J. Cent. South Univ. (2020) 27: 531-541

DOI: https://doi.org/10.1007/s11771-020-4314-1

Water invasion and remaining gas distribution in carbonate gas reservoirs using core displacement and NMR

GUO Cheng-fei(郭程飞)^{1, 2 ,3}, LI Hua-bin(李华斌)^{1, 2}, TAO Ye(陶冶)^{1, 2},LANG Li-yuan(郎丽媛)^{1, 2}, NIU Zhong-xiao(牛忠晓)⁴

1. State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation, Chengdu University of Technology, Chengdu 610059, China;

2. College of Energy Resources, Chengdu University of Technology, Chengdu 610059, China;

3. Puguang Branch, Sinopec Zhongyuan Oilfield Company, Dazhou 635002, China;

4. The Third Oil Production Plant, PetroChina Huabei Oilfield Company, Cangzhou 062450, China

Central South University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract: Water invasion is a common phenomenon in gas reservoirs with active edge-and-bottom aquifers. Due to high reservoir heterogeneity and production parameters, carbonate gas reservoirs feature exploitation obstacles and low recovery factors. In this study, combined core displacement and nuclear magnetic resonance (NMR) experiments explored the reservoir gas-water two-phase flow and remaining microscopic gas distribution during water invasion and gas injection. Consequently, for fracture core, the water-phase relative permeability is higher and the co-seepage interval is narrower than that of three pore cores during water invasion, whereas the water-drive recovery efficiency at different invasion rates is the lowest among all cores. Gas injection is beneficial for reducing water saturation and partially restoring the gas-phase relative permeability, especially for fracture core. The remaining gas distribution and the content are related to the core properties. Compared with pore cores, the water invasion rate strongly influences the residual gas distribution in fracture core. The results enhance the understanding of the water invasion mechanism, gas injection to resume production and the remaining gas distribution, so as to improve the recovery factors of carbonate gas reservoirs.

Key words: core displacement; gas-water two-phase flow; recovery factor; nuclear magnetic resonance (NMR); remaining gas distribution

Cite this article as: GUO Cheng-fei, LI Hua-bin, TAO Ye, LANG Li-yuan, NIU Zhong-xiao. Water invasion and remaining gas distribution in carbonate gas reservoirs using core displacement and NMR [J]. Journal of Central South University, 2020, 27(2): 531-541. DOI: https://doi.org/10.1007/s11771-020-4314-1.

1 Introduction

Most of the gas reservoirs discovered and developed in China are water-drive gas reservoirs, among which the reservoirs with active edge- and-bottom aquifers account for 40%-50% [1]. Water invasion is a common occurrence in gas reservoirs with active edge-and-bottom aquifers, resulting in recovery factors of water-drive gas reservoirs lower than those for constant-volume gas reservoirs by 30%-40%. Therefore, water invasion greatly affects the production and development of gas reservoirs [2]. The study area in this paper is a carbonate water-drive gas reservoir located in the Northeast Sichuan basin, China, which features high heterogeneity and a diverse pore structure. The pore space of this reservoir mainly consists of intergranular dissolved pores, intragranular dissolved pores, and a few fractures, as shown in Figure 1. Because the heterogeneity of carbonate gas reservoirs is higher than that of sandstone gas reservoirs, the water invasion behavior is more complex [3-5].

Figure 1 Casting thin sections of Well 104-1 in work area:

At present, two wells have been shut due to water flooding, causing the production capacity to decrease by 40×10⁴ m³/d. It is predicted that 7 wells will be flooded and closed in the future, leading to a total loss of gas reserves of 68.25×10⁸ m³. Therefore, the water invasion behavior of the gas reservoirs was studied to reduce the influence of water invasion in the work area.

Core simulation experiments are among the most important methods for studying the behavior of water invasion [6]. Based on the seepage law and the similarity principle, the method simulates the development process of water-drive gas reservoirs using a core to reveal the behavior of fluid flow in porous media [7, 8]. Reservoir properties and production parameters were examined in the experiments to analyze the fluid flow characteristics and the gas recovery factor during water invasion [9-12]. Fortunately, some of the key work has been performed, such as aquifer, fracture, permeability and production rate, to understand the water invasion mechanism in carbonate gas reservoirs, and the gas-water two-phase relative permeability curves were calibrated to be used in the engineering calculation [13-15]. To improve the recovery factor of water-drive gas reservoirs, it is necessary to plug the watering-out gas wells and eliminate the water-lock effect near the well area by gas injection, restoring the gas well’s productivity [16-17]. However, injecting gas to restore production after water invasion is not fully understood, and many uncertainties in this process remain.

