Synthesis and plugging behavior of fluorescent polymer microspheres as a kind of conformance control agent in reservoirs

The fluorescent polymer microsphere is a newly developed chemical agent for conformance control in reservoirs. In this paper, one kind of fluorescent polymer microspheres P(AM-BA-RhB) was synthesized via the inverse suspension polymerization method with Rhodamine B as a fluorescence functional monomer. Laboratory experiments were performed to characterize the morphology, fluorescent property, swelling property and plugging behavior of fluorescent polymer microspheres. The experimental results showed that the polymer microspheres P(AM-BA-RhB) displayed stable fluorescence performance in solutions containing metal ions at pH values between 3.0 and 10.0. The swelling property was not dramatically affected by the Rhodamine B embedded in the polymer microspheres by grafting. Both a visual micromodel test and sand-pack tubes experiment demonstrated that the fluorescent polymer microspheres could pass directly or by deformation through porous media and get into the in-depth formation. The injection pressure showed the phenomenon of “Wave-type Variation”. Three plugging behaviors such as piston plugging, protruding plugging and fingering plugging were put forward. The introduction of fluorescent polymer microspheres could provide one method to research the conformance control and EOR mechanism of polymer microspheres in the reservoirs.


Introduction
According to previous reports, reservoirs with induced fractures or high permeability channels, commonly called thief zones or streaks due to extensive water ooding, are quite common in mature oil reservoirs. Low efficiency of the injected water, which would lead to excessive water production and rapid production decline, has become one of the most crucial problems in the late stage of development for mature oilelds. 1-3 Therefore, it is important for the petroleum industry to develop more reliable techniques such as "green" water shut-off or conformance control. Gels have been introduced as water plugging agents to mitigate this problem. 1,[4][5][6][7][8] The gel application plug fractures and redirects water from high permeable zones to lowpermeability areas. A new trend in gel treatments is using polymer microspheres since they can overcome some distinct drawbacks inherent in in situ gelation systems. [9][10][11][12] The polymer microsphere is a kind of viscoelastic conformance control agent with three-dimensional structure and can absorb the formation water and migration in the porous medium of the reservoirs under the injection pressure. [13][14][15] Field test results in Shengli, Dagang and Jidong Oilelds in China showed that polymer microsphere was a promising conformance control agent in the development of heterogeneous reservoirs, especially the fractured reservoirs. [16][17][18] Hua and Lin et al. investigated the shape, size, rheological properties, plugging properties, prole control mechanism and oil displacement mechanism of the nanoscale polymer microspheres. They found that polymer microspheres could reduce water permeability because the microspheres adsorbed, accumulated and bridged in the pore throat, and the adsorbed layers would be collapsed under the pressure, entering deep into the reservoir due to the good deformation properties of the microspheres. 13 Yao and Wang et al. researched the effects of ionic strength on the transport and retention of polyacrylamide microspheres in porous media. 19 Yang and Kang et al. researched the mechanism and inuencing factors on the initial particle size and swelling capability of polymer microspheres from the synthesis and reservoir condition. 20 Yang and Xie et al. optimized the injection parameters of polymer microspheres and polymer composite ooding system, and they found that the composite system could better improve polymer ooding at the displacement rate of 3.5 m d À1 and the injection volume of 530 mg L À1 PV. 21 During the application of polymer microspheres, polymer solution was injected into the reservoir along with polymer microspheres. 21,22 Due to the same amide functional groups in polymer microspheres and polymer solutions, the conventional concentration method of polymer microspheres such as starchcadmium iodide method, chemiluminescent nitrogen method, ammonia electrode method and organic carbon content method were not accurate. [23][24][25] So it is difficult to research the plugging behavior and conformance control mechanism from a quantitative perspective. For these reasons, uorescent polymer microspheres could overcome the limits of conventional methods and achieve the real-time detection of polymer microspheres concentration and the distribution of polymer microspheres in the reservoir during the ooding process. Fluorescent polymer microspheres have a dual function, one is the conformance agent and the other is the oil eld tracer. The oil eld tracer will be a new application of uorescent polymer microspheres in the development of oilelds.
