waterflooding in the thin post-CHOPS reservoirs

waterflooding in the thin post-CHOPS reservoirs

Fuel 231 (2018) 507–514 Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Full Length Article CO2-cycl...

967KB Sizes 0 Downloads 8 Views

Fuel 231 (2018) 507–514

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

CO2-cyclic solvent injection (CO2-CSI) and gas/waterflooding in the thin post-CHOPS reservoirs

T



Hongze Ma, Yongan Gu

Petroleum Technology Research Centre (PTRC), Petroleum Systems Engineering, Faculty of Engineering and Applied Science, University of Regina, Regina, Saskatchewan S4S 0A2, Canada

G R A P H I C A L A B S T R A C T 1000

15

3000

qo

3 qo (cm /h)

600

400 5 200

0

0 0

50

100

150

iGOR (sc cm3/cm3)

10

2500

800

2000

1500

1000

ΔP or Ppro (kPa)

GOR ΔP Ppro

500

0

200

Time (min)

A R T I C LE I N FO

A B S T R A C T

Keywords: CO2-cyclic solvent injection (CO2-CSI) Gasflooding Waterflooding Foamy-oil flow and solution-gas drive Cold heavy oil production with sand (CHOPS) reservoirs

In this paper, a new combined CO2-cyclic solvent injection (CO2-CSI) and gasflooding (GF) process is proposed as an effective enhanced oil recovery (EOR) process in the post cold heavy oil production with sand (CHOPS) reservoirs. The synergy of the CO2-CSI and GF is explored and compared with that of the CO2-CSI and waterflooding (WF). A total of ten sandpacked laboratory tests were conducted to study and compare the production performances of the WF, CO2-CSI, combined CO2-CSI and GF/WF after the primary production. In the combined CO2-CSI and GF/WF processes, CO2/water is injected post, at the same time with, and prior to the CO2-CSI in the CSI + GF/WF, simultaneous (CSI + GF/WF), and GF/WF + CSI, respectively. Also the effect of CO2 injection rate is studied by injecting CO2 at three different flow rates in the CSI + GF. It is found that CO2-CSI + GF or CO2-CSI + WF performs better than the CO2-CSI or WF alone due to the extended foamy-oil flow. The combined CO2-CSI and WF outperforms the combined CO2-CSI and GF in terms of the heavy oil recovery factor (RF), heavy oil production rate, and cumulative gas-oil ratio (GOR). CO2 channeling is hindered by the subsequently injected water. In the combined CO2-CSI and GF, however, strong free-gas flow adversely affects the foamy-oil production in the later cycles of the CSI and in the subsequent GF. The best fluid injection timing in terms of the heavy oil RF is to inject CO2/water immediately after each cycle of the CSI production. A moderate CO2 injection rate gives the highest heavy oil RF of the CSI + GF. In conclusion, the combined CO2-CSI and GF/WF process is capable of recovering the remaining oil in the post-CHOPS reservoir.

1. Introduction The Canadian heavy oil reserves are estimated to be 55 billion ⁎

Corresponding author. E-mail address: [email protected] (Y. Gu).

https://doi.org/10.1016/j.fuel.2018.05.100 Received 18 August 2017; Received in revised form 23 March 2018; Accepted 19 May 2018 0016-2361/ © 2018 Elsevier Ltd. All rights reserved.

barrels, almost 2/3 of which are located in Saskatchewan. Of the provincial proven heavy oil reserves, 97% have less than 10-m main pay zones (MPZs) and 55% have less than 5-m MPZs [1]. Cold heavy oil

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

the gas mobility and strengthen the foamy-oil flow by injecting a foaming agent [18]. The heavy oil RF in the FOAM H-n-P was increased by 43.3% in comparison with the methane huff-n-puff alone. In this paper, a new hybrid EOR process, namely, the combined CO2-CSI and GF, was proposed to recover the heavy oil in the thin postCHOPS reservoirs and maximize the synergy of the two processes. It is worthwhile to emphasize that the GF can not only maintain the reservoir pressure to prevent the quick solvent release but also displace the remaining foamy oil at the end of the CSI production. A series of ten sandpacked tests were conducted to identify the optimum combined CO2-CSI and GF and compare it with the combined CO2-CSI and WF. During each test, the heavy oil RF, instantaneous GOR, instantaneous water–oil ratio (WOR), injection and production pressures were measured. The experimental results were analyzed to evaluate and compare three different production schemes: CSI + GF/WF; simultaneous (CSI + GF/WF); and GF/WF + CSI. Three CSI + GF tests with three different CO2 injection rates were performed to examine the effect of CO2 injection rate on the enhanced heavy oil recovery process.

