- Email: [email protected]
Potential of alkaline flooding to enhance heavy oil recovery through water-in-oil emulsification
Fuel 104 (2013) 284–293
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Potential of alkaline flooding to enhance heavy oil recovery through water-in-oil emulsification Haihua Pei a, Guicai Zhang a,b,⇑, Jijiang Ge a, Luchao Jin a, Chao Ma a a b
College of Petroleum Engineering, China University of Petroleum, Qingdao 266555, People’s Republic of China State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266555, People’s Republic of China
h i g h l i g h t s " Alkaline flooding can improve sweep efficiency through water-in-oil emulsification. " Improvement of sweep efficiency increases with the alkaline concentration. " There is an optimum slug size that results in the highest tertiary oil recovery. " The injection pattern is recommended to the continuous alkaline injection. " There is an optimum injection rate that obtains the highest tertiary oil recovery.
a r t i c l e
i n f o
Article history: Received 23 September 2011 Received in revised form 12 June 2012 Accepted 14 August 2012 Available online 7 September 2012 Keywords: Alkaline flooding W/O emulsion Sweep efficiency Sandpack flood test Enhanced heavy oil recovery
a b s t r a c t Alkaline flooding has great potential for enhancing the recovery of heavy oil, especially for reservoirs in which thermal methods are not suitable. In this study, alkaline flooding tests were performed in micromodels and sandpacks to investigate the microscopic displacement mechanisms for enhancing heavy oil recovery and the effect of the injection parameters on displacement efficiency. The micromodel tests indicate that the penetration of the alkaline solution into the crude oil and the subsequent formation of a water-in-oil (W/O) emulsion reduce the mobility of the water phase and divert the injected water into the unswept region, thereby improving the sweep efficiency. The sandpack flood results show that the tertiary oil recovery can reach about 20% of the initial oil in place (IOIP) using 1.0% NaOH, and the tertiary oil recovery has been found to increase as the alkaline concentration increases. However, there is an optimum slug size and injection rate at which the highest tertiary oil recovery can be obtained during the alkaline flooding process. Continuous alkaline injection can provide a higher tertiary oil recovery compared with a cyclic alkaline injection pattern. These results show that the alkaline flooding, if properly designed and controlled, can lead to enhanced heavy oil recovery through the water-in-oil emulsification. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Many countries, notably Venezuela, Canada, the United States and China, possess abundant heavy oil resources [1]. With the depletion of light oil resources and rising energy demands, the successful recovery of heavy oil is becoming increasingly important. However, the high viscosity of heavy oil makes it difficult to recover. Currently, only about 3–10% of the initial oil in place (IOIP) in heavy oil reservoirs can be recovered under primary production [2]. In order to recover additional heavy oil, different fluids usually have to be injected in order to displace the oil to the production ⇑ Corresponding author at: College of Petroleum Engineering, China University of Petroleum, Qingdao 266555, People’s Republic of China. Tel.: +86 53286981178. E-mail address: [email protected] (G. Zhang). 0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.08.024
wells. However, mobility ratio concerns dominate the displacement of viscous oil, and most EOR processes focus on reducing the oil viscosity or improving the mobility ratio [3]. Unfortunately, many heavy oil reservoirs in China are relatively thin or deep, making these reservoirs poor candidates for expensive thermal methods. Therefore, the displacement mobility ratio in these reservoirs needs to be improved using an inexpensive process [4]. Water flooding is a common and inexpensive secondary oil recovery technique for the heavy oil reservoirs with oil viscosities ranging from 100 to 10,000 mPa s. However, the incremental recoveries by water flooding are quite low, due to the poor sweep efficiencies caused by the adverse mobility ratio between oil and water [5]. As a result, most of the residual oil at the end of water flooding is bypassed, which is greatly different from that of the residual oil trapped by capillary forces in conventional oil
285
H. Pei et al. / Fuel 104 (2013) 284–293
reservoirs [6,7]. Therefore, the key problem in heavy oil reservoir is inefficient sweep due to low mobility of the oil, not the residual oil in the swept region [8]. Previous researches have demonstrated that alkaline flooding can potentially increase the recovery of heavy oil through the mechanism of emulsification [9–11]. However, different views on the mechanism for alkaline flooding to improve heavy oil recovery are proposed. Some researchers proposed the formation of oil-inwater (O/W) emulsions as a possible recovery mechanism [12– 14]. In this application, oil is emulsified into water phase and the oil droplets either plug rock pores and give improved sweep efficiency, or are perhaps entrained along with the flowing aqueous phase. Alternatively, the formation of water-in-oil (W/O) emulsions has also been proposed as a possible recovery mechanism [15,16]. These emulsions should be more viscous than oil by itself, thus they could possibly lead to improvements in the mobility ratio and sweep efficiency of the flood [17,18]. The aforementioned researches indicate that the flow of W/O emulsion and O/W emulsion in porous media are two mechanisms of enhanced oil recovery by alkaline flooding. Generally, when alkaline concentration is low in the system with addition of surfactant, the mechanism of O/W emulsion prevails in the oil displacement process. When high alkaline concentration is applied, the mechanism of W/O emulsion is prominent. Compared with O/W emulsion, the W/O emulsion has much higher viscosity, which can effectively block water channels to improve sweep efficiency [19]. This mechanism is believed to be especially effective for heavy oil, where the sweep efficiency of water flooding is usually poor. Although alkaline flooding has been extensively implemented in light oil reservoirs, only a few alkaline flooding studies and field tests have been conducted in heavy oil reservoirs. This is mainly ascribed to the fact that multiphase flow in heavy oil reservoir is a more complicated process than that of the light oil reservoirs. Moreover, heavy oil has higher contents of natural petroleum acids [20], which resulting in the mechanism of alkaline flooding for improving oil recovery to be more complicated. Therefore, the effect of various operating parameters on the performance of alkaline flooding in heavy oil reservoirs is not well understood. This study is aimed at studying the potential of alkaline flooding and the microscopic displacement mechanisms for enhanced heavy oil recovery. The research is focused on the mechanism of improving the sweep efficiency through water-in-oil emulsification, and the effects of the injection parameters on the displacement efficiency are optimized by measuring the pressure drop and recovery efficiency in sandpack flood tests. The findings of this investigation can be utilized for optimizing various operational conditions, in order to maximize oil recovery, by injecting alkaline solution into the heavy oil reservoirs.
2. Experimental
The heavy oil has a viscosity of 2000 mPa s at 55 °C. The acid number value of the oil is 2.69 mg KOH/gram of sample. The formation brine has a salinity of 0.5%, and the concentrations of Ca2+ and Mg2+ in the brine are relatively low. Thus, all of the solutions used in this study were prepared with NaCl solutions with concentrations of 5000 mg/L. The alkaline agent used in this study was sodium hydroxide (NaOH). 2.2. Measurements of interfacial tension The interfacial tensions (IFTs) between the oil and the different alkaline solution systems were measured using a Model Texas-500 spinning drop interfacial tensiometer at 55 °C. In the cases where alkaline were added to the water phase, the rotation speed was sufficiently high that the length of the oil drop is larger than four times its diameter and the oil–water IFT was determined with an image–capture device and image-acquisition software according to the following equation:
r ¼ 1:2336ðqw qo Þx2
3 D ; n
L P4 D
ð1Þ
where r is the oil–water interfacial tension in mN/m, qw and qo is the density of water phases and oil phase in g/cm3, respectively, x is the rotational velocity in rpm, D is the diameter of the oil drop in 104 m, L is the length of the oil drop in 104 m, n is the refractive index of water phase. The values of the IFT determinations were repeatable within ±0.005 mN/m. 2.3. Micromodel flood studies A glass-etched micromodel was used to investigate the displacement mechanisms of alkaline flooding. The micromodel was constructed by etching a two-dimensional network of pores and throats on glass plates using a photochemical method. The pore network used in this study was patterned, based on the pore structure of a core obtained from the reservoir. The transparent nature of the micromodel allowed pore-scale multiphase displacements to be visually observed. To observe these phenomena easily during the flooding, 0.05% eosin was added to color the injected brine in the micromodel flood tests. The procedure for the micromodel test was as follows: after being vacuumed, the micromodel was prepared for flooding by filling it with brine, and the brine was subsequently displaced by crude oil until no more water was produced. Because the viscosity of the crude oil was higher than the resident water, almost all of the water present in the micromodel was displaced. After the heavy oil was saturated, the model was aged for 24 h. Next, brine with the same salinity as the alkaline solution was injected into the micromodel at a constant flow rate of 0.003 mL/min. Using a video recorder and camera apparatus, the micromodel flood was visualized during the different stages of fluid injection.
2.1. Fluids and chemicals Oil and formation brine samples were collected from the Binnan heavy oil reservoir in the Shengli oilfield in China. To remove the solids and water, the heavy oil was centrifuged at 10,000 rpm at reservoir temperature (55 °C) for 4 h. The viscosity, density, and acid number of the oil were analyzed and are listed in Table 1.
