Study on the Influence of Oxygen Concentration and Fracture Characteristics on In Situ Combustion Development of Low–Medium Maturity Shale Oil Reservoirs Authors: Xing Jin, Wen Fang, Yang Wen, Yuanyuan Bai, Zezhou Chen, Shuai Zhao, Tao Li, Wanfen Pu Published: 2025 ABSTRACT In situ combustion (ISC) technology has been proven feasible in low–medium mature shale oil reservoirs. However, the impact of the oxygen concentration and fracture characteristics on development effectiveness is still unclear. Combustion tubes, gas chromatography (GC), and GC–mass spectrometry (GC–MS) are used to conduct ISC experiments on low–medium maturity shale oil. The impact of millimeter-scale fractures, oxygen concentration, and fracture characteristics on reservoir recovery is quantitatively analyzed using nuclear magnetic resonance (NMR) technology and a newly proposed shale oil recovery calculation method. Research has found that there is difficulty in ignition during ISC in millimeter-sized fractures. However, with the increase of the oxygen concentration and complexity of fractures, the ISC reaction becomes more complete, significantly improving the oil recovery of the reservoir. When the complexity of fractures and oxygen concentration reached 40%, the conversion of kerogen reached 90.6%, and the oil recovery of the reservoir increased to 60.6%. This indicated that high oxygen concentration can accelerate the oxidation reaction of organic matter and improve combustion stability, while fracture complexity increases the contact area between oxygen and organic matter, improves heating efficiency, and provides channels for the flow of modified oil. In addition, ISC can enhance the quality of shale oil and promote the flow and production of modified oil. The significance of this study lies in revealing the key influencing factors of oxygen concentration and fracture characteristics on ISC development of low–medium maturity shale oil reservoirs and proposing methods to improve the efficiency of successful ignition and combustion front advancement through chemical ignition agents and optimized fracture morphology. The research results provide a theoretical basis and practical guidance for optimizing the ISC process of low–medium maturity shale oil reservoirs. Introduction China has abundant terrestrial shale oil resources, which are important replacement resources for conventional oil and gas. Its technically recoverable reserves are about 845 ∼ 1183 × 108 tons, which can be further divided into medium–high maturity shale oil and low–medium maturity shale oil according to the degree of thermal maturity evolution. The technically recoverable reserves of low–medium maturity shale oil are 700 ∼ 900 × 108 t, accounting for more than 75% of the recoverable reserves of continental shale oil in China. They are mainly distributed in the Triassic system of the Ordos Basin, the Permian system of the Junggar Basin, the Paleogene system of the Cangdong Sag in Bohai Bay Basin, and the Cretaceous system of the Songliao Basin. − Due to the characteristics of rapid sedimentary phase transition, low maturity, high clay mineral content, strong heterogeneity, and high viscosity/density of similar fluids in terrestrial shale oil reserves, the commercial development model of the reservoir is still in the research stage. , Zhao et al. believe that in situ conversion technology is the optimal method for the conversion and development of terrestrial shale oil in China. Shell’s in situ conversion technology, in situ pyrolysis conversion (ICP), is quite mature. An oil shale in situ conversion pilot test conducted in the South Mahogany test area of Green River shale in the United States has been successful by Shell. , However, ICP technology has high requirements for shale oil reservoir conditions. When evaluating oil shale developed in Jilin Province, the cooperation is abandoned due to parameters such as oil content abundance and net-to-ground ratio being lower than the lower limit standards of Shell’s ICP. In addition, Shell has also collaborated with China National Petroleum Corporation to propose horizontal well lightweighting technology for low–medium maturity shale oil reservoirs in the Ordos Basin, which involved upgrading solid organic matter into flowable oil and gas through electric heating after horizontal fracturing of shale oil reservoirs. However, the collaboration has been abandoned due to geological conditions. There are still difficulties in using ICP technology for the development of Chinese terrestrial low–medium maturity shale oil reserves. Taiyuan University of Technology, Jilin University, and Jilin Zhongcheng Company have conducted a series of studies on in situ conversion methods for Chinese terrestrial shale oil, such as hot steam injection, supercritical water injection­(SCW), local chemical reaction method, high-voltage power frequency electric heating method (HVF), autothermic pyrolysis in situ conversion process (ATS), and fracturing combustion method. The research indoor and in-field has shown that the fracturing combustion method is currently one of the most suitable methods for the in situ transformation and development of shale oil reservoirs. This method involves injecting air for combustion after fracturing shale oil reservoirs to achieve