In terms of the experimental research on the remaining gas distribution, the conventional method is to use a microscopic visualized physical model, which displays an image of the microscopic gas-water distribution directly [18-20]. However, because the model is relatively simple and cannot withstand high temperatures and pressures, the method fails to reflect the seepage law under real reservoir conditions, and the image only reflects the microscopic distribution of fluid in two- dimensional space. Nuclear magnetic resonance (NMR) technology is widely used in geological fields, such as for reservoir structure characterization, reservoir type classification, and mobile water saturation determination, by identifying the fluid distribution in three- dimensional space [21-23].

In this paper, core displacement experiments and NMR technology were combined to study the gas-water two-phase flow behavior and the microscopic distribution of the remaining gas in the reservoir of the work area. The results can provide a better understanding of water invasion, gas injection and the remaining gas distribution to improve the recovery factor for carbonate gas reservoirs.

2 Experiment

2.1 Experimental principle

2.1.1 Experimental principle of displacement

The experimental method of core water flooding was used to simulate the water invasion process of the gas reservoirs. When gas was not produced at the outlet, the core was considered to be watering-out. Then, injecting gas to displace the watering-out core was performed to simulate restoring production of a gas well. The relative permeability of the gas-water two-phase flow and the gas recovery factors for different displacement processes were calculated by measuring the gas production, water production and injection pressure [6].

2.1.2 Experimental principle of NMR

The basic principle of NMR is to transmit a Carr-Purcell-Meiboom-Gill (CPMG) pulse sequence into the rock to obtain the attenuation signals of the spin echo strings from the hydrogen element and then carry out a mathematical inversion of the signals, finally forming the transverse relaxation time spectrum (T₂ spectrum), which reflects the distributions of water in different pore sizes or structures [24]. The x-coordinate of the T₂ spectrum represents the relaxation time, where the larger the pore is, the longer the relaxation time will be. The y-coordinate of the T₂ spectrum is the amplitude, which represents the proportion of a certain pore size.

2.2 Conditions and samples

A displacement apparatus and NMR on-line monitoring system for high temperatures and pressures were used in the experiments, which could conduct on-line measurements of the T₂ spectrum after displacements. The experimental temperature was 80 °C, and the confining pressure was 30 MPa. The apparatus was connected according to the experimental procedures, as shown in Figure 2. The major testing parameters of NMR included the dominant frequency (2.1 MHz), echo spacing (T_E=0.6 ms), polarization time (T_W=6 s), scanning times (NS=128), and echo number (NECH=4096). Core samples that were drilled from a coring well in the work area were prepared, and the fundamental parameters were tested according to the standard SY/T5345-2007. The fundamental core parameters are listed in Table 1. A fracture core was made by the method of artificial fracturing from a pore core. The experimental water was prepared according to the composition of the formation water of the reservoir, as listed in Table 2. A nitrogen gas source (N₂) was used to replace the natural gas, and the purity of N₂ was 99.999%.

Figure 2 Displacement and NMR experimental system:

Table 1 Fundamental parameters of cores

Table 2 Simulated formation water composition and total dissolved solids (TDS) (mg·L^-1)

2.3 Experimental procedures

2.3.1 Preparation

The core was saturated with formation water by the vacuum method; the core was weighed, and the T₂ spectrum of the core after being saturated with water was detected.

2.3.2 Establishment of initial irreducible water saturation

The steps to establish the initial irreducible water saturation (S_wi) of the core were as follows: connect the experimental apparatus, establish the experimental back pressure, and use N₂ to drive the water-saturated core at a constant flow rate of 0.01 mL/min until water was no longer produced at the outlet. Moreover, measure the quality of water production, calculate the S_wi and original N₂ reserves, and detect the T₂ spectrum of the core.

