Introduction
Middle East carbonate oil and gas projects are the mainstay of the overseas business of CNPC, with nearly 80% of single wells requiring reservoir stimulation to obtain or accurately evaluate production capacity [1]. Acid fracturing treatment is safe, convenient and inexpensive, and a reservoir stimulation method that meets overseas development model featured by “high-speed production and rapid payback of investment” [2-3]. Acid fracturing treat-ment means to inject acid after fracturing in carbonate rocks. Different minerals react with acid at different rates, and acid can dissolve some minerals on the surface of induced fractures, so the fractures become effective channels whose conductivity depends on their support strength and surface shape [4].
Carbonate reservoirs in the Middle East are dominated by porous bioclastic limestones, which are usually large in scale and less than 3000 m in burial depth [5]. These reservoirs have three important characteristics [6]: (1) Natural fractures are not developed or less developed. With strong microscopic heterogeneity and wide range of throat size distribution, the reservoirs are susceptible to solid phase invasion and water lock damage; (2) The purity of calcite minerals is extremely high (generally higher than 95%), and it is extremely difficult to form rough type etchings with the conventional acid fracturing; (3) The rock strength is low, and most of the reservoirs are characterized by plasticity, and the conductivity will decay quickly after acid fracturing stimulation. The conventional acid fracturing treatment based on high viscous acid cannot create fractures with effective etched morphology. In addition, high viscous acid produces a large amount of residue, and the presence of spent acid is easy to induce water lock. Both of them may block oil and gas flow channels [7]. Therefore, this method isn’t applicable for porous bioclastic limestone. Recent studies have shown that multi-stage fracturing with alternating injection of low-viscosity and retarded acid can significantly improve the stimulation effect of these reservoirs [1,8]. However, because of commercial confidentiality, the characteristics of acid retardation, acid etching and flowback have not been reported in open literatures. Shen et al. [9] prepared low-viscosity surfactant micellar acid (surfactant acid) by adding cationic surfactants to hydrochloric acid, and a retardation rate of 71% was achieved by adsorption test on rock surface. The idea of adsorption retardation has been reported several times, suggesting that recovery enhancing agents such as microemulsions and nanoparticles can adsorb onto rock surface and slow down acid-rock reaction [10⇓-12]. Diluted microemulsion is an oil-in-water nanofluid prepared by diluting bicontinuous microemulsion with water [13]. It is consistent with the viscosity of the diluted phase, and there is no residue damage. It has the effect of imbibition and displacement of crude oil when added into fracturing fluid [14]. Diluted microemulsion has been widely used in hydraulic fracturing stimulation [15], but its suitability for acid fracturing has not been tested due to differences in pH and salinity of the diluted phase.
The multi-stage microemulsion acid fracturing technology has the advantages of convenient treatment, high conductivity, clean and easy flowback. This study developed a microemulsion acid which can be prepared onsite and pumped at high speed for carbonate reservoirs in the Middle East. In addition, its performance was compared with conventional acid systems by experiments, and the micro- etching characteristics and conductivity were clarified. Finally, the recovery degree of the relative permeability of the oil phase after spent acid invasion was evaluated.
1. Reaction kinetics of microemulsion acid and rock
1.1. Microemulsion acid system
The microemulsion acid is prepared by diluting 3% bicontinuous microemulsion (mother liquor) in 20% hydrochloric acid (Fig. 1 ). The mother liquor is a uniform mixture of nonionic emulsifiers (45% combination of nonionic surfactants and 20% short-chain alcohol cosurfactant), 15% oil phase (naphthenic hydrocarbons and naphthenic acids) and 20% deionized water. Similar to the study of Carvalho et al. [10], the structure of the microemulsion acid belongs to O/W type, which can significantly reduce the cost and increase the equivalent concentration of acid. The viscosity and frictional characteristics of the microemulsion acid are basically the same as those of hydrochloric acid, and the uniformly dispersed micelle molecules are about 9-22 nm at 70 °C. Experiments have proved that the particle size of the spent acid molecules almost unchanged after adding sufficient CaCO3 into microemulsion acid, so acid-rock reaction doesn’t affect the stability of the microemulsion structure.