in situ heating of the reservoir, thereby achieving transformation and development of kerogen. In the field of oil and gas reservoir development, this method is often referred to as the ISC method and has been successfully applied to the commercial development of heavy oil reservoirs. , In addition, Qin and Wan from Xinjiang Oilfield in China study the development effect of ISC of air injection between fractures in low–medium maturity shale oil reservoirs through numerical simulation methods, confirming the feasibility of ISC development in low–medium maturity shale oil reservoirs. Yuan et al. from Kazan Federal University study the ISC of natural fractured oil shale through combustion tube experiments. This natural fracture is formed by external pressure with a width less than 105 μm to simulate the combustion behavior of oil shale reservoirs with natural fractures. It is noteworthy that Yuan et al.’s research successfully confirmed the sustainable combustion of natural fractured oil shale reservoirs, with a stable thermal front. Since oil shale or low–medium maturity shale oil both belong to ultralow permeability reservoirs, they need to be fractured before production. The fractured fissures provide a good geological foundation for air injection, making in situ combustion in tight shale reservoirs feasible. Fractured fissures can often reach the millimeter scale, but research on the impact of fractured fissure scale on in situ combustion in low–medium maturity shale reservoirs has not yet been conducted. The fractured fissure zone is the key place for the reaction between oxygen-containing gas and underground organic matter during the ISC process, but the width, complexity of the fractured fissures, and the development effect of different oxygen concentrations are still unclear and need to be evaluated in the laboratory. Therefore, combustion tubes and NMR, GC, and GC–MS are used to study the physical simulation experiment of ISC of low–medium maturity shale oil with millimeter-scale fractures in this paper. A method for calculating the recovery of low–medium maturity shale oil is proposed to quantify the effects of different oxygen concentrations and fracture complexities on the ISC. The conversion of kerogen is quantitatively characterized by the T2 spectrum of NMR, and the changes in crude oil properties are quantitatively analyzed by GC–MS. The ISC in low–medium maturity shale oil reservoirs at the scale of fracturing fractures, as well as the influence of oxygen concentration and fracture characteristics, have been further clarified. Methods Combustion Tube Experiments The shale sample is gray-black from the southern margin of the Ordos Basin, with a total organic carbon (TOC) content of 10.89% and a R0 of 0.58%, belonging to low–medium maturity shale oil. The samples are prepared as two 38 mm × 60 mm rock cores and multiple irregularly sized (1 ∼ 3 cm) rock blocks, denoted as #1, #2, and #3, respectively. Samples #1 and #2 are achieved through linear cutting, with equal division from the center part. The remaining samples are crushed to around 100 mesh for rock pyrolysis testing. The rock pyrolysis parameters S0, S1, and S2 are 0.01, 5.25, and 43.57 mg/g, respectively, as tested by the rock pyrolysis analysis method according to GB/T 18602–2012. In addition, cores #1 and #2 are split into two pieces evenly and through linear cutting, with a 2 mm fracture reserved in the middle for filling 40 ∼ 70 mesh ceramics to simulate fracturing. Rock #3 blocks are used for direct loading and then filled with ceramics to simulate complex fracture morphology, making it have good connectivity. In addition, by adding shale oil and water to the fractures to simulate the fluids in the fractures after fracturing backflow, the shale oil from Jimsar Oilfield had a viscosity of 215.6 mPa·s. The injection end is filled with 1.5 cm of ceramics mixed with kerosene as the igniter, in which kerosene is about 5 mL. The injection end is equipped with an electric heater. The injection end is heated to 350 °C for 3 h until it reaches saturation, and then oxygen-containing gas is injected to achieve ignition. The reaction zone of the experimental device is a copper tube with a length of 8 cm and a diameter of 3.8 cm, sealed with a high-temperature-resistant graphite gasket. In addition, the experimental setup is equipped with an injection section, ring pressure section, external heating section, back pressure section, pipeline tracing, condensation section, gas–liquid separation system, liquid collection section, gas collection section, and gas detection section, as shown in Figure . This paper designs three experiments to investigate the effects of millimeter-scale fractures, oxygen concentration, and fracture characteristics on the ISC of low–medium maturity shale oil reservoirs, as shown in Table . Combustion tube physical simulation device. (a) Physical picture. (b) Connection diagram. 