production with sand (CHOPS) is widely used in the Western Canadian heavy oil reservoirs in the last two decades as the field-scale primary production process [2]. A typical CHOPS primary production process has a heavy oil recovery factor (RF) of 5–15% of the original oil-inplace (OOIP) due to the foamy-oil flow and sand production, which is higher than a conventional primary production [3]. The foamy oil is formed when the nucleated small gas bubbles are dispersed and trapped in the viscous heavy oil. In the foamy-oil flow, the gas mobility is substantially controlled [4]. Also, the sand production generates some extremely high-permeability channels or wormholes to greatly increase the foamy heavy oil production rate [5]. Nevertheless, there is still 85–95% of the OOIP left in many post-CHOPS reservoirs [6]. As an economical secondary improved oil recovery (IOR) method, over 200 heavy oil waterfloods (WFs) have been applied in the Western Canada in the past 60 years [7]. Nevertheless, only 2–8% incremental heavy oil is recovered during the WFs in the post-CHOPS reservoirs because of the strong water channeling [8]. Similarly, gasflooding (GF) or continuous gas injection (CGI) also suffers from severe viscous fingering and an early gas breakthrough (BT) due to an extremely adverse gas-to-oil mobility ratio [9]. The incremental heavy oil production usually cannot offset the high costs of gas acquisition, transportation, storage, compression and injection, which make the GF alone uneconomical for most post-CHOPS reservoirs [10]. Thermal-based enhanced oil recovery (EOR) methods, such as cyclic steam stimulation (CSS) and steam-assisted gravity drainage (SAGD), have been commercially applied in a number of thick heavy oil reservoirs [11,12]. However, the extensive heat losses to the overburden and underburden and/or bottom-water zone make the thermal-based methods unsuitable for many thin post-CHOPS reservoirs [13]. Among the solvent-based methods, cyclic solvent injection (CSI) has been studied in the laboratories and tested in the thin post-CHOPS reservoirs recently because of its high energy efficiency and low greenhouse gas (GHG) emissions [14]. In the CSI, solvent is injected into a well to dilute the heavy oil. The well is shut in to soak for a period of time after solvent is injected and dissolved into the heavy oil. Then the same well is open and the solvent-diluted heavy oil with a much reduced viscosity is produced. The CSI is repeated in many cycles until the heavy oil production rate becomes too low. The CSI uses the solvent to reduce the heavy oil viscosity and restore the foamy-oil flow by repressurizing the heavy oil reservoir. The major technical shortcoming is that the heavy oil viscosity is regained after the solvent is released from the heavy oil during the CSI pressure-depletion production period. In addition, the foamy heavy oil in the reservoir is pushed away from the production well by the subsequently injected solvent in the next solvent injection period. An unfavourable injected gas-to-heavy oil mobility ratio also causes an early gas BT [15]. In the past, some modified CSI processes have been investigated to remedy the afore-mentioned major technical shortcomings of the traditional CSI. For example, the enhanced cyclic solvent process (ECSP) was proposed to alleviate the heavy oil viscosity regainment, in which a volatile solvent (e.g., methane) was injected prior to the injection of a soluble solvent (e.g., ethane or propane) [16]. Thus there was the solution-gas drive due to the volatile solvent dissolution and there was the heavy oil viscosity reduction due to the soluble solvent dissolution. The experimental results showed that the ECSP gave higher heavy oil production rate and RF with a lower gas-oil ratio (GOR) than the traditional CSI. On the other hand, the gasflooding-assisted cyclic solvent injection (GA-CSI) was tested by using gasflooding to push the remaining foamy oil towards the producer post the CSI production period, which also reduces the so-called back-and-forth movement of the foamy oil [17]. It should be noted that the solvent injector and the heavy oil producer were two different wells during the GA-CSI. It was found from nine sandpacked tests that an incremental 10–20% heavy oil was recovered in the GA-CSI by using propane as an extracting solvent during the CSI and as a displacing solvent during the GF. In addition, foamy oil-assisted methane huff-n-puff (FOAM H-n-P) was examined to control

