J. Cent. South Univ. (2021) 28: 1412-1421

DOI: https://doi.org/10.1007/s11771-021-4712-z

Remaining oil distribution in models with different heterogeneities after CO₂ WAG injection: Visual research by nuclear magnetic resonance technique

WANG Ye-fei(王业飞)^{1, 2, 3}, LI Zong-yang(李宗阳)^{1, 4}, ZHANG Shi-ming(张世明)⁴,LIU De-xin(刘德新)^{1, 3}, DING Ming-chen(丁名臣)^{1, 2, 3}

1. Key Laboratory of Unconventional Oil & Gas Development, Ministry of Education,China University of Petroleum (East China), Qingdao 266580, China;

2. Shaanxi Key Laboratory of Advanced Stimulation Technology for Oil & Gas Reservoirs,Xi’an Shiyou University, Xi’an 710043, China;

3. Shandong Key Laboratory of Oilfield Chemistry, China University of Petroleum (East China),Qingdao 266580, China;

4. Exploration and Development Research Institute of Shengli Oilfield, Dongying 257100, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract: Water alternating gas (WAG) injection is a widely used strategy for enhancing oil recovery (EOR) during gas flooding, and the mechanisms, operating parameters, and influencing factors of which have been extensively studied. However, with respect to its capacity in expanding macroscopic sweep volume under varying heterogeneities, the related results appear inadequate. In this research, three cores with different heterogeneities were used and flooded by the joint water and CO₂ WAG, then the effects of heterogeneity on oil recovery were determined. More importantly, the cores after CO₂ WAG injection were investigated using the nuclear magnetic resonance (NMR) technique for remaining oil distribution research, which could help us to understand the capacity of CO₂ WAG in enlarging sweep volume at different heterogeneities. The results show that the presence of heterogeneity may largely weaken the effectiveness of water flooding, the more severe the heterogeneity, the worse the water flooding. The WAG injection of CO₂ performs well in EOR after water flooding for all the cores with different heterogeneities; however, it could barely form a complete or full sweep throughout the low-permeability region, and un-swept bypassed regions remain. The homogeneous core is better developed by the injection of the joint water and CO₂ WAG than the heterogeneous and fractured cases.

Key words: heterogeneous reservoir; enhancing oil recovery; CO₂ WAG injection; sweep volume; remaining oil; nuclear magnetic resonance

Cite this article as: WANG Ye-fei, LI Zong-yang, ZHANG Shi-ming, LIU De-xin, DING Ming-chen. Remaining oil distribution in models with different heterogeneities after CO₂ WAG injection: Visual research by nuclear magnetic resonance technique [J]. Journal of Central South University, 2021, 28(5): 1412-1421. DOI: https://doi.org/10.1007/ s11771-021-4712-z.

1 Introduction

CO₂ injection today helps to recover a substantial portion of the oil from low permeability or tight reservoirs, mainly due to its high microscopic displacement efficiency. The residual oil saturation in CO₂ swept zones has been found to be quite low. What makes CO₂ flooding so powerful is the well-known mechanisms: miscibility, light hydrocarbon extraction, oil viscosity reduction, volume expansion (by gas dissolution), etc. [1-4];　however, the volumetric sweep in CO₂ floods has always been a critical issue of concern [5]. The unfavorable mobility ratio between CO₂ and oil (caused by their significant viscosity contrast) usually results in reduced sweep volume in reservoirs. Furthermore, the widespread reservoir heterogeneity would further decrease this macroscopic sweeping efficiency [6-8]. With the lack of profile control, viscous fingering, channeling, early CO₂ breakthrough, and oil bypassing would impair the macroscopic sweep efficiency of CO₂ flood, resulting in significantly reduced oil recovery.