1Conditions and Parameters for Combustion Tube Experimentsrunair flow (L/min)ignition temperature (degC)mass of core model(g)mass of organic carbon(g)fill the fractures with oil(g)oxygen concentration (%)crack characteristics13300125.0113.611.32621A 2 mm single fracture23300132.5814.441.32640A 2 mm single fracture3330080.588.7812.540millimeter-level complex fractures Conditions and Parameters for Combustion Tube Experiments The specific experimental steps are as follows: (1) place the prepared core into the model and set the back pressure to 3 MPa; (2) turn on the in situ heating device to heat the inlet end to 350 °C and then inject air for in situ ignition; (3) continuously inject air. The CH4 and CO2 gases generated by in situ combustion are collected through the gas collection bag and detected by the gas chromatograph, while the produced crude oil and water fluids are measured by the graduated cylinder and high-precision balance, thereby obtaining the experimental dynamic curve; (4) NMR is used to test the solid samples before and after the reaction to determine the curve of organic matter content change; and (5) A GC–MS instrument is used to test the carbon number distribution characteristics of the produced oil. Nuclear Magnetic Resonance Spectroscopy NMR is commonly used for identifying crude oils with different pore sizes, , and T1T2 spectra are also used for quantitative analysis of changes in kerogen content. , In this paper, T2 is tested using a SPEC-RC035 nuclear magnetic resonance core analyzer to measure changes in kerogen content in shale, and eq is used to calculate the conversion of kerogen. The testing parameters for NMR are set to a main magnetic field frequency of 15.14, a main frequency bias of −1610.9, a pulse width of 25.2, 64 scan times, and 1024 sampling points. The detailed principle of relaxation time T2 can be seen reported by Wan et al. , 1 KGConversion=∫T2adt−∫T2bdt∫T2adt where KGConversion is the conversion rate of kerogen; T 2a is the test result of the transverse relaxation time T2 of the rock core before the experiment; and T 2b is the test result of the transverse relaxation time T2 of the rock core after the experiment. The NMR equipment used the RC035 series from SPEC of China. The GC uses the 8860 series from Agilent. The GC–MS uses the SQI series from THERMO FISHER. The NMR sample preparation method is as follows: the initial sample is dried in a vacuum oven at 100 °C for 24 h. After the reaction, the sample is cleaned with dichloromethane and an ultrasonic instrument and then dried in a vacuum oven at 100 °C for 24 h. Calculation Method for Recovery of Low–Medium Maturity Shale Oil Reservoirs To further quantify the aforementioned impacts, this paper proposes a calculation method for the recovery of low–medium maturity shale oil reservoirs. Because the main forms of organic matter in low–medium maturity shale oil reservoirs are free oil, adsorbed oil, and kerogen, the recovery calculated solely based on the oil extraction object is often higher. Therefore, this paper proposes to convert kerogen into oil and gas resources through rock pyrolysis and TOC testing, and use it together with free oil as the total oil and gas geological reserves for recovery calculation. The specific method is to test the remaining organic carbon content and rock pyrolysis parameters S1 and S2 of low–medium maturity shale samples after oil washing. At this time, S1 can be considered as an adsorbed oil that cannot be extracted by organic solvents. To distinguish it from traditional S1, it is referred to as S1a in this paper, and S2 is the oil produced by the pyrolysis of kerogen per unit of shale. Therefore, (S1a + S2) × Mass shale can be regarded as the oil and gas resources that can be converted from kerogen, and its sum with free oil is considered the total oil and gas resources. Therefore, the recovery of low–medium maturity shale oil reservoirs can be calculated using eq . 2 η=δl+δg(S1a+S2)×Massshale+Oil0×100 where η is the recovery of low–medium maturity shale oil reservoirs, %; δ l is liquid hydrocarbon production, g; δ g is gaseous hydrocarbon production, g; Oil 0 is free oil mass, g; S1a is the adsorbed oil mass per unit of shale, mg/g; and S2 is kerogen pyrolysis hydrocarbon production per unit of shale, mg/g. Results and Discussion The Influence of Fracture Scale on Compression Figure shows the results of Experiment #1. From Figure a, it can be seen that the CO2 content in the gas product first increases and then decreases before 20 min, reaching a maximum of around 12%. It indicates that the ignition agent filled successfully burned, but did not ignite the shale oil in the fractures filled or the kerogen in the shale, and did not achieve sustained combustion. Combining the rock photos after combustion in Figure d, it can be seen that shale does not undergo significant combustion effects, and there is still shale oil remaining in the rock, indicating an insufficient combustion reaction. Then the shale oil filled in the fractures is extracted under the action of air flooding, as shown in Figure b. In addition, the T2 spectrum test results before and after ISC of the core experiment show a decrease in the total hydrogen signal (Figure c). As the core is dried and is a low–medium maturity shale, the hydrogen signal mainly comes from the response of kerogen. Therefore, it is believed that some kerogen in the core undergoes a reaction and is lost in the form of free hydrocarbons, such as CO2 or CO. The kerogen conversion is 18.27% by eq , and the recovery is 7.88% by eq . Yuan et al. successfully conducted ISC experiments of fractured oil shale by injecting air using