Prediction Of Pressure Gradient Research Articles

Correlation between pressure gradient from echocardiography and peak-to-peak pressure gradient from cardiac catheterization or surgery in patients with coarctation of aorta.

The gold standard for assessing pressure gradients (PG) across coarctation involves measurements obtained through cardiac catheterization or surgical intervention. There has been ongoing discussion regarding the accuracy of non-invasive methods for estimating these gradients. This study sought to establish the correlation and agreement between the systolic blood pressure (SBP) gradient between the upper and lower extremities, as well as, the mean and maximum PG derived from echocardiography, in comparison to the peak-to-peak pressure gradient obtained from either cardiac catheterization or surgery. We conducted a retrospective study on patients < 18 years diagnosed with coarctation at Chiang Mai University Hospital from 2011 to 2022. The study involved the measurement of the SBP gradient between the upper and lower extremities, mean and maximum PG using echocardiography, peak-to-peak pressure gradient obtained from cardiac catheterization, and pressure gradient recorded during surgical procedures. The Spearman's correlation and Bland-Altman analysis were employed to assess correlation and agreement. Fifty-four patients with aortic coarctation were enrolled. The mean PG measured by echocardiography showed a significantly moderate correlation (r = 0.78, p < 0.001) and the highest level of agreement according to Bland Altman plots, in comparison to the peak-to-peak pressure gradient measured during both cardiac catheterization and surgical procedure. The max PG demonstrated a notable overestimation compared to the gold standard (mean difference + 13.14 with a slope of biases + 0.64, p < 0.001). The mean PG obtained through echocardiography has more potential to be applied in practical application in predicting pressure gradient in patients with coarctation.

Effect of field viscometer dial readings on the estimation of drilling fluid frictional pressure loss using Herschel–Bulkley model

The main objective of this study is to investigate how uncertainties in dial readings would affect the frictional pressure loss estimation of the flow of drilling fluid in drill pipe. In this study, a new calculation procedure for determination of frictional pressure loss using the Herschel–Bulkley model is presented and verified with experimental data. A Monte Carlo simulation study is conducted to investigate the sensitivity of each dial reading on the estimation of frictional pressure drop for different circulation velocities. The results show that for the low yield fluids, 1o dial reading error introduces more than 6% error to the frictional pressure gradient prediction, especially at low flow velocity. While the errors in dial readings result in much smaller error in pressure gradient estimation for fluids with high yield stress, typically less than 5%. In addition, the readings at lower rotational speeds, i.e., 3 and 6 RPM, have a significant effect on the estimation of the frictional pressure loss, especially for low flow velocity and low yield stress fluids. More importantly, the uncertainty in dial readings has less effect on the frictional pressure drop estimation at moderately high flow velocity. This research also emphasizes the importance of rheological model parameter estimation using all available dial readings with the assistance of computational tools to better capture the rheological response of drilling fluids for better hydraulic calculation.

Enhancing pressure gradient prediction in multi-phase flow through diverse well geometries of North American shale gas fields using deep learning

To ensure the efficient operation of production wells and to avoid flow assurance issues, accurately predicting pressure gradients, Δp/ΔL, in wells is crucial. This study explores the relationship between Δp/ΔL and production well geometries. The data from 29,047 wells across five shale gas basins is utilized: Bakken, Barnett, British Columbia, Marcellus, and Delaware. These basins have distinct shale gas reservoir properties and diverse well geometries. Each well's geometry was characterized by ten parameters, and k-medoids clustering was applied to identify geometry patterns. However, this method overlooked minor geometric variations impacting Δp/ΔL. An adaptive factorization network model, a deep learning approach, was employed to address this. This model demonstrated significant predictive accuracy, as evidenced by high determination coefficients, R2, comparing actual and predicted Δp/ΔL in test data (overall 0.904; Bakken 0.840; Barnett 0.848; BC 0.844; Marcellus 0.872; Loving-Delaware 0.914; Reeves-Delaware 0.880). The method's reliability in forecasting Δp/ΔL is supported by its R2 values of ∼0.8 and the mean absolute percentage error of ∼5 %, indicating its applicability in newly developing basins. This methodology offers reliable Δp/ΔL predictions for all possible well geometry scenarios in shale gas fields but also eliminates the need for additional multi-phase flow simulations, enhancing design and production plan.