To seek a better conformance control, the water alternating gas (WAG) injection strategy has been used since 1957 [9, 10]. It combines the advantages of water flooding in expanding sweep volume and CO₂ injection in high displacement efficiency. Furthermore, WAG may also contribute to bubble generation in formation, building up high flow resistance in the swept regions (by virtue of the Jiamin effect), thus, diverting more displacing fluid into those un-swept regions, leading to enhanced macroscopic sweep efficiency. By micro-scale experiment, QIN et al [11] found that the presence of gas in large pores indeed diverted the flow path of brine during the second water flooding cycle. As a result, the improved sweep volume together with the good microscopic displacing efficiency can make CO₂ WAG very effective. After several decades of research and field application, WAG technology has matured to become one of the most effective enhancing oil recovery (EOR) techniques in low-permeability reservoirs. Its main design parameters and influencing factors have been much evaluated and revealed at laboratory scale, including WAG ratio, number of WAG cycles, volume of each cycle, injection rate, oil-gas phase behavior (miscible or immiscible), gas type, rock properties, gravity segregation, permeability variation, etc. [12-23]. For instance, simulation and experimental results indicated that multiple WAG cycles at a WAG ratio of 1:1 resulted in optimum oil recovery [12]. By increasing the WAG ratio in a model from 0.5 to 1 and to 2, AFZALI et al [13] also noticed that the optimal operating condition was achieved at the WAG ratio of 1. ZEKRI et al [14] found that increasing heterogeneity resulted in reduced oil recovery by WAG, and reservoir heterogeneity should be taken into consideration when ascertaining the optimum CO₂-WAG ratio. By conducting WAG injection in visual micro-models, DONG et al [15] found that the oil recovery declined greatly with WAG cycles because of the increasingly high-water saturation and reduced discontinuity of the oil phase. KULKARNI et al [16] reported that miscible CO₂ WAG could recover the waterflood residual oil over 60% to 70% more than immiscible case. MIRKALAEI et al [17] thought that injection rate should be optimized for WAG schemes. Low or higher injection rate might cause gravity segregation or viscous fingering, which led to a lower oil recovery factor. In immiscible WAG processes, ZHANG et al [18] found that the presence of N₂ in the enriched flue gas stream (30% CO₂) could reduce the gas solubility, weaken oil swelling, and diminish the viscosity reduction, indicating a less effective N₂ WAG than CO₂ WAG. AFZALI et al [19] conducted a comprehensive review on EOR by WAG injection and concluded that CO₂ was the most common gas used in the WAG operations and it had advantages over N₂ or O₂. The use of high-pressure air was also recently proposed due to its abundance.

To date, nearly all the commercial gas injection projects use the WAG method to realize profile control and gas consumption reduction. The substantial research and in-situ tests result in better understanding and application of the WAG process; however, in spite of these improvements, the field performance of WAG process is sometimes disappointing with enhanced oil recovery factors of only 5%-10% (according to the data given by Ref. [9] after reviewing over 50 field projects involving WAG process). This reminds us the question as to whether WAG can sweep the whole low-permeability (or initially bypassed) region (in heterogeneous or fractured reservoirs)? Can it really extract the majority of oil therein? Based only on the aforementioned parameter optimization and influencing factor evaluations [12-23], this issue cannot be thoroughly answered. Therefore, in this research, three different types of cores (homogenous, heterogeneous, and fractured cores) were used for the primary water and secondary CO₂ WAG flooding. The oil recovery factors and water cuts were measured to identify the effects of heterogeneity on water and CO₂ WAG flooding. More importantly, the distribution of the remaining oil after CO₂ WAG injection was investigated using a NMR technique; then, the production characteristics of the oil in the low-permeability (or initially bypassed) regions could be determined and thus, whether the WAG could remove the majority of oil there or not could be revealed.

2 Experimental section

2.1 Materials

2.1.1 Fluid

A sample of light crude oil was collected from the Yanchang oil field in Shanxi Province, China. The density and viscosity of the crude oil sample were measured and found to be 843.4 kg/m³ (at 20°C, 0.101 MPa) and 5.2 mPa·s (at 60°C, 0.101 MPa), respectively. The CO₂ (99.99% pure) was supplied by the Antaike Gas Company, China. The minimum miscibility pressure (MMP) between oil and CO₂ was determined to be 17.8 MPa at 60 °C via the slim tube method. The CO₂ WAG injections conducted at 15.0 MPa in this study are all in immiscible states.

2.1.2 Core sample

The cores used in this research are a type of artificial model widely applied in laboratory oil-recovery tests: the surface of such models behaves water-wet property initially. Three types of cores

with different permeabilities and heterogeneities were used, the homogeneous, heterogeneous and fractured cores. Those core plugs were all cut to cylinders with the same diameter and length of 100 and 25 mm, respectively. Figure 1 shows the photographs of cores after water and CO₂ WAG flooding and their specific physical parameters.

