For over a century, water-flooding has been one of the most successful and efficient secondary oil recovery method and can be implemented in the early or late stage of oilfield development^[1,2,3]. Researches in recent years show waterflooding effect is strongly affected by the salinity and composition of the injected water. Quite a number of researchers have confirmed that compared with injecting sea water, smart water flooding, that is injecting sea water with modified salinity and ion composition can affect interactions between oil, brine and rock, enhance the microscopic displacement efficiency and finally oil recovery of carbonate reservoirs^{[1,2,3,4,5,6,7,8,9,10,11,12,13]}. The study of Zhang et al.^[14] showed that increasing the concentration of SO₄^2- could improve the wettability of chalk layers. The research of Fathi et al.^[9] showed that the ultimate oil recovery improved dramatically as the concentration of SO₄^2- in the sodium chloride-free seawater increased to 4 times that of ordinary seawater. Chandrasekhar and Mohanty et al.^[15] found that modified seawater containing Mg²⁺ and SO₄^2- could change wettability of calcite and make calcite more water-wet. Awolayo et al.^[6] carried out a series of tests on Middle East carbonate core plugs and observed that increasing the concentration of sulfate in the seawater could make the crude oil/brine/rock system less oil-wet. Apparently, injecting smart water can change the wettability of carbonate rock, making it shift from oil-wet condition to water-wet condition, and thereby enhancing oil recovery. But if the injected smart water is poorly compatible with the original formation water, minerals could precipitate to form scale, which in turn could plug the matrix pores or fractures in oil-bearing formations, perforation channels, and down hole equipment, leading to decrease of oil and gas production at last^{[16,17,18,19,20,21,22]}. The most common mineral scales in oilfields are divided into two categories, sulfate and carbonate scales. After injection, the SO₄^2- in the smart water would react with the Ca²⁺ in the formation water to form CaSO₄^{[23,24,25,26]}; the CO₃^2- in the smart water would react with the Ca²⁺ in the formation water to form CaCO₃^[27,28]. Generally, the formation water of carbonate reservoir has high salinity and high Ca²⁺ concentration. The mixing of smart water with high sulfate concentration and formation water would lead to the formation of CaSO₄ scale and much smaller amount of CaCO₃ scale. The CaSO₄ scale is a main kind of scale in oil industry that can bring major troubles.

Researchers have conducted some experiments to find out the impact of in-situ mineral scale formation on reservoir permeability. Moghadasi et al.^[23] found through experiments that the permeability of reservoirs after water injection could drop by less than 30% to more than 90% due to the effect of CaCO₃ or CaSO₄ scale; the degree of permeability decline depended on the composition of the formation water and injected water, initial permeability and temperature of the reservoir, water injection duration and rate etc. Bin Merdhah and Yassin^[25] performed core flooding tests to study the permeability decline caused by common oil field scales formed in the sandstone cores when injected sea water and formation water with high concentration of barium, strontium, and calcium ions mixed in different conditions. The results showed that the CaSO₄ and SrSO₄ scale increased as the temperature rose, but BaSO₄ scale decreased because the solubility of BaSO₄ scale increased with the rise of temperature; all of them increased at higher differential pressure and higher salinity of injected water and formation water. Tahmasbi et al.^[24] investigated the permeability decline of packed columns of carbonate grain and glass bead when injected waters with high concentrations of calcium and sulfate ions mixed. The results showed that with the increase of concentrations of these ions in the injected water, the CaSO₄ precipitation increased, and the permeability of the packed columns dropped continuously; the permeability reduction rate didn’t decrease with the increase of the injection rate; at a given concentration and injection rate, the rate of permeability decline became faster with the rise of temperature. Moghadasi et al.^[29] found that in-situ scale precipitation would result in significant drop of permeability of sandstone cores; the calcium sulfate and strontium sulfate scales increased with the rise of temperature, but the precipitation of barium scale decreased; and the precipitation of all of them increased with the rise of pressure. Naseri et al.^[30] found through study that the composite BaSO₄-CaSO₄ scaling was opposite to the BaSO₄ scaling, the higher the temperature, the more the amount of composite BaSO₄-CaSO₄ scale, and the more rapidly the permeability of porous media would decrease. It can be seen that factors affecting mineral scale and consequently permeability include temperature, pressure, type of scale, salinity of injected and formation waters, water injection rate and initial reservoir permeability etc. Based on published researches, most of the previous smart water flooding plans focused on EOR only but paid little attention to the formation damage and reduction in the productivity of reservoir caused by water injection.