Prediction of Venous Trans-Stenotic Pressure Gradient Using Shape Features Derived From Magnetic Resonance Venography in Idiopathic Intracranial Hypertension Patients.

Idiopathic intracranial hypertension (IIH) is a condition of unknown etiology associated with venous sinus stenosis. This study aimed to develop a magnetic resonance venography (MRV)-based radiomics model for predicting a high trans-stenotic pressure gradient (TPG) in IIH patients diagnosed with venous sinus stenosis. This retrospective study included 105 IIH patients (median age [interquartile range], 35 years [27-42 years]; female:male, 82:23) who underwent MRV and catheter venography complemented by venous manometry. Contrast enhanced-MRV was conducted under 1.5 Tesla system, and the images were reconstructed using a standard algorithm. Shape features were derived from MRV images via the PyRadiomics package and selected by utilizing the least absolute shrinkage and selection operator (LASSO) method. A radiomics score for predicting high TPG (≥ 8 mmHg) in IIH patients was formulated using multivariable logistic regression; its discrimination performance was assessed using the area under the receiver operating characteristic curve (AUROC). A nomogram was constructed by incorporating the radiomics scores and clinical features. Data from 105 patients were randomly divided into two distinct datasets for model training (n = 73; 50 and 23 with and without high TPG, respectively) and testing (n = 32; 22 and 10 with and without high TPG, respectively). Three informative shape features were identified in the training datasets: least axis length, sphericity, and maximum three-dimensional diameter. The radiomics score for predicting high TPG in IIH patients demonstrated an AUROC of 0.906 (95% confidence interval, 0.836-0.976) in the training dataset and 0.877 (95% confidence interval, 0.755-0.999) in the test dataset. The nomogram showed good calibration. Our study presents the feasibility of a novel model for predicting high TPG in IIH patients using radiomics analysis of noninvasive MRV-based shape features. This information may aid clinicians in identifying patients who may benefit from stenting.

Development and assessment of performance of artificial neural networks for prediction of frictional pressure gradients during two-phase flow

This paper presents a discussion on several possibilities to predict the frictional pressure gradient during two-phase flow, with both the application of artificial intelligence and the implementation of conventional correlations and predictive methods. To this purpose, a huge database of approximately 8000 data points has been collected from 49 sources available in scientific literature, including 23 working fluids and the following ranges of parameters: mass fluxes from 32.7 to 2000 kg/m2s, saturation temperatures from -190°C to +120°C (reduced pressures from 0.021 to 0.780), tube diameters from 0.5 to 14.0 mm.This consolidated database has been used to train several artificial neural networks (ANNs), by using only two hidden layers (shallow neural networks) and evaluating the effect of: training and testing datasets choice (either test data included or outside the training domain), the number of neurons for each hidden layer (from 1 to 50), the type of output (either dimensional or non-dimensional), the type and number (from 1 to 22) of input parameters.The best results (MAPE of 16.8% and 88% of data within ±30%) have been obtained by using the liquid-only two-phase multiplier ΦLO2 as non-dimensional output and 12 mixed input parameters. Compared to the statistics of well-established literature correlations for frictional pressure drop (best MAPE of 22% and 73% of data points predicted within a ±30% error range, provided by Mauro et al. mechanistic method), the ANN demonstrates therefore a higher general accuracy. However, the use of Artificial Neural Networks does not guarantee a physical trend, which is instead preserved with conventional prediction methods.

