Performance Evaluation and Site Application of a Hydrophobic Long-Chain Ester-Based CO2 Fracturing Fluid Thickener

Nowadays, there are a wide variety of thickeners developed for dry CO2 fracturing worldwide, but numerous problems remain during in situ testing. To address problems in CO2 fracturing fluid operation (high frictional drag, low viscosity, low proppant-carrying capacity, narrow reservoir fractures, etc.), the authors have synthesized the novel hydrophobic long-chain ester thickener, studied viscosity, frictional drag, and proppant-carrying capacity of CO2 fracturing fluid and core damage by CO2 fracturing fluid by varying the temperature, pressure, and level of injection of the novel thickener and explored the thickening mechanism for this thickener in CO2. Based on the study results, as the temperature, pressure, and amount of injected thickener increased, fracturing fluid viscosity increased steadily. In the case of shearing for 125min under conditions of 170 S, 40C, and 20MPa, when the thickener level increased from 1% to 2%, fracturing fluid viscosity increased and then decreased, varying within 50–150mPa·s, and the viscosity-enhancing effect was evident; under conditions of 20C and 12MPa, as the flow rate increased, drag reduction efficiency reached 78.3% and the minimal proppant settling speed was 0.09m/s; under conditions of 40C and 20MPa, drag reduction efficiency reached 77.4% and the proppant settling speed was 0.08m/s; with the increases in temperature, pressure, and injection amount, core damage rates of the thickener varied within 1.77%–2.88%, indicating that basically no damage occurred. This study is of significant importance to the development of CO2 viscosity enhancers and CO2 fracturing operation.


Introduction
Research on CO 2 dry fracturing technology has been carried out in the world in the past years. Dry CO 2 fracturing technology is more advantageous than conventional hydraulic fracturing technology in that dry CO 2 is water-free and residue-free and flows back more quickly; however, in fracturing practice, liquid CO 2 has such problems as high frictional drag and low proppant-carrying efficiency. To address this issue, we have to add a thickener into CO 2 fracturing fluid. Based on the systematic analysis of the current situation of CO 2 thickeners worldwide, CO 2 thickeners can be divided into fluorinated, siloxane, and hydrocarbon thickeners. Huang et al. [1] obtained polymer polyFAST by copolymerization between styrene and fluorinated acrylate, its CO 2 thickening effect was good, the solubility of polymer polyFAST in CO 2 decreased with increasing polymer molecular weight, and when n(styrene) : n(fluorinated styrene) is about 3 : 7, the prepared copolymer had the best CO 2 thickening effect. But fluorinated polymers do not apply to real production due to great environmental hazard and high cost; Shi et al. [2] prepared a thickener for dense CO 2 and found that, at thickener concentrations of 2%-4%, the viscosity of liquid CO 2 increased by a factor of 2-3. Cui et al. [3] developed a surfactant-based thickener for liquid CO 2 and simulated the liquid CO 2 thickening process by rheological testing. As the surfactant formed micelles in liquid CO 2 , the viscosity of liquid CO 2 increased by a factor of about 86-216 and the maximum viscosity was 21.6 mPa·s. Shen et al. [4] synthesized a polyvinyl acetate telomere through free radical polymerization catalyzed by azodiisobutyronitrile (AIBN) and then the telomere underwent polymerization reaction with styrene to generate a binary copolymer, followed by molecular structure characterization by FTIR and HNMR. The styrene-vinyl acetate copolymer itself has CO 2 -philic groups and styrene thickening groups, and thus, in theory, it can become a high-performance lowcost thickener that is friendly to the environment and causes little damage to the rock core. Bae and Irani [5] prepared a silicon-containing polymer from a siloxane and methane and studied its impact on liquid CO 2 viscosity. The study shows that the viscosity of a mixture of 4% silicon-containing polymer + 20%toluene + 76% CO 2 was 1.2 mPa·s; thus, the viscosity of liquid CO 2 increased by a factor of 30, but to thicken liquid CO 2 with the polymer, a great amount of toluene has to be added; thus, excessive cosolvent consumption leads to poor economics, making the wide application impossible. In summary, even though varieties of CO 2 thickeners have been developed to date, slow dissolution, need of a cosolvent, insignificant viscosityenhancing effect, environmental pollution of fluorinated thickeners, and other major issues are still present. In contrast, the hydrophobic long-chain ester thickener synthesized in this study, upon dissolution in liquid CO 2 , exhibits essential properties of a dry fracturing fluid, eliminating the need of any cosolvent; at room temperature, the novel thickener at a concentration of 1.0%-3.0%, if fully stirred, can be directly dispersed and dissolved in 3 min and viscosity enhancement is evident, with the viscosity varying within 50-150 mPa·s; after dissolution of the thickener, the resulting viscoelastic fluid exhibits good dynamic proppant-carrying capacity, with a proppant concentration of up to 30%, its shear thinning behavior is good, with frictional drag reduction efficiency of over 70%, and its core damage range is 1.77%-2.88%, indicating that it is basically pollution free.