Under most circumstances, uorescent polymer microspheres are physically dyed through absorption or embedment, which results to the problem of dye leakage and limits the eld applications of these uorescent polymer microspheres. To avoid dye leakage, dyes could be covalently incorporated into polymer microsphere by the inverse suspension polymerization. [26][27][28] Rhodamine B (RhB), which has been widely used in medicine, environmental protection, textiles, colored glass, and cosmetics etc., is a commonly used commercial uorescent dye. [29][30][31][32] Herein, we report one kind of uorescent polymer microspheres P(AM-BA-RhB) acted as a novel conformance control agent. The uorescent polymer microspheres were covalently dyed with Allyl Rhodamine B by inverse suspension polymerization. The polymer microsphere was characterized by uorescence microscope, environmental scanning electron microscope (ESEM) and scanning electronic microscopy (SEM). The effects of salinity and temperature on the swelling property of the polymer microsphere were studied by weighing method. In addition, in order to dene the scope of application, the uorescence spectroscopy was used to investigate the effects of pH and ionic species on the uorescence intensity of the polymer microsphere. In order to monitor the concentration of uorescent polymer microspheres, the relationship between uorescence intensity and concentration of uorescent polymer microspheres was established. On the basis of these results, the plugging behavior of the polymer microsphere was investigated using core ooding test and visual micromodel. These researches provide theoretical support for the further study of uorescent polymer microspheres in the application of oilelds.

Materials
Allyl Rhodamine B (RhB) was synthesized by the method illustrated in ref. 33

Synthesis and purication of uorescent polymer microspheres
The polymerization was conducted in a 250 mL four-neck ask. The ask was equipped with mechanical stirrer, reux condenser and a constant temperature water bath. The uorescent polymer microsphere P(AM-BA-RhB) was prepared via inverse suspension polymerization using cyclohexane as the continuous phase and Span-60 as the nonionic polymeric surfactant (Fig. 1). 0.07 g MBA, 14.06 g AM, and Na 2 CO 3 were dissolved in 40 mL distilled water; this solution was immediately poured into 60 mL of cyclohexane (containing 0.005 g Allyl Rhodamine B and 0.45 g Span 60) which was previously purged with dry nitrogen steam in the ask. Air was ushed from the reactor by introducing nitrogen until the entire process completed. Aer aqueous droplets in continuous phase appeared, 10 mL APS aqueous solution (75 g L À1 ) was added to continuous phase to initiate polymerization. The polymerization lasted for 4 h at 50 C. Aer the prescribed time, uorescent polymer microsphere appeared when stirring was stopped. Then, the precipitates were washed with large amounts of anhydrous ethanol and ltered with qualitative lter paper. The product was dried in a vacuum oven at 45 C for 24 h. The conventional polymer microsphere of P(AM-BA) was prepared by the same process as described above, without adding the Allyl Rhodamine B in the reaction system.

Characterization methods
FT-IR spectra of uorescent polymer microsphere P(AM-BA-RhB), conventional polymer microsphere P(AM-BA), and acrylamide monomer were measured with a Nicolet model NEXUS670 spectrometer with samples prepared on KBr pellets.
The surface morphology of these uorescent polymer microspheres was observed by inverse uorescence microscope (Leica DMI 3000B, Germany) and scanning electron microscopy (SU8010, HITACHT, Japan).
The average particle size of uorescent polymer microsphere before and aer swelling was determined at room temperature by RISE-2006 laser particle size analyzer provided by Jinan Runzhi Technology Co., Ltd of China.
Environment Scanning Electron Microscope (ESEM) images were taken by FEI Quanta 200 FEG (FEI Company, Holland) to observe the structure of uorescent polymer microsphere aer swelling.
To measure the swelling ratio of uorescent polymer microspheres, a tea bag (i.e., a 100 mesh nylon screen) containing pre-weighed dry samples was immersed entirely in brine water (1 g L À1 NaCl). Aer being hydrated in brine water for various times, the swollen uorescent polymer microspheres were allowed to drain by taking the teabag from the brine, followed by removal of excess surface water using lter paper. The swelling ratio, S w was calculated from the following equation: 20 where m 0 and m 1 are the weights of the dry and swollen uorescent polymer microspheres, respectively.

