Abstract:There are abundant shale oil resources in Jiyang Depression，and high-production industrial oil flows have been successively obtained in several wells such as Wells YYP1，FYP1，and BYP5 with peak production exceeding 100 t/d since 2019，which shows good prospects for shale oil reservoir exploration and development. Summarizing the geological characteristics of shale oil reservoirs in Jiyang Depression and analyzing their enrichment and high production regularities are of important guiding significance for the economic and effective development of shale oil reservoirs. Based on the review of shale oil reservoir exploration and development practices in Jiyang Depression，the main control factors for the enrichment and high production of shale oil reservoirs were recognized in this paper through systematic research on the petrographic characteristics，reservoir space，oil-bearing properties，and compressibility of shale oil reservoirs in Jiyang Depression. Research shows that shale oil resources in Jiyang Depression are abundant but scattered，characterized by high depths，high temperatures and pressure coefficients，and complex tectonics，petrography，ground stress，and fluid properties. The large thickness of hydrocarbon source rocks，high organic matter content，moderate thermal evolution，and the development of a carbonate-rich phase in the saline lake basin are the main factors for shale oil enrichment. The development of an inorganic pore-fissure network，high formation pressure coefficient，high brittle mineral content，good compressibility，hole expansion and energy enhancement by CO2，and multi-scale fracture network fracturing are the main factors for the high production of shale oil reservoirs in Jiyang Depression. The breakthrough of shale oil reservoirs in Jiyang Depression shows the broad prospect of exploration and development of shale oil reservoirs in continental rift basins，which is of great guiding and reference significance for the same type of reservoirs in China and abroad.

Abstract:The core analysis and gas logging are universally applied to research the mobility of shale oil，but the role of the logging in evaluating shale oil is not played fully. On the basis of the core analysis and elemental capture spectroscopy （ECS）logging calibration，the logging evaluation method and characterization means of shale oil geological characteristics are established by the logging curve overlap method. The results show that the total organic carbon（TOC）calculated by the improved ΔlogR method is more consistent with core analysis data. There is a good corresponding relationship of logging curve overlap and its combination characteristics with gas logging abnormality，which can qualitatively reflect the mobility of shale oil. With the 2D NMR data of Well BYP5，the fluid occurrence characteristics and oil-bearing properties of shale oil formation are revealed for the first time from the perspective of logging. The quantitative oil-bearing property evaluation method of shale oil formation is determined，and the NMR T2 cut-off value of shale oil movable porosity is obtained to quantitatively evaluate oil-bearing properties. Applications indicate that the logging evaluation results are basically consistent with the core analysis results. The logging curve overlap method and the NMR data can more comprehensively and accurately reflect the oil-bearing properties of shale oil formation. Thus，the role of logging data in shale oil evaluation is fully displayed.

Abstract:Fengcheng Formation of the Lower Permian in Junggar Basin is a set of important source rocks. The understanding of the sedimentary environment in this period is controversial in that it is understood as marine facies by some researchers and as lacustrine facies by others. This paper utilizes the outcrop，seismic，drilling，logging，and core data in the study area and its adjacent areas to identify the formation period of Bogda Mountains at the southern margin of Junggar Basin and the geological structures in the basin，restores the tectono-sedimentary framework of the Early Permian in Junggar Basin，and analyzes the petroleum geological characteristics of Fengcheng Formation. The results show that the East Bogda Mountains and the east margin of Eren Habirga Mountains separated Junggar Basin from Tuha Basin，and Junggar Basin became a lake with marine vestiges due to residual sea occlusion at the end of the Carboniferous. Shixi Sag in the north of the basin and Shiqiantan Sag mainly developed the alluvial fans during the sedimentation of Fengcheng Formation. The northwestern margin，the central depression，and Bogda area，as lakes with marine vestiges in terms of sedimentary environments，inherited the marine saline water environment and marine organisms. Four sedimentary centers were developed in the basin，and favorable source rocks were developed in the semi-deep to deep lacustrine subfacies in saline water environments under the control of the uplifts and underwater low upheavals in Junggar Basin. The various sags in the hinterland of Junggar Basin and the periphery of Bogda Mountains，where the Lower Permian is widely distributed and incredibly rich natural gas resources might exist in deep burial environments，are favorable exploration directions for the next step.