Table 1 Basic properties of the heavy oil sample. Density @ 55 °C (kg/m3)
Viscosity @ 55 °C (mPa s)
Acid number (mg KOH/g oil)
Resin (wt.%)
Asphaltene (wt.%)
947.2
2000
2.69
19.5
2.033
2.4. Sandpack flood studies All chemical flooding tests for heavy oil recovery were carried out using sandpacks. The sandpack used in this study was 2.35 cm in diameter and 19.4 cm in length. For each sandpack test, fresh quartz sand was wet-packed to ensure the same wettability for all the tests. The sandpack was packed as follows: fresh quartz sands with the size fractions of 80–100 and 100–200 meshes obtained from sieve screening were blended at a fixed weight ratio of 3:1. A coreholder filled with formation brine was positioned vertically, and the sand was added in several increments to fill the coreholder. In each step, the sand was shaken slightly after being poured in. During this process, the water surface was kept above
286
H. Pei et al. / Fuel 104 (2013) 284–293
3. Results and discussion
rium value of the dynamic IFT decreases first and subsequently increases with increasing alkaline concentration in the brine. When the NaOH concentration is 0.2%, the equilibrium value of the dynamic IFT reaches the minimum. This is ascribed that the IFT is influenced by pH in alkali/acidic oil systems. Rudin and Wasan [21–23] reported that pH is important for producing the IFT minimum because the un-ionized acid in the acidic oil contributes to the IFT in conjunction with the ionized acid and the extent of acid ionization is governed by pH. An optimal pH exists where IFT is minimized. In the limit of low pH, the IFT corresponds only to adsorption of un-ionized acid on the interface. In the limit of high pH, most of the un-ionized acid has been ionized and the IFT corresponds to adsorption of mostly ionized acid. As pH is increased, IFT decreases because of the ionized acid increase until CMC is reached. Then, the ionized acid remains constant and the un-ionized acid continues to decrease, resulting in an increase in IFT until most of the un-ionized acid is converted. At this point, IFT remains fairly constant with further pH increase.
3.1. IFT behavior of Binnan oil/alkaline systems
3.2. Micromodel flood study
To investigate the effectiveness of the alkaline solution in reducing the oil–water interfacial tension, the interfacial tension behavior of the Binnan oil/brine was studied with different NaOH concentrations ranging from 0.1% to 1.0%. Fig. 1 shows the dynamic IFT curves between the oil and brine with different NaOH concentrations. The results indicate that the dynamic IFT value between the oil and NaOH system varies in the range of 0.03–0.12 mN/m. From the dynamic IFT curves, it is also observed that the oil–water IFT has obvious dynamic effect. These curves can be divided into three stages: in the first stage, the IFT shows a transient decrease, and the lowest IFT occurs in this stage; in the second stage, the IFT increases sharply; and in the third stage, the IFT becomes constant. The dynamic effect of the IFT is related to the interfacial interaction between the alkaline solution and the heavy oil phase. When an acidic crude oil contacts an alkali in the alkaline flooding, the acidic components in the crude oil react with the alkali in the flooding water to produce in situ surfactant which lowers the oil–water IFT. Such surfactants accumulate at the oil–water interface and then, some of them gradually diffuse into the bulk aqueous phase and/or oil phase, depending on their affinities for the aqueous and oil phases. Thus, IFT between the oil phase and alkaline solution changes with time and finally reaches an equilibrium value. As shown in Fig. 1, it takes 1 h for the IFT between each oil drop and alkaline solution to reach an equilibrium value. The equilib-
To investigate the displacement mechanisms of alkaline flooding for improving heavy oil recovery, it is necessary to examine the distribution of the residual oil after water flooding. A micromodel flooding test with 0.5% NaCl was first conducted; images of the oil distribution during the water flooding process are shown in Fig. 2. Due to the glass-etched micromodel is a water-wet quarter 5-spot model and the injecting well and producing well are located on the diagonal line of the model, thus the injected water creeps along the pore wall (see Fig. 2a) and fingers through the quarter 5-spot model. As a result, the injected water advances diagonally across the micromodel (see Fig. 2b), and several connected water channels are created diagonally after water breaks through (see Fig. 2c). After that, little oil can be recovered by continued water flooding. Consequently, most of the oil is bypassed, due to the viscous fingering caused by the adverse mobility ratio between the oil and the water. The poor sweep efficiency of the water flooding leaves most of the micromodel area untouched (see Fig. 2d). In another micromodel flood test, an alkaline slug consisting of 1.0% NaOH + 0.5% NaCl was injected, and the microscopic images of oil distribution during alkaline flooding process are shown in Fig. 3. It is observed that the alkaline solution penetrates the heavy oil and creates large water drops that are coated with thin oil films (see Fig. 3a). After entering the pore space, these large water drops divide into small discontinuous water droplets inside the oil phase, as shown in Fig. 3b. As additional water droplets appear in the oil phase, a W/O emulsion bank eventually forms (see Fig. 3c). Fig. 3e shows the oil distribution of the entire micromodel when the alkaline solution reaches the outlet. At this point, a relatively uniform degree oil saturation is distributed over the entire model. This is because the viscosity of a W/O emulsion is much higher than the viscosity of the water phase and even higher than the viscosity of the oil phase [24], which lead to the significant increase of the resistance to flow of the alkaline solution in the porous media. As a result, oil is subsequently displaced in the form of the W/O emulsion with little viscous fingering. Therefore, it is the W/O emulsion that reduces the mobility of water phase and diverts the injected alkaline solution to the unswept region of micromodel to improve the sweep efficiency. This is the mechanism that will be studied in alkaline flooding for heavy oil to improve sweep efficiency and thus enhanced oil recovery. Although the high viscosity of the W/O emulsion can increase the resistance to flow of the water phase in the high permeability zone and thus leads to improved sweep efficiency, the high viscosity also causes poor mobility. As a result, a number of residual W/O
the top of the sand to ensure that air was not introduced into the sample. The sandpack flooding was conducted horizontally. The experimental procedure was briefly described as follows. After the permeability of the sandpack in the presence of the formation brine was measured, the wet-packed sandpack was subsequently saturated with the heavy oil. The oil injection was continued until water production almost ceased (the water cut was less than 1%). After the oil injection, the sandpack was waterflooded until the oil production became negligible (oil cut was less than 1%). For the tertiary chemical flood tests, 0.25–2.5 pore volume (PV) chemical slugs were injected. The chemical injection was followed by extended water flooding until the oil production became negligible. All of these tests were conducted at the reservoir temperature (55 °C), except where otherwise specified.
Fig. 1. Dynamic IFT curves between the Binnan oil and brine with different alkaline concentrations.
H. Pei et al. / Fuel 104 (2013) 284–293
287
Fig. 2. Microscopic images of oil distribution during the water flooding process. Images: (a) 0.10 PV brine injected; (b) 0.50 PV brine injected; (c) water breakthrough (1.0 PV brine injected); and (d) end of water flooding (10.0 PV brine injected).
emulsions remain after alkaline flooding, as shown in Fig. 3d, which lowers the displacement efficiency of the alkaline flooding in the swept area. Fig. 3f shows that a certain amount of residual oil is left throughout the model. Thus, the W/O emulsions with high viscosity have also negative influence on the recovery of heavy oil. 3.3. Sandpack flood study To evaluate the potential of alkaline flooding for enhanced heavy oil recovery through water-in-oil emulsification, 16 flood tests were conducted in sandpacks. The effects of alkaline concentrations, slug size, injection pattern, and injection rate on alkaline flooding were investigated in these tests. The parameters of the sandpacks, chemical slug compositions, and flood results are summarized in Table 2. For each sandpack flood test, the oil recovery behavior, water cut and pressure drop were monitored and analyzed. It should be noted that an extensive de-emulsification process was needed to accurately measure the volume of oil and water when the emulsions were produced. 3.3.1. Effect of alkaline concentration In order to examine the effectiveness of alkaline concentration in creating the in situ emulsification in the alkaline flooding for the Binnan heavy oil, a series of sandpack flood tests (Runs 1–6) were conducted with NaOH concentrations ranging from 0.1% to 1.0%. In all of these tests, alkaline solutions with slug size of 0.5 PV were injected at a rate of 0.5 ml/min after the initial water flooding. Fig. 4 illustrates the tertiary oil recovery as a function of alkaline concentration. It is observed that the tests with low alkaline concentration (0.1% in Run 1 and 0.2% in Run 2) have a small increase of oil recovery. The incremental oil recovery increases with the alkaline concentration, and it exhibits a significant
change when the alkaline concentration ranges from 0.1% to 0.4%; above this value, the increase becomes slight. The produced emulsions by the alkaline flooding in the sandpack flooding tests were examined by observing the droplet distribution and checking the droplet size under microscope. And ScopeImage Plus software was used to determine the droplet size distribution for all emulsions. Fig. 5 shows photomicrographs of droplet distribution of the produced emulsions in the sandpack flooding tests with different NaOH concentration (Runs 1–6). It is observed that the transparent water phase is in the form of dispersed droplet, which indicates that the emulsions are W/O type. Emulsions with lower alkaline concentration, such as the 0.1% and 0.2% NaOH, have smaller droplets and fewer large droplets (see Fig. 5a and b) compared to other emulsions with higher alkaline concentration, as shown in Fig. 5e and f. This means that the percentage of the large droplets in an emulsion increases with increasing alkaline concentration and vice versa for the small droplets. The droplet size distribution of the water droplet for the produced emulsions of 0.1%, 0.2%, 0.4%, 0.6%, 0.8%, and 1.0% NaOH were measured and are shown in Fig. 6. The droplet size distribution for the produced emulsion of 0.1% NaOH had a droplet size ranging from 0 to 2.652 lm, as illustrated in Fig. 6a, whereas Fig. 6f shows the droplet size distribution for the produced emulsion of 1.0% NaOH, in which the droplet size ranges from 0 to 6.857 lm. The average droplet sizes for the produced emulsions of 0.1%, 0.2%, 0.4%, 0.6%, 0.8%, and 1.0% NaOH were determined to be 1.012, 1.354, 1.637, 1.864, 2.016, and 2.785 lm, respectively. The volume fraction of dispersed water phase is defined as emulsion quality. Based on the microscope observations, it is found that the emulsion quality and droplet size increase with increasing alkaline concentration. In the literature, Arhuoma et al. [24] have carried out studies for determining viscosities of W/O emulsions in porous media. They
288
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 3. Pore-level images of the formation of W/O emulsions during the alkaline flooding process in a micromodel test. Injected alkalis: 1.0% NaOH + 0.5% NaCl. (a) Penetration into heavy oil by alkaline solutions; (b) formation of water drops inside the oil phase; (c) formation of W/O emulsion bank; (d) residual W/O emulsions after the alkaline flooding; (e) the alkali front at the outlet of the micromodel; and (f) after alkaline flooding.