Plugging behavior experiments of uorescent polymer microspheres
2.4.1 Visual micromodel test. The visual micromodel was used to study the migration mechanism of uorescent polymer microspheres. The visual micromodel is a kind of etched glass model. The etched glass model was prepared by laser etching, and pore throat diameter from 300 mm to 400 mm. The whole size of the visual micromodel in the experiment was 6.3 cm Â 6.3 cm Â 0.4 cm. The actual internal porous model size was 4.0 cm Â 4.0 cm Â 0.4 cm. Two pores are located on the diagonal of the model to simulate the injection well and production well. The basic experimental setup is illustrated in Fig. 2. An accumulator with magnetic stirrer was used to contain the dispersed uorescent polymer microspheres solution. The  procedures of the experiment are as follows: (1) saturating the visual micromodel with brine water; (2) injecting the dispersed uorescent polymer microspheres solution with concentration of 2000 mg L À1 into the visual micromodel, and the injection rate was 0.003 mL min À1 . Record the dynamic process of polymer microspheres in the porous medium; (3) clean the visual micromodel; (4) analysis the images and conclude the migration law.
2.4.2 Core ooding test. The multifunction displacement device was used to study the plugging mechanism of uorescent polymer microspheres. The experimental setup in this study (see Fig. 3) was constructed from a sand-pack 60 cm in length and 2.5 cm in diameter. It was packed with different sand grains to simulate the reservoir. An ISCO pump was used to inject the dispersed uorescent polymer microspheres solution and brine water from accumulators to the sand-pack model. An accumulator with magnetic stirrer was used to contain the dispersed uorescent polymer microspheres solution. Pressure gauge was mounted on the inlet to monitor the injection pressure during the whole injection progress. Cylinder mounted on the model outlet was used as a collector to record the volume of uorescent polymer microspheres and brine water production. The Fluoromax-4 spectrometer was used to evaluate the uorescence intensity and get the concentrations of uorescent polymer microspheres at different stages. The injection concentration of uorescent polymer microspheres P(AM-BA-RhB) used in this experiment was 2000 mg L À1 and experimental temperature was 50 C. The procedures of the experiment are as follows: (1) vacuuming and saturating the sandpack with brine water, then calculating the porosity of the sand-pack model; (2) measuring the permeability of the sandpack model with brine water (1% NaCl) at different injection ow rates (0.1, 0.5, 1.0 and 1.5 mL min À1 ) and getting the average permeability; (3) 2 PV (Pore Volume) dispersed swollen uorescent polymer microspheres solution were injected into the sand-pack model at a rate of 0.5 mL min À1 , and then injected the brine water at the same injection rate. Recorded the injection pressure and collect the production liquid; (4) measuring the uorescence intensities at different ooding stages and calculating the concentration of uorescent polymer microsphere according the relationship between uorescence intensity and uorescent polymer microsphere; (5) cleaning the sand-pack model and the pipe lines.

Results and discussion
3.1 Structure characterization of the uorescent polymer microspheres P(AM-BA-RhB) Fig. 4 shows FT-IR spectra of uorescent polymer microspheres P(AM-BA-RhB), conventional polymer microspheres P(AM-BA) and acrylamide (AM). The double peak, within the range of 3100 and 3500 cm À1 in the spectrum of AM, is attributed to the amide unit and is not available in the spectra of P(AM-BA) and P(AM-BA-RhB). The N-H stretching vibration peak appeared at 3400 cm À1 , indicating the presence of acrylamide units within the polymer microspheres composed of P(AM-BA) and P(AM-BA-RhB). Furthermore, the C]O bending vibration absorption peak shied from 1600 cm À1 for AM to 1740 cm À1 for P(AM-BA), suggesting that the amide is partially hydrolyzed to the carboxyl. 34 The peak of 670 cm À1 is attributed to the bending vibration of aromatic hydrogen. As the trace amount of Allyl Rhodamine B is on the side chains, other characteristic peaks of Allyl Rhodamine B are weak, such as the benzene ring skeleton, methyl and methylene.  The morphology of the polymer gel was researched by a uorescence microscope ( Fig. 5A and B), a scanning electron microscope ( Fig. 5C and D) and an environment scanning electron microscope (Fig. 6). Fig. 6 gives the particle size analysis and three-dimensional cross-linked networks of uorescent polymer microspheres P(AM-BA-RhB) before and aer swelling at 25 C. As showed in Fig. 5, the uorescent polymer microspheres P(AM-BA-RhB) are spherical particles and can swell many times due to absorbing a lot of water (see Fig. 5D and C). As the uorescent polymer microspheres P(AM-BA-RhB) contact with brine water, the water molecules can go into the internal network under the action of inside and outside osmotic pressure, stretching the molecular chain of the network structure, enlarging the volume of polymer microspheres. The threedimensional cross-linked networks can be obviously observed in Fig. 5C and 6. It can be seen that the average particle sizes of uorescent polymer microspheres P(AM-BA-RhB) before and aer swelling are 125.7 mm and 215.6 mm, respectively (see Fig. 6). Moreover, from Fig. 5A and B we can see that the swollen polymer microspheres P(AM-BA-RhB) have uorescent property. The polymer microspheres emit red uorescence under UV light and the polymer microspheres show some colourless and transparent balls under the ordinary light.