We carried out^[31] static compatibility tests of mixing brines with different compositions at the atmospheric pressure and the temperature of 90 °C to find out the impact of salinity and ionic concentration of smart water on CaSO₄ scaling by removing the HCO₃^- in sea water to simplify the sea water composition. In this work, core flooding experiments were carried out under newly designed water injection program and experimental parameters to model the reaction of smart water and original formation water in carbonate reservoir to find the effect of scaling on reservoir permeability more accurately. Experiments were carried out in this work to study risk of scaling when smart water was used as injected water to enhance oil recovery, how to lower the risk of scaling by adjusting the salinity and ion composition of the smart water and the reduction of permeability caused by composite CaSO₄ and CaCO₃ scaling, and analyze the crystal size and morphology of composite CaSO₄-CaCO₃ scale on the rock pore surface by SEM photographs.

Six kinds of high purity analytical grade salts provided by Merck company, including NaCl, KCl, Na₂SO₄, NaHCO₃, MgCl₂•6H₂O and CaCl₂•2H₂O were used to prepare the synthetic brine solutions. The predetermined amounts of the salts were dissolved in deionized water to prepare the brine solutions. The brines were stirred by a magnetic stirrer and set aside for 24 hours to ensure that the brines were stable. Also to ensure that there were no suspended particles in the solution, the brines were filtered with 0.45 μm filter paper. The synthetic Persian Gulf seawater (PGSW) and the synthetic Koupal field formation water (KFW) were prepared by the above preparation method, and their chemical compositions are shown in Table 1. Modified brines were made based on the composition of Persian Gulf seawater to study the composite CaSO₄ and CaCO₃ scale formed at different salinities and ion compositions of smart water and the effect of the scale on reservoir permeability. Two sets of modified brines were prepared to study the effect of salinity and ionic composition of the injected smart water on the permeability reduction caused by the composite CaSO₄ and CaCO₃ scale. One set of modified brines with constant salinity (Table 2) was used to study the effect of ionic composition of injected smart water of scaling. The PGSW_1Ca1Mg2S means the concentrations of Ca²⁺, Mg²⁺ and SO₄^2- of the modified brine are 1, 1 and 2 times those of the PGSW, and so on and so forth. The other set of modified brines (Table 3) was prepared by diluting the seawater to study the effect of injected water salinity on scaling. PGSW#5D means the PGSW is diluted 5 times, and so on so forth.

All the core samples were taken from the carbonate dolomite outcrop in the Babakoohi mountain of Shiraz, which were low in clay content. The core samples were 7 cm in length, 3.8 cm in diameter, 13.61% in average porosity and 33.37×10^-3 μm² in absolute permeability. All the core samples were cleaned at ambient temperature by toluene in a Soxhlet extractor, and dried in an oven at 100°C for 24 hours before use. The physical properties of all the core samples are listed in Table 4. X-ray diffraction (XRD) analysis of the rock core samples shows that the core samples have more than 99% of dolomite particles, as shown in Fig. 1.

Fig. 2 shows the core flooding apparatus. Before each core-flooding run, the core sample was first saturated with KFW and then the pore volume and initial porosity were calculated. Thereafter, the saturated core was flooded with KFW at variable flow rates and the corresponding differential pressures were recorded, the initial core permeability was calculated with Darcy’s Law, and the results are given in Table 4.

