Experimental study on the mechanism and development effect of multi-gas assisted steam huff and puff process in the offshore heavy oil reservoirs

Multi-gas assisted steam huff and puff process is a relatively new thermal recovery technology for offshore heavy oil reservoirs. Some blocks of Bohai oilfield have implemented multi-gas assisted steam huff and puff process. However, the development mechanism still requires further study. In this paper, high-temperature high-pressure (HTHP) PVT experiments and different huff and puff experiments of sand pack were carried out to reveal the enhanced production mechanism and evaluate the development effect of multi-gas assisted steam huff and puff process. The results indicated that viscosity reduction and thermal expansion still were the main development mechanism of multi-gas assisted steam huff and puff process. Specifically, CO2 easily dissolved in the heavy oil that made it mainly play the role of reducing oil viscosity, N2 was characteristics of small solubility and good expansibility, and it could improve formation pressure, increase steam sweep volume and even reduce the heat loss. Meanwhile, injecting multi-gas and steam could break the balance of heavy oil component that made the content of resin reduce and the content of saturates, aromatics and asphaltene increase so as to further reduce the viscosity of heavy oil. Compared with steam huff and puff process, multi-gas assisted steam huff and puff process increased the recovery by 2–5%. The optimal water–gas ratio and steam injection temperature were 4:6 and 300℃, respectively. The results suggested that multi-gas assisted steam huff and puff process would have wide application prospect for offshore heavy oil reservoirs.


Introduction
As one of main methods for producing heavy oil, steam huff and puff process is characterized of simple process and wide applicability for heavy oil reservoirs. However, steam huff and puff process is classified into primary oil recovery mainly depending on natural energy (Liu 2013;Wang et al. 2004;Sun et al. 2017Sun et al. , 2018aSun et al. , b, 2019Yoelin 1970;Bao et al. 2016). For the conventional steam huff and puff process, specifically in the later stage of steam huff and puff process, the thermal efficiency of steam is dramatically reduced that leads to the recovery decrease. Thus, an improvement recovery process is required.
Previous research results show that injecting gas such as CO 2 , N 2 or flue gas can increase the steam sweep volume, supplement the formation energy and further reduce the viscosity of heavy oil and residual oil saturation (Gao et al. 2003;Firouz and Torabi 2014;Mohsenzadeh et al. 2016;Wang et al. 2018, Fan et al. 2019. For example, at the same time as steam injection, injecting CO 2 can reduce the viscosity of heavy oil and thereby increase the recovery of heavy oil Liu et al. 2016;Wang et al. 2018;Dong and Huang 2002;Dong et al. 2014;Song and Yang 2017). Specifically, CO 2 is easily dissolved into crude oil, which increases the volume of crude oil by 10-30%, and then the viscosity of crude oil will appear a certain degree decline which depends on pressure, temperature and the properties of crude oil. Meanwhile, crude oil extraction by CO 2 is a process where the interfacial tension between CO 2 and crude oil continuously decreases. N 2 injection will help maintain formation pressure, prolong the huff and puff cycle, improve the steam sweep volume and increase the displacement efficiency in the immediate vicinity of wellbore (Wu et al. 2018;Liu et al. 2011;Dong et al. 2016). Considering that N 2 has lower thermal conductivity, after injecting N 2 into reservoir, steam could be bring into the further area of reservoir which increases the heating volume and sweep efficiency of steam. What's more, as the pressure decreases in the course of development, CO 2 and N 2 will extract from crude oil and generate dissolved gas drive in the later stage. In view of these, injection CO 2 and N 2 would have positive effect for production heavy oil in the course of steam injection (Hong and Ault 1984;Butler et al. 2000;Fan et al. 2016;Jin et al. 2017;Yan et al. 2020).
Heavy oil resource is abundant in the Bohai oilfield of China. Heavy oil with the viscosity of less than 350 mPa·s can be effectively developed by cold recovery technology, such as water flooding, chemical flooding and well pattern thickening. However, for heavy oil with the viscosity of greater than 350 mPa·s, just limited reserves could be developed with the cold recovery technology. At this time, steam huff and puff process will become an effective way of development. But the reservoir type of Bohai oilfield is complex and diverse, viscosity of heavy oil in some blocks can exceed 10000 mPa·s, the development effect of simple steam huff and puff process is not ideal for this kind of heavy oil, a more reasonable development method is needed. Injecting gas such as CO 2 , N 2 and flue gas during steam huff and puff process has been generally applied to develop heavy oil reservoirs in onshore oilfield and achieved success. Therefore, multi-gas assisted steam huff and puff process should improve the production effect of offshore heavy oil reservoirs like Bohai oilfield (Liu et al. 2011;Gu et al. 2007).
In this paper, firstly, the influence of CO 2 and N 2 injection on the solubility, density and viscosity of heavy oil was studied by HTHP PVT experiments based on the offshore heavy oil of Bohai oilfield. Then sand pack experiments were carried out to compare the development effect of different huff and puff process and the change of crude oil component so as to further reveal the development mechanism of multi-gas assisted steam huff and puff process. Finally, 10 groups of sand pack experiments were conducted to optimize the key parameters of water-gas ratio and steam injection temperature. All the study results could provide theoretical and application support for the multi-gas assisted steam huff and puff process in offshore heavy oil reservoirs like Bohai oilfield.