3.2 Fluorescent property of uorescent polymer microspheres P(AM-BA-RhB) Fig. 7 shows the uorescent emission spectra of Rhodamine B and uorescent polymer microspheres P(AM-BA-RhB) in brine water. It can be observed that the max emission peak of Rhodamine B and P(AM-BA-RhB) were 575 nm and 580 nm, respectively. We nd that the emission spectra of uorescent monomer Rhodamine B moved towards to the long-wavelength band when the Rhodamine B was graed to the structure of polymer microspheres P(AM-BA-RhB). This means red shi has taken place. The reason why red shi occurred was that when uorescent monomer Rhodamine B was graed to polymer microspheres, the conjugation effect of uorescent monomer Rhodamine B increased, the energy of electron transition decreased and the uorescence emission wavelength became longer. Martínez et al. also found that the uorescence spectrum will be changed when the uorescent dye was embedded in the polymer gel by graing or coating. 35     8a shows the effect of pH on the uorescence intensity of uorescent polymer microspheres P(AM-BA-RhB) in brine water at the wavelength l ex ¼ 505 nm. When pH < 3, the uorescence intensity of uorescent polymer microspheres P(AM-BA-RhB) solution varies with the change of pH. However, the uorescence intensity remains unchanged when the pH is in the range of 3-10. When pH equals to 1.0, the dispersed solution was strong acid solution and the carboxyl of RhB can't ionize. Because the carboxyl group, which was a strong electronwithdrawing group, may quench the uorescence of RhB. This caused weak uorescent property of RhB. The carboxyl of RhB began to ionize and the degree of ionization increased gradually with the increase of pH value. Thus, the electronic absorption ability became weak and the uorescence enhanced gradually.
When the pH value is close to 3.0, the carboxyl of RhB has been completely ionization, the quenching effect disappeared and uorescence intensity reached the maximum value (see Fig. 8a). When the pH was greater than 3.0, the uorescence intensity changed little with the increase of pH value. RhB uorescence intensity presented undulating uctuation due to the high sensitivity of uorescence spectrometer. It can conclude that the uorescent polymer microspheres P(AM-BA-RhB) can be used in the brine water environment with pH greater than 3.0.
Following the general procedure, the effect of ionic species on the uorescence intensity of uorescent polymer microspheres P(AM-BA-RhB) was studied (see Fig. 8b). The response of the uorescent polymer microspheres P(AM-BA-RhB) to solutions containing ionic species such as Na + , K + , Mg 2+ , Ca 2+ , Paper and Fe 3+ was investigated. The results indicated that the ionic species had little inuence on the uorescence intensity. When the uorescent monomer RhB was graed to the network structure of polymer microspheres, the uorescence properties were reduced by the inuence of the reservoir condition. So the uorescent polymer microspheres P(AM-BA-RhB) were suitable for the brine water containing these ionic species.