After each core-flooding test, the used core was restored and prepared for the next core-flooding as follows. (a) If the rock core sample contained a small amount of clay: (1) the core was flooded with 6 PV of distilled water; (2) followed by 2PV of toluene to completely clean the rock core sample; (3) finally, the core sample was dried in an Oven at 100 °C for 24 hours to be prepared for the next core-flooding run. (b) If the rock core sample contained a high percentage of clay: (1) the rock core sample was flooded with 6 PV of saline water without SO₄^2- and CO₃^2- ions; (2) followed by 2PV of toluene to completely clean the core; (3) finally, the core sample was dried in an Oven at 100 °C for 24 hours to be prepared for the next core-flooding test^[32].

Three groups of core-flooding experiments were conducted under normal pressure at 90 °C. In the first group, the synthetic PGSW was used to displace the core saturated with KFW to test the extent of permeability decline caused by composite CaSO₄ and CaCO₃ scale. The second group of core-flooding experiments were performed to investigate the effect of contact time (CT) of injected water and formation water in the porous media on the scale quality. CT was a new parameter designed in this work. The third group of core-flooding experiments were conducted to study the salinity and ionic composition of injected PGSW on the permeability decline due to formation of in-situ composite CaSO₄-CaCO₃ scale. Finally, in order to study the morphology of deposited scale on the rock pore surface, the SEM photographs of scaled core samples were analyzed visually.

A new experiment procedure was designed to model the flow of injected water and formation water in the real reservoir conditions. After the porosity and initial core permeability were measured by injection KFW into the core, the transfer cell containing KFW was switched to the transfer cell of smart water and placed inside the oven together with the core holder assembly, and heated to the temperature of 90 °C. The system was left for 4-5 hours to reach the desired temperature. All core-flooding experiments were carried out at a constant flow rate by setting the injection pumping rate at 0.5 mL/min. First, 2PV of the smart water was injected into the core sample. After the water injection, the contact time was set to leave the incompatible injected water and formation water to contact and react with each other fully. Afterward, 3PV of smart water was injected into the core sample and the differential pressure between the two ends of core sample was recorded continuously. The final permeability of core sample was calculated using Darcy's linear-flow equation after scale formation. After flooding, the scaled rock core sample was taken out, dried in an oven at 100°C and cut into smaller pieces for SEM analysis.

The thermodynamics of mineral scaling by mixing the sulfate-rich PGSW and KFW was first analyzed with the OLI ScaleChem software to find out the scaling minerals under the experimental conditions in this work. A chemistry simulation program was developed based on thermodynamic balance and solubility of ions in water to predict scaling during water injection in the reservoir^[22].

The OLI ScaleChem software was used to predict the type of scale could be formed, to calculate the scaling potential and the amount of precipitated scale at the temperature of 90 °C and atmospheric pressure. The chemical compositions of the mixed brines used to predict the potential of mineral scale formation are shown in Table 1. The potential of scaling was evaluated by the scaling trend (ST), when the ST is greater than 1.0, the scaling potential is high.

Fig. 3 shows the potential scaling tendency as a function of PGSW ratio. The ratio of PGSW is the ratio of PGSW in the mixed solution in volume, for example, the ratio of PGSW is 80% means in 100 L of the mixed solution, there is 80 mL PGSW and 20 mL of KFW. The results show the maximum ST of CaSO₄ and CaCO₃ are 4.9 and 2.8 respectively. Clearly, the CaSO₄ scaling is higher. Fig. 4 shows the amount of scale formed as a function of ratio of PGSW in the case formation water and seawater are mixed. The results show the amount of calcium sulfate scale was the maximum of 2060.9 mg/L and the amount of calcium carbonate was the maximum of 13.4 mg/L at 80% PGSW. Calcium sulfate is the main type of scale in the smart water flooding system described herein.

All the core-flooding experiments were carried out with single phase fluid at a constant flow rate of 0.5 mL/min. The final permeability of the core sample was calculated with the recorded differential pressure data by Darcy's linear-flow equation. The permeability ratio was then plotted vs. pore volume of injected brine. The experimental results are shown in Table 5. Finally, the scaled core sample was observed visually by SEM to find out the crystal size and morphology of composite CaSO₄-CaCO₃ scale on the rock pore surface.