Fig. 1. Schematic diagram of microemulsion acid preparation. |
Vigorous reactions between hydrochloric acid and calcite in reservoir, short etched distance and severe damage to near-well rocks often lead to the failure of acid fracturing treatment. Low-cost ways to reduce the H+ mass transfer rate include the viscosity-raising type represented by gelled acid and the contact-isolating type represented by emulsified acid, but both lead to an increase in acid viscosity above 50 mPa·s, which makes it difficult to create the conditions for “fingering” in multi-stage alternating acid fracturing operation [16]. Gelled acid is inexpensive but has severe solid phase residues, while emulsified acid is difficult to prepare and has high fri ctional loss, which limits the use of them. To clarify the inhibitory effect of electrostatic adsorption of microemulsion acid on the mass transfer rate of H+, hydrochloric acid, gelled acid, emulsified acid and surfactant acid with the same H+ concentration as the microemulsion acid (6.0 mol/L) were prepared, and acid-rock reaction experiments were carried out and compared. The formula and preparation methods are shown in Table 1 . The chemical suppliers are given in Table 2 .
Table 1. Acid formula and preparation methods |
| Acid | Formula | Viscosity/(mPa·s) | Preparation method |
|---|---|---|---|
| HCl | 20% HCl+1% corrosion inhibitor | 1.5 | Directly mix |
| Gelled acid | 20% HCl+0.6% gelled agent+1% corrosion inhibitor | 679.56γ-0.51 | See Reference [18] |
| Oil phase: 93% diesel + 7% emulsifier; Acid phase: 8.57 mol/L HCl + 1% corrosion inhibitor. The volume ratio of oil phase to acid phase is 3:7 | 571.61γ-0.439 | See Reference [19] | |
| Surfactant acid | 20% HCl + 3% HSC-25 (nonionic surfactant combination) + 1% corrosion inhibitor | 1.7 | Directly mix |
| Microemulsion acid | 20% HCl+3% diluted microemulsion+1% corrosion inhibitor | 1.8 | Directly mix |
Note: γ—shear rate, s-1 |
Table 2. Chemicals, product codes and suppliers |
| Description | Product code | Supplier |
|---|---|---|
| Concentrated hydrochloric acid | HCl | Tianjin Fuyu Fine Chemical Co. |
| Emulsifier | EEA | Beijing Hongyi Enze Co. |
| Microemulsion acid | SPRA | KMS Oil Field Chemicals & Technical Service Ltd. Beijing |
| Thickener | KMS20 | |
| Cleanup additive | HSC-25 | |
| Corrosion inhibitor | DCA-6 | |
| Drag reduction agent | cHVFR | |
| Sodium hydroxide | NaOH | Shanghai Maclean Biochemical Technology Co. |
| Calcium carbonate | CaCO3 | |
| Potassium chloride | KCl | |
| Phenol red indicator |
1.2. Reaction characteristics and retardation mechanism
The rotating disk method was used to calculate the reaction rate of microemulsion acid with calcite and to quantify the diffusion rate of H+. The experiments were conducted using Indiana limestone outcrop with more than 95% calcite minerals and 1.3%-2.1% clay minerals and showing water-wet characteristics. The experimental temperature was 70 °C, and the pressure was 7.5 MPa [2]. Each group of experiment was tested for 15 min, and the acid sample was collected every 3 min. The reacting amount of acid and rock was calculated by titration (see Reference [19] for detailed steps).
Fig. 2. Ca2+ concentration variation in acid-rock reaction for different acid systems. |
The acid mass transfer coefficient can be calculated by using a fitted line of 1/2 power of reaction rate and rotating rate. The H+ mass transfer coefficient of microemulsion acid was calculated to be 0.65×10−5 cm2/s under experimental conditions. Using the same method, the H+ mass transfer coefficients of the other four acids were calculated for comparison (Fig. 3 ). It should be noted that the core treated with hydrochloric acid fell off at high rotational rate, so that only the valid time point data were used for calculation. The results show that hydrochloric acid has the highest H+ mass transfer coefficient (2.4×10−5 cm2/s) and emulsified acid has the lowest mass transfer coefficient (0.03×10−5 cm2/s). Emulsified acid has the strongest inhibition to H+ mass transfer rate due to its isolated contact effect, followed by gelled acid with polymer added to increase viscosity. From the results of microemulsion acid and surfactant acid, adsorption has a significant inhibitory effect on the reaction rate, but the retardation performance is not as good as other conventional acids.