Experimental sample
The crude oil was collected from the Bohai oilfield which is a heavy oil reservoir with the formation depth of 1510.2 m and original pressure of 15.06 Mpa. The oil sample was cleaned by using a centrifuge to remove any sand particles and brine. Table 1 shows the physical properties of oil sample. Density and viscosity of the oil sample were, respectively, measured by a densitometer (DMA 4200 M, Anton Paar, Austria) and a rheometer (MCR302, Anton Paar, Austria). The asphaltene content of the crude oil was measured by using the standard ASTM D2007-03 method and filter papers with a pore size of 2.5 μm.   (Jha 1986;Srivastava et al. 1997;Kavousi et al. 2014;Telmadarreir et al. 2016;Hatzikiriakos 2014).

Experimental sample
The heavy oil sample was from Bohai oilfield, and the initial oil saturation of the reservoir was 72%. The physical properties of heavy oil sample are shown in Table 1 and Fig. 1. Purity of CO 2 and N 2 (Tianyuan, Co., Ltd., China) was 99.99 mol%. The property of the formation water is   Table 2, with a density of 1.01 g/cm 3 under standard condition.

Experimental apparatus
As shown in Fig. 4, the experimental apparatus mainly consisted of sand pack, steam generator, plunger pump, flowmeter, calorstat and beaker, etc. The maximum steam output temperature and pressure of steam generator were 400 ℃ and 35 MPa, respectively. Operating range of calorstat was 0-350 ℃. Working range of pressure meter was 0-30 MPa. Working range of plunger pump was 0-50 mL/ min, and maximum pressure was 25 MPa. The size of sand pack model was 48 mm × 340 mm, its maximum bearing pressure was 35 MPa.

Experimental procedure
The specific experimental procedures are as follows: (1) Filled sand: erected the model, filled simulated sand from the upper port and blocked the port after the sand reached the filling port, shook the model by small amplitude in the middle axis and static 1 h, opened the port and continued to fill the sand. Repeated the opera-tion several times until the sand was no longer reduced from the filling port.
(2) Saturated formation water: constant temperature 70 ℃, injected formation water into the sand pack at 25 mL/ min until the producing port met the water, continuously injected formation water 1 h. Stopped injecting water and kept the temperature. The sand pack had a permeability of water phase between 3574 and 3862 md, and porosity between 33.7 and 37.6%. (3) Saturated oil: Injected oil into the sand pack at 20 mL/ min, drained off water until the output liquid of model was not hydrated and continuously injected oil 1 h. Stopped injecting oil and kept the constant temperature. The oil saturation of sand pack was between 71 and 73%. (4) Different huff and puff experiment: injected hot water/ steam into the sand pack at 20 mL/min. During the multi-gas assisted steam huff and puff experiment, the water-gas ratio was 1:2.

Fig. 4 Schematic apparatus of huff and puff experiment
The results and discussion of experiment The results and discussion of HTHP PVT experiment The solubility of CO 2 and N 2 in the heavy oil Figures 5 and 6 show the relationship between solubility of CO 2 and N 2 in the heavy oil and pressure at different temperature. It can be seen that the solubility of CO 2 and N 2 in degassed heavy oil increased as the pressure increased. With the increase in temperature, the solubility of CO 2 decreased and the solubility of N 2 increased. When the pressure was fixed at 12 MPa, the solubility of CO 2 in heavy oil decreased from 60.2 to 26.8 Sm 3 /m 3 as the temperature increased from 80 to 200 ℃, but the solubility of N2 in heavy oil increased from 2.9 to 8.1 Sm 3 /m 3 in the process. Under the same condition, the solubility of CO 2 was obvious higher than that of N 2 . When the pressure was 3 MPa and temperature was 200 ℃, the solubility of CO 2 and N 2 in the heavy oil was 13.35 and 2.1 Sm 3 /m 3 , respectively. In summary, although a high temperature had a negative effect on the solubility of CO 2 in the heavy oil, CO 2 still had a better solubility in heavy oil at 200 ℃.