2.3.3 Simulating water invasion

The simulated formation water was used to displace the core at S_wi at a constant flow rate of 0.01 mL/min until gas was no longer produced at the outlet. Then, measure the quality of water production and the volume of gas production. The unsteady state method [6, 7] was used to calculate the relative permeability of the gas or water phase, the water-drive gas recovery factor was calculated, and the T₂ spectrum of the core after water invasion was detected.

The steps in Section 2.3.2 were repeated to establish S_wi under different pump velocity conditions to simulate various water invasion rates. The cores were reused, and the pump velocities were set as 0.02, 0.04, 0.08 and 0.16 mL/min.

2.3.4 Simulating production resumption by injecting gas

After water invasion at pump velocities of 0.01 mL/min and 0.16 mL/min, core 1-1 and core 2-1 were reused to inject gas at the same pump speeds. The quality of water production and the volume of gas production were measured. The unsteady state method [6, 7] was used to calculate the relative permeability of the gas or water phase, the irreducible water saturation after gas flooding (S_wg) was calculated, and the T₂ spectrum of the core was detected under the S_wg condition.

3 Results and discussion

3.1 Gas-water flow behavior during water invasion

3.1.1 Characteristics of relative permeability curves

The low-rate water invasion process was simulated by a pump speed of 0.01 mL/min. From the relative permeability curves of pore core 1-1 and fracture core 2-1 (Figure 3), when the water saturation (S_w) is lower than 40%, the relative permeability of the gas-phase (K_rg) in the fracture core is higher than that in the pore core. Then, with the increase in S_w, the K_rg in the fracture core decreases rapidly, while the K_rg in the pore core decreases gradually. Additionally, when S_w is higher than 47%, the K_rg in the fracture core is lower than that in the pore core. During the whole process of water invasion, the relative permeability of the water-phase (K_rw) in the fracture core is always higher than that in the pore core. When S_w is higher than 47%, the K_rw in the fracture core rises greatly, and the K_rw in the pore core rises rapidly at first and then tends to remain unchanged. These results indicate that although the water invasion rate is extremely low (0.01 mL/min), the water-phase relative permeability in the fracture core is higher than that in the pore core because the main seepage channels of the fracture core are fractures, and the gas-phase is almost completely driven out.

Figure 3 Relative permeability curves of pore core and fracture core

3.1.2 Effect of gas-water two-phase flow interval

Different water invasion velocities possess little influence on the two-phase seepage interval of a certain pore core. In general, the extent of the influence on any pore core is less than 2.6%. However, the impact on the infiltration interval decreases with the permeability, as shown in Figure 4. According to the experimental results of the three pore cores, the average two-phase flow intervals are 47.9%, 44.1% and 36.3%. This outcome indicates that the pore throats of the low-permeability core are usually small, having a strong binding effect on the fluid. So, the gas-water two-phase flow interval is relatively narrow and leads to a high displacement pressure.

Figure 4 Gas-water two-phase flow intervals

The fracture core presents the narrowest gas-water two-phase flow interval of all the cores at different displacement velocities. As the water invasion rate increases, the flow interval of the fracture core decreases by 14.2%, from 30.8% to 16.6%. Since fractures provide prior seepage channels for the water advance, the higher the water invasion rate is, the earlier the water breakthrough occurs, and the narrower the gas-water two-phase flow interval is.