Table 2 Summary of sandpack flood tests. Test #
Porosity (%)
Permeability (mD)
Initial oil saturation (%)
Waterflood recovery (%)
Chemical formula
Injection rate (ml/min)
Slug size (PV)
Tertiary recovery (%IOIP)
Final recovery (%IOIP)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
43.85 42.19 44.23 42.78 43.50 43.85 44.45 43.40 42.78 43.50 44.45 43.74 43.26 43.85 43.74 43.76
1964 2259 2260 2193 2193 2015 2071 2260 2193 1962 2130 2130 2130 2015 2071 1912
89.01 90.14 90.05 90.28 89.43 90.14 88.24 89.04 88.89 88.80 88.24 90.75 89.29 90.79 90.11 90.03
33.87 32.81 32.24 31.69 32.19 31.76 32.39 35.08 35.94 32.00 34.85 33.91 35.69 35.52 33.71 34.03
0.1% 0.2% 0.4% 0.6% 0.8% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0%
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.1 0.25 0.75 1.0
0.5 0.5 0.5 0.5 0.5 0.5 0.25 1.0 2.0 2.5 0.5 + 0.5 0.25 + 0.25 + 0.25 + 0.25 0.5 0.5 0.5 0.5
6.29 10.63 16.04 17.08 18.65 19.96 17.88 19.30 14.25 11.29 18.21 12.18 19.17 20.90 18.00 16.42
40.16 43.44 48.28 48.77 49.84 51.72 50.27 54.38 50.19 43.29 53.06 46.09 54.86 56.42 52.71 50.45
NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 4. Effect of alkaline concentration on tertiary oil recovery.
found that the viscosity of a W/O emulsion increased with increasing emulsion quality. Based on their experimental results, Wang et al. [25] derived the following correlation for the effective viscosity of a W/O emulsion in sandpacks:
logðle Þ ¼ kh þ logðlo Þ
ð2Þ
where le is the effective viscosity of the W/O emulsion, lo is the oil phase viscosity, and h is the emulsion quality. The value of the coefficient k is larger than unity and depends on the porosity and permeability of the sandpack. Besides, it is also confirmed through microscope observations (see Fig. 5) that the alkaline solution with higher concentration can create more water droplets in the effluent emulsion. As a result, the effluent emulsion quality increases with increasing alkaline concentration. Based on the Eq. (2), it is concluded that the effective viscosity of the W/O emulsion increases with increasing alkaline concentration. Thus, the viscosity of a W/
289
O emulsion becomes higher than the viscosity of the water phase and even higher than the viscosity of the oil phase. Because of the high viscosity of W/O emulsions, the resistance to water flow in the high water saturation zone can be increased significantly, which is also evident by the higher increments in the pressure drop. Therefore, it can be seen that the pressure drop along the sandpack increases with increasing NaOH concentrations, as shown in Fig. 7. From these figures, it can be concluded that for the same injection rate, the increase in oil recovery is accompanied by an increase in pressure drop, and a higher peak value of pressure drop results in a higher tertiary oil recovery. The built-up pressure drop suggests the penetration of alkaline solution into residual oil drops to form high viscosity W/O emulsions, which can block the high-permeability water channels and reduce the mobility of the water phase. As a result, the subsequent injected water is diverted to the unswept areas, thereby resulting in improved sweep efficiency and increased heavy oil recovery. On the basis of the sandpack flood results, 1.0% NaOH was found to be notably effective for enhanced heavy oil recovery by the in situ generated W/O emulsions. Based on these results, 1.0% NaOH concentration solutions were used in additional sandpack flood tests to investigate the effects of other parameters in the alkaline flooding scheme.
3.3.2. Effect of alkaline slug size Determining the smallest chemical slug size that is able to recover the maximum residual oil is an important factor in the optimization of the proposed EOR process. To evaluate the effect of alkaline slug size on tertiary oil recovery, a series of sandpack flood tests (Runs 6–10) were performed by increasing the alkaline slug size incrementally from 0.25 to 2.5 PV while keeping the NaOH concentration and the injection rate at 1.0% and 0.5 ml/min, respectively. The results of the tertiary oil recovery are plotted as a function of slug size in Fig. 8. The tertiary oil recovery first increases with increasing alkaline slug size and later decreases, and it reaches the maximum value at
Fig. 5. Photomicrograph of droplet distribution of the produced emulsion in the sandpack flooding tests: (a) 0.1% NaOH, (b) 0.2% NaOH, (c) 0.4% NaOH, (d) 0.6% NaOH, (e) 0.8% NaOH and (f) 1.0% NaOH.
290
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 6. Droplet size distribution of the produced emulsion in the sandpack flooding tests: (a) 0.1% NaOH, (b) 0.2% NaOH, (c) 0.4% NaOH, (d) 0.6% NaOH, (e) 0.8% NaOH and (f) 1.0% NaOH.
0.5 PV. It is indicated that a larger alkaline slug size does not correspond to higher incremental oil recovery and that there is an optimum slug size that results in the highest tertiary oil recovery. This result can be explained on the basis of the production of the in situ W/O emulsion. The W/O emulsion has both positive and
negative effects on enhancing oil recovery: on one hand, its higher viscosity plays a key role in the effective blockage of the water channel to divert the injected water to the unswept areas and increase the sweep efficiency; on the other hand, it is difficult for this emulsion to move because of its higher viscosity; therefore, there is
H. Pei et al. / Fuel 104 (2013) 284–293
291
meet a demand for effectively blocking the water channels caused by the initial water flooding. The subsequent water will be diverted to other oil-rich areas, leading to a sharp increase of oil recovery. However, when the alkaline slug size increases from 1.0 to 2.5 PV, the excessive alkaline solutions will not only form W/O emulsions in the water channels, but also they will advance into other oil-rich areas and produce lots of high viscosity W/O emulsions in these areas. Although some of W/O emulsions can block the preformed water channels, the negative influence caused by the excessive viscous W/O emulsions on poor mobility is strengthened, and therefore, the tertiary oil recovery decreases with an increasing alkaline slug size. On the basis of these results, a slug size of 0.5 PV is considered as the optimum for the above chemical formula and enhancing the recovery of Binnan heavy oil in homogeneous sandpacks.
Fig. 7. Pressure drop response as a function of fluid injected in the sandpack flooding tests.
Fig. 8. Effect of slug size on tertiary oil recovery.
always a certain amount of residual oil left after alkaline flooding (as seen in Fig. 3d and f). Therefore, when the alkaline slug size increases from 0.25 to 0.5 PV, the amount of W/O emulsions can only
Fig. 9. Cumulative oil recovery of three flood test (Runs 8, 11, 12) with different injection pattern.
