Application of Machine Learning Classification Algorithms for Two-Phase Gas-Liquid Flow Regime Identification

This research aims to identify the best machine learning (ML) classification techniques for classifying the flow regimes in vertical gas-liquid two-phase flow. Two-phase flow regime identification is crucial for many operations in the oil and gas industry. Processes such as flow assurance, well control, and production rely heavily on accurate identification of flow regimes for their respective systems' smooth functioning. The primary motivation for the proposed ML classification algorithm selection processes was drilling and well control applications in Deepwater wells. The process started with vertical two-phase flow data collection from literature and two different flow loops. One, a 140 ft. tall vertical flow loop with a centralized inner metal pipe and a larger outer acrylic pipe. Second, an 18-ft long flow loop, also with a centralized, inner metal drill pipe. After extensive experimental and historical data collection, supervised and unsupervised ML classification models such as Multi-class Support vector machine (MCSVM), K-Nearest Neighbor Classifier (KNN), K-means clustering, and hierarchical clustering were fit on the datasets to separate the different flow regions. The next step was fine-tuning the models' parameters and kernels. The last step was to compare the different combinations of models and refining techniques for the best prediction accuracy and the least variance. Among the different models and combinations with refining techniques, the 5- fold cross-validated KNN algorithm, with 37 neighbors, gave the optimal solution with a 98% classification accuracy on the test data. The KNN model distinguished five major, distinct flow regions for the dataset and a few minor regions. These five regions were bubbly flow, slug flow, churn flow, annular flow, and intermittent flow. The KNN-generated flow regime maps matched well with those presented by Hasan and Kabir (2018). The MCSVM model produced visually similar flow maps to KNN but significantly underperformed them in prediction accuracy. The MCSVM training errors ranged between 50% - 60% at normal parameter values and costs but went up to 99% at abnormally high values. However, their prediction accuracy was below 50% even at these highly overfitted conditions. In unsupervised models, both clustering techniques pointed to an optimal cluster number between 10 and 15, consistent with the 14 we have in the dataset. Within the context of gas kicks and well control, a well-trained, reliable two-phase flow region classification algorithm offers many advantages. When trained with well-specific data, it can act as a black box for flow regime identification and subsequent well-control measure decisions for the well. Further advancements with more robust statistical training techniques can render these algorithms as a basis for well-control measures in drilling automation software. On a broader scale, these classification techniques have many applications in flow assurance, production, and any other area with gas-liquid two-phase flow.

Vertical-Downward Two-Phase Flow Regime Identification by Probabilistic Neural Network (PNN) and Nonlinear Support Vector Machine (SVM)

The present study proposes a new feature extraction method based on non-stationary conductivity probe signals. Two types of discriminative network models, i.e., the probabilistic neural network (PNN) and nonlinear support vector machine (SVM), are established for flow regime identification using small sample sets. The eigenvectors are composed of 16 feature quantities obtained by wavelet packet decomposition (WPD) and 8 feature quantities in the time-domain derived from the reconstructed low-frequency signals. The 8 features include maximum, minimum, standard deviation, arithmetic mean, kurtosis, peak factor, impulse factor and margin factor. The signals are normalized based on features rather than samples before flow regime identification. In the current study, WPD results show that the conductivity probe signals in two-phase flow are mostly in low frequency. The identification accuracy of the nonlinear SVM is 90.47%, which is better than 83.33% by the PNN method. This study verifies the superiority of nonlinear SVM in solving small samples and nonlinear flow regime classification problems. However, the accuracy of flow regime identification near flow regime transitional boundaries still remains questionable and needs further improvement.

Flow Regime Identification in Vertical Upward Gas–Liquid Flow Using an Optical Sensor With Linear and Quadratic Discriminant Analysis

The accurate identification of gas–liquid flow regimes in pipes remains a challenge for the chemical process industries. This paper proposes a method for flow regime identification that combines responses from a nonintrusive optical sensor with linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA) for vertical upward gas–liquid flow of air and water. A total of 165 flow conditions make up the dataset, collected from an experimental air–water flow loop with a transparent test section (TS) of 27.3 mm inner diameter and 5 m length. Selected features extracted from the sensor response are categorized into feature group 1, average sensor response and standard deviation, and feature group 2 that also includes percentage counts of the calibrated responses for water and air. The selected features are used to train, cross validate, and test four model cases (LDA1, LDA2, QDA1, and QDA2). The LDA models produce higher average test classification accuracies (both 95%) than the QDA models (80% QDA1 and 45% QDA2) due to misclassification associated with the slug and churn flow regimes. Results suggest that the LDA1 model case is the most stable with the lowest average performance loss and is therefore considered superior for flow regime identification. In future studies, a larger dataset may improve stability and accuracy of the QDA models, and an extension of the conditions and parameters would be a useful test of applicability.

Experimental Investigation for Flow Regime Identification Using Probability Density Function of Void Fraction Signals

In this study, upward air–water two-phase flow tests were carried out in a 3 cm diameter pipe under atmospheric pressure, and over 3000 data points were collected from a wide range of superficial gas and liquid velocities (⟨jg⟩ ≈ 0.02–30 m/s and ⟨jf⟩ ≈ 0.02–2 m/s) for the investigation of flow regime identification. The probability density function (PDF) of transient void fraction signals and its full-width at half-maximum (FWHM) were obtained and used for analysis and data classification. Considering the features of PDF profiles, the flow conditions can be classified into four regions, which show a fair agreement with the existing flow regime maps in general trends. Furthermore, by examining the FWHM distributions, two more regions with high-FWHM (HF) values were identified as the transitions of higher-flow bubbly-to-slug and slug-to-churn flows as well as most portion of churn flow, and a valley region next to the HF regions can express the transition of churn-to-annular flows. Overall, six groups of flow conditions can be classified based on the present methodology, and each group can be corresponding to specific flow regimes or transition regions. This study can provide a simple and efficient way for flow regime identification.