Fig. 3. H+ mass transfer coefficients of different acids. |
Fig. 4. Results of acid etching using rotating disk. |
After removing the interference from impurities such as clay minerals, the surface of limestone without crude oil is water-wet and positively charged [21-22]. Ma et al. [21] showed that the adsorption of an anionic surfactant onto the surface of water-wet limestone was significantly higher than that of nonionic and cationic surfactants, indicating that the adsorption depends mainly on electrostatic forces, but it is not suitable for acid-rock reaction. The Ca2+ salt present in acid-rock reaction causes the hydration radius of anionic surfactant to be sharply compressed and the water solubility of surfactant molecules deteriorates to form precipitation or enter oil phase, resulting in destruction to the microemulsion structure (Fig. 5a ). This is also the reason why no anionic surfactant acid has been seen [9]. Similar to surfactant acid, microemulsion acid also creates a barrier between rock and acid by adsorption, which reduces the H+ mass transfer rate. However, microemulsion acid can add negatively charged polar additives (naphthenic acids) to oil core and makes the microemulsion negatively charged on the whole (Fig. 5b ), which electrostatically enhances its adsorption capacity on limestone surface. Negatively charged oil cores are not miscible with Ca2+ salt, and the nonionic surfactant has a certain isolation effect, so the overall structure of the microemulsion will not be destroyed by Ca2+ from continuous acid-rock reaction. The adsorption performance can be regulated and the stability of microemulsion structure can be kept through component optimization [1]. This is the reason why the H+ mass transfer coefficient of microemulsion acid in Fig. 3 is about 42.8% lower than that of surfactant acid.
Fig. 5. Adsorption mechanisms of microemulsion acid and anionic surfactant. |
2. Differential etching on fractures
The morphological characteristics of support points after acid fracturing can be divided into four etching types: rough, grooved, turbulent and uniform [18]. The rough type is the most common type with the lowest treatment requirements, but it requires 5%-15% of mineral composition insoluble in acid. The calcite content in most Middle Eastern bioclastic limestones is more than 95%, and only less than 2% of the minerals are insoluble in hydrochloric acid, which is not conducive to the formation of rough etching after acid fracturing. For this type of reservoir, it’s better to conduct alternating acid injection at high rate pumping to create grooves [23] and turbulent etching [18]. The viscosity of microemulsion acid is similar to that of hydrochloric acid and can be mixed with variable viscosity reducers to reduce viscosity [24]. It can meet the requirements of high-rate and low-friction acid injection in multistage alternating acid fracturing process [18-19].
2.1. Experimental scheme
Four rock slabs were made from Indiana limestone outcrop with permeability of about 1×10−3 μm2 and porosity of about 20%. The shape of the slabs follows the API standard conductivity chamber. Uniaxial rock mechanics tests showed the elasticity modulus is 10.9 GPa and the Poisson's ratio is 0.24, and the rock is plastic [25]. The rock slab is split from the middle [23] to simulate the surface of rough hydraulic fracture after fracturing. The sum of the slab thicknesses is 50.0 mm. The slab is placed in the conductivity chamber after applying gasket adhesive around it and then injected with grout adhesive to seal the slab. To ensure the fracture width is the same, a copper plate of 0.50 mm thick should be inserted at the inlet and the outlet of the conductivity chamber, and then the plate is taken out after the slab at the right place. The equipment and procedures follow Zhang et al. [26]. The experimental temperature is 70 °C. The prepad fluid is clean slick water with variable viscosity [24] (the primary component is cationic polyacrylamide with a molecular weight of about 6×106, and the viscosity ratio to microemulsion acid is more than 50 when the shear rate is higher than 170 s−1). Prepad fluid and acid were injected alternately at 100 mL/min in three stages, and each of which lasted 10 min. Four groups of etching experiments were conducted on four slabs and using four kinds of acid (Table 1 ). (The significant difference between surfactant acid and microemulsion acid is the flowback performance, so surfactant acid is not selected as the object in this section).
2.2. Etching and conductivity
Fig. 6. Etching difference distributions of different acid systems. |
The shape of etched channels is influenced by its initial roughness. The roughness in the left and right rock slabs after acid etching shows a large difference at the peak and the valley (Fig. 7 ). From the change of the etching difference at the scanning position, whether etched by hydrochloric acid or microemulsion acid, the etched amount in the peak area of the right rock slab is relatively larger, while the etched amount in the valley area of the left rock slab is relatively smaller. This is because the peak area in the rock slab has strong hindrance to acid flow, so the etched degree is large. In addition, the strong hindrance from the peak area makes the high-speed flow of acid bifurcate, which changes the flow path and accelerates local flows, resulting in a low etched degree in the valley area of the rock slab on the other side, and finally producing differential etching. This result is similar to that observed by Gou et al. [23].
