Low permeability-tight oil and gas resources play an important role in the global energy. Since the 21st century, the proportion of proved reserves of low permeability-tight oil and gas in the total annual new proved reserves has increased from 35% to 70% (according to statistics in 2014)^[1]. In the past five years, the proportion of low permeability-tight oil and gas reserves in proved oil and gas reserves in China has reached 70%-80%^[2,3]. Low permeability-tight oil and gas has gradually become the main part of oil and gas development in China, and its proportion in oil and gas production has been increasing year by year. However, the low permeability-tight reservoir has the characteristics of low porosity, low permeability, small pore throats, and complex pore structure, etc., resulting in difficult recovery, low recovery rate, and fast decline in production, etc., making exploration and development more difficult. At present, problems such as high start-up pressure gradient and insufficient natural energy are generally existed in the development of low permeability-tight oil reservoirs, and it is necessary to supplement formation energy by high-pressure water injection and advanced water injection^[4,5]. However, due to poor reservoir properties and other reasons, long-term water injection leads to continuous increase of formation pressure around the water injection well^[6], resulting in high-pressure and under-injection. Meanwhile, excessively high pressure would induce micro-fractures in the formation, and with the dynamic extension of the fractures, the oil well will be at risk of burst flooding, which will seriously affect the effect of oilfield development^[7].

In recent years, nanotechnology has developed rapidly and has been widely used in many fields such as biology, medicine, aviation, military and energy^[8,9]. Researchers at home and abroad have tried to apply nano-materials in many fields of petroleum industry, especially in the field of low permeability-tight oil and gas development^{[10,11,12,13,14,15,16,17,18,19]}. Miranda et al.^[18] studied the surface wettability of reservoirs and fluid diffusivity using molecular dynamics from the molecular level, then examined the stability and rheological property of nano-silica particles modified by different functional groups in media of different salinities, and finally analyzed the potential nanoparticle systems for oil displacement from the point of lowering interfacial tension between oil and nanoparticles. Ayatollahi et al.^[19] introduced the applications of nanotechnology in improving oil recovery. But at present, most of these technologies are mainly in the stage of laboratory research, focusing on the characteristics of nano-materials themselves, instead of in-depth research in oil displacement mechanism, and the amplitude of enhanced oil recovery is limited.

Aiming at the key technical problems of further enhancing oil recovery in low permeability-tight oil field, this paper studies the effect of nano-particle iNanoW1.0 on expanding swept volume of water flooding in ultra-low permeable core by using low-field nuclear magnetic resonance technology, and the mechanism of expanding swept volume of water flooding was analyzed through the experiments of oxygen spectrum nuclear magnetic resonance and capillary analysis.

The experimental cores were taken from the Jurassic outcrop sandstone in Sichuan Basin, with diameter of 2.5 cm, length of 4.8-5.0 cm, and gas permeability of (1.21-1.31)× 10^-3 µm². The sample parameters are shown in Table 1. Deuterium water was purchased from Beijing Funuo Technology Development Co., LTD, with a purity of 99.9%±0.02%. The simulated formation water was prepared by purified water or deuterium water with a salinity of 5 000 mg/L (NaCl:CaCl₂= 9:1). iNanoW1.0 nanoparticles, with particle size of 10-50 nm and interfacial tension of 1.65 mN/m, were self-developed.

LF-NMR core displacement device, mainly including MR- dd high-temperature and high-pressure displacement device (produced by Nantong Huaxing Petroleum Instrument Co., LTD.), MesoMR23-060H-HTHP core low-field nuclear magnetic analyzer (produced by Shanghai Newmark Electronic Technology Co., LTD). JNM-ECA600 nuclear magnetic resonance spectrometer (600 MHz) made by Japan Electronics Co., LTD. Capillarity analysis system, self-developed.

1.3.1. The principle

As shown in Fig. 1, LF-NMR core displacement device is mainly composed of a high-temperature and high-pressure displacement device, a NMR device, a control unit and a metering unit. NMR measurement of rock samples containing water (oil) with the LF-NMR core displacement device can obtain the NMR relaxation signal (T₂ spectrum) of ¹H-proton- bearing fluid in the pores of the rock samples^[20]. Because of different sizes of pores in the rock, the measured data is actually the result of the superposition of several transverse relaxation components.

