Search

The southern Junggar Basin, Xinjiang, has abundant coalbed methane (CBM) resources. Currently, the Baiyang River development pilot test area (BYR block for short) in the Fukang east block has achieved large-scale CBM development, but the productivity characteristics and its controlling factor are still unclear. Based on the field production data of the BYR block and experimental tests, this paper summarizes the gas and water production characteristics and presents the analysis results of the geological and geochemical responses to the productivity of CBM wells. The productivity of CBM wells in the BYR block was generally characterized as medium-to-low yield. The productivity was jointly controlled by the burial depth, structure condition, thickness and number of co-production coal seams, and hydrogeological conditions. The gas production first increased and then decreased with the increase in the burial depth of the coal seam, and a burial depth between 750 and 1000 m was the most beneficial to increasing the gas production due to the good gas preservation conditions and suitable permeability and stress conditions. The total thickness of the co-production coal seams had a positive effect on the productivity of gas wells, but the productivity was also affected by the number of co-production coal seams and interlayer interference. In the BYR block, the co-production of the nos. 41 and 42 coal seams was the most favorable combination form for CBM drainage. The productivity of CBM wells had a good response to the Na+, K+ and HCO3− concentrations but a poor response to δD-H2O and δ18O-H2O. Based on the concentrations of the main ions and TDSs of the coal seam water, a productivity response index δ* was established, and there was a good positive correlation between the productivity and δ*. [ABSTRACT FROM AUTHOR]

There are large, heavy oil reserves in Block X of the Xinjiang oilfields, China. Due to its large burial depth (1300 m) and low permeability (26.0 mD), the traditional steam-injection technology cannot be used to obtain effective development benefits. This paper conducts experimental and simulation research on the feasibility and mechanism of CO 2 -energized fracturing of horizontal wells and N 2 foam huff-n-puff in deep heavy oil reservoirs with low permeability in order to further explore the appropriate production technology. The foaming volume of the foaming agent at different concentrations and the oil displacement effect of N 2 foam at different gas/liquid ratios were compared by the experiments. The results show that a high concentration of foaming agent mixed with crude oil is more conducive to increasing the foaming volume and extending the half-life, and the best foaming agent concentration is 3.0∼4.0%. The 2D micro-scale visualization experiment results show that N 2 foam has a good selective blocking effect, which increases the sweep area. The number of bubbles per unit area increases as the gas/liquid ratio increases, with 3.0∼5.0 being the optimal gas/liquid ratio. Numerical simulation results show that, when CO 2 -energized fracturing technology takes into account the advantages of fracturing and crude oil viscosity reduction by CO 2 dissolution, the phased oil recovery factor in the primary production period can reach approximately 13.7%. A solvent pre-slug with N 2 foam huff-n-puff technology is applied to improve oil recovery factor following primary production for 5∼6 years, and the final oil recovery factor can reach approximately 35.0%. The methodology formulated in this study is particularly significant for the effective development of this oil reservoir with deeply buried depth and low permeability, and would also guide the recovery of similar oil deposits. [ABSTRACT FROM AUTHOR]

In the past, reservoir engineers used numerical simulation or reservoir engineering methods to predict oil production, and the accuracy of prediction depended more on the engineers' own experience. With the development of data science, a new trend has arisen to use deep learning to predict oil production from the perspective of data. In this study, a hybrid forecasting model (CNN-LSTM) based on a convolutional neural network (CNN) and a Long Short-Term Memory (LSTM) neural network is proposed and used to predict the production of fractured horizontal wells in volcanic reservoirs. The model solves the limitation of traditional methods that rely on personal experience. First, the production constraints and production data are used to form a feature space, and the abstract semantics of the feature time series are extracted through convolutional neural network, then the LSTM neural network is used to predict the time series. The certain hyperparameters of the whole model are optimized by Particle Swarm Optimization algorithm (PSO). In order to estimate the model, some production dynamics from the Xinjiang oilfield of China are used for comparative analysis. The experimental results show that the CNN-LSTM model is superior to traditional neural networks and conventional decline curves. [ABSTRACT FROM AUTHOR]