Improved Two-Fluid Model for Segregated Flow and Integrated Multiphase Flow Modeling for Downhole Pressure Predictions

This work presents an integrated multiphase flow model for downhole pressure predictions with the purpose of producing relatively more accurate predictions under wide-flowing conditions while maintaining a simple form. As a component of the integrated model, an improved two-fluid model for segregated flow is proposed. The new two-fluid model was developed by improving the modeling of the wetted perimeters and the liquid wall shear stress. It outperformed five other existing state-of-the-art models in predicting the liquid holdup and pressure gradient of 11 experimental datasets from the literature, with an average relative error of 37.6% for liquid holdup predictions and 24.0% for pressure gradient predictions. The model succeeds in capturing the effects of inclination angle, pressure or gas density, and liquid and gas superficial velocities on liquid holdup and pressure gradient. The integrated model incorporates the state-of-the-art onset of liquid loading predictive model and classifies the flow into two major categories based on the onset of liquid loading that are modeled by two different approaches, respectively. It outperformed the other multiphase flow models in predicting the downhole pressure of 313 field data points, especially the wells that have both segregated and intermittent (mixed) flows. The errors were reduced to 4.4% for the field cases with mixed flow regions and 5.1% for all the field data points.

Predicting pressure gradient using artificial intelligence for transcatheter aortic valve replacement

ObjectiveAfter transcatheter aortic valve replacement, the mean transvalvular pressure gradient indicates the effectiveness of the therapy. The objective is to develop artificial intelligence to predict the post–transcatheter aortic valve replacement aortic valve pressure gradient and aortic valve area from preprocedural echocardiography and computed tomography data. MethodsA retrospective study was conducted on patients who underwent transcatheter aortic valve replacement due to aortic valve stenosis. A total of 1091 patients were analyzed for pressure gradient predictions (mean age 76.8 ± 9.2 years, 57.8% male), and 1063 patients were analyzed for aortic valve area predictions (mean age 76.7 ± 9.3 years, 57.2% male). An artificial intelligence learning model was trained (training: n = 663 patients, validation: n = 206 patients) and tested (testing: n = 222 patients) to predict pressure gradient, and a separate artificial intelligence learning model was trained (training: n = 640 patients, validation: n = 218 patients) and tested (testing: n = 205 patients) for predicting aortic valve area. ResultsThe mean absolute error for pressure gradient and aortic valve area predictions was 3.0 mm Hg and 0.45 cm2, respectively. Valve sheath size, body surface area, and age were determined to be the top 3 predictors for pressure gradient, and valve sheath size, left ventricular ejection fraction, and aortic annulus mean diameter were identified to be the top 3 predictors of post–transcatheter aortic valve replacement aortic valve area. A training dataset size of more than 500 patients demonstrated good robustness of the artificial intelligence models for pressure gradient and aortic valve area. ConclusionsThe artificial intelligence–based algorithm has demonstrated potential in predicting post–transcatheter aortic valve replacement transvalvular pressure gradient predictions for patients with aortic valve stenosis. Further studies are necessary to differentiate pressure gradient between valve types.

Prediction of Liquid Holdup in Horizontal Gas Wells Based on Dimensionless Number Selection

Abstract Pressure gradient prediction is crucial in gas well analysis. The experiment is the most effective method of understanding the flow characteristics in horizontal gas wells. The greatest difference between experimental and high-pressure conditions is gas density, which could cause the established multiphase correlations unreliable when they are applied to high-pressure gas wells. Similarity numbers are widely employed in predicting flow behavior. Nevertheless, few studies focused on this area. In addition, gas wells are characterized as high gas–liquid ratio; the majority empirical correlations were developed for oil wells, which have more consideration in low gas–liquid ratio, influencing the precision of gas well models. An experimental examination of gas–liquid flow has been carried out in this study. First, the experimental test matrix was designed to meet each flow pattern. Next, the effect of gas velocity, liquid velocity, pipe diameter, water-cut, and inclined angle on liquid holdup was explored. Subsequently, the similarity numbers suggested have been investigated and assessed for pressure scaling up. Finally, a comprehensive model was established, which was developed to forecast pressure gradient in gas wells. Field data were supplied to assess the new correlation. The results demonstrated that the Duns–Ros and the modified Duns–Ros dimensionless numbers were improper for pressure scaling up, whereas the Hewitt–Robert Number performs best. Based on the field data, the new correlation with Hewitt–Robert Number was superior to extensively employed pressure drop correlations, showing that it can deal with predicting pressure gradient in gas wells.