Proppant Sample.
A novel low-density ceramsite was selected as the proppant, its particle size is ф0.3-0.6 mm or 20-40 mesh, and its bulk density is 1.33 g/cm 3 .

Experimental Cores.
Each of the natural long cores to be used in the experiment was jointed by a few short cores, a single fracture was artificially crafted on each core, and basic physical parameters of these cores are shown in Table 1.

Experimental Equipment.
A HAAKE RS6000 rheometer, frictional drag test apparatus for CO 2 fracturing fluid, visual high-temperature and high-pressure PVT test system, leakoff and damage test apparatus for CO 2 fracturing fluid, visual dissolubility and dispersibility test apparatus for CO 2 , and pco.dimax cs1 high-power microscope were used.

Synthesis of the Hydrophobic Long-Chain Ester
Thickener. The hydrophobic long-chain ester thickener was prepared from oil-soluble solvents (10%-15%) and CO 2 -philic solvent (45%-65%) as reaction solvents, highly CO 2 -philic monomer (1%-3%), structural monomers (3%-5%), viscosity-enhancing monomers (15%-20%), dissolutionassisting monomers (5%-10%) as copolymerization monomers, oil-soluble chain transfer agents (0.5%-1%), and the oil-soluble initiator (0.5%-1%), wherein all the contents are on a mass basis. After completion of the reaction, the resulting copolymer was directly dissolved in the CO 2 -philic solvents and the oil-soluble solvent without separation, removal, and other posttreatment processes, making its preparation convenient. The preparation procedure is detailed as follows: (1) adding solvents: while the reactor temperature was kept at 35°C, 18 g of petroleum ether, 18 g of white mineral oil, and 194.1 g of propylene carbonate were added successively into the reactor; (2) adding copolymerization monomers: while stirring was kept in the reactor, 3 g of allylmethyl carbonate, 6 g of styrene, 6 g of strearyl acrylate, 12 g of methyl methacrylate, 24 g of methyl acrylate, 9 g of n-propyl vinyl ether, and 6 g of ethyl propylene ether were added in turn such that the monomers were completely dissolved; and (3) adding the initiator and the chain transfer agents: 1.5 g of azobisisobutyronitrile was added to initiate polymerization reaction for 2 h, then, 1.2 g of dodecyl mercaptan and 1.2 g of butyl mercaptan were added, the reactor was closed and heated to 100°C, followed by 8 h isothermal reaction, and the internal pressure of the reactor was controlled throughout the process to be up to 0.4 MPa. After reaction completion, cooling water was circulated in the reactor jacket to lower the temperature to 30°C, resulting in white emulsion in the reactor; the emulsion was pumped into a plastic bucket to obtain the hydrophobic long-chain ester thickener, which is the novel thickener for site application experiment mentioned in Section 5.2.

Viscosity Test.
A rheology test method for CO 2 fracturing fluid was established using a HAAKE RS6000 rheometer: viscosities of liquid/supercritical CO 2 fracturing fluids at variable concentrations were determined after shearing at 170 S −1 for 125 min, the viscosity-enhancing effect of the viscosity enhancer was evaluated, and rheological behavior and viscosity variability patterns of CO 2 fracturing fluids under different influential factors were analyzed. Variations in rheological properties of the fluid under different conditions could be detected using a high-temperature closed 2 Geofluids system, and the fluid temperature/shear resistance test and pressure versus viscosity relationship test could be performed [6,7]. Specifically, a liquid CO 2 tank is fully filled with pressurized liquid (supercritical) CO 2 for layer use. A thicker sample at a preset test concentration was put into a mixing tank which was then closed. Thereafter, liquid (supercritical) CO 2 was introduced into the mixing tank till a preset volume. The stirrer was turned on such that the thickener became fully dissolved. The well-dissolved dry fracturing fluid was introduced into the measuring cylinder of the rheometer and sheared at preset temperature, pressure, and shear rate until being stable before data saving.