3.3 Swelling property of uorescent polymer microspheres P(AM-BA-RhB) Fig. 9a shows the swelling ratio of uorescent polymer microspheres P(AM-BA-RhB) in the brine water with different concentrations of NaCl (1, 5, 8, 10, 20 g L À1 ) at 90 C. Within 40 hours of hydration, the swelling ratio increased rapidly at an early stage and gradually leveled off. The swelling ratio reached to the maximum aer 175 hours. This indicates that brine water can penetrate into the matrix of the uorescent polymer microspheres. In fact, the uorescent polymer microspheres are cross-linked particles with a three-dimensional network structure and many free hydrophilic groups (-CONH 2 ) inside. Polar brine water molecules can be bond to these groups easily through hydrogen bonding, leading to a signicant increase in the hydrated particle size of uorescent polymer microspheres. While the concentrations of NaCl increased from 1 to 20 g L À1 , the swelling ratio decreased from 12 to 2 at the equilibrium stage. These results illustrate that the uorescent polymer microspheres P(AM-BA-RhB) maintain good salt tolerance at high temperature, 90 C in this case. Fig. 9b shows the effect of temperature on the swelling ratio of two different polymer microspheres P(AM-BA-RhB) and P(AM-BA) in the brine water with various temperature form 25 to 110 C. When the temperature increased from 25 to 110 C, the swelling ratio increased from 8 to Fig. 9 Effects of salinity and temperature on swelling ratio of polymer microspheres.
14 times at the equilibrium stage. The two polymer microspheres displayed excellent thermal stability because of the three-dimensional network structures in the polymer microspheres.
Furthermore, there is no obvious difference between the swelling ratios of the two kinds of polymer microspheres (see Fig. 9b). This result suggests that swelling property has not been dramatically affected by the uorescent functional groups embedded in the polymer microspheres by graing or coating. Essentially, this result could be attributed to the low content of the uorescent functional groups.

Method of measuring the concentration of uorescent polymer microspheres
Aer studying the swelling property and uorescence property of uorescent polymer microspheres, the detection method of measuring the concentration of uorescent polymer microspheres was discussed in this part. This could provide methods for the research of plugging behavior study of uorescent polymer microspheres.
3.4.1 Standard curve of measuring the concentration of uorescent polymer microspheres and detection limit. Different concentrations of uorescent polymer microspheres P(AM-BA-RhB) were moved to quartz sampling cells, the uorescence emission intensity curve was measured by the scanning slit width of 5 nm. Experimental result of uorescent polymer microspheres P(AM-BA-RhB) was shown in Fig. 10, the concentration of microspheres had a linear relationship with the uorescence intensity at the concentration range of 10-500 mg L À1 . The linear equation was y ¼ 0.981 + 0.171x, and the correlation coefficient r was 0.9997. Carrying out 12 blank determinations and getting the ratio of three times of uorescence value standard deviation and the slope of the standard curve. The ratio was the detection limit. The detection limit of uorescent polymer microspheres P(AM-BA-RhB) was 0.5 mg L À1 . According to this method, the concentration of uorescent polymer microspheres P(AM-BA-RhB) in the production liquid can be detected quantitatively through the uorescence intensity value.
3.4.2 Sample determination and accuracy. In order to verify the accuracy of the standard curve, some known concentration of uorescent polymer microspheres P(AM-BA-RhB) was prepared and then the content of the microspheres was calculated by the standard curve. Compared the actual concentration value and test concentration value of uorescent microspheres and obtained the relative error. The results were presented in Table 1. As seen from the Table 1, this method is especially accurate in measuring concentration of polymer microspheres with the relative error less than 2% to actual values. This relative error can satisfy the requirements of the measurement.