combustion tubes, with combustion temperatures reaching 400–500 °C24. This may be attributed to differences in experimental conditions. In Yuan et al.’s experiment, the simulated fractures are microscale, and real fracturing proppants are not used to fill the fractures. In this paper, ceramics are used to fill the fractures, and shale oil and water are used to fill them, making the fractures more fluid. This may lead to easier outflow of injected air, resulting in the inability to maintain continuous combustion. Second, pressure has a significant impact on the oxidation heat release of kerogen, and increasing pressure can significantly increase the heat release of kerogen. Yuan et al. used a back pressure of up to 10 MPa in their experiment, which resulted in sustained kerogen formation in shale. In this article, the back pressure is set to 3 MPa, mainly simulating ISC between fractures. The condition of high-pressure air injection is not simulated, which failed to ignite kerogen in the shale. Therefore, in the ISC of low–medium maturity shale reservoirs by injecting air between the fracture network due to the extremely poor permeability between the fractures and the matrix, when air is injected under lower back pressure conditions, the gas will mainly flow along the fractures. The rate of oxidation heat release between air and the shale oil in the fractures is low, making it difficult to ignite shale oil and kerogen in the fractures, resulting in difficulty in maintaining continuous ISC and achieving significant transformation of kerogen, leading to lower recovery. #1 Experimental results. (a) Dynamic curve of gaseous products, (b) dynamic curve of liquid products, (c) NMR test result before and after the core reaction, and (d) core photo before the reaction. The Influence of Oxygen Concentration Figure shows the results of experiment #2. From Figure a, it can be seen that the CO2 content in the gas product first increases and then decreases before 40 min, reaching a maximum of around 25%. Compared to experiment 1, the duration of the gas product is longer and the concentration of CO2 is higher, indicating that injecting 40% oxygen-enriched air can promote the rate of exothermic reaction of crude oil oxidation, thus successfully achieving in situ ignition. An increase in oxygen concentration can significantly enhance the combustion efficiency of heavy hydrocarbon components. This is because the increase in the oxygen partial pressure leads to a shorter oxidation induction period and a higher combustion intensity. Furthermore, increasing the oxygen concentration can also reduce the activation energy of their high-temperature oxidation reactions. Therefore, as the oxygen concentration increased, the results of experiment #2 are significantly better than those of experiment #1. From Figure d, it can be seen that there are obvious burn marks on the fracture end face of the core, and there are also damages and microcracks on the injection end and surface of the core. However, the overall appearance of the core is still gray-black, indicating that the kerogen in the core had not undergone continuous combustion or cracking. Due to Figure b, it can be observed that the crude oil production curve exhibits two peaks, with the first peak corresponding to the time of maximum CO2 production, indicating that the filling oil in the fractures is mainly extracted. The proportion of CO2 began to decrease after the first oil production peak but remained at around 10%, indicating that the kerogen in the core is undergoing oxidation and cracking reactions with oxygen. At the second peak of oil production, the displacement pressure increased significantly, and crude oil is extracted at 6 MPa. This may be due to the fact that this part of the modified oil is produced by the pyrolysis of kerogen and the high viscosity of the crude oil. It prevents the injection of air and the CO2 generated in situ from passing through, allowing for a more thorough oxidation reaction between oxygen and the front-end kerogen, resulting in the production of more CO2. Consequently, the injection pressure increased rapidly. Ultimately, as the crude oil is displaced, it forms a significantly similar peak with the CO2. By calculating the total hydrogen signal, it is found that the conversion of kerogen in sample #2 is 44.39%, and the recovery is 22.2%. Compared with the primary conversion and recovery of sample #1, they increase by 26.12% and 14.32%, respectively, indicating that oxygen-enriched air is more conducive to ISC of low–medium maturity shale oil reservoirs. #2 Experimental results. (a) Dynamic curve of gaseous products, (b) dynamic curve of liquid products, (c) NMR test result before and after the core reaction, and (d) core photo before the reaction. The Impact of Fracture Characteristics Figure shows the results of experiment #3. From Figure a, it can be seen that the CO2 content in the gas product first increased and then decreased between 5 to 65 min, and the CO2 content is greater than 20% between 20 to 60 min. When gas was injected into complex fractures, the increase in fracture bands made it easier for the injected gas to flow, but it also increased the contact area between crude oil and oxygen. After injecting oxygen-rich air for 5 min, a significantly