2. Experimental section 2.1. Materials In this study, the original heavy oil and brine samples were collected from the Colony formation in the Bonnyville area, Alberta, Canada. The dead heavy oil density and viscosity were measured to be ρo = 0.992 g/ cm3 and µo = 33,876 cP at the atmospheric pressure of Pa = 1 atm and reservoir temperature of Tres = 21 °C, respectively. The brine density and viscosity were measured to be ρw = 1.030 g/cm3 and µw = 1.2 cP at t Pa = 1 atm and Tres = 21 °C, respectively. Methane (Praxair, Canada) with the purity of 99.97 mol% was dissolved into the dead heavy oil to reconstitute the live heavy oil at the initial reservoir pressure of Pi = 3.0 MPa and Tres = 21 °C. The corresponding GOR was equal to 9.6 sc cm3/cm3. Carbon dioxide (Praxair, Canada) with the purity of 99.998 mol% was injected as the extracting and displacing gaseous solvent in all solvent-based EOR processes. The detail compositional analysis result of the heavy oil, physicochemical properties of the brine, and PVT data of the heavy oil-CH4/CO2 system can be found elsewhere [15]. 2.2. Combined CO2-CSI and GF/WF The experimental set-up for conducting the primary production, WF, CO2-CSI, combined CO2-CSI and GF/WF is schematically shown in Fig. 1. This experimental set-up consisted of a sandpacked physical model, a fluid injection system, and a fluid production system. The combined CO2-CSI and GF/WF was carried out in a two-well configuration. The primary production with the live heavy oil and the subsequent CSI were conducted by using the injection/production well at the centre on the left-hand side of the physical model, whereas CO2/ water was injected from the injection well at the centre on the righthand side of the physical model during the GF/WF. One CO2 or brine cylinder was used to inject CO2 or brine during the combined CO2-CSI and GF/WF. The detailed experimental set-up and procedure for preparing the sandpacked physical model were described in the previous study [15]. Table 1 summarizes the major physical properties of the 2-D sandpacked physical model used in Tests #1–10. The specific production schemes of the primary production and subsequent IOR/EOR processes in Tests #1–10 are summarized in Table 1. The same production scheme of the primary production was used in each test, which started at an initial reservoir pressure of Pi = 3.0 MPa. A constant pressure drawdown rate of dP/dt = 5.0 kPa/ min was used to model a pressure-depletion process in the actual CHOPS reservoir. The primary production process was terminated when the production pressure reached Pf = 0.2 MPa. In Test #1, about 1.0 PV brine was injected at qw = 0.5 cm3/min for 600 min post the 508

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

Fluid injection unit CO2 cylinder CO2 regulator

Digital pressure indicator

Brine CO2

Physical model Producer

Syringe pump

Injector

Back-pressure regulator

Vacuum pump

Syringe pump

Oil and water collector Bubblers

Fluid production unit

Fig. 1. Schematic diagram of the experimental set-up for conducting the primary production, subsequent CSI and/or water/gasflooding in Tests #1–10 at Tres = 21 °C.

injection well was shut in for CO2 to dissolve into the heavy oil and restore the foamy-oil flow. The CSI soaking period was terminated at ts = 24 h when the reservoir pressure reached an almost constant value of 2.7 MPa. The CSI production period was commenced at Ps = 2.7 MPa and stopped at Pe = 0.2 MPa with a constant pressure drawdown rate of dP/dt = 12.5 kPa/min. Tests #1 and #2 were used as the baselines for comparison with the hybrid IOR and EOR processes. Different combined CSI and GF/WF schemes were examined in Tests #3–10 as hybrid EOR processes. In Tests #3–5 (CSI + GF), the GF (CO2-flooding) was started at the end of each CSI production period with a CO2 injection rate of qCO2 = 0.05, 0.10, 0.20 cm3/min, respectively. In Test #6 (simultaneous CSI + GF) or Test #7 (GF + CSI), the GF was conducted simultaneously with each CSI production or prior to each CSI injection. Combined CSI and WF processes were conducted in Tests #8–10 by injecting the brine at qw = 0.5 cm3/min from the injection well, similar to the combined CSI and GF processes in Tests #4, #6, and #7.

primary production. At the end of the WF, the instantaneous water-oil ratio (WOR) was already over 10:1 in volume. In Tests #2–10, CO2 was injected to repressurize the reservoir pressure to Pinj = 3.0 MPa during the CSI injection period. No more CO2 could be injected into the physical model at the end of the injection period of tinj = 40 min. Then the

Table 1 Physical properties of the 2-D sandpacked physical model, heavy oil production schemes, gas/water injection rates (qCO2 / qw ), and injection times (tinj) during the CO2 flooding or waterflooding in Tests #1–10 at Tres = 21 °C. Test No.