3.2 Effect of water invasion on gas recovery factor

3.2.1 Gas recovery factor of pore core

With the increase in permeability, the water-drive gas recovery factor rises, as shown in Figure 5. For example, the average gas recovery factor of core 1-1 at different displace velocities is higher than that of core 1-3 by 7.2%. Moreover, the displacement velocity has a different effect on the recovery factors of the pore cores. For high-permeability core 1-1, the water flooding recovery factor rises monotonously with the displacement velocity, and when the water flooding rate is 0.16 mL/min, the recovery factor reaches a maximum value of 62.2%. The recovery factor of core 1-2, with medium permeability, increases rapidly to 60.0% (0.08 mL/min) at first and then decreases slightly to 59.2% (0.16 mL/min) later. In low-permeability core 1-3, a clear peak is associated with the water flooding recovery factor. When the displacement velocity increases from 0.01 to 0.02 mL/min, the recovery factor reaches a maximum value of 56.2%, increasing by 1.8%. However, with further improvements in the displacement velocity, the recovery factor gradually decreases to 52.3% (0.16 mL/min). Therefore, the permeability of the pore core is the main factor affecting recovery. For the high-permeability reservoirs, improving the water-drive rate may increase the fluid mobility and ultimately improve the water-drive gas efficiency. However, reservoirs with poor properties are affected by microscopic heterogeneities. With the increase in the water invasion rate, the gas-water interface become unstable, and the water advances rapidly along the main seepage channels [25], resulting in a decrease in the recovery factor. Thus, to obtain the ideal recovery factor of a gas reservoir area with poor properties, the gas production rate should be controlled reasonably to reduce water invasion. In contrast, the gas production rate should be enhanced in a gas reservoir area with good properties.

Figure 5 Water-drive recovery factors versus displacement velocity

3.2.2 Gas recovery factor of fracture core

The recovery factor of the fracture core is the lowest among the all cores at the different water flooding velocities, as shown in Figure 5. With the increase in displacement velocity, the gas recovery factor declines from 45.0% (0.01 mL/min) to 24.0% (0.16 mL/min), decreasing by 21.0%. Even when the displacement velocity is a low 0.01 mL/min, the gas recovery factor of the fracture core (matrix absolute permeability is 1.59 mD) is 9.4% lower than that of pore core 1-3 (absolute permeability is 0.78 mD). This result shows that the existence of fractures has a strong influence on the gas recovery. Moreover, with increase in displacement velocity, the waterline advances rapidly along the fractures, but little water flows into the matrix pores, leading to the lowest recovery factors of this type. Due to the rapid increase in K_rw, sharp decrease in K_rg and narrow two-phase seepage interval after water breaking through in the fracture core, the gas recovery factor of the fracture core decreases by 21.0% by water invasion, which is 10 times higher than that of the pore cores (2.1%). Considering that a fracture reservoir is seriously affected by water invasion, it is suggested that enough distance should be established between the perforated sections of a production well and gas-water interface to avoid the advance of early water breakthrough [13].

3.3 Gas-water two-phase flow behavior when injecting gas

3.3.1 Characteristics of relative permeability curves

To understand the seepage process of injecting gas in watering-out wells to resume production, core 1-1 and core 2-1 were converted to gas injection at pump speeds of 0.01 and 0.16 mL/min after water invasion. The gas-water two-phase flow curves of pore core 1-1 during gas flooding are illustrated in Figure 6. Compared with the invasion process, the phase permeability curve characteristics during the gas injection process present significant changes, among which K_rw is higher than that in water invasion, while K_rg is lower than that in water invasion. Due to the carbonate rock’s hydrophilicity, water is continuously distributed in the pores after the water influx, and the water flow is driven by gas immediately. Therefore, K_rw is improved during the gas injection process compared with that in the water invasion process. By contrast, the gas-phase is distributed in the core as a discontinuous phase, leading to K_rg in the gas injection process to be lower than that in the water invasion process. In addition, S_w at the isotonic permeability point decreases from 57% to 48% due to the increase in gas saturation in the core after gas flooding.

Figure 6 Relative permeability curves for pore core (0.01 mL/min)

For fracture core 2-1, the changes in relative permeability characteristics during gas flooding are similar to those of the pore core, but the decreases in fluid relative permeability are less than those of the pore core. This result indicates that restoring the seepage capacity is easier for the fracture core because the fractures are the main seepage channels, and the seepage resistance is low.

Therefore, gas injection is beneficial to drive out the mobile water, which is of great significance to reduce the water saturation and eliminate the water-lock effect near the well area, so as to restore the watering-out well’s productivity after plugging.

3.3.2 Changes in irreducible water saturation

The S_wg values of both the pore and fracture cores are higher than S_wi (Table 3), indicating that water invasion has an irreversible impact on the relative permeability of the different types of cores. Water displaces gas, and some gas will be trapped by the breaker, flow around and water lock [14, 26], leading to water retention in the core and improving S_wg; therefore, the process of injecting cannot completely restore the productivity to the stage before water invasion. In addition, the gas displacement velocity exhibits little influence on the difference between the two kinds of irreducible water saturation for pore core 1-1 but poses a great influence on fracture core 2-1. Therefore, for watering-out gas wells, injecting gas is beneficial to partially restore the relative permeability of the gas-phase and achieve success in resuming a well’s production.