The effect of CO 2 and N 2 on the density of heavy oil
Figures 7 and 8 show the relationship between density of heavy oil saturated with gas and pressure at different designated temperature. As the pressure and temperature increased, the density of heavy oil saturated with CO 2 or N 2 decreased. The reason might be when the temperature increased, the volume of heavy oil expanded and the space between molecular increased, so the density decreased.
When the pressure increased, more CO 2 or N 2 dissolved into the heavy oil that increased the space between molecular; hence, the density of heavy oil would also decreased. When the temperature was 200℃, the density of heavy oil Relationship between density of heavy oil saturated with N 2 and pressure at different temperature saturated with CO 2 decreased from 0.853 to 0.834 g/cm 3 as the pressure increased from 3 to 12 MPa, which of heavy oil saturated with N 2 decreased from 0.872 to 0.856 g/cm 3 as the pressure increased from 3 to 12 MPa.

The effect of CO 2 and N 2 on the viscosity of heavy oil
Figures 9 and 10 show the relationship between viscosity of heavy oil saturated with gas and pressure at a given temperature. Compare the change of density and viscosity with pressure and temperature in Figs. 7, 8, 9, and 10. It was found that the density and viscosity had the same change trend with pressure and temperature. The viscosity of heavy oil saturated with CO 2 or N 2 decreased with the increase in pressure and temperature. It was because the change of pressure and temperature influenced the solubility of CO 2 or N 2 in heavy oil and the space between molecular. When the temperature was 200℃, the viscosity of heavy oil saturated with CO 2 decreased from 18.4 to 12.1 mPa·s as the pressure increased from 3 to 12 MPa, which of heavy oil saturated with N 2 decreased from 21.6 to 15.7 mPa·s as the pressure increased from 3 to 12 MPa.
In summary, dissolution of CO 2 or N 2 in heavy oil will expand the volume of heavy oil and decrease the density and viscosity. The solubility of CO 2 in heavy oil is bigger than that of N 2 , so CO 2 is more easily dissolved in the heavy oil and mainly plays the role of reducing oil viscosity. Although the effect of N 2 injection on reducing density and viscosity is not obvious, N 2 has a large compression coefficient due to its lower critical temperature and minimal dependence on temperature. As shown in Fig. 11, the compression coefficient of N 2 increases with the increase in pressure, which is much higher than that of CO 2 , because the critical temperature of N 2 is low, and the compressibility coefficient of N 2 is less affected by temperature (Li 2017). So N 2 injection will help maintain formation pressure, improve the steam sweep volume and increase the displacement efficiency, in addition, considering the lower thermal conductivity of N 2 , which is 0.0228 W/(m K) At 0.1 MPa and 0 ℃. N 2 injection will promote steam migration, enhance the heat carrying capacity of steam and improve the heat utilization of steam.