1 2 3 4 5 6 7 8 9 10

k (D)

5.0 4.0 5.0 4.9 3.5 4.2 3.5 4.7 4.1 4.0

ϕ (%)

38.9 38.0 37.9 38.3 38.9 38.8 39.0 38.1 38.2 38.3

Soi (%)

98.9 98.7 99.0 98.0 98.9 99.1 95.8 99.1 98.5 98.6

IOR/EOR process

WF CSI CSI + GF CSI + GF CSI + GF Simul CSI + GF GF + CSI CSI + WF Simul CSI + WF WF + CSI

GF/WF

qCO2/ qw (cm3/min)

tinj (min)

0.50 – 0.05 0.10 0.20 0.10 0.10 0.50 0.50 0.50

600 – 200 200 200 200 200 200 200 200

3. Results and discussion 3.1. Combined CSI and GF/WF This study examined three combined CSI and GF processes, i.e., 509

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

viscosity reduction and oil-swelling effect through CO2 dissolution were weakened. The heavy oil production rate, instantaneous GOR, pressure drop and production pressure during the first cycle of the GF in Test #4 (CSI + GF) is shown in Fig. 3(a). The pressure drop is the difference between the pressures measured at the two ends (the injection and production wells) of the physical model. This figure indicates that the heavy oil production rate fluctuated around 5.0 cm3/h, which was similar to that at the end of the CSI Cycle #1 in Test #4. This is because that the foamy-oil flow and free-gas flow alternate during the GF. The instantaneous GOR was about 160 sc cm3/cm3 during the GF and was higher than that near the end of the CSI Cycle #1 in Test #4. The injected CO2 caused a stronger free-gas flow during the GF. It is also found from Fig. 3(a) that the pressure drop was decreased slowly during the GF. The injected CO2 might flow along the previously formed gas channels and break through quickly. Fig. 3(b) shows the variations of the heavy oil production rate, instantaneous WOR, pressure drop and production pressure with the WF time of Cycle #1 in Test #8 (CSI + WF). The heavy oil production rate fluctuated during the WF, which is probably caused by the alternation of the foamy-oil flow and water flow. Unlike the quick gas BT in the GF of Test #4 (CSI + GF), water BT occurred when about 0.05 PV brine was injected. The WOR was increased rapidly and then decreased near the end. The heavy oil was produced at a high WOR due to the expansion of the dispersed gas bubbles and the water imbibition [21]. The pressure drop was increased prior to the water BT and then decreased to

CSI + GF (Test #4), simultaneous (CSI + GF) (Test #6), and GF + CSI (Test #7), and three combined CSI and WF processes, i.e., CSI + WF (Test #8), simultaneous (CSI + WF) (Test #9), and WF + CSI (Test #10). Their heavy oil RFs and residual oil saturations are summarized in Table 2. The measured residual oil saturations at different locations of the sandpacked model at the end of the combined CSI and WF can be found in the previous study [15]. The combined CSI and GF/WF recovers much more heavy oil than the WF alone (8.0%). Accordingly, the residual oil saturation is reduced in the combined CSI and GF/WF due to the joint microscopic displacement efficiency and macroscopic sweep efficiency. The CSI + GF/WF has higher heavy oil RF and lower residual oil saturation than the simultaneous (CSI + GF/WF) because the latter production time is shorter. The combined CSI and WF performed better than the combined CSI and GF in terms of not only the heavy oil RF and residual oil saturation but also the heavy oil production rate and cumulative GOR, as shown in Figs. 2(a–c), respectively. The mobility ratio of the injected water to heavy oil is about 60 times smaller than that of the injected CO2 to heavy oil. Thus, more heavy oil was recovered and the residual oil saturation is lower. The heavy oil was displaced into the larger pores when the injected water was imbibed into the smaller pores, which further reduces the residual oil saturation. Moreover, the injected water had also effectively controlled CO2 mobility and CO2 gravity overriding. On the other hand, the remaining heavy oil was surrounded by the injected water at the end and was difficult to access and recover due to the so-called waterblocking effect [20]. Hence, the heavy oil

Table 2 Heavy oil recovery factors (RFs) and residual oil saturations (Sor) of the primary production, subsequent CSI and/or WF/GF in Tests #1–10 at Tres = 21 °C.