3.3.3. Effect of injection pattern To investigate the effect of injection pattern on tertiary oil recovery by alkaline flooding, three sandpack flooding tests (Runs 8, 11, 12) were carried out using different injection patterns. The alkaline concentration and the injection rate were set at 1.0% and 0.5 ml/min, respectively. The total injection volume was kept constant at 1.0 PV in all of these tests. After the initial water flooding, the alkaline solution was injected using different patterns. In Run 8, a slug of 1.0 PV NaOH solution was injected continuously and was followed by an extended water flooding. Runs 11 and 12 used a cyclic alkaline injection; the slug size of each injection was 0.5 PV and 0.25 PV, respectively. The results for all three runs are presented in Fig. 9. Although the total volume of the alkaline slug remains the same, the incremental oil recovery gradually decreases with the number of slugs. This indicates that the continuous alkaline injection can obtain higher tertiary oil recovery than the cyclic alkaline injection. Pressure drop responses as a function of fluid injected in these three runs are presented in Fig. 10. The test with the continuous alkaline injection pattern (Run 8) has the highest pressure drop and the broadest pressure hump during the alkaline injection. For the cyclic alkaline injection pattern with an injection of 0.25 PV, the lowest and thinnest pressure humps appear in the first slug, and almost no peak injection pressure appears in the following three slugs. This result is because more W/O emulsions are produced with the continuous alkaline injection pattern, and the flow resistance of injected water increases significantly, which leads to a long-term plugging effect. For the cyclic alkaline injection patterns, in contrast, the high concentrations of the alkaline solutions
Fig. 10. Effect of injection pattern on pressure drop as a function of fluid injected.
292
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 11. Effect of alkaline injection rate on tertiary oil recovery.
are more likely diluted, and consequently, fewer W/O emulsions are produced. Moreover, the W/O emulsion bank is easily broken down during the alternating process, which leads to a short-term plugging effect. Therefore, considering alkaline concentration loss in cyclic injection, the injection pattern is recommended to select the continuous alkaline injection for alkaline flooding.
3.3.4. Effect of injection rate In order to examine the effect of alkaline injection rate on the tertiary oil recovery by alkaline flooding, additional sandpack flood tests (Runs 13–16 and Run 6) were carried out. The injection rate of the alkaline slug was increased incrementally from 0.1 ml/min to 1.0 ml/min. The NaOH concentration was kept constant at 1.0%, and the alkaline slug size was fixed at 0.5 PV. The incremental oil recovery as a function of alkaline injection rate is plotted in Fig. 11. The incremental oil first increases with the alkaline injection rate and later decreases, and it reaches the maximum value at 0.25 ml/min. This is because the driving force is not high enough to produce many W/O emulsions when the injection rate is low (0.1 ml/min in Run 13). However, higher injection rates also reduce the contact area and contact time between the injected alkaline solution and the oil in the sandpack such that the alkaline solution cannot fully react with the acidic components in the heavy oil; therefore, the number of W/O emulsions generated by the reaction between the alkali and the oil sharply decreases. Moreover, a higher injection rate can also cause the W/O emulsion bank to break down quickly, resulting in early water breakthrough. All these factors could contribute to a lower sweep efficiency and lower tertiary oil recovery. Therefore, there is an optimum injection rate at which enough W/O emulsions can be produced to block the water channels effectively and increase the flow resistance. This is consistent with the results of the flooding tests conducted by Cooke et al. [10] and Dong et al. [13], in which the highest incremental oil recovery is obtained at the optimum injection rate.
4. Conclusion A series of micromodel flood tests and sandpack flood tests were performed to evaluate the potential of alkaline flooding to enhance the recovery of Binnan heavy oil through water-in-oil emulsification. The results indicate that
(1) The mechanism of water-in-oil emulsification plays a prominent role in the alkaline flooding process for enhanced heavy oil recovery. Initially, the alkaline solution can penetrate the heavy oil to form water drops inside the oil phase, and W/O emulsions are produced. Because of high viscosity of the W/O emulsions, the high permeability zone caused by the preceding water flooding can be blocked, and the subsequent injected water phase is diverted to the unswept regions, thereby resulting in improved sweep efficiency and increased heavy oil recovery. (2) The water droplets number and droplets size in W/O emulsions increases with the alkaline concentration increasing, and a higher alkaline concentration results in a higher tertiary oil recovery. However, there is an optimum slug size that gives the highest tertiary oil recovery. (3) When the W/O emulsions displace the residual oil from the waterflood, continuous alkaline injection can give higher tertiary oil recovery compared with the cyclic alkaline injection pattern. The effectiveness of alkaline flooding depends on the injection rate. The highest incremental oil recovery is obtained at the optimum injection rate. (4) The alkaline flooding method is a promising EOR process for enhanced heavy oil recovery through the water-in-oil emulsification mechanism.
Acknowledgements Financial support by the National Natural Science Foundation of China (Grant 51104170), the Fundamental Research Funds for the Central Universities (Grant 12CX06024A), the Outstanding Doctoral Dissertation Training Program of China University of Petroleum (Grant LW110203A), and the Fok Ying Tung Education Foundation for Young Teachers in the Higher Education Institutions of China (Grant 114016) is gratefully acknowledged.
References [1] Selby R, Alikhan AA, Farouq Ali SM. Potential of non-thermal methods for heavy oil recovery. J Can Petrol Technol 1989;28:45–59. [2] Mai A, Bryan J, Goodarzi N, Kantzas A. Insights into non-thermal recovery of heavy oil. J Can Petrol Technol 2009;48:27–35. [3] Farouq Ali SM. Prospects of EOR techniques in Saskatchewan oil reservoirs. J Can Petrol Technol 1986;25:64–7. [4] Farouq Ali SM, Thomas S. The promise and problems of enhanced oil recovery methods. J Can Petrol Technol 1996;35:59–63. [5] Mai A, Kantzas A. Heavy oil water flooding: effects of flow rate and oil viscosity. J Can Petrol Technol 2009;48:42–51. [6] Bryan J, Kantzas A. Enhanced heavy-oil recovery by alkali–surfactant flooding. In: SPE Paper 110738-MS. SPE Annual technical conference and exhibition, Anaheim, California, U.S.A; 11–14 November, 2007. [7] Bryan J, Kantzas A. Investigation into the processes responsible for heavy oil recovery by alkali-surfactant flooding. In: SPE Paper 113993-MS. SPE/DOE Improved oil recovery symposium, Tulsa, Oklahoma, U.S.A; 19–23 April, 2008. [8] Bryan J, Kantzas A. Improved recovery potential in mature heavy oil fields by alkali–surfactant flooding. In: SPE Paper 117649-MS. SPE International thermal operations and heavy oil symposium, Calgary, Alberta, Canada; 20–23 October 2008. [9] Jennings Jr HY, Johnson Jr CE, Mcaliliffe CD. A caustic water flooding process for heavy oils. J Petrol Technol 1974;26:1344–52. [10] Cooke Jr CE, Williams RE, Kolodzie PA. Oil recovery by alkaline water flooding. J Petrol Technol 1974;26:1365–74. [11] Johnson Jr CE. Status of caustic and emulsion methods. J Petrol Technol 1976;28:85–92. [12] Liu Q, Dong M, Yue X, Hou J. Synergy of alkali and surfactant in emulsification of heavy oil in brine. Colloid Surf A 2006;273:219–28. [13] Dong M, Ma S, Liu Q. Enhanced heavy oil recovery through interfacial instability: a study of chemical flooding for Brintnell heavy oil. Fuel 2009;88:1049–56. [14] Ma S, Dong M, Li Z. Evaluation of the effectiveness of chemical flooding using heterogeneous sandpack flood test. J Petrol Sci Eng 2007;55:294–300. [15] Dong M, Liu Q, Li A. Displacement mechanisms of enhanced heavy oil recovery by alkaline flooding. Particuology 2012;10:298–305.
H. Pei et al. / Fuel 104 (2013) 284–293 [16] Zhang H, Dong M, Zhao S. Which one is more important in chemical flooding for enhanced court heavy oil recovery, lowering interfacial tension or reducing water mobility. Energy Fuels 2010;24:1829–36. [17] Ding B, Zhang G, Ge J. Research on mechanisms of alkaline flooding for heavy oil. Energy Fuels 2010;24:6346–52. [18] Pei H, Zhang G, Ge J. Analysis of microscopic displacement mechanisms of alkaline flooding for enhanced heavy oil recovery. Energy Fuels 2011;25:4423–9. [19] Pei H, Zhang G, Ge J. Comparative effectiveness of alkaline flooding and alkaline–surfactant flooding for improved heavy-oil recovery. Energy Fuels 2012;26:2911–9. [20] Jada A, Salou A. Effects of the asphaltene and resin contents of the bitumens on the water–bitumen interface properties. J Petrol Sci Eng 2002;33:185–93.
293
[21] Rudin J, Wasan DT. Mechanisms for lowering of interfacial tension in alkali/ acidic oil systems 1, experimental studies. Colloid Surf A 1992;68:67–79. [22] Rudin J, Wasan DT. Mechanisms for lowering of interfacial tension in alkali/ acidic oil systems 2, theoretical studies. Colloid Surf A 1992;68:81–94. [23] Rudin J, Wasan DT. Surfactant-enhanced alkaline flooding: buffering at intermediate alkaline pH. SPE Reserv Eval Eng 1993;18:275–80. [24] Arhuoma M, Dong M, Yang D, Idem R. Determination of water-in-oil emulsion viscosity in porous media. Ind Eng Chem Res 2009;48:7092–102. [25] Wang J, Dong M, Arhuoma M. Experimental and numerical study of improving heavy oil recovery by alkaline flooding in sandpacks. J Can Petrol Technol 2010;49:51–8.