Figure 7 Relative permeability curves of fracture core (0.01 mL/min)

Table 3 Irreducible water saturation

3.4 NMR results

3.4.1 T₂ spectra of typical rocks

Figure 8(a) shows the T₂ spectrum of a typical sandstone, reflecting the pore structure characteristics. The T₂ spectrum has a bimodal structure, the left peak is higher than the right one, and the lower envelope area of T₂<10 ms accounts for 60% of the total area, indicating that the proportion of small-scale pores in the sandstone is large. Figure 8(b) shows the T₂ spectrum of a typical carbonate rock with three peaks, reflecting the complex characteristics of the small, medium and large pores in the carbonate rocks. The NMR results of the carbonate rocks show that the pore structure characteristics are the same as the casting thin section results in Figure 1, indicating that the carbonate reservoirs have high heterogeneity and that the pore scale of carbonate rocks is widely distributed. Therefore, the gas-water two-phase flow behavior in carbonate reservoirs is very complicated.

Figure 8 T₂ spectra of typical rocks:

3.4.2 Characteristics of T₂ spectra in work area

NMR signals were detected after the four processes of water saturation, initial irreducible water saturation establishment, water invasion and gas injection, as shown in Figure 9. The characteristics of the T₂ spectra that reflect the changes in fluid because of different processes are as follows.

1) From the water-saturated T₂ spectra, the better the properties of the core are, the larger the proportion of large pores (T₂>100 ms) is, and the smaller the proportion of small pores (T₂<1 ms) is, and vice versa [8]. For pore core 1-3 with poor properties, the T₂>1000 ms signals are not detected, but the signals are present in fracture core 2-1.

Figure 9 T₂ spectra for different processes (0.01 mL/min):

2) Due to the carbonate rock’s hydrophilicity, the initial irreducible water saturation T₂ spectrum contains T₂<1 ms signals that are equal to those under water-saturated conditions, reflecting the kinds of very small pores or throats that are not involved in seepage because of the capillary pressure [27]. In addition, signals of irreducible water exist in macrosized or medium-sized pores (10 ms2<100 ms) because some water adsorbs to the pore walls of hydrophilic carbonate rocks [24], as shown in Figure 10.

3) Water invasion possesses different effects on the distribution and content of the remaining gas in the cores. The amplitude of the residual water T₂ spectrum after water invasion minus that of the water-saturated T₂ spectrum reflects the distribution and content of the gas remaining after water invasion. Because of the high gas recovery factors for pore core 1-1 and core 1-2 with good properties, the remaining gas is distributed in pores of various scales or sizes (10 ms2<1000 ms), and the proportion of remaining gas is small. For pore core 1-3 and fracture core 2-1, whose water-drive gas recovery factors are low, more than 80% of the remaining gas is distributed in pores with 10 ms2<100 ms, and the proportion of remaining gas is large. Furthermore, the residual water T₂ after the water invasion of core 2-1 shows that no gas remains in the fractures, indicating that water invasion completely floods the fractures and makes it difficult for the gas in the matrix to be driven out by water. Therefore, the gas recovery factor for the fracture core is the lowest among all.

Figure 10 Distribution model of irreducible water in pores and throat:

4) For core 1-1 and core 2-1, the amplitudes of the irreducible water saturation T₂ spectra after gas injection are higher than those of the initial irreducible water saturation T₂ spectra, indicating that water invasion has irreversible effects on the carbonate rocks. Because the matrix permeability of core 2-1 is lower than that of core 1-1, the irreducible water saturation and distribution of core 2-1 are larger and wider.