Figure 1 Photograph showing experimental cores after water and CO₂ WAG flooding

2.2 Experimental set-up

The schematic diagram of the experimental set-up is illustrated in Figure 2. A high-pressure stainless steel core holder was used to accommodate the core sample which was mounted in its center using a piece of rubber sleeve. Three piston cylinders were used to store the experimental fluids (oil, water, and CO₂). An injection pump connected to the cylinders was used to inject the fluids into the core through steel lines during experiments. The injection pressure was measured through a pressure gage installed at the inlet of the core holder. The confining pressure of the core holder was applied using another manual pump to avoid the bypassing flow of the experimental　fluids around the core. A back-pressure regulator was set at the outlet of the holder to maintain a pre-specified experimental pressure inside the core holder. The produced oil and water were gathered in a collector installed downstream of the device. Finally, an air bath maintained a pre-specified temperature of 60 °C throughout the experiment.

Figure 2 Schematic of experimental set-up used in huff-n-puff tests

2.3 Experimental procedure

The first step is oil saturation. The core samples were placed in an airtight stainless-steel container and oil was directly injected into the container after vacuumization for 4 h. Oil injection was continued until a pressure of 20 MPa was reached inside the container; then, the pressure was maintained for 24 h at 60°C to help in achieving complete oil saturation and aging. Thereafter, the volume of saturated oil was determined according to the weight increment of a core (before and after oil saturation). The detailed physical parameters of the cores used are listed in Table 1.

Table 1 Properties of cores and corresponding oil recovery factors measured

Following the procedure above, the oil saturated core was placed into the core holder and the confining pressure was applied and dynamically maintained at 2.0 MPa above the pressure inside the core holder. Temperature of the apparatus was then heated to the pre-specified 60 °C. Water was then injected to displace oil there at a constant rate of 0.25 mL/min and simultaneously pressurize the core to a pressure of 15.0 MPa. The produced oil and water were then collected and measured for oil recovery and water cut calculation. Such primary water flooding was continued until a high water cut above 98.0% was obtained. Thereafter, CO₂ and water were alternatively injected using a volume ratio of 1:1 under a pressure of about 15.0 MPa. The volume of each CO₂ and water slug was designed to be 0.1 PV (pore volume) and the injection rate was maintained at 0.25 mL/min. Such alternative injection was repeated until the oil production became negligible and the water cut surpassed 98.0%. Then, the oil recovery by water and CO₂ WAG injection could be calculated according to the oil collected during flooding. At last, the core plugs with remaining oil in it were analyzed using an NMR technique for determination of the remaining oil distribution. In addition, it is noted that the water all mentioned in this research is a MnCl₂ solution with a concentration of 500 g/L. This allows the water signal to be shielded during nuclear magnetic scans, and only the signal pertaining to the remaining oil is found, making the results more accurate.

3 Experimental results and discussion

3.1 Effect of heterogeneity on water flooding

Water flooding is a most widely used strategy to replenish formation energy and displace crude oil therein after primary elastic development. Even for a low-permeability reservoir, high-pressure water injection is a first consideration for development, followed by other EOR strategies, i.e. CO₂ WAG. Therefore, water flooding tests were firstly conducted in those chosen models with different heterogeneities, thus, heterogeneity effects on oil recovery could be investigated. The measured cumulative oil recovery factor (CRF), water cut (fraction of the produced water in the total liquid) and incremental recovery factor (IRF) are demonstrated in Figure 3.

Water is more viscous than CO₂, and as a result, the underlying viscous fingering and the consequent fluid channeling may (theoretically) be reduced; however, influenced by heterogeneity, water flooding becomes much poorer (see Figure 3). To be more specific, in the homogeneous core, water flooding performs well in recovering oil there, showing the fastest and longest CRF growth (which tends to stabilize after 1.0 PV water injection), a slower rise in water cut, and the highest IRF of 41.7%. Once the heterogeneity emerges (even at a very weak level of a permeability contrast of 3.0), water flooding becomes significantly worse, the rise of CRF slows down and becomes shorter (which gradually stabilizes after only 0.4 PV of water injection), the water cut grows faster. As a result, a considerably reduced IRF of 22.2% is measured. In addition, it is clear that the produced oil in such heterogeneous core is mainly the oil initially saturated in the high-permeability region, so it can be predicted that, for real heterogeneous reservoirs, if the volume ratio of those high-permeability region to the whole reservoir is reduced (to less than the 1:2 in this experiment), then, the water flooding there would be even worse than that in this experiment. For the fractured core, the designed open fracture (formed by two layers of iron mesh between the two pieces of half core) leads it to the severest heterogeneity. As a result, the injected water mainly flows along this fracture channel, resulting in the slowest CFR growth, the fastest water cut increase, and lowest IRF of 7.9%. The recognized imbibition [24, 25] may be the underlying mechanism for oil production in this core. So, we find that such open fractures (connecting the injection and production well) in reservoirs may cause disastrous consequences for water flooding.