Contents lists available at SciVerse ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Potential of alkaline flooding to enhance heavy oil recovery through water-in-oil emulsification Haihua Pei a, Guicai Zhang a,b,⇑, Jijiang Ge a, Luchao Jin a, Chao Ma a a b
College of Petroleum Engineering, China University of Petroleum, Qingdao 266555, People’s Republic of China State Key Laboratory of Heavy Oil Processing, China University of Petroleum, Qingdao 266555, People’s Republic of China
h i g h l i g h t s " Alkaline flooding can improve sweep efficiency through water-in-oil emulsification. " Improvement of sweep efficiency increases with the alkaline concentration. " There is an optimum slug size that results in the highest tertiary oil recovery. " The injection pattern is recommended to the continuous alkaline injection. " There is an optimum injection rate that obtains the highest tertiary oil recovery.
a r t i c l e
i n f o
Article history: Received 23 September 2011 Received in revised form 12 June 2012 Accepted 14 August 2012 Available online 7 September 2012 Keywords: Alkaline flooding W/O emulsion Sweep efficiency Sandpack flood test Enhanced heavy oil recovery
a b s t r a c t Alkaline flooding has great potential for enhancing the recovery of heavy oil, especially for reservoirs in which thermal methods are not suitable. In this study, alkaline flooding tests were performed in micromodels and sandpacks to investigate the microscopic displacement mechanisms for enhancing heavy oil recovery and the effect of the injection parameters on displacement efficiency. The micromodel tests indicate that the penetration of the alkaline solution into the crude oil and the subsequent formation of a water-in-oil (W/O) emulsion reduce the mobility of the water phase and divert the injected water into the unswept region, thereby improving the sweep efficiency. The sandpack flood results show that the tertiary oil recovery can reach about 20% of the initial oil in place (IOIP) using 1.0% NaOH, and the tertiary oil recovery has been found to increase as the alkaline concentration increases. However, there is an optimum slug size and injection rate at which the highest tertiary oil recovery can be obtained during the alkaline flooding process. Continuous alkaline injection can provide a higher tertiary oil recovery compared with a cyclic alkaline injection pattern. These results show that the alkaline flooding, if properly designed and controlled, can lead to enhanced heavy oil recovery through the water-in-oil emulsification. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Many countries, notably Venezuela, Canada, the United States and China, possess abundant heavy oil resources [1]. With the depletion of light oil resources and rising energy demands, the successful recovery of heavy oil is becoming increasingly important. However, the high viscosity of heavy oil makes it difficult to recover. Currently, only about 3–10% of the initial oil in place (IOIP) in heavy oil reservoirs can be recovered under primary production [2]. In order to recover additional heavy oil, different fluids usually have to be injected in order to displace the oil to the production ⇑ Corresponding author at: College of Petroleum Engineering, China University of Petroleum, Qingdao 266555, People’s Republic of China. Tel.: +86 53286981178. E-mail address: [email protected] (G. Zhang). 0016-2361/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2012.08.024
wells. However, mobility ratio concerns dominate the displacement of viscous oil, and most EOR processes focus on reducing the oil viscosity or improving the mobility ratio [3]. Unfortunately, many heavy oil reservoirs in China are relatively thin or deep, making these reservoirs poor candidates for expensive thermal methods. Therefore, the displacement mobility ratio in these reservoirs needs to be improved using an inexpensive process [4]. Water flooding is a common and inexpensive secondary oil recovery technique for the heavy oil reservoirs with oil viscosities ranging from 100 to 10,000 mPa s. However, the incremental recoveries by water flooding are quite low, due to the poor sweep efficiencies caused by the adverse mobility ratio between oil and water [5]. As a result, most of the residual oil at the end of water flooding is bypassed, which is greatly different from that of the residual oil trapped by capillary forces in conventional oil
285
H. Pei et al. / Fuel 104 (2013) 284–293
reservoirs [6,7]. Therefore, the key problem in heavy oil reservoir is inefficient sweep due to low mobility of the oil, not the residual oil in the swept region [8]. Previous researches have demonstrated that alkaline flooding can potentially increase the recovery of heavy oil through the mechanism of emulsification [9–11]. However, different views on the mechanism for alkaline flooding to improve heavy oil recovery are proposed. Some researchers proposed the formation of oil-inwater (O/W) emulsions as a possible recovery mechanism [12– 14]. In this application, oil is emulsified into water phase and the oil droplets either plug rock pores and give improved sweep efficiency, or are perhaps entrained along with the flowing aqueous phase. Alternatively, the formation of water-in-oil (W/O) emulsions has also been proposed as a possible recovery mechanism [15,16]. These emulsions should be more viscous than oil by itself, thus they could possibly lead to improvements in the mobility ratio and sweep efficiency of the flood [17,18]. The aforementioned researches indicate that the flow of W/O emulsion and O/W emulsion in porous media are two mechanisms of enhanced oil recovery by alkaline flooding. Generally, when alkaline concentration is low in the system with addition of surfactant, the mechanism of O/W emulsion prevails in the oil displacement process. When high alkaline concentration is applied, the mechanism of W/O emulsion is prominent. Compared with O/W emulsion, the W/O emulsion has much higher viscosity, which can effectively block water channels to improve sweep efficiency [19]. This mechanism is believed to be especially effective for heavy oil, where the sweep efficiency of water flooding is usually poor. Although alkaline flooding has been extensively implemented in light oil reservoirs, only a few alkaline flooding studies and field tests have been conducted in heavy oil reservoirs. This is mainly ascribed to the fact that multiphase flow in heavy oil reservoir is a more complicated process than that of the light oil reservoirs. Moreover, heavy oil has higher contents of natural petroleum acids [20], which resulting in the mechanism of alkaline flooding for improving oil recovery to be more complicated. Therefore, the effect of various operating parameters on the performance of alkaline flooding in heavy oil reservoirs is not well understood. This study is aimed at studying the potential of alkaline flooding and the microscopic displacement mechanisms for enhanced heavy oil recovery. The research is focused on the mechanism of improving the sweep efficiency through water-in-oil emulsification, and the effects of the injection parameters on the displacement efficiency are optimized by measuring the pressure drop and recovery efficiency in sandpack flood tests. The findings of this investigation can be utilized for optimizing various operational conditions, in order to maximize oil recovery, by injecting alkaline solution into the heavy oil reservoirs.
2. Experimental
The heavy oil has a viscosity of 2000 mPa s at 55 °C. The acid number value of the oil is 2.69 mg KOH/gram of sample. The formation brine has a salinity of 0.5%, and the concentrations of Ca2+ and Mg2+ in the brine are relatively low. Thus, all of the solutions used in this study were prepared with NaCl solutions with concentrations of 5000 mg/L. The alkaline agent used in this study was sodium hydroxide (NaOH). 2.2. Measurements of interfacial tension The interfacial tensions (IFTs) between the oil and the different alkaline solution systems were measured using a Model Texas-500 spinning drop interfacial tensiometer at 55 °C. In the cases where alkaline were added to the water phase, the rotation speed was sufficiently high that the length of the oil drop is larger than four times its diameter and the oil–water IFT was determined with an image–capture device and image-acquisition software according to the following equation:
r ¼ 1:2336ðqw qo Þx2
3 D ; n
L P4 D
ð1Þ
where r is the oil–water interfacial tension in mN/m, qw and qo is the density of water phases and oil phase in g/cm3, respectively, x is the rotational velocity in rpm, D is the diameter of the oil drop in 104 m, L is the length of the oil drop in 104 m, n is the refractive index of water phase. The values of the IFT determinations were repeatable within ±0.005 mN/m. 2.3. Micromodel flood studies A glass-etched micromodel was used to investigate the displacement mechanisms of alkaline flooding. The micromodel was constructed by etching a two-dimensional network of pores and throats on glass plates using a photochemical method. The pore network used in this study was patterned, based on the pore structure of a core obtained from the reservoir. The transparent nature of the micromodel allowed pore-scale multiphase displacements to be visually observed. To observe these phenomena easily during the flooding, 0.05% eosin was added to color the injected brine in the micromodel flood tests. The procedure for the micromodel test was as follows: after being vacuumed, the micromodel was prepared for flooding by filling it with brine, and the brine was subsequently displaced by crude oil until no more water was produced. Because the viscosity of the crude oil was higher than the resident water, almost all of the water present in the micromodel was displaced. After the heavy oil was saturated, the model was aged for 24 h. Next, brine with the same salinity as the alkaline solution was injected into the micromodel at a constant flow rate of 0.003 mL/min. Using a video recorder and camera apparatus, the micromodel flood was visualized during the different stages of fluid injection.
2.1. Fluids and chemicals Oil and formation brine samples were collected from the Binnan heavy oil reservoir in the Shengli oilfield in China. To remove the solids and water, the heavy oil was centrifuged at 10,000 rpm at reservoir temperature (55 °C) for 4 h. The viscosity, density, and acid number of the oil were analyzed and are listed in Table 1.