high-temperature oxidation reaction occurs, successfully achieving in situ ignition. From Figure d, it can be seen that the react core appears gray white, indicating that the kerogen in the shale also participated in the high-temperature oxidation reaction. The crude oil production curve is consistent with no. 2 and also shows two peaks. The first peak corresponds to the maximum CO2 product time, indicating that the filling oil in the fractures is mainly extracted. The proportion of CO2 begins to decrease after the first oil production peak, and at the beginning of the second oil production stage, CH4 and H2 reach their production peaks, and the displacement pressure increases, indicating that the kerogen has undergone a pyrolysis reaction at this temperature, producing new oil and gas. As in sample #2, the high viscosity of the newly generated crude oil led to an increase in displacement pressure. However, the presence of multiple complex fractures reduces its flow resistance. Therefore, in experiment #3, although the displacement pressure increased, it quickly decreased to the previous level. In the second oil production stage, the proportion of CO2 in the gas phase further increases and is higher than in the first oil production stage. This is due to the high-temperature oxidation reaction between kerogen and dead carbon after cracking and oxygen. The conversion of kerogen in sample #3 is 90.63%, and the recovery is 60.6%. Compared to sample #1, the primary conversion of kerogen and recovery of shale oil increased by 72.36% and 52.72%, respectively, indicating that complex fracture morphology is more conducive to ISC of low–medium maturity shale oil reservoirs. #3 Experimental results. (a) Dynamic curve of gaseous products, (b) dynamic curve of liquid products, (c) NMR test result before and after the core reaction, and (d) core photo before the reaction. Figure shows kerogen conversion and shale oil recovery under different experimental conditions. It can be seen that the shale oil recovery increased with the increase of kerogen conversion, indicating that significant conversion of kerogen can be converted into shale oil and gas resources for exploitation. Injecting oxygen-enriched air can increase the in situ oxidation heat release rate, enhance combustion efficiency, and thus improve the recovery of low–medium maturity shale oil reservoirs. In complex fractures, it can further provide better flow channels for the crude oil produced by in situ conversion of kerogen, thereby significantly improving the recovery. Kerogen conversion and shale oil recovery under different experimental conditions. In addition, the traditional recovery calculation results by eq are compared with those proposed in this paper, as shown in Figure . It can be seen that the recovery obtained by the traditional calculation method is high, even reaching 123.68% in experiment #2, which is obviously unreasonable. The main reason for the high recovery ratio by this method is that it ignores the hydrocarbon generation potential of immature kerogen. In experiments or development, immature kerogen will produce additional oil and gas, thereby affecting the calculation of the recovery ratio. Therefore, the recovery ratio calculation method proposed in this study takes into account the kerogen content and conversion rate and has good applicability for the calculation of the recovery ratio in kerogen-rich shale oil reservoirs. 3 η=OilbOila where η is the recovery of oil reservoirs, %; Oil b is cumulative oil production, g; and Oil b is initial oil content weight, g. Comparison of recovery ratio calculation results. The Impact of ISC on Shale Oil Upgrading The density and carbon number distribution of the modified oil extracted from experiment #3 and the initial shale oil are tested to clarify the impact of ISC on the quality of crude oil. The viscosity test results show that the viscosity of the oil extracted from sample #3 is 64.12 mPa·s, with a viscosity reduction rate of 70.3% compared to the initial crude oil viscosity of 215.6 mPa·s. Figure shows the carbon number distribution of shale oil and the modified oil extracted from sample #3. It can be seen that the carbon number distribution of shale oil before the reaction is relatively uniform (C9 ∼ C35), mainly distributed between C12 ∼ C25, with a proportion of 84.95%, which is the reason for its poor fluidity at room temperature. The carbon number distribution of the modified oil extracted from sample #3 is mainly between C10 and C18, accounting for 84.03%. Due to the increase in the proportion of low-carbon components, the modified oil has excellent fluidity at room temperature, which is more conducive to the production of modified oil. The research results are consistent with the understanding that the ISC of heavy oil can improve crude oil quality. , In addition, Yuan et al. also found that injecting air into oil shale for ISC can improve crude oil quality. The research results indicate that ISC can not only achieve in situ hydrocarbon generation conversion of type II kerogen but also improve the viscosity of the modified oil, making it more fluid and easier to extract. Comparison of carbon number distribution of crude oil before ISC in medium to low maturity shale oil reservoirs. The Influence of