The development effect of different huff and puff experiments
In order to compare the development effect of different huff and puff process and further reveal the development mechanism of multi-gas assisted steam huff and puff process, different huff and puff experiments were carried out at different designed temperature; namely, hot water huff and puff process, steam huff and puff process, multi-gas assisted steam huff and puff process were studied at 150 ℃ and 250 ℃, respectively.  Heavy oil is a complex and multi-component mixture that includes saturates, aromatics, resin and asphaltene (SARA, also called four components). Different development methods may destroy the component balance of heavy oil, which results in the change of four components content in the developed heavy oil (Kokal et al. 1992;Monin and Audibert 1987). In order to further study the development mechanism of different huff and puff process, as shown in Table 3, the SARA was analyzed after each huff and puff experiment. The results show that the content of saturates, aromatics and asphaltene increased, but the content of resin reduced from hot water huff and puff process to steam huff and puff process, and then to multi-gas assisted steam huff and puff process at each experiment temperature. Meanwhile, a higher content of asphaltene and lower content of resin appeared with the temperature increase.
Thermal expansion and viscosity reduction are still the main mechanism of thermal recovery. As shown in Fig. 1, the viscosity of heavy oil was very sensitive to the temperature. The viscosity of degassed heavy oil decreased exponentially as the temperature increased. But when the temperature was more than 150 ℃, the viscosity of heavy oil changed little with the increase in temperature. Figure 12 shows the relation between huff and puff cycle and recovery of hot water huff and puff process. Figure 13 shows the change of four components content after hot water huff and puff process at different temperature. As shown in Fig. 13, when the temperature rose from 150 ℃ to 250 ℃, the four components of heavy oil changed little after hot water huff and puff process. It indicated that heavy oil had not occurred obvious physicochemical reaction, so the main mechanism of hot water huff and puff process was viscosity reduction and thermal expansion. Figure 14 compares the recovery between hot water huff and puff process and steam huff and puff process. As shown  . 12 The recovery of hot water huff and puff process at different temperature Fig. 13 The SARA change of hot water huff and puff process at different temperature Fig. 14 The Recovery of hot water/steam huff and puff process at different temperature in Fig. 14, compared with hot water huff and puff process, the recovery of steam huff and puff process had obviously increased. Figures 15 and 16 show the change of four components content after hot water huff and puff process and steam huff and puff process at different temperature. In the course of steam huff and puff, when the temperature increased from 150 to 250 ℃, the content of saturates, aromatics and asphaltene increased, but the content of resin reduced. It indicates that the aquathermal cracking reaction had occurred in the steam huff and puff process. In the later period of experiment, the condensate of liquid hydrocarbon appeared in the beaker wall, which proved that steam distillation also had occurred. Therefore, in addition to viscosity reduction and thermal expansion, the recovery mechanism of steam huff and puff process also consisted of hydrothermal cracking reaction and steam distillation. Figure 17 shows the recovery of three different huff and puff experiments. As shown in Fig. 17, the recovery of multi-gas assisted steam huff and puff process was further improved compared with hot water/steam huff and puff process due to the synergistic effect of multi-gas and steam. The recovery of multi-gas assisted steam huff and puff at 150 ℃ and 250 ℃ was 11.9% and 22.3%, respectively, which was 7.9% and 2% higher than that of hot water huff and puff and steam huff and puff at 150 ℃, and 13.4% and 5% higher than that of hot water huff and puff and steam huff and puff at 250 ℃. Figures 18 and 19 compare the change of four components content after three different huff and puff experiments at different temperature. As shown in Figs. 18 and 19, compared with hot water/steam huff and puff process, the four components content in the developed heavy oil of multigas assisted steam huff and puff process had exhibited more obvious change. What's more, at the same temperature, the content of asphaltene in the developed heavy oil of multigas assisted steam huff and puff process was the highest (Clark et al. 1984(Clark et al. , 1987aClark and Hyne 1990). With the increase in temperature from 150 to 250 ℃, the content of saturates, aromatics and asphaltene obviously increased, but which of resin reduced. The result indicates that multi-gas injection in the process of steam huff and puff process had further broken the balance of heavy oil components and the Fig. 15 The SARA change of hot water/steam huff and puff process at 150℃ Fig. 16 The SARA change of hot water/steam huff and puff process at 250℃ Fig. 17 The recovery of three different huff and puff process at different temperature Fig. 18 The SARA change of three different huff and puff process experiment at 150℃ aquathermal cracking reaction had played an important role during the multi-gas assisted steam huff and puff process. In addition, in the later period of multi-gas assisted steam huff and puff experiments, the condensate of liquid hydrocarbons also appeared in the beaker wall, and the amount was obviously more than that of steam huff and puff experiments. It indicates that injecting multi-gas had enhanced the steam distillation. What's more, from the HTHP PVT experiment, we have known that the CO 2 and N 2 could increase the expansion ratio, reduce the density and viscosity by dissolving in heavy oil, and N 2 could maintain formation pressure, improve heating range and heat utilization of steam. All of the development mechanism further improved the recovery of multi-gas assisted steam huff and puff process.