Table 1 Basic properties of the heavy oil sample. Density @ 55 °C (kg/m3)
Viscosity @ 55 °C (mPa s)
Acid number (mg KOH/g oil)
Resin (wt.%)
Asphaltene (wt.%)
947.2
2000
2.69
19.5
2.033
2.4. Sandpack flood studies All chemical flooding tests for heavy oil recovery were carried out using sandpacks. The sandpack used in this study was 2.35 cm in diameter and 19.4 cm in length. For each sandpack test, fresh quartz sand was wet-packed to ensure the same wettability for all the tests. The sandpack was packed as follows: fresh quartz sands with the size fractions of 80–100 and 100–200 meshes obtained from sieve screening were blended at a fixed weight ratio of 3:1. A coreholder filled with formation brine was positioned vertically, and the sand was added in several increments to fill the coreholder. In each step, the sand was shaken slightly after being poured in. During this process, the water surface was kept above
286
H. Pei et al. / Fuel 104 (2013) 284–293
3. Results and discussion
rium value of the dynamic IFT decreases first and subsequently increases with increasing alkaline concentration in the brine. When the NaOH concentration is 0.2%, the equilibrium value of the dynamic IFT reaches the minimum. This is ascribed that the IFT is influenced by pH in alkali/acidic oil systems. Rudin and Wasan [21–23] reported that pH is important for producing the IFT minimum because the un-ionized acid in the acidic oil contributes to the IFT in conjunction with the ionized acid and the extent of acid ionization is governed by pH. An optimal pH exists where IFT is minimized. In the limit of low pH, the IFT corresponds only to adsorption of un-ionized acid on the interface. In the limit of high pH, most of the un-ionized acid has been ionized and the IFT corresponds to adsorption of mostly ionized acid. As pH is increased, IFT decreases because of the ionized acid increase until CMC is reached. Then, the ionized acid remains constant and the un-ionized acid continues to decrease, resulting in an increase in IFT until most of the un-ionized acid is converted. At this point, IFT remains fairly constant with further pH increase.
3.1. IFT behavior of Binnan oil/alkaline systems
3.2. Micromodel flood study
To investigate the effectiveness of the alkaline solution in reducing the oil–water interfacial tension, the interfacial tension behavior of the Binnan oil/brine was studied with different NaOH concentrations ranging from 0.1% to 1.0%. Fig. 1 shows the dynamic IFT curves between the oil and brine with different NaOH concentrations. The results indicate that the dynamic IFT value between the oil and NaOH system varies in the range of 0.03–0.12 mN/m. From the dynamic IFT curves, it is also observed that the oil–water IFT has obvious dynamic effect. These curves can be divided into three stages: in the first stage, the IFT shows a transient decrease, and the lowest IFT occurs in this stage; in the second stage, the IFT increases sharply; and in the third stage, the IFT becomes constant. The dynamic effect of the IFT is related to the interfacial interaction between the alkaline solution and the heavy oil phase. When an acidic crude oil contacts an alkali in the alkaline flooding, the acidic components in the crude oil react with the alkali in the flooding water to produce in situ surfactant which lowers the oil–water IFT. Such surfactants accumulate at the oil–water interface and then, some of them gradually diffuse into the bulk aqueous phase and/or oil phase, depending on their affinities for the aqueous and oil phases. Thus, IFT between the oil phase and alkaline solution changes with time and finally reaches an equilibrium value. As shown in Fig. 1, it takes 1 h for the IFT between each oil drop and alkaline solution to reach an equilibrium value. The equilib-
To investigate the displacement mechanisms of alkaline flooding for improving heavy oil recovery, it is necessary to examine the distribution of the residual oil after water flooding. A micromodel flooding test with 0.5% NaCl was first conducted; images of the oil distribution during the water flooding process are shown in Fig. 2. Due to the glass-etched micromodel is a water-wet quarter 5-spot model and the injecting well and producing well are located on the diagonal line of the model, thus the injected water creeps along the pore wall (see Fig. 2a) and fingers through the quarter 5-spot model. As a result, the injected water advances diagonally across the micromodel (see Fig. 2b), and several connected water channels are created diagonally after water breaks through (see Fig. 2c). After that, little oil can be recovered by continued water flooding. Consequently, most of the oil is bypassed, due to the viscous fingering caused by the adverse mobility ratio between the oil and the water. The poor sweep efficiency of the water flooding leaves most of the micromodel area untouched (see Fig. 2d). In another micromodel flood test, an alkaline slug consisting of 1.0% NaOH + 0.5% NaCl was injected, and the microscopic images of oil distribution during alkaline flooding process are shown in Fig. 3. It is observed that the alkaline solution penetrates the heavy oil and creates large water drops that are coated with thin oil films (see Fig. 3a). After entering the pore space, these large water drops divide into small discontinuous water droplets inside the oil phase, as shown in Fig. 3b. As additional water droplets appear in the oil phase, a W/O emulsion bank eventually forms (see Fig. 3c). Fig. 3e shows the oil distribution of the entire micromodel when the alkaline solution reaches the outlet. At this point, a relatively uniform degree oil saturation is distributed over the entire model. This is because the viscosity of a W/O emulsion is much higher than the viscosity of the water phase and even higher than the viscosity of the oil phase [24], which lead to the significant increase of the resistance to flow of the alkaline solution in the porous media. As a result, oil is subsequently displaced in the form of the W/O emulsion with little viscous fingering. Therefore, it is the W/O emulsion that reduces the mobility of water phase and diverts the injected alkaline solution to the unswept region of micromodel to improve the sweep efficiency. This is the mechanism that will be studied in alkaline flooding for heavy oil to improve sweep efficiency and thus enhanced oil recovery. Although the high viscosity of the W/O emulsion can increase the resistance to flow of the water phase in the high permeability zone and thus leads to improved sweep efficiency, the high viscosity also causes poor mobility. As a result, a number of residual W/O
the top of the sand to ensure that air was not introduced into the sample. The sandpack flooding was conducted horizontally. The experimental procedure was briefly described as follows. After the permeability of the sandpack in the presence of the formation brine was measured, the wet-packed sandpack was subsequently saturated with the heavy oil. The oil injection was continued until water production almost ceased (the water cut was less than 1%). After the oil injection, the sandpack was waterflooded until the oil production became negligible (oil cut was less than 1%). For the tertiary chemical flood tests, 0.25–2.5 pore volume (PV) chemical slugs were injected. The chemical injection was followed by extended water flooding until the oil production became negligible. All of these tests were conducted at the reservoir temperature (55 °C), except where otherwise specified.
Fig. 1. Dynamic IFT curves between the Binnan oil and brine with different alkaline concentrations.
H. Pei et al. / Fuel 104 (2013) 284–293
287
Fig. 2. Microscopic images of oil distribution during the water flooding process. Images: (a) 0.10 PV brine injected; (b) 0.50 PV brine injected; (c) water breakthrough (1.0 PV brine injected); and (d) end of water flooding (10.0 PV brine injected).
emulsions remain after alkaline flooding, as shown in Fig. 3d, which lowers the displacement efficiency of the alkaline flooding in the swept area. Fig. 3f shows that a certain amount of residual oil is left throughout the model. Thus, the W/O emulsions with high viscosity have also negative influence on the recovery of heavy oil. 3.3. Sandpack flood study To evaluate the potential of alkaline flooding for enhanced heavy oil recovery through water-in-oil emulsification, 16 flood tests were conducted in sandpacks. The effects of alkaline concentrations, slug size, injection pattern, and injection rate on alkaline flooding were investigated in these tests. The parameters of the sandpacks, chemical slug compositions, and flood results are summarized in Table 2. For each sandpack flood test, the oil recovery behavior, water cut and pressure drop were monitored and analyzed. It should be noted that an extensive de-emulsification process was needed to accurately measure the volume of oil and water when the emulsions were produced. 3.3.1. Effect of alkaline concentration In order to examine the effectiveness of alkaline concentration in creating the in situ emulsification in the alkaline flooding for the Binnan heavy oil, a series of sandpack flood tests (Runs 1–6) were conducted with NaOH concentrations ranging from 0.1% to 1.0%. In all of these tests, alkaline solutions with slug size of 0.5 PV were injected at a rate of 0.5 ml/min after the initial water flooding. Fig. 4 illustrates the tertiary oil recovery as a function of alkaline concentration. It is observed that the tests with low alkaline concentration (0.1% in Run 1 and 0.2% in Run 2) have a small increase of oil recovery. The incremental oil recovery increases with the alkaline concentration, and it exhibits a significant
change when the alkaline concentration ranges from 0.1% to 0.4%; above this value, the increase becomes slight. The produced emulsions by the alkaline flooding in the sandpack flooding tests were examined by observing the droplet distribution and checking the droplet size under microscope. And ScopeImage Plus software was used to determine the droplet size distribution for all emulsions. Fig. 5 shows photomicrographs of droplet distribution of the produced emulsions in the sandpack flooding tests with different NaOH concentration (Runs 1–6). It is observed that the transparent water phase is in the form of dispersed droplet, which indicates that the emulsions are W/O type. Emulsions with lower alkaline concentration, such as the 0.1% and 0.2% NaOH, have smaller droplets and fewer large droplets (see Fig. 5a and b) compared to other emulsions with higher alkaline concentration, as shown in Fig. 5e and f. This means that the percentage of the large droplets in an emulsion increases with increasing alkaline concentration and vice versa for the small droplets. The droplet size distribution of the water droplet for the produced emulsions of 0.1%, 0.2%, 0.4%, 0.6%, 0.8%, and 1.0% NaOH were measured and are shown in Fig. 6. The droplet size distribution for the produced emulsion of 0.1% NaOH had a droplet size ranging from 0 to 2.652 lm, as illustrated in Fig. 6a, whereas Fig. 6f shows the droplet size distribution for the produced emulsion of 1.0% NaOH, in which the droplet size ranges from 0 to 6.857 lm. The average droplet sizes for the produced emulsions of 0.1%, 0.2%, 0.4%, 0.6%, 0.8%, and 1.0% NaOH were determined to be 1.012, 1.354, 1.637, 1.864, 2.016, and 2.785 lm, respectively. The volume fraction of dispersed water phase is defined as emulsion quality. Based on the microscope observations, it is found that the emulsion quality and droplet size increase with increasing alkaline concentration. In the literature, Arhuoma et al. [24] have carried out studies for determining viscosities of W/O emulsions in porous media. They
288
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 3. Pore-level images of the formation of W/O emulsions during the alkaline flooding process in a micromodel test. Injected alkalis: 1.0% NaOH + 0.5% NaCl. (a) Penetration into heavy oil by alkaline solutions; (b) formation of water drops inside the oil phase; (c) formation of W/O emulsion bank; (d) residual W/O emulsions after the alkaline flooding; (e) the alkali front at the outlet of the micromodel; and (f) after alkaline flooding.