3.5 Plugging behavior of uorescent polymer microspheres 3.5.1 Micro migration and migration mechanism. The visual micromodel was used to simulate channel distribution under reservoir conditions. Fig. 11a illustrates the migration behavior of uorescent polymer microspheres in the porous medium of visual micromodel. As is shown in Fig. 11a, some uorescent polymer microspheres were accumulated and squeezed in the pore-throat of the model, playing a role of plugging, so that other uorescent polymer microspheres cannot move forward, which results in an immobile formation prole being formed under the current pressure. As long as this immobile formation prole is strong enough, following uorescent polymer microspheres will be blocked by it and then more immobile formation prole will be formed. Plugging effect of the polymer gel to pore-throat is a joint action of those immobile formation proles. Denitely, the immobile formation proles are not absolutely immovable. They are dynamically changing. The movement of a small number of uorescent polymer microspheres will destroy the original layer sand and form new immobile formation prole, i.e., uorescent polymer microspheres can deform, breakthrough and transport to the next pore-throat so long as sufficient supply of pressure. Fig. 11b illustrated the migration schematic diagram of uorescent polymer microspheres. When uorescent polymer microspheres were being injected, the particles preferentially entered the high permeability zone. Fluorescent polymer microspheres can pass directly or by deformation through the porous media and enter the in-depth formation. When the size of the uorescent polymer microsphere is much larger than the pore throat size, the uorescent polymer microspheres will be  trapped at the entrance of the pore throat and directly plug the high permeability channel. When the size of uorescent polymer microsphere is a little larger than the pore throat size but not very much, the polymer gel can be deformed and pass through the pore throat. Moreover, the deformable polymer microspheres can revert to their original shape aer entering a larger pore. Additionally, when the size of uorescent polymer microsphere is smaller than the pore throat size, two or more uorescent polymer microspheres will be stranded in the pore space and bridge onto the pore throat. Once the bridge is formed and consolidated, the newly arriving uorescent polymer microspheres accumulate upstream from bridged pores (see Fig. 11b), thus decreasing the following uid ow rate and yielding uid diversion effects. When the size of uorescent polymer microsphere is too smaller than the pore throat size, uorescent polymer microspheres will easily pass through the pore throat and enter the in-depth formation. In conclusion, the uorescent polymer microspheres can be deformed and pass through pores easily, so they can enter the in-depth formation and plug the high permeability channels. 3.5.2 Plugging property and plugging mechanism. Three kinds of sand-pack tubes with different permeability were used to study the plugging property of uorescent microspheres in the porous media. In this part, the concentration of the polymer microspheres was monitored according to the uorescence intensity of production uid. According the relationship among permeability, porosity and pore size, 15 for the sand-pack with permeability of 12, 27, 36 mm 2 and porosity of 30.8%, 31.5%, 31.9%, the average pore size were 88.2, 130.9, 150.2 mm. The results of the plugging experiments are shown in Fig. 12.
Sand-pack tube 1. Fig. 12a shows the production concentration curve and injection pressure curve of uorescent polymer microspheres in sand-pack tube 1 with permeability of 12 mm 2 . It can be seen from Fig. 12a that the subsequent brine water injection pressure is constantly rising and uctuating aer 2 PV injections of uorescent polymer microspheres, and nally the injection pressure reaches a peak of 1.9 MPa at 7.3 PV. Then the injection pressure drops slowly in the uctuation, and at the same time the occurrence of uorescent microspheres was monitored at the outlet of the tube. With the continuous injection of brine water, the injection pressure begins to decrease slowly, and the concentration of uorescent polymer microspheres at the outlet increases rst and then decreases. Through the observation of the produced concentration curve of the uorescent polymer microspheres, it can be found that when the injection volume gets to 8-14 PV, the microspheres concentration rises from 0 to 70 mg L À1 ; from 14 PV to 17 PV, the microspheres concentration consists a platform with concentration of 70 mg L À1 ; aer 18 PV injection, the uorescent polymer microspheres concentration begins to decline.
The injection pressure keeps rising in a general trend and drops occasionally within a large range at the brine water injection stage, which is dened as "Wave-type Variation". 18 In fact, when the swollen uorescent polymer microspheres were injected into the tube, they were adsorbed and accumulated on sand surfaces at the beginning. The pressure increased due to blocking pores near the inlet. Moreover, with higher injection volume of uorescent polymer microspheres, uorescent polymer microspheres accumulated faster, leading to more rapid pressure increase. Since the swollen uorescent polymer microspheres have a certain degree of deformation, they could deform under pressure and move to the next pore. This is indicated by the uctuation when the pressure increased during swollen uorescent polymer microspheres injection and the pressure continued to rise with large uctuations. 36 Sand-pack tube 2. Fig. 12b shows the production concentration curve and injection pressure curve of uorescent polymer microspheres in sand-pack tube 2 with permeability of 27 mm 2 . It can be seen from Fig. 12b that aer 2 PV injections of uorescent polymer microspheres, the subsequent water injection pressure is constantly rising and uctuating as well, whereas the peak value of injection pressure is 1.5 MPa. Then the injection pressure drops slowly again in the uctuation, and before the injection pressure getting the peak value, the occurrence of uorescent polymer microspheres was monitored at the outlet of the tube. Through observation of the produced concentration curve of the microspheres, we can nd that the concentration curve of uorescent polymer microspheres shows a sharp peak. It is different from the production curve in Fig. 12a.