Prediction of pressure gradient and hold-up in horizontal liquid-liquid pipe flow

This paper aims to propose correlations to predict pressure gradient, friction factor and fluid phase hold-up in liquid-liquid horizontal pipe flow. To develop the correlations, experiments are conducted using high viscous oils (202 and 630 mPa⋅s) in a steel pipe of length 11.25 m and length-to-diameter ratio of 708. In addition, the experimental data from the literature comprising wide range of flow and fluid properties is analyzed. For the analysis, the liquid-liquid pipe flow data is categorized into two as: stratified and dispersed. The existing friction factor correlations are modified to incorporate the effects of viscosity of the oil phase, interfacial curvature (contact/wetting angle-in lieu of material of the pipe) and fluid phase fraction. In the two-fluid model of stratified flow, the wall stress and interfacial stress correlations are substituted with superficial velocities of fluids and superficial Reynolds numbers of fluid phases replacing fluid phase velocities and fluid Reynolds numbers. Similarly, for dispersed flow, an effective Reynolds number is described as the sum of superficial Reynolds number of oil and water phases. Substituting the generally employed mean or mixture Reynolds number with the effective Reynolds number into the existing single-phase turbulent flow friction factor correlation, an effective friction factor for oil-water flow is proposed. Employing the proposed correlations, the pressure gradient across the oil-water flow and hold-up volume fraction are predicted with significant reduction in error compared with that of conventionally employed correlations. The average error and standard deviation values of −7.06%, 20.72% and 0.31%, 18.79% are found for stratified flow and dispersed flow respectively.

Experimental study of swirling flow pneumatic liquid-carrying characteristics in the vortex tool inserted tube under liquid loading conditions

Liquid unloading by vortex tool is one of the most important applications of swirling flow pneumatic conveying. However, due to swirl decay, the evolution of gas–liquid flow patterns and the dynamic characteristics of pressure drop downstream of the vortex tool are much more complex than those in non-swirling flow. In this study, the gas–liquid flow patterns and pressure drop inside the vortex tool inserted tube (ID = 60 mm) under liquid loading conditions were experimentally studied and compared with those of the plain tube. With the superficial gas velocities ranging from 0.5 to 6.0 m/s and superficial liquid velocities of 0.01, 0.02, and 0.04 m/s, the effects of the structural parameters and stages of vortex tools on the pressure drop were discussed. The experimental results reveal that three distinct swirling flow patterns can be identified downstream of the vortex tool, namely swirling gas column flow, swirling churn flow and swirling annular flow. The insertion of vortex tool facilitates an earlier transition of the churn flow to the annular flow, and the induced swirling flow can restrain the liquid from falling back. The helix angle of the corkscrew deflector has a more significant effect on the pressure drop than its length. By reducing the helix angle from 54° to 45°, the total pressure drop and downstream pressure gradient are significantly decreased, even lower than those of the plain tube. But the series combination of 54° and 45° vortex tools results in a significant increase in the downstream pressure gradient. Additionally, at low gas flow rate, negative frictional pressure drop occurs in both plain tube and vortex tool inserted tube. Hence, the new friction factor correlations were established respectively for the positive and negative friction regions in the cases of the tube with and without vortex tool, and most of the frictional pressure gradient predictions are within an error band less than 20%.