The T₂ spectrum obtained by NMR reflects the spatial distribution of ¹H-proton-bearing fluid in the rock sample. The longer the relaxation time, the larger the pore diameter is, and the shorter the relaxation time, the smaller the pore diameter is. The peak area enclosed by signal amplitude and relaxation time represents the volume of ¹H-proton-bearing fluid contained in the pore. The larger the peak area, the larger the volume of ¹H-proton-bearing fluid contained in the pore is; conversely, the smaller the volume of ¹H-proton-bearing fluid^[21]. Fig. 2 shows the NMR T₂ spectrum of extra-low permeability outcrop core saturated with water. The area of the left peak (P₁ peak) represents the volume of water in small pores, the area of the middle peak (P₂ peak) represents the volume of water in medium pores, and the area of the right peak (P₃ peak) represents the volume of water in large pores. Therefore, in this experiment, the distribution and content of ¹H-proton- bearing fluid in different pores were obtained by measuring the T₂ spectrum of ¹H-proton-bearing fluid in the core.

The principle of online NMR test is to obtain a series of T₂ spectra by real-time test of ¹H-proton-bearing fluid in the core during displacement process using the LF-NMR technology. Real-time saturation of ¹H-proton-bearing fluid in core can be analyzed based on T₂ spectrum. Since deuterium (²H) protons in deuterium water have no signal in LF-NMR, deuterium water can be distinguished from water as the medium saturating core. The experiments were carried out on the MesoMR23- 060H-HTHP core nuclear magnetic analyzer. The main test parameters were: magnetic field intensity of 0.5 T, resonance frequency of 21 MHz, probe coil diameter of 70 mm, echo time of 0.3 ms, recovery time of 3 000 ms, cumulative number of 16, echo number of 8000, T₂ spectrum fitting points of 100.

1.3.2. Experimental procedures

(1) Deuterium water and purified water were used to prepare simulated formation water with a salinity of 5000 mg/L. The simulated formation water prepared with purified water was used to prepare iNanoW1.0 nano-sized oil-displacement agent with a mass fraction of 0.1%; (2) The simulated formation water prepared with deuterium water was injected into the core at a constant rate of 0.05 mL /min till the core was saturated, and the stable injection pressure was obtained; (3) The deuterium water in the core was displaced by the simulated formation water prepared with purified water at the steady injection pressure obtained in step (2) until the T₂ NMR spectrum curve no longer changed and the signal quantity stop increasing; (4) On the basis of step (3), the iNanoW1.0 nano-sized oil-displacement agent was continuously injected at the constant pressure until the NMR T₂ spectrum curve no longer changed and the signal quantity no longer increased.

1.4.1. The principle

Water in nature does not exist in single molecular form, but forms a multi-molecular network structure through hydrogen bond association^[22]. The width of the ¹⁷O-NMR spectral line can reflect the average relative size of the water molecule network structure. The wider the spectral line, the larger the network structure is, and the stronger the association of hydrogen bonds is; The narrower the spectral line, the smaller the network structure is, and the weaker the association of hydrogen bonds is^[23]. The NMR spectrum is not at a certain frequency, but presents a distribution with a certain width. The width of the spectral line is measured by the full width at the semi-maximum intensity (i.e., the half-peak width). The ¹⁷O-NMR spectral line width of pure water is shown in Fig. 3.

1.4.2. The procedures

(1) The IiNanoW1.0 dispersion with mass fractions of 0.005%, 0.010%, 0.050%, 0.100%, 0.300% and 0.500%, respectively were prepared with purified water; (2) ¹⁷O-NMR spectrum diagram of iNanoW1.0 dispersion with different mass fractions were tested, and compare with ¹⁷O-NMR spectrum diagram of pure water, with the nuclear magnetic resonance frequency of 82 MHz and temperature of 25 °C.

1.5.1. The principle

As shown in Fig. 4, the self-developed capillary analysis system can be used to measure the variation of injection differential pressure and flow rate in the process of capillary (beam) displacement. It is mainly compose of an injection system, a capillary (beam) model, a microscopic observation system, a miacroflow measuring system and a data processing system^[24]. The capillary (beam) consists of several capillaries with equal or unequal diameters in parallel, which can be used to simulate reservoirs with different permeabilities by changing the size and arrangement of the capillaries.

