Fracturing Productivity Prediction Model and Optimization of the Operation Parameters of Shale Gas Well Based on Machine Learning

Tan, Chaodong; Yang, Junzheng; Cui, Mingyue; Wu, Hua; Wang, Chunqiu; Hanwen, Deng; Song, Wenrong; Xiong, Hao

doi:10.2113/2021/2884679

Cited by 8 publications

(4 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For large displacement values and small numbers of samples, the displacementprediction effect of the XGBoost algorithm is better than that of the other two prediction methods in the sliding process of landslides. It is consistent with existing research results [46,47].…”

Section: Discussionsupporting

confidence: 94%

Landslide Displacement Prediction during the Sliding Process Using XGBoost, SVR and RNNs

Xu¹,

Jiang²,

Yang³

2022

Applied Sciences

View full text Add to dashboard Cite

In order to promptly evacuate personnel and property near the foot of the landslide and take emergency treatment measures in case of sudden danger, it is very necessary to select suitable forecasting methods for conduct short-term displacement predictions in the slope-sliding process. In this paper, we used Python to develop the landslide displacement-prediction method based on the eXtreme Gradient Boosting (XGBoost) algorithm, and optimized the hyperparameters through a genetic algorithm to solve the problem of insufficient short-term displacement-prediction accuracy for landslides. We compared the deviation, relative error (RE) and median of RE of predicted values obtained using XGBoost, SVR and RNNs, and the actual value of landslide displacement. The results show that the accuracies of slope displacement prediction using XGBoost and SVR are very high, and that using RNNs is very low during the sliding process. For large displacement values and small numbers of samples, the displacement-prediction effect of XGBoost algorithm is better than that of SVR and RNNs in the sliding process of landslide. There are generally only fewer data samples collected during the landslide sliding process, so RNNs is not suitable for displacement prediction in this scenario. If the number of data samples is large enough, using RNNs to predict the long-term displacement of the slope may also have a much higher accuracy.

show abstract

Section: Discussionsupporting

confidence: 94%

Landslide Displacement Prediction during the Sliding Process Using XGBoost, SVR and RNNs

Xu¹,

Jiang²,

Yang³

2022

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Nevertheless, given the complexity of the problem, continued exploration of new methods is deemed crucial. [20] In this research, various techniques were employed by the authors, including six machine learning algorithms: "random forest (RF), back propagation (BP) neural network, support vector regression (SVR), extreme gradient boosting (XGBoost), light gradient boosting machine (Light GBM), and multivariable linear regression (MLR)." Evaluation results indicate that the production prediction model of the XGBoost algorithm was found to be the most effective, with an R 2 value of 0.90.…”

Section: Related Workmentioning

confidence: 99%

Oil and Gas Production Forecasting Using Decision Trees, Random Forst, and XGBoost

A. Al shabaan,

N. Nemer

2024

J. Al-Qadisiyah Comp. Sci. Math.

View full text Add to dashboard Cite

Oil and gas production forecasting has always been a hot topic in the petroleum industry. Production forecasting in this sector aims to estimate future production rates, facilitating operational planning, production optimization, and resource allocation for companies. Scientists have traditionally attempted to forecast oil and gas production using methods such as Numerical Reservoir Simulation (NRS) and Decline Curve Analysis (DCA). However, these methods present challenges including time-consuming processes lasting hours or even days, uncertain accuracy, reliance on accurate static models, and uncertainty in dynamic model parameters. In this research, aim to address these limitations by leveraging machine learning models for production forecasting. These models enable faster and more precise decision-making by accurately predicting future outcomes based on historical data. Our study employs three models: Decision Trees (DTR), Random Forest (RFR), and XGBoost. In this reserch utilize the Python programming language and a dataset from wells in New York State, USA. Experimental results demonstrate that the RFR model achieves the highest accuracy (99%) for oil and gas production compared to the XGBoost and DTR models.

show abstract

“…Owing to the complex transport mechanism, different geological conditions, and drilling designs, prediction of shale gas well production and estimate ultimate recovery (EUR) shows great uncertainty. Due to the advantage of few hypotheses in physical mechanism, AI-based EUR prediction models have been established. − The most common ML-based models for prediction of gas production performance include decision tree model, random forest regression, gradient boosted machine, support vector machine, etc. Excellent prediction of gas production can be achieving using AI-based methods with input of actual shale production data and geological and engineering design parameters .…”

Section: Shale Structure Characterization and Gas Transport Predictio...mentioning

confidence: 99%

Review of Shale Gas Transport Prediction: Basic Theory, Numerical Simulation, Application of AI Methods, and Perspectives

et al. 2023

View full text Add to dashboard Cite

The gas transport mechanism in shale reservoirs is extremely complex and is a typical multiscale and multiphysics coupled transport process, considering the complex shale rock structure, wide distribution of micropores and nanopores in shale gas reservoirs, diverse gas occurrence forms, and large pore size spans. An accurate understanding of the shale gas transport process and mechanism is important for effective exploration of shale gas reservoirs. In this work, a review of the recent progress in the prediction of shale gas transport in porous media is presented. The basic theory of gas transport in nanopores is discussed. The gas transport in organic and inorganic matter and the gas adsorption effect are covered. Then, gas transport simulations with conventional multiscale numerical methods, including molecular dynamics and lattice Boltzmann simulations, are reviewed, and the multiscale modeling methods are discussed. Furthermore, the application of artificial intelligence (AI) methods in shale gas transport research is discussed. The focus is on the characterization of the shale porous geometry, including porosity, tortuosity, pore size distribution, and reconstruction of the shale porous medium. The application of AI-based methods such as neural networks and machine learning for the prediction of porous flow properties is discussed. This study intends to provide a comprehensive review of shale gas transport characteristics and to enable the accessibility of AI tools in the research of shale gas.

show abstract

Fracturing Productivity Prediction Model and Optimization of the Operation Parameters of Shale Gas Well Based on Machine Learning

Cited by 8 publications

References 13 publications

Landslide Displacement Prediction during the Sliding Process Using XGBoost, SVR and RNNs

Landslide Displacement Prediction during the Sliding Process Using XGBoost, SVR and RNNs

Oil and Gas Production Forecasting Using Decision Trees, Random Forst, and XGBoost

Review of Shale Gas Transport Prediction: Basic Theory, Numerical Simulation, Application of AI Methods, and Perspectives

Contact Info

Product

Resources

About