Abstract:Ultra-deep carbonate gas reservoirs are deeply buried，with complex stress states and extremely strong heterogeneity，and the change laws of pore and throat structures before and after stress changes is not clear. In this paper，the cores from Deng4 gas reservoir in Gaoshiti-Moxi block were selected，and CT experiments were conducted to obtain the size distribution of pores and throats，the proportion of pores and vugs，and the connectivity of various reservoirs before and after the stress experiment and fracturing experiment. In addition，the effects of stress and fracturing on the pore and throat structures of ultra-deep carbonate reservoirs were studied. The research results reveal that the pores and throats in the pore-vug and fracture-vug cores decrease obviously after recovery from stress，and most of them are the micropores and mesopores with a radius of less than1 mm and the micro-throats with a radius of less than 0.04 mm. The average pore radius，average throat radius，average throat length，as well as the proportion of fractures and vugs all rise significantly，while the total volume of pores and throats tends to decline. The volume ratio of connected pores and throats in the pore-vug cores decreases significantly，while the volume ratio of connected pores and throats in the fracture-vug cores remains at the original level due to the development of fractures. After fracturing，the pores in the pore-vug cores decrease sharply and the reduced pores are mainly micropores and mesopores with a radius below 0.5 mm. The average pore radius greatly increases，while the most increased pore radii are in 0.5-0.8 mm. There is an upward trend in the pore volume，the throats，average throat radius，average throat length as well as total throat volume，and the proportion of fractures and vogs and the volume of connected pores and throats increase drastically. The experiments indicate that after recovery from stress，the flow capacity of the ultra-deep carbonate reservoir does not decrease but greatly increases，and fracturing mainly improves the connectivity of pores and throats by increasing the proportion of fractures in the pore-throat space.

Abstract:Induced fractures，cataclastic rocks，or fault rocks will be formed during the active periods of faults connecting the source rocks and reservoirs，which play a transporting role in oil and gas migration. Affected by factors such as fault activity intensity，lithologies on both fault walls，and shale contents of fault rocks，different parts of faults（fault sections）have different transport capacities for fluids. Quantitative evaluation of fault transport differences is helpful for predicting favorable areas of fault-controlled tight oil and gas reservoirs. In this paper，parameters including the growth index of faults，fracturing ratios of faults，and rock fracturing pressure were used to develop a calculation method for the fault transport index，and the fault-controlled tight oil and gas accumulation mechanism and pattern were explicated. The source rock coefficient was calculated according to the thickness，maturity，and type of the source rock. In addition，a comprehensive analysis was performed on the transport index，source rock coefficient，pressure differences between source rock and reservoirs，and favorable reservoir facies belt for each section of Fault Boshen4 of Bonan Sag in Jiyang Depression. Then，a classification standard for favorable accumulation areas of fault-controlled tight oil and gas was constructed. The predicted favorable accumulation areas of fault-controlled tight oil and gas are consistent with the actual exploration results. The results reveal that the calculated fault transport indexes by section can quantitatively assess the differences in fault transport capacity and that the large pressure difference between source rock and reservoirs，fault transport difference，and lateral sealing of fault rocks are the main factors in the fault-controlled tight oil and gas accumulation.

Abstract:The seismic response of the carboniferous ultra-deep thin sand layer in Tahe Oilfield is complex. The interval transit time and density logging curve are less sensitive to reservoirs，and the seismic attribute and the existing inversion methods cannot meet the requirements of fine development of oil and gas reservoirs in terms of reservoir prediction accuracy. Therefore，this paper proposes a pseudo-acoustic wave log constraint inversion method with seismic bipolarity transform based on bio-inspired algorithm. Specifically，the paper uses the bio-inspired algorithm to perform bipolarity transform on seismic data with different phases to make the date bipolar，which enhances the visual resolution of the data. In addition，most of the thin sand layers have corresponding seismic reflection events，which improves the relationships between thin sand layers and seismic responses. Then，the paper reconstructs pseudo-acoustic wave by self potential and natural gamma log curves to strengthen the reservoir sensitivity of inversion parameters. As a result，the inversion process only requires low-level log constraint，and the inversion results have a positive correlation with seismic data，which decreases the ambiguity and increases the reliability of reservoir prediction. The carboniferous ultra-deep thin sand layer，with a depth of 5 000 m and a thickness of more than 3 meters of Well S9702 area in Tahe Oilfield，can be identified by the proposed method and the error is small. The narrow channel morphology is relatively clear，with the distribution of sand bodies in line with the geological law.