Frictional Drag Test.
On a CO 2 fracturing fluid frictional drag test apparatus, differential pressures of liquid/supercritical CO 2 fracturing fluids in the pipeline subject to thickener addition were determined by high-pressure long-tube flow differential pressure test, where the highpressure long tube was 10 m long and 10 mm in diameter. And drag reduction efficiency was calculated using Dr = ðΔP 1 −ΔP 2 Þ/ΔP 1 × 100% (Dr is the drag reduction efficiency, ΔP 1 is the differential pressure of neat liquid CO 2 when passing through the long tube, and ΔP 2 is the differential pressure of CO 2 fracturing fluid fed with the novel thickener when passing through the long tube). In this experiment, at 20°C/12 MPa and 40°C/20 MPa, drag reduction efficiencies of the fluid fed with the thickener at different concentrations at variable flow rates were determined [8].

Proppant-Carrying Capacity Test.
Assisted by a hightemperature high-pressure visual PVT experimental system, this experiment studies the settling patterns of the proppant in liquid and supercritical CO 2 fracturing fluids, determines settling speeds of the proppant in liquid and supercritical CO 2 fracturing fluids versus thickener concentration and proppant concentration as well as their variability patterns, and analyzes the settling speed of the proppant in each CO 2 fracturing fluid. On a proppant settling test apparatus, the experiment was carried out according to a procedure detailed as follows: (1) Fill the thickener at a volume ratio into an autoclave and put the ceramsite on top of the visual autoclave (2) Heat the autoclave by a water bath surrounding it such that a desired test temperature is achieved (3) Open the CO 2 cylinder and fill CO 2 into the autoclave via an air compressor such that withinreactor pressure reaches a desired pressure (4) Set the stirrer speed to 1000 rpm such that the thickener can be fully dissolved in CO 2 (5) Rotate the manual ball releaser on top of the autoclave such that the ceramsite falls inside the autoclave and record the ceramsite settling time (6) Calculate ceramsite settling speed 2.3.5. Core Damage Test. The long core leak-off damage evaluation test was performed on a CO 2 fracturing fluid leak-off damage test apparatus, and leak-off quantity Q, leak-off coefficient C, and leak-off velocity V were used to explore the leak-off behavior of liquid/supercritical CO 2 fracturing fluids; by substituting permeability values prior to and after damage of the experimental core into core damage rate calculation formula D d = ðK 1 − K 2 Þ/K 1 × 100% (predamage and postdamage permeability values are K 1 and K 2 mD, respectively), the core damage rate can be obtained [9]; hence, the degree of core damage of the CO 2 fracturing fluid was evaluated by the fracturing fluid performance evaluation method according to SY/T 5107-1995 National Standards for Oil and Gas Industry in the People's Republic of China (see the flowchart of the apparatus operation in Figure 1).

Explorative
Test on the Thickening Mechanism. The visual carbon dioxide dissolution dispersion test apparatus and pco.dimax cs1 high-power microscope were employed to observe the dissolution and dispersion state of the viscosity enhancer in liquid/supercritical CO 2 fluids, and the CO 2 fracturing fluid thickening mechanism was discussed according to the solvent-solute molecular dissolution theory [10].