Flow behavior of nano-silica solutions in cement cracks/fractures

One of the most prevalent causes of failure in oil and gas wells is the formation of cracks in the wellbore cement. The cracks can lead to loss of zonal isolation and fluid leakage to the surface. Compared to cement slurry which is difficult to squeeze into the cracks due to its high-solid content, nanoparticle (NP) gels have gained significant attraction for sealing fractures/cracks in oil and gas wells. It is highly desirable to know the flow behavior of NP solutions in the cracks. This study investigated the flow of nano-silica solutions through narrow rectangular cross-sections simulating sealant flow in cracks of oil well cement sheath. Pressure gradient was tested at 4 flow rates for various aspect ratios of the rectangular cross-sections. For a given cross-section of conduit, pressure drop increased with flow rate. For a given conduit height, the pressure drop decreased with conduit width due to increased flow area. Pressure drop increased as the conduit height reduced. These characteristics of nan-particle solution flow are like those found for conventional fluids. The pressure gradient of nano-silica solution flow in a conduit of a given rectangular cross-section is nearly proportional to flow velocity at tested flow rates, indicating laminar flow conditions in the tested flow conditions. The proportionality factor (rate of increase) increases as the aspect ratio of the cross-section decreases (slot becomes narrower). This is because a system with a lower aspect ratio has four boundaries that generate a 2-dimensional velocity profile, while a system with a higher aspect ratio has essentially two boundaries that generate a 1-dimensional velocity profile. More boundaries mean more frictional resistance to fluid flow. Hydraulics models for conventional fluid flow through conduits of rectangular cross-sections under-predict the pressure gradient of nano-silica solution. The discrepancy/difference is over 90% in pressure gradient prediction in general. The hydraulics model presented by Guo et al. (2022) with aspect ratio is better than the classical hydraulics model with hydraulic diameter. However, both models fail to predict the pressure gradient along the conduits with slot-like cross sections. The excessive pressure drop is attributed to the interactions between the nano-particles and conduit walls. Further investigations are required to gain a thorough understanding of nano-silica flow.

Two-phase stratified flow in horizontal pipes: A CFD study to improve prediction of pressure gradient and void fraction

Improved understanding of flow regime effects on design-influencing engineering quantities is of primary importance. This work is focused on the numerical prediction of pressure gradient and void fraction in a horizontal pipe where gas-liquid stratified flow is present in different operating conditions. The problem was modeled with unsteady, multiphase CFD (computational fluid dynamics) simulations. Volume Of Fluid (VOF) method was used as multiphase model. To define the numerical methodology, this study provides details on the influence of discretization grid and turbulence model on the simulation accuracy. It shows that mesh density on pipe cross-section is the most important grid parameter to focus on. Different turbulence models are required depending on the gas velocity and on its turbulence flow regime. Transition SST is able to model all the operating conditions but Realizable k-ε is adopted to further increase the accuracy of the results. A general underestimation of the pressure gradient is reported with an average error of − 6.43 % and − 16.21 % for a liquid superficial velocity of 0.04 m/s and 0.06 m/s respectively. Comparison with two-fluid 1D models shows that CFD simulations are the most accurate tools for predicting the pressure gradient at gas superficial velocities lower than 1.3 m/s. The implementation of a drift flux model shows a good agreement between experimental results and CFD simulations concerning void fraction estimation. CFD results are also used to underline the physical phenomena limiting the performance of 1D models.

PRESSURE GRADIENT COMPUTATION FOR FOAMS WITH DIFFERENT GEOMETRIC PROPERTIES: BASED ON ANN AND SVR MACHINE LEARNING MODEL AND TRAINED BY CFD SIMULATIONS