Abstract:sandstone and mudstone identification tasks are usually based on logging curves and rely on traditional methods such as empirical formulas，field core sampling，cross plots，and cluster analysis，but these methods fail to make full use of the sandstone and mudstone features contained in the logging curves. At the same time，these traditional methods have low accuracy and slow efficiency and are greatly affected by human factors. To address the above problems，this paper uses logging data as the basis，combines the key technical difficulties of sandstone and mudstone identification, and conducts sensitivity analysis on logging parameters，so as to select appropriate influencing factors and construct a complete training data set through several pre-processing operations. In addition，the paper introduces the Focal Loss function and proposes the FL-XGBoost model according to the sparsity of logging labels and carries out sandstone and mudstone identification in Niuzhuang area of Shengli Oilfield. The experimental results show that the sandstone and mudstone identification model using the FL-XGBoost algorithm achieves an accuracy of 0.827 in identifying the sandstone and mudstone in the study area. Finally，the strong generalization ability of the FL-XGBoost algorithm in the identification classification field is verified through five publicly classified dataset design comparison experiments.

Abstract:Carbonate fracture-cavity reservoirs are characterized by strong heterogeneity. Single seismic attribute prediction and conventional seismic fusion methods do not take additional information such as mud leakage into consideration，which could lead to large errors. According to well seismic calibration，a fracture-cavity reservoir prediction method based on well-controlled multi-attribute machine learning was proposed to finely predict fracture-cavity reservoirs，with the attributes of mud-leakage points taken as the constraints. Firstly，in accordance with the results of well seismic calibration，different sensitive seismic attribute values at the leakage points were extracted as input data，and the reservoir types defined by the features of leakage points were output to form training set data. Then，based on the support vector machine（SVM）method，model training was conducted on the data to obtain a prediction model highly consistent with prior seismic attributes. Finally，the model was applied to predict the Ordovician fracture-cavity reservoir in Shunbei Oilfield，Tarim Basin.The prediction results show that this method can reflect the real reservoir characteristics and fits well with drilling features.

Abstract:Logging while drilling（LWD）technologies are employed in many oilfields to reduce the impact of mud intrusion on logging curves，which require the prediction of logging curves for undrilled formations as it is of great guiding significance to LWD. Therefore，a method based on the gated recurrent unit（GRU）neural network was applied to predict the logging curves of undrilled formations. The model combines the input gate and forget gate of long short-term memory（LSTM）into an update gate and turns the input gate into a reset gate，which makes its structure simple and not prone to overfitting.Meanwhile，it retains the long-term memory function of the LSTM model and can effectively alleviate the problem of gradient vanishing or explosion. Taking real logging data from vertical wells in Xinjiang Oilfield and LWD data in western South China Sea Oilfield as examples，this study selected the five logging curves of drilled formations and adjoining wells，namely，the curves of the natural gamma ray，deep induction resistivity，acoustic time difference，density，and well diameter，as training samples and input into the LSTM model and GRU model for learning training. The trained models were then used to predict the logging curves of undrilled formations. The application results indicate that the average correlation coefficients of predicted logging curves for Xinjiang Oilfield and western South China Sea Oilfield by the GRU model are 13.78% and 12.13% higher than that of the LSTM model，and the mean root mean square errors are decreased by 27.08% and 42.17%，respectively. The GRU model can accurately predict the variation trend of logging curves for undrilled formations.

Abstract:The gas reservoirs of the second Member of Dengying Formation both in Gaoshiti-Moxi area and Taihe gas area have strong heterogeneity and multi-scale fractures and pores（caves），showing the characteristics of fracture-cavity gas reservoirs with bottom water. It is difficult to develop such reservoirs due to the rapid intrusion of bottom water along the fractures and caves and no effective technologies for reference. In order to scientifically develop such fractured gas reservoirs with bottom water and provide technical reference for other gas reservoirs with bottom water，this paper takes the development characteristics of typical gas reservoirs bottom with water both in China and abroad as the research object，deeply analyzes the development status，geological characteristics and development technologies of various gas reservoirs with bottom water，clarifies the main controlling factors affecting the the reservoir recovery and summarizes the technologies for controlling water in the reservoirs. In addition，the paper puts forward appropriate strategies for developing fracture-cavity bottom water gas reservoirs with strong heterogeneity. The study found that the the fractures，interlayers and water energy are the main geological factors that affect the development effect of the reservoirs. Furthermore，the reservoir recovery can be improved by deepening the understanding of gas reservoir characteristics. Reasonable technical technologies can slow down non-uniform bottom water intrusion. Balanced production is the key to realizing the efficient development of fracturecavity gas reservoirs with bottom water. Specific technical strategies are as follows. ①The distribution law of fractures and caves in the reservoirs shall be precisely depicted by a well-seismic combination，and the understanding of the geological characteristics of the reservoirs shall be continuously deepened under dynamic and static conditions. ②The reasonable technical limits shall be made to strengthen the monitoring and management of the reservoirs. ③The prediction and simulation systems for the water intrusion in the reservoirs shall be constructed to develop a water control strategy for the whole life cycle of the reservoirs.

