The development of low-permeability reservoirs and tight-oil reservoirs is often unsatisfactory due to problems such as small pore radius and low permeability. The poor flow conditions lead to difficulties in water injection development. Laboratory experiments and field applications have shown that the injection of CO₂ into low-permeability reservoirs has a significant ability to enhance oil. In CO₂ injection, oil mixed with CO₂ has much increased mobility and reduces the interfacial tension. Therefore, CO₂ flooding could greatly improve oil recovery. CO₂ injection has been recognized as the second-largest enhanced oil recovery (EOR) process in the world.

CO₂ has attracted the attention of experts around the world because of its excellent oil displacement capacity. There is a long research history on CO₂ flooding in theory research, and it has been frequently used in applications in the field. Since the 1950s, a large amount of theoretical research and field application research has been carried out (Guo et al., 2002; Jiang et al., 2010; He et al., 2011). In the 1970s, the application of CO₂ flooding made great progress (Cao et al., 2020). China started research on CO₂ flooding theory and application in the 1960s. This technology has not been widely promoted in China because of the lack of CO₂.

Global warming is a major issue for the world, as increased amounts of CO₂ and other greenhouse gases in the atmosphere have resulted in increases in sea levels and changes in the climate. Many countries have proposed policies to reduce carbon emissions. In CO₂ flooding, the injected CO₂ is trapped underground, and it is an important way to reduce carbon emissions. Therefore, CO₂ flooding has the dual purpose of oil recovery and CO₂ reduction, and has broad application prospects.

Laboratory experiments of CO₂ flooding are applied to support the implementation of CO₂ flooding in the field (Xie, 1991; Cheng et al., 2008; Ma et al., 2020). Many experts have made great contributions in this field. The influence of CO₂ on the viscosity and volume of oil was studied by Hao et al. (2005) through laboratory experiments, and the mechanism of CO₂ EOR was researched. Miller and Jones studied the effect of CO₂ on the physical properties of heavy oil (Miller and Jones, 1981; Sankur and Emanuel, 1983). Gao et al. (2020) analyzed the influence of development parameters on the CO₂ EOR through experiments (Gao et al., 2020). Wang et al. (2016) discussed periodic CO₂ injection through experiments in low-permeability porous media, and the influence of injection slug and velocity is revealed in their work (Wang et al., 2016). The differences between miscible flooding and immiscible flooding were compared by Zhao et al. (2018), and field experiments were conducted to support the research. The above works mainly focus on the mechanism of CO₂ EOR, and the effect of fractures was not studied.

Oil development of low permeability and tight reservoirs is often affected by fractures. On the one hand, by fracturing the injection well, the injection rate of CO₂ is improved and the reservoir energy can be recovered in a short time. On the other hand, the oil production well is usually fractured to improve the production rate in low-permeability reservoirs. As a high-conductivity channel, fractures are closely related to important parameters such as the swept volume and gas channeling during reservoir development. It is necessary to carry out experimental research on the effect of fractures on CO₂ flooding. In order to explore the effect of fractures on CO₂ flooding, Huang used cores with annulus to represent fractures (Huang et al., 2008). An experimental study of CO₂ huff and puff was conducted, showing that fractures had a great influence on the CO₂ huff and puff. In order to optimize the injection methods of CO₂ flooding after water flooding, CO₂ compound flooding experiments with single fracture and complex fracture in a five-spot well pattern were implemented, respectively (Shi et al., 2021). The physical model of fracture radial flow was used in their work. To develop fractured tight reservoirs effectively, Wang (2017) designed a CO₂ flooding experiment to analyze pressure transfer in fractured porous media (Wang, 2017). In this experiment, the fractures were formed with different apertures. By comparing the CO₂ flooding results with the control experiment, it was shown that the gas channeling caused by fracture was the key factor in the production of oil in the matrix. Existing experiments show that fractures have a significant impact on the development of CO₂ flooding in low-permeability and tight reservoirs. However, further research is required on pressure distribution, fluid composition, and fluid displacement.