Table 2 Summary of sandpack flood tests. Test #
Porosity (%)
Permeability (mD)
Initial oil saturation (%)
Waterflood recovery (%)
Chemical formula
Injection rate (ml/min)
Slug size (PV)
Tertiary recovery (%IOIP)
Final recovery (%IOIP)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
43.85 42.19 44.23 42.78 43.50 43.85 44.45 43.40 42.78 43.50 44.45 43.74 43.26 43.85 43.74 43.76
1964 2259 2260 2193 2193 2015 2071 2260 2193 1962 2130 2130 2130 2015 2071 1912
89.01 90.14 90.05 90.28 89.43 90.14 88.24 89.04 88.89 88.80 88.24 90.75 89.29 90.79 90.11 90.03
33.87 32.81 32.24 31.69 32.19 31.76 32.39 35.08 35.94 32.00 34.85 33.91 35.69 35.52 33.71 34.03
0.1% 0.2% 0.4% 0.6% 0.8% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0% 1.0%
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.1 0.25 0.75 1.0
0.5 0.5 0.5 0.5 0.5 0.5 0.25 1.0 2.0 2.5 0.5 + 0.5 0.25 + 0.25 + 0.25 + 0.25 0.5 0.5 0.5 0.5
6.29 10.63 16.04 17.08 18.65 19.96 17.88 19.30 14.25 11.29 18.21 12.18 19.17 20.90 18.00 16.42
40.16 43.44 48.28 48.77 49.84 51.72 50.27 54.38 50.19 43.29 53.06 46.09 54.86 56.42 52.71 50.45
NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH NaOH
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 4. Effect of alkaline concentration on tertiary oil recovery.
found that the viscosity of a W/O emulsion increased with increasing emulsion quality. Based on their experimental results, Wang et al. [25] derived the following correlation for the effective viscosity of a W/O emulsion in sandpacks:
logðle Þ ¼ kh þ logðlo Þ
ð2Þ
where le is the effective viscosity of the W/O emulsion, lo is the oil phase viscosity, and h is the emulsion quality. The value of the coefficient k is larger than unity and depends on the porosity and permeability of the sandpack. Besides, it is also confirmed through microscope observations (see Fig. 5) that the alkaline solution with higher concentration can create more water droplets in the effluent emulsion. As a result, the effluent emulsion quality increases with increasing alkaline concentration. Based on the Eq. (2), it is concluded that the effective viscosity of the W/O emulsion increases with increasing alkaline concentration. Thus, the viscosity of a W/
289
O emulsion becomes higher than the viscosity of the water phase and even higher than the viscosity of the oil phase. Because of the high viscosity of W/O emulsions, the resistance to water flow in the high water saturation zone can be increased significantly, which is also evident by the higher increments in the pressure drop. Therefore, it can be seen that the pressure drop along the sandpack increases with increasing NaOH concentrations, as shown in Fig. 7. From these figures, it can be concluded that for the same injection rate, the increase in oil recovery is accompanied by an increase in pressure drop, and a higher peak value of pressure drop results in a higher tertiary oil recovery. The built-up pressure drop suggests the penetration of alkaline solution into residual oil drops to form high viscosity W/O emulsions, which can block the high-permeability water channels and reduce the mobility of the water phase. As a result, the subsequent injected water is diverted to the unswept areas, thereby resulting in improved sweep efficiency and increased heavy oil recovery. On the basis of the sandpack flood results, 1.0% NaOH was found to be notably effective for enhanced heavy oil recovery by the in situ generated W/O emulsions. Based on these results, 1.0% NaOH concentration solutions were used in additional sandpack flood tests to investigate the effects of other parameters in the alkaline flooding scheme.
3.3.2. Effect of alkaline slug size Determining the smallest chemical slug size that is able to recover the maximum residual oil is an important factor in the optimization of the proposed EOR process. To evaluate the effect of alkaline slug size on tertiary oil recovery, a series of sandpack flood tests (Runs 6–10) were performed by increasing the alkaline slug size incrementally from 0.25 to 2.5 PV while keeping the NaOH concentration and the injection rate at 1.0% and 0.5 ml/min, respectively. The results of the tertiary oil recovery are plotted as a function of slug size in Fig. 8. The tertiary oil recovery first increases with increasing alkaline slug size and later decreases, and it reaches the maximum value at
Fig. 5. Photomicrograph of droplet distribution of the produced emulsion in the sandpack flooding tests: (a) 0.1% NaOH, (b) 0.2% NaOH, (c) 0.4% NaOH, (d) 0.6% NaOH, (e) 0.8% NaOH and (f) 1.0% NaOH.
290
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 6. Droplet size distribution of the produced emulsion in the sandpack flooding tests: (a) 0.1% NaOH, (b) 0.2% NaOH, (c) 0.4% NaOH, (d) 0.6% NaOH, (e) 0.8% NaOH and (f) 1.0% NaOH.
0.5 PV. It is indicated that a larger alkaline slug size does not correspond to higher incremental oil recovery and that there is an optimum slug size that results in the highest tertiary oil recovery. This result can be explained on the basis of the production of the in situ W/O emulsion. The W/O emulsion has both positive and
negative effects on enhancing oil recovery: on one hand, its higher viscosity plays a key role in the effective blockage of the water channel to divert the injected water to the unswept areas and increase the sweep efficiency; on the other hand, it is difficult for this emulsion to move because of its higher viscosity; therefore, there is
H. Pei et al. / Fuel 104 (2013) 284–293
291
meet a demand for effectively blocking the water channels caused by the initial water flooding. The subsequent water will be diverted to other oil-rich areas, leading to a sharp increase of oil recovery. However, when the alkaline slug size increases from 1.0 to 2.5 PV, the excessive alkaline solutions will not only form W/O emulsions in the water channels, but also they will advance into other oil-rich areas and produce lots of high viscosity W/O emulsions in these areas. Although some of W/O emulsions can block the preformed water channels, the negative influence caused by the excessive viscous W/O emulsions on poor mobility is strengthened, and therefore, the tertiary oil recovery decreases with an increasing alkaline slug size. On the basis of these results, a slug size of 0.5 PV is considered as the optimum for the above chemical formula and enhancing the recovery of Binnan heavy oil in homogeneous sandpacks.
Fig. 7. Pressure drop response as a function of fluid injected in the sandpack flooding tests.
Fig. 8. Effect of slug size on tertiary oil recovery.
always a certain amount of residual oil left after alkaline flooding (as seen in Fig. 3d and f). Therefore, when the alkaline slug size increases from 0.25 to 0.5 PV, the amount of W/O emulsions can only
Fig. 9. Cumulative oil recovery of three flood test (Runs 8, 11, 12) with different injection pattern.
