The integrity of natural gas pipelines will decrease with an increase in operating time, thus causing pipeline leaks and accidents. However, it is challenging to improve the precision and automation of existing sensors to raise leak prediction and classification precision. Therefore, based on deep learning, a 1D convolutional neural network (CNN) incorporating the channel attention mechanism is proposed for recognizing and classifying the type of natural gas pipeline leakage. Firstly, the data reconstruction of the leaked acoustic signals, which have been classified by energy modes, is performed by feature augmentation and Bessel filtering. Subsequently, a lightweight CNN is proposed, and an attention mechanism is introduced to optimize the model performance. The results show that the training performance of the network with the attention mechanism is superior to that of the original network and the network with batch normalization. The attention mechanism network is then used to train the leakage signals with different features of engineering parameters. Finally, the test accuracy achieves 97.81%, validating the effectiveness of the proposed method for identifying and classifying natural gas leaks. It presents new ideas for the implementation of deep learning in the natural gas and chemical industries.

Combining the advantages of multicontinuum and multifracture representations provides an easy-to-use tool to adequately capture the characteristic of the multiscaled fracture system in shale gas reservoir. A hybrid model is established on the basis of simplified conceptual productivity assumption, where the matrix volume is divided into two sub-domains (triple-porosity model and dual-depletion flowing model) and the fracture volume is represented by discrete finite conductivity fracture. In addition, the mechanisms of instant desorption, viscous flow and dual-depletion in matrix are taken into account. The rate transient responses are then obtained by use of semi-analytical approach. Based on the model, type curves are plotted and verified by comparing with alternative reliable methods. Different flow regimes in shale gas reservoirs can be identified and detected. The Generalized Likelihood Uncertainty Estimation methodology, based on probabilistic aggregation theory, is employed to integrating those two productivity models together such that the production can be predicted more accurately. A field example is applied to validate the applicability of this new model. Finally, it is concluded that the proposed model can predict the rate and cumulative rate more easily and practically.