In this study, artificial cores were made according to the formation parameters in Shengli Oilfield. Fracture length and aperture are the primary factors affecting CO₂ flooding. Therefore these fractures were made with different lengths and apertures. During CO₂ flooding, the pressure along the core and the components at the outlet was recorded. Then the influence of fractures on CO₂ flooding was investigated.

The experimental equipment contained UQT-120-70 high temperature and high pressure long-core test system, Bruker 450 GC, and Agilent 7890A GC as the main components. The pressure was measured with SENEX DG2113-C-70/B. The pressure measurement signal was automatically recorded by the computer data acquisition system and was set to be recorded every 2 min. The cores used in this experiment were artificially produced according to the formation parameters of Shengli Oilfield. After cleaning and drying, the basic parameters of the cores are shown in Table 1. The fractures were made by wire cutting, and quartz sand was used as a filler to support the fractures as shown in Figure 1.

Since it is difficult to obtain a single core longer than 0.5m, the commonly used method is splicing conventional cores in a certain arrangement. Filter paper was applied to connect short cores to eliminate the terminal effect. The arrangement of each core was as follows.

1) The harmonic mean permeability k ¯ of the n cores was calculated using Eq. 1. A comparison was made between k ¯ and the permeability of the n cores, to find the core whose permeability was closest to k ¯ . Then this core was put in the first place at the outlet of the experimental device:

Four kinds of cores were considered in this experiment: the control group (without fracture in cores), a core with fracture length 10 cm and aperture 0.05mm, a core with fracture length 20 cm and aperture 0.05 mm, and a core with fracture length 20 cm and aperture 0.1 mm. By carrying out CO₂ flooding experiments on different types of cores, the influences of fracture length and aperture on the pressure distribution, composition, and displacement effect during CO₂ flooding were studied. The fluid used in this experiment was obtained by compounding the oil and gas samples from the surface separator of a well in the block.

The experimental steps mainly included the following: ① Firstly, the core was cleaned with petroleum ether and dried with nitrogen, then dead oil was injected with a pump to fully saturate the core, and keep the pressure and temperature at formation conditions for some time. After the core was fully saturated, the amount of dead oil was recorded. ② The pressure and temperature at the outlet were set to 20 MPa and 50°C, respectively. Then the prepared live oil was used to displace the dead oil in the core until the gas-oil ratio at the inlet and outlet was the same to ensure that the core was completely saturated with live oil. ③ Industrial-purity CO₂ (99.9%) was injected at a constant rate to displace the oil in the core until no more oil was driven out at the outlet. The data record changed over time, including oil production, gas production, pressure at the measuring points, etc. The produced gas and oil samples were collected for chromatographic analysis. ④ After each experiment was finished, the cores were cleaned, dried with nitrogen, evacuated with a vacuum pump, and steps ① to ③ were repeated for the next set of experiments.

A CO₂ flooding experiment in a long core was designed. The artificial cores and live oil were prepared according to the reservoir parameters in Shengli Oilfield. The inlet was injected with CO₂ at a constant speed of 0.2 ml/min. The pressure distribution along the core and components of the gas phase and oil phase at the outlet was tested. The results demonstrate the following:

1) The CO₂ flooding experiments on the matrix cores show that the pressure of each measuring point decreased rapidly with time in the early stage. After CO₂ broke through the core, the injection-production pressure difference decreased with the decrease of flow resistance. It indicates that constant speed displacement could be achieved without high injection pressure after the gas broke through. The proportion of CH₄ in the gas phase at the outlet exceeded 70% when CO₂ broke through the core. Then the proportion of methane continued to decrease, and the proportion of CO₂ continued to increase until the gas component at the outlet was mainly CO₂. The proportion of light and intermediate components increased after CO₂ broke through the core.