3.4.3 Remaining gas distribution

Figure 11 shows the T₂ variations in the residual water after water flooding at different displacement rates (0.01 and 0.16 mL/min) for core 1-1 and core 2-1, reflecting the influence of the displacement velocity on the water-drive gas efficiency and remaining gas distribution. The right peak of the residual water T₂ in core 1-1 at a rate of 0.01 mL/min moves slightly to the left at a rate of 0.16 mL/min, but the lower envelope area remains unchanged. This result indicates that the increase in displacement velocity has little influence on the gas recovery factor and its distribution. However, with the increase in water flooding velocity, the middle peak of the remaining water T₂ in core 2-1 moves to the left and down, and the lower envelope area significantly decreases. Therefore, to obtain the ideal gas recovery factor, the water invasion rate or gas production rate should be reasonably controlled to delay water breakthrough in a fracture reservoir. In addition, more than 66% of the remaining gas in the fracture core is distributed in pores with 10 ms2<100 ms. Therefore, the area where gas wells are shut in because of water flow along fractures is an important potential target to improve gas recovery.

Figure 11 Residual water T₂ spectra for pore and fracture cores

4 Conclusions

In this paper, four core samples, divided into pore type and fracture type, were designed to simulate the water invasion process, followed by the injecting of gas into the cores to simulate the process of restoring watering-out gas wells. Core displacement experiments and NMR technology were combined to investigate the gas-water two-phase flow behavior and the microscopic distribution of the remaining gas in carbonate gas reservoirs during the water invasion and gas injection. The conclusions are summarized as follows.

1) During water invasion, the water-phase relative permeability of the fracture core is higher and the co-seepage interval is narrower than that of the three pore cores.

2) The type of core, water invasion rate and permeability affect the water-drive recovery efficiency. Because of fractures providing a prior seepage channel, the fracture core presents the lowest efficiency among the all cores at different water invasion rates; additionally, as the invasion rate increases, the recovery factor decreases. For the pore cores, the water invasion rate slightly impacts the recovery; however, the lower the permeability is, the less the recovery factor.

3) Gas injection is beneficial for reducing the water saturation and for partially restoring the gas-phase relative permeability, especially for the fracture core.

4) The remaining gas distribution and content are related to the core properties. The water invasion rate yields a great impact on the distribution of the remaining gas in the fracture core, but a minor influence on that in the pore cores. Furthermore, a large amount of gas remains in the fracture core matrix, which is an important target to improve the gas recovery of reservoirs.

References

[1] LI Yang, KANG Zhi-jiang, XUE Zhao-jie, ZHENG Song-qing. Theories and practices of carbonate reservoirs development in China [J]. Petroleum Exploration and Development, 2018, 45(4): 152-162. DOI: 10.1016/S1876- 3804(18)30074-0.

[2] DURUCAN S, AHSAN M, SHI Ji-quan, SYED A, KORRE A. Two phase relative permeabilities for gas and water in selected European coals [J]. Fuel, 2014, 134: 226-236. DOI: 10.1016/j.fuel.2014.05.040.

[3] GUO Ping, ZHANG Hui-min, DU Jian-fen, WANG Zhou-hua, ZHANG Wan-bo, REN Hong-mei. Study on gas-liquid relative permeability experiments of fractured- porous reservoirs [J]. Petroleum, 2017, 3(3): 348-354. DOI:10.1016/j.petlm.2017.02.003.

[4] YE Zu-yang, LIU Hui-hai, JIANG Qing-hui, ZHOU Chuang-bing. Two-phase flow properties of a horizontal fracture: The effect of aperture distribution [J]. Advances in Water Resources, 2015, 76: 43-54. DOI: 10.1016/ j.advwatres.2014.12.001.

[5] WU Yu-shu, PAN L, PRUESS K. A physically based approach for modeling multiphase fracture-matrix interaction in fractured porous media [J]. Advances in Water Resources, 2004, 27(9): 875-887. DOI: 10.1016/j.advwatres.2004. 07.002.

[6] FANG Jian-long, GUO Ping, XIAO Xiang-jiao, DU Jian-fen, DONG Chao, XIONG Yu-ming, LONG Fang. Gas-water relative permeability measurement of high temperature and high pressure tight gas reservoirs [J]. Petroleum Exploration and Development, 2015, 42(1): 92-96. DOI: 10.1016/ S1876-3804(15)60010-6.

[7] BUCKLEY S E, LEVERETT M C. Mechanism of fluid displacement in sands [J]. Transactions of the AIME, 1942, 146(1): 107-116. DOI: 10.2118/942107-G.