Figure 3 Measured parameters during water flooding in cores with different heterogeneities:

In general, for homogenous, heterogeneous and fractured cores, water flooding all works in recovering crude oil but to different extents. Such effectiveness becomes significantly poorer in the presence of heterogeneity, the more severe the heterogeneity, the worse the water flooding. So, for all the cores used here, further EOR methods (i.e. CO₂ WAG) are needed.

3.2 Effect of heterogeneity on CO₂ WAG flooding

The WAG pattern is proposed firstly to inhabit the viscous fingering in gas flooding; at the same time, the bubble formed during WAG injection can also enlarge the macroscopic sweep volume by the well-known Jiamin effect, reducing the heterogeneity effects on sweep efficiency [11]. In order to determine the adaptability and performance of CO₂ WAG for EOR under different heterogeneities, flooding tests were continued after the previous water injection. The CRF, water cut, and IRF are measured in Figure 4.

As illustrated in Figure 4, the CRF in the homogenous core is enhanced by CO₂ WAG injection from the initial 41.7% to 66.7% with an IRF of 25.0%, and the water cut could also be reduced from 100.0% to its lowest value of 75.0%. This confirms that CO₂ WAG injection could indeed overcome the deficiency of the single water flooding technique, improve the microscopic displacement efficiency, thus, increasing oil recovery. For the heterogeneous core, on the one hand, CO₂ WAG could improve the displacement efficiency in the initially swept high-permeability regions (as it behaves in the aforementioned homogeneous core); on the other hand, expand the sweep volume by virtue of the Jiamin effect [11, 26]. Such dual actions increase the CRF significantly from 22.2% to 49.5% with an IRF of 27.3%, and the water cut drops to its lowest value of 14.3%. This demonstrates the good adaptability of the CO₂ WAG method in removing oil from reservoirs (with such heterogeneity) after water flooding. For the fractured core, it is interesting to find such excellent performance of CO₂ WAG, the　CRF is increased from 7.9% to 51.3% with a significant IRF of 43.4%. The underlying mechanisms for this maybe include higher flow resistance in fracture (two layers of iron mesh), water imbibition, and CO₂-oil interactions (i.e. dissolution and extraction), etc. [11, 24-26]; however, this oil production process seems to be very long with a lower rate of oil production (Figure 4(a)) than that in the aforementioned homogeneous and heterogeneous cores. This may be attributed to the fact that the oil removal by water imbibition and CO₂-oil interaction mechanisms is usually very slow and time consuming [24-28].

Figure 4 Measured parameters during CO₂ WAG flooding in cores with different heterogeneities:

In summary, CO₂ WAG injection works well in recovering crude oil from the adopted models with their different heterogeneities, and is a promising method for EOR after previous water flooding. In heterogeneous and fractured cores, it even contributes to more oil production than previous water flooding, however, we cannot conclude in general that CO₂ WAG will perform as well as it does in experiments (when applied to in-situ tests, especially in fractured reservoirs), because its effectiveness may be influenced by the other factors, i.e. the permeability and matrix size. To be more specific, the oil recovery mechanisms of water imbibition and CO₂-oil interaction may be reduced with decreased permeability and increased matrix size [27-29], thus, influencing the WAG performance. So, such influences of permeability, matrix size, etc. should be considered before CO₂ WAG is used on a real reservoir.

3.3 Performance of joint water and CO₂ WAG flooding

3.3.1 Total oil recovery

The total oil recovery factor (TRF) obtained after the joint two-step water and CO₂ WAG flooding is demonstrated in Figure 5. It is found that, firstly, water or CO₂ WAG may perform differently according to the prevailing heterogeneity. To be　 more specific, water flooding contributes to more oil production than CO₂ WAG does under homogeneous condition, while under heterogeneous and fractured conditions the later outperforms the former. Secondly, the homogeneous core could be better developed by the joint water and CO₂ WAG, and returns the highest TRF of 66.7% over the heterogeneous and fractured cases (TRFs: 49.5% and 51.3%). Although the oil recovery factor could be significantly enhanced by two-step water and CO₂ WAG injection, a large amount of oil remains trapped (especially in the heterogeneous and fractured cores) and further EOR methods are needed, i.e. gel treatment.