In this research work, a combination of computational fluid dynamics (CFD) simulation and artificial intelligence (AI) methods are conducted to study the effects of geometric properties of aluminum foams on airflow and to compute and predict pressure gradients in foams with such varied geometric parameters as porosity (65-90&amp;#37;) and pore diameter (200-2000 &amp;mu;m). The 3D foam structures are created by the Laguerre-Voronoi tessellations method. Based on the CFD results, pressure gradient for 114 different foams can be calculated in terms of inlet flow velocity (in the range 0.1-8 m/s). Foam pressure gradient is found to increase with increasing inlet flow velocity but with decreasing pore diameter and porosity. Comparisons reveal that the results obtained in the present study for pressure gradient are consistent with the data reported in the literature. It is, therefore, concluded that CFD simulation is a useful tool for pressure gradient estimation in a variety of foam types. Unique simulations are, however, needed each time foam structural properties change, which entails significant increases in the associated computation costs. This drawback may, nonetheless, be at least partially addressed by taking advantage of soft computing methods such as machine learning (ML). Artificial neural network (ANN) and support vector regression (SVR) as subsets of AI are designed (models with input variables inlet velocity and the foam structural parameters: porosity, pore diameter, and strut diameter) and trained using CFD results to predict pressure gradients in a large number of foams. When applied to new foam samples, the ML models exhibit an acceptable performance in predicting pressure gradients. Using such provisions, the method can be effectively used for predicting pressure gradient in various porous media at minimum computation costs.

Unified Modeling of Pseudo-Slug and Churn Flows for Liquid Holdup and Pressure Gradient Predictions in Pipelines and Wellbores

Summary Pseudo-slug flow is a coherent flow pattern that occurs in horizontal or inclined pipes surrounded by segregated and conventional slug flows. Recent studies have demonstrated that this flow pattern can occupy a large area in the flow pattern map and cannot be simply ignored. Churn flow commonly occurs in upward vertical or slightly deviated wellbores characterized by a chaotic intermittent flow behavior. Pseudo-slug and churn flows are generally considered as two different flow patterns mainly because of their visual differences. However, some recent studies have shown that they share many similarities. There are several hydraulic models for churn flow, and the models for pseudo-slug flow have also emerged in recent years. However, these models only work for a certain inclination angle range, and none work for both flow patterns. This paper proposes a new unified hydraulic model that is applicable for both flow patterns. The liquid holdup prediction is based on the drift-flux model, while the pressure gradient is predicted using a two-fluid model that considers wall frictions from both gas and liquid phases. The new model better captures the effects of gas and liquid flow rates, gas density, liquid viscosity, inclination angle, and pipe diameter on the liquid holdup and pressure gradient compared with other models. Model evaluations show that the new model gives the best predictions compared with other available pseudo-slug or churn models. The new model reduces the total average absolute relative error to 14.0% for the liquid holdup prediction and 19.1% for the pressure gradient prediction.

Experimental and Modeling Investigation on Gas-Liquid Two-Phase Flow in Horizontal Gas Wells

Abstract Wellbore pressure gradient in gas wells is significant in designing deliquification technologies and optimizing production. At present, no model has yet to be established specifically for gas wells at a wide gas flowrate range. When calculating pressure gradient in a specific gas field, engineers must evaluate these widely used models and get the best performance model at a certain range. To establish a more comprehensive model in horizontal gas wells, an experimental study was conducted to investigate the flow behavior of liquid-gas two-phase flow at different gas and liquid velocities and inclined angles in a 50 mm visual pipe. The evaluation of these widely used models against the experimental data shows that no model can predict liquid holdup at different gas velocity ranges, and huge deviations due to several reasons can be observed. After conducting a comprehensive analysis, a new liquid holdup correlation was proposed based on the Mukherjee–Brill model by correlating from the experimental results, which have parametric ranges closer to the production of gas wells. This new model adopts a new dimensionless gas velocity number to characterize flow similarities and better scale up pressure from the experiment to the gas wells. By validating against experimental data and field data, the results indicate that the new two-phase flow model has stable performance and can accurately predict pressure gradients at different ranges of pressure and gas/liquid velocities.