1 2 3 4 5 6 7 8 9 10

Heavy oil RF (%)

20

15

Primary production

1st cycle

RF (%)

RFCSI (%)

10.0 11.4 10.1 10.7 9.7 10.3 10.8 11.2 10.1 10.9

– 6.5 5.9 6.3 5.6

2nd cycle RFGF/WF (%) – – 4.8 5.9 6.3 8.1 6.1 4.2 8.0 6.2

4.1 6.1 6.0

RFCSI (%)

– 5.7 1.1 0.8 0.6 0.3 3.8 0.7

3rd cycle RFGF/WF (%) – – 1.3 2.2 0.3 2.9 3.2 3.2 6.9 3.5

Subtotal RFCSI (%)

RFCSI (%)

– 0.7 0.6 0.4 0.8 0.3 0.8 0.7

Subtotal RFGF/WF (%)

Total RF (%)

Sor (%)

18.0 24.3 23.9 26.3 23.6 21.3 24.8 30.1 25.9 28.9

81.1 74.7 75.3 72.2 75.6 78.0 72.0 69.3 73.0 70.1

RFGF/WF (%) – – 0.05 0.03 0.3 0.02 0.03 0.8 0.9 0.9

– 12.9 7.6 7.5 7.0

8.0 – 6.2 8.1 6.9 11.0 9.3 8.2 15.8 10.6

4.7 10.7 7.4

15 RFWF RFGF RFCSI

Average oil production rate qo (cm3/h)

Test No.

RFsimul (CSI + GF) RFsimul (CSI + WF)

10

5

10

5

0

0 4

6

7

8

9

10

4

6

7

8

9

10

Test number

Test number

Fig. 2b. Measured average heavy oil production rates (qo) in the CSI + GF (Test #4), simultaneous (CSI + GF) (Test #6), GF + CSI (Test #7), CSI + WF (Test #8), simultaneous (CSI + WF) (Test #9), and WF + CSI (Test #10).

Fig. 2a. Measured heavy oil RFs in the CSI + GF (Test #4), simultaneous (CSI + GF) (Test #6), GF + CSI (Test #7), CSI + WF (Test #8), simultaneous (CSI + WF) (Test #9), and WF + CSI (Test #10). 510

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

1000

reach a plateau. The reservoir pressure was maintained at a higher pressure than that in GF Cycle #1 of Test #4 (CSI + GF).

cGOR (sc cm3/cm3)

800

3.2. Foamy-oil flow versus free-gas flow

600

As the reservoir pressure declines during the CSI, gas is nucleated to become the small gas bubbles inside the heavy oil. The dispersed gas bubbles are transported through the porous media by the heavy oil, which forms the foamy-oil flow. As the reservoir pressure is further reduced, some dispersed gas bubbles coalesce and grow large enough to become a free-gas phase at certain time. Some dispersed gas bubbles become continuous free gas that flows along the gas channels. During the subsequent GF, the remaining foamy oil is displaced by the injected gas. In this study, the produced foamy oil as shown in Fig. 4 indicates the foamy-oil flow during the GF in the CSI + GF [16–19]. The foamyoil flow was observed in some pore-scale micromodel tests under the similar test conditions [21]. The important differences between the foamy-oil flow and free-gas flow are the gas distribution and flow. In the foamy-oil flow, the small gas bubbles are formed through the nucleation at the widely distributed

400

200

0

4

6

7

8

9

10

Test number Fig. 2c. Measured cumulative GORs in the CSI + GF (Test #4), simultaneous (CSI + GF) (Test #6), GF + CSI (Test #7), CSI + WF (Test #8), simultaneous (CSI + WF) (Test #9), and WF + CSI (Test #10).

Fig. 3a. Measured heavy oil production rate (qo), instantaneous GOR, pressure drop (ΔP), and production pressure (Ppro) during the GF of Cycle #1 in Test #4 (CSI + GF).

Fig. 3b. Measured heavy oil production rate (qo), instantaneous WOR, pressure drop (ΔP), and production pressure (Ppro) during the WF of Cycle #1 in Test #8 (CSI + WF). 511

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

increased and the instantaneous GOR was decreased, which indicates a stronger foamy-oil flow. The pressure drop became larger due to the high isothermal compressibility of the dispersed gas. At a later time, the dispersed gas bubbles coalesced continuously and became the free gas at the end. Therefore, the free-gas flow became dominant again, which results in a lower heavy oil production rate and a higher instantaneous GOR. The foamy-oil flow and free-gas flow alternated during the subsequent GF, as described in the previous section. The heavy oil RFs, heavy oil production rates, and cumulative GORs of the CSI and GF in Test #4 (CSI + GF) are compared in Figs. 6(a–c). From the first cycle to the third cycle of the CSI or GF, the heavy oil RF and heavy oil production rate were decreased and the cumulative GOR was increased. The gas saturation became higher as more and more heavy oil was produced, which led to the dominant free-gas flow and more severe gas channeling. In the first cycle, the CSI and GF had almost the same heavy oil RFs and heavy oil production rates. The foamyoil flow was effective in the both processes. GF Cycle #2 gave higher heavy oil RF and heavy oil production rate and a lower cumulative GOR than CSI Cycle #2. The free-gas flow was strong from the beginning of the CSI Cycle #2. Most foamy oil was displaced during GF Cycle #2. The heavy oil RF and production rate were much lower and the cumulative GOR was far higher during GF Cycle #3. The gas channeling became so severe that most heavy oil was bypassed and thus remained untouched.