3.3.3. Effect of injection pattern To investigate the effect of injection pattern on tertiary oil recovery by alkaline flooding, three sandpack flooding tests (Runs 8, 11, 12) were carried out using different injection patterns. The alkaline concentration and the injection rate were set at 1.0% and 0.5 ml/min, respectively. The total injection volume was kept constant at 1.0 PV in all of these tests. After the initial water flooding, the alkaline solution was injected using different patterns. In Run 8, a slug of 1.0 PV NaOH solution was injected continuously and was followed by an extended water flooding. Runs 11 and 12 used a cyclic alkaline injection; the slug size of each injection was 0.5 PV and 0.25 PV, respectively. The results for all three runs are presented in Fig. 9. Although the total volume of the alkaline slug remains the same, the incremental oil recovery gradually decreases with the number of slugs. This indicates that the continuous alkaline injection can obtain higher tertiary oil recovery than the cyclic alkaline injection. Pressure drop responses as a function of fluid injected in these three runs are presented in Fig. 10. The test with the continuous alkaline injection pattern (Run 8) has the highest pressure drop and the broadest pressure hump during the alkaline injection. For the cyclic alkaline injection pattern with an injection of 0.25 PV, the lowest and thinnest pressure humps appear in the first slug, and almost no peak injection pressure appears in the following three slugs. This result is because more W/O emulsions are produced with the continuous alkaline injection pattern, and the flow resistance of injected water increases significantly, which leads to a long-term plugging effect. For the cyclic alkaline injection patterns, in contrast, the high concentrations of the alkaline solutions
Fig. 10. Effect of injection pattern on pressure drop as a function of fluid injected.
292
H. Pei et al. / Fuel 104 (2013) 284–293
Fig. 11. Effect of alkaline injection rate on tertiary oil recovery.
are more likely diluted, and consequently, fewer W/O emulsions are produced. Moreover, the W/O emulsion bank is easily broken down during the alternating process, which leads to a short-term plugging effect. Therefore, considering alkaline concentration loss in cyclic injection, the injection pattern is recommended to select the continuous alkaline injection for alkaline flooding.
3.3.4. Effect of injection rate In order to examine the effect of alkaline injection rate on the tertiary oil recovery by alkaline flooding, additional sandpack flood tests (Runs 13–16 and Run 6) were carried out. The injection rate of the alkaline slug was increased incrementally from 0.1 ml/min to 1.0 ml/min. The NaOH concentration was kept constant at 1.0%, and the alkaline slug size was fixed at 0.5 PV. The incremental oil recovery as a function of alkaline injection rate is plotted in Fig. 11. The incremental oil first increases with the alkaline injection rate and later decreases, and it reaches the maximum value at 0.25 ml/min. This is because the driving force is not high enough to produce many W/O emulsions when the injection rate is low (0.1 ml/min in Run 13). However, higher injection rates also reduce the contact area and contact time between the injected alkaline solution and the oil in the sandpack such that the alkaline solution cannot fully react with the acidic components in the heavy oil; therefore, the number of W/O emulsions generated by the reaction between the alkali and the oil sharply decreases. Moreover, a higher injection rate can also cause the W/O emulsion bank to break down quickly, resulting in early water breakthrough. All these factors could contribute to a lower sweep efficiency and lower tertiary oil recovery. Therefore, there is an optimum injection rate at which enough W/O emulsions can be produced to block the water channels effectively and increase the flow resistance. This is consistent with the results of the flooding tests conducted by Cooke et al. [10] and Dong et al. [13], in which the highest incremental oil recovery is obtained at the optimum injection rate.
4. Conclusion A series of micromodel flood tests and sandpack flood tests were performed to evaluate the potential of alkaline flooding to enhance the recovery of Binnan heavy oil through water-in-oil emulsification. The results indicate that
(1) The mechanism of water-in-oil emulsification plays a prominent role in the alkaline flooding process for enhanced heavy oil recovery. Initially, the alkaline solution can penetrate the heavy oil to form water drops inside the oil phase, and W/O emulsions are produced. Because of high viscosity of the W/O emulsions, the high permeability zone caused by the preceding water flooding can be blocked, and the subsequent injected water phase is diverted to the unswept regions, thereby resulting in improved sweep efficiency and increased heavy oil recovery. (2) The water droplets number and droplets size in W/O emulsions increases with the alkaline concentration increasing, and a higher alkaline concentration results in a higher tertiary oil recovery. However, there is an optimum slug size that gives the highest tertiary oil recovery. (3) When the W/O emulsions displace the residual oil from the waterflood, continuous alkaline injection can give higher tertiary oil recovery compared with the cyclic alkaline injection pattern. The effectiveness of alkaline flooding depends on the injection rate. The highest incremental oil recovery is obtained at the optimum injection rate. (4) The alkaline flooding method is a promising EOR process for enhanced heavy oil recovery through the water-in-oil emulsification mechanism.
Acknowledgements Financial support by the National Natural Science Foundation of China (Grant 51104170), the Fundamental Research Funds for the Central Universities (Grant 12CX06024A), the Outstanding Doctoral Dissertation Training Program of China University of Petroleum (Grant LW110203A), and the Fok Ying Tung Education Foundation for Young Teachers in the Higher Education Institutions of China (Grant 114016) is gratefully acknowledged.
References [1] Selby R, Alikhan AA, Farouq Ali SM. Potential of non-thermal methods for heavy oil recovery. J Can Petrol Technol 1989;28:45–59. [2] Mai A, Bryan J, Goodarzi N, Kantzas A. Insights into non-thermal recovery of heavy oil. J Can Petrol Technol 2009;48:27–35. [3] Farouq Ali SM. Prospects of EOR techniques in Saskatchewan oil reservoirs. J Can Petrol Technol 1986;25:64–7. [4] Farouq Ali SM, Thomas S. The promise and problems of enhanced oil recovery methods. J Can Petrol Technol 1996;35:59–63. [5] Mai A, Kantzas A. Heavy oil water flooding: effects of flow rate and oil viscosity. J Can Petrol Technol 2009;48:42–51. [6] Bryan J, Kantzas A. Enhanced heavy-oil recovery by alkali–surfactant flooding. In: SPE Paper 110738-MS. SPE Annual technical conference and exhibition, Anaheim, California, U.S.A; 11–14 November, 2007. [7] Bryan J, Kantzas A. Investigation into the processes responsible for heavy oil recovery by alkali-surfactant flooding. In: SPE Paper 113993-MS. SPE/DOE Improved oil recovery symposium, Tulsa, Oklahoma, U.S.A; 19–23 April, 2008. [8] Bryan J, Kantzas A. Improved recovery potential in mature heavy oil fields by alkali–surfactant flooding. In: SPE Paper 117649-MS. SPE International thermal operations and heavy oil symposium, Calgary, Alberta, Canada; 20–23 October 2008. [9] Jennings Jr HY, Johnson Jr CE, Mcaliliffe CD. A caustic water flooding process for heavy oils. J Petrol Technol 1974;26:1344–52. [10] Cooke Jr CE, Williams RE, Kolodzie PA. Oil recovery by alkaline water flooding. J Petrol Technol 1974;26:1365–74. [11] Johnson Jr CE. Status of caustic and emulsion methods. J Petrol Technol 1976;28:85–92. [12] Liu Q, Dong M, Yue X, Hou J. Synergy of alkali and surfactant in emulsification of heavy oil in brine. Colloid Surf A 2006;273:219–28. [13] Dong M, Ma S, Liu Q. Enhanced heavy oil recovery through interfacial instability: a study of chemical flooding for Brintnell heavy oil. Fuel 2009;88:1049–56. [14] Ma S, Dong M, Li Z. Evaluation of the effectiveness of chemical flooding using heterogeneous sandpack flood test. J Petrol Sci Eng 2007;55:294–300. [15] Dong M, Liu Q, Li A. Displacement mechanisms of enhanced heavy oil recovery by alkaline flooding. Particuology 2012;10:298–305.
H. Pei et al. / Fuel 104 (2013) 284–293 [16] Zhang H, Dong M, Zhao S. Which one is more important in chemical flooding for enhanced court heavy oil recovery, lowering interfacial tension or reducing water mobility. Energy Fuels 2010;24:1829–36. [17] Ding B, Zhang G, Ge J. Research on mechanisms of alkaline flooding for heavy oil. Energy Fuels 2010;24:6346–52. [18] Pei H, Zhang G, Ge J. Analysis of microscopic displacement mechanisms of alkaline flooding for enhanced heavy oil recovery. Energy Fuels 2011;25:4423–9. [19] Pei H, Zhang G, Ge J. Comparative effectiveness of alkaline flooding and alkaline–surfactant flooding for improved heavy-oil recovery. Energy Fuels 2012;26:2911–9. [20] Jada A, Salou A. Effects of the asphaltene and resin contents of the bitumens on the water–bitumen interface properties. J Petrol Sci Eng 2002;33:185–93.
293
[21] Rudin J, Wasan DT. Mechanisms for lowering of interfacial tension in alkali/ acidic oil systems 1, experimental studies. Colloid Surf A 1992;68:67–79. [22] Rudin J, Wasan DT. Mechanisms for lowering of interfacial tension in alkali/ acidic oil systems 2, theoretical studies. Colloid Surf A 1992;68:81–94. [23] Rudin J, Wasan DT. Surfactant-enhanced alkaline flooding: buffering at intermediate alkaline pH. SPE Reserv Eval Eng 1993;18:275–80. [24] Arhuoma M, Dong M, Yang D, Idem R. Determination of water-in-oil emulsion viscosity in porous media. Ind Eng Chem Res 2009;48:7092–102. [25] Wang J, Dong M, Arhuoma M. Experimental and numerical study of improving heavy oil recovery by alkaline flooding in sandpacks. J Can Petrol Technol 2010;49:51–8.