Results and Discussion
3.1. Results of the Viscosity Test. Thickener concentration versus CO 2 viscosity curves and fracturing fluid viscosity versus shear time curves at variable concentrations of the thickener in the fluid were obtained by the HAAKE RS6000 rheometer. As shown in Figure 2, CO 2 fluid at 20°C/8 MPa was in a liquid state, while at 40°C/15 MPa and 45°C/7.5 MPa, CO 2 fluid was in a supercritical state, as the injection concentration of the thickener increased, the viscosities of both liquid CO 2 and supercritical CO 2 fracturing fluids tended to increase; in particular, when the thickener injection concentration was greater than 1.2%, the viscosity of liquid CO 2 fracturing fluid was always greater than that of supercritical CO 2 fracturing fluid. As shown in Figures 3-5, at 3 Geofluids 40°C/20 MPa, when thickener concentration was 1.0%, the viscosity of CO 2 fracturing fluid tended to increase and then decrease over time and remained at 30 mPa·s or so at 30 min till the end of experiment; likewise, when thickener concentration increased to 1.5%, the viscosity remained at 60 mPa·s or so over time; when thickener concentration increased to 2%, the maximum fracturing fluid viscosity was 140 mPa·s, and at the later stage of the experiment, the viscosity decreasing trend was shown but overall viscosity was greater than 100 mPa·s. After shearing at 170 s −1 for 125 min, this thickener remained highly stable and had good viscosity-enhancing effect, indicating strong temperature/shear resistance. Given a constant temperature, viscosity of CO 2 fracturing fluid increased with increasing pressure, indi-cating that pressure variation has indeed significant impact on viscosity of CO 2 fracturing fluid and pressure increase leads to more violent momentum exchange between thickener molecule and CO 2 molecule and higher density of CO 2 fluid such that thickener solute becomes more soluble in CO 2 , further increasing viscosity of the fracturing fluid system. Therefore, increasing the thickener concentration and test pressure has important impact on the increase in viscosity of liquid or supercritical CO 2 [11].

Results of Frictional Drag Test.
On a frictional drag test apparatus for CO 2 fracturing fluid, the high-pressure long-tube flow differential pressure test was conducted to determine differential pressures of CO 2 fracturing fluid in a pipeline subject to thickener addition and a formula for drag reduction efficiency calculation was used to obtain drag reduction efficiency; then, drag reduction efficiency   As can be seen in Figures 6 and 7, with the increases in the flow rate and concentration of injected thickener, drag reduction efficiency increased significantly. Given a constant temperature, frictional drag coefficient increased gradually with increasing pressure because higher pressure leads to smaller intermolecular distance and stronger friction force. When the temperature, pressure, and pipe diameter were constant, the faster the flow rate of the fluid in pipeline, the greater the degree of turbulence and the friction loss and drag reduction effect became more significant at higher Reynolds numbers. As temperature and pressure varied, drag reduction efficiency of the thickener ranged within 59.1%-78.3%. When thickener concentration was 2% and flow rate was above 2.0 m/s, drag reduction efficiency reached 78.3%.

3.3.
Results of the Proppant-Carrying Capacity Test. Settling times and settling speeds of single-particle proppant and the proppant at a proppant concentration of 5% in liquid/supercritical CO 2 fracturing fluids were determined by a proppant settling test apparatus, as shown in Tables 2 and 3.
As inferred in Tables 2 and 3, supercritical CO 2 is generally superior to liquid CO 2 in proppant-carrying capacity: in the case of a single-particle proppant, the minimum settling speed in liquid CO 2 was 0.09 m/s while the minimum settling speed in supercritical CO 2 was 0.08 m/s; in the case of 5% proppant concentration, the minimum settling speed in liquid CO 2 was 0.08 m/s while the minimum settling speed in supercritical CO 2 was 0.07 m/s, because the hydrophobic long-chain ester thickener dissolved in CO 2 is markedly viscoelastic and thus has excellent proppant-carrying capacity. Given a constant amount of the injected thickener, increases in temperature and pressure enabled higher fracturing fluid viscosity such that the proppant settling speed became slower. In comparison with single-particle settling, the proppant settling speed at 5% proppant concentration was lower than the free single-particle settling speed due to intermolecular interference and the single-particle settling will cause upward flow of the surrounding liquid which impedes the surrounding fluid from sinking; thus a greater drag will be present at a proppant concentration of 5% and the increase in proppant concentration is, in fact, equivalent to increases in proppant buoyancy and settling drag, manifested as a slower settling speed [12].