Abstract:In order to identify and evaluate the rock wettability of shale oil reservoirs and deepen the understanding of the crude oil producing in the reservoirs，this paper takes rock samples from the shale oil reservoirs in the West 233 area of Ordos Basin as the research object and carries out two-dimensional nuclear magnetic resonance（NMR）T1-T2 map measurement as well as spontaneous imbibition wettability experiment based on NMR technology in the state of fully saturated water and irreducible water. In addition，the paper establishes the basic method for evaluating wettability based on the two-dimensional NMR T1-T2 maps and selects parallel rock samples for core water flooding experiment based on NMR technology，so as to analyze and evaluate the influence of wettability on the crude oil production characteristics during the water flooding. The research results show that after the oil flooding water（i.e.，in the irreducible water state），the force between the water and the pore throat wall is significantly reduced，and the water exhibits the characteristics of bulk relaxation in the state of free fluid in the pore throat space. As a result，the wettability of the rock samples is oil-wet，which is consistent with the results of the spontaneous imbibition wettability experiment based on NMR technology. Therefore，it can be seen that the wettability of the rock samples can be evaluated by comparing the two-dimensional NMR T1-T2 maps in different states. During the water flooding，the crude oil is mainly produced in the medium and large pore throats of the rock samples but poorly produced in small pore throats. When the capillary number of injected water is increased，and the oil-wet characteristics of the rock in reservoirs are weakened，the crude oil producing in the small pore throats is significantly improved.Therefore，the capillary number of injected water and wettability of the rock should be comprehensively considered during water flooding development，so as to improve the water flooding efficiency of shale reservoirs.

Abstract:The imbibition can improve the crude oil production of unconventional reservoirs by the oil-water displacement.Currently，the spontaneous imbibition predominates in the experimental research，which simulates the imbibition process in cores under the capillary force at atmospheric pressure. However，there are fluid pressures during the actual fracturing in the reservoirs，and the traditional experimental methods cannot reflect the forced imbibition characteristics. Taking the argillaceous dolomite of Qianjiang Formation of Qianjiang Sag in Jianghan Basin as an example，this paper designed an forced imbibition experiment based on low field NMR（LF-NMR）technology and studied the imbibition laws at fluid pressures.The results show that according to the T2 pore classification，98.14%-99.49% of the pores in the shale oil core samples are the small holes and medium holes，and the medium holes are the main reservoir spaces. Compared with spontaneous imbibition，the additional fluid pressure makes more use of smaller pores，and the recovery of 5 MPa and 10 MPa forced imbibition is increased by 24.32% and 62.59% respectively. The imbibition causes fractures on the surface of cores with high clay content，increases the contact area，and improves the oil-water displacement efficiency. At the same time，the imbibition can not only improve physical properties but also damage reservoirs.

Abstract:As a new oil displacement technology for low-permeability reservoirs，nanofluid flooding has advantages such as good injectivity，low formation damage，great displacement ability，and intelligent response，which has attracted more and more attention from researchers. Compared with spherical nanoparticles，Janus nanosheets have lower interfacial free energy and more restricted rotation，which can further prevent the diffusion of internal and external molecules and form solid particle films with higher interfacial strength，exhibiting excellent oil displacement performance. Researchers have developed multiple preparation methods for Janus nanosheets，whose application in oil displacement mainly includes nano-emulsion technology and nano-foam stabilization technology. In addition，the characteristic of intelligent response gives such materials more potential. When one side of the sheet nanomaterials is hydrophilic while the other is hydrophobic，they will show amphipathicity and “Pickering”emulsion effect simultaneously，which greatly improves their ability as an efficient nanofluid flooding. With low-cost，large-scale，green，and controllable preparation and the improvement in temperature and salt resistance in the future，Janus nanosheets are expected to significantly enhance oil recovery and provide innovative solutions for stable production and stimulation of oilfields.

Abstract:The well pattern adjustment is an effective method to further enhance recovery after the waterflooding reservoirs enter the ultra-high water cut stage. On the basis of the experimental device for three-dimensional（3D）physical simulation，waterflooding physical simulation experiments for positive-rhythm reservoirs were carried out under the basic well pattern，thinned well pattern，and infilled well pattern，and the development effects with and without adjustment of well patterns（well pattern thinning and well pattern infilling）were compared. It is found that compared with the basic well pattern，the thinned well pattern and the infilled well pattern can increase the final recoveries，and the increased recovery under thinned well pattern is greater than that under infilled well pattern. When the cumulative injected water volumes are the same，a higher injection rate leads to a higher waterflooding sweep coefficient and greater recovery. In addition，the final recovery with increased liquid production after well pattern adjustment is higher than that without increased liquid production. The waterflooding development effect under the thinned well pattern is superior to that under the basic well pattern and infilled well pattern，and increased liquid production after well pattern adjustment can give rise to enhanced oil recovery.

