Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Machine Learning Based Predictive Models for CO2 Corrosion in Pipelines With Various Bending Angles
Yang, Huihui (Shell International Exploration and Production, Inc.) | Lu, Ligang (Shell International Exploration and Production, Inc.) | Tsai, Kuochen (Shell International Exploration and Production, Inc.)
Abstract Predicting CO2 corrosion in fluid transmission pipelines is crucial for oil/gas company in upstream applications. This paper applies Light Gradient Boosting Machine (LightGBM) and Multiple Layer Perceptron Neural Network (MLPNN) models for the prediction of CO2 corrosion in aqueous pipelines with different pipe bending angles. To build the predictive models, a data set with total of 77,745 data points was generated parametrically by a computational fluid dynamics (CFD) model. Since different environmental conditions and geometries of the pipeline may cause non-uniform corrosion, a total of seven variables, including flow velocity, pH value, CO2 concentration, pipe inner diameter, pipe bend angle, radius and temperature are taken as the input features with the corrosion rate as the target variable. The CFD model was then used to compute the electrochemical processes occurring at the metal surfaces to predict the corrosion rate. Knowing that these features have nonlinear relationship with the target, tree based LightGBM, and neural network based MLPNN were chosen. LightGBM can control the overfitting issues, deal with comparative scales of the features and learn non-linear decision boundaries via boosting. The most significant findings are that these two types of machine learning (ML) algorithms have higher efficiency and can predict new results in microseconds in contrast to hours or even days using CFD. The R square of the LightGBM model is 0.9985, which is slightly higher than that of the MLPNN model at 0.9931. The k-fold cross validation results also show the stability of the two models. These ML models are 5 to 6 orders of magnitude faster than CFD models with similar accuracy therefore significantly saving time and cost. We further built a web application based on these predictive models as a tool for pipeline design and monitoring applications.
- Well Completion > Well Integrity > Subsurface corrosion (tubing, casing, completion equipment, conductor) (1.00)
- Facilities Design, Construction and Operation > Pipelines, Flowlines and Risers > Materials and corrosion (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (1.00)
Summary This paper investigates the computational behaviors of simple-to-use, relatively fast, and versatile machine learning (ML) methods to predict apparent viscosity, a key rheological property of nanoparticle-surfactant-stabilized CO2 foam in unconventional reservoir fracturing. The first novelty of our study is the investigation of the predictive performance of ML approaches as viable alternatives for predicting the apparent viscosity of NP-Surf-CO2 foams. The predictive and computational performance of five nonlinear ML algorithms were first compared. Support vector regression (SVR), K-nearest neighbors (KNN), classification and regression trees (CART), feed-forward multilayer perceptron neural network (MLPNN), and multivariate polynomial regression (MPR) algorithms were used to create models. Temperature, foam quality, pressure, salinity, shear rate, nanoparticle size, nanoparticle concentration, and surfactant concentration were identified as relevant input parameters using principal component analysis (PCA). A data set containing 329 experimental data records was used in the study. In building the models, 80% of the data set was used for training and 20% of the data set for testing. Another unique aspect of this research is the examination of diverse ensemble learning techniques for improving computational performance. We developed meta-models of the generated models by implementing various ensemble learning algorithms (bagging, boosting, and stacking). This was done to explore and compare the computational and predictive performance enhancements of the base models (if any). To determine the relative significance of the input parameters on prediction accuracy, we used permutation feature importance (PFI). We also investigated how the SVR model made its predictions by utilizing the SHapely Additive exPlanations (SHAP) technique to quantify the influence of each input parameter on prediction. This workโs application of the SHAP approach in the interpretation of ML findings in predicting apparent viscosity is also novel. On the test data, the SVR model in this work had the best predictive performance of the single models, with an R of 0.979, root mean squared error (RMSE) of 0.885 cp, and mean absolute error (MAE) of 0.320 cp. Blending, a variant of the stacking ensemble technique, significantly improved this performance. With an R of 1.0, RMSE of 0.094 cp, and MAE of 0.087 cp, an SVR-based meta-model ensembled with blending outperformed all single and ensemble models in predicting apparent viscosity. However, in terms of computational time, the blended SVR-based meta-model did not outperform any of its constituent models. PCA and PFI ranked temperature as the most important factor in predicting the apparent viscosity of NP-Surf-CO2 foams. The ML approach used in this study provides a comprehensive understanding of the nonlinear relationship between the investigated factors and apparent viscosity. The workflow can be used to evaluate the apparent viscosity of NP-Surf-CO2 foam fracturing fluid efficiently and effectively.
- Asia (1.00)
- North America > United States > California (0.46)
- North America > United States > Massachusetts (0.28)
This paper presents a toolbox for optimizing geotechnical design of subsea foundations. The geotechnical design challenge of subsea shallow foundations is to withstand greater dead and operational loads on soft seabeds without increasing the footprint size or weight. The motivation is to reduce costs associated with installation โ for example eliminating the need for a heavy-lift vessel to place foundation units alone if handling limits of pipe-laying vessels are exceeded โ whilst providing acceptable in-service reliability. The tools presented focus on prediction of undrained seabed response and are intended for deep water developments on fine grained seabeds, as this scenario presents a significant challenge in terms of minimizing subsea foundation footprints. The toolbox addresses optimization of geotechnical subsea foundation performance through four aspects: (i) optimizing the analysis methodology, (ii) modifying the foundation configuration, (iii) improving the site characterisation data as input to the design, and (iv) altering the basis of design. The research presented derives from a combination of physical model testing in a geotechnical centrifuge, numerical analysis and theoretical modelling. The methods, procedures and processes are presented in terms of design equations, theoretical frameworks or design charts, many of which are freely available as web-based applications. Worked examples throughout the paper demonstrate the efficiencies in terms of footprint area to be realized through adoption of these tools.