1.5.2. The procedures

Distilled water and iNanoW1.0 nano-sized oil-displacement agent with a mass fraction of 0.5% (filtered repeatedly through a 0.45 m pore size filter) were injected into two hydrophilic capillary tubes with 60 cm in length and 2.0 μm in inner diameter respectively, at a constant rate of 0.1 mL/min and room temperature until liquid flowed out of the end of the capillary tube, and the change of injection differential pressure with time was recorded.

Fig. 5 shows the injection pressure tracking curve of core LA-10-1 at the constant flow rate of 0.05 mL/min when the core was saturated with deuterium water. It can be seen from the curve that the injection pressure increases with the increase of injected deuterium water. After about 8 h, the core injection pressure finally stabilized at 0.49 MPa, which was determined as the injection pressure of the core constant pressure displacement experiment. The simulated formation water was injected into the core LA-10-1 saturated by deuterium water at a constant pressure of 0.49 MPa, and the T₂ spectrum obtained by online monitoring was shown in Fig. 6. In the T₂ spectrum, the peak with relaxation time less than 10 ms (P₁ peak) represents the small pores in the core. The peak (P₂ peak) with a relaxation time between 10 and 100 ms represents the medium pores in the core. The peak (P₃ peak) corresponding to the relaxation time greater than 100 ms represents the large pores in the core. It can be seen from the spectrum, when the simulated formation water injection volume increased from 0.25 PV (injected pore volume multiple) to 1.50 PV, the strength of P₁, P₂ and P₃ peaks of core LA-10-1 all showed an increasing trend. The peak strength of P₁, P₂ and P₃ did not change after the injection volume was greater than 1.50 PV. This indicates that when the injection volume is 1.50 PV, core LA-10-1 reaches displacement equilibrium; at the injection pressure of 0.49 MPa, the swept volume of simulated formation water in core LA-10-1 no longer increases as injection volume increases further. The peak area of P₁ is much larger than that of peak P₂ or P₃, mainly because core LA-10-1 is of ultra-low permeability (1.21×10^-3 µm²) with mainly small pores. In addition, it can be seen from the T₂ spectrum that when the injection volume reaches 0.50 PV, the P₃ peak signal intensity corresponding to the large pores woudln’t increase anymore. When the injection volume reaches 1.25 PV, the signal intensity of P₂ peak corresponding to the medium pores wouldn’t increase anymore. According to the above data, it can be inferred that the simulated formation water injected into the core reaches displacement equilibrium in the order of large pores, medium pores and small pores.

On the basis of simulated formation water displacement reaching balance, the iNanoW1.0 nano-sized oil-displacement agent with a mass fraction of 0.1% was injected into the core LA-10-1 at a constant pressure of 0.49 MPa, and the T₂ spectrum obtained by online monitoring of LF-NMR was shown in Fig. 7. Increasing iNanoW1.0 nano-sized oil-displacement agent volume from 0.5 PV to 1.5 PV, the peak signal intensity of P₂ and P₃ in the T₂ spectrum did not change, while the peak signal intensity of P₁ increased. When the injection volume was greater than 1.5 PV, the signal strength of P₁, P₂ and P₃ peaks no longer changed anymore, indicating that when the displacement pressure is 0.49 MPa and the iNanoW1.0 nano-sized oil-displacement agent injection volume is 1.5 PV, the core LA-10-1 has reached the displacement equilibrium. The increase of P₁ peak signal strength indicates that iNanoW1.0 nano-sized oil-displacement agent can further expand the swept volume in small pores on the basis of simulated formation water displacement balance.