Figure 5 Measured total oil recovery factor when using joint water and CO₂ WAG flooding in cores with different heterogeneities

3.3.2 Remaining oil distribution measured by nuclear magnetic resonance

In order to better understand the remaining oil distribution after injecting the joint water and CO₂ WAG, the cores were subjected to nuclear magnetic resonance (NMR) analysis after CO₂ WAG flooding. The measured results are shown in Figure 6 where the color reflects the signal produced by hydrogen atoms during NMR measurement which is proportional to the oil saturation level. Therefore, the dominant yellow coloration represents a high oil saturation level, while blue denotes a low level.

Figure 6 NMR images showing distribution of remaining oil after water and CO₂ WAG flooding

It is found that, for the homogeneous core, the color of the longitudinal and transverse sections is mainly blue and light yellow, implying a relatively low oil saturation therein. More importantly, its color distribution is uniform. This implies that the whole core volume has been sufficiently swept, and there are no bypassed regions with relatively enriched remaining oil, but for those heterogeneous and fractured cores, their color distribution differs significantly from that observed in homogenous specimens. To be more specific, for the heterogeneous core, the upper part (with high permeability) appears light blue, indicating a low residual oil saturation therein, while its lower part (with low permeability) is mainly light-yellow, and the oil saturation there is relatively high. This uniformity in color and remaining oil distribution imply that, although CO₂ WAG flooding is able to expand the macroscopic sweep volume, then, enhancing oil recovery (Figure 4) could barely form a complete or full sweep throughout the low-permeability region, and un-swept bypassed regions remain. Further evidence to support this lies in the fractured core, where the color in areas adjacent to the fracture is bright-blue with low residual oil saturation there, while the color in regions far away from the fracture is light-yellow with more oil trapped therein. This proves that such regions have not yet been fully swept by CO₂ WAG flooding. So as discussed above, CO₂ WAG method is found to be able to enhance oil recovery after water flooding, but it cannot form a complete sweep in those regions initially bypassed when using water. This might be responsible for the low oil recovery factors (5%-10%) observed in some field applications [10]. It also reminds us that further EOR methods are needed after CO₂ WAG flooding, i.e. gel treatment or a CO₂ huff-n-puff strategy [30-36].

4 Conclusions

In this research, water and CO₂ WAG flooding were used to recover oil from models with different heterogeneities. The effects of heterogeneity on water and CO₂ WAG flooding were determined, then, the capacity of CO₂ WAG in expanding sweep efficiency was assessed using visual NMR to ascertain the remaining oil distribution. The main conclusions can be summarized as follows:

1) Water flooding is able to recover oil from cores with different heterogeneities, but its effectiveness diminishes in the presence of heterogeneity, the more severe the heterogeneity, the poorer the efficacy of water flooding.

2) CO₂ WAG significantly enhances the oil recovery after water flooding, and is a promising EOR method for cores with different heterogeneities; however, it could barely form a complete or full sweep in those regions initially bypassed by water. Further EOR operations are needed and are deemed worthwhile after CO₂ WAG processing.

3) The homogeneous core could be better developed by the joint water flooding and CO₂ WAG techniques, and returns the highest TRF of 66.7% over the heterogeneous and fractured cases.

Contributors

WANG Ye-fei and DING Ming-chen provided the concept and edited the draft of manuscript. LI Zong-yang conducted the specific experiments. WANG Ye-fei conducted the literature review. ZHANG Shi-ming, LIU De-xin and DING Ming-chen wrote the first draft of the manuscript. All authors replied to reviewers’ comments and revised the final version.

Conflict of interest

WANG Ye-fei, LI Zong-yang, ZHANG Shi-ming, LIU De-xin and DING Ming-chen declare that they have no conflict of interest.

References

[1] HOLM L W, JOSENDAL V A. Mechanisms of oil displacement by carbon dioxide [J]. Journal of Petroleum Technology, 1974, 26 (12): 1427-1438. DOI: https://doi.org/ 10.2118/4736-PA.

[2] DING Ming-chen, WANG Ye-fei, WANG Wei, ZHAO Hai-long, LIU De-xin, GAO Miao. Potential to enhance CO₂ flooding in low permeability reservoirs by alcohol and surfactant as co-solvents [J]. Journal of Petroleum Science and Engineering, 2019, 182: 106305. DOI: https://doi.org/ 10.1016/j.petrol.2019.106305.