[8] LI C, ZHOU X, YOU S, IBRAGIMOV J J. Analysis of two-phase gas-water flow in carbonate reservoirs [J]. Journal of Mining Science, 2018, 53(4): 643-654. DOI: 10.1134/S1062739117042631.

[9] HAN Yu-jiao, ZHOU Can-can, YU Jun, LI Chao-liu. Experimental investigation on the effect of wettability on rock-electricity response in sandstone reservoirs [J]. Fuel, 2019, 239: 1246-1257. DOI: 10.1016/j.fuel.2018.11.072.

[10] SHEN Wei-jun, XU Yan-mei, LI Xi-zhe, HUANG Wei-gang, GU Jiang-rui. Numerical simulation of gas and water flow mechanism in hydraulically fractured shale gas reservoirs [J]. Journal of Natural Gas Science & Engineering, 2016, 35: 726-735. DOI: 10.1016/j.jngse.2016.08.078.

[11] SHI Lei, GAO Shu-sheng, XIONG Wei. Physical simulation of the mechanism for operation of water-encroached (flooded) underground gas storage facilities [J]. Chemistry and Technology of Fuels and Oils, 2012, 47(6): 426-433. DOI: 10.1007/s10553-012-0319-2.

[12] FOURAR M, BORIES S, LENORMAND R, PERSOFF P. Two-phase flow in smooth and rough fractures: Measurement and correlation by porous-medium and pipe flow models [J]. Water Resources Research, 1993, 29(11): 3699-3708. DOI: 10.1029/93WR01529.

[13] LI Cheng-hui, LI XI-zhe, GAO Shu-sheng, LIU Hua-xun, YOU Shi-qiang, FANG Fei-fei, SHEN Wei-jun. Experiment on gas-water two-phase seepage and inflow performance curves of gas wells in carbonate reservoirs: A case study of Longwangmiao Formation and Dengying Formation in Gaoshiti-Moxi block, Sichuan Basin, SW China [J]. Petroleum Exploration and Development, 2017, 44(6): 983-992. DOI: 10.1016/S1876-3804(17)30110-6.

[14] FANG Fei-fei, SHEN Wei-jun, GAO Shu-sheng, LIU Hua-xun, WANG Qing-fu, LI Yang. Experimental study on the physical simulation of water invasion in carbonate gas reservoirs [J]. Applied Sciences, 2017, 697(7): 1-12. DOI: 10.3390/app7070697.

[15] FANG Fei-fei, SHEN Wei-jun, LI Xi-zhe, GAO Shu-sheng, LIU Hua-xun, LI Jie. Experimental study on water invasion mechanism of fractured carbonate gas reservoirs in Longwangmiao Formation, Moxi block, Sichuan Basin [J]. Environmental Earth Sciences, 2019, 316: 1-11. DOI: 10.1007/s12665-019-8325-x.

[16] CAI Jian-chao, YU Bo-ming, ZOU Ming-qing, LUO Liang. Fractal characterization of spontaneous co-current imbibition in porous media [J]. Energy & Fuels, 2010, 24(3): 1860-1867. DOI: 10.1021/ef901413p.

[17] FATH A H, POURANFARD A R, PARANDVAR R, POURHADI S. An investigation of different gas injection scenarios as enhanced condensate recovery method in a naturally fractured gas-condensate reservoir [J]. Petroleum Science and Technology, 2016, 34(3): 295-301. DOI: 10.1080/10916466.2015.1098663.

[18] SHI Lei, WANG Jie-ming, LIAO Guang-zhi, XIONG Wei, GAO Shu-sheng. Mechanism of gas-water flow at pore-level in aquifer gas storage [J]. Journal of Central South University, 2013, 20(12): 3620-3626. DOI: 10.1007/s11771-013- 1888-x.

[19] ZHOU H W, YUE Z Q, THAM L G, XIE H. The shape of moving boundary of fluid flow in sandstone: Video microscopic investigation and stochastic modeling approach [J]. Transport in Porous Media, 2003, 50(3): 343-370. DOI: 10.1023/a:1021183100914.