Core Damage Evaluation.
The long core leak-off and damage test was carried out on a CO 2 fracturing fluid leakoff and damage test apparatus at a temperature of 55°C and a pressure of 14 MPa. The core leak-off performance         Table 4, the parameters of damage evaluation of fracturing fluid are shown in Table 5, and the core damage performance of the CO 2 fracturing fluid was shown in Table 6. As shown in Table 4, under the condition of 45°C/10 MPa, the leak-off velocity of supercritical CO 2 fluid in the long core was 0.037 m·min −1 , while the leakoff velocity of supercritical CO 2 fracturing fluid containing a 1.5% viscosity enhancer injected into the long core was 0.012 m·min −1 , and in comparison, the leak-off coefficient of CO 2 fracturing fluid into which the viscosity enhancer was injected was markedly lower than that of pure CO 2 fluid, indicating that this viscosity enhancer reduces the leak-off property of supercritical CO 2 fluid to some extent. As CO 2 fluid is residue free, no filter cake will be formed at the front end of the core in a long core leak-off test, indicating a greater leak-off coefficient and a faster leakoff velocity; after a thickener was incorporated, the CO 2 fracturing fluid will form gradually flocculent filter cakes at the front end of the rock core, resulting in lower leakoff coefficient, indicating that CO 2 fluid incorporated with a thickener has pronounced effect of leak-off reduction after entering fractures and thus ensures fracture forming efficiency of a fracturing fluid. And in Table 6, where core damage rate D d was calculated by substituting the leak-off coefficient into the calculation formula for D d , it can be found that the CO 2 thickener had low core damage rates ranging within 1.81%-2.88%, and in comparison with the core damage criteria for fracturing fluids in Table 5, its damage to the rock core is almost negligible; moreover, CO 2 is slightly soluble in water and readily soluble in the in-place oil; therefore, the thickened CO 2 fracturing fluid system would be well compatible with the formation.

Discussion on the Thickening Mechanism of the Hydrophobic Long-Chain Ester Thickener in CO 2 Fracturing Fluids
The dissolution and dispersion states of the viscosity enhancer in liquid/supercritical CO 2 fluids were examined by visual dissolution and the dispersion test apparatus and the pco.dimax cs1 high-power microscope. The results of the sample dispersion test are shown in Table 7, and the dissolution status of the thickener in CO 2 is seen in Figure 8. Intermolecular interactions between viscosity enhancer solute and CO 2 solvent include (1) solvent-solute intermolecular interaction: solute-to-solvent aggregates will form due to strong interaction between solute and solvent molecules and (2) solvent-solvent intermolecular interaction: in highly compressible dilute supercritical fluid, apart from solvent-solute aggregates, there are solvent-solvent aggregates. Dissolution and dispersion of the hydrophobic long-chain ester thickener in CO 2 were examined using a pco.dimax cs1 high-frequency microscopic camera, results of the dispersion test of the same amount of thickener injected at different temperatures and pressures are shown in Table 7, and dissolution details are seen in Figure 8. As shown in Figure 8, the hydrophobic long-chain ester thickener dissolves instantly in CO 2 ; at a thickener concentration of 2%, it can be directly dispersed and dissolved in CO 2 in less than 2 min, so it is readily dispersible and soluble in liquid CO 2 and supercritical CO 2 . In terms of synthetic composition, propylene carbonate is a CO 2 -philic solvent, white mineral oil and petroleum ether are oil-soluble solvents, allylmethyl carbonate is a strong CO 2 -philic monomer, and ethyl propenyl ether and n-propyl vinyl ether are cosolvent monomers, while methyl acrylate and methyl methacrylate are viscosity-enhancing monomers. When the thickener is dissolved in CO 2 , it can be quickly dispersed and dissolved due to effects of the CO 2 -philic solvent and oil-soluble solvents, as the copolymer components have been directly dissolved in the CO 2 -philic and oil-soluble solvents first, making the polymeric molecular chains prestretched, and during dissolution, the viscosity-enhancing groups on the molecular chains dissolve to enhance viscosity, meanwhile multiple association effects between them and intramolecular and intermolecular structural groups via the hydrogen bond and dispersion force as well as Lewis acidbase reaction contribute to structural viscosity; thereby, the viscosity-enhancing effect is greatly increased [13][14][15][16][17][18].

