Shale gas plays an important role in supplementing energy demand and reducing carbon footprint. A precise and effective prediction of shale gas production is important for optimizing completion parameters. This paper established a gated recurrent unit and multilayer perceptron combined neural network (GRU-MLP model) to forecast multistage fractured horizontal shale gas well production. A nondominated sorting genetic algorithm II (NSGA II) was introduced into the model to enable its automatic architectural optimization. In addition, embedded discrete fracture models (EDFM) and a reservoir simulator were used to generate training datasets. Meanwhile, a sensitivity analysis was carried out to find the variable’s importance and support the history matching. The results illustrated that the GRU-MLP model can precisely and efficiently predict the productivity of multistage fractured horizontal shale gas in a rapid and effective manner. Additionally, the model fits better at peak values of shale gas production. The GRU-MLP hybrid model has a higher accuracy within an acceptable computational time range compared to recurrent neural networks (RNN), long short-term memory (LSTM), and GRU models. The mean absolute percentage error (MAPE) and root mean square percentage error (RMSPE) for shale gas production generated by GRU-MLP model were 3.90% and 3.93%, respectively, values 84.87% and 84.88% smaller than those of the GRU model. Consequently, compared with a purely data-driven method, the physics-constrained data-driven method behaved better. The main results of the study will hopefully contribute to the intelligent development of shale gas production prediction.
The influence of complex natural fractures on multiple shale-gas well performance with varying well spacing is poorly understood. It is difficult to apply the traditional local grid refinement with structured or unstructured gridding techniques to accurately and efficiently handle complex natural fractures. In this study, we introduced a powerful non-intrusive embedded discrete fracture model (EDFM) technology to overcome the limitations of exiting methods. Through this unique technology, complex fracture configurations can be easily and explicitly embedded into structured matrix blocks. We set up a field-scale two-phase reservoir model to history match field production data and predict long-term recovery from Marcellus. The effective fracture properties were determined thorough history matching. In addition, we extended the single-well model to include two horizontal wells with and without including natural fractures. The effects of different numbers of natural fractures on two-well performance with varying well spacing of 200 m, 300 m, and 400 m were examined. The simulation results illustrate that gas productivity almost linearly increases with the number of two-set natural fractures. Furthermore, the difference of well performance between different well spacing increases with an increase in natural fracture density. A larger well spacing is preferred for economically developing the shale-gas reservoirs with a larger natural fracture density. The findings of this study provide key insights into understanding the effect of natural fractures on well performance and well spacing optimization.
The deep shale gas reservoirs of the southern Sichuan Basin exhibit high temperatures, high pressure, large stress differentials, and complex natural fracture systems. Since 2019, hydraulic fracturing technology in this region has evolved through four stages: exploratory fracturing, intensive limited-volume fracturing, tight spacing with controlled fluid and proppants, and balanced fracturing that combines long-section temporary plugging with short-section intensive cutting. Despite these advances, production remains suboptimal due to inefficient reserve utilization, a lack of quantitative methods for residual gas evaluation, and unclear identification of the remaining reserves. To address these challenges, we developed an integrated workflow combining dynamic production analysis, geomechanical modeling, and numerical simulation to evaluate representative fracturing techniques. Fracture propagation in the well group was modeled in the in-house hydraulic fracture simulator, ZFRAC, to assess fracture geometry, while production history and geological data were used to build calibrated reservoir simulation models. This enabled quantitative assessment of effective fracture parameters, reserve utilization, and residual gas distribution. The results show significant intra-stage heterogeneity driven by stress interference, effective fracture half-lengths of 60–105 m, and a cut-off ratio (proportion of effective fracture half-length to wetted fracture half-length) of 60–93%. Reserve utilization peaked at 60% for intensive limited-volume fracturing, while the efficacy of long-section temporary plugging was limited. These findings offer critical insights for optimizing infill strategies and enhancing sustainable shale gas development in southern Sichuan.