2.2.1. Degree of permeability damage

The synthetic PGSW was injected into core sample saturated with KFW to study the degree of permeability reduction caused by the composite CaSO₄ and CaCO₃ scale. In this test, the contact time was 0 h, that is initial contact, corresponding to the experiment No.1 in Table 5.

Fig. 5 shows the permeability of the core decreased by about 61% due to the formation of CaSO₄-CaCO₃ scale. The permeability decline trend changed with the pore volume of injected PGSW. During the initial period of PGSW injection (one pore volume injected), the permeability decreased sharply. Then the decrease rate of permeability slowed down, and permeability gradually became stable after the sharp drop. This is because in the early period of PGSW injection, the composite CaSO₄- CaCO₃ became saturated fast and then precipitated in the rock pores into scale, blocking the pores.

After the nucleation of CaSO₄ and CaCO₃ scale, at the beginning, the permeability reduction is mostly caused by crystal growth. Later, as crystals deposit on the pore surface continuously, scale occupies more and more pore spaces. Crystal growth and crystal deposition are slower than scale nucleation. Thus, the reservoir damage includes three stages, scale nucleation (or precipitation), crystal growth, and in-situ crystal deposition^[33].

Fig. 5 shows that the permeability drop has fluctuations. This is because several suspended scale particles may bridge across a pore throat, and a single scale particle may block a pore throat. The blocking and bridging of pore throat by suspended scale particles can cause the differential pressure on two side of pore throat to increase, until the particles passing the pore throat and the bridge break by supplied pressure, and consequently the differential pressure decreases. This phenomenon occurs repeatedly during the injection of PGSW^[15].

2.2.2. Effect of CT on scale precipitation

As mentioned above, the experiment was conducted to find out the CT between injected water and formation water when the amount of composite CaSO₄-CaCO₃ scale reached the maximum. To be specific, four experiments at the contact time of 0, 12, 24, and 48 hours were performed, corresponding to core-flooding runs 1, 2, 3, and 4, respectively in Table 5. In these experiments, 2PV of PGSW was injected first, after reaching the set contact time, more PGSW was injected and the pressure difference of the two ends of the core sample was recorded to calculate the final permeability. Fig. 6 shows with the increase of contact time, the amount of composite CaSO₄ and CaCO₃ scale after mixing of PGSW and KFW increases and the ratio of final permeability to initial permeability decreases. After the injected water mixed with the formation water for 12 hours within the core, the composite CaSO₄-CaCO₃ scale had caused severe reservoir damage, leading to higher permeability loss. When the contact time extended to 24 h and 48 h, the permeability loss didn’t increase significantly. The next other core- flooding runs were all completed at the contact time of 12 hours.

2.2.3. Effect of ionic composition of injected water on scaling

The chemical composition of injection water is one of the most important factors affecting the scaling tendency. To examine the effect of ion composition of injected water on permeability decline, a group of core-flooding tests were carried out, namely experiments 1, 5, 6, and 7 in Table 5. The chemical compositions of PGSW are shown in Table 2.

Fig. 7 shows the amount of composite CaSO₄-CaCO₃ scale decreases with the increase of Mg²⁺ concentration in the injected smart water, leading to the increase of the ratio of final permeability to initial permeability. Ions do not exist totally as free ions in the solution, and ion pairing in the solution is very complex. If bound ions are not available for scale formation and thus they increase the solubility of mineral salts by decreasing the reactants’ available concentration^[31]. In these core-flooding tests, the chemical composition of the injected brines contains magnesium ion, sulfate ion, and carbonate ion, and the magnesium ion can make complex or undergo ion pairing with sulfate ion and carbonate ion, lowering the concentrations of sulfate ion and carbonate ion, so less composite CaSO₄-CaCO₃ scale will be formed.

Fig. 8 shows with the increase of sulfate ion concentration in the smart water, the amount of composite CaSO₄-CaCO₃ scale formed increases, and the ratio of final permeability to initial permeability decreases. This is because super-saturation increases dramatically when the concentration of sulfate ion in the smart water exceeds that in PGSW, reactivity speed of the sulfate ions with calcium ions increases as well, and the amount of calcium sulfate scale formed increases, leading to severe reservoir damage and permeability loss.

