Omar S. Alolayan , Samuel J. Raymond , Justin B. Montgomery , John R. Williams
{"title":"Towards better shale gas production forecasting using transfer learning","authors":"Omar S. Alolayan , Samuel J. Raymond , Justin B. Montgomery , John R. Williams","doi":"10.1016/j.upstre.2022.100072","DOIUrl":null,"url":null,"abstract":"<div><h3>Abstract</h3><p>Deep neural networks<span> can generate more accurate shale gas<span> production forecasts in counties with a limited number of sample wells by utilizing transfer learning. This paper provides a way of transferring the knowledge gained from other deep neural network models trained on adjacent counties into the county of interest. The paper uses data from more than 6000 shale gas wells across 17 counties from Texas Barnett and Pennsylvania Marcellus shale formations to test the capabilities of transfer learning. The results reduce the forecasting error between 11% and 47% compared to the widely used Arps decline curve model.</span></span></p></div>","PeriodicalId":101264,"journal":{"name":"Upstream Oil and Gas Technology","volume":"9 ","pages":"Article 100072"},"PeriodicalIF":2.6000,"publicationDate":"2022-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"10","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Upstream Oil and Gas Technology","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S266626042200010X","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENERGY & FUELS","Score":null,"Total":0}
引用次数: 10
Abstract
Deep neural networks can generate more accurate shale gas production forecasts in counties with a limited number of sample wells by utilizing transfer learning. This paper provides a way of transferring the knowledge gained from other deep neural network models trained on adjacent counties into the county of interest. The paper uses data from more than 6000 shale gas wells across 17 counties from Texas Barnett and Pennsylvania Marcellus shale formations to test the capabilities of transfer learning. The results reduce the forecasting error between 11% and 47% compared to the widely used Arps decline curve model.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