1) Water-gas ratio
Ensured that other parameters remain unchanged, which was to ensure the size of sand pack, initial pressure, initial porosity, initial saturation, steam injection temperature 250 ℃ and so on were the same. Six different water-gas volume ratio (2:8, 3:7, 4:6, 5:5, 6:4, 7:3) experiments of multi-gas assisted steam huff and puff process with 6 cycles were studied. The each cycle recovery and cumulative recovery of multi-gas assisted steam huff and puff process are shown in Figs. 20 and 21. It can be seen that with the increase in steam injection, cumulative recovery increased. When the water-gas ratio reached 4:6, the cumulative recovery began to decrease and the maximum cumulative recovery was acquired when the water-gas ratio was 4:6.
It could be known from the results of previous experimental studies that CO 2 and N 2 could dissolve into the heavy oil and reduce the viscosity and interfacial tension; furthermore, N 2 had a low thermal conductivity that increased the sweep volume and heat utilization of steam. As a result, multigas assisted steam huff and puff process could more effectively improve the micro-displacement efficiency of steam and reduce residual oil saturation than steam huff and puff process. In the process of experiments, with the increase in steam ratio, the heat injection increased greatly, which increased the steam sweep volume and viscosity reduction area, thus effectively improved the development effect of thermal recovery. However, as the water-gas ratio increased, the Jia Min effect would be formed at the throat to prevent the flow of gas and increase the pressure in immediate vicinity of wellbore due to the heat loss and steam transforming to hot water, which made the injection difficult and a large amount of fluid cannot be effectively injected. At the same time, due to the decrease of the amount of injected gas, the dissolved gas of oil was reduced, which weakened the expansion energy and viscosity reduction of heavy oil, so the cumulative recovery would reduce. Therefore, the optimal water-gas ratio was 4:6.
2) Steam injection temperature Fig. 19 The SARA change of three different huff and puff process experiment at 250℃ Fig. 20 Each cycle recovery of multi-gas assisted steam huff and puff process with different water-gas ratio Fig. 21 The cumulative recovery of multi-gas assisted steam huff and puff process with different water-gas ratio Ensured that other parameters remain unchanged, which was to ensure the size of sand pack, initial pressure, initial porosity, initial saturation, water-gas ratio 4:6 and so on were the same. Four different steam injection temperature (250 ℃, 280 ℃, 300 ℃, 350 ℃) experiments of multi-gas assisted steam huff and puff process with 6 cycles were studied.
The each cycle recovery and cumulative recovery are shown in Figs. 22 and 23. As the steam injection temperature rose, the cumulative recovery increased. The cumulative recovery of 4 groups of experiments was 24.93%, 27.61%, 29.32% and 30.35%, respectively, the cumulative recovery was the highest at 350 ℃, but the cumulative recovery was little different between 300 and 350 ℃. The reason could be that as the steam injection temperature increased, both viscosity reduction degree of heavy oil and oil flow capacity in the formation increased; in addition, the thermal expansion of heavy oil increased the oil saturation in the formation. However, constantly increasing steam injection temperature would reduce the dissolved gas and weaken the amplitude of volume expansion and viscosity reduction of heavy oil. Therefore, after reaching a certain steam injection temperature, although steam injection temperature still increased, volume expansion and viscosity reduction of heavy oil were not obvious. And the excessive temperature easily caused the heavy components of heavy oil to coke, which leaded to the blockage of the pore, and reduced the permeability and cumulative recovery. So the optimal steam injection temperature was 300 ℃.

Conclusions
1. After injecting CO 2 or N 2 , the solubility of gas in degassed heavy oil increased as the pressure increased.
With the increase in temperature, the solubility of CO 2 decreased and the solubility of N 2 increased. Under the same condition, the solubility of CO 2 was obvious higher than that of N 2 . Due to the dissolving of N 2 and CO 2 , the density and viscosity of heavy oil decreased as the pressure increased. 2. Injecting multi-gas and steam could break the balance of heavy oil component. With the increase in temperature, the content of resin reduced and the content of saturates, aromatics and asphaltene increased. Except for thermal expansion and viscosity reduction, the hydrothermal cracking reaction and steam distillation were also the important mechanism of enhanced oil recovery for multi-gas assisted steam huff and puff process. 3. Multi-gas assisted steam huff and puff process increased the recovery by 2-5% compared with steam huff and puff process. The optimal water-gas ratio and steam injection temperature which were 4:6 and 300 ℃, respectively, were optimized by experiment. Considering that multigas assisted steam huff and puff process had higher recovery, it is suggested to further study the optimal value of injection-production parameters to excavate the greatest potential of multi-gas assisted steam huff and puff process for developing the offshore heavy oil reservoirs.
Funding This study was funded by the Department of well completion at the SINOPEC Research Institute of Petroleum Engineering.

Fig. 22
Each cycle recovery of multi-gas assisted steam huff and puff process at different injection temperature Fig. 23 The cumulative recovery of multi-gas assisted steam huff and puff process at different injection temperature