[3] HOLM L W. Status of CO₂ and hydrocarbon miscible oil recovery methods [J]. Journal of Petroleum Technology, 1976, 28(1): 76-84. DOI: https://doi.org/10.2118/5560-PA.

[4] LI Ze-ya, GU Yong-an. Soaking effffect on miscible CO₂ flooding in a tight sandstone formation [J]. Fuel, 2014, 134: 659-668. DOI: https://doi.org/10.1016/j.fuel.2014.06.024.

[5] HINDERAKER L, UTSETH R H, HUSTAD O S, KVANVIK B A, PAULSEN J E. RUTH-A comprehensive Norwegian R&D program on IOR [C]// SPE European Petroleum Conference. Milan, Italy, 1996: SPE 36844. DOI: https:// doi.org/10.2118/36844-MS.

[6] GAO Yun-cong, ZHAO Mi-fu, WANG Jian-bo, ZONG Chang. Performance and gas breakthrough during CO₂ immiscible flooding in ultra-low permeability reservoirs [J]. Petroleum Exploration and Development, 2014, 41(1): 88-95. DOI: 10.1016/S1876-3804(14)60010-0. (in Chinese)

[7] DING Ming-chen, YUAN Fu-qing, WANG Ye-fei, XIA Xi-ran, CHEN Wu-hua, LIU De-xin. Oil recovery from a CO₂injection in heterogeneous reservoirs: The influence of permeability heterogeneity, CO₂-oil miscibility and injection pattern [J]. Journal of Natural Gas Science and Engineering, 2017, 44: 140-149. DOI: https://doi.org/10.1016/j.jngse. 2017.04.015.

[8] AL-BAYATI D, SAEEDI A, MYERS M, WHITE C, XIE Q. CLENNELL B. Insight investigation of miscible SCCO₂ Water Alternating Gas (WAG) injection performance in heterogeneous sandstone reservoirs [J]. Journal of CO₂ Utilization, 2018, 28: 255-263. DOI: https://doi.org/ 10.1016/j.jcou.2018.10.010.

[9] DING Ming-chen, WANG Ye-fei, HAN Yu-gui, GAO Miao, WANG Ren-zhuo. Interactions in bypassed oil-CO₂ systems and their utilization in enhancing the recovery of bypassed oil [J]. Fuel, 2019, 237: 1068-1078. DOI: https://doi.org/ 10.1016/j.fuel.2018.10.019.

[10] CHRISTENSEN J R, STENBY E H, SKAUGE A. Review of the WAG field experience [C]// 1998 SPE International Petroleum Conference and Exhibition. Villahermosa, Mexico, 1998: SPE 71203. DOI: https://doi.org/10.2118/ 39883-MS.

[11] QIN Z, ARSHADI M, PIRI M. Micro-scale experimental investigations of multiphase flow in oil-wet carbonates. II. Tertiary gas injection and WAG [J]. Fuel, 2019, 257: 116012. DOI: https://doi.org/10.1016/j.fuel.2019.116012.

[12] PANDA M, AMBROSE J G, BEUHLER G, MCGGUIRE P L. Optimized EOR design for the eileen west end area, greater prudhoe bay [J]. SPE Reservoir Evaluation & Engineering, 2009, 12(1): 25-32. DOI: https://doi.org/10.2118/123030-PA.

[13] AFZALI S, GHAMARTALE A, REZAEI N, ZENDEHBOUDI S. Mathematical modeling and simulation of water-alternating-gas (WAG) process by incorporating capillary pressure and hysteresis effects [J]. Fuel, 2020, 263: 116362. DOI: https://doi.org/10.1016/j.fuel.2019.116362.

[14] ZEKRI A Y, NASR M S, ALSHOBA KYH A S. Evaluation of oil recovery by water alternating gas (WAG) injection oil-wet and water-wet systems [C]// SPE Enhanced Oil Recovery Conference. Kuala, Malaysia, 2011: SPE 143438. DOI: https://doi.org/10.2118/143438-MS.

[15] DONG M, FORAIE J, HUANG S, CHATZIS I. Analysis of immiscible water-alternating-gas (WAG) injection using micromodel [C]// Petroleum Society’s Canadian International Petroleum Conference. Calgary, Canada, 2002: Paper 2002-158. DOI: https://doi.org/10.2118/2002-158.