[20] LI Jun-jian, JIANG Han-qiao, WANG Chuan, ZHAO Yu-yun, GAO Ya-jun, PEI Yan-li, WANG Chen-chen, DONG Hu. Pore-scale investigation of microscopic remaining oil variation characteristics in water-wet sandstone using CT scanning [J]. Journal of Natural Gas Science and Engineering, 2017, 48: 36-45. DOI: 10.1016/j.jngse.2017. 04.003.

[21] WANG Heng, TABORDA E A, ALVARADO V, CORTES F B. Influence of silica nanoparticles on heavy oil microrheology via time-domain NMR T₂ and diffusion probe [J]. Fuel, 2019, 241: 962-972. DOI: 10.1016/j.fuel.2018. 12.103.

[22] VINCENT B, FLEURY M, SANTERRE Y, BRIGAUD B. NMR relaxation of neritic carbonates: An integrated and petrographical approach [J]. Journal of Applied Geophysics, 2011, 74(1): 38-58. DOI: 10.1016/j.jappgeo.2011.03.002.

[23] XIAO Liang, MAO Zhi-qiang, WANG Zhao-nian, JIN Yan. Application of NMR logs in tight gas reservoirs for formation evaluation: A case study of Sichuan basin in China [J]. Journal of Petroleum Science & Engineering, 2012, 81: 182-195. DOI: 10.1016/j.petrol.2011.12.025.

[24] JIANG Guan-cheng, LI Ying-ying, ZHANG Min. Evaluation of gas wettability and its effects on fluid distribution and fluid flow in porous media[J]. Petroleum Science, 2013, 10(4): 515-527. DOI: 10.1007/s12182-013-0303-4.

[25] LI Gao, REN Wen-xi, MENG Ying-feng, WANG Chuan-lei, WEI Na. Micro-flow kinetics research on water invasion in tight sandstone reservoirs [J]. Journal of Natural Gas Science and Engineering, 2014, 20: 184-191. DOI: 10.1016/ j.jngse.2014.06.024.

[26] HERRING A L, SHEPPARD A, ANDERSSON L, WILDENSCHILD D. Impact of wettability alteration on 3D nonwetting phase trapping and transport [J]. International Journal of Greenhouse Gas Control, 2016, 46: 175-186. DOI: 10.1016/j.ijggc.2015.12.026.

[27] PINI R, KREVOR S C M, BENSON S M. Capillary pressure and heterogeneity for the CO₂/water system in sandstone rocks at reservoir conditions [J]. Advances in Water Resources, 2012, 38: 48-59. DOI: 10.1016/j.advwatres.2011. 12.007.

(Edited by YANG Hua)

中文导读

碳酸盐岩气藏水侵模拟与剩余气分布的核磁共振实验研究

摘要：边底水活跃气藏发生水侵是一种常见现象。研究区为碳酸盐岩气藏，因储层非均质性强、生产任务重，气藏发生严重水侵，影响开发效果，并将导致气藏采收率低。本文采用岩心驱替实验和核磁共振技术相结合的方法，研究了碳酸盐岩气藏水侵过程和注气复产过程中，气液两相渗流规律以及剩余气微观分布规律。研究结果表明，在水侵过程中，裂缝型岩心较孔隙型岩心的液相相对渗透率高、共渗区窄；在不同水侵速度下，裂缝型岩心的采收率最低。在注气复产过程中，注气对裂缝型岩心降低含水饱和度、恢复部分气相相对渗透率的效果突出。剩余气量及其分布与岩心物性相关，水侵速度对裂缝型岩心剩余气的分布影响较大，但对孔隙型岩心的影响小。研究结果进一步提高了对碳酸盐岩气藏水侵机理、注气复产以及剩余气分布的认识，为提高该类型气藏采收率提供依据。

关键词：岩心驱替实验；气液两相渗流；采收率；核磁共振技术；剩余气分布

Foundation item: Project(2016ZX05017) supported by the China National Science and Technology Major Project

Received date: 2019-05-29; Accepted date: 2019-08-26

Corresponding author: LI Hua-bin, PhD, Professor; Tel: +86-13320963635; E-mail: lihuab1111@163.com; ORCID: 0000-0003- 2357-8297