of relatively decreasing pressure, which is expected to show itself as an increase in permeability.
In the following, by starting the injection of diluted formation water (DFW), with investigation and analysis, it can be said that the inhibitor effect disappeared and in fact, the swelling of clays is at this stage. With diluted seawater  injection (DSW) and due to the lack of high salinity reduction compared to the previous injection fluid "DFW", it can be claimed that the changes seen in the pressure diagrams are due to the nanoparticle effect in the previous stage, delay in migration compare to swelling, and in fact, due to the migration of released clays. As expected, and explained in the pressure plots, with NF injection we saw an increase in permeability Table (12) and a decrease in the formation damage index (Fig. 26). However, with DFW injection, we see a decrease in permeability and an increase in the formation damage index due to the gradual disappearance of the SI effect and the onset of swelling. Due to the salinity remaining constant from DFW to DSW, we do not expect further swelling, but due to the leaching of nanoparticles (NF), controlling migration in the previous stages, and the delay of migration relative to swelling, we see a slight increase in permeability. We have seen the effect of swelling inhibitor (SI) injection, which can be suggested using this type of swelling inhibitor (one percent zirconium oxychloride in 20 times diluted seawater) in LSWF scenarios.  Fig. 27. The red line trend of Test No. 5 shows the FW-DFW-DSW-NF-SI injection sequence and the blue trend of Test No. 7 shows the FW-NF-SI-DFW-DSW injection sequence. The blue line trend shows that when the effect of the controllers is removed from the micromodel by continuing to inject DFW and DSW, the formation damage index returns to the previous range of test No. 5 and increases the formation damage. Conversely, when they were used at the last stages, in Experiment 5 after DFW and DSW injections, they did not succeed in controlling the formation damage and did not show a significant effect. In general, the migration controllers (NF) and swelling inhibitors (SI) are very good at controlling formation damage in these conditions and should be present in the injected fluid (Fig. 28).

Conclusions
Microfluidic tests were performed with the presence of kaolinite and montmorillonite clays in the vicinity of smart water designed and engineered in this study. Figure 29 shows the flowchart of this research and the results of each step.
1. Gradual reduction of water salinity in the smart water injection methods can be useful in reducing the formation damage caused by swelling or migration of clays.
2. At the time of injection of smart water, despite the simultaneous phenomenon of migration and swelling at the same time, swollen clays have a greater impact on the formation damage and eliminate the positive effect of clay migration. 3. During water injection, when the critical salinity concentration is high, more clay is removed from the system and the permeability is improved, and as a result, the damage is reduced when the salinity reaches less than the critical concentration. 4. In a 5% mixture of kaolinite and montmorillonite clays, migration has more effect and is effective in controlling damage to the formation because of migration of clays to the out-of-porous media and lack of a barrier for clay entrapment in pore throats in comparison with the clay coated micromodel with 10% mixture of clays. 5. Nanofluid composition (NF) controls migration and in the case of mixing clays with both the migration and swelling abilities, increases the formation damage therefore, it is better not to use it in the conditions of simultaneous swelling and migration. It is recommended to use the combination of NF and SI in smart water injection where there is a combination of swelling and migration. 6. The Swelling inhibitor (SI) has been very effective in controlling swelling compared to other injected fluids. 7. In reservoirs where the amount of a mixture of swollen and migratory clays is high, more than 5% was considered here, it is better to allow the migration of clays to the point of leaving the reservoir by digging production wells in the vicinity of injection wells and using only "swelling control compound". 8. In reservoirs with more swellable clays, the use of migration control fluid or nanofluid (NF) due to con- trol of their movement clays will stop and settle in the pore throats and block the flow paths and as a result will intensify the damage to the formation.