Fig. 7. The rock slab before etching and variation of etching difference. |
The rock slabs were characterized quantitatively using surface tortuosity [27] (Table 3 ). It can be found that: (1) The surface tortuosity of the rock slabs decreased by about 13.6% and 8.1% after etched by emulsified acid and gelled acid, respectively, indicating that the originally rough parts on the fracture surface were etched and the surface became relatively smooth, which is consistent with the results of Pournik [18]. Emulsified acid and gelled acid etched the rock slabs relatively uniform, so the decrease of the surface tortuosity is relatively large. (2) The surface tortuosity slightly increased after etched by microemulsion acid and hydrochloric acid because microemulsion acid has the ability to enter the deep part of the fracture wall to form a net-like etching, while hydrochloric acid has the largest H+ mass transfer coefficient, fast acid-rock reaction, low viscosity, and strong flow ability, so it can induce differential etching [28], which improves the surface tortuosity of the rock slab to a certain extent.
Table 3. Variation of surface tortuosity |
| Acid | Before/after reaction | Surface tortuosity | Acid | Surface tortuosity | |
|---|---|---|---|---|---|
| Emulsified acid | Before reaction | 1.324 | Microemulsion acid | Before reaction | 1.430 |
| After reaction | 1.144 | After reaction | 1.442 | ||
| Gelled acid | Before reaction | 1.260 | Hydrochloric acid | Before reaction | 1.410 |
| After reaction | 1.158 | After reaction | 1.413 |
During multi-stage alternating acid injection, viscous fingering, inlet jet and initial roughness disturb the flow field. The differential flow field induces differential acid-rock reaction, which is important to improve the conductivity of hydraulic fractures. Gelled acid and emulsified acid have higher viscosity and relatively uniform flow. In addition, due to the relatively low acid-rock reaction rate, even if there is a difference in acid flow velocity, it is not enough to induce a dominant acid flow channel, so gelled acid and emulsified acid fail to show differential etching phenomenon.
Fig. 8. Fracture conductivity of rock slabs after etching at different closure pressure. |
3. Spent acid flowback
The mainstream throat radius of Middle Eastern porous bioclastic limestone is smaller than that of sandstone with the same permeability, but the former’s throat radius varies more widely, resulting in stronger water lock and greater solid-phase residue damage [5,29]. From the cases in Eagle Ford and other places, clean and easy-to-flow-back fracturing fluid can increase the production of crude oil by more than 20%, but using conventional acid fracturing fluid cannot prevent water lock and reduce solid phase residue [15]. Surfactant and microemulsion are chemical reagents that are clean and can prevent water block. Li et al. [22] used a microscopic model to visually reveal the mechanism of surfactant micelle solution to improve flowback efficiency of tight reservoirs, but the pH and salinity of spent acid flowback environment are special. Therefore, the flowback mechanism after microemulsion acid fracturing needs further understanding.
3.1. Experimental design
Four limestone cores from layer M of Halfaya oil field in Iraq were used to evaluate the flowback effect of different acids. The core diameter is 3.81 cm, the length is 8.0 cm, the permeability is about 0.7×10−3 μm2, and the porosity is about 18 %. The experimental crude oil taken from Tarim limestone reservoir is dewatered and degassed, with a density of 0.879 g/cm3, wax, colloid and asphaltene content of 9.7% and viscosity of 3.91 mPa·s (at 70 °C). The wettability of the limestone pore throat was converted using naphthenic acid to simulate the oil-wet characteristics of the subsurface core [21]. The acid was prepared in the same way as in Table 1 . The spent acid was prepared by using CaCl2 to replace HCl in the acid formula and adjusting the pH to 4-5. Due to the high viscosity of gelled acid, its flowback ability is worse than that of hydrochloric acid, so gelled acid was not chosen for our experiment.
The flowback experimental setup is shown in Fig. 9 , and the experimental steps are as follows: (1) Put the core into the core holder, add the confining pressure to 8 MPa and outlet back pressure to 3 MPa to simulate the pore pressure [7], and vacuum from the inlet for 48 h. (2) At a rate of 0.05 mL/min, 10.00 PV (pore volume multiple) crude oil is injected from the outlet to saturate the core. (3) Inject 0.15 PV spent acid at a constant rate of 0.05 mL/min from the inlet to simulate spent acid intrusion. (4) Inject 2.00 PV crude oil at a constant rate of 0.05 mL/min from the outlet to simulate flowback process. (5) Record pressure data at real time, and quantitatively evaluate the change of saturation field in core by CT in-situ scanning with the method provided by Su et al. [30].