[16] KULKARNI M M, RAO D N. Experimental investigation of miscible and immiscible water-alternating-gas (WAG) process performance [J]. Journal of Petroleum Science and Engineering, 2005, 48: 1-20. DOI: https://doi.org/10.1016/ j.petrol.2005.05.001.

[17] MIRKALAEI S M M, HOSSEINI S J, MASOUDI R, ATAEI A, DEMIRAL B M R, KARKOOTI H. Investigation of different I-WAG schemes toward optimization of displacement efficiency [C]// SPE Enhanced Oil Recovery Conference. Kuala, Malaysia, 2011: SPE 144891. DOI: https://doi.org/10.2118/144891-MS.

[18] ZHANG Y P, SAYEGH S, HUGNG S. Enhanced heavy oil recovery by immiscible WAG injection [C]// Petroleum Society’s 7th Canadian International Petroleum Conference. Calgary, Canada, 2006: Paper 2001-014. DOI: https://doi.org/ 10.2118/2006-014.

[19] AFZALI S, REZAEI N, ZENDEHBOUDI S. A comprehensive review on enhanced oil recovery by water alternating gas (WAG) injection [J]. Fuel, 2018, 227: 218-246. DOI: https://doi.org/10.1016/j.fuel.2018.04.015.

[20] HUANG E, HOLM L. Effect of WAG injection and rock wettability on oil recovery during CO₂ flooding [J]. SPE Reservoir Engineering, 1988, 3(1): 119-129. DOI: https://doi.org/10.2118/15491-PA.

[21] CHEN S, LI H, YANG D, TONTIWACHWUTHIKUL P. Optimal parametric design for water alternating-gas (WAG) process in a CO₂-miscible flooding reservoir [J]. Journal of Canadian Petroleum Technology, 2010, 49(10): 75-82. DOI: https://doi.org/10.2118/141650-PA.

[22] KHAN M Y, MANDAL A. Analytical model for gravity segregation in WAG displacement recovery of inclined stratified reservoirs [J]. Journal of Petroleum Science and Engineering, 2020, 186: 106722. DOI: https://doi.org/ 10.1016/j.petrol.2019.106722.

[23] VALEEV A, SHEVELEV A. Design of WAG parameters [C]// SPE Russian Petroleum Technology Conference. Moscow, Russia, 2017: SPE-187843-MS. DOI: https://doi.org/10.2118/ 187843-MS.

[24] YANG Bin, KANG Yi-li, LU Xiao-feng, YOU Li-jun, ZHANG Hao, CHEN Zhang-xin. Experimental investigation of the pore shape factor in fluid imbibition model-taking the Longmaxi shale in Sichuan Basin as examples [J]. Journal of Petroleum Science and Engineering, 2020, 193: 107327. DOI: https://doi.org/10.1016/j.petrol.2020.107327.

[25] BENAVENTE D, LOCK P, CURA M, GARCIA D, ORDONEZ S. Predicting the capillary imbibition of porous rocks from microstructure [J]. Transport in Porous Media, 2002, 49(1): 59-76. DOI: https://doi.org/10.1023/A:10160 47122877.

[26] CROGH N A, EIDE K, MORTERUD S E. WAG injection at the statfjord field, a success story [C]// SPE 13th European Petroleum Conference. Aberdeen, UK, 2002: SPE 78348. DOI: https://doi.org/10.2118/78348-MS.

[27] DING Ming-chen, WANG Ye-fei, WANG Yang, GAO Miao, LIU De-xin, CHEN Wu-hua. Experimental investigation of bypassed-oil recovery via CO₂soaking and huff and puff injection: Effects of miscibility and bypassed-oil size [J]. Fuel, 2019, 248: 152-160. DOI: https://doi.org/10.1016/j.fuel.2019. 03.088.

[28] CAI Jian-chao, LI Cao-xiong, SONG Kao-ping, ZOU Shuang-mei, YANG Zhang-ming, SHEN Ying-hao, MENG Qing-bang, LIU Yang. The influence of salinity and mineral components on spontaneous imbibition in tight sandstone [J]. Fuel, 2020, 269: 117087. DOI: https://doi.org/10.1016/j.fuel. 2020.117087.

