A deep-learning-based prediction method of the estimated ultimate recovery (EUR) of shale gas wells

Liu, Yuyang; Ma, Xiaobin; Zhang, Xiaowei; Guo, Wei; Kang, Lixia; Yu, Rong; Sun, Yize

doi:10.1016/j.petsci.2021.08.007

Cited by 57 publications

(15 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although it is simple, the processing capacity is limited. Therefore, the activation function commonly used in the neural network contains nonlinear factors, and the commonly used activation functions include sigmoid, tanh, ReLU, softmax, etc., to improve the expressive ability of the model . According to the layer division, the neural network inside the DNN can be divided into: input layer, hidden layer, and output layer.…”

Section: Methodsmentioning

confidence: 99%

“…The output result is the value. Furthermore, the forward propagation is used to calculate the output of the training sample, and the loss function is used to measure the loss between the output calculated by the training sample and the real training sample label. , The back propagation algorithm of DNN uses gradient descent method to iteratively optimize the loss function. Try to find the appropriate hidden layer and output layer corresponding to the linear coefficient matrix W , bias vector b , so that all the training sample input calculated output as much as possible equal to or close to the sample label.…”

Section: Methodsmentioning

confidence: 99%

“…Therefore, the activation function commonly used in the neural network contains nonlinear factors, and the commonly used activation functions include sigmoid, tanh, ReLU, softmax, etc., to improve the expressive ability of the model. 16 According to the layer division, the neural network inside the DNN can be divided into: input layer, hidden layer, and output layer. Generally, the first layer is the input layer, the last layer is the output layer, and the middle layers are hidden layer.…”

Section: Deep Neural Networkmentioning

confidence: 99%

See 2 more Smart Citations

Toward Production Forecasting for Shale Gas Wells Using Transfer Learning

et al. 2023

View full text Add to dashboard Cite

Accurate prediction of shale gas well production and estimated ultimate recovery (EUR) is always a difficult and hot spot in shale gas development. In particular, the production and EUR prediction of shale gas wells in new production blocks are faced with the lack of field gas well data and the difficulty of model development. In view of the above problems, this study proposes a new deep transfer learning strategy, which uses transfer component analysis (TCA) and deep neural network (DNN) to achieve shale gas well production and EUR prediction across formations/blocks. The feature extractor based on TCA can narrow the input feature distribution of the source and the target domains. The neural network model can be used to establish a domainadaptive transfer learning model without the prediction performance degradation caused by distribution offset. Validity and accuracy of the model were analyzed using gas well data from Weiyuan and Luzhou blocks in Sichuan Basin, China. The results appear that the reasonable application of TCA can greatly improve the prediction performance of shale gas well transfer learning model. For data sets of the same size, compared with the transfer learning model developed by classical machine learning algorithms, the proposed neural network-based transfer learning model can significantly improve the accuracy of production prediction across formations/ blocks. In addition, the proposed model can also be extended to other types of oil and gas production prediction tasks cross formations/blocks.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Deep Neural Networkmentioning

confidence: 99%

See 1 more Smart Citation

Toward Production Forecasting for Shale Gas Wells Using Transfer Learning

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Yuyang et al 42 developed a model to forecast the flowback rate of shale gas wells through the use of deep learning based on the characteristic factors of flowback in the Weiyuan formation. Liu et al 43 designed and performed a deep learning-based model for the evaluation of the estimated ultimate recovery of shale gas wells using geological data, hydraulic fracturing data, and production data. Niu et al 44 developed K-nearest neighbor (KNN), SVM, RF, and gradient boosting decision tree (GBDT) machine learning models to predict the estimated ultimate recovery of shale gas.…”

Section: ■ Introductionmentioning

confidence: 99%

Hybrid Machine Learning-Mathematical Programming Approach for Optimizing Gas Production and Water Management in Shale Gas Fields

López-Flores

Lira-Barragán

Rubio-Castro

et al. 2023

ACS Sustainable Chem. Eng.

View full text Add to dashboard Cite

This paper presents a novel mathematical programming approach that simultaneously incorporates a mixed-integer nonlinear programming formulation with machine learning models to determine the operating conditions, gas production, and optimal water management for the completion phase in shale gas fields. The dataset for the development of an artificial neural network model has been collected from the Eagle Ford Texas formation. The total cumulative gas production and flowback water generated in shale gas wells are selected as output variables. The mathematical optimization model considers machine learning models for each well, mass balances, treatment, storage, reuse, and disposal options as well as well location selection and associated costs and revenues from the sale of the shale gas produced. A case study has been used to illustrate the benefits of the proposed approach. The compromise solution offers a water consumption per produced energy of 6.71 L/GJ in addition to the fact that 27% of the total fracture water required can be obtained by reusing the flowback water with attractive economic indicators.

show abstract

“…The main source of uncertainty is the lack of available geological data. Depending on the quantity and quality of the available data, different methods are used for the evaluation of the EUR [1][2][3]. For example, in the initial stage of development of the hydrocarbon deposit, there is very little information available; therefore, approximate estimates are usually made using analog or volumetric calculations.…”

Section: Introductionmentioning

confidence: 99%

Production Data Analysis Techniques for the Evaluation of the Estimated Ultimate Recovery (EUR) in Oil and Gas Reservoirs

Elmabrouk

Mahmud

2022

HighTech. Innov. J.

View full text Add to dashboard Cite

The calculation of oil reserves (estimate ultimate recovery, EUR) is required for reservoir management. It is important to differentiate between oil reserves and oil resources. The latter is roughly defined as the sum of recoverable and unrecoverable volumes of oil in place; whereas, the oil reserves can be defined as those amounts of oil anticipated to be commercially recoverable from a given date under defined conditions. However, there is always uncertainty when making reserve estimates, and the main source of uncertainty is the lack of available geological data. Depending on the quantity and quality of the available data, different methods are used for the evaluation of the EUR. A number of essentially straight-line extrapolation techniques (production data analysis) have been proposed to estimate the EUR for oil and gas wells. Thus, a detailed analysis of past performance of oil and water production data is required in order to predict the future performance of the oil and gas wells. This work utilized seven straight-line extrapolation techniques to estimate and compare the values of EUR of three oil wells from the same reservoir. The comparison shows very similar estimated EUR. Doi: 10.28991/HIJ-2022-03-01-09 Full Text: PDF

show abstract

A deep-learning-based prediction method of the estimated ultimate recovery (EUR) of shale gas wells

Cited by 57 publications

References 32 publications

Toward Production Forecasting for Shale Gas Wells Using Transfer Learning

Toward Production Forecasting for Shale Gas Wells Using Transfer Learning

Hybrid Machine Learning-Mathematical Programming Approach for Optimizing Gas Production and Water Management in Shale Gas Fields

Production Data Analysis Techniques for the Evaluation of the Estimated Ultimate Recovery (EUR) in Oil and Gas Reservoirs

Contact Info

Product

Resources

About