Pressure gradient prediction of gas-liquid two-phase pipe flow in horizontal gas wells

Accurate prediction of pressure gradient in gas wells is the theoretical basis for dynamic analysis, optimization of production and deliquification technologies design. Due to the complicated wellbore trajectories and flow behavior of horizontal gas wells, the accuracy of any existing prediction model of pressure gradient alone is difficult to meet engineering requirement within a wide range of gas-liquid ratio. Hence, a semi-empirical model of liquid holdup in a vertical wellbore is proposed based on three transition points: annular-churn, churn-slug and slug-bubble transition points. Based on the corresponding relationship between liquid holdups in the vertical pipe and the inclined pipe, a prediction method of liquid holdup in the inclined pipe is proposed, and thus a prediction model of pressure gradient in horizontal wells is established. The results show that the most important step in predicting pressure gradient of gas-liquid two-phase flow is the accurate prediction of liquid holdup. The liquid holdup increases first and then decreases with the increase of the inclination angle. Liquid holdup in inclined pipes can be predicted based on the variation of liquid holdup in the vertical pipe with the inclined angle. The relative performance factor (RPF) of the proposed model is the smallest compared with the gas-well logging data of the reference models, indicating the new model can meet the requirements of engineering calculation accuracy. It is concluded that the new model can be applied to the prediction of gas-liquid two-phase pressure gradient in horizontal gas wells with different production ranges, and can provide technical support for field production analysis.

Real-time prediction of formation pressure gradient while drilling

Accurate real-time pore pressure prediction is crucial especially in drilling operations technically and economically. Its prediction will save costs, time and even the right decisions can be taken before problems occur. The available correlations for pore pressure prediction depend on logging data, formation characteristics, and combination of logging and drilling parameters. The objective of this work is to apply artificial neural networks (ANN) and adaptive neuro-fuzzy inference system (ANFIS) to introduce two models to estimate the formation pressure gradient in real-time through the available drilling data. The used parameters include rate of penetration (ROP), mud flow rate (Q), standpipe pressure (SPP), and rotary speed (RS). A data set obtained from some vertical wells was utilized to develop the predictive model. A different set of data was utilized for validating the proposed artificial intelligence (AI) models. Both models forecasted the output with a good correlation coefficient (R) for training and testing. Moreover, the average absolute percentage error (AAPE) did not exceed 2.1%. For validation stage, the developed models estimated the pressure gradient with a good accuracy. This study proves the reliability of the proposed models to estimate the pressure gradient while drilling using drilling data. Moreover, an ANN-based correlation is provided and can be directly used by introducing the optimized weights and biases, whenever the drilling parameters are available, instead of running the ANN model.

Prediction of pressure gradient for oil-water flow: A comprehensive analysis on the performance of machine learning algorithms

Pressure gradient (PG) in liquid-liquid flow is one of the key components to design an energy-efficient transportation system for wellbores. This study aims to develop five robust machine learning (ML) algorithms and their fusions for a wide range of flow patterns (FP) regimes. The MLs include Support Vector Machine (SVM), Gaussian Process (GP), Random Forest (RF), Artificial Neural Network (ANN), k-Nearest Neighbor (kNN), and fusions of these five MLs. A total of eleven hundred experimental data points for nine FPs (two stratified and seven dispersed patterns) in horizontal wellbores are used to develop the MLs. The MLs' performance is evaluated using the metrics including mean absolute percentage error (MAPE), median absolute percentage error (MdAPE), coefficient of variation of root mean squared error (CV-RMSE), and adjusted coefficient of determination. The evaluation metrics are cross-validated using a repeated train-test split strategy. Seven important predictor variables are identified using a supervised feature selection approach: oil and water velocities, FP, input diameter, oil and water density, and oil viscosity. The results show that the high PG prediction accuracy can be achieved using GP compared to other MLs except for the ML-fusions (p < 0.05). A Friedman's test and Wilcoxon Sign-Rank post hoc analysis with Bonferroni correction show that PG prediction errors using GP are significantly lower than using the ANN model (p < 0.05). The values are 18.44 % and 23.9 % for CV-RMSE, 11.6 % and 10.06 % for MAPE, and 7.5 % and 6.75 % for MdAPE, using ANN and GP, respectively. While the previous studies mostly used ANN to demonstrate the capability of MLs to predict PG over the mechanistic or correlation-based models, the present research has shown that GP is even better than ANN using a wide range of FPs and a large data set. • Machine-learning algorithms are developed to predict the pressure gradient. • The Gaussian Process model produces significantly lower prediction error. • The median absolute percentage error using Gaussian Process is 6.75 %. • More than 95 % variation of pressure gradient can be explained using seven variables.