Basic Geology of Experimental Well.
Well Hong 87-X is located at block M in oilfield J. This block is in the southern part of the Rangzijing structure in the central depression area of the southern Songliao Basin, neighboring the Gudian reverse fault of the Fuxin uplift zone in the east and SWdipping slope zone of the Xinli structure in the north, and the slope zone is subjected to regional compressive torsional stress and forms a NNW-trending fold zone associated with a number of fault zones parallel axially to the fold zone. The oil reservoir belongs to the fault rock-type oil reservoir in the setting of the westward-dipping slope. The bed of interest is 3300 m deep, the reservoir temperature is 104°C, the gradient of formation pressure is 0.  7 Geofluids reservoir lithology is dominated by greyish-brown siltstone and fine sandstone, mainly including lithic arkose or feldspathic litharenite. Particle sizes of the rocks are generally 0.03-0.25 mm, and the clay mineral content is 14.2%.

Site Operation Process. The fracturing operation of Well
Hong 87-X in block M of oilfield J with old thickener Znj01 did not get desirable effect. Later, the CO 2 dry fracturing operation with the novel thickener was conducted, where 835 m 3 of liquid CO 2 was consumed, ceramsite of 20-40 mm in the size was chosen as the proppant, pumping rates were 5.2-8.2 m3/min, proppant loading was 9 m 3 , proppant concentration was 4%, operation pressures were 21-37 MPa, and pump-shutoff pressure was 20.8 MPa. Prior to operation, the pipeline was circulated with CO 2 to test the pressure, ensuring that all the CO 2 -related lines are free of liquid buildup; the ratio of liquid CO 2 to the thickener was kept at 100 : 1; the prepad fluid volume and proppant transport program were adjusted based on the results of the prefracturing test; the site should support continuous proppant loading and continuous liquid supply. Figure 9 shows the dry fracturing operation curve of Well Hong 87-X.   8 Geofluids reservoir, stable pressure and superior proppant-carrying capacity were noted; in the case of the old thickener Znj01, the pumping rate was 5 m 3 /min and the proppant concentration was 2.5% at the stage of proppant loading, and after entering the formation, the pressure surged and eventually led to sand plug. Given the equivalent pumping rate, compared with the old thickener, the novel thickener had more stable proppant-carrying capacity during proppant loading. As a result of the site application, when the old thickener was used, the daily oil output of this well was up to 3.86 t; while after the novel thickener was used, the maximum daily oil output reached 7.4 t, with an increment of 3.54 t.

Conclusions
(1) The novel hydrophobic long-chain ester thickener synthesized in this study is a linear block copolymer supplied in the form of white emulsion with a molecular weight ranging within about 500,000-800,000. Based on performance testing, its density was 1.05 g/cm 3 and its pH was 7.5; the thickener at 2.0 wt% could be dissolved in liquid CO 2 in 3 min to form a fracturing fluid at a viscosity of 80 mPa·s, and its drag reduction efficiency was 78.3%  (2) The viscosity-enhancing effect, frictional drag, proppant-carrying capacity, and core damage of the novel thickener in CO 2 fracturing fluid were experimentally determined, and the thickening mechanism of this novel thickener in CO 2 was discussed. As a result, with the increase in the amount of the injected thickener, the viscosity of CO 2 fracturing fluid increased gradually, and when the concentration of the injected thickener was greater than 2%, the viscosity-enhancing effect in liquid CO 2 fracturing fluid was superior to that in supercritical CO 2 fracturing fluid; the drag reduction effect was notable when the thickener concentration was 2.0% and the flow rate of the fracturing fluid was more than 2.0 m/s. The thickener, if dissolved in CO 2 , could form multiorder structures and become significantly viscoelastic, resulting in higher viscosity; thus, it has excellent proppant-carrying capacity. After the novel thickener was added into CO 2 fracturing fluid, the maximum core damage rate was 2.88%, which is negligible (3) After CO 2 dry fracturing of Well Hong 87-X in oilfield J using the novel thickener, the fracturing effect was remarkable and the oil output of the well was substantially elevated; given the equivalent pumping rate, compared with the old thickener, the novel thickener had more stable proppant-carrying capacity during proppant loading, indicating that the novel thickener has better effect of viscosity enhancing and proppant carrying; moreover, the CO 2 dry fracturing technology is of significant importance to CO 2 sequestration and the reduction of greenhouse gas emission; it increases the oil output and protects the environment as well, resulting in a win-win situation in line with the China national strategy of "carbon peak and carbon neutralization"

Data Availability
The raw data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.

Conflicts of Interest
The authors declare no competing financial interest.