Fig. 9. Flowback experiment with CT in-situ scanning. |
3.2. Mechanism and law of spent acid flowback
Fig. 10. Change of oil saturation field during water intrusion and flowback of spent acid. |
Fig. 11. Change of spent acid retention at different stages. |
Fig. 12. Relationship between flowback pressure and flowback volume. |
In the early stage of flowback, water phase was continuously produced from the core, and the oil and water relative permeabilities were influenced by water saturation. The water saturation hardly changed at the late stage of flowback, so the recovery degree of the oil phase relative permeability at the late stage of flowback can be estimated using the method from Longoria et al. [31]. Fig. 13 shows the relationship between oil phase relative permeability and flowback volume. It can be seen that the oil phase relative permeability is the highest for microemulsion acid and the lowest for emulsified acid after flowback. When the flowback volume reaches 2.0 PV, the oil phase relative permeability for microemulsion acid, surfactant acid, hydrochloric acid and emulsified acid can be recovered to 0.37, 0.26, 0.21 and 0.19, respectively, indicating that the flowback performance of spent microemulsion acid is much better than the other three acids.
Fig. 13. Oil phase relative permeability vs. flowback volume. |
The spent acid and crude oil were mixed thoroughly in the test tube by the volume ratio of 1:1, kept at a constant temperature of 70 °C, and the emulsion morphology was observed by taking samples regularly (Fig. 14 ). From the photos, it can be seen that: (1) The spent emulsified acid and crude oil formed a more stable water-in-oil emulsion, and water droplets tended to aggregate and became large after 3 d, but the emulsion was not broken; (2) Oil-in-water emulsion was formed at the beginning of the spent surfactant acid mixing with crude oil, but the emulsion gradually aggregated and only a small amount of water droplets remained in the crude oil, and the emulsion was completely broken after 3 d; (3) The spent microemulsion acid mixed with crude oil was in a nearly miscible state under ultra-low interfacial tension, and then the crude oil and water droplets in the mixture gradually separated and aggregated, until the color of the mixture tended to be homogeneous after 3 d.
Fig. 14. Micrographs of spent acids mixed with crude oil. |
The above experiment shows that: (1) Compared with spent hydrochloric acid, crude oil is a wetting phase, and it is easy to gather the crude oil in the reservoir pores and throats through capillary imbibition to form a continuous phase, which occupies the flow channel and is not conducive to the spent acid flowback, resulting in low flowback rate; (2) The spent emulsified acid is a nonuniform emulsion, of which the water phase is easy to gather into large droplets in pores and induce water lock, so the spent acid flowback rate is the lowest due to great flowback resistance; (3) The spent surfactant acid and crude oil formed an oil-in-water emulsion, but the emulsion dissipated after a short time, so that almost no water lock occurred, but its wetting reversal may lead to the formation of oil droplets which may be stuck in pore throats, raising flowback resistance, and resulting in low recovery of oil relative permeability; (4) The spent microemulsion acid and crude oil can be miscible to form microemulsion with ultra-low interfacial tension among microemulsion, oil and water, almost miscible, which is conducive to flowback, bringing the highest spent acid flowback rate, and the highest recovery of oil relative permeability.
4. Conclusions
After adding a negatively charged polar agent, microemulsion oil core can be negatively charged, and its adsorption capacity onto limestone surface can be enhanced. The adsorption can establish an isolating barrier between rock and acid to realize the reduction of the H+ mass transfer rate. The negatively charged oil core is not miscible with Ca2+ salt, and the overall structure of the microemulsion will not be destroyed by Ca2+ salt generated from acid-rock reaction. The adsorption performance is adjustable and the microemulsion structure is stable.
Microemulsion acid has a certain retardation ability and low viscosity, which has a strong ability to penetrate deep into the fracture wall, and forming channels by differential etching and realizing net-like etching, which can greatly improve the permeability of the reservoir around the fracture. Meanwhile, it can maintain high conductivity under high closure pressure.
The mixture of spent microemulsion acid and crude oil shows the characteristics of ultra-low interfacial tension. The microemulsion, oil and water are almost in a miscible state without any droplets or water lock effect. The flowback resistance is low, the flowback rate of spent acid is high, and the recovery of crude oil relative permeability is high.