Fig. 4. Produced foamy oil in the oil collector during the first cycle of the GF in Test #4 (CSI + GF).

3.3. Effect of CO2 injection rate sites in the heavy oil reservoir. Thus, the gas bubbles can easily remain as a dispersed phase in the continuous heavy oil. In the free-gas flow, however, the gas preferentially flows as a separate phase through the low-resistance channels from the high-pressure zones to the low-pressure zones [22]. Fig. 5 shows the measured heavy oil production rate, instantaneous GOR, pressure drop and production pressure during the first cycle of the CSI production in Test #4 (CSI + GF). In the beginning of the CSI production, the oil production rate was extremely low because the foamy oil after the primary production was pushed away from the producer by the newly injected CO2. A large amount of CO2 was near the producer and the free-gas flow was dominant. Hence, the instantaneous GOR was high and the pressure drop was almost zero. As the reservoir pressure was depleted, the heavy oil production rate was

In this study, three different CO2 injection rates of qCO2 = 0.05, 0.10, 0.20 cm3/min were tested during the GF in Tests #3–5 (CSI + GF), respectively. The heavy oil RFs of the CSI + GF in Tests #3–5 are plotted and compared in Fig. 7(a). The heavy oil RFs of the CSI + GF are slightly higher than 12.9% of the CSI alone (Test #2), which were 13.8, 15.6, and 13.9% in Tests #3–5, respectively. The remaining foamy oil after the previous cycle of the CSI production was quickly recovered by the injected CO2. During the CSI + GF, most heavy oil was produced in the first cycle, which helped to increase the heavy oil production rate at the early stage. Only 3.1, 3.4, and 2.0% heavy oil were recovered during the second and third cycles of the CSI + GF in Tests #3–5, respectively. In general, the heavy oil RF of the CSI was quickly decreased with the cycle number in each test because the oil saturation became lower and lower in the late cycles. Less than 1%

Fig. 5. Measured heavy oil production rate (qo), instantaneous GOR, pressure drop (ΔP), and production pressure (Ppro) during the CSI production of Cycle #1 in Test #4 (CSI + GF).

512

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

20

7000 6000

15

cGOR (sc cm3/cm3)

Heavy oil RF (%)

CSI GF

10

CSI GF

5000 4000 3000 2000

5 1000 0

0 1

2

1

3

2

Cycle number

Fig. 6c. Measured cumulative GORs during the three cycles of the CSI and GF in Test #4 (CSI + GF).

Fig. 6a. Measured heavy oil RFs during the three cycles of the CSI and GF in Test #4 (CSI + GF).

The variations of the measured average heavy oil production rates of GF Cycles #1–3 in Tests #3–5 (CSI + GF) are plotted in Fig. 7(b). During the first cycle of the GF, the average heavy oil production rate was slightly increased with the CO2 injection rate. It reduced dramatically during the second cycle of the GF, especially in Test #5 with the highest CO2 injection rate. Almost no heavy oil was produced during the last cycle of the GF. Fig. 7(c) shows the measured cumulative GORs of GF Cycles #1 and #2 in Tests #3–5 (CSI + GF). Tests #3–5 show the lower cumulative GORs of 146–171 sc cm3/cm3 during GF Cycle #1 than those of 500–1000 sc cm3/cm3 obtained from the field-scale CO2 flooding projects [24]. The low cumulative GOR is mainly attributed to the foamy-oil flow during the GF. The dispersed gas bubbles were trapped in the viscous heavy oil, which effectively controlled the gas mobility. Besides, the foamy oil increased the oil saturation and also the oil relative permeability because the dispersed gas bubbles expanded when the reservoir pressure was declined. Also, the foamy oil swelled and occupied some gas channels so that the gas channeling was weakened. The cumulative GORs of the GFs were increased when more cycles were conducted. This is because that the gas saturation became higher and higher as the oil production continued, especially near the producer. The cumulative GORs of GF Cycle #3 were increased to 3326, 6556, and 1157 sc cm3/cm3 in Tests #3–5, respectively. They are too