- Europe (1.00)
- North America > United States > Texas (0.67)
- Research Report > New Finding (0.46)
- Research Report > Experimental Study (0.46)
- Reservoir Description and Dynamics (1.00)
- Facilities Design, Construction and Operation > Pipelines, Flowlines and Risers > Offshore pipelines (1.00)
- Facilities Design, Construction and Operation > Offshore Facilities and Subsea Systems (1.00)
- Data Science & Engineering Analytics > Information Management and Systems (1.00)
This paper presents a series of numerical investigations of the capacity of the foundations of a spudcan in clay after it penetrates through a sand layer and under planar combined loading. The model is validated against available relationships for uniaxial capacity of spudcan foundations on clay soils. The model is then used to investigate the uniaxial capacity and combined VH, VM, and HM capacity of spudcan foundations in the underlying clay of sand overlying clay soils. The results are presented in the form of failure envelopes. The effects of the embedment depth, undrained shear strength of clay, and sand plug thickness on the uniaxial and combined capacity are investigated and discussed.
- North America > United States (1.00)
- Oceania > Australia (0.93)
- Europe > United Kingdom > North Sea (0.89)
- Europe > Norway > North Sea (0.89)
- Europe > Netherlands > North Sea (0.89)
- (2 more...)
A New Tool for Searching Sweet Spots by Using Gradient Boosting Decision Trees and Generative Adversarial Networks
Tang, Jizhou (Harvard University) | Fan, Bo (Avigilon, A Motorola Solutions Company) | Xu, Ganchuan (CNPC Chuanqing Drilling Company) | Xiao, Lizhi (China University of Petroleum, Beijing) | Tian, Shouceng (China University of Petroleum, Beijing) | Luo, Shaocheng (CNPC Logging Company) | Weitz, David (Harvard University)
High-density completions prevail in shale oil formation in China due to the difficulty of identifying the sweet spot with high accuracy. Knowing the location of sweet spots benefits the horizontal well drilling and the selection of perforation clusters. Generally, field engineers determine sweet spots from the well logging interpretation. However, a group of prevalent classifiers based on gradient boosting decision trees were introduced to automatically determine sweet spots according to datasets from the well logging. Based on boosted tree algorithms, Extreme Gradient Boosting (XGBoost), Unbiased boosting with categorical features (CatBoost) and Light Gradient Boosting Machine (LightGBM) are utilized to control the over-fitting issues. Compared with linear support vector machines (SVMs) or kernel machine, these robust algorithms can deal with comparative scales of the features and learn non-linear decision boundaries via boosting. Moreover, they are less influenced by the presence of outliers. Another prevailing approach, named Generative Adversarial Networks (GANs), was implemented to augment the training dataset by using a small number of training samples. In terms of the training purpose, we randomly selected 60 horizontal wells. In each well, tens to hundreds of datasets of different formation intervals were collected. Features, such as resistivity, interval transit time, layer thickness, shale content, porosity, permeability, oil saturation, and coordinates in three dimensions, were extracted from well-logging datasets and regarded as inputs for classifiers. Datasets of remaining wells were used for testing. Compared with conventional SVMs, the prediction accuracies of sweet spots by XGBoost and CatBoost were significantly improved to 81.61% and 82.5%, respectively. Additionally, GANs, as an unsupervised machine learning tool, have been attempted to augment the dataset by utilizing a relatively small number of training samples. A generative model is used for capturing the data distribution, and a discriminative model aims at predicting a label to which that data created by the generative model. Without special pre-processing of the input datasets and fine tuning CTGANs model, the fake dataset could still bring 68.58% accuracy for all detections and 59.01% of the label corresponding with oil formation that showing its potential in data augmentation. This paper illustrates a new tool for categorizing the reservoir quality by using gradient boosting decision trees and GANs methods, which further helps search and identify sweet spots. An extensive application has been built for the field cases in a certain oilfield. This tool provides a guideline for covering more sweet spots during the drilling and completion treatment, which immensely decreases the exploration cost.
- North America > United States (1.00)
- Asia (1.00)
- Geophysics > Seismic Surveying (1.00)
- Geophysics > Borehole Geophysics (1.00)
- North America > United States > West Virginia > Appalachian Basin (0.99)
- North America > United States > Virginia > Appalachian Basin (0.99)
- North America > United States > Tennessee > Appalachian Basin (0.99)
- (10 more...)
- Reservoir Description and Dynamics > Formation Evaluation & Management > Open hole/cased hole log analysis (1.00)
- Management (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Neural networks (1.00)
- Data Science & Engineering Analytics > Information Management and Systems > Artificial intelligence (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Ensemble Learning (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Decision Tree Learning (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Support Vector Machines (0.54)