Sand-pack tube 3. Fig. 12c shows the production concentration curve and injection pressure curve of uorescent polymer microspheres in sand-pack tube 3 with permeability of 36 mm 2 . It can be seen from Fig. 12c that the subsequent water injection pressure is constantly rising with uctuation as well aer 2 PV injections of uorescent polymer microspheres, whereas the peak value of injection pressure is 1.25 MPa. Then the injection pressure drops slowly again in the uctuation, and before the injection pressure getting the peak value, the occurrence of uorescent polymer microspheres was also monitored at the outlet of the tube. Through observation of the produced concentration curve of the microspheres, it is found that the concentration curve of uorescent polymer microspheres shows three sharp peaks. It is different from the production curves in Fig. 12a and b.
It is noteworthy that the uorescent polymer microspheres production concentration reaches a platform from 14 PV to 17 PV in the sand-pack tube 1, which indicates that the uorescent polymer microspheres are produced at a certain highconcentration for a period of time. And it is assumed that the microspheres mainly move parallel to the piston in the sandpack tube 1 (see Fig. 12d). This is named as "piston plugging". In the sand-pack tube 2, the concentration curve of uorescent polymer microspheres shows one peak at 6 PV, indicating that the uorescent polymer microspheres move along a large-pore channel, and this is called "protruding plugging" (see Fig. 12d). However, in the sand-pack tube 3, there are three peaks in the production concentration of uorescent polymer microspheres, indicating that the uorescent polymer microspheres are transported through three large channels, which is called "ngering plugging" (see Fig. 12d). Due to the difference of fracture distribution in the reservoirs, uorescent polymer microspheres have different plugging behavior. The slug design of polymer microspheres should fully consider the plugging behavior in the reservoirs.

Conclusions
In this study, a kind of uorescent polymer microspheres P(AM-BA-RhB) acted as a novel conformance control agent was proposed. Laboratory experiments have been conducted to understand the structure characterization, uorescent property, swelling property and plugging behavior of uorescent polymer microspheres. The major conclusions that can be drawn from this study are as follows: (1) The uorescent polymer microspheres P(AM-BA-RhB), which can uoresce under ultraviolet irradiation, was synthesized via the inverse suspension polymerization method. The uorescence functional monomer of Rhodamine B was uniformly distributed inside the polymer microspheres. When the Rhodamine B was graed on the structure of polymer microspheres P(AM-BA-RhB), the red shi occurred. The uorescent polymer microspheres retained their uorescent properties aer swelling in the metal ions containing solution at the pH 3.0-10.0.
(2) The uorescent polymer microspheres P(AM-BA-RhB) were spherical particles with an average particle size of 125.7 mm. This kind of microspheres was composed of threedimensional cross-linked networks and had good swelling property in high salinity and high temperature conditions. With the increase of salinity, the swelling ratio decreased. And the swelling ratio increased with the increase of the formation temperature. The swelling property was not dramatically affected by the uorescent functional groups embedded in the polymer microspheres by graing or coating.
(3) The linear equation between uorescence intensity and concentration of uorescent polymer microspheres P(AM-BA-RhB) was constructed. According to this linear equation, the concentration of uorescent polymer microspheres in the Paper production liquid can be detected quantitatively through the uorescence intensity value.
(4) Visual micromodel test and sand-pack tubes experiment of uorescent polymer microspheres showed that the uorescent polymer microspheres can pass directly or by deformation through porous media and enter the in-depth formation. The injection pressure showed the phenomenon of "Wave-type Variation". Three plugging behaviors such as piston plugging, protruding plugging and ngering plugging were put forward. These results were useful for the slug design of polymer microspheres in the eld application.

Conflicts of interest
The authors declare no competing nancial interest.