Heat transfer coefficient, pressure gradient, and flow patterns of R1233zd(E) and R1336mzz(Z) evaporating in a microchannel tube

This paper presents the heat transfer coefficient, pressure gradient, and flow patterns of R1233zd(E) and R1336mzz(Z) in a microchannel tube. Both heat transfer coefficient and pressure gradient are measured simultaneously and presented in the same plot in this paper. The experiment was conducted on a 24-port microchannel tube with an average hydraulic diameter of 0.643 mm. Experiment covers mass flux from 100 to 200 kg-m−2s−1, heat flux from 0 to 6 kW-m−2, vapor quality from 0 to 1. The saturation temperature of R1233zd(E) is from 30 to 50 °C, and R1336mzz(Z) is from 40 to 60 °C. The pressure gradient is a strong function of mass flux and saturation temperature (properties). Heat flux has an insignificant effect on the pressure gradient because of low accelerating pressure drop. The heat transfer coefficient is strongly affected by mass flux, heat flux, and saturation temperature. When mass flux is higher, the flow has more waves and turbulence, and the HTC and dP/dz are higher. R1233zd(E) and R1336mzz(Z) have much smaller HTC and higher dP/dz than R1234yf and R1234ze(E) under the same condition. R1234yf has almost the same HTC as R1234ze(E) at high quality (x>0.5), but higher at low quality (x<0.5). Hwang and Kim (2006) is recommended for predicting pressure gradient of the two low pressure fluids (R1233zd(E): MAE=14.1%, ME=-3.7%, Dev=16.8%; R1336mzz(Z): MAE=16.7%, ME=-2.0%, Dev=22.0%). Liu and Winterton (1991) is recommended for predicting HTC of R1336mzz(Z) (MAE=15.2%, ME=-4.6%, Dev=17.9%). None model has good agreement with the measured HTC of R1233zd(E).

Drift-Flux Parameters for High-Viscous-Oil/Gas Two-Phase Upward Flow in a Large and Narrow Vertical and Inclined Annular Duct

Abstract Proper sizing of flow lines in the upstream energy industry depends on accurate modeling of gas–liquid flow, which has a common occurrence in production wells and has been studied thoroughly for many decades. However, data of flow in duct geometries different from circular pipes and when the liquid viscosity is much higher than that of water are scarce. Proper prediction of pressure gradient, heat and mass transfer, and corrosion depends on the accuracy of the model used to calculate the volumetric phase fraction. In pumped directional wells with inverted-shroud gravitational separators, there is flow through an annular duct formed between the wells’ casing and the separator itself that can have some tens of meters. The present work is an investigation on upward vertical/inclined high-viscous-oil/gas flow in a large and narrow annulus (30 mm hydraulic diameter with an outer diameter equal to 155 mm), using a radial geometry comparable with those found in real production systems. Air–water and air–oil mixtures, the latter with two oil viscosity ranges, were used as working fluids. The experimental test section used was 9.67 m long positioned at 90 deg (vertical) and 45 deg and made of two concentric pipes. Flow pattern transitions from the literature were analyzed and compared with the collected experimental data. Drift-flux parameters were obtained from multiple working conditions. These drift-flux parameters were used in the development of a novel flow pattern-independent correlation, compared against the present data and other data sets from the literature in which other geometries and fluids were used. The predictions of the proposed drift-flux correlation are significantly superior in comparison with correlations selected from the literature in all cases.