heavy oil was recovered during the third cycle of the CSI. Also, no light to mediate hydrocarbons were extracted by CO2 under the immiscible conditions because the heavy oil does not contain any hydrocarbons under C9 according to the gas chromatography (GC) analysis results [15]. In the first cycle of the GF, the heavy oil RF was slightly increased when CO2 injection rate was increased. A higher CO2 injection rate leads to a higher pressure gradient that accelerates the foamy-oil flow and heavy oil production. In addition, more CO2 can be dissolved into the heavy oil at a higher reservoir pressure. Thus, the heavy oil viscosity is further reduced and the oil-swelling effect is stronger. However, during the second cycle of the GF, the lowest heavy oil RF of 0.3% was achieved in Test #5 with the highest CO2 injection rate of 0.20 cm3/ min. This implies that more high-permeability channels were formed because of the gas BT at a higher CO2 injection rate. The subsequently injected CO2 flows through the established gas channels without displacing the foamy heavy oil [23]. During the third cycle of the GF, only 0.05, 0.03, and 0.3% heavy oil were recovered due to the severe gas channeling, corresponding to three respective CO2 injection rates. In summary, the heavy oil was quickly produced but the total heavy oil RF was low when CO2 injection rate is 0.20 cm3/min. Test #4 gave the highest heavy oil RF of 15.6% with an intermediate CO2 injection rate of 0.10 cm3/min.

20

RFGF RFCSI

15 CSI GF

Heavy oil RF (%)

Average oil production rate qo (cm3/h)

3

Cycle number

10

15

10

5

5

0 3

0

1

2

4

5

Test number

3

Cycle number

Fig. 7a. Measured heavy oil RFs during the three cycles of the CSI + GF in Tests #3–5 with the respective CO2 injection rates of qCO2 = 0.05, 0.10, and 0.20 cm3/min.

Fig. 6b. Measured average heavy oil production rates (qo) during the three cycles of the CSI and GF in Test #4 (CSI + GF). 513

Fuel 231 (2018) 507–514

H. Ma, Y. Gu

Average oil production rate qo (cm3/h)

15

ratio (GOR) than the combined CSI and GF. CO2 mobility is effectively controlled and CO2 gravity overriding is weakened because of the injected water. The foamy-oil flow and free-gas flow alternate during the CSI and subsequent GF. The free-gas flow becomes dominant as the gas saturation is increased in the late cycles of the CSI and GF. A high CO2 injection rate during the GF is detrimental to the heavy oil production as the free-gas flow is too strong. In summary, the GF/WF can be combined with the CO2-CSI as an effective EOR process in the thin post cold heavy oil production with sand (CHOPS) reservoirs.

10

Cycle #1

Cycle #1 Cycle #1

5

Acknowledgements

Cycle #2 Cycle #2 Cycle #3

Cycle #3

0

3

4

Cycle #2 Cycle #3

The authors wish to acknowledge the innovation fund from the Petroleum Technology Research Centre (PTRC) and the discovery grant from the Natural Sciences and Engineering Research Council (NSERC) of Canada to Dr. Yongan Gu. The technical assistance of the research group members in carrying out the nonstop experimental tests is highly appreciated.

5

Test number

Fig. 7b. Measured average heavy oil production rates (qo) during the three cycles of the CSI + GF in Tests #3–5 with the respective CO2 injection rates of qCO2 = 0.05, 0.10, and 0.20 cm3/min.