2.2.4. Effect of salinity of injected water on scaling

Previous laboratory studies showed that injecting low salinity brine had a significant effect on oil recovery^{[1, 8, 26]}. To investigate the potential damage of injecting low salinity brine, a group of experiments were designed, namely, experiments 1, 8 and 9 in Table 5. The chemical composition of the injected water in the experiments is shown in Table 3.

Fig. 9 shows changing the injected water salinity has significant impact on the amount of scale formed and in turn permeability. The salinity of injected fluid can be lowered by dilution. When the PGSW was diluted 5 times, the ratio of final permeability to initial permeability increased to some extent, but with further dilution, the ratio decreased. Through an earlier study^[31], we found that the salinity and concentration of active ions of solution controlled the solubility of mineral salts. Salinity has impact on scale precipitation opposite to the concentration of active ions. The amount of scale precipitated decreases first and then increases with the increase of the salinity of solution, and there is an optimum salinity (PGSW diluted 5 times under the experimental condition in this work) at which the mineral salts have the highest solubility. The concentrations of reagent ions such as Ca²⁺, CO₃^2- and SO₄^2- in the solution reduced by diluting PGSW 5 times with deionized water (10% salinity of seawater), so the activity product of reagent ions decreased. In addition, at this salinity (5 time diluted PGSW), the salinity of injected PGSW is still the effective parameter compared with the concentration of reagent ions, and leads to a reduction in the composite CaSO₄-CaCO₃ scale formed. However, when the PGSW is further diluted (i.e. 10 times), the effects of salinity and concentration of reagent ions will reverse, the activity product of reagent ions will increase, and the activity of anions and cations will enhance, resulting in the formation of more composite CaSO₄-CaCO₃ scale, and in turn reduction of permeability.

2.2.5. Scanning electron microscopy (SEM) analysis

SEM was used to study the shape and particle size, deposition sites, and distribution of scale crystals on the rock pore surface. Take the experiment # 7, the core sample with scale was removed at the end of water-flooding, dried and cut into smaller sections for SEM analysis.

Fig. 10 shows the SEM images of the core sample before scaling. Fig. 11 shows the SEM image of the composite CaSO₄-CaCO₃ scale crystals in the rock pores, it can be seen some crystals gather and are perpendicular to the pore surface. Compared with scale crystals on the rock surface, scale crystals perpendicular to the pore surface have greater resistance to flow of formation water. Therefore, the permeability loss of core samples is mainly caused by these crystals. The scale crystals are irregular in shape, some are tabular and some plate-like. They deposit on top of each other and grow perpendicular to the pore wall. Also, both single and conglomeration of crystals are observed.

When there are bivalent ions in the formation water of carbonate reservoir, injecting smart water has the risk of scaling. The salinity and ionic composition of the injected water have strong impacts on the extent of permeability loss. Precipitation of composite CaSO₄-CaCO₃ scale is the main cause of permeability decline of carbonate reservoir. The permeability decline caused by composite CaSO₄- CaCO₃ scale formation in the porous media ranged from 61% to 79.1% of the initial permeability under the experimental conditions in this work. The extent of permeability decline depends on the ion composition and salinity of the injected water, and contact time of the injected water and formation water.

The results show that increasing the magnesium ions concentration or decreasing SO₄^2- concentration of the injected smart water can reduce the amount of composite CaSO₄-CaCO₃ scale formed, and in turn the permeability loss.

Changing the salinity of injected seawater has a significant effect on the scale precipitation in porous media, and in turn on the permeability decline. The salinity of seawater can be reduced through dilution. When the seawater was diluted 5 times, the ratio of final permeability to initial permeability increased somewhat, but the ratio decreased with further dilution.

SEM images confirm composite CaSO₄-CaCO₃ deposits and grows on the rock pore surface, and the composite CaSO₄-CaCO₃ scale crystals often grow perpendicular to the pore wall. The permeability loss of core is mainly caused by crystals growing perpendicular to the pore wall.