[29] KHOSRAVI M, BAHRAMIAN A, EMADI M, ROSTAMI B, ROAYAIE E. Mechanistic investigation of bypassed-oil recovery during CO₂ injection in matrix and fracture [J]. Fuel, 2014, 117: 43-49. DOI: https://doi.org/10.1016/j.fuel.2013. 09.082.

[30] GUO Jian-chun, LI Ming, CHEN Chi, TAO Liang, LIU Zhuang, ZHOU De-sheng. Experimental investigation of spontaneous imbibition in tight sandstone reservoirs [J]. Journal of Petroleum Science and Engineering, 2020, 193: 107395. DOI: https://doi.org/10.1016/j.petrol.2020.107395.

[31] YANG Zhao-peng, GAO Shu-sheng, YUE Xiang-an, LU Yu-gang, SUI Hong-gang, WEI Hao-guang. New method for plugging channeling in oil field by nitrogen foam [J]. Journal of Central South University, 2014, 21(2): 677-683. DOI: 0.1007/s11771-014-1988-2.

[32] FERNO M A, STEINSBO M, EIDE O, AHMED A, AHMED K, GRAUE A. Parametric study of oil recovery during CO2injections in fractured chalk: Influence of fracture permeability, diffusion length and water saturation [J]. Journal of Natural Gas Science and Engineering, 2015, 27: 1063-1073. DOI: https://doi.org/10.1016/j.jngse.2015.09. 052.

[33] ABDURRAHMAN M, HIDAYAT F, HUSNA U Z, ARSAD A. Determination of optimum CO₂water alternating gas (CO₂-WAG) ratio in Sumatera Light Oilfield [J]. Materials Today: Proceedings, 2021, 39: 970-974. DOI: https://doi.org/ 10.1016/j.matpr.2020.04.495.

[34] KHAN M Y, MANDAL A. Analytical model for gravity segregation in WAG displacement recovery of inclined stratified reservoirs [J]. Journal of Petroleum Science and Engineering, 2020, 186: 106722. DOI: https://doi.org/10. 1016/j.petrol.2019.106722.

[35] WANG G, PICKUP G, SORBIE K, MACKAY E, SKAUGE A. Numerical study of CO₂ injection and the role of viscous crossflow in near-miscible CO₂-WAG [J]. Journal of Natural Gas Science and Engineering, 2020, 74: 103112. DOI: https://doi.org/10.1016/j.jngse.2019.103112.

[36] GROENENBOOM J, KECHUT N I, MAR-OR A. Foam-assisted WAG: Injection strategies to optimize performance [C]// SPE/IATMI Asia Pacific Oil & Gas Conference and Exhibition. Jakarta, Indonesia, 2017: SPE-186991-MS. DOI: https://doi.org/10.2118/186991-MS.

(Edited by FANG Jing-hua)

中文导读

不同非均质性模型中CO₂水气交替注入后剩余油分布：核磁共振的可视化研究

摘要：水气交替(WAG)是注气提高采收率(EOR)中应用较为广泛的注入方式，其主要驱油机理、参数优化和影响因素等已有较多的研究。然而，有关其在不同非均质性条件下扩大波及体积能力的报道不多。在三块非均质性不同的岩心模型中，依次开展了水驱和CO₂-WAG驱，研究了非均质性对WAG提高采收率的影响。更为关键的是，对CO₂-WAG注入后利用核磁共振技术(NMR)对模型中剩余油的分布进行扫描，能够对不同非均质性模型中CO₂-WAG驱扩大波及体积的效果有直观判断。 结果表明，非均质性的出现会严重削弱水驱的有效性：非均质性越强，水驱效果越差。水驱后，CO₂-WAG驱在不同非均质性条件下均能较好地提高采收率，但是并不能对低渗区域形成完全的波及，驱替后仍然存在无法驱扫的区域。水驱和CO₂-WAG驱对均质岩心模型的开发效果较非均质模型好。

关键词：非均质油藏；提高采收率；CO₂水气交替；波及体积；剩余油；核磁共振

Foundation item: Project(KFJJ-TZ-2019-3) supported by the Open Project of Shaanxi Key Laboratory of Advanced Stimulation Technology for Oil & Gas Reservoirs, China; Projects(51504275, 51974344) supported by the National Natural Science Foundation of China

Received date: 2020-06-04; Accepted date: 2020-10-19

Corresponding author: DING Ming-chen, PhD, Associate Professor; Tel: +86-17854267636; E-mail: Dingmc@upc.edu.cn; ORCID: https://orcid.org/0000-0001-6558-6485