Abstract:The chemical combination flooding is one of the key techniques for enhanced heavy oil recovery. More importance is attached to the emulsification and viscosity reduction mechanism during the development of such combination systems，and some efficient emulsification systems are formed. The“oil recovery increment”caused by strong emulsification，however，is still unclear，which results in difficulties in determining the actual contributions of emulsification to oil displacement.In this study，three combination systems with significantly different properties were collected，namely，the ultralow interfacial tension system（#1），the strong emulsification system（#2），and the system with both ultra-low interfacial tension and strong emulsification（also called dual-effect system）（#3）. The comparative studies of interfacial tension，emulsification performance，and oil displacement at different water-oil viscosity ratios were carried out. The results reveal that the second and third systems（#2 and #3）are more capable of stabilizing heavy oil emulsions than the ultra-low interfacial tension system（#1），and the contributions of emulsification to the combination flooding in heavy oil reservoirs varies with the water-oil viscosity ratios. When the water-oil viscosity ratio is less than 0.200，the oil recovery of the dual-effect system（#3）is 3.6%-6.7% higher than that of the ultra-low interfacial tension system（#1），which indicates that emulsification can enhance the oil displacement capacity of the system. When the ratio is equal to or greater than 0.200，however，the oil displacement effect of the three systems is similar，and the impact of emulsification is significantly reduced or can even be ignored. The foam combination flooding could significantly raise the oil recovery increment in comparison with the binary combination flooding，and more importantly，it can reduce the water-oil viscosity ratio limits required by combination system emulsification performance for heavy oil displacement from 0.200 to 0.150. It can be seen that the performance of combination systems of combination flooding in heavy oil reservoirs should be determined according to the difference in water-oil viscosity ratios.

Abstract:Most mature onshore oilfields in China have entered the high water-cut stage，and the field experience shows that some oil wells have a lower water cut after a certain shut-in period，the reason for which is the remaining oil re-enrichment. The remaining oil re-enrichment in the lower potential closed zones after water flooding under the effect of gravity and capillary force is one of the important factors affecting the of remaining oil re-distribution. In this paper，the water saturation classification standard for waterflooded reservoirs was established by means of the water cut classification and fractional flow curve. On this basis，the indexes for the re-enrichment and the potential producing of remaining oil were constructed to characterize and evaluate the remaining oil re-enrichment. Based on the laboratory-scale physical simulation experiment，The influence of different factors on remaining oil re-enrichment was simulated separately under the dominant conditions of gravity or hydrodynamics. The results show that the remaining oil re-enrichment mainly occurs in the upper part of a reservoir and the lower part of the interlayer under the dominant condition of gravity，and the re-enrichment effect is better at a high-pressure gradient. Under the dominant condition of hydrodynamics，the high-pressure gradient can promote the remaining oil re-enrichment，and a farther distance from well pattern to the top of the structure indicates a larger space for remaining oil re-enrichment and higher recovery.

Abstract:At present，the prediction of the hydraulic fracturing effect of oil production wells in Daqing Oilfield mostly relies on experience or simple models such as multiple linear regression，which leads to poor stability of prediction results and low prediction accuracy. With Block N23 of Daqing Oilfield as an example，the correlation between the fracturing effect of oil production wells and influencing factors is analyzed by the mathematical statistics. The influence of those factors on the hydraulic fracturing effect in Block N23 is studied by a random forest algorithm. Additionally，the principles and implementation methods of meta learning，Bayesian optimization，and model ensemble in automatic machine learning are presented，and a prediction model of a data-driven hydraulic fracturing effect based on the automatic machine learning technology is constructed. Meanwhile，the model is compared with three common machine learning algorithms：random forest，support vector machine and neural network. The proposed model is employed to design and optimize the hydraulic fracturing of Block N23. The results show that the production parameters before fracturing exert an important influence on predicting the effect of oil production well after fracturing. The model constructed by the automatic machine learning algorithm has higher accuracy than other algorithms. The determination coefficient on the test set is 0.695，and the average relative prediction error is 18.96%，which is 57.53% lower than the current level. Compared with the original one，the fracturing scheme optimized by the model can increase the economic benefit by about 3.2×104-27.4×104 yuan per well.