High-Temperature Oxidation on the Kerogen-Rich Shale To further clarify the impact of high-temperature oxidation on shale rich-kerogen, the shale samples, size is 1.5 mm × 1.5 mm, before and after oxidation are scan by CT. Through the reconstruction processing with Avizo software, the three-dimensional distribution characteristics of the pore structure of the shale samples before and after the reaction are obtained, as shown in Figure . As can be seen from the figure, before the reaction, the pores in the shale samples are scattered and have extremely poor connectivity. After the reaction, the number of pores significantly increased and the pore volume expanded. The porosity distribution before the reaction is between 2.9% and 6.3%, with an average porosity of 4.3%. After the reaction, the porosity shows excellent connectivity, the number of microcracks increases and the scale becomes larger, and the porosity distribution is between 5.1% and 28.8%, with an average porosity of 19.1%. The reasons for the increase in the number of shale matrix pores after the oxidation reaction are 3-fold. First, due to the oxidation reaction of solid kerogen under air conditions, the solid organic carbon is converted into gas products such as CO, CO2, and H2O, forming a large number of organic pores. Second, some rock minerals in the shale react under the high-temperature oxidation environment, forming some inorganic pores. Third, due to the thermal decomposition of solid kerogen in the micrometer-nanometer pores, the internal pressure of the pores increases, thereby generating microcracks. With the significant increase in the pores and microcracks in the shale matrix, the connectivity of the shale matrix pores is further improved. Distribution of porosity before and after oxidation. (a) The three-dimensional pore distribution of the sample before the reaction. (b) Calculation results of sample porosity before reaction. (c) The three-dimensional pore distribution of the sample after the reaction. (b) Calculation results of sample porosity after reaction. To further analyze the fluidity of the air injection after the oxidation reaction, Avizo software was used to further examine the gas flow characteristics of the samples under certain pressures after the reaction, thereby proving the improvement effect of high-temperature oxidation on the tight shale reservoir. Figure further informatively shows the distribution of the pressure field and velocity streamlines of the oxidized shale samples at a displacement pressure of about 1 MPa, as it can be seen that with the increase in injection pressure, the pressure in the connected pores gradually rises, and the gas begins to flow along the connected pores in the shale medium. This result indicates that after the shale is oxidized at high temperature, a large number of connected pores can be formed, and with the increase in injection pressure, the gas can enter the shale matrix to participate in the reaction. Pressure field and streamline distribution characteristics of shale after oxidation at 1 MPa. Conclusion In this paper, the influence of millimeter-scale fractures on the ISC development of low–medium maturity shale oil reservoirs by air injection is studied using combustion tubes and GC, and GC–MS devices. Based on NMR and the proposed calculation method for low–medium maturity shale oil recovery, the effects of oxygen concentration and fracture characteristics on recovery are quantitatively analyzed. The research results found that the presence of fractures made it easy for injected gas to flow along the fractures, and injecting air can easily cause ignition failure, resulting in low conversion of kerogen and recovery of shale oil reservoirs. Rich oxygen air can ignite shale oil in fractures and promote the transformation of kerogen in shale, but when kerogen does not participate in combustion, its conversion is 44.39%, and the shale oil recovery is 22.2%. The results of complex fracture experiments show that the conversion rate of kerogen and shale oil recovery can be significantly improved only when the shale oil in the fracture and the kerogen in the shale are ignited and continuously burned. An efficient underground in situ ignition agent should be developed to meet the needs of in situ ignition in this type of reservoir. In addition, an increase in oxygen concentration can accelerate the exothermic reaction rate of organic matter oxidation and raise reservoir temperature to promote kerogen cracking and hydrocarbon generation, thereby improving oil recovery. Complex fractures can increase the contact area between organic matter and oxygen, improve the heat release efficiency per unit volume of the reservoir, and provide good channels for the flow and recovery of modified oil. This study suggests that chemical ignition agents and enhanced oxygen concentration should be used to ensure the success of in situ ignition and the sustainable advancement of the combustion front during the ISC of fractured shale reservoirs with low–medium maturity. The complexity of the fractures has a favorable effect on their recovery. Therefore, the optimal fracture morphology should be selected for reservoir modification before development based on the development process, such as throughput or interfracture injection.