In order to further compare the displacement efficiency of iNanoW1.0 nano-sized oil-displacement agent and simulated formation water, the offline test method was adopted to eliminate the interference of pressure, residual liquid and other factors, and the T₂ spectrum of iNanoW1.0 nano-sized oil-displacement agent and simulated formation water when the displacement of core LA-10-1 has reached equilibrium under constant pressure were collected (Fig. 8). By comparing the T₂ spectra, it can be seen that the peaks of P₃ basically coincide. The signal intensities of P₁ and P₂ of iNanoW1.0 nano-sized oil-displacement agent are both higher than those of simulated formation water, and the signal intensity increment of P₁ is more obvious. The above results show that iNanoW1.0 nano-sized oil-displacement agent has a larger swept volume in the small pores of core LA-10-1 than the simulated formation water, which means that it can increase the swept volume of small pores more significantly. The offline T₂ spectral peak area data of simulated formation water and iNanoW1.0 nano-sized oil-displacement agent during displacement are in Table 2. According to equation (1), it was calculated that iNanoW1.0 nano-sized oil-displacement agent increased displacement swept volume by 21.5% than simulated formation water:

The core LA-5-1 and LA-5-2 were displaced with iNanoW1.0 nano-sized oil-displacement agent, similar calculation was carried out for the effect of expanding swept volume. The results showed that the swept volume of the cores increased by 10.5% and 18.9% respectively than simulated formation water (Table 3). Due to heterogeneity, the cores LA-5-1 and LA-5-2 with the same gas permeability have different swept volumes.

2.2.1. Oxygen spectrum nuclear magnetic resonance experiments

Low field nuclear magnetic resonance core displacement experimental results show that the iNanoW1.0 nano-sized oil-displacement agent can increase swept volume of small pores on the basis of the simulated formation water flooding. In order to further study the mechanism of this effect, from the perspective of network structure of water molecules, the ¹⁷O-NMR test was applied to analyze the influence of iNanoW1.0 nano-sized oil-displacement agent on hydrogen bond association of water molecules. The results showed that the iNanoW1.0 dispersion with a mass fraction of 0.1% can reduce the ¹⁷O-NMR spectral line half-peak width of pure water from 123.94 Hz to 65.13 Hz (Fig. 9), and the half-peak width reflected the strength of water molecular hydrogen bond association^[23]. Thus it can be seen that iNanoW1.0 nanoparticles can effectively weaken the hydrogen bond association between water molecules, thus changing the structure of water molecular network, producing "small molecular water" that can enter the small pores of ultra-low permeability reservoirs, and further expanding the swept volume on the basis of conventional water flooding.

Fig. 10 shows the weakening effects on hydrogen bond association by iNanoW1.0 dispersion with different mass fractions. It can be seen, when the mass fraction of iNanoW1.0 increases from 0.01% to 0.10%, the corresponding half-peak width of ¹⁷O-NMR spectral line gradually narrows and shows a declining trend. When the concentration reaches 0.10%, the corresponding half-peak width of ¹⁷O-NMR spectral line tends to be stable with the increase of iNanoW1.0 mass fraction.

2.2.2. Capillary analysis experiments

In order to further verify the above conclusions, the capillary analysis system was used to compare the changes of the differential pressure with time during the injection of iNanoW1.0 dispersion (mass fraction of 0.50%) and distilled water in the capillary (Fig. 11). As the injection time gets longer, both the injection differential pressure of iNanoW1.0 dispersion and distilled water increase, mainly because they need to overcome the capillary resistance. In the process of the injection, the injection differential pressure of iNanoW1.0 dispersion is always lower than that of distilled water, indicating that the ability to reduce capillary resistance of iNanoW1.0 dispersion is stronger, which is good for expanding the swept volume. The results indicate the change in the structure of water molecule network, and further verify the LF-NMR results that the iNanoW1.0 nano-sized displacement agent can increase swept volume of small pores.

On the basis of simulated formation water reaching displacement equilibrium, iNanoW1.0 nano-sized oil-displacement agent can further increase 10%-20% of the swept volume in medium and small pores in extra low permeability core, and the reason is that iNanoW1.0 nanoparticles can effectively weaken the hydrogen bonding between water molecules, change the network structure, and make water easier to enter the low permeability area can’t be swept by conventional water flooding. The ability to weaken the association of hydrogen bonds improves with the increase of the mass fraction of iNanoW1.0 nanoparticles, and tends to be stable after the mass fraction of iNanoW1.0 nanoparticles reaching 0.10%. The study result has laid a theoretical foundation for expanding swept volume with nano-sized oil-displacement agent, and provided an important theoretical and experimental reference for water flooding development of low permeability-tight reservoirs.

I—swept volume increase rate, %;

S_H, S_N—T₂ spectral peak area of simulated formation water and nano-sized oil displacement agent, dimensionless;

T₂—transverse relaxation time, ms.