Research on prediction methods of formation pore pressure based on machine learning
Abstract Formation pressure is the most fundamental data in oil and gas drilling and production; it has an important position in the entire cycle of oil and gas extraction. However, most current prediction methods are limited to parametric methods with fixed models; such that the accuracy does not m...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-06-01
|
| Series: | Energy Science & Engineering |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/ese3.1112 |
| _version_ | 1811258229068398592 |
|---|---|
| author | Honglin Huang Jun Li Hongwei Yang Biao Wang Reyu Gao Ming Luo Wentuo Li Geng Zhang Liu Liu |
| author_facet | Honglin Huang Jun Li Hongwei Yang Biao Wang Reyu Gao Ming Luo Wentuo Li Geng Zhang Liu Liu |
| author_sort | Honglin Huang |
| collection | DOAJ |
| description | Abstract Formation pressure is the most fundamental data in oil and gas drilling and production; it has an important position in the entire cycle of oil and gas extraction. However, most current prediction methods are limited to parametric methods with fixed models; such that the accuracy does not meet requirements. This is especially true for deeper layers of marine sedimentary basins where the safety density window is extremely narrow. In this study, we propose a novel method to predict pore pressure using machine learning techniques. For the first time, the effective stress (direct output variable) was accurately predicted by a combination of four input variables (2900 sets of data, of which 90% is the training subset and 10% is the testing subset), including longitudinal velocity, porosity, mud content, and density. As such, an accurate prediction of the formation pressure was achieved based on the effective stress theorem. The performance of machine learning techniques was verified by comparing and analyzing the prediction results with traditional parametric single and multivariate models; whereby the best algorithm was chosen by structural optimization and comparative analysis of five algorithms (multilayer perceptron neural network, radial basis neural network, support vector machine, random forest, and gradient boosting machine). Compared with the methods based on parametric one‐dimensional and multivariate models, the machine learning‐based method was determined to possess high accuracy, adequate self‐adaptation, and high fault tolerance (D2 = 0.9981, RMSE = 0.00718 g/cm3). Moreover, the multilayer perceptual neural network algorithm outperformed other machine learning algorithms in terms of goodness of fit, generalization, and prediction accuracy, with D2 = 0.9981 and RMSE = 0.00709 g/cm3. The formation pressure prediction model developed in this study is not affected by the mechanical depositional environment and is applicable to sandy mudstone formations, such that it can be a useful and highly accurate alternative to the traditional formation pressure prediction methods with fixed parameter forms. |
| first_indexed | 2024-04-12T18:11:07Z |
| format | Article |
| id | doaj.art-70dbe241c817437c93d4e17c63da32d3 |
| institution | Directory Open Access Journal |
| issn | 2050-0505 |
| language | English |
| last_indexed | 2024-04-12T18:11:07Z |
| publishDate | 2022-06-01 |
| publisher | Wiley |
| record_format | Article |
| series | Energy Science & Engineering |
| spelling | doaj.art-70dbe241c817437c93d4e17c63da32d32022-12-22T03:21:50ZengWileyEnergy Science & Engineering2050-05052022-06-011061886190110.1002/ese3.1112Research on prediction methods of formation pore pressure based on machine learningHonglin Huang0Jun Li1Hongwei Yang2Biao Wang3Reyu Gao4Ming Luo5Wentuo Li6Geng Zhang7Liu Liu8College of Petroleum Engineering China University of Petroleum Beijing ChinaCollege of Petroleum Engineering China University of Petroleum Beijing ChinaCollege of Petroleum Engineering China University of Petroleum Beijing ChinaCollege of Petroleum Engineering China University of Petroleum Beijing ChinaCollege of Petroleum Engineering China University of Petroleum Beijing ChinaDrilling Technology Department CNOOC China Limited Zhanjiang ChinaDrilling Technology Department CNOOC China Limited Zhanjiang ChinaCollege of Petroleum Engineering China University of Petroleum Beijing ChinaSchool of Foreign Languages, China University of Petroleum Beijing ChinaAbstract Formation pressure is the most fundamental data in oil and gas drilling and production; it has an important position in the entire cycle of oil and gas extraction. However, most current prediction methods are limited to parametric methods with fixed models; such that the accuracy does not meet requirements. This is especially true for deeper layers of marine sedimentary basins where the safety density window is extremely narrow. In this study, we propose a novel method to predict pore pressure using machine learning techniques. For the first time, the effective stress (direct output variable) was accurately predicted by a combination of four input variables (2900 sets of data, of which 90% is the training subset and 10% is the testing subset), including longitudinal velocity, porosity, mud content, and density. As such, an accurate prediction of the formation pressure was achieved based on the effective stress theorem. The performance of machine learning techniques was verified by comparing and analyzing the prediction results with traditional parametric single and multivariate models; whereby the best algorithm was chosen by structural optimization and comparative analysis of five algorithms (multilayer perceptron neural network, radial basis neural network, support vector machine, random forest, and gradient boosting machine). Compared with the methods based on parametric one‐dimensional and multivariate models, the machine learning‐based method was determined to possess high accuracy, adequate self‐adaptation, and high fault tolerance (D2 = 0.9981, RMSE = 0.00718 g/cm3). Moreover, the multilayer perceptual neural network algorithm outperformed other machine learning algorithms in terms of goodness of fit, generalization, and prediction accuracy, with D2 = 0.9981 and RMSE = 0.00709 g/cm3. The formation pressure prediction model developed in this study is not affected by the mechanical depositional environment and is applicable to sandy mudstone formations, such that it can be a useful and highly accurate alternative to the traditional formation pressure prediction methods with fixed parameter forms.https://doi.org/10.1002/ese3.1112effective stressformation physical property and lithologylogging datamachine learningpore pressure |
| spellingShingle | Honglin Huang Jun Li Hongwei Yang Biao Wang Reyu Gao Ming Luo Wentuo Li Geng Zhang Liu Liu Research on prediction methods of formation pore pressure based on machine learning Energy Science & Engineering effective stress formation physical property and lithology logging data machine learning pore pressure |
| title | Research on prediction methods of formation pore pressure based on machine learning |
| title_full | Research on prediction methods of formation pore pressure based on machine learning |
| title_fullStr | Research on prediction methods of formation pore pressure based on machine learning |
| title_full_unstemmed | Research on prediction methods of formation pore pressure based on machine learning |
| title_short | Research on prediction methods of formation pore pressure based on machine learning |
| title_sort | research on prediction methods of formation pore pressure based on machine learning |
| topic | effective stress formation physical property and lithology logging data machine learning pore pressure |
| url | https://doi.org/10.1002/ese3.1112 |
| work_keys_str_mv | AT honglinhuang researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT junli researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT hongweiyang researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT biaowang researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT reyugao researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT mingluo researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT wentuoli researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT gengzhang researchonpredictionmethodsofformationporepressurebasedonmachinelearning AT liuliu researchonpredictionmethodsofformationporepressurebasedonmachinelearning |