References [1] Annual Reservoir. Saskatchewan energy and mines. Regina, Saskatchewan, Canada; 2000. [2] Dusseault MB. Comparing Venezuelan and Canadian heavy oil and tar sands. Canadian international petroleum conference. Petroleum Society of Canada; 2001. [3] Han G, Bruno M, Dusseault MB. How much oil you can get from CHOPS. J Can Petrol Technol 2007;46:24–32. [4] Abusahmin BS, Karri RR, Maini BB. Influence of fluid and operating parameters on the recovery factors and gas oil ratio in high viscous reservoirs under foamy solution gas drive. Fuel 2017;197:497–517. [5] Haddad AS, Gates I. Modelling of cold heavy oil production with sand (CHOPS) using a fluidized sand algorithm. Fuel 2015;158:937–47. [6] Chang J, Ivory J. Field-scale simulation of cyclic solvent injection (CSI). J Can Petrol Technol 2013;52:251–65. [7] Miller KA. Improving the state of the art of Western Canadian heavy oil waterflood technology. J Can Petrol Technol 2006;45:7–11. [8] Dong M, Huang S. Methane pressure-cycling process with horizontal wells for thin heavy-oil reservoirs. SPE Res Eval Eng 2006;9:154–64. [9] Jha KN. A laboratory study of heavy oil recovery with carbon dioxide. J Can Petrol Technol 1986;25:54–63. [10] Sankur V, Emanuel AS. A laboratory study of heavy oil recovery with CO2 injection. SPE California regional meeting. Society of Petroleum Engineers; 1983. [11] Butler RM, McNab GS, Lo HY. Theoretical studies on the gravity drainage of heavy oil during in-situ steam heating. Can J Chem Eng 1981;49:22–33. [12] Vittoratos E, Scott GR, Beattie CI. Cold Lake cyclic steam stimulation: a multiwell process. SPE Res Eval Eng 1990;5:19–24. [13] Ma H, Yu G, She Y, Gu Y. A parabolic solvent chamber model for simulating the solvent vapor extraction (VAPEX) heavy oil recovery process. J Petrol Sci Eng 2017;149:465–75. [14] Rangriz-Shokri A, Babadagli T. Field scale modeling of CHOPS and solvent/thermal based post CHOPS EOR applications considering non-equilibrium foamy oil behavior and realistic representation of wormholes. J Petrol Sci Eng 2016;137:144–56. [15] Ma H, Huang D, Yu G, She Y, Gu Y. Combined cyclic solvent injection (CSI) and waterflooding (WF) in the post-cold heavy oil production with sand (CHOPS) reservoirs. Energy Fuels 2017;31:418–28. [16] Yadali Jamaloei B, Dong M, Mahinpey N. Enhanced cyclic solvent process (ECSP) for heavy oil and bitumen recovery in thin reservoirs. Energy Fuels 2012;26:2865–74. [17] Jia X, Zeng F, Gu Y. Gas flooding-assisted cyclic solvent injection (GA-CSI) for enhancing heavy oil recovery. Fuel 2015;140:344–53. [18] Sun X, Dong M, Zhang Y, Maini BB. Enhanced heavy oil recovery in thin reservoirs using foamy oil-assisted methane huff-n-puff method. Fuel 2015;159:962–73. [19] Lu T, Li Z, Li S, Li B, Liu S. Performances of different recovery methods for Orinoco Belt heavy oil after solution gas drive. Energy Fuels 2013;27:3499–507. [20] Bedrikovetsky P. WAG displacements of oil-condensates accounting for hydrocarbon ganglia. Transp Porous Med 2003;52:229–66. [21] Lu T, Li Z, Li S, Wang P, Wang Z. Enhanced heavy oil recovery after solution gas drive by water flooding. J Petrol Sci Eng 2016;137:113–24. [22] Sheng JJ, Maini BB, Hayes RE, Tortike WS. Critical review of foamy oil flow. Transp Porous Med 1999;35:157–87. [23] Jiang T, Zeng F, Jia X, Gu Y. A new solvent-based enhanced heavy oil recovery method: cyclic production with continuous solvent injection. Fuel 2014;115:426–33. [24] Dyer SB, Farouq Ali SM. The potential of the immiscible carbon dioxide flooding process for the recovery of heavy oil. Technical meeting/petroleum conference of the South Saskatchewan section. Petroleum Society of Canada; 1989.

1000 Cycle #2

cGOR (sc cm3/cm3)

800

600 Cycle #2

400

200

Cycle #1

Cycle #1

0

Cycle #2

3

4 Test number

Cycle #1

5

Fig. 7c. Measured cumulative GORs during Cycles #1 and #2 of the GF in Tests #3–5 with the respective CO2 injection rates of qCO2 = 0.05, 0.10, and 0.20 cm3/min.

high to be plotted in this figure. In GF Cycles #1 and #2, a higher CO2 injection rate caused a higher cumulative GOR. More CO2 was injected and the free-gas flow became stronger at a higher CO2 injection rate. 4. Conclusions In this paper, a new enhanced oil recovery (EOR) process, namely, combined CO2-cyclic solvent injection (CO2-CSI) and gasflooding (GF) in a two-well configuration is proposed and compared with the combined CO2-CSI and waterflooding (WF). The joint enhanced oil recovery (EOR) and improved oil recovery (IOR) mechanisms are examined by conducting the experimental tests of the CSI + GF/WF, simultaneous (CSI + GF/WF), and GF/WF + CSI, in which the CO2/water is injected post, at the same time with, and prior to the CSI, respectively. The highest heavy oil recovery factor (RF) is obtained in the CSI + WF due to the enhanced microscopic displacement efficiency (CSI) and improved volumetric sweep efficiency (WF). The simultaneous (CSI + WF) has the highest average heavy oil production rate and the lowest cumulative gas-oil ratio (GOR) because the reservoir pressure is well maintained. The combined CSI and WF gives a higher heavy oil RF with a higher heavy oil production rate and a lower cumulative gas-oil

514